Drug–Drug

Interactions
Edited by
Dong Hyun Kim and Sangkyu Lee
Printed Edition of the Special Issue Published in Pharmaceutics

www.mdpi.com/journal/pharmaceuitcs
Drug–Drug Interactions
Drug–Drug Interactions

Editors
Dong Hyun Kim
Sangkyu Lee

MDPI • Basel • Beijing • Wuhan • Barcelona • Belgrade • Manchester • Tokyo • Cluj • Tianjin
Editors
Dong Hyun Kim Sangkyu Lee
Inje University College of Medicine Kyungpook National University
Korea Korea

Editorial Office
MDPI
St. Alban-Anlage 66
4052 Basel, Switzerland

This is a reprint of articles from the Special Issue published online in the open access journal
Pharmaceutics (ISSN 1999-4923) (available at: https://www.mdpi.com/journal/pharmaceutics/
special issues/DDIs).

For citation purposes, cite each article independently as indicated on the article page online and as
indicated below:

LastName, A.A.; LastName, B.B.; LastName, C.C. Article Title. Journal Name Year, Volume Number,
Page Range.

ISBN 978-3-0365-2035-3 (Hbk)

ISBN 978-3-0365-2036-0 (PDF)

© 2021 by the authors. Articles in this book are Open Access and distributed under the Creative
Commons Attribution (CC BY) license, which allows users to download, copy and build upon
published articles, as long as the author and publisher are properly credited, which ensures maximum
dissemination and a wider impact of our publications.
The book as a whole is distributed by MDPI under the terms and conditions of the Creative Commons
license CC BY-NC-ND.
Contents

About the Editors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Preface to ”Drug–Drug Interactions” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

Young-Guk Na, Jin-Ju Byeon, Hyun Wook Huh, Min-Ki Kim, Young G. Shin, Hong-Ki Lee
and Cheong-Weon Cho
Effect of Ticagrelor, a Cytochrome P450 3A4 Inhibitor, on the Pharmacokinetics of Tadalafil
in Rats
Reprinted from: Pharmaceutics 2019, 11, 354, doi:10.3390/pharmaceutics11070354 . . . . . . . . . 1

Sanjita Paudel, Aarajana Shrestha, Piljoung Cho, Riya Shrestha, Younah Kim, Taeho Lee,
Ju-Hyun Kim, Tae Cheon Jeong, Eung-Seok Lee and Sangkyu Lee
Assessing Drug Interaction and Pharmacokinetics of Loxoprofen in Mice Treated with
CYP3A Modulators
Reprinted from: Pharmaceutics 2019, 11, 479, doi:10.3390/pharmaceutics11090479 . . . . . . . . . 11

Hyeon-Cheol Jeong, Soo Hyeon Bae, Jung-Woo Bae, Sooyeun Lee, Anhye Kim,
Yoojeong Jang and Kwang-Hee Shin
Evaluation of the Effect of CYP2D6 Genotypes on Tramadol and O-Desmethyltramadol
Pharmacokinetic Profiles in a Korean Population Using Physiologically-Based
Pharmacokinetic Modeling
Reprinted from: Pharmaceutics 2019, 11, 618, doi:10.3390/pharmaceutics11110618 . . . . . . . . . 25

Yu Fen Zheng, Soo Hyeon Bae, Zhouchi Huang, Soon Uk Chae, Seong Jun Jo,
Hyung Joon Shim, Chae Bin Lee, Doyun Kim, Hunseung Yoo and Soo Kyung Bae
Lack of Correlation between In Vitro and In Vivo Studies on the Inhibitory Effects of
(-)-Sophoranone on CYP2C9 Is Attributable to Low Oral Absorption and Extensive Plasma
Protein Binding of (-)-Sophoranone
Reprinted from: Pharmaceutics 2020, 12, 328, doi:10.3390/pharmaceutics12040328 . . . . . . . . . 41

So-Young Park, Phi-Hung Nguyen, Gahyun Kim, Su-Nyeong Jang, Ga-Hyun Lee,
Nguyen Minh Phuc, Zhexue Wu and Kwang-Hyeon Liu
Strong and Selective Inhibitory Effects of the Biflavonoid Selamariscina A against CYP2C8
and CYP2C9 Enzyme Activities in Human Liver Microsomes
Reprinted from: Pharmaceutics 2020, 12, 343, doi:10.3390/pharmaceutics11070354 . . . . . . . . . 59

Young Hee Choi

Interpretation of Drug Interaction Using Systemic and Local Tissue Exposure Changes
Reprinted from: Pharmaceutics 2020, 12, 417, doi:10.3390/pharmaceutics12050417 . . . . . . . . . 73

Jung Hwan Ahn, Junhyeong Kim, Naveed Ur Rehman, Hye-Jin Kim, Mi-Jeong Ahn and
Hye Jin Chung
Effect of Rumex Acetosa Extract, a Herbal Drug, on the Absorption of Fexofenadine
Reprinted from: Pharmaceutics 2020, 12, 547, doi:10.3390/pharmaceutics12060547 . . . . . . . . . 93

Gracia Santos-Dı́az, Ana Marı́a Pérez-Pico, MiguelÁngel Suárez-Santisteban,

Vanesa Garcı́a-Bernalt, Raquel Mayordomo and Pedro Dorado
Prevalence of Potential Drug–Drug Interaction Risk among Chronic Kidney Disease Patients in
a Spanish Hospital
Reprinted from: Pharmaceutics 2020, 12, 713, doi:10.3390/pharmaceutics12080713 . . . . . . . . . 107

v
Yoshihiro Noguchi, Tomoya Tachi and Hitomi Teramachi
Subset Analysis for Screening Drug–Drug Interaction Signal Using
Pharmacovigilance Database
Reprinted from: Pharmaceutics 2020, 12, 762, doi:10.3390/pharmaceutics12080762 . . . . . . . . . 119

Malavika Deodhar, Sweilem B Al Rihani, Meghan J. Arwood, Lucy Darakjian, Pamela Dow,
Jacques Turgeon and Veronique Michaud
Mechanisms of CYP450 Inhibition: Understanding Drug-Drug Interactions Due to
Mechanism-Based Inhibition in Clinical Practice
Reprinted from: Pharmaceutics 2020, 12, 846, doi:10.3390/pharmaceutics12090846 . . . . . . . . . 129

Dominique A. Garrison, Zahra Talebi, Eric D. Eisenmann, Alex Sparreboom and

Sharyn D. Baker
Role of OATP1B1 and OATP1B3 in Drug-Drug Interactions Mediated by Tyrosine
Kinase Inhibitors
Reprinted from: Pharmaceutics 2020, 12, 856, doi:10.3390/pharmaceutics12090856 . . . . . . . . . 147

Milica Markovic, Moran Zur, Inna Ragatsky, Sandra Cvijić and Arik Dahan
BCS Class IV Oral Drugs and Absorption Windows: Regional-Dependent Intestinal
Permeability of Furosemide
Reprinted from: Pharmaceutics 2020, 12, 1175, doi:10.3390/pharmaceutics12121175 . . . . . . . . 165

Florian Klomp, Christoph Wenzel, Marek Drozdzik and Stefan Oswald

Drug–Drug Interactions Involving Intestinal and Hepatic CYP1A Enzymes
Reprinted from: Pharmaceutics 2020, 12, 1201, doi:10.3390/pharmaceutics12121201 . . . . . . . . 181

Yunzhou Fan, Zhengxuan Liang, Jinghui Zhang and Guofeng You

Oral Proteasomal Inhibitors Ixazomib, Oprozomib, and Delanzomib Upregulate the Function of
Organic Anion Transporter 3 (OAT3): Implications in OAT3-Mediated Drug-Drug Interactions
Reprinted from: Pharmaceutics 2021, 13, 314, doi:10.3390/pharmaceutics13030314 . . . . . . . . . 207

Alexander V. Dmitriev, Anastassia V. Rudik, Dmitry A. Karasev, Pavel V. Pogodin,

Alexey A. Lagunin, Dmitry A. Filimonov and Vladimir V. Poroikov
In Silico Prediction of Drug–Drug Interactions Mediated by Cytochrome P450 Isoforms
Reprinted from: Pharmaceutics 2021, 13, 538, doi:10.3390/pharmaceutics13040538 . . . . . . . . . 223

vi
About the Editors
Dong Hyun Kim
Ph.D. (Professor) acquired his B.S. degree in Pharmacy in 1982 from Seoul National University
in Seoul, Korea, and Ph.D. in Toxicology in 1988 from Korea Advanced Institute of Science and
Technology. He was trained at Vanderbilt University as a postdoc under supervision by Dr. Fred
Guengerich in 1988–1990. He then joined the Doping Control Center, Korea Institute for Science
and Technology (KIST), Seoul, from 1990 and served as Director of the center during 2004–2008. In
2011, he joined College of Medicine, Inje University, in Busan and currently serves as Professor in the
Department of Pharmacology. His main research interests are drug metabolism, drug interactions,
and metabolomics. Alongside his academic research, he has been collaborating with numerous
pharmaceutical companies on new drug development. He has published approximately 200
peer-reviewed articles in internationally recognized journals.

Sangkyu Lee
Ph.D. (Associate Professor) acquired his B.S. degree in Pharmacy in 2002 from Yeungnam
University in Gyeongsan, Korea, and a Ph.D. in Toxicology in 2007 from the same university. He
then joined Doping Control Center, Korea Institute for Science and Technology (KIST), Seoul, in
2007–2009 and Ben May Department for Cancer Research, The University of Chicago, Chicago,
USA, in 2009–2011, where he was a postdoctoral fellow. At KIST, he studied drug metabolism
and drug interactions with his supervisor, Prof. Dong Hyun Kim. In 2011, he joined College of
Pharmacy, Kyungpook National University in Daegu, Korea, and currently serves as an Associate
Professor. As a toxicologist, his research has focused on using mass spectrometry, and he is conducts
research on drug metabolism, drug interactions, and the pharmacokinetics of new drug candidates
or naturally active substances. He is also an expert in mass spectrometry-based toxicometabolomics
and toxicoproteomics research. He has published close to 190 peer-reviewed articles. Dr. Lee’s
research is supported by the National Research Foundation (NRF) of Korea grant founded by the
Korea government. He is currently Vice Director of BK21 FOUR Community-Based Intelligent Novel
Drug Discovery Education Unit, having started in this position in 2020, and he also serves on the
Editorial Board of Mass Spectrometry Letters (Scopus).

vii
Preface to ”Drug–Drug Interactions”
Drug–drug interactions (DDIs) cause a drug to affect other drugs, leading to reduced drug
efficacy or increased toxicity of the affected drug. Some well-known interactions are known to be
the cause of adverse drug reactions (ADRs) that are life threatening to the patient. Traditionally, DDI
have been evaluated around the selective action of drugs on specific CYP enzymes. The interaction
of drugs with CYP remains very important in drug interactions but, recently, other important
mechanisms have also been studied as contributing to drug interaction including transport- or
UDP-glucuronyltransferase as a Phase II reaction-mediated DDI. In addition, novel mechanisms of
regulating DDIs can also be suggested. In the case of the substance targeted for interaction, not only
the DDIs but also the herb–drug or food–drug interactions have been reported to be clinically relevant
in terms of adverse side effects.
This Special Issue serves to highlight the current progress in research on drug–drug interactions
and contains eleven outstanding research articles and five review articles. Firstly are the results of
in vitro, animal, and human evaluations of drug–drug interactions and herb–drug interactions that
induce representative changes in CYP activity. In addition, the results of studies on DDIs that may
occur depending on the transporter, plasma protein binding, and human genotype are included.
Practical research results on DDI evaluation using In Silico Prediction and Pharmacovigilance
Database are presented. This Special Issue contains results related to the main mechanisms by which
DDIs can cause ADRs. In the review article, the DDIs and their relation to CYP1A is systematically
arranged, and findings related to the role of CYP in mechanism-based inhibition in clinical settings
are summarized. As additional insightful content, the interaction of monoclonal antibodies, the
occurrence of DDI according to systemic and local tissue exposure, and the effects of OATP1B1 and
1B3 on DDI are systematically reviewed.
This Special Issue aims to highlight current progress in understanding the drug interactions
of commercial drugs, both clinically and nonclinically, and in new discoveries regarding the
mechanisms of drug interactions that cause ADR. We expect that this Special Issue will provide
insights into drug–drug interactions related to adverse drug reactions and contribute to advancement
of the relevant research areas.

Dong Hyun Kim, Sangkyu Lee

Editors

ix
 pharmaceutics

Article
Effect of Ticagrelor, a Cytochrome P450 3A4 Inhibitor,
on the Pharmacokinetics of Tadalafil in Rats
Young-Guk Na † , Jin-Ju Byeon † , Hyun Wook Huh, Min-Ki Kim, Young G. Shin, Hong-Ki Lee *
and Cheong-Weon Cho *
College of Pharmacy and Institute of Drug Research and Development, Chungnam National University,
99, Daehak-ro, Yuseong-gu, Daejeon 34134, Korea
* Correspondence: dvmlhk@gmail.com (H.-K.L.); chocw@cnu.ac.kr (C.-W.C.); Tel.: +82-42-821-7301 (H.-K.L.);
+82-42-821-5934 (C.-W.C.)
† These authors contributed equally to this work.

Received: 26 June 2019; Accepted: 17 July 2019; Published: 20 July 2019

Abstract: Tadalafil is a cytochrome P450 (CYP) 3A4 substrate. Because there are few data on
drug-drug interactions, it is advisable to take sufficient consideration when co-administering tadalafil
with CYP3A4 inducers or inhibitors. This study was conducted to assess the effect of ticagrelor, a
CYP3A4 inhibitor, on the pharmacokinetic properties of tadalafil after oral administration to rats. A
total of 20 Sprague–Dawley male rats were randomly divided into the non-pretreated group and
ticagrelor-pretreated group, and tadalafil was orally administered to each group after pretreatment
with or without ticagrelor. Blood samples were collected at predetermined time points after oral
administration of tadalafil. As a result, systemic exposure of tadalafil in the ticagrelor-pretreated
group was significantly increased compared to the non-pretreated group (1.61-fold), and the clearance
of tadalafil in the ticagrelor-pretreated group was significantly reduced than the non-pretreated group
(37%). The prediction of the drug profile through the one-compartment model could explain the
differences of pharmacokinetic properties of tadalafil in the non-pretreated and ticagrelor-pretreated
groups. This study suggests that ticagrelor reduces a CYP3A-mediated tadalafil metabolism and
that tadalafil and a combination regimen with tadalafil and ticagrelor requires dose control and
specific pharmacotherapy.

Keywords: tadalafil; ticagrelor; drug-drug interaction; pharmacokinetics; plasma concentration; CYP3A4

1. Introduction
Erectile dysfunction (ED), the most prevalent complaint in males, is the persistent inability to
maintain an erection [1]. Various factors, such as age and presence of cardiovascular diseases, influence
the incidence of ED [2,3]. Particularly, vascular diseases, such as coronary artery disease (CAD), are
related to a high prevalence of ED [4–6]. It has been reported that 42% of patients between the ages of
40 and 60 years are affected by this condition [7–9]. In addition, a high rate of ED prevalence (~75%)
has been investigated in CAD patients [10,11]. The most commonly prescribed oral drugs for ED
are the 5-phosphodiesterase (PDE5) inhibitors, and the drugs for CAD are the platelet aggregation
inhibitor [12,13]. Therefore, PDE5 may be co-administered to CAD patients already receiving platelet
aggregation inhibitors.
Tadalafil (Cialis® ), the approved PDE5 inhibitor, is used in the treatment of ED and is one of the
most frequently prescribed PDE5 inhibitors [14]. In addition, tadalafil exhibits a longer clinical efficacy
(up to 36 h) than sildenafil or vardenafil [15,16]. Because of the long duration of action, a once-a-day
dose regimen improves the life quality of patients. Tadalafil is classified as a cytochrome P450 (CYP)
3A4 substrate and is mainly metabolized by CYP3A4 to catechol, which is extensively bound to form
methyl catechol glucuronide, a major circulating metabolite of tadalafil through methylation [17].

Pharmaceutics 2019, 11, 354; doi:10.3390/pharmaceutics110703541 www.mdpi.com/journal/pharmaceutics

Pharmaceutics 2019, 11, 354

Ticagrelor (Brilinta® ), the platelet aggregation inhibitor, belongs to P2Y12 receptor antagonists
and is used for the treatment of CAD [18,19]. It provides the superior and more sustained inhibition
of platelet aggregation than clopidogrel, another P2Y12 receptor antagonist [20]. Recent studies have
reported that ticagrelor acts as an inhibitor of CYP3A4 [21,22]. When ticagrelor was co-administered
with atorvastatin, ticagrelor acted as a CYP inhibitor, increasing the maximal plasma concentration
of atorvastatin and the area under the plasma concentration-time curve from 0 to infinity by 23%
and 36%, respectively [23]. Therefore, drug-drug interaction by co-administration of tadalafil and
ticagrelor can inhibit the metabolism of CYP3A4 substrates, such as tadalafil. Despite the possibility of
co-administration, the pharmacokinetic interactions between ticagrelor and tadalafil are still uncertain.
Because ticagrelor is an inhibitor of CYP3A4 that accounts for about 15–30% of the total CYP
enzyme in humans, the co-administration potentially affects efficacy and safety by altering tadalafil
exposure [24]. In healthy male volunteers, the plasma concentrations of tadalafil have been shown to
increase with CYP3A4 inhibitors, such as ritonavir [25]. Although tadalafil is well tolerated, patients
have experienced side effects, such as headaches, stomachaches, back pain, muscle aches, nasal
congestion, redness, limb pain, dizziness, or blurred vision due to the high exposure of tadalafil [26,27].
In other studies, the side effects were increased when doses of tadalafil were higher, indicating that
side effects were somewhat related to blood concentration of tadalafil [28,29]. Thus, the combination of
tadalafil with ticagrelor also warranted investigation and a drug-drug interaction study should be
conducted. However, pharmacokinetic interactions between tadalafil and ticagrelor have not been
reported in vivo models.
The objective of our study was to evaluate the effect of ticagrelor on the plasma concentration-time
profiles of tadalafil in rats. This study was performed with a parallel design consisting of a
non-pretreated group and a ticagrelor-pretreated group. The ticagrelor-pretreated group received oral
administration of ticagrelor for seven days of the pretreatment period to inhibit CYP3A. Tadalafil was
then orally administered on the seventh day. The non-compartment analysis was performed to analyze
the pharmacokinetic profile and the one-compartment model was successfully applied to compare the
pharmacokinetics between the two groups. This study assumed that potential drug-drug interactions
between ticagrelor and tadalafil could have a clinical impact on the patients.

2. Materials and Methods

2.1. Chemicals and Reagents

Tadalafil, ticagrelor, and paclitaxel (internal standard (IS)) were kindly obtained from Korea
United Pharma Inc. (Seoul, Korea). Dimethyl sulfoxide, formic acid, and polyethylene glycol 400
(PEG 400) were purchased from Sigma–Aldrich (St. Louis, MO, USA). Acetonitrile and methanol were
purchased from J.T. Baker (Phillipsburg, NJ, USA). Analytical grade reagents were used throughout
this study. Overall, distilled water was used.

2.2. LC-MS/MS Analysis of Tadalafil

The concentrations of tadalafil were analyzed by a liquid chromatography tandem-mass
spectrometry (LC-MS/MS) system equipped with Agilent 1290 series and Agilent 6495 Triple Quad
LC/MS (Agilent Technologies, Santa Clara, CA, USA). A YMC-Triart C18 column (50 × 2.0 mm, 1.9 μm;
YMC Inc., Wilmington, NC, USA) was used. The mobile phase was a mixture of 0.2% formic acid
in acetonitrile and 0.2% formic acid in distilled water (50:50, v/v), and the flow rate was 0.4 mL/min.
The temperature of the column and autosampler were set as 30 ◦ C and 4 ◦ C, respectively. The positive
ion mode using Agilent jet stream electrospray ionization (AJS-ESI) was applied to record the scan
mass spectra. The ion transitions of tadalafil and IS were set as 390.4→268.1 m/z and 876.4→308.1 m/z,
respectively, and detected with a multiple reaction monitoring (MRM) mode. The collision energies for
tadalafil and IS were 10 V and 30 V, respectively. The cell accelerator voltage was 5 V and the dwell
time was set as 200 ms. The source parameters were set as follows: Gas temperature 200 ◦ C, gas flow

2
Pharmaceutics 2019, 11, 354

14 L/min, nebulizer 20 psi, sheath gas heater 250 ◦ C, sheath gas flow 11 L/min, capillary 3000 V, and
nozzle voltage 1500 V.
In this analysis, the most abundant ion transition of tadalafil (390.4→268.1 m/z) was selected to
determine the lowest limit of quantification (LLOQ), and the LLOQ of tadalafil was 3 ng/mL. The range
of calibration curve of tadalafil was set to 3–6670 ng/mL. The curve was written with a weighted linear
regression (1/x) and showed excellent linearity with R2 > 0.997. The method has shown accurate and
reproducible results within acceptable tolerances (less than 20% coefficient of variation (CV) at LLOQ
and less than 15% CV at all other concentrations) [30]. The acquired LC-MS/MS data were processed
with Agilent analysis software (Agilent MassHunter Quantitative Software Version B.07.00, Agilent
Technologies, Santa Clara, CA, USA).

2.3. Animals
All animal experiments were carried out in accordance with the protocol (No. CNU-01167) and the
“Guidelines in Use of Animal” approved by Chungnam National University Institutional Animal Care
and Use Committee (Daejeon, Korea, 2019). Male Sprague–Dawley rats (aged 7–8 weeks, bodyweight
250–300 g) were purchased from Nara-Biotec (Seoul, Korea). All animals were housed in a dark-light
cycle of 12 h at 22 ◦ C and were allowed free access to water and food.

2.4. Pharmacokinetic Study

The experiment was performed in a parallel design. A total of 20 rats were randomly divided into
two groups. The control group, Group N (n = 10, non-pretreated rats), was orally administered for
7 days only with the vehicle (normal saline/polyethylene glycol 400/dimethyl sulfoxide = 40:20:20).
Group T (n = 10, ticagrelor-pretreated rats), an experimental group, received 10 mg/kg of ticagrelor in
the vehicle once a day for 7 days orally. Thereafter, on the seventh day, tadalafil (2 mg/kg) was orally
administered 30 min after the final administration of the vehicle or ticagrelor. Doses of all samples were
calculated according to the weight of the rats and administered using gavage. The dose of ticagrelor
and tadalafil given to rats were determined by converting the human doses (90 mg for ticagrelor and
20 mg for tadalafil) to animal doses with body surface area, according to the United States Food and
Drug Administration guidelines [31]. Blood (0.3 mL) was collected from the retro-orbital plexus or the
jugular vein at predetermined time-points (0, 0.33, 0.67, 1, 1.5, 2, 4, 6, 8, 12, and 24 h) after the oral
administration. Samples were centrifuged at 15,000× g for 5 min at 4 ◦ C. The plasma was collected and
stored at −20 ◦ C until the concentration of tadalafil was analyzed by LC-MS/MS.

2.5. Sample Preparation for LC-MS/MS Analysis

Protein precipitation was applied to extract tadalafil from plasma. Briefly, 200 μL of acetonitrile
(0.2 v/v% formic acid) containing 500 ng/mL of IS was added to 20 μL of the plasma sample. After
shaking for 5 min, the mixture was centrifuged at 15,000× g for 5 min. The supernatant (150 μL) was
transferred and 5 μL of the sample was injected into the LC-MS/MS.

2.6. Pharmacokinetic Data Analysis

Pharmacokinetic data were analyzed based on the non-compartment analysis of WinNonlin
software 8.1 (Pharsight Corp., Sunnyvale, CA, USA). The maximum plasma concentration of tadalafil
(maximum concentration (Cmax )) and the time to reach the maximal plasma concentration (Tmax ) of
tadalafil were determined from the plasma concentration-time profiles. The area under the plasma
concentration vs time curve from 0 to 24 h (AUC0–24 ) was calculated by the linear trapezoidal rule
and the area under the plasma concentration vs time curve from 0 h to infinite time (AUC0–∞ ) was
estimated by extrapolating time to infinity. The elimination half-life (T1/2 ) and apparent total clearance
(CL/F) were determined from the ln 2/elimination rate constant and dose/AUC0–∞ , respectively [32].

3
Pharmaceutics 2019, 11, 354

2.7. Pharmacokinetic Modeling

Pharmacokinetic modeling of the plasma concentration-time profile of tadalafil was performed
using a one-compartment model to investigate the absorption and elimination rates. The parameters
of the one-compartment model of tadalafil were estimated based on data from Group N and Group T.
In this model, the shift of the tadalafil amount in the gut compartment and the central compartment
is described by the following equations [33]:

dG
= −Ka ·G
dt
dC
= Ka ·G − Ke ·C
dt
where G and C represent the amount of tadalafil in the gut compartment and the central compartment,
respectively. Ka indicates the absorption rate constant from the gut compartment to the central
compartment. Ke indicates the elimination rate constant from the central compartment to the gut
compartment. Each equation was applied to both Group N and Group T data through the classic
model of WinNolin software 8.1 (Pharsight Corp., Sunnyvale, CA, USA).
Predicted plasma concentrations of tadalafil (Conpred ) using the one-compartment model were
estimated by the following equation [34]:

Conpred = C/(V/F)

where V/F indicates the central volume of distribution of tadalafil.

The fold error versus concentration or time was calculated using the following equation [35]:

Conpred − Conobs
Fold error =
Conobs

where Conpred indicates the predicted concentration of tadalafil from the one-compartment
pharmacokinetic modeling and Conobs indicates the observed concentration of tadalafil from the
pharmacokinetic experiment. The fold error within ±2-fold is an acceptable range [35].

2.8. Statistical Analysis

Values are represented as mean ± standard deviation (SD). A student’s t-test was applied for
statistical significance of differences (p < 0.05) and the statistical analysis was performed using
GraphPad Prism 8 software (GraphPad Software Inc., La Jolla, CA, USA).

3. Results and Discussion

3.1. Pharmacokinetic Data

The purpose of our study was to assess the effect of ticagrelor on plasma concentration-time
profiles of tadalafil in rats. Drug-drug interactions may occur when CYP inducers or inhibitors are
co-administered with the drug metabolized by the CYP enzyme. Several studies have demonstrated
that ticagrelor is a CYP3A4 inhibitor and tadalafil is a CYP3A4 substrate [23,25]. Therefore, potential
drug-drug interactions between ticagrelor and tadalafil can occur, affecting the pharmacokinetic profile,
the efficacy for treatment of erectile dysfunction, and the frequency of side effects.
To identify the potential drug-drug interactions between ticagrelor and tadalafil in rats, ticagrelor
was administered once a day for seven days to inhibit CYP3A in rats only for Group T. On the last
day of the study, Group N and Group T received 2 mg/kg tadalafil via oral administration to assess
the pharmacokinetics effect of ticagrelor on tadalafil. The plasma concentrations of ticagrelor were

4
Pharmaceutics 2019, 11, 354

sufficient to competitively inhibit CYP3A, and the effects of CYP3A inhibition on systemic exposure of
tadalafil could be determined [36].
The pharmacokinetic profiles of tadalafil for Group N and Group T are shown in Figure 1.
The parameters of the non-compartment analysis are listed in Table 1. The non-compartmental
analysis is the model-independent method, which is based on the time course of drug concentrations.
Pharmacokinetic parameters from the non-compartmental analysis are mainly used to evaluate drug
exposure in oral administration of a drug. The systemic exposure of tadalafil was increased in Group
T compared with Group N. Group T showed higher Cmax , AUC0–24 , and AUC0-∞ than the Cmax ,
AUC0–24 , and AUC0-∞ of Group N. These results showed significant increases in the values of AUC0–24
(1.61-fold, p < 0.05) and AUC0–∞ (1.66-fold, p < 0.05). The Cmax of Group T slightly increased 1.15-fold
compared to that of Group N, but there was no significant difference (p = 0.3332). The ratio of ACU0–24
and AUC0–∞ between Group N and Group T exceeded the range of 0.8–1.25, where no pre-specified
pharmacological effect was observed [37].

$ %
 
*URXS1 *URXS1
3ODVPDFRQFHQWUDWLRQ

3ODVPDFRQFHQWUDWLRQ

*URXS7 *URXS7
 
QJP/

QJP/


 



 
           
7LPHK 7LPHK

Figure 1. Plasma concentration-time profiles of tadalafil after oral administration to non-pretreated

rats (Group N) and ticagrelor-pretreated rats (Group T). Values are represented as mean ± SD (n
= 10). Dotted marks and dashed lines indicate the observed plasma concentration and the fitted
pharmacokinetic profile from the one-compartment model, respectively. (A) Linear scale; (B) log scale.

Table 1. Non-compartment analysis parameters of tadalafil after oral administration to non-pretreated

rats (Group N) and ticagrelor-pretreated rats (Group T). Values are represented as mean ± SD (n = 10).

Parameters Group N Group T Ratio a p-Value

Cmax (ng/mL) 375.94 ± 72.81 432.71 ± 119.79 1.15 0.3332
AUC0–24 (ng·h/mL) 3046.88 ± 732.50 4895.74 ± 1592.87 1.61 0.0124
AUC0–∞ (ng·h/mL) 3070.95 ± 743.01 5095.04 ± 1800.30 1.66 0.0134
Tmax (h) 1.45 ± 0.50 3.22 ± 1.30 2.22 0.0030
T1/2 (h) 3.15 ± 0.23 4.47 ± 0.89 1.42 0.0018
CL/F (L/h/kg) 0.68 ± 0.15 0.43 ± 0.13 0.63 0.0035
Value o f Group T
a Ratio = Value o f Group N ; maximum concentration (Cmax ); area under the plasma concentration vs. time curve from
0 to 24 h (AUC0–24 ); area under the plasma concentration vs. time curve from 0 to infinity (AUC0–∞ );time to reach
maximal concentration (Tmax ); half-life (T1/2 ); apparent total clearance (CL/F).

As a result of CYP inhibition by ticagrelor, the plasma concentration of tadalafil decreased more
slowly after co-administration of ticagrelor and tadalafil than after tadalafil alone. Tmax , T1/2 , and
CL/F showed statistically significant differences between Groups T and Group N. In the case of Tmax ,
the absorption of tadalafil in Group T was delayed (Tmax of 3.22 ± 1.30 h) compared to that in Group
N (Tmax of 1.45 ± 0.50 h) (p < 0.005). Thus, the tadalafil plasma concentration remained high until
8 h. This result indicated that the reduced metabolism of tadalafil by CYP3A inhibition caused the
absorption of the drug for a longer period of time [38]. In particular, Group T showed an increased T1/2
compared to Group N (p < 0.005) and the CL/F value of Group T was significantly lower than that of
Group N (p < 0.005). These results showed that the inhibition of hepatic CYP3A metabolism and the
reduction of the first pass effect increased the exposure of tadalafil [39].

5
Pharmaceutics 2019, 11, 354

In general, the non-compartment analysis parameters of tadalafil showed statistically significant

differences between Group N and Group T (Figure 2). In the present study, co-administration with
ticagrelor increased tadalafil exposure and affected the AUC0–24 , AUC0-∞ , Tmax , and half-life of
tadalafil. In particular, the half-life of tadalafil increased from 3.15 h in Group N to 4.47 h in Group T,
resulting in the increased AUC0–24 (1.61-fold), AUC0–∞ (1.66-fold), and Tmax (2.22-fold). In addition,
the result of 37% reduction of tadalafil clearance in Group T compared with Group N supported our
findings. These data suggest that ticagrelor inhibits tadalafil metabolism due to drug-drug interaction
with tadalafil.

$ % & 
p   p p



$8&±QJ⋅KP/

$8&±’QJ⋅KP/

&PD[QJP/

 
 

 

 

  
*URXS1 *URXS7 *URXS1 *URXS7 *URXS1 *URXS7

'  ( ) 
p  p p


&/)/KNJ
  
7PD[K

7K



 



  
*URXS1 *URXS7 *URXS1 *URXS7 *URXS1 *URXS7

Figure 2. Comparisons of the non-compartment analysis parameters of tadalafil from non-pretreated

rats (Group N) and ticagrelor-pretreated rats (Group T). Boxes mean 25th and 75th percentiles of data
and whiskers mean 5th and 95th percentiles of data. The median and mean values are displayed as a
solid line (–) and a plus mark (+) in boxes, respectively. (A) Cmax ; (B) AUC0–24 ; (C) AUC0–∞ ; (D) Tmax ;
(E) T1/2 ; (F) CL/F. Maximum concentration (Cmax ); area under the plasma concentration vs. time
curve from 0 to 24 h (AUC0–24 ); area under the plasma concentration vs. time curve from 0 to infinity
(AUC0–∞ );time to reach maximal concentration (Tmax ); half-life (T1/2 ); apparent total clearance (CL/F).

3.2. Pharmacokinetic Modeling

The one-compartment model was successfully applied to describe each group in terms of the
plasma concentration-time profile after a single oral administration of 2 mg/kg tadalafil. Among the
various compartment models, the one-compartment model, which is the simplest and best suited to
the observed pharmacokinetic profile of tadalafil, was used to compare the observed pharmacokinetic
profile with the fitted pharmacokinetic profile. The one-compartment modeling supports the results of
the non-compartment analysis that changed by the co-administration of ticagrelor. The modeling-based
comparison allows for greater confidence in the metabolic inhibition effect by comparing the observed
pharmacokinetic profile with the fitted pharmacokinetic profile, as well as by matching with the
parameters of the non-compartment analysis. Additionally, the model-based approach was used to
evaluate a more specific and quantified absorption, elimination, and distribution of tadalafil in vivo.
As shown in Figure 1, the dashed lines and dotted marks mean the predicted plasma concentration
and observed plasma concentration, respectively. The dashed lines showed similar patterns to
the dotted marks. Figure 3 indicates the residual plot of the fold error versus time or observed
plasma concentration. It shows the acceptable residual values (fold error <2) at most of the analyzed
points, which indicates that the one-compartment model is suitable for the explanation of the plasma
concentration profile of tadalafil [35,40].

6
Pharmaceutics 2019, 11, 354

Figure 3. Residual plots written with the pharmacokinetic modeling values of non-pretreated rats
(Group N) and ticagrelor-pretreated rats (Group T). (A) Fold error vs. time; (B) fold error vs.
plasma concentration.

Table 2 lists the parameters of the one-compartment model. In the one-compartment model, Ke and
V/F were decreased significantly in Group T compared to Group N. The Ke of Group T (0.13 ± 0.03 h−1 )
were decreased compared with that of Group N (0.17 ± 0.05 h−1 ) (p < 0.05). Moreover, the V/F of Group
T (3.36 ± 0.95 L/kg) was significantly decreased compared with that of Group N (4.39 ± 0.78 L/kg)
(p < 0.05). These reductions led to a decrease in the clearance of tadalafil. In addition, the Ka of Group T
was also decreased, but there was no statistical significance (p = 0.1933) (Figure 4). However, this slight
reduction was due to the delay in absorption, and it is clear that the absorption increased because of
the high Cmax . These results were consistent with those from the non-compartment analysis, indicating
that the more absorption and the low clearance caused the increased exposure of tadalafil.

Table 2. The one-compartment model parameters of tadalafil in non-pretreated rats (Group N) and
ticagrelor-pretreated rats (Group T). Values are represented as mean ± SD (n = 10).

Parameters Group N Group T Ratio a p-Value

Ka (1/h) 1.71 ± 0.91 1.14 ± 0.73 0.67 0.1933
Ke (1/h) 0.17 ± 0.03 0.13 ± 0.03 0.77 0.0166
V/F (L/kg) 4.39 ± 0.76 3.36 ± 0.95 0.77 0.0438
a Ratio = Value o f Group T
Value o f Group N ; Absorption rate constant from the gut compartment to the central compartment (Ka );
elimination rate constant from the central compartment to the gut compartment (Ke ); the central volume of
distribution (V/F).

$ % &
  
p  p
 p
 

9)/NJ
.DK

.HK

  


 


  
*URXS1 *URXS7 *URXS1 *URXS7 *URXS1 *URXS7

Figure 4. Comparisons of the one-compartment model parameters of tadalafil in non-pretreated rats

(Group N) and ticagrelor-pretreated rats (Group T). Boxes mean 25th and 75th percentiles of data and
whiskers mean 5th and 95th percentiles of data. The median and mean values are displayed as a solid
line (–) and a plus mark (+) in boxes, respectively. (A) Ka ; (B) Ke ; (C) V/F. Absorption rate constant
from the gut compartment to the central compartment (Ka ); elimination rate constant from the central
compartment to the gut compartment (Ke ); the central volume of distribution (V/F).

In addition, the results from the modeling indicate that ticagrelor is a weak CYP3A inhibitor.
The weak inhibitor is defined as a substance that increases the AUC value of the CYP substrate by

7
Pharmaceutics 2019, 11, 354

1.25-fold to 2-fold, or that it reduces the clearance of the CYP substrate by 20–50%, and ticagrelor is also
classified as this inhibitor [22,41]. Changes in the AUC (1.61-fold) and clearance (37%) of tadalafil by
co-administration with ticagrelor were within the range, supporting the above classification. Therefore,
we inferred that ticagrelor acts as a weak CYP3A inhibitor, affecting the pharmacokinetic profiles
of tadalafil, a CYP3A substrate. Furthermore, co-administration of tadalafil with ticagrelor or other
CYP3A inhibitors, such as ketoconazole and diltiazem, would also be expected to increase tadalafil
exposure [42].
In a clinical aspect, tadalafil shows robust safety and tolerability [12]. However, the
co-administration of tadalafil with ticagrelor to elderly men may cause unexpected side effects. The
half-life of tadalafil in normal healthy men is reported to be approximately 17.5 h after administration,
whereas that of elderly men is about 21.6 h [43]. This half-life in elderly men is likely to increase further
due to co-administration with ticagrelor, which may increase the incidence of side effects, such as
headaches and dyspepsia. In addition, patients receiving nitrate medication for treatment of angina
are recommended to postpone the nitrate treatment for at least 48 h after administration of tadalafil,
but this may need to be longer to prevent serious hypotension [44]. Therefore, most patients receiving
these drugs are older, so co-administration of tadalafil should be carefully considered.

4. Conclusions
In our study, the plasma concentration-time profile of tadalafil was significantly changed by
co-administration with ticagrelor. A ticagrelor-inhibited CYP3A-mediated tadalafil metabolism and
the systemic exposure of tadalafil increased by pretreatment with ticagrelor. Co-administration of
tadalafil with ticagrelor increased the AUC of tadalafil by approximately 61% and decreased the
clearance of tadalafil by 37%. These results suggest that the co-administration of tadalafil and ticagrelor
may need dose control and specific drug therapy to avoid side effects from drug-drug interactions.
Therefore, CAD patients receiving ticagrelor should be closely monitored when administering tadalafil
concomitantly. For further studies, an additional experiment is expected to identify the variation of
clinical efficacy of tadalafil by co-administration with ticagrelor, a CYP3A4 inhibitor.

Author Contributions: Conceptualization, Y.-G.N.; methodology, Y.-G.N. and J.-J.B.; software, Y.-G.N. and J.-J.B.;
validation, H.-W.H. and J.-J.B.; formal analysis, J.-J.B. and H.-W.H.; investigation, Y.-G.N. and J.-J.B.; resources,
M.-K.K.; data curation, Y.-G.N.; writing—original draft preparation, Y.-G.N. and J.-J.B.; writing—review and
editing, Y.-G.S., H.-K.L., and C.-W.C.; supervision, C.-W.C.
Funding: This work was supported by the Basic Science Research Program (2016R1A2B4011294) through the
National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science, and Technology.
Acknowledgments: The materials were kindly supported by Korea United Pharma Inc.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Hatzimouratidis, K.; Amar, E.; Eardley, I.; Giuliano, F.; Hatzichristou, D.; Montorsi, F.; Vardi, Y.; Wespes, E.
Guidelines on male sexual dysfunction: Erectile dysfunction and premature ejaculation. Eur. Urol. 2010, 57,
804–814. [CrossRef] [PubMed]
2. Laumann, E.O.; Nicolosi, A.; Glasser, D.B.; Paik, A.; Gingell, C.; Moreira, E.; Wang, T. Sexual problems among
women and men aged 40–80 y: Prevalence and correlates identified in the global study of sexual attitudes
and behaviors. Int. J. Import. Res. 2005, 17, 39–57. [CrossRef] [PubMed]
3. Dégano, I.R.; Marrugat, J.; Grau, M.; Salvador-González, B.; Ramos, R.; Zamora, A.; Martí, R.; Elosua, R. The
association between education and cardiovascular disease incidence is mediated by hypertension, diabetes,
and body mass index. Sci. Rep. 2017, 7, 12370. [CrossRef] [PubMed]
4. Gandaglia, G.; Briganti, A.; Jackson, G.; Kloner, R.A.; Montorsi, F.; Montorsi, P.; Vlachopoulos, C. A systematic
review of the association between erectile dysfunction and cardiovascular disease. Eur. Urol. 2014, 65,
968–978. [CrossRef] [PubMed]

8
Pharmaceutics 2019, 11, 354

5. Montorsi, P.; Ravagnani, P.M.; Galli, S.; Rotatori, F.; Briganti, A.; Salonia, A.; Dehò, F.; Montorsi, F. Common
grounds for erectile dysfunction and coronary artery disease. Curr. Opin. Urol. 2004, 14, 361–365. [CrossRef]
[PubMed]
6. Jackson, G.; Boon, N.; Eardley, I.; Kirby, M.; Dean, J.; Hackett, G.; Montorsi, P.; Montorsi, F.; Vlachopoulos, C.;
Kloner, R.; et al. Erectile dysfunction and coronary artery disease prediction: Evidence-based guidance and
consensus. Int. J. Clin. Pract. 2010, 64, 848–857. [CrossRef]
7. Schwarz, E.R.; Rastogi, S.; Kapur, V.; Sulemanjee, N.; Rodriguez, J.J. Erectile dysfunction in heart failure
patients. J. Am. Coll. Cardiol. 2006, 48, 1111–1119. [CrossRef]
8. Speel, T.G.; Van Langen, H.; Meuleman, E.J. The risk of coronary heart disease in men with erectile dysfunction.
Eur. Urol. 2003, 44, 366–371. [CrossRef]
9. Montorsi, P.; Ravagnani, P.M.; Galli, S.; Rotatori, F.; Veglia, F.; Briganti, A.; Salonia, A.; Deho, F.; Rigatti, P.;
Montorsi, F.; et al. Association between erectile dysfunction and coronary artery disease. Role of coronary
clinical presentation and extent of coronary vessels involvement: The COBRA trial. Eur. Heart J. 2006, 27,
2632–2639. [CrossRef]
10. Kloner, R.A.; Mullin, S.H.; Shook, T.; Matthews, R.; Mayeda, G.; Burstein, S.; Peled, H.; Pollick, C.;
Choudhary, R.; Rosen, R.; et al. Erectile dysfunction in the cardiac patient: How common and should we
treat? J. Urol. 2003, 170, S46–S50. [CrossRef]
11. Nascimento, E.R.; Maia, A.C.; Pereira, V.; Soares-Filho, G.; Nardi, A.E.; Silva, A.C. Sexual dysfunction and
cardiovascular diseases: A systematic review of prevalence. Clinics 2013, 68, 1462–1468. [CrossRef]
12. Coward, R.M.; Carson, C.C. Tadalafil in the treatment of erectile dysfunction. Ther. Clin. Risk Manag. 2008, 4,
1315–1330. [CrossRef]
13. Clappers, N.; Brouwer, M.A.; Verheugt, F.W. Antiplatelet treatment for coronary heart disease. Heart 2007,
93, 258–265. [CrossRef]
14. Huang, S.A.; Lie, J.D. Phosphodiesterase-5 (PDE5) inhibitors in the management of erectile dysfunction.
Pharm. Ther. 2013, 38, 407–419.
15. Gong, B.; Ma, M.; Xie, W.; Yang, X.; Huang, Y.; Sun, T.; Luo, Y.; Huang, J. Direct comparison of tadalafil with
sildenafil for the treatment of erectile dysfunction: A systematic review and meta-analysis. Int. Urol. Nephrol.
2017, 49, 1731–1740. [CrossRef]
16. Blount, M.A.; Beasley, A.; Zoraghi, R.; Sekhar, K.R.; Bessay, E.P.; Francis, S.H.; Corbin, J.D. Binding of tritiated
sildenafil, tadalafil, or vardenafil to the phosphodiesterase-5 catalytic site displays potency, specificity,
heterogeneity, and cGMP stimulation. Mol. Pharmacol. 2004, 66, 144–152. [CrossRef]
17. Rezvanfar, M.A.; Rahimi, H.R.; Abdollahi, M. ADMET considerations for phosphodiesterase-5 inhibitors.
Expert Opin. Drug Metab. Toxicol. 2012, 8, 1231–1245. [CrossRef]
18. Na, Y.G.; Byeon, J.J.; Wang, M.; Huh, H.W.; Son, G.H.; Jeon, S.H.; Bang, K.H.; Kim, S.J.; Lee, H.J.; Lee, H.K.;
et al. Strategic approach to developing a self-microemulsifying drug delivery system to enhance antiplatelet
activity and bioavailability of ticagrelor. Int. J. Nanomed. 2019, 14, 1193–1212. [CrossRef]
19. Son, G.H.; Na, Y.G.; Huh, H.W.; Wang, M.; Kim, M.K.; Han, M.G.; Byeon, J.J.; Lee, H.K.; Cho, C.W. Systemic
design and evaluation of ticagrelor-loaded nanostructured lipid carriers for enhancing bioavailability and
antiplatelet activity. Pharmaceutics 2019, 11, E222. [CrossRef]
20. Bliden, K.P.; Tantry, U.S.; Storey, R.F.; Jeong, Y.H.; Gesheff, M.; Wei, C.; Gurbel, P.A. The effect of ticagrelor
versus clopidogrel on high on-treatment platelet reactivity: Combined analysis of the ONSET/OFFSET and
RESPOND studies. Am. Heart J. 2011, 162, 160–165. [CrossRef]
21. Teng, R.; Butler, K. The effect of ticagrelor on the metabolism of midazolam in healthy volunteers. Clin. Ther.
2013, 35, 1025–1037. [CrossRef]
22. Zhou, D.; Andersson, T.B.; Grimm, S.W. In vitro evaluation of potential drug-drug interactions with ticagrelor:
Cytochrome P450 reaction phenotyping, inhibition, induction, and differential kinetics. Drug Metab. Dispos.
2011, 39, 703–710. [CrossRef]
23. Teng, R.; Mitchell, P.D.; Butler, K.A. Pharmacokinetic interaction studies of co-administration of ticagrelor
and atorvastatin or simvastatin in healthy volunteers. Eur. J. Clin. Pharmacol. 2013, 69, 477–487. [CrossRef]
24. Klein, K.; Zanger, U.M. Pharmacogenomics of cytochrome P450 3A4: Recent progress toward the “missing
heritability” problem. Front. Genet. 2013, 4, 1–12. [CrossRef]

9
Pharmaceutics 2019, 11, 354

25. Garraffo, R.; Lavrut, T.; Ferrando, S.; Durant, J.; Rouyrre, N.; MacGregor, T.R.; Sabo, J.P.; Dellamonica, P.
Effect of tipranavir/ritonavir combination on the pharmacokinetics of tadalafil in healthy volunteers. J. Clin.
Pharmacol. 2011, 51, 1071–1078. [CrossRef]
26. Fraunfelder, F.W. Visual side effects associated with erectile dysfunction agents. Am. J. Oppthalmol. 2005, 140,
723–724. [CrossRef]
27. Montorsi, F.; Verheyden, B.; Meuleman, E.; Jünemann, K.P.; Moncada, I.; Valiquette, L.; Casabe, A.; Pacheco, C.;
Denne, J.; Knight, J.; et al. Long-term safety and tolerability of tadalafil in the treatment of erectile dysfunction.
Eur. Urol. 2004, 45, 339–345. [CrossRef]
28. Porst, H.; Giuliano, F.; Glina, S.; Ralph, D.; Casabé, A.R.; Elion-Mboussa, A.; Shen, W.; Whitaker, J.S.
Evaluation of the efficacy and safety of once-a-day dosing of tadalafil 5 mg and 10 mg in the treatment of
erectile dysfunction: Results of a multicenter, randomized, double-blind, placebo-controlled trial. Eur. Urol.
2006, 50, 351–359. [CrossRef]
29. Hatzichristou, D.; Gambla, M.; Rubio-Aurioles, E.; Buvat, J.; Brock, G.B.; Spera, G.; Rose, L.; Lording, D.;
Liang, S. Efficacy of tadalafil once daily in men with diabetes mellitus and erectile dysfunction. Diabet. Med.
2008, 25, 138–146. [CrossRef]
30. Tiwari, G.; Tiwari, R. Bioanalytical method validation: An updated review. Pharm. Methods 2010, 1, 25–38.
[CrossRef]
31. US Food and Drug Administration. Guidance for Industry and Reviewers: Estimating the Maximum
Safe Starting Dose in Initial Clinical Trials for Therapeutics in Adult Healthy Volunteers. 2005. Available
online: https://www.fda.gov/regulatory-information/search-fda-guidance-documents/estimating-maximum-
safe-starting-dose-initial-clinical-trials-therapeutics-adult-healthy-volunteers (accessed on 24 August 2018).
32. Bang, K.H.; Na, Y.G.; Huh, H.W.; Hwang, S.J.; Kim, M.S.; Kim, M.; Lee, H.K.; Cho, C.W. The delivery strategy
of paclitaxel nanostructured lipid carrier coated with platelet membrane. Cancers 2019, 11, 807. [CrossRef]
33. Kim, M.S.; Baek, I.H. Effect of dronedarone on the pharmacokinetics of carvedilol following oral administration
to rats. Eur. J. Pharm. Sci. 2018, 111, 13–19. [CrossRef]
34. Kim, J.S.; Kim, M.S.; Baek, I.H. Enhanced bioavailability of tadalafil after intranasal administration in beagle
dogs. Pharmaceutics 2018, 10, 187. [CrossRef]
35. Sjögren, E.; Thorn, H.; Tannergren, C. In silico modeling of gastrointestinal drug absorption: Predictive
performance of three physiologically based absorption models. Mol. Pharm. 2016, 13, 1763–1778. [CrossRef]
36. Teng, R. Ticagrelor: Pharmacokinetic, pharmacodynamic and pharmacogenetic profile: An update. Clin.
Pharmacokinet. 2015, 54, 1125–1138. [CrossRef]
37. Kothare, P.A.; Seger, M.E.; Northrup, J.; Mace, K.; Mitchell, M.I.; Linnebjerg, H. Effect of exenatide on
the pharmacokinetics of a combination oral contraceptive in healthy women: An open-label, randomised,
crossover trial. BMC Clin. Pharmacol. 2012, 12, 8. [CrossRef]
38. Doligalski, C.T.; Tong Logan, A.; Silverman, A. Drug interactions: A primer for the gastroenterologist.
Gastroenterol. Hepatol. 2012, 8, 376–383.
39. Jetter, A.; Kinzig-Schippers, M.; Walchner-Bonjean, M.; Hering, U.; Bulitta, J.; Schreiner, P.; Sörgel, F.; Fuhr, U.
Effects of grapefruit juice on the pharmacokinetics of sildenafil. Clin. Pharmacol. Ther. 2002, 71, 21–29. [CrossRef]
40. Lee, D.S.; Kim, S.J.; Choi, G.W.; Lee, Y.B.; Cho, H.Y. Pharmacokinetic–pharmacodynamic model for the
testosterone-suppressive effect of leuprolide in normal and prostate cancer rats. Molecules 2018, 23, 909. [CrossRef]
41. Wagner, C.; Pan, Y.; Hsu, V.; Grillo, J.A.; Zhang, L.; Reynolds, K.S.; Sinha, V.; Zhao, P. Predicting the effect of
cytochrome P450 inhibitors on substrate drugs: Analysis of physiologically based pharmacokinetic modeling
submissions to the US Food and Drug Administration. Clin. Pharmacokinet. 2015, 54, 117–127. [CrossRef]
42. Teng, R.; Butler, K. Effect of the CYP3A inhibitors, diltiazem and ketoconazole, on ticagrelor pharmacokinetics
in healthy volunteers. J. Drug Assess. 2013, 2, 30–39. [CrossRef]
43. Fransis, S.H.; Corbin, J.D. Molecular mechanisms and pharmacokinetics of phosphodiesterase-5 antagonists.
Curr. Urol. Rep. 2003, 4, 457–465. [CrossRef]
44. Kloner, R.A.; Hutter, A.M.; Emmick, J.T.; Mitchell, M.I.; Denne, J.; Jackson, G. Time course of the interaction
between tadalafil and nitrates. J. Am. Coll. Cardiol. 2003, 42, 1855–1860. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

10
 pharmaceutics

Article
Assessing Drug Interaction and Pharmacokinetics of
Loxoprofen in Mice Treated with CYP3A Modulators
Sanjita Paudel 1 , Aarajana Shrestha 2 , Piljoung Cho 1 , Riya Shrestha 1 , Younah Kim 1 , Taeho Lee 1 ,
Ju-Hyun Kim 2 , Tae Cheon Jeong 2 , Eung-Seok Lee 2 and Sangkyu Lee 1, *
1 BK21 Plus KNU Multi-Omics based Creative Drug Research Team, College of Pharmacy, Research Institute
of Pharmaceutical Sciences, Kyungpook National University, Daegu 41566, Korea;
sanjitapdl99@gmail.com (S.P.); whvlfwjd@naver.com (P.C.); riya.shrestha07@gmail.com (R.S.);
younah86@naver.com (Y.K.); tlee@knu.ac.kr (T.L.)
2 College of Pharmacy, Yeungnam University, Gyeongsan 38541, Korea; aarajanashrestha1@gmail.com (A.S.);
jhkim@yu.ac.kr (J.-H.K.); taecheon@ynu.ac.kr (T.C.J.); eslee@ynu.ac.kr (E.-S.L.)
* Correspondence: sangkyu@knu.ac.kr; Tel.: +82-53-950-8571; Fax: +82-53-950-8557

Received: 29 July 2019; Accepted: 12 September 2019; Published: 16 September 2019

Abstract: Loxoprofen (LOX) is a non-selective cyclooxygenase inhibitor that is widely used for
the treatment of pain and inflammation caused by chronic and transitory conditions. Its alcoholic
metabolites are formed by carbonyl reductase (CR) and they consist of trans-LOX, which is active,
and cis-LOX, which is inactive. In addition, LOX can also be converted into an inactive hydroxylated
metabolite (OH-LOXs) by cytochrome P450 (CYP). In a previous study, we reported that CYP3A4 is
primarily responsible for the formation of OH-LOX in human liver microsomes. Although metabolism
by CYP3A4 does not produce active metabolites, it can affect the conversion of LOX into trans-/cis-LOX,
since CYP3A4 activity modulates the substrate LOX concentration. Although the pharmacokinetics
(PK) and metabolism of LOX have been well defined, its CYP-related interactions have not been
fully characterized. Therefore, we investigated the metabolism of LOX after pretreatment with
dexamethasone (DEX) and ketoconazole (KTC), which induce and inhibit the activities of CYP3A,
respectively. We monitored their effects on the PK parameters of LOX, cis-LOX, and trans-LOX in
mice, and demonstrated that their PK parameters significantly changed in the presence of DEX or
KTC pretreatment. Specifically, DEX significantly decreased the concentration of the LOX active
metabolite formed by CR, which corresponded to an increased concentration of OH-LOX formed
by CYP3A4. The opposite result occurred with KTC (a CYP3A inhibitor) pretreatment. Thus, we
conclude that concomitant use of LOX with CYP3A modulators may lead to drug–drug interactions
and result in minor to severe toxicity even though there is no direct change in the metabolic pathway
that forms the LOX active metabolite.

Keywords: Loxoprofen; drug-drug interaction; CYP3A; Dexamethasone; Ketoconazole

1. Introduction
Loxoprofen (LOX) is an anti-inflammatory prodrug (NSAID) with potential antipyretic and
analgesic properties, but its side effects are less defined when compared to other NSAIDs [1–3]. It is one
of the most clinically prescribed NSAIDs in Japan, and it is also popular in Eastern Asia, the Middle
East, Latin America, and Africa [1]. It is a well-known cyclo-oxygenase (COX1 and COX2) inhibitor
that reduces the synthesis of inflammatory agents such as prostaglandins [2,4,5]. It is also widely used
for the management of pain and inflammation in chronic and transient conditions (e.g., toothache,
headache, menstrual cramps, common cold, etc.) [1,2,6].

11
Pharmaceutics 2019, 11, 479; doi:10.3390/pharmaceutics11090479 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2019, 11, 479

Although LOX has fewer sides effects than other NSAIDs, many adverse effects have been
reported such as GI disorders (erosive gastritis and bleeding), renal disorders, cardiovascular disorders,
headaches, anaphylaxis, and abdominal pain [1,2,6,7]. Moreover, the use of LOX with drugs
such as methotrexate, warfarin, aspirin, and valacyclovir is contraindicated [6,8–10]. Generally,
the pharmacokinetic (PK) parameters of many drugs are influenced by the inhibition or induction of
cytochrome P450 enzymes (CYPs) [11], which can result in various drug interactions, toxicity, and either
an increase or decrease in drug activity [12].
Previous studies have extensively investigated the metabolic activities of LOX in various organisms
such as rats, mice, monkeys, rabbits, and humans [1,13,14]. However, to the best of our knowledge,
there is still a large gap in knowledge regarding the interactions of LOX with other drugs that may cause
simple to severe adverse drug interactions. The alcoholic metabolites of LOX (cis-LOX and trans-LOX)
are mainly produced by carbonyl reductase (CR), and trans-LOX is the only active metabolite derived
from LOX, which LOX is a pharmacologically inactive drug unless it is metabolized to trans-LOX [1,5].
In addition to CR, LOX is also metabolized by CYP450. In our previous study, we reported that the
CYP-mediated metabolism of LOX was catalyzed by CYP3A4 and CYP3A5 to form hydroxylated
LOX (OH-LOX). We also found that an inactive metabolite of LOX catalyzed by CYP was significantly
higher in dexamethasone (DEX)-induced liver microsomes [15].
It is well known that CYP3A4 is most abundant in the liver, and it can catalyze approximately
50% of commercially available drugs [16,17]. Many commercial drugs are known to regulate the
activity of CYP3A, and they may be clinically administered with LOX. The CYP3A/LOX metabolic
pathway produces an inactive metabolite and competes with another pathway for LOX that produces
the active metabolite. If LOX is co-administered with a drug that modulates the activity of CYP3A,
there may be possible drug interactions affecting the concentration of LOX or the active metabolite
of LOX. Thus, we investigated the interaction of LOX with CYP3A by monitoring the effects of DEX
(a CYP3A inducer) or ketoconazole (KTC, a CYP3A inhibitor) pretreatment on the PK parameters of
LOX, cis-LOX, and trans-LOX in an ICR mouse model.

2. Materials and Methods

2.1. Materials
LOX (2-(4-((2-Oxocyclopentyl)methyl)phenyl)propanoic acid, C15 H18 O3 , CAS ID 68767-14-5) was
procured from Tokyo Chemicals Industry (Tokyo, Japan). LOX was used to synthesize cis-LOX and
trans-LOX ((2S)-2-[4-[[(1R,2S)-2-hydroxycyclopentyl]methyl]phenyl]propanoic acid, C15 H20 O3 , CAS
ID 83648-76-4) with purities of 96.5% and 97.9%, respectively [18]. DEX, KTC, dextromethorphan,
and phenacetin were purchased from Sigma-Aldrich Co., LLC. (St. Louis, MO, USA) while midazolam
was procured from Bukwang Pharmaceutical Co., Ltd. (Seoul, Korea). Mass spectrometry (MS) grade
water and acetonitrile (ACN) were obtained from Fischer Scientific (Pittsburgh, PA, USA).

2.2. Animal Treatments and Sample Preparations

Male ICR mice (36 mice of 5 weeks) were purchased from Orient Co. (Seongnam, Korea) andwere
randomly divided into 4–6 mice per cage. Then, the mice were acclimatized for 1 week in a controlled
environment (relative humidity: 60%, temperature: 25 ◦ C) under a 12-h/12-h light/dark cycle and
supplied standard rodent chow and tap water freely. All animal handling procedures followed
protocols issued by the Society of Toxicology (USA, 1989) and were approved on March 21, 2019 by the
Institutional Review Board of Kyungpook National University (project ID # 2019-41).
The acclimatized animals were divided into eight groups, with each group containing 3 mice:
groups I and II, vehicle (corn oil); groups III and IV, DEX-treated; groups V and VI, vehicle (10% Ethanol);
and groups VII and VIII, KTC-treated group. Briefly, groups I and II were intraperitoneally (i.p.) treated
with corn oil for 3 days while DEX (dissolved in corn oil and administered at 40 mg/kg) was given to
groups III and IV for 3 consecutive days [19]. Then, animals were fasted for 12 hours with free access

12
Pharmaceutics 2019, 11, 479

to water before starting the experiment. After fasting, LOX (20 mg/kg) was administered orally to
groups I, II, III, and IV. Ethanol (10%) and KTC (60 mg/kg) were administered i.p. to the vehicle-treated
groups (V and VI) and KTC-treated groups (VII and VIII), respectively [20], and after 3 min, LOX was
given orally (20 mg/kg).
After the oral administration of LOX, blood from groups I, III, V, and VII was collected from the tail
at 0, 5, 10, 15, 30, 60, 120, and 240 min post-administration and placed into sodium heparin-containing
tubes. After the last blood collection, the mice were sacrificed by cervical dislocation. Plasma was then
prepared by centrifuging the blood at 4000× g for 15 min at 4 ◦ C and stored at −80 ◦ C until analysis.
For analysis, each sample was prepared by mixing plasma (10 μL) and 90 μL of ACN containing 0.1%
formic acid and 5 μM of tolbutamide as an internal standard (IS). The samples were then vortexed and
centrifuged at 13,000× g for 10 min at 4 ◦ C. Finally, 10 μL of sample supernatant was injected into the
LC-MS/MS system. The PK parameters (the maximum plasma concentration [Cmax ], time to reach
the maximum plasma concentration [Tmax ], elimination half-life [T1/2 ], and area under the plasma
concentrations [AUC]) were analyzed by WinNonlin software (Version 2.1, Scientific Consulting,
Louisville, KY, USA).
Similarly, blood from groups II, IV, VI, and VIII was collected from the hepatic portal vein 10 min
post-LOX administration to identify metabolites of LOX and to analyze drug–drug interactions. Plasma
from these samples was prepared as explained above. Then, 300 μL of ACN having 0.1% formic acid
and 5 μM of the IS were mixed with 100 μL of each plasma sample. Next, the samples were vortexed
and centrifuged at 13,000× g for 10 min at 4 ◦ C. Supernatants were transferred into tubes and dried
using a Labconco Speed Vac (Labconco Corporation, Kansas City, MO, USA). The dried samples were
reconstituted using 100 μL of 50% methanol and centrifuged at 13,000× g for 10 min at 4 ◦ C. Each
supernatant was transferred into an LCMS vial, and 5 μL were injected into a high-resolution mass
spectrometer (HRMS).

2.3. CYP Activities in the Mouse Liver

Male ICR mice were divided into 4 groups with each group containing 3 mice: group I, vehicle
(corn oil); group II, DEX-treated; group III, vehicle (10% Ethanol); and group IV, KTC-treated group.
Group I was treated with corn oil for 3 days while DEX (dissolved in corn oil and administered at
40 mg/kg) was administered i.p. to group II for 3 consecutive days [19]. Similarly, group III and
group IV were treated once with 10% ethanol or KTC (60 mg/kg), respectively [20]. Twenty-four hours
after the last treatment, the liver was excised and homogenized with three volumes of ice-cold 0.1 M
potassium phosphate buffer (pH 7.4). The supernatant fraction was then separated as the S9 fraction
from the mixture by centrifugation at 9000× g at 4 ◦ C and stored at −80 ◦ C until use.
To characterize the CYP activities in the harvested livers, phenacetin O-demethylation (using 80 μM
of phenacetin) for CYP1A, dextromethorphan O-demethylation (using 5 μM of dextromethorphan) for
CYP2D, and midazolam 1’-hydroxylation (using 5 μM of midazolam) for CYP3A were used as cocktail
probe reactions [21,22]. The level of protein in the S9 fraction was determined by Bradford assay [23].
The cocktail probes were then incubated with 10 μL of each S9 fraction, potassium phosphate buffer
(pH 7.4), and an NADPH generating system (NGS) in a final volume of 100 μL. After incubation for
30 min at 37 ◦ C, ice-cold ACN containing 0.1% formic acid and IS (5 μM) was added to stop the reaction.
Samples were then vortexed and centrifuged at 13,000× g for 10 min at 4 ◦ C. Finally, the supernatants
were transferred to LC-MS/MS vials for analysis.

2.4. Instrument and Data Acquisition

A Shimadzu Prominence UFLC system (Kyoto, Japan) connected to a TSQ vantage triple
quadrupole mass spectrometer with a HESI-II spray source incorporated with a DGU-20A5 degasser,
an LC-20AD pump, a SIL-20A autosampler, and a CTO-20A column oven was used for the analyses.
A shim-pack GIS C18 column (150 × 3.0 mm, 3 μM) was used to separate analytes in the samples. Mobile
phases A and B were composed of water with formic acid and ACN with 0.1% formic acid, respectively,

13
Pharmaceutics 2019, 11, 479

and the flow rate was 0.50 mL/min at 40 ◦ C. The gradient conditions were as follows: 20% of B between
0 and 0.25 min, 20–80% of B between 0.25 and 9.75 min, 80–20% of B between 9.75 and 10 min, and 20%
of B between 10 and 13 min. The MS was operated under the following conditions: electrospray
ionization in negative mode at 3.0 kV, capillary temperature at 350 ◦ C, vaporizer temperature at 300 ◦ C,
sheath gas pressure at 35 Arb, and auxiliary gas pressure at 10 Arb. Finally, Xcalibur software (Thermo
Fisher Scientific Inc., Waltham, MA, USA) was used for data analysis.
HRMS coupled with ultrahigh performance liquid chromatography (UHPLC) was used to detect
hydroxy-LOX and other metabolites of LOX. The UHPLC system, Dionex Ultimate 3000 (Dionex Softron
GmbH, Germering, Germany) consisted of an HPG-3200SD Standard binary pump, a WPS 3000 TRS
analytical autosampler, and a TCC-3000 SD column compartment. In this experiment, the HRMS was
a Q Exactive Focus quadrupole-Orbitrap MS (Thermo Fisher Scientific, Bremen, Germany) equipped
with a heated electrospray ionization (HESI-II) ion probe.
LOX was detected as a deprotonated ion [M-H]− at m/z 245.1175 in negative ion mode. Therefore,
negative ion mode was used with the following optimized conditions for LOX: spray voltage of 2.5 kV,
capillary temperature of 320 ◦ C, auxiliary gas at 12 aux unit, aux gas heater temperature of 200 ◦ C,
sheath gas at 35 aux units, and S-lens RF level of 50. LOX and its metabolites were separated using
a 150 mm × 2.1 mm, 2.6-μm reverse-phase liquid chromatography column, Kinetex® C18 column
(Phenomenex, CA, USA) at 40 ◦ C. Furthermore, MS-grade solvents were used as the mobile phase
in gradient elution mode: 0.1% aqueous formic acid as Solvent A and 0.1% formic acid in ACN as
Solvent B. The flow was set to 0.22 mL/min with a gradient condition of Solvent B as 10% between 0
and 0.5 min, 10–50% between 0.5 and 21.5 min, 50–95% between 21.5 and 22.5 min, 95% between 22.5
and 25.5 min, 95–10% between 25.5 and 25.6 min, and 10% between 25.6 and 30 min.

2.5. Method Validation

Stock solutions of LOX, cis-LOX, and trans-Lox (40 mg/mL) were prepared in methanol to generate
a calibration curve and linearity. The concentrations of LOX in plasma used to generate the calibration
curve were 0.1, 0.2, 0.5, 1.0, 5.0, 10.0, 20.0, and 40.0 μg/mL. The concentrations of cis-LOX and trans-LOX
used to generate their calibration curves were 0.2, 0.5, 1.0, 5.0, 10.0, 20.0, and 40.0 μg/mL. An amount of
10 μL of these samples was processed as mentioned above for LC-MS/MS analysis. Area peak ratios of
analytes/IS versus concentration of samples were used to prepare the calibration curves. The calibration
equation of LOX was y = 8 × 10−7 x + 0.0002 (R2 = 0.997). The calibration equations for cis-LOX and
trans-LOX were y = 2 × 10−6 x + 0.0002 (R2 = 0.996) and y = 1 × 10−6 x − 0.0002 (R2 = 0.997), respectively.
To evaluate the accuracy and precision of LOX measurements, mouse plasma was spiked with a
known concentration of LOX or QC samples at 0.2, 1.0, 10.0, 40.0 μg/mL (n = 5). Similarly, the accuracy
and precision of cis-LOX and trans-LOX measurements were evaluated by spiking mouse plasma
with a known concentration of either compound or QC samples at 0.5, 5.0, and 40.0 μg/mL (n = 5).
Moreover, the accuracy and precision of intraday and interday were analyzed on the same day and
five consecutive days at each concentration.

2.6. Statistical Analysis

The results are presented as the mean and the standard error of the mean. A Student’s unpaired
t-test was applied in the statistical analyses of the obtained results using Graph Pad Prism. Results
with a p-value ≤ 0.05, ≤ 0.01 and ≤0.001 were considered statistically significant.

3. Results

3.1. Identification of Loxoprofen and Its Metabolites

To determine the plasma concentration of LOX and its metabolites via LC-MS/MS analysis,
the produced product ions were checked and optimized for each compound. The ion intensities of LOX,
cis-LOX, and trans-LOX were high in the negative mode of ionization; therefore, all the conditions

14
Pharmaceutics 2019, 11, 479

for analysis used the negative mode for LC-MS/MS analysis. The MRM transitions chosen for LOX,
cis-LOX, and trans-LOX were m/z 245.0 → 83.1, 247.1 → 202.2, and 247.1 → 203.1, respectively [15].
Representative MRM chromatograms of LOX, cis-LOX, trans-LOX, and the IS in mouse plasma are
presented in Figure S1. LOX, cis-LOX, trans-LOX, and the IS were eluted at 8.5, 8.0, 8.2, and 8.9 min,
respectively. No endogenous sources of interference were observed. LOX and its two metabolites were
evaluated for linearity, precision, and accuracy. The calibration curves calculated within the range
of 0.1–40.0 μg/mL for LOX and within the range of 0.2–40.0 μg/mL for cis-LOX and trans-LOX were
linear. The precision (RSD %) range of LOX, cis-LOX, and trans-LOX was 1.8–12.9%, and the accuracy
(RE %) range was less than 14.7% (Table S1). Thus, the values were within the acceptable range and the
method was accurate and precise.

3.2. Evaluation Model for Determination of LOX–Drug Interaction

To evaluate the CYP3A-induced LOX interaction, we prepared an experimental model that
regulated CYP3A activity by administering an inducer (DEX) and an inhibitor (KTC) of CYP3A into
mice. Briefly, DEX in corn oil was administered up to 3 consecutive days to induce CYP3A. Inhibition
of CYP3A was induced by a single dose of KTC in 10% ethanol. Only vehicle groups were administered
with their respective solvents without the addition of a CYP3A inducer or inhibitor. The induction and
inhibition of CYP3A were confirmed with CYP assay using five different probe substrates (Figure S2).
DEX increased the metabolism of CYP3A substrate (midazolam) by approximately 10-fold when
compared to the vehicle (VH) group. KTC significantly decreased the metabolism of CYP3A4 substrate
(midazolam) when compared to the VH group (Figure S2). However, the metabolic activities of other
CYP enzymes were unaffected between VH and treated groups. Thus, we validated our method of
specifically regulating CYP3A activity via DEX and KTC.

3.3. Pharmacokinetic Analysis

The validated method was applied to determine the concentration of LOX, cis-LOX, and trans-LOX
in mice pretreated with DEX or KTC. After pretreatment, LOX was orally administered to mice
(20 mg/kg) after a 12-h fasting period. The plasma concentrations of LOX, cis-LOX, and trans-LOX
were significantly decreased in the DEX pretreated group (Figure 1). In the KTC pretreated group,
the plasma concentrations of cis-LOX, and trans-LOX were significantly increased and the plasma
concentration of LOX was also increased but not significantly (Figure 1).
The PK parameters of LOX, cis-LOX, and trans-Lox in the VH- (corn oil) and DEX-treated groups
are shown in Table 1. Although blood was collected for up to 240 min, LOX and its metabolites were not
detected after 60 min. The Cmax , AUC(0–60) , and AUC(0–∞) of all three compounds were significantly
lowered in the DEX-pretreated group as opposed to the VH group. In the DEX-treated group, the
Cmax of LOX, cis-LOX, and trans-LOX was 2.5 ± 0.2, 1.1 ± 0.2, and 2.1 ± 0.2 μg/mL, respectively,
and the AUC(0–60) of the three compounds was 53.5 ± 6.1, 29.9 ± 4.4, and 67.6 ± 5.7 μg·min/mL,
respectively. However, the Tmax from all three compounds did not show any statistical difference
when compared to the VH group. In contrast to the VH group, AUC(0–∞) of LOX, cis-LOX, and
trans-Lox in the DEX-treated group indicated a lower area under plasma concentration (56.2 ± 6.9,
31.5 ± 4.4, and 85.8 ± 5.0, respectively) over an extended time period. Moreover, the elimination
T1/2 of LOX, cis-LOX, and trans-Lox in the VH group was 14.9 ± 0.6, 12.3 ± 0.3, and 18.2 ± 0.6 min,
respectively, which is significantly different from the respective elimination T1/2 (12.0 ± 0.7, 13.9 ± 0.6,
and 26.4 ± 1.6 min) generated by the DEX-pretreated group.

15
Pharmaceutics 2019, 11, 479

Figure 1. Mean plasma concentration versus time profiles of LOX, cis-LOX, and trans-LOX in either the
presence of a CYP3A4 inducer (DEX) or inhibitor (KTC) with their respective vehicle. (A) Mean plasma
concentration versus time profiles after i.p. administration of either VH (corn oil) or DEX (40 mg/kg)
for 3 consecutive days. The plasma concentrations of all the compounds in the VH and DEX groups
showed significant decrement up to 60 min in the DEX-treated group as compared to its VH group.
The bars represent standard error (SE) (n = 3). (B) Mean plasma concentration versus time profiles after
a single dose i.p. administration of either VH (10% ethanol) or KTC (60 mg/kg). In the KTC group,
LOX, cis-LOX, and trans-LOX showed increased mean plasma concentrations as opposed to the VH
group. The bars indicate standard error (SE) (n = 3).

Furthermore, the PK parameters for LOX, cis-LOX, and trans-LOX in the VH and KTC groups
are represented in Table 2. The Cmax of cis-LOX, and trans-LOX in the VH group (1.2 ± 0.1, and
2.1 ± 0.2 μg/mL, respectively) were significantly lower than those in the KTC-treated group (1.6 ± 0.1,

16
Pharmaceutics 2019, 11, 479

and 3.1 ± 0.3 μg/mL, respectively). However, the Tmax of these three compounds did not show
significant variation between the VH and KTC groups. In contrast to the Tmax , the elimination T1/2 of
trans-LOX in the VH group (26.0 ± 0.5 min) was statistically different from that in the KTC-treated
group (19.8 ± 0.7 min). Moreover, The AUC(0–60) for, cis-LOX, and trans-LOX in the KTC-treated
group was 49.0 ± 5.9, and 80.4 ± 9.6 μg·min/mL, respectively, which were higher than the respective
values generated by the VH group. Altogether, the PK data generated by the DEX- and KTC-treated
groups and their respective VH indicate that DEX and KTC significantly affected the PK of cis-LOX,
and trans-LOX but PK of LOX was only affected by DEX, even though the formation of cis-LOX and
trans-LOX was regulated by CR.

Table 1. Pharmacokinetic parameters of loxoprofen (LOX), cis-LOX, and trans-LOX in VH- (corn oil)
and dexamethasone (DEX)-treated groups.

Analytes Parameters VH (Corn Oil) DEX

Cmax (μg/mL) 4.8 ± 0.9 2.5 ± 0.2 **
Tmax (min) 5.0 ± 0.0 5.0 ± 0.0
LOX AUC(0–60) (μg·min/mL) 95.7 ± 14.5 53.5 ± 6.1 **
T1/2 (min) 14.9 ± 0.6 12.0 ± 0.7 *
AUC(0–∞) (μg·min/mL) 102.2 ± 15.0 56.2 ± 6.9 ***
Cmax (μg/mL) 2.4 ± 0.3 1.1 ± 0.2 **
Tmax (min) 10.4 ± 0.7 10.0 ± 1.0
cis-LOX AUC(0–60) (μg·min/mL) 72.7 ± 12.4 29.9 ± 4.4 **
T1/2 (min) 12.3 ± 0.3 13.9 ± 0.6 *
AUC(0–∞) (μg·min/mL) 75.8 ± 12.8 31.5 ± 4.4 **
Cmax (μg/mL) 4.4 ± 0.5 2.1 ± 0.2 **
Tmax (min) 9.1 ± 1.2 10.4 ± 0.9
trans-LOX AUC(0–60) (μg·min/mL) 137.4 ± 19.0 67.6 ± 5.7 **
T1/2 (min) 18.2 ± 0.6 26.4 ± 1.6 **
AUC(0–∞) (μg·min/mL) 154.8 ± 19.2 85.8 ± 5.0 **
All data are expressed as the mean ± standard error (SE) (n = 3). Cmax : maximum plasma concentration; AUC(0–60) :
area under the plasma concentration-time curve (AUC) from 0 to 60 min; Tmax : time to reach maximum plasma
concentration; T1/2 : elimination half-life; AUC(0–∞) : area under the plasma concentration-time curve from 0 to
infinite time. * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

Table 2. Pharmacokinetic parameters of LOX and its metabolites in 10% ethanol (VH) and KTC-
treated groups.

Analytes Parameters VH (10% Ethanol) KTC

Cmax (μg/mL) 2.7 ± 0.3 3.5 ± 0.5
Tmax (min) 5.0 ± 0.0 5.0 ± 0.0
LOX AUC(0–60) (μg·min/mL) 54.7 ± 4.6 66.6 ± 6.6
T1/2 (min) 14.6 ± 1.1 12.2 ± 0.9
AUC(0–∞) (μg·min/mL) 59.1 ± 5.6 68.9 ± 6.5
Cmax (μg/mL) 1.2 ± 0.1 1.6 ± 0.1 *
Tmax (min) 13.3 ± 0.8 11.6 ± 0.8
cis-LOX AUC(0–60) (μg·min/mL) 36.0 ± 3.9 49.0 ± 5.9 *
T1/2 (min) 13.3 ± 0.9 13.4 ± 0.6
AUC(0–∞) (μg·min/mL) 38.4 ± 4.4 51.9 ± 6.1 *
Cmax (μg/mL) 2.1 ± 0.2 3.1 ± 0.3 **
Tmax (min) 11.6 ± 0.8 11.7 ± 0.8
trans-LOX AUC(0–60) (μg·min/mL) 69.9 ± 6.0 80.4 ± 9.6 *
T1/2 (min) 26.0 ± 0.5 19.8 ± 0.7 **
AUC(0–∞) (μg·min/mL) 90.9 ± 7.1 94.7 ± 11.0
All data are expressed as the mean ± standard error (SE) (n = 3). Cmax : maximum plasma concentration; AUC(0–60) :
area under the plasma concentration-time curve (AUC) from 0 to 60 min; Tmax : time to reach maximum plasma
concentration; T1/2 : elimination half-life; AUC(0–∞) : area under the plasma concentration-time curve from 0 to
infinite time. * p ≤ 0.05, ** p ≤ 0.01 and *** p ≤ 0.001.

17
Pharmaceutics 2019, 11, 479

3.4. Metabolism and Metabolite Identification of LOX During DEX or KTC Treatment
The purpose of this study was to identify changes made by DEX and/or KTC on CR-mediated
LOX metabolites (cis-LOX and trans-LOX) and also on CYP-mediated LOX metabolites (OH-LOX).
During this PK study, OH-LOX could not be detected. Therefore, a full scan in Q Exactive Focus
was used to identify all the metabolites present in plasma. A parallel reaction monitoring (PRM)
mode was applied to confirm the metabolites through fragmentation patterns using collision induced
dissociation (CID). Seven metabolites were confirmed after comparing the EICs of the test samples
with blank plasma (Figures S3 and S4). LOX and all of its metabolites were detected in the negative
ionization mode. To confirm the metabolites of LOX, their MS/MS fragmentation was checked
(Figure S5). LOX (C15 H17 O3 ) was detected at a retention time of 18 min with only one major fragment
ion 83.0492 (C5 H7 O). Trans-LOX (M1, C15 H19 O3 ) with an m/z ratio of 247.1339, eluted at 17.1 min with
major fragment ions 233.1181 (C14 H17 O3 , –CH2 ), 217.1230 (C14 H17 O2 , –CH2 O), 201.1279 (C14 H17 O),
and 191.1071 (C12 H15 O2 , –C3 H4 O). Cis-LOX (M2, C15 H19 O3 ), having the m/z ratio 247.1336, was
detected at a retention time of 17.5 min. Its major fragments were 217.1230 (C14 H17 O2 , –CH2 O) and
191.1071 (C12 H15 O2 , –C3 H4 O). M3 and M4 are OH-LOX (C15 H17 O4 ), having an m/z ratios of 261.1138
and 261.1133, and they were eluted at a retention time of 11.8 and 12.5 min, respectively. The MS/MS
spectra of these metabolites were 99.0441 (C5 H7 O2 , –C10 H10 O2 ) and 81.0335 (C5 H5 O, –C10 H12 O3 )
for M3 and only a single major product ion 99.0441 (C5 H7 O2 ) for M4. M5 is hydroxy trans-LOX
(C15 H19 O4 ), having an m/z ratio of 263.1288, and it was detected at a retention time of 11 min. Its major
fragment ions were 233.1181 (C14 H17 O3 , –CH2 O), 207.1022 (C12 H15 O3 , –C3 H4 O), 133.0650 (C9 H9 O,
–C6 H10 O3 ), and 99.0442 (C5 H7 O2 , –C10 H12 O2 ). M6 was identified as a taurine conjugate (C17 H24 O5 NS)
whose m/z ratio was 354.1382, and it was detected at a retention time of 13.3 min. Its major fragments
were 149.9859 (C3 H4 O4 NS, –C14 H20 O), 124.0065 (C2 H6 O3 NS, –C15 H18 O2 ), and 106.9798 (C2 H3 O3 S,
–C15 H21 O2 N). M7 was a glucuronide conjugate (C21 H25 O9 ), having an m/z ratio 421.1514, and it eluted
at a retention time of 14 min. Its major fragments were 245.1182 (C15 H17 O3 ), 193.0348 (C6 H9 O7 ),
175.0242 (C6 H7 O6 ), and 83.0492 (C5 H7 O). Parent compounds and their fragments detected during the
study are represented in Table 3.

Table 3. Identified metabolites of LOX in mouse plasma using HRMS.

Parent Ions Elemental Error

Compounds Product Ions (m/z) Description
(m/z) Composition (ppm)
Lox 245.1179 C15 H17 O3 0.4 83.0492 LOX
M1 247.1339 C15 H19 O3 2.0 233.1181, 217.1230, 201.1279, 191.1071 Trans-LOX
M2 247.1336 C15 H19 O3 0.5 217.1230, 191.1071 Cis-LOX
M3 261.1138 C15 H17 O4 4.2 99.0441, 81.0335 OH-LOX
M4 261.1133 C15 H17 O4 2.3 99.0441 OH-LOX
M5 263.1288 C15 H19 O4 1.9 233.1181, 207.1022, 133.0650, 99.0442 OH-trans-LOX
M6 354.1382 C17 H24 O5 NS 2.0 149.9859, 124.0065, 79.9563 Taurine conjugate
Glucuronide
M7 421.1514 C21 H25 O9 3.6 245.1182, 193.0348, 175.0242, 83.0492
conjugate

We identified the effects of CYP3A induction and inhibition on seven different known metabolites
of LOX. The general characteristics of LOX, its metabolites, and their concentration in different
groups are described in (Table S2). In this study, we found that the concentration of LOX, trans-LOX
(M1), and cis-LOX (M2) significantly decreased in the DEX-treated group (74.1 ± 6.3%, 80.1 ± 1.2%,
and 61.9 ± 3.9%, respectively) and increased in the KTC-treated group (178.2 ± 8.3%, 158.9 ± 11.9%,
and 173.1 ± 5.8%, respectively). Furthermore, the concentrations of M3, M4, and M5 significantly
increased in the DEX-treated group (160.5 ± 4.1%, 440.4 ± 8.3%, and 286.3 ± 11.4%, respectively), and
only the concentrations of M4 and M5 decreased in the KTC-treated group (93.6 ± 1.9% and 90.2 ± 2.7%,
respectively). The taurine conjugate (M6) decreased in both the DEX- (65.3 ± 2.84%) and KTC-treated
(91.2 ± 2.0%) groups. In contrast, the glucuronide conjugate increased in both the DEX- (174.4 ± 6.5%)
and KTC-treated (275.7 ± 14.1%) groups. M6 and M7 are phase 2 metabolites, which indicates that

18
Pharmaceutics 2019, 11, 479

CYPs are not the main enzyme involved in the formation of these metabolites. The concentration of
LOX and its metabolites in the VH, DEX-treated, and KTC-treated groups is represented graphically
in Figure 2. DEX and KTC are well-known CYP3A modulators, and in this study, we found that the
pretreatment of DEX or KTC had a significant effect on the concentration of both CYP-mediated and
CR-mediated metabolites.

Figure 2. Comparison of loxoprofen (LOX) and its metabolites in male ICR mice treated with VH,
DEX, or KTC. (A) The relative concentration of LOX and its metabolites after DEX administration (i.p.
40 mg/kg for consecutive 3 days, n = 3) compared to VH (i.p. corn oil for consecutive 3 days, n = 3).
(B) The relative concentration of LOX and its metabolites after KTC administration (single dose i.p.
60 mg/kg, n = 3) compared to VH (10% ethanol, n = 3).

DEX and KTC affected the concentration and the PK of CYP-mediated metabolites, which, in turn,
influenced the concentration and the PK of cis- and trans-LOX. The metabolites detected in this study
are summarized in Figure 3.

Figure 3. Metabolic pathway of loxoprofen and its metabolites.

4. Discussion
LOX is a nonselective COX inhibitor that is administered for the management of pain and
inflammation, and it is well tolerated by patients [2,24,25]. Although the pharmacokinetics and
metabolism of LOX are well defined, its interaction(s) with CYP enzymatic pathways have not been
fully characterized [3,6–10,13,15,26–29]. However, CYP3A regulating commercial drugs are widely in

19
Pharmaceutics 2019, 11, 479

clinical use [16,17] and there may be the chance to use CYP3A-regulating drugs and LOX together.
Although CYP3A does not generate the LOX active metabolite, it may be possible to influence its
generation by regulating CYP3A activity. To that end, we evaluated whether the rate of LOX active
metabolite formation is influenced by CYP3A activity. We also investigated the effects of DEX and KTC,
which are widely used CYP3A modulators, on the metabolism and PK of LOX, cis-LOX, and trans-LOX.
DEX is a steroidal drug used in the treatment of many conditions such as skin diseases, allergies,
rheumatic disorder, asthma, and certain autoimmune diseases [30,31]. It is also effective in treating
various cancers such as central nervous system tumors, brain metastases, advanced melanoma,
leukemia, lymphoma, and multiple myeloma. Additionally, DEX is effective at combating the side
effects of chemotherapy [32,33]. DEX is metabolized by CYP3A4 and CYP17A, but the CYP17A metabolic
pathway plays no major role in its in vivo metabolism [34–36]. DEX has been used as a CYP3A4
inducer in clinical studies. Interestingly, rifampicin, rifabutin, phenytoin, phenobarbital, primidone,
carbamazepine, etc., are strong CYP3A4 inducers [16]. It has also been reported that a single dose
(50 mg/kg) of DEX can induce the activity of CYP3A. However, persistent administration of DEX could
make stable induction of CYP3A activity, which was independent of inducer administration [37,38].
Moreover, chronic DEX administration likely leads to the autoinduction of CYP3A, which has a direct
impact on the PK parameters of DEX [38]. Nevertheless, patients may suffer from an increase in
substrate concentration after they stop using the CYP3A inducer if they take both the substrate (LOX)
and the inducer (DEX) of CYP3A [37].
KTC is a broad-spectrum antifungal drug used in the treatment of many fungal infections such as
blastomycosis, coccidioidomycosis, histoplasmosis, etc. [39,40]. KTC is a known reversible inhibitor
of CYP3A4 [16] and displays hepatotoxicity through immune-allergic mechanisms, which limits its
therapeutic use [41,42]. Usually, CYP3A4 inhibitors are divided into reversible inhibitors (such as KTC,
itraconazole, terfenadine, astemizole, and quinidine) [43] and irreversible inhibitors (such as gestodene
and levonorgestrel) [44]. Many studies also reported the biotransformation of KTC through oxidation,
O-dealkylation, hydroxylation, and FMO (via the UGT1A4 metabolic pathway) [42,45]. Furthermore,
one study revealed that KTC showed dose-dependent kinetics, indicating that its metabolism occurs in
the liver [46]. A recent study also revealed that KTC metabolism is similar in humans and mice, which
may help to resolve the issue regarding drug metabolism and toxicology [42]. Nevertheless, a sudden
increase or decrease in the Cmax of drugs and toxic metabolites could result from either the induction
or inhibition of metabolizing enzymes [47].
CYP3A4 is the major CYP involved in LOX metabolism, but it does not form its active
metabolite [15]. The active metabolite of LOX (trans-LOX) is formed via CR pathway [1,5]. Similarly to
previous studies, we also found that DEX and KTC were strong CYP3A activity modulators [36,48,49].
In addition, we determined that the Cmax , AUC(0–60) , and AUC(0–∞) of LOX, cis-LOX, and trans-LOX
were significantly decreased in DEX-treated mice. However, the Cmax , AUC(0–60) , and AUC(0–∞) of
cis-LOX, and trans-LOX are significantly increased in KTC-treated mice. Furthermore, DEX increased
the concentration of the OH-LOX metabolite and decreased the concentration of the active metabolite.
This may lead to a decrease in the activity of trans-LOX during the subsequent administration of LOX
and DEX. We also observed that changes in CYP pathway activity can modulate CR pathway activity
regarding LOX, even though CYP3A4 does not participate in LOX active metabolite formation.
These results indicate that the co-administration of LOX and DEX may lead to a decrease
in the pharmacological activity of LOX by decreasing the concentration of the active metabolite.
We also observed that the concentration of active metabolites catalyzed by CR increased and that the
concentration of inactive metabolites decreased during the co-administration of LOX and the CYP3A
inhibitor, KTC. Therefore, if LOX and KTC are administered together, the pharmacological activity
of LOX may be enhanced by increasing the active metabolite. Subsequently, this may also increase
toxicity. In a previous study, LOX was associated with an increase in small bowel mucosal injury,
erosive gastritis, gastroduodenal ulcers, etc., during concomitant use of a proton pump inhibitor such
as lansoprazole [1,50,51]. Since lansoprazole is a known inhibitor of CYP3A and CYP2C19 [52,53],

20
Pharmaceutics 2019, 11, 479

these side effects may have been caused by changes in the blood levels of LOX and its active metabolite
caused by the concomitant use of LOX and lansoprazole. It has been reported that LOX slightly
inhibited the metabolism of tacrolimus by interacting with CYP3A in human liver microsomes [54].
Similarly, tizanidine is typically used to manage muscle spasticity and pain [55]. However, it has
been reported that combination therapy of tizanidine and LOX for neck pain increased the risk of
irreversible symptomatic bradycardia via CYP inhibition [56].
These previous studies clearly show that LOX can interact with drugs that modulate CYP activity
even though its active metabolite is not formed by CYP enzymes. Considering the pharmacokinetics
and metabolism of LOX and its metabolites, the use of LOX with CYP modulators (e.g., DEX and KTC)
may result in the decreased pharmacological action of LOX or may cause from minor toxicity to major
toxicity by interacting with CYP3A.

5. Conclusions
The pharmacokinetic parameters of LOX and its active metabolite were significantly altered when
LOX was co-administered with CYP3A activity modulators. In clinical practice, LOX is administered
concurrently with many other drugs. Therefore, more studies are needed to assess the possible
interactions of LOX with CYP enzymes. In this study, pharmacodynamic interactions were not
evaluated; however, they will be evaluated in future studies.

Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4923/11/9/479/s1,

Figure S1: Representative chromatograms: (A) blank plasma (B), standard loxoprofen (LOX), and its metabolites
in plasma. LOX and its metabolites in plasma after administering 20 mg/mL LOX to the (C) VH (Corn oil) group
(n = 3), (D) DEX-treated group (n = 3), (E) VH (10% Ethanol) group (n = 3), and (F) KTC-treated group (n = 3),
Figure S2: CYP induction by dexamethasone (n = 3) (A), and CYP inhibition by ketoconazole (n = 3) (B), Figure S3:
Representative EICs of Loxoprofen (LOX) and its metabolites in mouse plasma: VH (Corn oil) group (n = 3) (A),
DEX-treated group (n = 3) (B), VH (10% Ethanol) group (n = 3) (C), and KTC-treated group (n = 3) (D), Figure S4:
Representative EICs of loxoprofen (LOX) and its metabolites in mouse blank plasma (A) as well as standard LOX
and tolbutamide (IS) (B), Figure S5: MS/MS spectra of Loxoprofen (A), M1 (B), M2 (C), M3 (D), M4 (E), M5 (F),
M6 (G), and M7 (H), Table S1: Method validation table (n = 3), Table S2: Characterization of loxoprofen (LOX)
metabolites identified in mouse plasma along with their average % of area (n = 3).
Author Contributions: Conceptualization, S.L.; Data Curation, S.P. and S.L.; Formal Analysis, S.P.; P.C.; R.S. and
Y.K., Funding Acquisition, T.L. and S.L.; Methodology, S.P.; Resources, A.S., T.C.J., and E.-S.L., Writing – Original
Draft Preparation, S.P.; Writing – Review & Editing, J.-H.K. and S.L.
Funding: This research was supported by a grant from the Korea Health Technology R&D Project through the
Korea Health Industry Development Institute, funded by the Ministry of Health & Welfare, Republic of Korea
(HI16C1501), and the Korea Basic Science Institute (KBSI) National Research Facilities & Equipment Center (NFEC)
grant funded by the Korea Government (Ministry of Education) (No.2019R1A6C1010001).
Conflicts of Interest: The authors report no conflicts of interest.

References
1. Greig, S.L.; Garnock-Jones, K.P. Loxoprofen: A review in pain and inflammation. Clin. Drug Investig. 2016,
36, 771–781. [CrossRef] [PubMed]
2. Wan, D.; Zhao, M.; Zhang, J.; Luan, L. Development and In Vitro-In Vivo Evaluation of a Novel
Sustained-Release Loxoprofen Pellet with Double Coating Layer. Pharmaceutics 2019, 11, 260. [CrossRef]
[PubMed]
3. Helmy, S.A. Pharmacokinetics and Bioequivalence Evaluation of 2 Loxoprofen Tablets in Healthy Egyptian
Male Volunteers. Clin. Pharmacol. Drug Dev. 2013, 2, 173–177. [CrossRef] [PubMed]
4. Yamakawa, N.; Suemasu, S.; Watanabe, H.; Tahara, K.; Tanaka, K.-I.; Okamoto, Y.; Ohtsuka, M.; Maruyama, T.;
Mizushima, T. Comparison of pharmacokinetics between loxoprofen and its derivative with lower ulcerogenic
activity, fluoro-loxoprofen. Drug Metab. Pharmacok. 2013, 28, 118–124. [CrossRef]
5. Kang, H.-A.; Cho, H.-Y.; Lee, Y.-B. Bioequivalence of hana loxoprofen sodium tablet to dongwha Loxonin®
tablet (Loxoprofen Sodium 60 mg). J. Pharm. Investig. 2011, 41, 117–123. [CrossRef]
6. Moore, N.; Pollack, C.; Butkerait, P. Adverse drug reactions and drug–drug interactions with over-the-counter
NSAIDs. Ther. Clin. Risk. Manag. 2015, 11, 1061–1075. [PubMed]

21
Pharmaceutics 2019, 11, 479

7. Sawamura, R.; Kazui, M.; Kurihara, A.; Izumi, T. Absorption, distribution, metabolism and excretion of
loxoprofen after dermal application of loxoprofen gel to rats. Xenobiotica 2014, 44, 1026–1038. [CrossRef]
8. Takeda, M.; Khamdang, S.; Narikawa, S.; Kimura, H.; Hosoyamada, M.; Cha, S.H.; Sekine, T.; Endou, H.
Characterization of methotrexate transport and its drug interactions with human organic anion transporters.
J. Pharmacol. Exp. Ther. 2002, 302, 666–671. [CrossRef]
9. Yue, Z.; Shi, J.; Jiang, P.; Sun, H. Acute kidney injury during concomitant use of valacyclovir and loxoprofen:
Detecting drug-drug interactions in a spontaneous reporting system. Pharmacoepidemiol. Drug Saf. 2014, 23,
1154–1159. [CrossRef]
10. Takahashi, H.; Kashima, T.; Kimura, S.; Murata, N.; Takaba, T.; Iwade, K.; Abe, T.; Tainaka, H.; Yasumori, T.;
Echizen, H. Pharmacokinetic interaction between warfarin and a uricosuric agent, bucolome: Application of
in vitro approaches to predicting in vivo reduction of (S)-warfarin clearance. Drug Metab. Dispos. 1999, 27,
1179–1186.
11. Dinger, J.; Meyer, M.R.; Maurer, H.H. Development of an in vitro cytochrome P450 cocktail inhibition assay
for assessing the inhibition risk of drugs of abuse. Toxicol. Lett. 2014, 230, 28–35. [CrossRef] [PubMed]
12. Ab Rahman, N.S.; Abd Majid, F.A.; Wahid, A.; Effendy, M.; Zainudin, A.N.; Zainol, S.N.; Ismail, H.F.;
Wong, T.S.; Tiwari, N.K.; Giri, S. Evaluation of Herb-Drug Interaction of Synacinn™ and Individual
Biomarker through Cytochrome 450 Inhibition Assay. Drug Metab. Lett. 2018, 12, 62–67. [CrossRef]
[PubMed]
13. Tanaka, Y.; Nishikawa, Y.; Hayashi, R. Species differences in metabolism of sodium 2-[4-(2-
oxocyclopentylmethyl)-phenyl] propionate dihydrate (loxoprofen sodium), a new anti-inflammatory agent.
Chem. Pharm. Bull. 1983, 31, 3656–3664. [CrossRef] [PubMed]
14. Jhee, O.H.; Lee, M.H.; Shaw, L.M.; Lee, S.E.; Park, J.H.; Kang, J.S. Pharmacokinetics and bioequivalence study
of two brands of loxoprofen tablets in healthy volunteers. Arzneimittelforschung 2007, 57, 542–546. [CrossRef]
[PubMed]
15. Shrestha, R.; Cho, P.; Paudel, S.; Shrestha, A.; Kang, M.; Jeong, T.; Lee, E.-S.; Lee, S. Exploring the Metabolism
of Loxoprofen in Liver Microsomes: The Role of Cytochrome P450 and UDP-Glucuronosyltransferase in Its
Biotransformation. Pharmaceutics 2018, 10, 112. [CrossRef]
16. Zhou, S.-F. Drugs behave as substrates, inhibitors and inducers of human cytochrome P450 3A4.
Curr. Drug Metab. 2008, 9, 310–322. [CrossRef] [PubMed]
17. Basheer, L.; Kerem, Z. Interactions between CYP3A4 and dietary polyphenols. Oxid. Med. Cell. Longev.
2015, 2015. [CrossRef]
18. Naruto, S.; Terada, A. Synthesis of the eight possible optically active isomers of 2-[4-(2-
hydroxycyclopentylmethyl) phenyl] propionic acid. Chem. Pharm. Bull. 1983, 31, 4319–4323. [CrossRef]
19. Seervi, M.; Lotankar, S.; Barbar, S.; Sathaye, S. Assessment of cytochrome P450 inhibition and induction
potential of lupeol and betulin in rat liver microsomes. Drug Metab. Pers. Ther. 2016, 31, 115–122. [CrossRef]
20. Seneca, N.; Zoghbi, S.S.; Shetty, H.U.; Tuan, E.; Kannan, P.; Taku, A.; Innis, R.B.; Pike, V.W. Effects of
ketoconazole on the biodistribution and metabolism of [11 C] loperamide and [11 C] N-desmethyl-loperamide
in wild-type and P-gp knockout mice. Nucl. Med. Biol. 2010, 37, 335–345. [CrossRef]
21. Shrestha, R.; Kim, J.-H.; Nam, W.; Lee, H.S.; Lee, J.-M.; Lee, S. Selective inhibition of CYP2C8 by fisetin and
its methylated metabolite, geraldol, in human liver microsomes. Drug Metab. Pharmacok. 2018, 33, 111–117.
[CrossRef] [PubMed]
22. Sim, J.; Nam, W.; Lee, D.; Lee, S.; Hungchan, O.; Joo, J.; Liu, K.-H.; Han, J.Y.; Ki, S.H.; Jeong, T.C.
Selective induction of hepatic cytochrome P450 2B activity by leelamine in vivo, as a potent novel inducer.
Arch. Pharmacal. Res. 2015, 38, 725–733. [CrossRef] [PubMed]
23. Bradford, M.M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing
the principle of protein-dye binding. Anal. Biochem. 1976, 72, 248–254. [CrossRef]
24. Hamaguchi, M.; Seno, T.; Yamamoto, A.; Kohno, M.; Kadoya, M.; Ishino, H.; Ashihara, E.; Kimura, S.;
Tsubakimoto, Y.; Takata, H. Loxoprofen sodium, a non-selective NSAID, reduces atherosclerosis in mice by
reducing inflammation. J. Clin. Biochem. Nutr. 2010, 47, 138–147. [CrossRef] [PubMed]
25. Sekiguchi, H.; Inoue, G.; Nakazawa, T.; Imura, T.; Saito, W.; Uchida, K.; Miyagi, M.; Takahira, N.; Takaso, M.
Loxoprofen sodium and celecoxib for postoperative pain in patients after spinal surgery: A randomized
comparative study. J. Orthop. Sci. 2015, 20, 617–623. [CrossRef] [PubMed]

22
Pharmaceutics 2019, 11, 479

26. Nagashima, H.; Tanaka, Y.; Watanabe, H.; Hayashi, R.; Kawada, K. Optical inversion of (2R)-to (2S)-isomers
of 2-[4-(2-oxocyclopentylmethyl)-phenyl] propionic acid (loxoprofen), a new anti-inflammatory agent, and its
monohydroxy metabolites in the rat. Chem. Pharm. Bull. 1984, 32, 251–257. [CrossRef] [PubMed]
27. Foti, R.S.; Dalvie, D.K. Cytochrome P450 and non–cytochrome P450 oxidative metabolism: Contributions to
the pharmacokinetics, safety, and efficacy of xenobiotics. Drug Metab. Dispos. 2016, 44, 1229–1245. [CrossRef]
28. Tanaka, Y.; Nishikawa, Y.; Matsuda, K.; Yamazaki, M.; Hayashi, R. Purification and some properties of
ketone reductase forming an active metabolite of sodium 2-[4-(2-oxocyclopentylmethyl)-phenyl] propionate
dihydrate (loxoprofen sodium), a new anti-inflammatory agent, in rabbit liver cytosol. Chem. Pharm. Bull.
1984, 32, 1040–1048. [CrossRef]
29. Sawamura, R.; Kazui, M.; Kurihara, A.; Izumi, T. Pharmacokinetics of loxoprofen and its active metabolite
after dermal application of loxoprofen gel to rats. Pharm. 2015, 70, 74–80.
30. Verhoef, C.M.; van Roon, J.A.; Vianen, M.E.; Lafeber, F.P.; Bijlsma, J.W. The immune suppressive effect of
dexamethasone in rheumatoid arthritis is accompanied by upregulation of interleukin 10 and by differential
changes in interferon γ and interleukin 4 production. Ann. Rheum. Dis. 1999, 58, 49–54. [CrossRef]
31. Liu, D.; Ahmet, A.; Ward, L.; Krishnamoorthy, P.; Mandelcorn, E.D.; Leigh, R.; Brown, J.P.; Cohen, A.; Kim, H.
A practical guide to the monitoring and management of the complications of systemic corticosteroid therapy.
Allergy Asthma Clin. Immunol. 2013, 9, 30. [CrossRef] [PubMed]
32. Cook, A.M.; McDonnell, A.M.; Lake, R.A.; Nowak, A.K. Dexamethasone co-medication in cancer
patients undergoing chemotherapy causes substantial immunomodulatory effects with implications for
chemo-immunotherapy strategies. Oncoimmunology 2016, 5, e1066062. [CrossRef] [PubMed]
33. Pufall, M.A. Glucocorticoids and Cancer. Adv. Exp. Med. Biol. 2015, 872, 315–333. [CrossRef] [PubMed]
34. Al Katheeri, N.; Wasfi, I.; Lambert, M.; Albo, A.G.; Nebbia, C. In vivo and in vitro metabolism of
dexamethasone in the camel. Vet. J. 2006, 172, 532–543. [CrossRef] [PubMed]
35. Tomlinson, E.; Maggs, J.; Park, B.; Back, D. Dexamethasone metabolism in vitro: Species differences.
J. Steroid. Biochem. 1997, 62, 345–352. [CrossRef]
36. Tomlinson, E.; Lewis, D.; Maggs, J.; Kroemer, H.; Park, B.; Back, D. In vitro metabolism of dexamethasone
(DEX) in human liver and kidney: The involvement of CYP3A4 and CYP17 (17, 20 LYASE) and molecular
modelling studies. Biochem. Pharmacol. 1997, 54, 605–611. [CrossRef]
37. Iwanaga, K.; Honjo, T.; Miyazaki, M.; Kakemi, M. Time-dependent changes in hepatic and intestinal
induction of cytochrome P450 3A after administration of dexamethasone to rats. Xenobiotica 2013, 43, 765–773.
[CrossRef]
38. Li, J.; Chen, R.; Yao, Q.-Y.; Liu, S.-J.; Tian, X.-Y.; Hao, C.-Y.; Lu, W.; Zhou, T.-Y. Time-dependent
pharmacokinetics of dexamethasone and its efficacy in human breast cancer xenograft mice:
A semi-mechanism-based pharmacokinetic/pharmacodynamic model. Acta Pharmacol. Sin. 2018, 39,
472–481. [CrossRef]
39. Sohn, C.A. Evaluation of ketoconazole. Clin. Pharm. 1982, 1, 217–224.
40. Jones, H.E. Ketoconazole. Arch. Dermatol. 1982, 118, 217–219. [CrossRef]
41. Rodriguez, R.J.; Acosta, D. Metabolism of ketoconazole and deacetylated ketoconazole by rat hepatic
microsomes and flavin-containing monooxygenases. Drug Metab. Dispos. 1997, 25, 772–777. [PubMed]
42. Kim, J.-H.; Choi, W.-G.; Lee, S.; Lee, H. Revisiting the metabolism and bioactivation of ketoconazole in
human and mouse using liquid chromatography-mass spectrometry-based metabolomics. Int. J. Mol. Sci.
2017, 18, 621. [CrossRef] [PubMed]
43. Zhou, S.; Chan, S.Y.; Goh, B.C.; Chan, E.; Duan, W.; Huang, M.; McLeod, H.L. Mechanism-based inhibition of
cytochrome P450 3A4 by therapeutic drugs. Clin. Pharmacok. 2005, 44, 279–304. [CrossRef] [PubMed]
44. Thummel, K.; Wilkinson, G. In vitro and in vivo drug interactions involving human CYP3A. Annu. Rev.
Pharmacol. Toxicol. 1998, 38, 389–430. [CrossRef] [PubMed]
45. Bourcier, K.; Hyland, R.; Kempshall, S.; Jones, R.; Maximilien, J.; Irvine, N.; Jones, B. Investigation into
UDP-glucuronosyltransferase (UGT) enzyme kinetics of imidazole-and triazole-containing antifungal drugs
in human liver microsomes and recombinant UGT enzymes. Drug Metab. Dispos. 2010, 38, 923–929.
[CrossRef] [PubMed]
46. Huang, Y.; Colaizzi, J.; Bierman, R.; Woestenborghs, R.; Heykants, J. Pharmacokinetics and dose
proportionality of ketoconazole in normal volunteers. Antimicrob. Agents Chemother. 1986, 30, 206–210.
[CrossRef]

23
Pharmaceutics 2019, 11, 479

47. Hukkanen, J. Induction of cytochrome P450 enzymes: A view on human in vivo findings. Expert Rev.
Clin. Pharmacol. 2012, 5, 569–585. [CrossRef]
48. Gentile, D.M.; Tomlinson, E.S.; Maggs, J.L.; Park, B.K.; Back, D.J. Dexamethasone metabolism by human
liver in vitro. Metabolite identification and inhibition of 6-hydroxylation. J. Pharmacol. Exp. Ther. 1996, 277,
105–112.
49. Kaeser, B.; Zandt, H.; Bour, F.; Zwanziger, E.; Schmitt, C.; Zhang, X. Drug-drug interaction study of
ketoconazole and ritonavir-boosted saquinavir. Antimicrob. Agents Chemother. 2009, 53, 609–614. [CrossRef]
50. Fujimori, S.; Hanada, R.; Hayashida, M.; Sakurai, T.; Ikushima, I.; Sakamoto, C. Celecoxib Monotherapy
Maintained Small Intestinal Mucosa Better Compared with Loxoprofen Plus Lansoprazole Treatment.
J. Clin. Gastroenterol. 2016, 50, 218–226. [CrossRef]
51. Mizukami, K.; Murakami, K.; Yamauchi, M.; Matsunari, O.; Ogawa, R.; Nakagawa, Y.; Okimoto, T.;
Kodama, M.; Fujioka, T. Evaluation of selective cyclooxygenase-2 inhibitor-induced small bowel injury:
Randomized cross-over study compared with loxoprofen in healthy subjects. Dig. Endosc. 2013, 25, 288–294.
[CrossRef] [PubMed]
52. Cascorbi, I. Drug interactions-principles, examples and clinical consequences. Dtsch. Arztebl. Int. 2012, 109,
546–556. [PubMed]
53. Li, X.-Q.; Andersson, T.B.; Ahlström, M.; Weidolf, L. Comparison of inhibitory effects of the proton
pump-inhibiting drugs omeprazole, esomeprazole, lansoprazole, pantoprazole, and rabeprazole on human
cytochrome P450 activities. Drug Metab. Dispos. 2004, 32, 821–827. [CrossRef] [PubMed]
54. Iwasaki, K. Metabolism of tacrolimus (FK506) and recent topics in clinical pharmacokinetics.
Drug Metab. Pharmacok. 2007, 22, 328–335. [CrossRef]
55. Milanov, I.; Georgiev, D. Mechanisms of tizanidine action on spasticity. Acta Neurol. Scand. 1994, 89, 274–279.
[CrossRef] [PubMed]
56. Li, X.; Jin, Y. Irreversible profound symptomatic bradycardia requiring pacemaker after tizanidine/loxoprofen
combination therapy: A case report. J. Int. Med.Res. 2018, 46, 2466–2469. [CrossRef] [PubMed]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

24
 pharmaceutics

Article
Evaluation of the Effect of CYP2D6 Genotypes on
Tramadol and O-Desmethyltramadol Pharmacokinetic
Profiles in a Korean Population Using
Physiologically-Based Pharmacokinetic Modeling
Hyeon-Cheol Jeong 1 , Soo Hyeon Bae 2 , Jung-Woo Bae 3 , Sooyeun Lee 3 , Anhye Kim 4 ,
Yoojeong Jang 1 and Kwang-Hee Shin 1, *
1 College of Pharmacy, Research Institute of Pharmaceutical Sciences, Kyungpook National University,
Daegu 41566, Korea; houkiboshi01@knu.ac.kr (H.-C.J.); kersy@daum.net (Y.J.)
2 Korea Institute of Radiological & Medical Sciences, Seoul 01812, Korea; shbae@kirams.re.kr
3 College of Pharmacy, Keimyung University, Daegu 42601, Korea; jwbae11@kmu.ac.kr (J.-W.B.);
sylee21@kmu.ac.kr (S.L.)
4 Department of Clinical Pharmacology and Therapeutics, CHA Bundang Medical Center, CHA University,
Seongnam 13496, Korea; ahkim1@cha.ac.kr
* Correspondence: kshin@knu.ac.kr; Tel.: +82-53-950-8582

Received: 6 October 2019; Accepted: 15 November 2019; Published: 17 November 2019

Abstract: Tramadol is a μ-opioid receptor agonist and a monoamine reuptake inhibitor.

O-desmethyltramadol (M1), the major active metabolite of tramadol, is produced by
CYP2D6. A physiologically-based pharmacokinetic model was developed to predict changes
in time-concentration profiles for tramadol and M1 according to dosage and CYP2D6 genotypes in
the Korean population. Parallel artificial membrane permeation assay was performed to determine
tramadol permeability, and the metabolic clearance of M1 was determined using human liver
microsomes. Clinical study data were used to develop the model. Other physicochemical and
pharmacokinetic parameters were obtained from the literature. Simulations for plasma concentrations
of tramadol and M1 (after 100 mg tramadol was administered five times at 12-h intervals) were
based on a total of 1000 virtual healthy Koreans using SimCYP® simulator. Geometric mean
ratios (90% confidence intervals) (predicted/observed) for maximum plasma concentration at
steady-state (Cmax,ss ) and area under the curve at steady-state (AUClast,ss ) were 0.79 (0.69–0.91)
and 1.04 (0.85–1.28) for tramadol, and 0.63 (0.51–0.79) and 0.67 (0.54–0.84) for M1, respectively.
The predicted time–concentration profiles of tramadol fitted well to observed profiles and those of M1
showed under-prediction. The developed model could be applied to predict concentration-dependent
toxicities according to CYP2D6 genotypes and also, CYP2D6-related drug interactions.

Keywords: CYP2D6; O-desmethyltramadol; pharmacokinetics; physiologically-based pharmacokinetics;

tramadol

1. Introduction
Tramadol is an orally available, centrally acting, weak-opioid analgesic drug [1]. The anti-nociceptive
effect of tramadol is due to two mechanisms: an opioid mechanism, and a non-opioid mechanism [2].
Tramadol acts as a μ-opioid receptor agonist, like traditional opioids. It also has analgesic effects by
inhibiting reuptake of monoamine neurotransmitters, such as 5-hydroxytryptamine (5-HT) and
noradrenaline [3,4]. These mechanisms lead to reduced pain signal conduction and analgesic
effects. Tramadol is a racemate, and analgesic mechanisms differ depending on the isomer:

25
Pharmaceutics 2019, 11, 618; doi:10.3390/pharmaceutics11110618 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2019, 11, 618

(−)-tramadol exhibited ≈ 10-fold higher noradrenaline reuptake inhibitory activity than (+)-tramadol,
and (+)-tramadol showed ≈ 4-fold higher 5-HT reuptake inhibitory activity than (−)-tramadol [5].
Tramadol is predominantly metabolized in the liver. Approximately 10%–30% of administered
tramadol is excreted, unmetabolized, in the urine. The well-known metabolic pathway of tramadol
is divided into three major pathways: O-desmethyltramadol (M1) is produced by cytochrome P450
(CYP) 2D6, and N-desmethyltramadol (M2) is produced by CYP2B6 and CYP3A4. These metabolites
are converted to either N,O-didesmethyltramadol (M5) and other inactive metabolites by CYP2D6,
CYP2B6, and CYP3A4, or converted to glucuronides by UDP-glucuronosyltransferase (UGT) 1A8
and UGT2B7 [6]. M1, a major active metabolite of tramadol, has about 700-fold higher affinity for
μ-opioid receptors than tramadol [7]. M5 also has 24-fold higher affinity for μ-opioid receptors than
tramadol [3]. Tramadol is mainly considered to inhibit monoamine reuptake, while M1 and M5 bind to
μ-opioid receptors and exhibit analgesic effects. Thus, genetic polymorphism of CYP2D6 could have
an effect on the risk of adverse events during tramadol administration [4].
A physiologically-based pharmacokinetic (PBPK) approach is a bottom-up approach that requires
data about the physicochemical properties and pharmacokinetic (PK) parameters (i.e., absorption,
distribution, metabolism, and excretion; ADME) of the target drug [8,9]. In addition, the PBPK
model considers bodyweight, height, organ volume, blood flow, and inter-individual variation for
metabolizing enzymes and transporters [10]. Therefore, the PBPK model can be used to predict plasma
concentration–time profiles more closely than traditional compartmental PK models [11]. In this regard,
PBPK modeling can predict human PK profiles using in vitro or preclinical study data from drug
development. Further, such modeling is also used to investigate the interaction potential with other
drugs or food, and to predict PK profiles in special populations, such as pregnant women, geriatric
patients, or children [9,12].
Studies predicting the PK profile of tramadol using PK modeling have been reported previously.
Many articles used a population PK approach with nonlinear mixed-effects modeling (NONMEM),
however a PBPK approach was rarely used to predict the PK profile of tramadol [13–15]. When tramadol
was administered, M1 also had impact on efficacy and toxicity [16]. Therefore, M1 should be integrated
for PBPK model of tramadol for better interpretation.
The aims of this study were to develop a PBPK model that could predict the concentration–time
profiles of tramadol and M1 in Koreans, and to investigate effects of the CYP2D6 genotype on PK
profiles at routinely administered doses. The developed PBPK model was applied to predict the effects
of CYP2D6 genotype and tramadol dosage on the plasma concentration profiles of tramadol and M1 in
a healthy Korean population.

2. Materials and Methods

2.1. Clinical Study Design

The clinical study was approved by the Institutional Review Board of Keimyung University (Deagu,
Republic of Korea, approval number: 40525-201509-BR-70-02, 23 February 2016) and Kyungpook
National University Hospital (Daegu, Republic of Korea, approval number of clinical trial: 2016-08-005,
24 August 2016) and carried out at the Kyungpook National University Hospital Clinical Trial Center
(Daegu, Republic of Korea). A total of 23 subjects participated who voluntarily agreed to take part
in the clinical study and signed their informed consent. Subject characteristics are presented in
Table 1. All subjects received a 100-mg tramadol hydrochloride tablet (Tridol® extended-release
(ER); Yuhan Pharmaceutical, Seoul, Korea) five times at 12-h intervals. Whole blood was collected
in an anticoagulant tube at pre-dose, and at 0.5, 1, 1.5, 2, 2.5, 4, 6, 8, 10, 12, 24, 48, and 72 h after
administration. The obtained whole blood was used for CYP2D6 genotyping, and the plasma was
separated for determination of tramadol and M1 [17].

26
Pharmaceutics 2019, 11, 618

Table 1. Demographic characteristics (n = 23).

Characteristic Mean (SD)

Age (years) 24.78 (4.80)
Height (cm) 176.51 (5.64)
Weight (kg) 71.61 (8.87)
CYP2D6 genotypes (no. of subjects)
Wild-type 14
*5/*5 1
*10/*10 8
SD, standard deviation.

2.2. CYP2D6 Genotyping

The determination of CYP2D6 genotype was performed for CYP2D6*2 (normal function),
CYP2D6*5 (no function), and CYP2D6*10 (decreased function). Genotyping for CYP2D6*2 and
CYP2D6*10 was performed using pyrosequencing. CYP2D6*5 was sequenced by long polymerase
chain reaction (PCR) because of deletion of a specific sequence. Pyrosequencing was performed using
Pyromark Q96 ID and Pyromark Gold Q96 reagents (Qiagen, Hilden, Germany). The conditions of PCR
(total 35 cycles) were: denaturation (94 ◦ C for 30 s), annealing (56 ◦ C for 30 s), and polymerization (72 ◦ C
for 30 s). The processes were finished by extension at 72 ◦ C for 5 min. CYP2D6*5 and duplication were
determined by long-PCR, as previously described [18,19]. The CYP2D6 phenotype was determined
based on genotype and activity score [20–23]. CYP2D6*5 was a non-functional allele and homozygous
CYP2D6*5 was classified as a poor metabolizer (PM). Homozygous CYP2D6*10 was classified as an
intermediate metabolizer (IM).

2.3. Determination of Tramadol and O-Desmethyltramadol (M1) Using LC-MS/MS

The obtained whole blood samples were immediately centrifuged at 4 ◦ C, 3000 rpm for 10 min.
The isolated plasma samples were transferred to a new microcentrifuge tube and kept at −70 ◦ C until
analysis. The plasma samples were completely thawed, then 10 μL of internal standard (tramadol 13 C-d3
for tramadol, and M1-d6 for M1) were added to 100 μL samples and mixed briefly. A total of 300 μL
of acetonitrile was added, and then samples were mixed thoroughly for 30 s prior to centrifugation
at 2500 rpm for 10 min. The organic solvent layer was transferred to a new polypropylene tube and
evaporated under nitrogen gas at 40 ◦ C. Methanol 200 μL was added into tubes containing pellets
for reconstitution, and 5 μL of reconstituted sample were analyzed. Analyses were carried out on
API3200 tandem mass spectrometry system (AB SCIEX, Framingham, MA, USA) equipped with an
Agilent 1260 series HPLC system (Agilent Technologies, Santa Clara, CA, USA). Separation of tramadol
and M1 was conducted using a Luna C18 column (5.0 μm, 2.0 × 50 mm; Phenomenex, Torrance,
CA, USA). Five millimoles ammonium formate and 0.1% formic acid in methanol (A), and 5 mM
ammonium formate solution (B), were used for the mobile phase. The used gradient method was as
follows: 0–2 min (97%–5% B), 2–4 min (5% B), 4–5 min (5%–97% B), and 5–8 min (97% B). Electrospray
ionization-positive ion mode was used for mass detection. The mass transitions (m/z) used were
264.2→58.1 for tramadol, 268.2→58.1 for tramadol internal standard (IS), 250.2→58.2 for M1, and
256.2→64.1 for M1 IS. To obtain pharmacokinetic parameters, non-compartmental analysis (NCA) was
performed using Phoenix (Certara Inc., Princeton, NJ, USA).

2.4. Parallel Artificial Membrane Permeability Assay (PAMPA)

To determine the permeability of tramadol, a parallel artificial membrane permeability assay
(PAMPA) was performed [24]. Gentest™ Pre-coated PAMPA Plate System (Corning, Tewksbury,
MA, USA) was used for the permeability assay. All the processes were followed according to the
manufacturer’s protocol. Tramadol hydrochloride, dimethyl sulfoxide (DMSO), phosphate-buffered
saline (PBS), and acetonitrile were purchased from Sigma-Aldrich (St. Louis, MO, USA). Tramadol

27
Pharmaceutics 2019, 11, 618

stock solution (1 mM) was prepared using 100% DMSO and diluted to 15 μM using PBS (pH 7.4). The
PAMPA plate was equilibrated for 30 min at room temperature before performing the permeability
assay. PBS 200 μL was dispensed on the acceptor side and 300 μL of working solution was dispensed
on the donor side. Incubation was carried out at room temperature for 5 h, and the acceptor and donor
side buffers were analyzed using liquid chromatography-tandem mass spectrometry (LC-MS/MS).
A total of 12 replicated samples were assayed and mean permeability was calculated and applied to
the model.

2.5. Assessment of Intrinsic Clearance of M1

For the experiments, O-desmethyltramadol HCl (M1), glucose 6-phosphate, glucose 6-phosphate
dehydrogenase, MgCl2 , β-nicotinamide adenine dinucleotide phosphate (NADP), chlorpropamide,
Trizma® base, Trizma® hydrochloride, DMSO and formic acid were obtained from Sigma-Aldrich
(St. Louis, MO, USA). To evaluate the intrinsic clearance of M1 by CYPs, metabolic stability studies
under NADPH system were conducted in human liver microsoms (HLM) [25,26]. For details,
NADPH-generating system (1.3 mM NADP+ , 3.3 mM glucose 6-phosphate, 3.3 mM MgCl2 , and
0.4 unit/mL glucose-6-phosphate dehydrogenase) and HLM 0.25 mg/mL were added and preincubated
at 37 ◦ C for 5 min. Then, 20 μM M1 was added and reacted at 37 ◦ C for 0, 1, 5, 10, 20, 30, and 40 min,
respectively. The total volume of the reaction mixture was 100 μL. After each reaction, the reaction
was terminated by addition of 150 μL of acetonitrile containing an internal standard (100 ng/mL
chlorpropamide). All experiments were repeated duplicated and the samples were vortexed for 5 min
and centrifuged (13,000 rpm, 4 ◦ C) for 15 min. Then, supernatants were injected into LC-MS/MS and
M1 were analyzed. The concentration of M1 (20 μM) at 0 min was used to evaluate the metabolism of
CYP through the change of drug concentration over time.

2.6. Qunatitaion Methods of M1 in In Vitro Experiments Using LC-QTOF

High-performance liquid chromatography (HPLC)-grade acetonitrile and deionized water were
obtained from Berdick and Jackson (Muskegon, MI, USA). HLM (50 donors pooled) were purchased
from Corning. Analyses were carried out on Sciex QTOF 5600 plus (AB SCIEX, Framingham, MA,
USA) equipped with an Agilent 1260 series HPLC system (Agilent Technologies, Santa Clara, CA,
USA). For quantitation of M1, the compounds were separated on a Poroshell 120 (4.6 × 50 mm, 2.7 μm;
Agilent Technologies) with an isocratic mobile phase consisting of acetonitrile and 0.1% aqueous formic
acid (70:30 v/v) at a flow rate of 0.5 mL/min. The overall run time was 5 min per sample. The column
and autosampler temperatures were maintained at 40 ◦ C and 4 ◦ C, respectively. Time-of-flight mass
spectrometry analysis was selected in positive ion mode for the sample analysis. The quantitative
analytical data were processed using PeakView® (Version 2.2.0; AB SCIEX, Framingham, MA, USA)
and MultiQuant® (Version 3.0.2; Framingham, MA, USA), and the formulas C15 H23 NO2 (M1), and
C10 H13 ClN2 O3 S (chlorpropamide) were used for quantitation.

2.7. Development of PBPK Model for Tramadol and M1

PBPK model development was performed using SimCYP® simulator version 18 (Certara, Sheffield,
UK). Most of the parameters for tramadol and M1 were entered with reference to the literature.
According to previous reports, tramadol was not substrate for P-glycoprotein (P-gp) (ABCB1) and the
role of proton-based pumps such as OATP for tramadol permeability was unclear [27,28]. Therefore,
in this study, the permeability of tramadol was determined by PAMPA assay. The advanced drug
absorption and metabolism model was used to consider the ER formulation, and the dissolution profiles
of Tridol® ER 100 mg (Yuhan Pharmaceutical, Seoul, Korea) were applied. The elimination profile for
M1 applied the intrinsic clearance by HLM. Kp scalar of tramadol was set to match the observed Vss
and the predicted Vss value in the model, and Kp scalar of M1 was obtained from parameter estimation.
In clinical study, the Vss of tramadol was calculated to 2.6 L/kg by non-compartmental analysis. That
of M1 was not calculated because the exact dose of M1 is unknown. Intrinsic clearance involved in

28
Pharmaceutics 2019, 11, 618

tramadol metabolism was estimated using retrograde model option, and human liver microsomal
intrinsic clearance was applied to M1 library. The renal clearance of tramadol and M1 were applied for
the predicted value which is the closest to the observed blood concentration–time profile by parameter
estimation. The PBPK model was evaluated so that it could effectively predict PK profiles for tramadol
and M1 when observed mean plasma concentrations fitted to the predicted plasma concentration–time
profile and its 90% confidence interval (CI). The other evaluation criteria were geometric mean ratio
for peak plasma concentration at steady-state (Cmax,ss ), and area under the plasma concentration–time
curve at last observation at steady-state (AUClast,ss ), and the 90% CI for these values. If the geometric
mean ratio and its 90% CI were within the range 0.7–1.43, the model was considered to fit well.

2.8. Prediction of Changes in Concentration–Time Profiles for Tramadol and M1 According to CYP2D6
Genotype and Dosing Regimen
The therapeutic range and toxic range of tramadol and M1 were determined by reference to the
literature. Because the manufacturer’s recommended acceptable maximum dose of Tridol® ER was
400 mg per day, the tramadol ER tablet was administered to a virtual healthy Korean population
at 100 and 200 mg (5 times at 12-h intervals) to simulate the change of concentration–time profiles
for tramadol and M1. This simulation assumes linear PK properties for multiple doses of tramadol
100 mg and 200 mg [29]. The effect of CYP2D6 genotype was also simulated for tramadol 100 mg and
200 mg for populations consisting of CYP2D6 poor metabolizers (PM), intermediate metabolizers (IM),
extensive metabolizers (EM), and ultra-rapid metabolizers (UM).

3. Results

3.1. Clinical Study

A total of 23 subjects, aged 20–34 years, were enrolled in the clinical study. The average plasma
concentration–time profiles for tramadol and M1 are shown in Figure 1. The allele (phenotype)
frequencies for CYP2D6 were: Wild-type (EM, 14 subjects), *10/*10 (IM, 8 subjects), and *5/*5 (PM,
1 subject).

Figure 1. Cont.

29
Pharmaceutics 2019, 11, 618

Figure 1. The average plasma concentration–time profiles after five times oral administration (τ = 12 h)
of 100 mg tramadol for (a) tramadol and (b) O-desmethyltramadol (M1). Solid blue line, average for all
subjects (n = 23); solid black line, CYP2D6 wild-type subjects (n = 14); short dashed black line, CYP2D6
*5/*5 subject (n = 1); long dashed black line, CYP2D6 *10/*10 (n = 8).

3.2. Metabolism Assay for O-Desmethyltramadol (M1)

We conducted metabolic stability study of M1 and confirmed that M1 was metabolized mainly by
CYPs and partially by UGTs. In control sample without NADPH or HLM, more than 96% of tramadol
and M1 remained during the incubation time, indicating that the disappearance of tramadol and M1
were mainly caused by CYP enzymes. Metabolic stability of M1 by CYP was assessed using results of
the disappearance test for M1. The slope of linear regression was calculated and intrinsic clearance
(CLint ) of the drug in the in vitro microsome system was calculated:

CLint,mic (μL/min/mg protein) = k × Vincubation /Cincubation (1)

where Vincubation is incubation volume, Cincubation is concentration of microsomal protein.

To apply the elimination profile to O-desmethyltramadol (M1), an HLM assay was performed.
The M1 disappearance test showed an HLM intrinsic clearance (CLint,HLM ) of 52.95 μL/min/mg protein
(Figure 2). CLint,HLM for M1 was applied to the M1 PBPK model.

30
Pharmaceutics 2019, 11, 618

Figure 2. The plot of remaining rate of O-desmethyltramadol (M1) after incubation with human
liver microsoms (HLM). Each point (obtained by duplicate measurements) represents the mean value.
The intrinsic clearance by HLM (CLint,mic ) was calculated as 52.92 μL/min/mg protein.

3.3. PAMPA Results

Results of the PAMPA assay for 12 tramadol samples diluted to 15 μM showed permeability
ranges from 9.14 × 10−6 cm/s to 11.5 × 10−6 cm/s. The mean ± standard deviation permeability was
calculated as 10.4 × 10−6 ± 0.056 cm/s. The calculated mean PAMPA permeability was applied to the
tramadol PBPK model.

3.4. Development of the PBPK Model for Tramadol and M1

The input parameters for tramadol and M1 and demographic characteristics for virtual population
in the PBPK model are presented in Tables 2 and 3. Data for the healthy Korean population were
obtained from the Certara repository. Ten virtual trials, including 100 virtual subjects in each virtual
trial (total 1000 subjects), were performed for tramadol and M1. In the tramadol model, most of
the observation profiles were within the 5th and 95th percentile range, and the predicted mean
tramadol concentration in plasma was similar to the observed profile (Figure 3a). The geometric
mean ratios of Cmax,ss and AUClast,ss for tramadol were 0.79 and 1.04, respectively (Table 4). Most
observed concentration–time profiles were included in the 5th and 95th percentiles of the predicted
concentration–time profiles. In addition, the mean predicted plasma tramadol concentration was well
fitted to the observed tramadol concentration. In the concentration–time profiles for M1, most of the
observations were also within the 5th and 95th percentile range of the predicted profile (Figure 3b). The
range of 90% CI for Cmax,ss and AUClast,ss were included in the range of 0.7–1.43 (30% range of geometric
mean ratio); however, the geometric mean ratio was predicted to be relatively low (Table 4). Both
tramadol and M1 were predominantly distributed in the liver. The second most common distribution
sites were the spleen (tramadol) and heart (M1; Table 5).

31
Pharmaceutics 2019, 11, 618

Table 2. Input parameters for tramadol and O-desmethyltramadol (M1) in the physiologically-based
pharmacokinetic (PBPK) model.

Tramadol M1 *
Parameters
Value Source Value Source
Physicochemical properties and blood binding
Molecular weight
263.4 [30] 249.354 [31]
(g/mol)
Log P 1.35 [30] 2.26 [32]
9.41 (Monoprotic 9.62 (Monoprotic
pKa [30] [32]
base) base)
Predicted in
fup 0.8 [33] 0.525
SimCYP
Absorption
Absorption type PAMPA - n/a -
Papp (×10−6 cm/s) 10.2 Experimental data n/a -
Distribution
Kp scalar 0.946 Adjusted using Vss 0.107 Estimated
Vss (L/kg) 2.6 Observed data 0.628 Estimated
Elimination
CLint CYP2D6: 0.447;
52.92
(μL/min/pmol or CYP2B6: 0.028; Retrograde model Experimental data
(WOMC–HLM)
mg protein) CYP3A4: 0.020
CLR 1.850 Estimated 0.481 Estimated
CLint : intrinsic clearance; CLR : renal clearance; CYP: cytochrome P450 superfamily; fup : unbound fraction in plasma;
HLM: human liver microsomes; Kp: plasma-tissue partition coefficient; PAMPA: parallel artificial membrane
permeability assay; Papp : apparent permeability; Vss : volume of distribution in steady-state; WOMC: whole organ
metabolic clearance, n/a: not applicable. * Metabolite model does not take account of absorption.

Table 3. The demographic characteristics of the participated subjects for virtual Korean population
(n = 1000).

Parameters Mean (Range)

Age (years) 28.9 (20.2–40.0)
Height (cm) 166.2 (147.6–188.0)
Weight (kg) 62.3 (42.9–93.9)
The percentage of female 50%

32
Pharmaceutics 2019, 11, 618

Figure 3. The observed (each symbol, n = 23) and simulated mean (solid dark green line) plasma
concentration–time profiles after administration of 100 mg tramadol extended-release (ER) tablet twice
daily (total five times) for (a) tramadol, and (b) O-desmethyltramadol (M1); blue dashed line represents
5th and 95th percentiles.

Table 4. Observed and simulated pharmacokinetic (PK) parameters for tramadol and
O-desmethyltramadol (M1) after oral administration of 100 mg tramadol ER tablet twice daily (five
times in total).

Parameters Observed (Range) Simulated (Range) Ratio (90% CI)

Tramadol
Geometric mean Cmax,ss (ng/mL) 643.8; (294.0–942.1) 508.4; (122.1–1226) 0.79; (0.69–0.91)
Geometric mean AUClast,ss
8965; (4127–16,038) 9346; (1217–42,462) 1.04; (0.85–1.28)
(ng/mL·h)
M1
Geometric mean Cmax,ss (ng/mL) 103.8; (29.8–176.7) 65.68; (1.07–368.1) 0.63; (0.51–0.79)
Geometric mean AUClast,ss
1775; (445.3–2875) 1.187; (6.236–7522) 0.67; (0.54–0.84)
(ng/mL·h)
AUClast : area under the curve from 48 h to 120 h at steady-state; CI: confidence interval; Cmax,ss : maximum drug
concentration in plasma at steady-state.

33
Pharmaceutics 2019, 11, 618

Table 5. Maximum simulated concentrations at steady-state for tramadol and O-desmethyltramadol

(M1) in each organ (Cmax,ss ) after oral administration of 100 mg tramadol ER tablet twice daily (five
times in total).

Maximum Concentration at Steady-State in Each Organ (ng/mL)

Organ
Tramadol M1
Adipose tissue 400.6 12.89
Bone 899.8 23.88
Brain 1034 19.74
Gut 2741 88.38
Heart 763.0 94.97
Kidney 1452 78.46
Liver 3034 157.5
Lung 991.0 21.09
Muscle 2411 79.13
Pancreas 2000 56.25
Skin 1355 42.03
Spleen 2821 88.53

3.5. Prediction of Changes in Concentration–Time Profiles for Tramadol and M1 According to CYP2D6
Genotype and Dosage
To investigate the effect of CYP2D6 genotype and dosage on PK profiles, simulations were
performed for the administration of 100 and 200 mg of tramadol every 12 h (total 5 times). The
tramadol/M1 concentration–time profiles were captured from the pre-dose (0 h) to 120 h. The differences
on PK profiles according to CYP2D6 genotypes were assessed in the general Korean population in
CYP2D6 groups: PM, IM, EM, and UM. As a result, plasma concentration–time profiles for tramadol
were within the therapeutic range in all groups after administration of 100 mg tramadol ER. Predicted
plasma M1 concentrations were very low in the PM group (mean Cmax,ss 0.643 ng/mL) compared to the
CYP2D6 IM, EM, and UM groups (mean values 40.93, 83.80, and 126.8 ng/mL, respectively).
The plasma concentration–time profiles for tramadol and M1, and changes in PK parameters, in
the various CYP2D6 genotype groups following oral administration of 100 and 200 mg tramadol ER
tablet twice daily (total five times) are shown in Figure 4 and Table 6 (the plasma concentration–time
profiles for each CYP2D6 phenotype after administration of 100 and 200 mg of tramadol were presented
in Supplementary Materials Figures S1 and S2). Following tramadol 100 mg administrations, the
Cmax,ss of tramadol in CYP2D6 PMs reached to toxic range. For CYP2D6 UMs, the Cmax,ss of M1
exceeded the therapeutic margin (Supplementary Materials Figure S1). Following tramadol 200 mg
administrations, the Cmax,ss of tramadol were reached to the toxic range in all CYP2D6 metabolizer
groups. For M1, the Cmax,ss exceeded the therapeutic margin in the CYP2D6 IMs, EMs, and UMs
(Figure S2). In Table 7, observed and predicted Cmax,ss and AUClast,ss values, and predicted/observed
geometric mean ratios are presented. The CYP2D6 UM group was excluded from this table because UM
subjects did not exist in the clinical study. In the CYP2D6 EM and IM groups, the predicted/observed
geometric mean ratios for Cmax,ss and AUClast,ss for tramadol satisfied the acceptance criteria (0.7–1.43);
however, the tramadol AUClast,ss ratio for the CYP2D6 PM group was overestimated at 1.95. The
prediction results for M1 showed that AUClast,ss satisfied the acceptance criteria in the CYP2D6 EM
group; however, Cmax,ss and AUClast,ss values were underestimated in both the CYP2D6 IM and PM
groups, where the predicted values were much lower than observed values.

34
 Table 6. Predicted geometric mean Cmax,ss and AUClast,ss values for tramadol and O-desmethyltramadol (M1) following oral administration of 100 and 200 mg
tramadol ER tablet twice daily (total five times) in various CYP2D6 metabolizer groups.

UM EM IM PM
Parameters
Tramadol M1 Tramadol M1 Tramadol M1 Tramadol M1
Tramadol 100 mg
Cmax,ss (ng/mL) 357.2 126.8 469.6 83.80 593.8 40.93 721.3 0.6433
Pharmaceutics 2019, 11, 618

(range) (72.43–927.6) (21.39–449.0) (122.2–1117) (12.62–368.1) (165.8–1379) (5.511–240.9) (209.5–1675) (0.0975–5.312)

AUClast,ss (ng/mL·h) 5353 1881 8206 1445 12,049 813.5 16,795 2.919
(range) (648.7–25,267) (278.6–8560) (1217–34,213) (175.2–7522) (1932–42,462) (79.54–5234) (2682–61,319) (0.3351–34.42)
Tramadol 200 mg
Cmax,ss (ng/mL) 714.3 253.5 # 939.1 * 167.6 1188 ** 81.86 1443 ** 1.287
(range) (144.9–1855) (42.77–898.0) (244.3–2235) (25.24–736.3) (331.7–2758) (11.02–481.7) (418.9–3349) (0.1950–10.62)
AUClast,ss (ng/mL·h) 10,706 3761 16,411 2890 24,097 1627 33,591 5.839
(range) (1297–50,533) (557.1–17,119) (2434–68,426) (350.3–15,044) (3864–84,923) (159.1–10,467) (5365–122,637) (0.67–68.83)
AUClast,ss : area under the curve from 48 h to 120 h at steady-state; Cmax,ss : maximum drug concentration in plasma at steady-state; EM: extensive metabolizer; IM: intermediate
metabolizer; PM: poor metabolizer; UM: ultra-rapid metabolizer. * Above the therapeutic range for tramadol (>800 ng/mL); ** in toxic range for tramadol (1000–2000 ng/mL); # above
maximum therapeutic range for M1 (>200 ng/mL).

35
Table 7. Predicted and observed geometric mean PK parameters for tramadol and M1 according to CYP2D6 genotype following oral administration of 100 mg
tramadol ER tablet twice daily (total five times).

EM IM PM
Tramadol
Observed Predicted Ratio Observed Predicted Ratio Observed Predicted Ratio
(n = 13) (n = 1000) (90% CI) (n = 8) (n = 1000) (90% CI) (n = 1) (n = 1000) (90% CI)
Cmax,ss (ng/mL) 551.2 469.6 0.85 828.5 593.8 0.72 721.3
751.10 0.96
(range) (294.0–904.4) (122.2–1117) (0.72–1.01) (676.6–942.1) (165.8–1379) (0.59–0.87) (209.5–1675)
AUClast,ss (ng/mL·h) 7116 8206 1.15 13,501 12,049 0.89 16,795
8591.72 1.95
(range) (4127–9345) (1217–34,213) (0.90–1.48) (10,527–16,038) (1932–42,462) (0.66–1.20) (2682–61,319)
EM IM PM
M1
Observed Predicted Ratio Observed Predicted Ratio Observed Predicted Ratio
(n = 13) (n = 1000) (90% CI) (n = 8) (n = 1000) (90% CI) (n = 1) (n = 1000) (90% CI)
Cmax,ss (ng/mL) 125.0 83.80 0.67 87.79 40.93 0.47 0.6433
29.8 0.02
(range) (81.8–176.7) (12.62–368.1) (0.52–0.86) (66.0–114.1) (5.511–240.9) (0.32–0.69) (0.0975–5.312)
AUClast,ss (ng/mL·h) 1996 1445 0.72 1718 813.5 0.47 2.919
445.3 0.01
(range) (1373–2875) (175.2–7522) (0.56–0.94) (1223–2199) (79.54–5234) (0.31–0.72) (0.3351–34.42)
AUClast,ss : area under the curve from 48 h to 120 h at steady-state; CI: confidence interval; Cmax,ss : maximum drug concentration in plasma at steady-state; EM: extensive metabolizer; IM:
intermediate metabolizer; PM: poor metabolizer; Ratio = predicted/observed. Since the observed data for the PM group are for 1 subject, the CI value cannot be obtained.
Pharmaceutics 2019, 11, 618

Figure 4. The predicted mean concentration–time profiles after administration of 100 mg and 200 mg
tramadol ER tablet twice daily (total five times) for tramadol (a and c), and O-desmethyltramadol
(b and d), respectively. Gray areas in (a) and (c) represent the therapeutic concentration range
(100–800 ng/mL); checked gray area in (c) represents the toxic range (above 1000 ng/mL) for tramadol;
and the gray area in (b) and (d) represents the maximum therapeutic range for M1 (up to 200 ng/mL).

4. Discussion
PBPK models for tramadol and M1 were developed. Tramadol plasma concentration–time profiles
were well predicted from the proposed model. Prediction results for M1 included values in the
5th to 95th percentiles of most observed plasma concentration–time values, and the predicted mean
plasma concentration was also similar to the observed concentration–time profile. However, geometric
mean Cmax,ss and AUClast,ss ratios were under-predicted (0.63 and 0.67 for Cmax,ss and AUClast,ss ,
respectively). To predict concentration-dependent toxicities, the therapeutic range (100–800 ng/mL
for tramadol, and up to 200 ng/mL for M1) and the tramadol toxic range and lethal concentration
(>1000 ng/mL, and >2000 ng/mL, respectively) were obtained from the literatures [34,35]. In general,
the recommended dose of tramadol is up to 400 mg per day for immediate-release formulations and
300 mg per day for ER formulations [2]. Simulations were performed for 100 and 200 mg with 12-h
intervals (5 times) according to CYP2D6 genotypes. After administration of 100 mg of tramadol, the
predicted Cmax,ss of tramadol reached to toxic range in CYP2D6 PMs and exceeded therapeutic range
in some IMs, and the predicted Cmax,ss of M1 exceeded therapeutic margin in CYP2D6 UMs. After
tramadol 200 mg administrations, the predicted tramadol Cmax,ss reached to toxic ranges in all CYP2D6
metabolizer groups, even in some EMs and the predicted M1 Cmax,ss exceeded the therapeutic margin
in CYP2D6 IMs, EMs, and UMs. The concentrations exceeded the therapeutic margins or reached to
the toxic range might be related to potential toxicities after tramadol administrations, even though
recommended doses of tramadol were administered.
PBPK modeling is useful for predicting PK profiles for rare genotypes in the population.
The frequency of CYP2D6 UM in the Korean population has been reported as approximately 1.25% [36].
In the clinical study used for our PBPK model development, there was only one PM subject, and no UM
subject was found. The model developed in this study could predict the plasma concentration–time

36
Pharmaceutics 2019, 11, 618

profiles of tramadol and M1 for these two groups. Using the developed model, plasma tramadol/M1
concentration–time profiles for CYP2D6 UM, a very rare genotype in Koreans, were also predicted.
Tramadol inhibits reuptake of 5-HT and norepinephrine. M1 binds to μ-opioid receptors and
exhibits analgesic effects. Due to these actions, the side effects of tramadol differ depending on
CYP2D6 genotype. In the PM group, a high risk of side effects due to tramadol, such as serotonin
syndrome, can be expected; and in the UM group, a high risk of μ-opioid receptor-related side effects,
such as respiratory depression, can be expected relative to other CYP2D6 genotypes [16]. In our
simulation, the plasma concentrations of tramadol and M1 exceeded the therapeutic concentration
range, even after administration of recommended doses. These results suggest that the frequency
of concentration-related adverse drug reactions may be reduced by optimizing the dosing regimen
according to CYP2D6 genotype of the patient or population.
Tramadol and M1 distribution in each tissue were estimated using the PBPK model, and tramadol
and M1 were distributed most to the liver. In cases of fatal intoxication due to tramadol, the highest
concentration of tramadol was evident in the liver, after the blood and urine. These distribution
characteristics are considered due to the hepatic metabolism of tramadol and its metabolites [37].
The distribution of tramadol to adipose tissue differed from that for M1. Indeed, tramadol is considered
to distribute widely to lipid-rich tissues because of its higher affinity for lipids than M1 (logD for
tramadol and M1: 1.13 and 0.4, respectively) [38]. Further research is needed about the distribution
characteristics of tramadol and M1 to each tissue.
The predicted plasma M1 concentration–time profiles were under-predicted due to a lack of
information about distribution and elimination properties. Since M1 is produced by tramadol
metabolism, elimination profiles (intrinsic clearance by CYP, renal clearance and additional clearance)
of tramadol were adjusted to improve the M1 model; however, there were no significant changes in M1
concentration–time profiles. This might be due to poor distribution of M1 from liver to plasma, or to
exaggeration of elimination. For improvement, the M1 model was built using parameter estimation by
observed plasma concentration–time profiles as distribution and elimination profiles (tissue–plasma
partition coefficient, additional clearance, renal clearance, bile clearance). When estimating several
parameters, the predicted plasma M1 concentration–time profiles changed significantly when values
of the unbound fraction in incubated microsomes (fumic ) and active hepatic scalar were changed. Thus,
the plasma M1 concentration–time profile might be greatly influenced by metabolism. More detailed
information and parameters for M1 metabolism are needed for more accurate predictions of plasma
M1 concentration–time profiles.
Regarding limitations of our study, tramadol is metabolized not only to M1, but also to
N-desmethyltramadol (M2) by CYP2D6, CYP2B6, and CYP3A4. In accordance with the literature,
the toxicity of tramadol and M1 can be determined using M1/M2 ratio [34]. Therefore, an M2 model
could improve the predictability of concentration-related adverse drug reactions after tramadol
administration. Moreover, organic cation transporter 1 (OCT1) and multidrug resistance protein 1
(MDR1) influence the disposition of tramadol and M1. Significant differences in drug disposition
according to OCT1 and MDR1 genotypes have been shown, even in same CYP2D6 phenotype [39–41].
Due to lack of information of transporter kinetic parameter for each organ, the transport kinetic
parameters for M1 were excluded for the model. For elaborate model prediction, OCT1 and MDR1
genotypes (OCT*1, *2, *3, *4, *5, and MDR1 C3435T) could be incorporated.

5. Conclusions
In summary, our PBPK model for tramadol and M1 was developed and predicted
concentration–time profiles after multiple administrations of a tramadol ER formulation in the
Korean population. Differences in PK profiles and concentration-dependent toxicities were predicted
according to CYP2D6 phenotype and dosage. Most modeling studies of tramadol used a population
PK approach, and the literature using PBPK modeling focused on the PK profile of tramadol itself.
However, this study developed a model with predictive power for tramadol and M1, the major active

37
Pharmaceutics 2019, 11, 618

metabolite. This model could be applied to predict concentration-dependent toxicity profiles in cases
of tramadol overdose or abuse and also, CYP2D6-related drug interactions.

Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4923/11/11/618/s1,

Figure S1: The predicted mean tramadol and O-desmethyltramadol concentration-time profiles after administration
of 100 mg tramadol ER tablet twice daily (five times in total) for CYP2D6 poor metabolizer, intermediate metabolizer,
extensive metabolizer and ultra-rapid metabolizer, respectively; Figure S2: The predicted mean tramadol and
O-desmethyltramadol concentration-time profiles after administration of 200 mg tramadol ER tablet twice daily
(five times in total) for CYP2D6 poor metabolizer, intermediate metabolizer, extensive metabolizer and ultra-rapid
metabolizer, respectively.
Author Contributions: J.-W.B., S.L. and K.-H.S. conceived and designed the experiments; H.-C.J., S.H.B. and
K.-H.S. performed the experiments, simulations, and analyzed the data; H.-C.J., Y.J., S.H.B., A.K. and K.-H.S.
reviewed the data and wrote the paper. All authors approved the final manuscript.
Funding: This research was supported by the Bio and Medical Technology Development Program of the
National Research Foundation (NRF) and was funded by the Korean government (MSIP and MOHW;
No. NRF-2015M3A9E1028327) and the Korea Health Technology R&D Project through the Korea Health
Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare, Republic of Korea
(grant number: HI17C0927).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Shipton, E. Tramadol—Present and future. Anaesth. Intensive Care 2000, 28, 363–374. [CrossRef] [PubMed]
2. Miotto, K.; Cho, A.K.; Khalil, M.A.; Blanco, K.; Sasaki, J.D.; Rawson, R. Trends in Tramadol: Pharmacology,
Metabolism, and Misuse. Anesth. Analg. 2017, 124, 44–51. [CrossRef] [PubMed]
3. Lassen, D.; Damkier, P.; Brøsen, K. The Pharmacogenetics of Tramadol. Clin. Pharmacokinet. 2015, 54, 825–836.
[CrossRef] [PubMed]
4. Kaye, A.D. Tramadol, pharmacology, side effects, and serotonin syndrome: A review. Pain Physician 2015,
18, 395–400.
5. Leppert, W. Tramadol as an analgesic for mild to moderate cancer pain. Pharmacol. Rep. 2009, 61, 978–992.
[CrossRef]
6. Lehtonen, P.; Sten, T.; Aitio, O.; Kurkela, M.; Vuorensola, K.; Finel, M.; Kostiainen, R. Glucuronidation of
racemic O-desmethyltramadol, the active metabolite of tramadol. Eur. J. Pharm. Sci. 2010, 41, 523–530.
[CrossRef]
7. Grond, S.; Sablotzki, A. Clinical pharmacology of tramadol. Clin. Pharmacokinet. 2004, 43, 879–923. [CrossRef]
8. Kostewicz, E.S.; Aarons, L.; Bergstrand, M.; Bolger, M.B.; Galetin, A.; Hatley, O.; Jamei, M.; Lloyd, R.;
Pepin, X.; Rostami-Hodjegan, A. PBPK models for the prediction of in vivo performance of oral dosage
forms. Eur. J. Pharm. Sci. 2014, 57, 300–321. [CrossRef]
9. Zhao, P.; Zhang, L.; Grillo, J.; Liu, Q.; Bullock, J.; Moon, Y.; Song, P.; Brar, S.; Madabushi, R.; Wu, T.
Applications of physiologically based pharmacokinetic (PBPK) modeling and simulation during regulatory
review. Clin. Pharmacol. Ther. 2011, 89, 259–267. [CrossRef]
10. Abbiati, R.A.; Manca, D. A modeling tool for the personalization of pharmacokinetic predictions.
Comput. Chem. Eng. 2016, 91, 28–37. [CrossRef]
11. Price, P.S.; Conolly, R.B.; Chaisson, C.F.; Gross, E.A.; Young, J.S.; Mathis, E.T.; Tedder, D.R. Modeling
interindividual variation in physiological factors used in PBPK models of humans. Crit. Rev. Toxicol. 2003,
33, 469–503. [CrossRef] [PubMed]
12. Marsousi, N.; Desmeules, J.A.; Rudaz, S.; Daali, Y. Usefulness of PBPK Modeling in Incorporation of Clinical
Conditions in Personalized Medicine. J. Pharm. Sci. 2017, 106, 2380–2391. [CrossRef] [PubMed]
13. T’Jollyn, H.; Snoeys, J.; Colin, P.; Van Bocxlaer, J.; Annaert, P.; Cuyckens, F.; Vermeulen, A.; Van Peer, A.;
Allegaert, K.; Mannens, G.; et al. Physiology-Based IVIVE Predictions of Tramadol from in Vitro Metabolism
Data. Pharm. Res. 2015, 32, 260–274. [CrossRef] [PubMed]
14. Salman, S.; Sy, S.K.; Ilett, K.F.; Page-Sharp, M.; Paech, M.J. Population pharmacokinetic modeling of tramadol
and its O-desmethyl metabolite in plasma and breast milk. Eur. J. Clin. Pharmacol. 2011, 67, 899–908.
[CrossRef]

38
Pharmaceutics 2019, 11, 618

15. Garrido, M.J.; Habre, W.; Rombout, F.; Trocóniz, I.F. Population Pharmacokinetic/Pharmacodynamic
Modelling of the Analgesic Effects of Tramadol in Pediatrics. Pharm. Res. 2006, 23, 2014–2023. [CrossRef]
16. Faria, J.; Barbosa, J.; Moreira, R.; Queirós, O.; Carvalho, F.; Dinis-Oliveira, R. Comparative pharmacology
and toxicology of tramadol and tapentadol. Eur. J. Pain 2018, 22, 827–844. [CrossRef]
17. Lee, J.; Yoo, H.D.; Bae, J.W.; Lee, S.; Shin, K.H. Population pharmacokinetic analysis of tramadol and
O-desmethyltramadol with genetic polymorphism of CYP2D6. Drug Des. Dev. Ther. 2019, 13, 1751.
[CrossRef]
18. Yu, H.; Hong, S.; Jeong, C.H.; Bae, J.W.; Lee, S. Development of a linear dual column HPLC–MS/MS method
and clinical genetic evaluation for tramadol and its phase I and II metabolites in oral fluid. Arch. Pharmacal Res.
2018, 41, 288–298. [CrossRef]
19. Byeon, J.Y.; Kim, Y.H.; Lee, C.M.; Kim, S.H.; Chae, W.K.; Jung, E.H.; Choi, C.I.; Jang, C.G.; Lee, S.Y.; Bae, J.W.
CYP2D6 allele frequencies in Korean population, comparison with East Asian, Caucasian and African
populations, and the comparison of metabolic activity of CYP2D6 genotypes. Arch. Pharmacal Res. 2018,
41, 921–930. [CrossRef]
20. Byeon, J.Y.; Kim, Y.H.; Na, H.S.; Jang, J.H.; Kim, S.H.; Lee, Y.J.; Bae, J.W.; Kim, I.S.; Jang, C.G.;
Chung, M.W.; et al. Effects of the CYP2D6* 10 allele on the pharmacokinetics of atomoxetine and its
metabolites. Arch. Pharmacal Res. 2015, 38, 2083–2091. [CrossRef]
21. Doki, K.; Homma, M.; Kuga, K.; Kusano, K.; Watanabe, S.; Yamaguchi, I.; Kohda, Y. Effect of CYP2D6
genotype on flecainide pharmacokinetics in Japanese patients with supraventricular tachyarrhythmia. Eur. J.
Clin. Pharmacol. 2006, 62, 919–926. [CrossRef] [PubMed]
22. Findling, R.L.; Nucci, G.; Piergies, A.A.; Gomeni, R.; Bartolic, E.I.; Fong, R.; Carpenter, D.J.; Leeder, J.S.;
Gaedigk, A.; Danoff, T.M. Multiple dose pharmacokinetics of paroxetine in children and adolescents with
major depressive disorder or obsessive–compulsive disorder. Neuropsychopharmacology 2006, 31, 1274.
[CrossRef] [PubMed]
23. Yoo, H.D.; Lee, S.N.; Kang, H.A.; Cho, H.Y.; Lee, I.K.; Lee, Y.B. Influence of ABCB1 genetic polymorphisms
on the pharmacokinetics of risperidone in healthy subjects with CYP2D6* 10/* 10. Br. J. Pharmacol. 2011,
164, 433–443. [CrossRef] [PubMed]
24. Chen, X.; Murawski, A.; Patel, K.; Crespi, C.L.; Balimane, P.V. A Novel Design of Artificial Membrane for
Improving the PAMPA Model. Pharm. Res. 2008, 25, 1511–1520. [CrossRef]
25. Bae, S.H.; Kwon, M.J.; Park, J.B.; Kim, D.; Kim, D.H.; Kang, J.S.; Kim, C.G.; Oh, E.; Bae, S.K.
Metabolic drug-drug interaction potential of macrolactin A and 7-O-succinyl macrolactin A assessed
by evaluating cytochrome P450 inhibition and induction and UDP-glucuronosyltransferase inhibition
in vitro. Antimicrob. Agents Chemother. 2014, 58, 5036–5046. [CrossRef]
26. Li, A.P. In vitro approaches to evaluate ADMET drug properties. Curr. Top. Med. Chem. 2004, 4, 701–706.
[CrossRef]
27. Kanaan, M.; Daali, Y.; Dayer, P.; Desmeules, J. Uptake/Efflux Transport of Tramadol Enantiomers and
O-Desmethyl-Tramadol: Focus on P-Glycoprotein. Basic Clin. Pharmacol. Toxicol. 2009, 105, 199–206.
[CrossRef]
28. Saarikoski, T.; Saari, T.I.; Hagelberg, N.M.; Backman, J.T.; Neuvonen, P.J.; Scheinin, M.; Olkkola, K.T.; Laine, K.
Effects of terbinafine and itraconazole on the pharmacokinetics of orally administered tramadol. Eur. J.
Clin. Pharmacol. 2015, 71, 321–327. [CrossRef]
29. Mattia, C.; Coluzzi, F. Once-daily tramadol in rheumatological pain. Expert Opin. Pharmacother. 2006,
7, 1811–1823. [CrossRef]
30. T’jollyn, H.; Vermeulen, A.; Van Bocxlaer, J. PBPK and its virtual populations: The impact of physiology on
pediatric pharmacokinetic predictions of tramadol. AAPS J. 2019, 21, 8. [CrossRef]
31. Pubchem. O-Desmethyltramadol. Available online: https://pubchem.ncbi.nlm.nih.gov/compound/9838803
(accessed on 31 January 2019).
32. Wojsławski, J.; Białk-Bielińska, A.; Stepnowski, P.; Dołżonek, J. Leaching behavior of pharmaceuticals and
their metabolites in the soil environment. Chemosphere 2019, 231, 269–275. [CrossRef] [PubMed]
33. T’jollyn, H.; Snoeys, J.; Vermeulen, A.; Michelet, R.; Cuyckens, F.; Mannens, G.; Van Peer, A.; Annaert, P.;
Allegaert, K.; Van Bocxlaer, J.; et al. Physiologically Based Pharmacokinetic Predictions of Tramadol Exposure
Throughout Pediatric Life: An Analysis of the Different Clearance Contributors with Emphasis on CYP2D6
Maturation. AAPS J. 2015, 17, 1376–1387. [CrossRef] [PubMed]

39
Pharmaceutics 2019, 11, 618

34. Barbera, N.; Fisichella, M.; Bosco, A.; Indorato, F.; Spadaro, G.; Romano, G. A suicidal poisoning due to
tramadol. A metabolic approach to death investigation. J. Forensic Leg. Med. 2013, 20, 555–558. [CrossRef]
[PubMed]
35. Perdreau, E.; Iriart, X.; Mouton, J.B.; Jalal, Z.; Thambo, J.B. Cardiogenic shock due to acute tramadol
intoxication. Cardiovasc. Toxicol. 2015, 15, 100–103. [CrossRef]
36. Lee, S.Y.; Sohn, K.M.; Ryu, J.Y.; Yoon, Y.R.; Shin, J.G.; Kim, J.W. Sequence-based CYP2D6 genotyping in the
Korean population. Ther. Drug Monit. 2006, 28, 382–387. [CrossRef]
37. Vazzana, M.; Andreani, T.; Fangueiro, J.; Faggio, C.; Silva, C.; Santini, A.; Garcia, M.; Silva, A.; Souto, E.
Tramadol hydrochloride: Pharmacokinetics, pharmacodynamics, adverse side effects, co-administration of
drugs and new drug delivery systems. Biomed. Pharmacother. 2015, 70, 234–238. [CrossRef]
38. Costa, I.; Oliveira, A.; Guedes de Pinho, P.; Teixeira, H.M.; Moreira, R.; Carvalho, F.; Jorge Dinis-Oliveira, R.
Postmortem Redistribution of Tramadol and O-Desmethyltramadol. J. Anal. Toxicol. 2013, 37, 670–675.
[CrossRef]
39. Tzvetkov, M.V.; Saadatmand, A.R.; Lötsch, J.; Tegeder, I.; Stingl, J.C.; Brockmöller, J. Genetically polymorphic
OCT1: Another piece in the puzzle of the variable pharmacokinetics and pharmacodynamics of the
opioidergic drug tramadol. Clin. Pharmacol. Ther. 2011, 90, 143–150. [CrossRef]
40. Stamer, U.M.; Frank, M.; Stuber, F.; Brockmoller, J.; Steffens, M.; Tzvetkov, M.V. Loss-of-function
polymorphisms in the organic cation transporter OCT1 are associated with reduced postoperative tramadol
consumption. Pain 2016, 157, 2467–2475. [CrossRef]
41. Slanar, O.; Nobilis, M.; Kvétina, J.; Matousková, O.; Idle, J.R.; Perlík, F. Pharmacokinetics of tramadol is
affected by MDR1 polymorphism C3435T. Eur. J. Clin. Pharmacol. 2007, 63, 419–421. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

40
 pharmaceutics

Article
Lack of Correlation between In Vitro and In Vivo
Studies on the Inhibitory Effects of (-)-Sophoranone
on CYP2C9 Is Attributable to Low Oral Absorption
and Extensive Plasma Protein Binding of
(-)-Sophoranone
Yu Fen Zheng 1,† , Soo Hyeon Bae 2,3,† , Zhouchi Huang 3 , Soon Uk Chae 3 , Seong Jun Jo 3 ,
Hyung Joon Shim 3 , Chae Bin Lee 3 , Doyun Kim 3,4 , Hunseung Yoo 4 and Soo Kyung Bae 3, *
1 School of Basic Medicine and Clinical Pharmacy, China Pharmaceutical University, 639 Longmian Road,
Jiangning District, Nanjing 211198, China; 1020172557@cpu.edu.cn
2 Q-fitter, Inc., Seoul 06578, Korea; sh.bae@qfitter.com
3 College of Pharmacy and Integrated Research Institute of Pharmaceutical Sciences, The Catholic University,
Korea, Bucheon 14662, Korea; hzc0826@catholic.ac.kr (Z.H.); zldtnseoz@naver.com (S.U.C.);
sungjun6734@naver.com (S.J.J.); tony6533@naver.com (H.J.S.); aribri727@catholic.ac.kr (C.B.L.);
doyun325@naver.com (D.K.)
4 Life Science R&D Center, SK Chemicals, 310 Pangyo-ro, Sungnam 13494, Korea; hs.yoo@sk.com
* Correspondence: baesk@catholic.ac.kr; Tel.: +82-2-2164-4054
† These authors contributed equally to this work.

Received: 8 March 2020; Accepted: 5 April 2020; Published: 7 April 2020

Abstract: (-)-Sophoranone (SPN) is a bioactive component of Sophora tonkinensis with various

pharmacological activities. This study aims to evaluate its in vitro and in vivo inhibitory potential
against the nine major CYP enzymes. Of the nine tested CYPs, it exerted the strongest inhibitory
effect on CYP2C9-mediated tolbutamide 4-hydroxylation with the lowest IC50 (Ki ) value of 0.966
± 0.149 μM (0.503 ± 0.0383 μM), in a competitive manner. Additionally, it strongly inhibited other
CYP2C9-catalyzed diclofenac 4 -hydroxylation and losartan oxidation activities. Upon 30 min
pre-incubation of human liver microsomes with SPN in the presence of NADPH, no obvious shift
in IC50 was observed, suggesting that SPN is not a time-dependent inactivator of the nine CYPs.
However, oral co-administration of SPN had no significant effect on the pharmacokinetics of diclofenac
and 4 -hydroxydiclofenac in rats. Overall, SPN is a potent inhibitor of CYP2C9 in vitro but not in vivo.
The very low permeability of SPN in Caco-2 cells (Papp value of 0.115 × 10−6 cm/s), which suggests
poor absorption in vivo, and its high degree of plasma protein binding (>99.9%) may lead to the lack
of in vitro–in vivo correlation. These findings will be helpful for the safe and effective clinical use
of SPN.

Keywords: (-)-sophoranone; CYP2C9; potent inhibition; in vitro; in vivo; drug interaction; low
permeability; high plasma protein binding

1. Introduction
(-)-Sophoranone (SPN; Figure 1), a major bioactive flavonoid isolated from the roots of Sophora
tonkinensis, is used in traditional Chinese medicine for the treatment of acute pharyngolaryngeal
infections and sore throat [1–3]. It exhibits anti-inflammatory effects by inhibiting nitric oxide production
in macrophages [4] and 5-lipoxygenase activity [3]. Several studies have also demonstrated its other
biological activities, such as anti-cancer [5], anti-diabetic diabetic [6], and immunomodulatory [7]
activities. In our previous study, after orally administering 12.9 mg/kg SPN to rats, the maximum

41
Pharmaceutics 2020, 12, 328; doi:10.3390/pharmaceutics12040328 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 328

plasma concentration (Cmax ) was approximately 13.1 ng/mL at 60 min [8]. Thus, although conclusive
results are lacking, SPN is likely to be a promising drug candidate.

Figure 1. Chemical structure of (-)-sophoranone (SPN).

Drug–drug interactions can increase the likelihood of treatment failure or the frequency and
severity of adverse events [9]. Thus, drug–drug interaction assessment is a critical component of
new drug discovery and development as well as clinical practice [9,10]. The majority of known
drug interactions occur because of inhibition of drug-metabolizing enzymes [11–13]. Among all
drug-metabolizing enzymes, the cytochrome P450 (CYP) superfamily plays an important role in the
oxidation of almost 90% of currently used drugs [14]. Among at least 57 human cytochrome P450
enzymes identified to date, 9 hepatic P450 enzymes (CYP1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and
3A4) have shown to play predominant roles in the metabolism of drugs and other xenobiotics [12].
Therefore, the inhibitory potential of SPN on the nine major CYP enzymes should also be investigated.
There are a few reports on the in vitro and in vivo inhibitory effects of SPN on CYP enzymes. In rats,
oral administration of 5 g/kg S. tonkinensis extract over 14 days was found to increase the plasma
concentrations of metoprolol, omeprazole, and bupropion. This might be attributed to the inhibition of
the activities of rat CYP enzymes, CYP2D6, CYP2C19, and CYP2B6 [15]. However, these results could
not directly reflect the in vivo inhibitory potential of SPN on CYP enzymes due to multiple components
of the extract. Several flavonoids, including SPN, have been found to inhibit CYP3A4-mediated
reactions in vitro [16].
However, currently, there is limited information about SPN’s in vitro inhibitory potentials,
especially on the other eight CYP enzymes, thereby warranting further in vitro and in vivo
investigations to improve our understanding of drug interactions with SPN. Using human liver
microsomes in this study, we evaluated SPN’s potential to inhibit CYP1A2, CYP2A6, CYP2B6, CYP2C8,
CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4 in a reversible and time-dependent manner. We
report herein that SPN is a potent inhibitor of CYP2C9 in vitro but not in vivo. To explain this lack of
correlation between in vitro and in vivo results, we performed plasma protein binding of SPN and
permeability test using Caco-2 cells.

2. Materials and Methods

2.1. Chemicals and Reagents

Pooled human liver microsomes from 150 donors (75 males; 75 females) were purchased from
Corning Life Sciences (Woburn, MA, USA), and (-)-sophoranone (99.7% purity; SPN) was supplied
by SK Chemicals Ltd. (Sungnam, Gyeonggi-do, Korea). β-Nicotinamide adenine dinucleotide
phosphate disodium salt (NADP), glucose 6-phosphate disodium salt hydrate, glucose 6-phosphate
dehydrogenase, MgCl2 , and all chemicals including the specific substrates, its metabolites, and
well-known inhibitors of nine P450s were purchased from Sigma–Aldrich Corporation (St. Louis,
MO, USA), Santa Cruz Biotechnology (Dallas, TX, USA), or Cayman Chemicals (Ann Arbor, MI, USA)
unless stated otherwise. The purity of all purchased compounds was higher than 97.0%. HPLC-grade
acetonitrile and methanol were obtained from Burdick & Jackson Company (Morristown, NJ, USA).

42
Pharmaceutics 2020, 12, 328

Caco-2 cells were supplied by the Korean Cell Line Bank (Seoul, Korea) and cultured according to
the supplier’s recommendations. Transwell (24-well, 6.5 mm polycarbonate inserts, 0.4-μm pore) and
cell culture reagents were purchased from Corning Life Sciences. Heparinized human plasma was
obtained from donors at the Severance Hospital of Yonsei University Health System (Seoul, Korea) and
stored at −80 ◦ C prior to use.

2.2. Reversible Inhibition of (-)-Sophoranone towards the Nine CYP Isoforms in Human Liver Microsomes
The inhibitory effects of SPN on CYP1A2, CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6,
CYP2E1, and CYP3A4 were evaluated in pooled human liver microsomes through the use of specific
CYP probe substrates (cocktail assay), as previously described [17,18] with a slight modification.
Concentrations of each CYP probe in Table 1 were used close to their reported Km values [17,18].
Briefly, a 90-μL incubation mixture, including pooled human liver microsomes (final concentration
0.1 mg/mL), 50 mM phosphate buffer (pH 7.4), each CYP-probe substrate cocktail set, and SPN
(0–50 μM), was pre-incubated for 5 min at 37 ◦ C. SPN was dissolved in methanol and spiked into the
incubation mixture to a final concentration of 0.5% methanol. All P450-selective substrates (except
coumarin due to solubility) were dissolved in methanol and serially diluted with methanol to the
required concentrations, and the organic solvent was subsequently evaporated under a gentle stream
of N2 gas to minimize the effects of organic solvents on CYP activities. On the other hand, coumarin
dissolved in 50 mM phosphate buffer (pH 7.4) was directly added into the mixed tube. The reaction
was initiated by adding 10-μL aliquot of NADPH-generating system (1.3 mM NADP+ , 3.3 mM glucose
6-phosphate, 3.3 mM MgCl2 , and 0.4 unit/mL glucose-6-phosphate dehydrogenase) before 15 min
incubation at 37 ◦ C in a shaking water bath. After incubation, the reactions were stopped by adding
200 μL of ice-cold acetonitrile containing 2 μM chlorpropamide as an internal standard. The incubation
mixtures were centrifuged (16,000× g, 15 min) and 5 μL of the supernatant was injected into the
LC-MS/MS system. All incubations were performed in triplicate, and the data are shown as the mean
± standard deviation. Incubation samples containing well-known CYP inhibitors for each isozyme
(Table 2) in parallel were included to compare inhibitory effects, all of which appear on the US FDA list
of recommended or accepted in vitro inhibitors [12,19–21].
Additionally, to determine whether the inhibition of CYP2C9 by SPN was substrate specific, we
also examined SPN’s inhibitory effects on other CYP2C9-specific biotransformation pathways (i.e.,
diclofenac 4 -hydroxylation and losartan oxidation) in human liver microsomes [22,23]. Diclofenac and
losartan were used at 5 μM, respectively, and other procedures were similar to those of cocktail assays.

2.3. Determination of the Ki of (-)-Sophoranone on CYP2C9 Activity in Human Liver Microsomes

Among the nine tested CYP enzymes, SPN showed the lowest IC50 value for CYP2C9 (Table 2).
Based on the IC50 values, the Ki values of SPN on CYP2C9 activity were determined. Briefly, Ki
values were obtained by incubating various concentrations of two CYP2C9 probe substrates (50, 100,
and 150 μM tolbutamide; or 2, 5, and 10 μM diclofenac) in the presence of 0−5 μM SPN or 0−2 μM
sulfaphenazole, a well-known typical CYP2C9 inhibitor. Other procedures were similar to those of the
reversible inhibition studies. All incubations were performed in triplicate, and the data are shown as
the mean ± standard deviation.

43
 Table 1. Optimized mass parameters for the detection of metabolites of the nine P450-probe substrates and internal standard used in the cocktail assays.

CYPs Probe Substrates Km (μM) Metabolite ESI a Q1 Ion (m/z) Q3 Ion (m/z) Q1 Pre-bias (V) CE b (eV) Q3 Pre-bias (V)
1A2 Phenacetin 50 Acetaminophen + 152 110.2 −14 −12 −19
2A6 Coumarin 5 7-Hydroxycoumarin + 163 107 −15 −35 −15
2B6 Bupropion 50 6-Hydroxybupropion + 256 238 −15 −35 −15
2C8 Rosiglitazone 10 p-Hydroxyrosiglitaonze + 374 151 −15 −35 −15
2C9 Tolbutamide 100 4-Hydroxytolbutamide + 287 87 −15 −35 −15
Pharmaceutics 2020, 12, 328

2C19 Omeprazole 20 5-Hydroxyomeprazole + 362 214 −13 −13 −22

2D6 Dextromethorphan 5 Dextrorphan + 258 157 −15 −35 −15
2E1 Chlorzoxazone 50 6-Hydroxychlorzoxazone − 184 119.9 18 15 24
3A4 Midazolam 2 1 -Hydroxymidazolam + 342 203 −15 −35 −15
Chlorpropamide (Internal + 277 111 −15 −20 −15
2
standard) − 275 190 15 35 15
The optimized ion spray voltage was 4 kV and a nebulizing gas flow of 3 L/min, heating gas flow of 10 L/min, an interface temperature of 300 ◦ C, desolvation line temperature of 250 ◦ C,
heating block temperature of 400 ◦ C, and a drying gas flow rate of 10 L/min. a ESI, electrospray ionization mode; b CE, collision energy.

44
Pharmaceutics 2020, 12, 328

Table 2. IC50 values of well-known CYP inhibitors and SPN in reversible inhibition studies using a
cocktail assay (n = 3).

IC50 Values (μM)

CYPs
Well-Known Inhibitors SPN
1A2 α-Naphthoflavone 0.0458 ± 0.00694 >50 a
2A6 Tryptamine 2.98 ± 0.635 >50 a
2B6 Ticlopidine 2.19 ± 0.513 >50 a
2C8 Quercetin 8.51 ± 0.958 13.6 ± 3.15
2C9 Sulfaphenazole 0.677 ± 0.109 0.966 ± 0.149
2C19 S-benzylnirvanol 0.215 ± 0.0228 16.8 ± 3.21
2D6 Quinidine 0.127 ± 0.0192 >50
2E1 Diethyldithiocarbamate 12.0 ± 3.67 >50 a
3A4 Ketoconazole 0.0404 ± 0.00821 >50
Data represent the mean ± standard deviation of triplicate. a The remaining activities at the highest concentration
tested, 50 μM, were greater than 80%.

2.4. Time-Dependent Inactivation of (-)-Sophoranone toward the Nine CYP Isoforms in Human
Liver Microsomes
Pooled human liver microsomes (1 mg/mL) were incubated with SPN (0−50 μM) for 30 min at
37 ◦ C in the absence or presence of an NADPH-generating system (i.e., the “inactivation incubation”).
After inactivation incubation, aliquots (10 μL) were transferred into fresh incubation tubes (final
volume 100 μL) containing an NADPH-generating system and each P450-selective substrate cocktail
set. The reaction mixtures were incubated for 15 min at 37 ◦ C in a shaking water bath. After incubation,
the reactions were stopped by adding 200 μL of ice-cold acetonitrile containing 2 μM chlorpropamide,
as an internal standard. The incubation mixtures were centrifuged (16,000× g, 15 min) and 5 μL of the
supernatant was injected into the LC-MS/MS system. All incubations were performed in triplicate, and
the data are shown as the mean ± standard deviation.

2.5. Caco-2 Cell Permeability of (-)-Sophoranone

Caco-2 cell permeability was assessed to predict the oral absorption of SPN. Cell culture and
transport studies were performed as previously described [24,25]. Briefly, for the bi-directional
transport studies, the cells were seeded at a density of 1 × 105 cells/well, and the cell medium was
replaced until they formed confluent monolayers. On the 25th day, the cell monolayers were washed
with pre-warmed HBSS buffer. The bi-directional permeability assay was instigated by adding 10 μM
for propranolol, or 10 μM and 50 μM for SPN in HBSS to an apical well (200 μL) for apical (A) to
basolateral (B) transport or to a basolateral insert (800 μL) for the B to A transport. Before the experiment,
the integrity of the cell monolayers was evaluated by measuring the transepithelial electrical resistance
using a Millicell ohmmeter. After 2 h incubation at 37 ◦ C, samples were withdrawn from both sides,
respectively. All samples were stored at −80 ◦ C until LC-MS/MS analysis, and all experiments were
performed in triplicate.
The apparent permeability coefficient (Papp ) was calculated using the following equation.

Papp = (Vr /C0 ) × (1/A) × ([Drug]/t)

where, Vr is the volume of medium in the receiver chamber, C0 is the donor compartment concentration
at time zero, A is the area of the cell monolayer, t is the treatment time of the drug, and [Drug] is the
drug concentration in the receiver chamber.

2.6. Effects of (-)-Sophoranone on the Pharmacokinetics of Diclofenac in Rats

In this study, we investigated whether SPN, an in vitro potent inhibitor of CYP2C9, affects the
pharmacokinetics of diclofenac in rats. Male Sprague–Dawley rats (8 weeks, 270–290 g) were purchased

45
Pharmaceutics 2020, 12, 328

from Orient Bio (Sungnam, Gyeonggi-do, Korea), and the protocol for pharmacokinetic interaction
studies in rats was approved by the Institutional Animal Care and Use Committee (IACUC-CUK) at
The Catholic University of Korea (Approval No. 2019-021, approved 31 May 2019). The procedures
used for housing and handling were previously reported [18]. Before administration, rats were fasted
for 12 h with free access to water. The carotid arteries of each rat were cannulated with a polyethylene
tube (Clay Adams, Franklin Lakes, NJ, USA) for blood sampling. Each rat was individually housed
in a rat metabolic cage and allowed to recover from anesthesia for 4–5 h prior to the start of the
experiment. The rats were divided into two groups: (1) diclofenac alone (n = 6) and (2) SPN and
diclofenac co-administration (n = 6). SPN was suspended in dimethylsulfoxide:PEG400:distilled water
(5:60:35, v/v/v) and administered by oral gavage at a dose of 75 mg/kg in a volume of 5 mL/kg. Fifteen
minutes after oral administration of SPN, 2 mg/kg diclofenac was dissolved in normal saline and
administered by oral gavage. Approximately 0.25 mL of blood from each rat was collected into an
Eppendorf tube before diclofenac dosing (0 min), and at 3, 5, 10, 15, 30, 45, 60, 90, 120, 180, 240, 360,
and 480 min post-dosing. The blood samples were immediately centrifuged at 13,000× g for 5 min at
4 ◦ C. The plasma samples were divided into two Eppendorf tubes by 50 μL and stored at −80 ◦ C until
LC-MS/MS analysis. After the experiments, the rats were euthanized with CO2 .

2.7. Determination of the Unbound Fraction of (-)-Sophoranone in Plasma and Human Liver Microsomes
The plasma or liver microsomal protein bindings were performed using a rapid equilibrium
dialysis device and cellulose membranes with a molecular weight cutoff of 8000 (Thermo Scientific,
Rockford, IL, USA) [17]. The rat and human plasma samples (200 μL) containing SPN at 10 and 50 μM,
respectively, were dialyzed against a dialysis buffer, phosphate-buffered saline (PBS, 400 μL). The
loaded dialysis plate was covered with sealing tape, placed on an orbital shaker at approximately
200 rpm, and incubated at 37 ◦ C for 4 h. Thereafter, samples (100 μL) from both PBS and plasma
chambers were collected and mixed with an equal volume of blank plasma and PBS, respectively. All
samples were stored at −80 ◦ C until LC-MS/MS analysis. The unbound fraction of SPN in human (or
rat) plasma was calculated by dividing the SPN concentration in PBS by that in plasma.
The human liver microsomal incubation mixtures (final concentration 0.1 mg/mL) without NADPH
generating system were used to determine the unbound fraction of SPN. Other procedures were similar
to those of plasma protein binding assay.

2.8. LC-MS/MS Analysis

2.8.1. In Vitro Samples

Metabolites of nine P450-selective substrates were analyzed using a Shimadzu Nexera X2 UPLC
system coupled to an LCMS-8050 triple quadruple mass spectrometer (Shimadzu Corporation, Kyoto,
Japan) equipped with an electrospray ionization interface as previously described with a slight
modification [17,18]. Separation was performed on a reversed-phase column (Luna C18 , 50 mm × 2.0
mm i.d.; 3 μm particle size; Phenomenex, Torrance, CA, USA) maintained at 40 ◦ C. The mobile phase
consisted of distilled water containing 0.1% formic acid (A) and acetonitrile containing 0.1% formic
acid (B), with a flow rate of 0.5 mL/min. The gradient elution program used was as follows: (1) Mobile
phase A was set to 95% at 0 min, (2) a linear gradient was run to 5% in 2.6 min, and (3) a linear
gradient was run to 95% in 3.0 min and re-equilibrated for 2 min. The total run time was 5 min. The
optimized compound-dependent parameters of the metabolites of the nine P450-selective substrates
and the internal standard are listed in Table 1. Three-day validations were performed to confirm
the effectiveness of the LC-MS/MS system for simultaneous determination of the nine P450-selective
substrate metabolites at the respective ranges of 0.01–10 μM in blank microsomal incubation mixtures.
We found that the precision (≤12.1%) and accuracy (95.4–110.2%) values were within acceptable
ranges. Supplemental Figure S1 shows the representative LC-MS/MS chromatograms of a human liver
microsomal incubation sample containing nine P450-selective metabolites and an internal standard.

46
Pharmaceutics 2020, 12, 328

The auto-optimized mass transitions were m/z 312 > 231 and m/z 437 > 207.1 for quantification of
4 -hydroxydiclofenac and losartan carboxylic acid, respectively. HPLC conditions were the same as
those in the cocktail assay.

2.8.2. In Vivo Samples

The plasma concentrations of diclofenac and 4 -hydroxydiclofenac were determined by a
previously reported LC-MS/MS method [26] with some modifications. Briefly, 50 μL aliquots of plasma
were extracted with 300 μL aliquots of acetonitrile containing chlorpropamide (internal standard),
followed by LC-MS/MS (Shimadzu Corporation). Chromatographic separation was performed on
a Phenomenex Luna C18 column (100 × 2.00 mm; 3.0 μm). The isocratic mobile phase consisted of
0.1% formic acid in distilled water (A) and 0.1% formic acid in acetonitrile (B) (45:55, v/v), with a
flow rate of 0.3 mL/min. The transitions were m/z 296.0 > 214.0 for diclofenac, m/z 312 > 231 for
4 -hydroxydiclofenac, and m/z 277 > 111 for the internal standard. The data acquisition was computed
using LabSolutions LCMS Ver.5.6 (Shimadzu Corporation). The calibration curves for diclofenac and
4 -hydroxydiclofenac were linear (r ≥ 0.996) over the concentration range of 20–5000 ng/mL.
The LC-MS/MS condition for the determination of SPN in plasma was the same with a previously
reported method [27]. The calibration curve for SPN was linear (r ≥ 0.995) over the concentration
range of 1–250 ng/mL.

2.9. Analysis of Inhibition Kinetics and Pharmacokinetic Parameters

The IC50 values were calculated via nonlinear least-squares regression analysis from logarithmic
plots of inhibitor concentration versus percentage of activity remaining after inhibition, using SigmaPlot
(ver. 14.0; Systat Software Inc, Chicago, IL, USA). The Ki values were determined from the equations for
a single substrate single inhibitor model and the software available in the SigmaPlot Enzyme Kinetics
module. Competitive, non-competitive, uncompetitive, or mixed inhibition models were evaluated
and ranked according to the best fit based on Akaike Information Criterion (AIC) values. For visual
inspection, the data were presented as Dixon plots.
Pharmacokinetic parameters were calculated by a non-compartmental analysis using WinNonlin
Professional software (version 5.2, Pharsight Corp., Mountain View, CA, USA) that used the total area
under the plasma concentration–time curve from time zero to infinity (AUC∞ ) or the last measured
time (AUCt ). The logarithmic trapezoidal rule was used during the declining plasma level phase and
the linear trapezoidal rule was used for the rising plasma-level phase. The peak plasma concentration
(Cmax ) and time to reach Cmax (Tmax ) were read directly from the experimental data. Statistically
significant differences were recognized at p < 0.05.

3. Results

3.1. Reversible Inhibition of (-)-Sophoranone toward the Nine CYP Isoforms in Human Liver Microsomes
The inhibitory effects of SPN on the activities of nine CYP isozymes (CYP1A2, CYP2A6, CYP2B6,
CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4) in human liver microsomes are shown
in Figure 2, and the IC50 values are listed in Table 2. The IC50 values for the positive controls
used in the reversible inhibition studies were in an acceptable degree of accuracy with published
values [12,19–21]. Of the P450 isoforms tested, SPN exerted the strongest inhibitory effect on
CYP2C9-catalyzed tolbutamide hydroxylation, with an IC50 value of 0.966 ± 0.149 μM (Table 2). SPN
showed weak inhibitory effects toward CYP2C8 and CYP2C19, with IC50 values of 13.6 ± 3.15 μM and
16.8 ± 3.21 μM, respectively. However, SPN had no apparent inhibitory effects toward the other CYPs
tested (Table 2); the residual enzyme activities at the highest tested concentration (50 μM) were greater
than 80%, except for CYP2D6 (53.9 ± 3.53%) and CYP3A4 (53.3 ± 4.00%) (Figure 2).

47
Pharmaceutics 2020, 12, 328

(A) CYP1A2 (B) CYP2A6 (C) CYP2B6

120 120 120

90 90 90

60 60 60

30 30 30

0 0 0
0.05 0.5 5 50 0.05 0.5 5 50 0.05 0.5 5 50
% of control activity

120 (D) CYP2C8 120 (E) CYP2C9 120

(F) CYP2C19

90 90 90

60 60 60

30 30 30

0 0 0
0.05 0.5 5 50 0.05 0.5 5 50 0.05 0.5 5 50

(G) CYP2D6 (H) CYP2E1 (I) CYP3A4

120 120 120

90 90 90

60 60 60

30 30 30

0 0 0
0.05 0.5 5 50 0.05 0.5 5 50 0.05 0.5 5 50

SPN ( M)
Figure 2. Inhibition curves of SPN on the nine major P450 activities in human liver microsomes
using substrate cocktails including CYP1A2 for phenacetin O-deethylase (A), CYP2A6 for coumarin
7-hydroxylase (B), CYP2B6 for bupropion hydroxylase (C), CYP2C8 for rosiglitazone p-hydroxylase (D),
CYP2C9 for tolbutamide 4-hydroxylase (E), CYP2C19 for omeprazole 5-hydroxylase (F), CYP2D6 for
dextromethorphan O-demethylase (G), CYP2E1 for chlorzoxazone 6-hydroxylase (H), and CYP3A4 for
midazolam 1 -hydroxylase (I). The activity is expressed as a percentage of remaining activity compared
with the control, no containing SPN. Data are the mean ± standard deviation of triplicate incubations.
The dashed lines represent the best fit to the data with non-linear regression.

To determine whether the inhibitory effects of SPN on CYP2C9 was substrate specific, we
examined the inhibitory effects on other CYP2C9-specific biotransformation pathways (i.e., diclofenac
4 -hydroxylation and losartan oxidation) and found that SPN also markedly inhibited their activities,
with IC50 values of 0.879 ± 0.0888 μM and 0.455 ± 0.0486 μM, respectively, (Figure 3).

48
Pharmaceutics 2020, 12, 328

(A) Diclofenac hydoxylation (B) Losartan oxidation

120 120

% of control activity
% of control activity
90 90

60 60

30 30

0 0
0.05 0.5 5 50 0.05 0.5 5 50
SPN ( M) SPN ( M)

Figure 3. Inhibition curves of SPN on the CYP2C9-catalyzed diclofenac 4 -hydroxylation (A) and
losartan oxidation (B) activities in human liver microsomes. Data are the mean ± standard deviation of
triplicate incubations. The dashed lines represent the best fit to the data with non-linear regression.

3.2. Determination of the Ki of (-)-Sophoranone for CYP2C9 Activity

Based on the lowest IC50 value for CYP2C9, to characterize the type of reversible inhibition of
CYP2C9 by SPN, enzyme kinetic experiments were performed in the presence of various concentrations
of SPN and tolbutamide, or diclofenac. Otherwise, identical samples containing a known potent
CYP2C9 inhibitor (sulfaphenazole), were included in the analysis. Representative Dixon plots of
CYP2C9 inhibition by SPN and sulfaphenazole in human liver microsomes are shown in Figure 4,
and the Ki values are summarized in Supplemental Table S1. Using a nonlinear regression analysis,
SPN demonstrated competitive inhibition against CYP2C9-catalyzed tolbutamide hydroxylation or
diclofenac hydroxylation, with calculated Ki values of 0.503 ± 0.0383 μM and 0.587 ± 0.0470 μM
(Figure 4A,B). Sulphafenazole competitively inhibited CYP2C9 with a Ki value of 0.267 ± 0.0170 μM
(Figure 4C), which was similar to a previously reported value [28].

(A) (B) (C)

Formation of 4-hydroxytolbutamide
Formation of 4-hydroxytolbutamide

50 M
Formation of 4'-hydroxydiclofenac

50 M 2 M
1.0 25 1.0
100 M 5 M 100 M
150 M 10 M 150 M
1/v (nmol/min/mg)
1/v (nmol/min/mg)

1/v (pmol/min/mg)

0.8 20 0.8

0.6 15 0.6

0.4 10 0.4

0.2 5 0.2

-2 0 2 4 6 -2 0 2 4 6 -0.5 0.0 0.5 1.0 1.5 2.0

[SPN] ( M) [SPN] ( M) [Sulfaphenazole] ( M)

Figure 4. Dixon plots to determine Ki values of SPN on the CYP2C9 enzyme activity, using tolbutamide
(A) or diclofenac (B) as substrates. The well-known inhibitor of CYP2C9, sulfaphenazole, is used
as a positive control (C) using tolbutamide as a substrate. The concentrations of tolbutamide were
determined 50 (•), 100 (), and 150 () μM, respectively; diclofenac was used at 2 (•), 5 (), and 10
() μM, respectively. v represents formation rate of 4-hydroxytolbutamide (nmol/min/mg protein) or
4 -hydroxydiclofenac (pmol/min/mg protein). Data are the mean ± standard deviation of triplicate
incubations. The dashed lines of SPN (A,B) and sulfaphenazole (C) all fit well to competitive
inhibition types.

49
Pharmaceutics 2020, 12, 328

3.3. Time-Dependent Inactivation of (-)-Sophoranone towards the Nine CYP Isoforms in Human
Liver Microsomes
The IC50 shift method incorporating a dilution is one of the most efficient and convenient methods
for evaluating time-dependent inhibitory effects. A shift in IC50 to a lower value (“shift”) with
pre-incubation indicates time-dependent inactivation [29–31]. After 30 min pre-incubation of SPN
with human liver microsomes in the presence of NADPH, no obvious shift in IC50 was observed for
inhibition of the nine CYPs (Figure 5), suggesting that SPN is not a time-dependent inactivator for the
nine CYPs.

(A) CYP1A2 (B) CYP2A6 (C) CYP2B6

120 120 120

90 90 90

60 60 60

30 with NADP H 30 with NADP H 30 with NADP H

without NADP H without NADP H without NADP H
0 0 0
0.1 0.5 5 50 0.1 0.5 5 50 0.1 0.5 5 50

(D) CYP2C8 (E) CYP2C9 (F) CYP2C19

% of control activity

120 120 120

90 90 90

&<3&
60 60 60

30 with NADP H 30 with NADP H 30 with NADP H

without NADP H without NADP H without NADP H
0 0 0
0.1 0.5 5 50 0.1 0.5 5 50 0.1 0.5 5 50

(G) CYP2D6 (H) CYP2E1 (I) CYP3A4

120 120 120

90 90 90

60 60 60

30 with NADP H 30 with NADP H 30 with NADP H

without NADP H without NADP H
,
without NADP H
0 0 0
0.1 0.5 5 50 0.1 0.5 5 50 0.1 0.5 5 50

SPN ( M)
Figure 5. Time-dependent inhibition curves of SPN on the nine major P450 activities in human liver
microsomes using substrate cocktails after 30 min pre-incubation with the presence (•) or absence ()
of an NADPH-generating system. Data are the mean ± standard deviation of triplicate incubations.

3.4. Caco-2 Cell Permeability of (-)-Sophoranone

A bi-directional permeability assay using Caco-2 monolayer cells was performed to predict the
intestinal absorption of SPN. SPN showed very low permeability in both directions (from A-to-B and
B-to-A). The calculated Papp values of SPN from A-to-B were (0.115 ± 0.0369) × 10−6 cm/s and (0.172
± 0.0488) × 10–6 cm/s at 10 μM and 50 μM, respectively, (n = 3, each). These results indicated that
SPN is poorly absorbed in vivo. The Papp values from B-to-A were (0.101 ± 0.00444) × 10−6 cm/s at
10 μM (n = 3) and (0.152 ± 0.0353) × 10−6 cm/s at 50 μM (n = 3). SPN was not a substrate for efflux
transporters, that is, P-gp and BCRP, as the efflux ratio (B-to-A/A-to-B) is less than 2. The Papp of

50
Pharmaceutics 2020, 12, 328

propranolol, a reference high permeable compound, from A-to-B and B-to-A were (26.8 ± 3.31) × 10−6
cm/s and (21.5 ± 2.19) × 10−6 cm/s, respectively, (n = 3, each), similar to the reported values [24,25].

3.5. Effects of (-)-Sophoranone on the Pharmacokinetics of Diclofenac in Rats

We conducted pharmacokinetic studies to investigate the effects of SPN on the pharmacokinetics
of diclofenac in rats. Findings in the literature on the dried S. tonkinensis herbs indicate that a
recommended daily dose for an adult human with the body weight of 60 kg were to be 6–10 g [32],
which correlated to the equivalent dose ranges in rats, 0.620–1.03 g/kg [33]. He et al. [2] reported
that the average contents of SPN in various S. tonkinensis samples were found to be approximately
2.53 mg/g (0.0253%). Reflecting this content, the dosage in rats, 0.620–1.03 g/kg of the dried herb, might
be consistent with 15.7–26.1 mg/kg in terms of SPN. Thus, in this study, the SPN dose of 75 mg/kg was
used in rats, which is approximately 2.87- to 4.87-fold greater than the recommended human dose.
The mean plasma concentration-time profiles of diclofenac and 4-hydroxydiclofenac after oral
administration of diclofenac (2 mg/kg) in the absence or presence of oral co-administration of SPN
(75 mg/kg) in rats are illustrated in Figure 6, and the relevant pharmacokinetic parameters are shown
in Table 3. The plasma levels of diclofenac and 4-hydroxydiclofenac were similar in both groups
(Figure 6A,B). Likewise, no significant differences were observed in any other pharmacokinetic
parameter of diclofenac and 4 -hydroxydiclofenac (Table 3). The in vivo marker for CYP2C9 activity,
expressed as the molar AUC ratio of 4 -hydroxydiclofenac to diclofenac, was not significant (0.799
± 0.167 versus 0.904 ± 0.0534; p value of 0.215) in the presence or absence of SPN (Table 3). In the
treatment group with co-administration of SPN, the Cmax of SPN was found to be 33.7 ± 14.8 ng/mL
(0.0732 ± 0.0321 μM) at approximately 60–75 min post-dose (Figure 6C). Given the Ki values of SPN on
CYP2C9 activity (0.503 ± 0.0383 μM for tolbutamide hydroxylation and 0.587 ± 0.0470 μM for diclofenac
hydroxylation), the plasma concentrations of SPN are too low to inhibit CYP2C9-mediated metabolism
of diclofenac in vivo. Overall, the co-administration of SPN did not alter the pharmacokinetics of
diclofenac and 4 -hydroxydiclofenac.

(A) Diclofenac (B) 4'-Hydroxydiclofenac (C) SPN

1200 300 60
Plasma concentrations (ng/mL)

without SPN without SPN

with SPN with SPN

900 45
200

600 30

100
300 15

0 0 0
0 120 240 360 480 0 120 240 360 480 0 120 240 360 480
Time (min)
Figure 6. Mean plasma concentrations of diclofenac (A) and 4 -hydroxydiclofenac (B) after oral
administration of diclofenac at a dose of 2 mg/kg without (•, n = 6) or with (, n = 6) oral dosing of SPN
(75 mg/kg) to rats. Mean plasma concentrations of SPN (C) after co-administration of SPN (75 mg/kg)
and diclofenac (2 mg/kg) to rats (, n = 6). Vertical bars mean standard deviation.

51
Pharmaceutics 2020, 12, 328

Table 3. Mean (± standard deviations) pharmacokinetic parameters of diclofenac and 4 -hydroxydiclofenac

after oral administration of diclofenac at a dose of 2 mg/kg without or with oral administration of SPN
(75 mg/kg) to rats.

Parameters Without SPN (n = 6) With SPN (n = 6)

Diclofenac
AUCt (μg min/mL) a 63.8 ± 6.28 69.4 ± 2.98
AUC∞ (μg min/mL) b 71.7 ± 9.16 80.8 ± 7.78
t1/2 (min) c 153 ± 60.1 173 ± 41.5
Cmax (ng/mL) d 882 ± 245 787 ± 104
Tmax (min) e 5 (3–5) 5 (3–5)
4 -hydroxydiclofenac
AUCt (μg min/mL) 44.8 ± 6.38 44.5 ± 7.24
AUC∞ (μg min/mL) 68.6 ± 12.1 67.6 ± 12.9
t1/2 (min) 278 ± 70.0 296 ± 59.7
Cmax (ng/mL) 180 ± 55.2 173 ± 40.8
Tmax (min) 10 (10–30) 15 (10–30)
Metabolic conversion ratio f
AUC∞, 4 -hydroxydiclofenac /AUC∞, diclofenac 0.904 ± 0.0534 0.799 ± 0.167
No significant differences were observed in all pharmacokinetic parameters of diclofenac and 4 -hydroxydiclofenac.
a Total area under the plasma concentration–time curve from time zero to time last sampling time; b total area under

the plasma concentration–time curve from time zero to infinity; c terminal half-life; d peak plasma concentration;
max . Median (ranges); the metabolic conversion ratio, AUC∞,4 -hydroxydiclofenac /AUC∞,diclofenac , was
e time to reach C f

calculated based on a molar basis.

3.6. Determination of the Unbound Fraction of (-)-Sophoranone in Plasma and Human Liver Microsomes
SPN was extensively bound to plasma proteins, regardless of species. The free fractions (%) of
SPN at 10 and 50 μM in human plasma were 0.0457 ± 0.00612% and 0.0927 ± 0.0400%, respectively,
(n = 3, each). Similarly, when 10 and 50 μM SPN were added to the rat plasma, the free fractions were
0.0380 ± 0.0102% and 0.0531 ± 0.0149%, respectively, (n = 3, each). After adding 10 and 50 μM SPN
to rat and human plasma, free fractions remained relatively unchanged, suggesting that SPN has no
binding saturation in plasma.
SPN also exhibited marked non-specific bindings to human liver microsomes, although to a lesser
extent than those in human plasma. The unbound fractions of SPN at 10 and 50 μM were calculated to
be 0.621 ± 0.0405% and 0.724 ± 0.170%, respectively (n = 3, each), at a microsomal protein concentration
of 0.1 mg/mL.

4. Discussion
This study focused on the in vitro and in vivo inhibitory effects of SPN on human CYPs, especially
CYP2C9. We screened the inhibitory effects of SPN on the major human CYP isoforms (CYP1A2,
CYP2A6, CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1, and CYP3A4) in human liver
microsomes. Of the nine tested CYP isoforms, SPN exerted the strongest inhibitory effect on CYP2C9
activity, with the lowest IC50 value of 0.966 ± 0.149 μM (Table 2; Figure 2). In addition to CYP2C9, SPN
mildly inhibited several CYP enzymes, with potency ranked in the order CYP2C8 > CYP2C19; the
IC50 values were 13.6 ± 3.15 μM and 16.8 ± 3.21 μM, respectively (Table 2; Figure 2). Although the
IC50 values could not been calculated, SPN also appears to weakly inhibit CYP2D6 and CYP3A4; the
residual enzyme activities at the highest tested concentration (50 μM) were 53.9 ± 3.53% and 53.3 ±
4.00%, respectively (Figure 2). No apparent inhibition of the other CYPs (CYP1A2, CYP2A6, CYP2B6,
and CYP2E1) was observed (Figure 2). SPN also strongly inhibited other CYP2C9-catalyzed diclofenac
4 -hydroxylation and losartan oxidation activities (Figure 3). The inhibition mechanisms of SPN on
CYP2C9-catalyzed tolbutamide 4-hydroxylation and diclofenac 4 -hydroxylation activities were both
competitive, with Ki values of 0.503 ± 0.0383 μM and 0.587 ± 0.0470 μM, respectively. Pre-incubation

52
Pharmaceutics 2020, 12, 328

of SPN for 30 min with human liver microsomes and an NADPH-generating system did not alter the
inhibition potencies against the nine CYPs, suggesting that SPN is not a time-dependent inactivator.
The reversible inhibition of SPN-mediated CYP3A4 activity was less consistent with the published
literature. Li et al. [16] reported that among 44 tested flavonoids, SPN inhibited CYP3A4-catalyzed
bufalin 5 -hydroxylation activity with a Ki value of 2.17 ± 0.29 μM. They only focused on the in vitro
inhibitory potentials of several flavonoids against CYP3A4 activity. To the best of our knowledge,
to date, bufalin has not been used as the in vitro probe substrate for the CYP3A4 activity, and the
reference material of 5 -hydroxybufalin is not commercially available. Because of the presence of
several binding regions within the CYP3A4 active site, multiple probe substrates are often used
for in vitro CYP3A4-mediated drug–drug interaction studies, including midazolam, nifedipine, and
testosterone [34]. In that study, when other CYP3A4 substrates were tested, the ranges of IC50 values by
SPN were reported to be 5.62–38.4 μM [16]. Additionally, we examined the inhibitory effect on another
CYP3A4-catalyzed testosterone 6β-hydroxylation and found that SPN also inhibited the activity with
an IC50 value of 31.5 ± 4.79 μM, which showed a higher percentage inhibition compared to midazolam
(data not shown). Altogether, the in vitro CYP3A4 inhibition by SPN seemed to be substrate-specific.
Generally, alterations in the activities of hepatic CYPs through in vitro inhibition or induction
represent the major mechanisms underlying pharmacokinetic drug–drug interactions [11–13]. It has
been estimated that CYP2C9 is responsible for the metabolic clearance of up to 15–20% of all drugs
undergoing phase I metabolism, including clinically important drugs such as S-warfarin, phenytoin,
tolbutamide, losartan, and several anti-inflammatory drugs [23,35]. Considering that SPN is a potent
CYP2C9 inhibitor in vitro, there may be potential for herb–drug interactions between SPN and CYP2C9
substrates after concomitant oral administration.
Using the in vitro reversible inhibition results, a clinical drug–drug interaction risk was initially
predicted by the basic static model approach, as recommended by the EMA [36] and US FDA [37] with
calculating the R1 value (R1 = 1 + [Imax,u /Ki,u ]), which representing the predicted AUC ratio in the
presence or absence of inhibitor. Where, Imax,u (Cmax,u ) is maximal free plasma concentration of the
inhibitor and Ki,u is the unbound in vitro inhibition constant. However, little information is yet to be
reported on the Cmax values of SPN after oral administration of SPN. As stated in the Introduction, from
our previous study, the Cmax of SPN was reported to be 13.1 ng/mL in rats after oral dosing of 12.9 mg/kg
SPN in rats [8]. Thus, we investigated whether SPN affects the pharmacokinetics of diclofenac and
4 -hydroxydiclofenac, produced by hepatic CYP2C9 enzyme, in rats. In the group that received
co-administration of SPN (75 mg/kg), the Cmax of SPN was found to be 33.7 ± 14.8 ng/mL (0.0732
± 0.0321 μM) at 60–75 min (Figure 6C). These results suggest that SPN has low oral bioavailability.
The calculated values of Imax, u and Ki,u for SPN used in this study were 0.0420 ± 0.0184 nM and
3.39 ± 0.258 nM (3.95 ± 0.316 nM for diclofenac 4 -hydroxylation), respectively. Considering these
values, the R1 value of SPN for the inhibition of CYP2C9 in vitro was calculated as 1.0124 (Ki , u for
tolbutamide 4-hydroxylation) or 1.0106 (Ki, u for diclofenac 4 -hydroxylation) which are both below
the EMA and US FDA cut-off criteria of R1 , 1.02 [36,37], indicating that the potential for clinically
relevant drug interaction-mediated CYP2C9 inhibition by SPN may be low and no clinical interaction
studies are warranted. In our results, also no significant differences were observed in any of the other
pharmacokinetic parameters of diclofenac and 4 -hydroxydiclofenac in rats in the absence or presence
of oral co-administration of SPN at a dose of 75 mg/kg (Table 3). Furthermore, the molar metabolic
conversion ratio, expressed as AUC4 -hydroxydiclofenac /AUCdiclofenac , which indicated a causal factor for
the evaluation of the capacities of CYP2C9 activity in vivo, did not show significant differences (0.799
± 0.167 versus 0.904 ± 0.0534) in both groups (Table 3).
To explain the lack of in vitro–in vivo correlation, we assessed two factors that could limit the
accuracy of in vitro models in predicting metabolic drug interactions in vivo, which were SPN’s degree
of plasma protein binding and its permeability in Caco-2 cells. We found that SPN was extensively
bound in both human and rat plasma proteins (>99.9%) with a mean unbound fraction value of 0.0574%
in the range of 10 and 50 μM. Thus, taking the plasma protein binding of SPN into account, the unbound

53
Pharmaceutics 2020, 12, 328

maximum concentrations of SPN in plasma might be 0.0420 ± 0.0184 nM, which is much lower than
the unbound Ki values of SPN in vitro. Some drugs that indicate in vitro–in vivo discrepancy because
of high plasma protein bindings have been reported [38–40]. Tolfenamic acid strongly inhibited
CYP1A2 in vitro but not in vivo because of high plasma protein binding (99.7%) [38]. Montelukast
is a very potent inhibitor of CYP2C8 in vitro with Ki values ranging from 0.0092–0.15 μM [41].
However, in humans, montelukast has had no effect on the pharmacokinetics of the CYP2C8 substrates,
pioglitazone [39] and rosiglitazone [40]. The high degree of protein binding of montelukast in plasma
(>99.7%) is similar to that of tolfenamic acid and explicitly explains the lack of its in vivo interaction,
irrespective of its strong inhibitor potency in vitro. The Caco-2 cell model is widely used to predict the
absorption across the intestinal barrier, and a good correlation between its oral absorption in humans
and its apparent permeability (Papp ) across the Caco-2 cell barrier has been shown [24,25]. A recent
study has provided some updated guidelines on how permeability values might correlate with human
oral absorption: Low permeability (0–20% absorbed) is correlated to Papp values < 1–2 × 10−6 cm/s;
moderate permeability (20–80% absorbed) to Papp values < 2–10 × 10−6 cm/s; and high permeability
(80–100% absorbed) to Papp values > 10 × 10−6 cm/s [42]. Propranolol had >90% human absorption and
exhibited high permeability with a Papp value of (26.8 ± 3.31) × 10−6 cm/s in our assay. SPN exhibited
a very low permeability with mean Papp values of 0.115 × 10−6 cm/s (0.429% of propranolol Papp ) and
0.172 × 10−6 cm/s (0.642% of propranolol Papp ) at 10 and 50 μM, respectively, indicating that it is poorly
absorbed in vivo. SPN was not a substrate for efflux transporters, that is, P-gp and BCRP, as the efflux
ratio (B-to-A/A-to-B) is less than 2.
Overall, SPN is a potent inhibitor of CYP2C9 in vitro but not in vivo. This apparent discrepancy is
due to the extensive plasma protein binding and very low permeability of SPN, which resulted in poor
oral absorption. These approaches could help in making more reliable in vitro–in vivo extrapolations
about the risk of in vivo inhibition potential. In conclusion, these findings have provided useful
information on the safe and effective use of SPN in clinical practice.

Supplementary Materials: The following is available online at http://www.mdpi.com/1999-4923/12/4/328/s1,

Figure S1: Typical LC-MS/MS chromatograms for the formed metabolites of the nine CYP-specific probe in human
liver microsomes and their internal standard in the positive electrospray ionization mode. Table S1: Ki values and
inhibition types for CYP2C9 by SPN and sulfaphenazole in human liver microsomes (n = 3).
Author Contributions: Conceptualization, Y.F.Z., S.H.B., and S.K.B.; methodology, Z.H., S.U.C., C.B.L., and D.K.;
software, Z.H. and S.U.C.; validation, Y.F.Z., Z.H., S.U.C., C.B.L., S.J.J., and H.J.S.; formal analysis, Y.F.Z., S.U.C.,
and S.J.J.; investigation, Z.H., C.B.L., D.K., and H.Y.; resources, D.K. and H.Y.; data curation, Y.F.Z., S.H.B., and
S.K.B.; writing—original draft preparation, Y.F.Z. and S.H.B.; writing—review and editing, S.H.B. and S.K.B.;
visualization, S.J.J. and H.J.S.; supervision, S.K.B.; project administration, S.K.B.; funding acquisition, S.K.B. All
authors have read and agreed to the published version of the manuscript.
Funding: This research was supported by Industrial Core Technology Development Program funded by the
Ministry of Trade, Industry and Energy (No. 10063475) and Technology Program for establishing biocide safety
management funded by the Ministry of Environment (No. 1485016231) of the Korean government, and Research
Fund of The Catholic University of Korea (2018).
Conflicts of Interest: The authors declare no conflict of interest. Soo Hyeon Bae is the employee of the company
Q-fitter, Inc. Doyun Kim, Hunseung Yoo are the employees of the company SK Chemicals. The companies had no
role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript,
and in the decision to publish the results.
Previous Presentation of Information: A portion of the information in this manuscript has been presented as a
poster in a conference previously: Soo Hyun Jang, Yu Fen Zheng, Chae Bin Lee, Soo Kyung Bae. 2019. Cytochrome
P450 2C9 inhibition by sophoranone in human liver microsomes. 15th Euro-Global Summit on Toxicology and
Applied Pharmacology, Berlin, Germany.

References
1. Ding, P.L.; He, C.M.; Cheng, Z.H.; Chen, D.F. Flavonoids rather than alkaloids as the diagnostic constituents
to distinguish Sophorae Flavescentis Radix from Sophorae Tonkinensis Radix et Rhizoma: An HPLC fingerprint
study. Chin. J. Nat. Med. 2018, 16, 951–960. [CrossRef]

54
Pharmaceutics 2020, 12, 328

2. He, C.M.; Cheng, Z.H.; Chen, D.F. Qualitative and quantitative analysis of flavonoids in Sophora tonkinensis
by LC/MS and HPLC. Chin. J. Nat. Med. 2013, 11, 690–698. [CrossRef]
3. Yoo, H.; Kang, M.; Pyo, S.; Chae, H.S.; Ryu, K.H.; Kim, J.; Chin, Y.W. SKI3301, a purified herbal extract from
Sophora tonkinensis, inhibited airway inflammation and bronchospasm in allergic asthma animal models
in vivo. J. Ethnopharmacol. 2017, 206, 298–305. [CrossRef]
4. Lee, J.W.; Lee, J.H.; Lee, C.; Jin, Q.; Lee, D.; Kim, Y.; Hong, J.T.; Lee, M.K.; Hwang, B.Y. Inhibitory constituents
of Sophora tonkinensis on nitric oxide production in RAW 264.7 macrophages. Bioorg. Med. Chem. Lett. 2015,
25, 960–962. [CrossRef]
5. Kajimoto, S.; Takanashi, N.; Kajimoto, T.; Xu, M.; Cao, J.; Masuda, Y.; Aiuchi, T.; Nakajo, S.; Ida, Y.; Nakaya, K.
Sophoranone, extracted from a traditional Chinese medicine Shan Dou Gen, induces apoptosis in human
leukemia U937 cells via formation of reactive oxygen species and opening of mitochondrial permeability
transition pores. Int. J. Cancer 2002, 99, 879–890. [CrossRef]
6. Yang, X.; Deng, S.; Huang, M.; Wang, J.; Chen, L.; Xiong, M.; Yang, J.; Zheng, S.; Ma, X.; Zhao, P.; et al.
Chemical constituents from Sophora tonkinensis and their glucose transporter 4 translocation activities. Bioorg.
Med. Chem. Lett. 2017, 27, 1463–1466. [CrossRef] [PubMed]
7. Atta-Ur-Rahman; Haroone, M.S.; Tareen, R.B.; Mohammed Ahmed Hassan, O.M.; Jan, S.; Abbaskhan, A.;
Asif, M.; Gulzar, T.; Al-Majid, A.M.; Yousuf, S.; et al. Secondary metabolites of Sophora mollis subsp. griffithii
(Stocks) Ali. Phytochem. Lett. 2012, 5, 613–616. [CrossRef]
8. Jang, S.M.; Bae, S.H.; Choi, W.K.; Park, J.B.; Kim, D.; Min, J.S.; Yoo, H.; Kang, M.;
Ryu, K.H.; Bae, S.K. Pharmacokinetic properties of trifolirhizin, (-)-maackiain, (-)-sophoranone and
2-(2,4-dihydroxyphenyl)-5,6-methylenedioxybenzofuran after intravenous and oral administration of Sophora
tonkinensis extract in rats. Xenobiotica 2015, 45, 1092–1104. [CrossRef] [PubMed]
9. Rekić, D.; Reynolds, K.S.; Zhao, P.; Zhang, L.; Yoshida, K.; Sachar, M.; Piquette Miller, M.; Huang, S.M.;
Zineh, I. Clinical drug-drug interaction evaluations to inform drug use and enable drug access. J. Pharm. Sci.
2017, 106, 2214–2218. [CrossRef]
10. Bjornsson, T.D.; Callaghan, J.T.; Einolf, H.J.; Fischer, V.; Gas, L.; Grimm, S.; Kao, J.; King, S.P.; Miwa, G.; Ni, L.;
et al. The conduct of in vitro and in vivo drug-drug interaction studies: A pharmaceutical research and
manufactures of America (PhRMA) perspective. Drug Metab. Dispos. 2003, 31, 815–832. [CrossRef]
11. Lin, J.H.; Lu, A.Y. Inhibition and induction of cytochrome P450 and the clinical implications. Clin.
Pharmacokinet. 1998, 35, 361–390. [CrossRef] [PubMed]
12. Peng, Y.; Wu, H.; Zhang, X.; Zhang, F.; Qi, H.; Zhong, Y.; Wang, Y.; Sang, H.; Wang, G.; Sun, J. A comprehensive
assay for nine major cytochrome P450 enzymes activities with 16 probe reactions on human liver microsomes
by a single LC/MS/MS run to support reliable in vitro inhibitory drug-drug interaction evaluation. Xenobiotica
2015, 45, 961–977. [CrossRef]
13. Wienkers, L.C.; Heath, T.G. Predicting in vivo drug interactions from in vitro drug discovery data. Nat. Rev.
Drug Discov. 2005, 4, 825–833. [CrossRef] [PubMed]
14. Chen, Q.; Zhang, T.; Wang, J.F.; Wei, D.Q. Advances in human cytochrome P450 and personalized medicine.
Curr. Drug Metab. 2011, 12, 436–444. [CrossRef] [PubMed]
15. Cai, J.; Ma, J.; Xu, K.; Gao, G.; Xiang, Y.; Lin, C. Effect of Radix Sophorae Tonkinensis on the activity of
cytochrome P450 isoforms in rats. Int. J. Clin. Exp. Med. 2015, 8, 9737–9743.
16. Li, Y.; Ning, J.; Wang, Y.; Wang, C.; Sun, C.; Huo, X.; Yu, Z.; Feng, L.; Zhang, B.; Tian, X.; et al. Drug interaction
study of flavonoids toward CYP3A4 and their quantitative structure activity relationship (QSAR) analysis
for predicting potential effects. Toxicol. Lett. 2018, 294, 27–36. [CrossRef]
17. Cho, D.Y.; Bae, S.H.; Lee, J.K.; Kim, Y.W.; Kim, B.T.; Bae, S.K. Selective inhibition of cytochrome P450 2D6 by
Sarpogrelate and its active metabolite, M-1, in human liver microsomes. Drug Metab. Dispos. 2014, 42, 33–39.
[CrossRef]
18. Zheng, Y.F.; Bae, S.H.; Choi, E.J.; Park, J.B.; Kim, S.O.; Jang, M.J.; Park, G.H.; Shin, W.G.; Oh, E.; Bae, S.K.
Evaluation of the in vitro/in vivo drug interaction potential of BST204, a purified dry extract of ginseng, and its
four bioactive ginsenosides through cytochrome P450 inhibition/induction and UDP-glucuronosyltransferase
inhibition. Food Chem. Toxicol. 2014, 68, 117–127. [CrossRef]
19. Li, G.; Huang, K.; Nikolic, D.; van Breemen, R.B. High-throughput cytochrome P450 cocktail inhibition assay
for assessing drug-drug and drug-botanical interactions. Drug Metab. Dispos. 2015, 43, 1670–1678. [CrossRef]

55
Pharmaceutics 2020, 12, 328

20. Kim, H.J.; Lee, H.; Ji, H.K.; Lee, T.; Liu, K.H. Screening of ten cytochrome P450 enzyme activities with 12
probe substrates in human liver microsomes using cocktail incubation and liquid chromatography-tandem
mass spectrometry. Biopharm. Drug Dispos. 2019, 40, 101–111. [CrossRef]
21. Valicherla, G.R.; Mishra, A.; Lenkalapelly, S.; Jillela, B.; Francis, F.M.; Rajagopalan, L.; Srivastava, P.
Investigation of the inhibition of eight major human cytochrome P450 isozymes by a probe substrate cocktail
in vitro with emphasis on CYP2E1. Xenobiotica 2019, 49, 1396–1402. [CrossRef] [PubMed]
22. Kumar, V.; Wahlstrom, J.L.; Rock, D.A.; Warren, C.J.; Gorman, L.A.; Tracy, T.S. CYP2C9 inhibition: Impact of
probe selection and pharmacogenetics on in vitro inhibition profiles. Drug Metab. Dispos. 2006, 34, 1966–1975.
[CrossRef] [PubMed]
23. Yasar, U.; Tybring, G.; Hidestrand, M.; Oscarson, M.; Ingelman-Sundberg, M.; Dahl, M.L.; Eliasson, E. Role of
CYP2C9 polymorphism in losartan oxidation. Drug Metab. Dispos. 2001, 29, 1051–1056. [PubMed]
24. Elsby, R.; Surry, D.D.; Smith, V.N.; Gray, A.J. Validation and application of Caco-2 assays for the in vitro
evaluation of development candidate drugs as substrates or inhibitors of P-glycoprotein to support regulatory
submissions. Xenobiotica 2008, 38, 1140–1164. [CrossRef] [PubMed]
25. Markowska, M.; Oberle, R.; Juzwin, S.; Hsu, C.P.; Gryszkiewicz, M.; Streeter, A.J. Optimizing Caco-2 cell
monolayers to increase throughput in drug intestinal absorption analysis. J. Pharmacol. Toxicol. Methods 2001,
46, 51–55. [CrossRef]
26. Cho, M.A.; Yoon, J.G.; Kim, V.; Kim, H.; Lee, R.; Lee, M.G.; Kim, D. Functional characterization of
pharmcogenetic variants of human cytochrome P450 2C9 in Korean populations. Biomol. Ther. (Seoul) 2019,
27, 577–583. [CrossRef] [PubMed]
27. Yoo, H.; Ryu, K.H.; Bae, S.K.; Kim, J. Simultaneous determination of trifolirhizin, (-)-maackiain,
(-)-sophoranone, and 2-(2,4-dihydroxyphenyl)-5,6-methylenedioxybenzofuran from Sophora tonkinensis in
rat plasma by liquid chromatography with tandem mass spectrometry and its application to a pharmacokinetic
study. J. Sep. Sci. 2014, 37, 3235–3244. [CrossRef]
28. Bourrié, M.; Meunier, V.; Berger, Y.; Fabre, G. Cytochrome P450 isoform inhibitors as a tool for the investigation
of metabolic reactions catalyzed by human liver microsomes. J. Pharmacol. Exp. Ther. 1996, 277, 321–332.
29. Obach, R.S.; Walsky, R.L.; Venkatakrishnan, K. Mechanism-based inactivation of human cytochrome P450
enzymes and the prediction of drug-drug interactions. Drug Metab. Dispos. 2006, 35, 246–255. [CrossRef]
30. Parkinson, A.; Kazmi, F.; Buckley, D.B.; Yerino, P.; Paris, B.L.; Holsapple, J.; Toren, P.; Otradovec, S.M.;
Ogilvie, B.W. An evaluation of the dilution method for identifying metabolism-dependent inhibitors of
cytochrome P450 enzymes. Drug Metab. Dispos. 2011, 39, 1370–1387. [CrossRef]
31. Stresser, D.M.; Mao, J.; Kenny, J.R.; Jones, B.C.; Grime, K. Exploring concepts of in vitro time-dependent CYP
inhibition assays. Expert Opin. Drug Metab. Toxicol. 2014, 10, 157–174. [CrossRef] [PubMed]
32. Xu, L.; Wang, W. Herbs for clearing heat. In Chinese Materia Medica: Combinations and Applications, 1st ed.;
Donica Publishing Ltd.: St. Albans, UK, 2002; p. 115.
33. US Food and Drug Administration. Guidance for Industry: Estimating the Maximum Safe Starting
Dose in Initial Clinical Trials for Therapeutics in Adult Healthy Volunteer. 2005. Available online:
https://www.fda.gov/media/72309/download (accessed on 31 March 2020).
34. Foti, R.S.; Rock, D.A.; Wienkers, L.C.; Wahlstrom, J.L. Selection of alternative CYP3A4 probe substrates for
clinical drug interaction studies using in vitro data and in vivo simulation. Drug Metab. Dispos. 2010, 38,
981–987. [CrossRef] [PubMed]
35. Van Booven, D.; Marsh, S.; McLeod, H.; Carrillo, M.W.; Sangkuhl, K.; Klein, T.E.; Altman, R.B. Cytochrome
P450 2C9-CYP2C9. Pharmacogenet. Genom. 2010, 20, 277–281. [CrossRef] [PubMed]
36. European Medicine Agency. Guideline on the Investigation of Drug Interactions. 2012.
Available online: https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-investigation-
drug-interactions-revision-1_en.pdf (accessed on 31 March 2020).
37. US Food and Drug Administration. Guidance for industry: In Vitro Drug Interaction Studies-Cytochrome
P450 Enzyme-and Transporter-Mediated Drug Interactions. 2020. Available online: https://www.fda.gov/
media/134582/download (accessed on 31 March 2020).
38. Karjalainen, M.J.; Neuvonen, P.J.; Backman, J.T. Tolfenamic acid is a potent CYP1A2 inhibitor in vitro but
does not interact in vivo: Correction for protein binding is needed for data interpretation. Eur. J. Clin.
Pharmacol. 2007, 63, 829–836. [CrossRef] [PubMed]

56
Pharmaceutics 2020, 12, 328

39. Jaakkola, T.; Backman, J.T.; Neuvonen, M.; Niemi, M.; Neuvonen, P.J. Montelukast and zafirlukast do not
affect the pharmacokinetics of the CYP2C8 substrate pioglitazone. Eur. J. Clin. Pharmacol. 2006, 62, 503–509.
[CrossRef] [PubMed]
40. Kim, K.A.; Park, P.W.; Kim, K.R.; Park, J.Y. Effect of multiple doses of montelukast on the pharmacokinetics
of rosiglitazone, a CYP2C8 substrate, in humans. Br. J. Clin. Pharmacol. 2006, 63, 339–345. [CrossRef]
41. Walsky, R.L.; Obach, R.S.; Gaman, E.A.; Gleeson, J.P.; Proctor, W.R. Selective inhibition of human cytochrome
P4502C8 by montelukast. Drug Metab. Dispos. 2005, 33, 413–418. [CrossRef]
42. Press, B.; di Grandi, D. Permeability for intestinal absorption: Caco-2 assay and related issues. Curr. Drug
Metab. 2008, 9, 893–900. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

57
 pharmaceutics

Article
Strong and Selective Inhibitory Effects of the
Biflavonoid Selamariscina A against CYP2C8
and CYP2C9 Enzyme Activities in Human
Liver Microsomes
So-Young Park 1,2 , Phi-Hung Nguyen 3 , Gahyun Kim 2 , Su-Nyeong Jang 1,2 , Ga-Hyun Lee 1,2 ,
Nguyen Minh Phuc 2,4 , Zhexue Wu 2 and Kwang-Hyeon Liu 1,2, *
1 BK21 Plus KNU Multi-Omics based Creative Drug Research Team, College of Pharmacy and Research
Institute of Pharmaceutical Sciences, Kyungpook National University, Daegu 41566, Korea;
soyoung561021@gmail.com (S.-Y.P.); wts1424@naver.com (S.-N.J.); lgh2710@gmail.com (G.-H.L.)
2 College of Pharmacy and Research Institute of Pharmaceutical Sciences, Kyungpook National University,
Daegu 41566, Korea; hyunlove9137@naver.com (G.K.); phucnguyen0606@gmail.com (N.M.P.);
wuzhexue527@gmail.com (Z.W.)
3 Institute of Natural Products Chemistry, Vietnam Academy of Science and Technology, 18-Hoang Quoc Viet,
Cau Giay, Hanoi 100000, Vietnam; nguyenphihung1002@gmail.com
4 Vietnam Hightech of Medicinal and Pharmaceutical JSC, Group 11 Quang Minh town,
Hanoi 100000, Vietnam
* Correspondence: dstlkh@knu.ac.kr; Tel.: +82-53-950-8567; Fax: +82-53-950-8557

Received: 25 March 2020; Accepted: 9 April 2020; Published: 10 April 2020

Abstract: Like flavonoids, biflavonoids, dimeric flavonoids, and polyphenolic plant secondary metabolites
have antioxidant, antibacterial, antiviral, anti-inflammatory, and anti-cancer properties. However, there is
limited data on their effects on cytochrome P450 (P450) and uridine 5 -diphosphoglucuronosyl transferase
(UGT) enzyme activities. In this study we evaluate the inhibitory potential of five biflavonoids against
nine P450 activities (P450s1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A) in human liver microsomes
(HLMs) using cocktail incubation and liquid chromatography-tandem mass spectrometry (LC–MS/MS).
The most strongly inhibited P450 activity was CYP2C8-mediated amodiaquine N-dealkylation with
IC50 ranges of 0.019~0.123 μM. In addition, the biflavonoids—selamariscina A, amentoflavone,
robustaflavone, cupressuflavone, and taiwaniaflavone—noncompetitively inhibited CYP2C8 activity
with respective Ki values of 0.018, 0.083, 0.084, 0.103, and 0.142 μM. As selamariscina A showed
the strongest effects, we then evaluated it against six UGT isoforms, where it showed weaker
inhibition (UGTs1A1, 1A3, 1A4, 1A6, 1A9, and 2B7, IC50 > 1.7 μM). Returning to the P450 activities,
selamariscina A inhibited CYP2C9-mediated diclofenac hydroxylation and tolbutamide hydroxylation
with respective Ki values of 0.032 and 0.065 μM in a competitive and noncompetitive manner. However,
it only weakly inhibited CYP1A2, CYP2B6, and CYP3A with respective Ki values of 3.1, 7.9, and 4.5 μM.
We conclude that selamariscina A has selective and strong inhibitory effects on the CYP2C8 and
CYP2C9 isoforms. This information might be useful in predicting herb-drug interaction potential
between biflavonoids and co-administered drugs mainly metabolized by CYP2C8 and CYP2C9.
In addition, selamariscina A might be used as a strong CYP2C8 and CYP2C9 inhibitor in P450
reaction-phenotyping studies to identify drug-metabolizing enzymes responsible for the metabolism
of new chemicals.

Keywords: biflavonoid; cytochrome P450; drug interactions; selamariscina A; uridine 5 -diphosph-

oglucuronosyl transferase

59
Pharmaceutics 2020, 12, 343; doi:10.3390/pharmaceutics11070354 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 343

1. Introduction
Flavonoids are polyphenolic secondary metabolites that are common in the plant kingdom and
are ingested by humans in their food [1]. Flavonoids are grouped into various classes based on
structure. These classes are: anthocyanidins, chalcones, flavanones, flavones, flavonols, isoflavonoids,
and biflavonoids [2]. Many pharmacological benefits have been ascribed to flavonoids, including
antioxidant, anti-inflammatory, anti-cancer, antiviral, and hepatoprotective effects [3,4].
Having flavonoids in your diet may reduce the risk of atherosclerosis, cardiovascular disease,
diabetes mellitus, osteoporosis, and certain cancers [4,5]. Because of flavonoids’ benefits and wide
distribution, their intake has risen steadily in recent years in the West and Asia. Daily intake of
flavonoids has been estimated at 100 mg/day in the Asian population because of the high consumption
of soy products [6,7]. On the other hand, daily intake of flavonoids has been estimated to be in the
range of 20–50 mg/day in Western populations [8]. Further intake of flavonoids through dietary
supplements and plant extracts with prescribed drugs is common. The vast body of literature describes
the significant interactions between flavonoid herbs and therapeutic drugs.
Several flavonoids are substrates for cytochrome P450 (P450) and uridine 5 -diphosphoglucuronosyl
transferase (UGT) enzymes [2], suggesting that flavonoids could inhibit the activities of these enzymes.
A number of studies have demonstrated that flavonoids are potent inhibitors of CYP1A2, CYP3A,
and UGT1A1 in vitro [5,8]. For example, the flavone tangeretin competitively inhibits the activity
of CYP1A2 with a Ki value as low as 68 nM in human liver microsomes (HLMs) [9]. It also
inhibits UGT1A1-mediated estradiol glucuronidation with an IC50 value of 1 μM [10]. The flavonols
quercetin and kaempferol inhibit the metabolism of nifedipine and felodipine by CYP3A4 in HLMs
at concentrations larger than 10 μM. [11]. Animal studies show that oral quercetin increases the
bioavailability of oral doxorubicin [12]. These results can be attributed to the reduced first-pass
metabolism of doxorubicin due to quercetin-induced inhibition of CYP3A and/or enhanced doxorubicin
absorption in the gastrointestinal tract via quercetin-induced inhibition of P-glycoprotein (P-gp). Surya
Sandeep et al. (2014) reported that naringenin significantly increases the bioavailability of orally
administered felodipine, a P-glycoprotein and CYP3A4 substrate drug, in rats, through the inhibition
of intestinal P-gp and CYP3A4 [13]. Alnaqeeb et al. (2019) reported that quercetin and guava leaf
extracts in combination with warfarin exert a greater increase on warfarin’s Cmax and International
Normalized Ratio values than when used alone, indicating the inhibition of CYP2C8, 2C9 and 3A4,
major warfarin-metabolizing enzymes [14]. Biflavonoids, formed by the covalent bond between two
monoflavonoids, are a subclass of flavonoid [15]. They are secondary metabolites, but are limited to
several species in plants such as Ginkgo biloba, Selaginella species, Hypericum perforatum, and Garcinia
kola [16]. Befitting their status as flavonoids, they have anti-cancer, anti-microbial, antiviral, and
anti-inflammatory properties [16]. In contrast to the extensive studies on drug interaction with
flavonoids, data on the inhibitory effects of biflavonoids on P450 and UGT enzymes are rare, though
biflavonoids are taken in the form of dietary supplements (e.g., Ginkgo biloba extract [17]). The inhibitory
potential of amentoflavone, the major biflavonoid in Cupressus funebris, against P450 and UGT enzymes
was only recently reported [18,19].
In this study, we evaluate the inhibitory effects of five biflavonoids—selamariscina A, amentoflavone,
robustaflavone, cupressuflavone, and taiwaniaflavone (Figure 1)—on nine P450 enzymes using HLMs.
We further investigate the ability of selamariscina A, which most strongly inhibited CYP2C8 and
CYP2C9 activities, to inhibit six UGT isoforms. Furthermore, the inhibition mechanism and kinetic
parameters (Ki ) were determined for selamariscina A and compared with those of montelukast,
a well-known selective CYP2C8 inhibitor [20].

60
Pharmaceutics 2020, 12, 343

Figure 1. Chemical structures of biflavonoids from Selaginella tamariscina: selamariscina A (A),

amentoflavone (B), robustaflavone (C), cupressuflavone (D) and taiwaniaflavone (E).

2. Materials and Methods

2.1. Chemicals and Reagents

We purchased acetaminophen, alamethicin, amodiaquine, bupropion, chenodeoxycholic acid,
chlorzoxazone, dextromethorphan, estrone glucuronide, glucose-6-phosphate (G6P), glucose- 6-phosphate
dehydrogenase (G6PDH), hydroxybupropion, magnesium chloride (MgCl2), N-acetylserotonin,
β-nicotinamide adenine dinucleotide phosphate (NADP+), N-desethylamodiaquine, omeprazole,
phenacetin, trifluoperazine, trimipramine, and uridine 5 -diphosphoglucuronic acid (UDPGA)
from Sigma-Aldrich (St. Louis, MO, USA). 4-Hydroxydiclofenac, 4-hydroxytolbutamide, 5-hydr-
oxyrosiglitazone, coumarin, diclofenac, midazolam, montelukast, mycophenolic acid, rosiglitazone,
and tolbutamide came from Toronto Research Chemicals (Toronto, ON, Canada). We obtained
1 -hydroxymidazolam from Cayman Chemical (Ann Arbor, MI, USA), while 7-ethyl-10-hydrox-
ycomptothecine (SN-38) was provided by Santa Cruz Biotechnology (Dallas, TX, USA). All solvents
were LC–MS grade (Fisher Scientific, Pittsburgh, PA, USA). All the other reagents were of analytical or
LC–MS grade and are commercially available. We purchased the pooled human liver microsomes
(XTreme 200) from XenoTech (Lenexa, Kansas City, KS, USA). In this study, we used selamariscina
A, amentoflavone, robustaflavone, cupressuflavone, and taiwaniaflavone identified from Selaginella
tamariscina (Beauv.), which were collected at Yen Tu Mountain, Uong Bi town, Quang Nihn province,
Vietnam. The information regarding the identification of their chemical structures was described in
our previously published paper [21,22].
We isolated selamariscina A, amentoflavone, robustaflavone, cupressuflavone, and taiwaniaflavone
from Selaginella tamariscina (Beauv.), which were collected at Yen Tu Mountain, Uong Bi town, Quang
Nihn province, Vietnam. The five compounds were purified and examined by HPLC to get 95% purity.
Their chemical structures were identified by analyzing their NMR data, which were in good agreement
with those published in a previous report [21,22].

2.2. Inhibitory Effect of Five Biflavonoids against Human Cytochrome P450 Activity
The inhibitory potential of the five biflavonoids on the metabolism of nine P450 probe substrates
was evaluated using previously developed methods with minor modifications [23,24]. Biflavonoids
were dissolved in methanol. The final concentration of methanol in the incubation mixture was
1.0% (v/v). We used these P450 probe substrates: phenacetin for CYP1A2, coumarin for CYP2A6,
bupropion for CYP2B6, amodiaquine for CYP2C8, diclofenac for CYP2C9, omeprazole for CYP2C19,
dextromethorphan for CYP2D6, chlorzoxazone for CYP2E1 and midazolam for CYP3A (Table 1).
The incubation mixtures containing pooled human liver microsomes (HLMs, XTreme 200, XenoTech),

61
Pharmaceutics 2020, 12, 343

P450 probe substrates, and inhibitor (0~20 μM) were pre-incubated at 37 ◦ C for 5 min. The concentration
range of the inhibitor varied (0, 0.002, 0.005, 0.02, 0.05, and 0.2 μM for CYP2C8; 0, 0.02, 0.05, 0.2, 0.5,
and 2 μM for CYP2C9; 0, 0.5, 2, 5, 10, and 20 μM for other P450 isoforms). After pre-incubation,
a reduced nicotinamide adenine dinucleotide phosphate (NADPH) generation system containing
1 unit/ml G6PDH, 1.3 mM β- nicotinamide adenine dinucleotide phosphate (β- NADP+ ), 3.3 mM
MgCl2 , and 3.3 mM G6P was added to initiate a reaction, and further incubated for 10 min at 37 ◦ C.
The reaction was stopped by adding 50 μL of ice-cold acetonitrile containing 7 nM trimipramine
(internal standard, IS). After centrifugation at 18,000 g (5 min, 4 ◦ C), aliquots of supernatants were
analyzed by LC–MS/MS (Shimadzu LCMS 8060 system, Shimadzu, Kyoto, Japan). All microsomal
incubations were conducted in triplicate.

Table 1. Selected reaction monitoring (SRM) condition for the major metabolites of the nine cytochrome
P450 probe substrates and internal standard (IS).

P450 Concentration SRM Transition Collision Energy

Substrates Metabolites Polarity
Enzyme (μM) (m/z) (eV)
1A2 Phenacetin 100 Acetaminophen 152 > 110 ESI+ 25
2A6 Coumarin 5 7-Hydroxycoumarin 163 > 107 ESI+ 17
2B6 Bupropion 50 6-Hydroxybupropion 256 > 238 ESI+ 20
Amodiaquine 1 N-Desethylamodiaquine 328 > 283 ESI+ 17
2C8
Rosiglitazone 5 p-Hydroxyrosiglitazone 374 > 151 ESI+ 17
Tolbutamide 100 4-Hydroxytolbutamide 287 > 89 ESI+ 60
2C9
Diclofenac 10 4-Hydroxydiclofenac 312 > 231 ESI+ 15
2C19 Omeprazole 20 5-Hydroxyomeprazole 362 > 214 ESI+ 10
2D6 Dextromethorphan 5 Dextrorphan 258 > 157 ESI+ 35
2E1 Chlorzoxazone 50 6-Hydroxychlorzoxazone 184 > 120 ESI− 18
3A Midazolam 5 
1 -Hydroxymidazolam 342 > 203 ESI+ 25
IS Trimipramine 0.007 295 > 100 ESI+ 17
ESI: Electrospray ionization (ESI) interface to generate protonated ions [M+H]+ or deprotonated ion [M−H]−.

2.3. Kinetic Characterization of Five Biflavonoids on CYP2C8 in Human Liver Microsomes

To determine the inhibition mechanism and constants (Ki values) of the five biflavonoids against
CYP2C8 activity, different concentrations of biflavonoids (0, 0.002, 0.005, 0.02, 0.05, and 0.2 μM for
selamariscina A; 0, 0.05, 0.02, 0.05, 0.2 and 0.5 μM for the other four biflavonoids) were added to
reaction mixtures containing different concentrations of amodiaquine (0.1, 0.4 and 1 μM). The other
conditions were the same as in the cytochrome P450 inhibition study.

2.4. Kinetic Characterization of Selamariscina A on Five P450 Enzymes in Human Liver Microsomes
We used HLMs to determine the mechanisms and constants (Ki values) for selamariscina A
inhibition of CYP1A2, CYP2B6, CYP2C8, CYP2C9 and CYP3A. The selamariscina A (0~50 μM) was
added into the reaction mixtures, each of which contained concentrations of phenacetin (20, 50,
and 100 μM), bupropion (20, 50, and 100 μM), amodiaquine (0.1, 0.4, and 1 μM), rosiglitazone (2, 5,
and 10 μM), diclofenac (1, 4, and 10 μM), tolbutamide (50, 100, and 200 μM), and midazolam (0.5, 2,
and 5 μM). The substrates were used at concentrations approximately near to their respective Km
values [25–27]. The concentration range of selamariscina A varied (0, 0.002, 0.005, 0.02, 0.05, and 0.2 μM
for CYP2C8; 0, 0.05, 0.02, 0.05, 0.2, and 0.5 μM for CYP2C9; 0, 0.2, 0.5, 2, 5, and 20 μM for CYP3A; 0, 0.5,
2, 5, 20, and 50 μM for CYP1A2 and CYP2B6). The other conditions were the same as in the cytochrome
P450 inhibition study.

2.5. Time-Dependent Inhibition Assay

The time-dependent inhibition of selamariscina A against CYP2C8 and CYP2C9 enzymes was
evaluated using an IC50 shift method. Selamariscina A was pre-incubated at six concentrations (0, 0.002,
0.005, 0.02, 0.05, and 0.2 μM) with HLMs in the presence of an NDAPH generation system for 30 min

62
Pharmaceutics 2020, 12, 343

at 37 ◦ C. The reaction was initiated by adding 1 μM amodiaquine or 10 μM diclofenac and further

incubated for 10 min. Incubation was terminated by adding 50 μL of ice-cold acetonitrile containing
7 nM trimipramine. After centrifugation, aliquots of supernatants were analyzed by LC–MS/MS.

2.6. Inhibitory Effect of Selamariscina A against Human UGT Activity

The ability of selamariscina A to inhibit the metabolism of six UGT enzyme probe substrates
was evaluated using previously developed methods with minor modifications [28]. The microsomal
incubation was performed by dividing the non-interactive substrate cocktail sets (set A included SN-38
for UGT1A1, CDCA for UGT1A3 and TFP for UGT1A4 while set B included N-SER for UGT1A6, MPA
for UGT1A9 and NX for UGT2B7) (Table 2). In brief, HLMs (0.25 mg/mL) were activated by incubation
in the presence of alamethicin (25 μg/mL) for 15 min on ice. After the addition of UGT probe substrates
and inhibitor (0, 0.5, 2, 5, 20 and 50 μM), the incubation mixtures were pre-incubated at 37 ◦ C for 5 min.
After pre-incubation, 5 mM UDPGA was added to initiate a reaction, and further incubated for 60 min
at 37 ◦ C. The reaction was stopped by adding 50 μL of ice-cold acetonitrile containing 250 nM estrone
glucuronide (IS). After centrifugation at 18,000 g (5 min, 4 ◦ C), aliquots of supernatants were analyzed
by LC–MS/MS. All microsomal incubations were conducted in triplicate.

Table 2. Selected reaction monitoring (SRM) condition for the major metabolites of the six uridine
5 -diphosphoglucuronosyl transferase (UGT) enzyme substrates and internal standard (IS).

Collision
UGT Concentration SRM
Substrates Metabolites Polarity Energy
Enzyme (μM) Transition (m/z)
(eV)
1A1 SN-38 0.5 SN-38 glucuronide 569 > 393 ESI+ 30
1A3 Chenodeoxycholic acid (CDCA) 2 CDCA-24 glucuronide 567 > 391 ESI− 20
1A4 Trifluoperazine (TFP) 0.5 TFP N-glucuronide 584 > 408 ESI+ 30
1A6 N-Acetylserotonin (N-SER) 1 N-SER glucuronide 395 > 219 ESI+ 10
1A9 Mycophenolic acid (MPA) 0.2 MPA 7-O-glucuronide 495 > 319 ESI− 25
2B7 Naloxone (NX) 0.2 NX 3-glucuronide 504 > 310 ESI+ 30
IS Estrone glucuronide 0.25 445 > 269 ESI− 35

2.7. LC–MS/MS Analysis

All metabolites and the IS were separated on a Kinetex XB-C18 column (100 × 2.10 mm, 2.6 μm,
100 Å; Phenomenex, Torrance, CA, USA) and analyzed using a Shimadzu LCMS 8060 triple-quadrupole
mass spectrometer coupled with a Nexera X2 ultra high-performance liquid chromatography system
(Shimadzu, Kyoto, Japan) equipped with an electrospray ionization interface. The mobile phase
consisted of 0.1% formic acid in water (A) and 0.1% formic acid in acetonitrile (B). The elution condition
was set as 8% B (0–0.5 min), 8%→60% B (0.5–5 min), 60% B (5–6 min), 60%→8% B (6–6.1 min) and 8% B
(6.1–9 min) for the analysis of metabolites of P450 probe substrates and set as 0%→40% B (0–1 min),
40%→50% B (1–5 min), 50%→0% B (5–5.1 min), and 0% B (5.1–8 min) for the analysis of metabolites
of UGT probe substrates. The flow rate was 0.2 mL/min. Electrospray ionization was performed in
positive-ion mode at 4000 V or in negative-ion mode at −3500 V. The optimum operating conditions
were determined as follows: vaporizer temperature, 300 ◦ C; capillary temperature, 350 ◦ C; collision
gas (argon) pressure, 1.5 mTorr. Quantitation was conducted in selected reaction monitoring (SRM)
modes with the precursor-to-product ion transition for each metabolite (Tables 1 and 2).

2.8. Data Analysis

We analyzed the data with Shimadzu LabSolution LC–MS software. The IC50 values were
calculated by WinNonlin software (Pharsight, Mountain View, CA, USA). The type of inhibition and
the apparent kinetic parameters for inhibitory activity (Ki) were determined by following several
criteria: visual inspection of Dixon plots, Lineweaver–Burk double reciprocal plots, and secondary
plots of Lineweaver–Burk plots versus biflavonoid concentrations, the size of the sum of squares of the
residuals, Akaike Information Criteria values, the S.E. and 95% confidence interval of the parameter

63
Pharmaceutics 2020, 12, 343

estimates from the nonlinear regression analysis [29] using the WinNonlin software. The models tested
included competitive, competitive partial, noncompetitive, noncompetitive partial, uncompetitive,
uncompetitive partial, and mixed-type inhibition.

3. Results and Discussion

3.1. Inhibition of Cytochrome P450 Enzymes Activities by Five Biflaovnoids

The inhibitory potential of the five biflavonoids against cytochrome P450 enzyme activity
was evaluated using HLMs (Table 3). We found selamariscina A, amentoflavone, robustaflavone,
cupressuflavone, and taiwaniaflavone strongly inhibit CYP2C8 activity with respective IC50 values of
0.019, 0.084, 0.083, 0.083, and 0.12 μM. They also show strong inhibition on CYP2C9 activity with IC50
values of 0.047, 0.15, 0.15, 0.21, and 0.20 μM. Their inhibition of the other seven P450 isoforms was
much lower (IC50 ≥ 1.2 μM) than on CYP2C8 and CYP2C9 (IC50 ≤ 0.21 μM). The IC50 value of the
inhibition of diclofenac hydroxylase activity by amentoflavone that we found (0.15 μM) is 4.3 times
higher than the 0.035 μM reported by von Moltke et al. (2004) [19]. The reason could be differences in
incubation conditions, such as CYP2C19 probe substrate and concentrations (diclofenac 10 μM versus
S-mephenytoin 25 μM) [30]. The inhibitory potential (IC50 = 1.3 μM) of amentoflavone on CYP3A was
similar to the previously reported value (IC50 = 4.3 μM) [19].

Table 3. Inhibitory effects of five biflavonoids and montelukast against nine cytochrome P450 isoforms.

P450 IC50 (μM)

Enzyme Substrate
Selamaris-Cina A Amento-Flavone Robusta-Flavone Cupressu-Flavone Taiwania-Flavone Montelukast
1A2 Phenacetin 7.4 4.4 4.5 5.9 6.8 >50
2A6 Coumarin 11.6 11.9 11.8 >20 10.6 >50
2B6 Bupropion 5.3 7.1 5.7 6.7 6.4 >50
2C8 Amodiaquine 0.019 0.084 0.083 0.083 0.12 0.52
2C9 Diclofenac 0.047 0.15 0.15 0.21 0.20 9.73
2C19 Omeprazole 13.3 3.4 6.4 3.0 5.0 >50
2D6 Dextromethorphan 10.6 2.6 2.2 2.7 3.2 >50
2E1 Chlorzoxazone >20 3.3 2.9 2.3 6.0 >50
3A Midazolam 2.7 1.3 1.2 1.5 1.2 >50

As the five flavonoids strongly inhibited microsomal CYP2C8 activity, we sought to clarify the
mechanism of inhibition. The Lineweaver–Burk plots, Dixon plots and secondary reciprocal plots
indicated that selamariscina A, amentoflavone, robustaflavone, cupressuflavone, and taiwaniaflavone
noncompetitively inhibited CYP2C8-mediated amodiaquine N-deethylation activity with Ki values
of 0.018, 0.083, 0.084, 0.103, and 0.142 μM, respectively (Table 4), which are lower than those of the
well-known CYP2C8 inhibitors zafirlukast (0.39 μM) [31] and quercetin (4.72 μM) [32].

Table 4. Ki values for inhibition of CYP2C8-catalyzed amodiaquine N-deethylation in human liver

microsomes by five biflavonoids.

P450 Mode of
Substrate Inhibitor Ki (μM) a
Enzyme Inhibition
Selamariscina A 0.018 ± 0.002 Noncompetitive
Amentoflavone 0.083 ± 0.009 Noncompetitive
CYP2C8 Amodiaquine Robustaflavone 0.084 ± 0.016 Noncompetitive
Cupressuflavone 0.103 ± 0.017 Noncompetitive
Taiwaniaflavone 0.142 ± 0.026 Noncompetitive
a Values represent the average ± standard error in triplicate.

64
Pharmaceutics 2020, 12, 343

Of the five biflavonoids, selamariscina A most strongly inhibited CYP2C8-mediated amodiaquine

N-deethylation activity with a Ki value of 0.018 μM, which is similar to the IC50 value of the
known strong CYP2C8 inhibitor montelukast (0.020 μM) [31]. Further, its inhibitory potential against
CYP2C8 was much stronger than other known CYP2C8 inhibitors axitinib (Ki = 0.17 μM [33]),
clotrimazole (IC50 = 0.78 μM [31]), felodipine (IC50 = 1.20 μM [31]), nilotinib (Ki = 0.10 μM [33]),
and quercetin (Ki = 1.56 μM [34]). We further evaluated the inhibition mechanism of selamariscina
A, which showed the strongest CYP2C8 inhibition, for the other P450 enzymes. The inhibitory
potential (Ki) of selamariscina A against P450 enzyme activities was in the order: CYP2C8 > CYP2C9
> CYP1A2 > CYP3A > CYP2B6 (Table 5, Figure 2 and Figure S1). To determine whether inhibition
by selamariscina A was substrate-specific, we evaluated its inhibitory effects on CYP2C8-mediated
rosiglitazone 5-hydroxylation. We found that it showed strong inhibition with a Ki value of 0.010 μM
in a substrate-independent manner. Selamariscina A also inhibited CYP2C9-mediated diclofenac
and tolbutamide hydroxylation with Ki values of 0.032 and 0.065 μM, respectively, in a substrate-
independent manner. Its inhibitory potential against CYP2C9 was much stronger than other known
CYP2C8 inhibitors sulfaphenazole (Ki = 0.12–0.7 μM [35]), fluvoxamine (Ki = 0.63–16 μM [35]),
fluconazole (Ki = 0.28 μM [36]), and fluoxetine (Ki = 19–87 μM [35]). Its inhibitory potential for CYP2C8
and CYP2C9 was much stronger than other P450s.

Table 5. Ki values for the inhibition of CYP1A2-catalyzed phenacetin O-deethylation, CYP2B6-catalyzed

bupropion hydroxylation, CYP2C8-catalyzed amodiaquine N-deethylation, CYP2C8-catalyzed
rosiglitazone 5-hydroxylation, CYP2C9-catalyzed diclofenac 4-hydroxylation, CYP2C9-catalyzed
tolbutamide 4-hydroxylation, and CYP3A-catalyzed midazolam 1 -hydroxylation in human liver
microsomes by selamariscina A.

P450
Substrate Ki (μM) a Mode of Inhibition
Enzyme
1A2 Phenacetin 3.1 ± 0.6 Competitive
2B6 Bupropion 7.9 ± 1.1 Noncompetitive
Amodiaquine 0.018 ± 0.002 Noncompetitive
2C8
Rosiglitazone 0.010 ± 0.003 Noncompetitive, partial
Diclofenac 0.032 ± 0.007 Competitive
2C9
Tolbutamide 0.065 ± 0.01 Noncompetitive
3A Midazolam 4.5 ± 0.5 Noncompetitive
a Values represent the average ± standard error in triplicate.

In addition, several P450 inhibitors including azamulin, clopidogrel, methoxalene, and ticlopidine
[37–39] have been shown to be time-dependent inhibitors of cytochrome P450. We investigated the effect
of incubation time on IC50 values of selamariscina A using the CYP2C8 substrate amodiaquine and the
CYP2C9 substrate diclofenac. The inhibitory potential of selamariscina A against CYP2C8-mediated
amodiaquine O-deethylase activity and CYP2C9-mediated diclofenac hydroxylase activity in HLMs
pre-incubated in the presence of an NADPH generation system (IC50 values of 0.031 and 0.092 μM,
respectively) was a bit weaker than in untreated HLMs (IC50 values of 0.019 and 0.054 μM, respectively).
This suggests that selamariscina A is not a time-dependent inhibitor (data are not shown).

65
Pharmaceutics 2020, 12, 343

Figure 2. Representative Dixon plots obtained from a kinetic study of CYP1A2-catalyzed phenacetin
O-deethylation (A), CYP2B6-catalyzed bupropion hydroxylation (B), CYP2C8-catalyzed amodiaquine
N-deethylation (C), CYP2C8-catalyzed rosiglitazone 5-hydroxylation (D), CYP2C9-catalyzed diclofenac
4-hydroxylation (E), and CYP3A-catalyzed midazolam 1 -hydroxylation (F) in the presence of different
concentrations of selamariscina A in pooled human liver microsomes (XTreme 200, XenoTech). Each data
point shown represent the mean ± standard error in triplicate for the samples.

3.2. Inhibition of UGT Enzymes Activities by Selamariscina A

The inhibitory potential of selamariscina A against uridine 5 -diphosphoglucuronosyl transferase
(UGT) activity was evaluated using HLMs (Table 6). Selamariscina A inhibited UGT1A1 and UGT1A4

66
Pharmaceutics 2020, 12, 343

activity with IC50 values of 1.7 and 7.7 μM, respectively. However, its inhibition of UGT1A1 and
UGT1A4 isoforms was much weaker than that of CYP2C8 (IC50 = 0.019 μM). The inhibitory potential
of selamariscina A for UGT1A3, UGT1A6, UGT1A9, and UGT2B6 was negligible (IC50 > 40 μM).
The IC50 value of the inhibition of UGT1A1 activity by amentoflavone found in our study (1.7 μM) is
similar to its previously reported value (IC50 = 0.78 μM) [18].

Table 6. Inhibitory effects of selamariscina A against six uridine 5 -diphosphoglucuronosyl transferase

(UGT) isoforms.

UGT
Substrate IC50 (μM) a
Enzyme
1A1 SN-38 * 1.7 ± 0.5
1A3 Chenodeoxycholic acid >50
1A4 Trifluoperazine 7.7 ± 1.9
1A6 N-Acetylserotonin 46.1 ± 11.7
1A9 Mycophenolic acid 40.4 ± 11.1
2B7 Naloxone >50
* SN-38: 7-Ethyl-10-hydroxy camptothecin; a values represent the average ± standard error in triplicate.

3.3. Comparison of the Selectivity of Selamariscina A and Montelukast for CYP2C8 Inhibition
Montelukast has been used to inhibit CYP2C8 in reaction-phenotyping studies [20].
We re-evaluated its inhibitory potential against the nine P450 isoforms in this study using HLMs
(XTreme 200, XenoTech). Montelukast strongly inhibited CYP2C8 activity with an IC50 value
of 0.52 μM, but it showed weak inhibition on the other eight P450 enzymes (IC50 > 9.73 μM)
(Table 3). The IC50 value for the CYP2C8 isoform (IC50 = 0.52 μM at 0.25 mg/mL microsomal protein
concentration) was similar to previously reported values (IC50 = 0.18 μM at 0.3 mg/mL microsomal
protein concentration) [20]. However, montelukast showed more than 25 times weaker inhibition
than selamariscina A (IC50 = 0.019 μM at 0.25 mg/mL microsomal protein concentration). At 0.5 μM
selamariscina A concentration, approximately 25 times greater than the Ki value, selamariscina A was
found to inhibit CYP2C8 and CYP2C9 by 92.8% and 88.6% respectively, and only slightly affected
the enzyme activities of the other P450 isoforms (Figure 3). Selamariscina A at 0.5 μM concentration
inhibited none of the other P450 isoform-specific activities above 21.8% in HLMs, indicating that
selamariscina A could be used as a selective CYP2C8 and CYP2C9 inhibitor in P450 phenotyping
studies. Montelukast at 0.5 μM concentration, a well-known selective CYP2C8 inhibitor [20], showed
moderate inhibition on CYP2C8 by 52.7% in pooled HLMs. At 5 μM concentration, montelukast
inhibited CYP2C8 by 86.1% in HLMs; however, it also inhibited CYP2C9 and CYP2B6 activities by
31.0% and 20.4%, respectively in pooled HLMs. Montelukast (5 μM) showed negligible inhibition on
the other six P450 isoforms. Selamariscina A could be useful as a strong CYP2C8 and CYP2C9 inhibitor
in P450 reaction-phenotyping studies.

67
Pharmaceutics 2020, 12, 343

Figure 3. Inhibitory effects of selamariscina A (0.5 μM, ) and montelukast (0.5 μM, ; 5 μM, ) on the
enzymatic activities of nine P450 isoforms in pooled human liver microsomes (0.25 mg/mL, XTreme 200,
XenoTech). Phenacetin (100 μM), coumarin (5 μM), bupropion (50 μM), amodiaquine (1 μM), diclofenac
(10 μM), omeprazole (20 μM), dextromethorphan (5 μM), chlorzoxazone (50 μM), and midazolam
(5 μM) were used as the respective substrates of P450s 1A2, 2A6, 2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A.
The data are means of the average ± standard error in triplicate.

3.4. Evaluation of Drug Interaction Potential of Selamariscina A

It was estimated that an in vivo interaction potential via the inhibition of P450 would likely occur
if the ratio of inhibitor Cmax /Ki exceeded one and would be possible if it was between 0.1 and 1.0 [40,41].
Based on amentoflavone’s maximum concentrations (0.041 and 0.063 μM) in rat blood after a single
oral dose of Selaginella doderleinii Hieron extracts (200 mg/kg; contents: 103.82 mg/g amentoflavone,
37.52 mg/g robustaflavone, 44.4 mg/g 2,”3 -dihydro-3 ,3”-biapigenin, 53.4 mg/g 3 ,3”-binaringenin, and
35.1 mg/g delicaflavone) [42] and Selaginella doderleinii Hieron extracts (600 mg/kg) [43], the respective
values of Cmax /Ki were 0.49 and 0.76 from the data of pooled HLMs (Ki = 0.083 μM), indicating that
amentoflavone has possible drug interaction potential with CYP2C8 substrate drugs [44]. Recently,
nanotechnology-based delivery systems such as liposomes have been developed for improving
oral bioavailability [42]. The values of Cmax (0.22 μM) of amentoflavone after administration of
liposome-based Selaginella doderleinii Hieron extracts (200 mg/kg) were 5.4 times higher than those of
the control [42], resulting in a Cmax /Ki value of 2.65, indicating that amentoflavone has drug interaction
potential. In the case of selamariscina A, the present study provides the first published data on its
pharmacokinetics in animals and humans. Therefore, it is difficult to estimate the drug interaction
potential of selamariscina A for humans. However, selamariscina A might have drug interactions with
CYP2C8 substrate drugs such as cerivastatin [45], paclitaxel [46], and rosiglitazone [47] because its
CYP2C8 inhibitory potential was more than 4.5 times stronger than that of amentoflavone. Therefore,
in vivo studies are necessary to determine whether drug interactions between selamariscina A and
CYP2C8 or CYP2C9 substrates have clinical relevance.

4. Conclusions
In conclusion, we report that selamariscina A is a strong CYP2C8 and CYP2C9 inhibitor.
When evaluated for amodiaquine O-deethylation and diclofenac hydroxylation inhibitory activity
against CYP2C8 and CYP2C9, as well as seven other P450s, it exhibited above 50-fold selectivity.
Like montelukast and sulfaphenazole, selamariscina A could be useful as a strong CYP2C8 and CYP2C9
inhibitor in P450 phenotyping studies when HLMs are the enzyme source. Additionally, selamariscina
A might cause clinically relevant pharmacokinetic drug interactions with other co-administered drugs

68
Pharmaceutics 2020, 12, 343

metabolized by CYP2C8 or CYP2C9. These in vitro findings provide primary data for future in vivo
animal and clinical studies on risk prediction related to the interaction of drugs with herbs.

Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4923/12/4/343/s1,

Figure S1: Representative Lineweaver–Burk plots obtained from a kinetic study of CYP1A2-catalyzed
phenacetin O-deethylation (A), CYP2B6-catalyzed bupropion hydroxylation (B), CYP2C8-catalyzed amodiaquine
N-deethylation (C), CYP2C8-catalyzed rosiglitazone 5-hydroxylation (D), CYP2C9-catalyzed diclofenac
4-hydroxylation (E), and CYP3A-catalyzed midazolam 1 -hydroxylation (F) in the presence of different
concentrations of selamariscina A. Each data point shown represent the mean ± standard error in triplicate samples.
Author Contributions: S.-Y.P. and K.-H.L. conceived and designed the experiments; S.-Y.P., P.-H.N., G.K., S.-N.J.,
G.-H.L., and N.M.P. performed experiments; S.-Y.P., Z.W., and K.-H.L. analyzed the data; S.-Y.P. and K.-H.L. wrote
the paper. All authors have read and agreed to the published version of the manuscript.
Funding: This study was supported by the National Research Foundations of Korea, Ministry of Science and
ICT, Republic of Korea [NRF-2019R1A2C1008713], a Korea Basic Science Institute National Research Facilities &
Equipment Center grant funded by the Korea government (Ministry of Education) [2019R1A6C1010001], and the
National Foundation for Science and Technology Development of Vietnam, Ministry of Science and Technology
[NAFOSTED-104.01-2017.50].
Conflicts of Interest: The authors declare no conflict of interest. N.M.P. is from Vietnam Hightech of Medicinal
and Pharmaceutical JSC, the company lease state had no role in the design of the study; in the collection, analyses,
or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.

References
1. Bravo, L. Polyphenols: Chemistry, dietary sources, metabolism, and nutritional significance. Nutr. Rev. 1998,
56, 317–333. [CrossRef] [PubMed]
2. Cermak, R.; Wolffram, S. The potential of flavonoids to influence drug metabolism and pharmacokinetics by
local gastrointestinal mechanisms. Curr. Drug Metab. 2006, 7, 729–744. [CrossRef] [PubMed]
3. Kumar, S.; Pandey, A.K. Chemistry and biological activities of flavonoids: An overview. Sci. World J. 2013,
2013, 162750. [CrossRef]
4. Miron, A.; Aprotosoaie, A.C.; Trifan, A.; Xiao, J. Flavonoids as modulators of metabolic enzymes and drug
transporters. Ann. N. Y. Acad. Sci. 2017, 1398, 152–167. [CrossRef] [PubMed]
5. Kopecna-Zapletalova, M.; Krasulova, K.; Anzenbacher, P.; Hodek, P.; Anzenbacherova, E. Interaction
of isoflavonoids with human liver microsomal cytochromes P450: Inhibition of CYP enzyme activities.
Xenobiotica 2017, 47, 324–331. [CrossRef] [PubMed]
6. Manach, C.; Scalbert, A.; Morand, C.; Remesy, C.; Jimenez, L. Polyphenols: Food sources and bioavailability.
Am. J. Clin. Nutr. 2004, 79, 727–747. [CrossRef] [PubMed]
7. Tikkanen, M.J.; Adlercreutz, H. Dietary soy-derived isoflavone phytoestrogens. Could they have a role in
coronary heart disease prevention? Biochem. Pharm. 2000, 60, 1–5. [CrossRef]
8. Cermak, R. Effect of dietary flavonoids on pathways involved in drug metabolism. Expert Opin. Drug Metab.
Toxicol. 2008, 4, 17–35. [CrossRef]
9. Obermeier, M.T.; White, R.E.; Yang, C.S. Effects of bioflavonoids on hepatic P450 activities. Xenobiotica 1995,
25, 575–584. [CrossRef]
10. Williams, J.A.; Ring, B.J.; Cantrell, V.E.; Campanale, K.; Jones, D.R.; Hall, S.D.; Wrighton, S.A. Differential
modulation of UDP-glucuronosyltransferase 1A1 (UGT1A1)-catalyzed estradiol-3-glucuronidation by the
addition of UGT1A1 substrates and other compounds to human liver microsomes. Drug Metab. Dispos. 2002,
30, 1266–1273. [CrossRef]
11. Miniscalco, A.; Lundahl, J.; Regardh, C.G.; Edgar, B.; Eriksson, U.G. Inhibition of dihydropyridine metabolism
in rat and human liver microsomes by flavonoids found in grapefruit juice. J. Pharmacol. Exp. Ther. 1992, 261,
1195–1199. [PubMed]
12. Choi, J.S.; Piao, Y.J.; Kang, K.W. Effects of quercetin on the bioavailability of doxorubicin in rats: Role of
CYP3A4 and P-gp inhibition by quercetin. Arch. Pharm. Res. 2011, 34, 607–613. [CrossRef] [PubMed]
13. Surya Sandeep, M.; Sridhar, V.; Puneeth, Y.; Ravindra Babu, P.; Naveen Babu, K. Enhanced oral bioavailability
of felodipine by naringenin in Wistar rats and inhibition of P-glycoprotein in everted rat gut sacs in vitro.
Drug Dev. Ind. Pharm. 2014, 40, 1371–1377. [CrossRef] [PubMed]

69
Pharmaceutics 2020, 12, 343

14. Alnaqeeb, M.; Mansor, K.A.; Mallah, E.M.; Ghanim, B.Y.; Idkaidek, N.; Qinna, N.A. Critical pharmacokinetic
and pharmacodynamic drug-herb interactions in rats between warfarin and pomegranate peel or guava
leaves extracts. BMC Complement. Altern. Med. 2019, 19, 29. [CrossRef] [PubMed]
15. Tabares-Guevara, J.H.; Lara-Guzman, O.J.; Londono-Londono, J.A.; Sierra, J.A.; Leon-Varela, Y.M.;
Alvarez-Quintero, R.M.; Osorio, E.J.; Ramirez-Pineda, J.R. Natural Biflavonoids Modulate Macrophage-
Oxidized LDL Interaction In Vitro and Promote Atheroprotection In Vivo. Front. Immunol. 2017, 8, 923.
[CrossRef] [PubMed]
16. Kim, H.P.; Park, H.; Son, K.H.; Chang, H.W.; Kang, S.S. Biochemical pharmacology of biflavonoids:
Implications for anti-inflammatory action. Arch. Pharm. Res. 2008, 31, 265–273. [CrossRef]
17. Liu, H.; Ye, M.; Guo, H. An updated review of randomized clinical trials testing the improvement of cognitive
function of ginkgo biloba extract in healthy people and Alzheimer’s patients. Front. Pharmacol. 2019, 10,
1688. [CrossRef]
18. Lv, X.; Zhang, J.B.; Wang, X.X.; Hu, W.Z.; Shi, Y.S.; Liu, S.W.; Hao, D.C.; Zhang, W.D.; Ge, G.B.; Hou, J.; et al.
Amentoflavone is a potent broad-spectrum inhibitor of human UDP-glucuronosyltransferases. Chem. Biol.
Interact. 2018, 284, 48–55. [CrossRef]
19. Von Moltke, L.L.; Weemhoff, J.L.; Bedir, E.; Khan, I.A.; Harmatz, J.S.; Goldman, P.; Greenblatt, D.J. Inhibition
of human cytochromes P450 by components of Ginkgo biloba. J. Pharm. Pharmacol. 2004, 56, 1039–1044.
[CrossRef]
20. Walsky, R.L.; Obach, R.S.; Gaman, E.A.; Gleeson, J.P.; Proctor, W.R. Selective inhibition of human cytochrome
P4502C8 by montelukast. Drug Metab. Dispos. 2005, 33, 413–418. [CrossRef]
21. Nguyen, P.H.; Ji, D.J.; Han, Y.R.; Choi, J.S.; Rhyu, D.Y.; Min, B.S.; Woo, M.H. Selaginellin and biflavonoids as
protein tyrosine phosphatase 1B inhibitors from Selaginella tamariscina and their glucose uptake stimulatory
effects. Bioorg. Med. Chem. 2015, 23, 3730–3737. [CrossRef] [PubMed]
22. Zhang, Y.X.; Li, Q.Y.; Yan, L.L.; Shi, Y. Structural characterization and identification of biflavones in
Selaginella tamariscina by liquid chromatography-diode-array detection/electrospray ionization tandem
mass spectrometry. Rapid Commun. Mass Spectrom. 2011, 25, 2173–2186. [CrossRef] [PubMed]
23. Heo, J.K.; Nguyen, P.H.; Kim, W.C.; Phuc, N.M.; Liu, K.H. Inhibitory effect of selaginellins from Selaginella
tamariscina (Beauv.) spring against cytochrome p450 and uridine 5 -diphosphoglucuronosyltransferase
isoforms on human liver microsomes. Molecules 2017, 22, 1590. [CrossRef] [PubMed]
24. Kim, H.J.; Lee, H.; Ji, H.K.; Lee, T.; Liu, K.H. Screening of ten cytochrome P450 enzyme activities with 12
probe substrates in human liver microsomes using cocktail incubation and liquid chromatography-tandem
mass spectrometry. Biopharm. Drug Dispos. 2019, 40, 101–111. [CrossRef] [PubMed]
25. Perloff, E.S.; Mason, A.K.; Dehal, S.S.; Blanchard, A.P.; Morgan, L.; Ho, T.; Dandeneau, A.; Crocker, R.M.;
Chandler, C.M.; Boily, N.; et al. Validation of cytochrome P450 time-dependent inhibition assays: A two-time
point IC50 shift approach facilitates kinact assay design. Xenobiotica 2009, 39, 99–112. [CrossRef]
26. Kim, S.J.; You, J.; Choi, H.G.; Kim, J.A.; Jee, J.G.; Lee, S. Selective inhibitory effects of machilin A isolated from
Machilus thunbergii on human cytochrome P450 1A and 2B6. Phytomedicine 2015, 22, 615–620. [CrossRef]
27. Krishnan, S.; Moncrief, S. An evaluation of the cytochrome p450 inhibition potential of lisdexamfetamine in
human liver microsomes. Drug Metab. Dispos. 2007, 35, 180–184. [CrossRef]
28. Joo, J.; Lee, B.; Lee, T.; Liu, K.H. Screening of six UGT enzyme activities in human liver microsomes using
liquid chromatography/triple quadrupole mass spectrometry. Rapid Commun. Mass Spectrom. 2014, 28,
2405–2414. [CrossRef]
29. Shin, J.G.; Soukhova, N.; Flockhart, D.A. Effect of antipsychotic drugs on human liver cytochrome P-450 (CYP)
isoforms in vitro: Preferential inhibition of CYP2D6. Drug Metab. Dispos. 1999, 27, 1078–1084. [PubMed]
30. Lee, E.; Wu, Z.; Shon, J.C.; Liu, K.H. Danazol Inhibits Cytochrome P450 2J2 Activity in a Substrate-independent
Manner. Drug Metab. Dispos. 2015, 43, 1250–1253. [CrossRef]
31. Walsky, R.L.; Gaman, E.A.; Obach, R.S. Examination of 209 drugs for inhibition of cytochrome P450 2C8.
J. Clin. Pharmacol. 2005, 45, 68–78. [CrossRef] [PubMed]
32. Cao, L.; Kwara, A.; Greenblatt, D.J. Metabolic interactions between acetaminophen (paracetamol) and
two flavonoids, luteolin and quercetin, through in-vitro inhibition studies. J. Pharm. Pharmacol. 2017, 69,
1762–1772. [CrossRef] [PubMed]

70
Pharmaceutics 2020, 12, 343

33. Wang, Y.; Wang, M.; Qi, H.; Pan, P.; Hou, T.; Li, J.; He, G.; Zhang, H. Pathway-dependent inhibition of
paclitaxel hydroxylation by kinase inhibitors and assessment of drug-drug interaction potentials. Drug Metab.
Dispos. 2014, 42, 782–795. [CrossRef] [PubMed]
34. Li, X.Q.; Bjorkman, A.; Andersson, T.B.; Ridderstrom, M.; Masimirembwa, C.M. Amodiaquine clearance
and its metabolism to N-desethylamodiaquine is mediated by CYP2C8: A new high affinity and turnover
enzyme-specific probe substrate. J. Pharmacol. Exp. Ther. 2002, 300, 399–407. [CrossRef] [PubMed]
35. Khojasteh, S.C.; Prabhu, S.; Kenny, J.R.; Halladay, J.S.; Lu, A.Y. Chemical inhibitors of cytochrome P450
isoforms in human liver microsomes: A re-evaluation of P450 isoform selectivity. Eur. J. Drug Metab.
Pharmacol. 2011, 36, 1–16. [CrossRef] [PubMed]
36. Brown, H.S.; Galetin, A.; Hallifax, D.; Houston, J.B. Prediction of in vivo drug-drug interactions from
in vitro data: Factors affecting prototypic drug-drug interactions involving CYP2C9, CYP2D6 and CYP3A4.
Clin. Pharmacol. 2006, 45, 1035–1050. [CrossRef]
37. Richter, T.; Murdter, T.E.; Heinkele, G.; Pleiss, J.; Tatzel, S.; Schwab, M.; Eichelbaum, M.; Zanger, U.M. Potent
mechanism-based inhibition of human CYP2B6 by clopidogrel and ticlopidine. J. Pharmacol. Exp. Ther. 2004,
308, 189–197. [CrossRef]
38. Palacharla, R.C.; Molgara, P.; Panthangi, H.R.; Boggavarapu, R.K.; Manoharan, A.K.; Ponnamaneni, R.K.;
Ajjala, D.R.; Nirogi, R. Methoxsalen as an in vitro phenotyping tool in comparison with 1-aminobenzotriazole.
Xenobiotica 2019, 49, 169–176. [CrossRef]
39. Stresser, D.M.; Broudy, M.I.; Ho, T.; Cargill, C.E.; Blanchard, A.P.; Sharma, R.; Dandeneau, A.A.; Goodwin, J.J.;
Turner, S.D.; Erve, J.C.; et al. Highly selective inhibition of human CYP3Aa in vitro by azamulin and evidence
that inhibition is irreversible. Drug Metab. Dispos. 2004, 32, 105–112. [CrossRef]
40. Bjornsson, T.D.; Callaghan, J.T.; Einolf, H.J.; Fischer, V.; Gan, L.; Grimm, S.; Kao, J.; King, S.P.; Miwa, G.;
Ni, L.; et al. The conduct of in vitro and in vivo drug-drug interaction studies: A Pharmaceutical Research
and Manufacturers of America (PhRMA) perspective. Drug Metab. Dispos. 2003, 31, 815–832. [CrossRef]
41. Lee, H.; Heo, J.K.; Lee, G.H.; Park, S.Y.; Jang, S.N.; Kim, H.J.; Kwon, M.J.; Song, I.S.; Liu, K.H. Ginsenoside rc
is a new selective ugt1a9 inhibitor in human liver microsomes and recombinant human ugt isoforms. Drug
Metab. Dispos. 2019, 47, 1372–1379. [CrossRef] [PubMed]
42. Chen, B.; Wang, X.; Lin, D.; Xu, D.; Li, S.; Huang, J.; Weng, S.; Lin, Z.; Zheng, Y.; Yao, H.; et al. Proliposomes
for oral delivery of total biflavonoids extract from Selaginella doederleinii: Formulation development,
optimization, and in vitro-in vivo characterization. Int. J. Nanomed. 2019, 14, 6691–6706. [CrossRef]
[PubMed]
43. Chen, B.; Wang, X.; Zou, Y.; Chen, W.; Wang, G.; Yao, W.; Shi, P.; Li, S.; Lin, S.; Lin, X.; et al. Simultaneous
quantification of five biflavonoids in rat plasma by LC-ESI-MS/MS and its application to a comparatively
pharmacokinetic study of Selaginella doederleinii Hieron extract in rats. J. Pharm. Biomed. Anal. 2018, 149,
80–88. [CrossRef] [PubMed]
44. Liu, D.; Wu, J.; Xie, H.; Liu, M.; Takau, I.; Zhang, H.; Xiong, Y.; Xia, C. Inhibitory effect of hesperetin
and naringenin on human udp-glucuronosyltransferase enzymes: Implications for herb-drug interactions.
Biol. Pharm. Bull. 2016, 39, 2052–2059. [CrossRef]
45. Wang, J.S.; Neuvonen, M.; Wen, X.; Backman, J.T.; Neuvonen, P.J. Gemfibrozil inhibits CYP2C8-mediated
cerivastatin metabolism in human liver microsomes. Drug Metab. Dispos. 2002, 30, 1352–1356. [CrossRef]
46. Vaclavikova, R.; Horsky, S.; Simek, P.; Gut, I. Paclitaxel metabolism in rat and human liver microsomes is
inhibited by phenolic antioxidants. Naunyn. Schmiedebergs Arch. Pharmacol. 2003, 368, 200–209. [CrossRef]
[PubMed]
47. Baldwin, S.J.; Clarke, S.E.; Chenery, R.J. Characterization of the cytochrome P450 enzymes involved in the
in vitro metabolism of rosiglitazone. Br. J. Clin. Pharmacol. 1999, 48, 424–432. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

71
 pharmaceutics

Review
Interpretation of Drug Interaction Using Systemic
and Local Tissue Exposure Changes
Young Hee Choi
College of Pharmacy and Integrated Research Institute for Drug Development, Dongguk University_Seoul,
32 Dongguk-lo, Ilsandong-gu, Goyang-si 10326, Gyeonggi-do, Korea; choiyh@dongguk.edu;
Tel.: +82-31-961-5212

Received: 10 April 2020; Accepted: 30 April 2020; Published: 2 May 2020

Abstract: Systemic exposure of a drug is generally associated with its pharmacodynamic (PD) effect
(e.g., efficacy and toxicity). In this regard, the change in area under the plasma concentration-time
curve (AUC) of a drug, representing its systemic exposure, has been mainly considered in evaluation
of drug-drug interactions (DDIs). Besides the systemic exposure, the drug concentration in the tissues
has emerged as a factor to alter the PD effects. In this review, the status of systemic exposure, and/or
tissue exposure changes in DDIs, were discussed based on the recent reports dealing with transporters
and/or metabolic enzymes mediating DDIs. Particularly, the tissue concentration in the intestine,
liver and kidney were referred to as important factors of PK-based DDIs.

Keywords: drug interaction; pharmacokinetics; tissue-specific; systemic exposure

1. Introduction
Drug-drug interactions (DDIs) are described as the pharmacokinetic (PK) or pharmacodynamic
(PD) influence of a perpetrator drug on a victim drug, resulting in an unexpected effect. In practice,
DDIs have gained much attention due to their changes of pharmacologic effects (i.e., the loss of efficacy
or unintentional toxicity) [1]. The importance of DDIs is well-recognized by the fact that mismanaged
DDIs constitute one of the major causes of drug withdrawal from the market (e.g., mibefradil,
terfenadine, and cisapride) [2–4]. Numerous drug withdrawal cases, as well as serious side-effects
caused by DDIs, including PK-based interactions, can be the result of accompanying clinically relevant
PD interactions [1–4]. The DDIs that can change PK profiles involve: (1) Gastrointestinal absorption,
(2) protein binding in plasma and/or tissue, (3) carrier-mediated transport across plasma membranes
(e.g., hepatic or renal uptake and biliary or urinary secretion), and (4) metabolism. PD interactions,
such as acting an agonist or antagonist at the receptor may also increase or decrease the effects of a
drug [5].
Since the drug concentration at the target site drives its efficacy and toxicity [6], accurate
measurement or prediction of drug concentrations at the target sites is necessary to evaluate DDIs [7–11].
Generally, it is assumed that the plasma drug concentration reflects its concentration in tissue, and drug
concentrations in plasma and tissue are mostly thought to be similar. The direct measurement of
drug concentration in tissue is practically limited, and the plasma drug concentration has been used
as a surrogate for drug concentrations in tissue for PK-PD investigations [11–13]. Among possibly
changed PK profiles in DDIs, transporter- and/or metabolic enzyme-mediated alteration of plasma
concentration of a victim drug by a perpetrator drug is a major proportion to be considered in
PK-based DDI evaluations [12,13]. Of course, plasma drug concentrations do not always reflect
tissue concentrations [6,14–17], and moreover, a closer correlation of tissue concentration, not plasma
concentration, with PD effects of drugs has emerged (e.g., an increased hepatic concentration of
metformin is closely associated with improving its glucose-lowering effect [18,19]; a decreased hepatic
concentration of pravastatin is closely associated with a reduction of its lipid-lowering effect [20,21]).

73
Pharmaceutics 2020, 12, 417; doi:10.3390/pharmaceutics12050417 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 417

Owing to the accumulation of scientific knowledge for understanding PK-related mechanisms of

DDIs and awareness of DDIs, it has become a key issue when assessing DDIs to determine how a victim
drug concentration changes at the target (pharmacological action) sites or in the whole body [22–24].
Regulatory agencies, such as the United States Food and Drug Administration (FDA) and European
Medicines Agency (EMA) have published guidance for drug interaction studies, and they recommend
that evaluation of PK-based DDIs be conducted for drugs under development and on the market [22,23].
As a result, detailed information about DDIs via metabolic enzymes and transporters have become
available at the time of market approval. Nevertheless, the clinically relevant DDIs, leading to the
loss of efficacy or advent of toxicity, have been increasing because polypharmacotherapy is becoming
progressively more common. In addition, the underlying mechanisms of DDIs have been variously
interpreted, which makes it difficult to explain a correlation of PK and PD changes in DDIs, as well as
to predict their clinical relevance [25–29]. In other words, if the alteration of a PD effect is negligible
even through a PK change of a victim drug was observed later, the DDI would be ignored at the time.
In some cases, if the PK and PD changes of a victim drug are not related, it is likely that the critical
underlying mechanism, that triggers a PK change of a victim drug, is mainly associated with a PD
effect has not been investigated [25].
Considering that drug concentrations in plasma and tissue are important parameters to investigate
DDIs in aspects of PK, as well as PD, we summarized how transporters or metabolic enzymes are
involved in the changes of systemic exposure (represented by an area under the plasma concentration
versus time curve (AUC)) and/or local tissue concentration of a victim drug in PK-based DDIs.

2. Transporter and/or Metabolic Enzymes as Main Determinant Factors in PK-Based DDIs

The individual chemical properties of a victim drug or a perpetrator drug (i.e., molecular size,
lipophilicity, ionization, and binding affinity, etc.) can determine its own PK characteristics and affect
the PK change of counterpart individual drugs in DDIs [30–32]. In addition to the physicochemical
properties of drugs, other factors determine the drug concentrations in plasma and tissue, like regulation
of inward or outward flow of a drug into blood vessels or tissues by passive diffusion and/or transporters,
metabolism by phase I and phase II metabolic enzymes, and/or transporter-mediated excretion (e.g.,
renal or biliary excretion) [25,33,34]. First of all, a perpetrator drug can affect transporter-mediated
uptake and/or efflux of a victim drug across cell membranes of tissues, as well as its excretion from
the tissue (e.g., to the blood or outside of the body). Also, a perpetrator drug is able to induce or
inhibit metabolic enzymes and consequently change the enzyme-mediated elimination of a victim
drug [25,35–37]. A transporter-mediated metabolite and/or parent form of a victim drug excretion can be
altered by a perpetrator drug as well in DDIs. In other words, the interplay of transporters and metabolic
enzymes in enterocytes, hepatocytes, and renal proximal tubules is shown in Figure 1, and their
effects on systemic exposure and/or tissue concentration of a victim drug have been increasingly
reported [25,33,34,38].
The functions of transporters are to uptake drugs into cells and to export drugs or drug metabolites
from the cells. Each transporter has specific expression pattern in individual tissues. In particular,
transporters expressed in the small intestine, liver, and kidneys are important for drug disposition and
DDIs [39–42]. We summarized the possible transporters in the human intestine, liver, and kidneys
that are able to cause clinically relevant DDIs in Table 1. Considering the absorption, distribution,
metabolism and excretion (ADME) process of a drug, efflux transporters such as P-glycoprotein
(P-gp; gene symbol ABCB1), the multidrug resistance protein 2 (MRP2; gene symbol ABCC2), and the
breast cancer resistance protein (BCRP; gene symbol ABCG2) are localized to the apical membrane of
enterocytes (Figure 1A), thereby regulating the bioavailability of orally administered substrate drugs.
The concomitantly administered perpetrator drugs inhibited these efflux transporters, resulting in an
increase in the bioavailability of the victim drug. In contrast, the bioavailability of the drug was reduced
when these efflux transporters were induced by perpetrator drugs. After the absorbed drugs pass
through the portal vein and reach the basolateral (i.e., sinusoidal) membrane of hepatocytes (Figure 1B),

74
Pharmaceutics 2020, 12, 417

uptake transporters in the basolateral membrane of hepatocytes mediate the drugs into hepatocytes as
the most important site of drug metabolism. For example, organic anion-transporting polypeptides
(OATPs) such as OATP1B1 (gene symbol SLCO1B1) and organic cation transporters (OCTs), such as
OCT1 (gene symbol SLC22A1), mediate the transport of organic anions and organic cations, respectively.
After which efflux transporters localized in the canalicular membrane of hepatocytes [(e.g., P-gp, BCRP,
multidrug and toxin extrusion protein 1 (MATE1, gene symbol SLC47A1), multidrug resistance protein
2 (MRP2), and bile salt export pump (BSEP, gene symbol ABCB11)] mediate drug transport into the
bile [34,43]. In hepatocytes, the drugs can be metabolized by phase I or phase II metabolic enzymes,
and some of them are transported into bile [34,43]. Sometimes, for a small proportion, the efflux of the
drug or metabolites across the basolateral membrane into can occur [34].
Drug transporters also play a major role in drug secretion from the renal proximal tubular cells into
the urine (Figure 1C). Transporter-mediated uptake across the basolateral membrane of the proximal
tubular cells and efflux across the luminal membrane mainly coordinate the renal secretion of a drug,
which affects the drug concentration in the blood and kidneys. For the secretion of organic cation
drugs, OCT2 (gene symbol SLC22A2), localized in the basolateral membrane uptakes the drug and
subsequently, MATE1 and MATE2-K (gene symbol SLC47A2), localized in the luminal membrane,
efflux the drug in the proximal tubular cells. Similarly, the uptake and efflux transporters for organic
anions are also expressed in the kidney. Alternation of these processes by a concomitantly administered
drug leads to reduced or increased renal clearance of the victim drug [43].
After uptake of a victim drug into the tissue, metabolic pathways are involved in PK-based DDIs
as an elimination pathway. Cytochrome P450s (CYPs) and UDP-glucuronosyltransferases (UGTs) are
the primary metabolic enzymes in phase I and phase II metabolism, as shown in Figure 1 [44,45].
CYP enzymes play a major role in drug elimination through oxidation, reduction, and hydroxylation.
They are mainly present on the smooth endoplasmic reticulum and mitochondria of the hepatocytes
and small intestinal epithelia, and to a lesser extent in the proximal tubules of the kidneys [46].
Their by-products of metabolism known as metabolites can be either, pharmacologically active or
inactive [47]. Among the CYP enzymes in the human liver, CYP3A4 is the most abundant, followed
by CYP2E1 and CYP2C9, representing approximately 22.1%, 15.3%, and 14.6% of the total CYP450s
(based on protein content), respectively [48], among which the highly expressed CYP isoforms show
high potential to cause metabolic DDIs. In addition, drugs or metabolites formed from phase I
metabolism are conjugated with a hydrophilic compound with the help of transferase enzymes during
the process of phase II metabolism. The most common phase II drug-metabolizing enzymes are UGTs,
sulfotransferases (SULTs), N-acetyltransferases (NATs), glutathione S-transferases (GSTs), thiopurine
S-methyltransferase (TPMT), and catechol O-methyltransferase (COMT), and glucuronidation by UGTs
have been reported as major phase II enzyme-mediated DDIs [5,49]. Glucuronide conjugated products
are mostly hydrophilic and are readily excreted from the body, mainly through efflux transporters in the
liver, intestines, and kidneys (e.g., biliary excretion and renal excretion) [5,49,50]. UGTs are normally
highly expressed in the liver and intestine, and their substrates are relatively more overlapping with
each other compared to substrate for CYPs. Especially, UGT1A1, 1A3, 1A9, 2A1, or 2B7-mediated DDIs
have commonly occurred. These pathways, including interplay of metabolic enzymes and transporters
during the ADME of drugs, and any alterations of them may result in changes in the PK and PD of a
victim drug [5,45].

75
Pharmaceutics 2020, 12, 417

Figure 1. Representative transporters and metabolic enzymes in enterocytes (A), hepatocytes (B) and
renal proximal renal tubules (C), which are possibly involved in DDIs.

Table 1. Characteristics of transporters mediating the occurrence of clinically relevant DDIs.

Protein Location Direction Ref.

Apical membrane in enterocyte efflux
P-gp (MDR1) Canalicular membrane in hepatocyte efflux [43,51]
1 Luminal membrane in renal proximal tubule cell efflux
MDR3 Canalicular membrane in hepatocyte efflux [43]
BSEP Canalicular membrane in hepatocyte efflux [43]
Apical membrane in enterocyte efflux
[43]
BCRP Canalicular membrane in hepatocyte efflux
1 Luminal membrane in renal proximal tubule cell efflux
MRP1 Basolateral membrane in renal proximal tubule cell efflux
Apical membrane in enterocyte efflux
MRP2 Canalicular membrane in hepatocyte efflux
1 Luminal membrane in renal proximal tubule cell efflux [6,43]

Basolateral membrane in enterocyte uptake

MRP3 Basolateral membrane in hepatocyte efflux
Basolateral membrane in renal proximal tubule cell efflux
Basolateral membrane in hepatocyte efflux
MRP4 [38,43]
1 Luminal membrane in renal proximal tubule cell efflux
MRP5,6 Basolateral membrane in hepatocyte efflux [43]
Basolateral membrane in enterocyte uptake
OATP1B1
Basolateral membrane in hepatocyte uptake
[43,52]
Basolateral membrane in enterocyte uptake
OATP1B3
Basolateral membrane in hepatocyte uptake

76
Pharmaceutics 2020, 12, 417

Table 1. Cont.

Protein Location Direction Ref.

Basolateral membrane in enterocyte uptake
OATP2B1 [43,53]
Basolateral membrane in hepatocyte uptake
OAT1 Basolateral membrane in renal proximal tubule cell uptake [43]
Basolateral membrane in hepatocyte uptake
OAT2 [38,43]
Basolateral membrane in renal proximal tubule cell uptake
OAT3 Basolateral membrane in renal proximal tubule cell uptake
[43]
OAT4 1 Luminal membrane in renal proximal tubule cell efflux/uptake 2
Basolateral membrane in enterocyte uptake
OCT1
Basolateral membrane in hepatocyte uptake
OCT2 Basolateral membrane in renal proximal tubule cell uptake
[43,54–56]
Basolateral membrane in enterocyte uptake
OCT3 Canalicular membrane in hepatocyte efflux/uptake
Basolateral membrane in renal proximal tubule cell uptake
Canalicular membrane in hepatocyte efflux/uptake
MATE1
1 [43,51]
Luminal membrane in renal proximal tubule cell efflux/uptake 2
MATE2-K 1 Luminal membrane in renal proximal tubule cell efflux/uptake 2

Apical membrane in enterocyte uptake

PEPT1
1 Luminal membrane in renal proximal tubule cell uptake 2 [43,57]
PEPT2 1 Luminal membrane in renal proximal tubule cell uptake 2
1 Luminal membrane means apical membrane. 2 Uptake in apical membrane in kidneys represents a
reabsorption pathway.

3. Effects of Transporter- or Metabolic Enzyme Mediated Changes of Systemic Exposure and/or

Local Tissue Concentration on PD Effects
Regarding the effects of PK profile changes on DDIs, it is important to consider which PK
parameters (or concepts) are primarily or closely associated with PD effects in DDIs [1]. The accurate
measurement or prediction of the unbound (free) drug concentration at the target site is essential for
evaluating DDIs [6,8–11]. When considering the protein binding of a drug, it is assumed that only free
drugs, not bound to proteins and lipids in the blood and tissues, are metabolized and distributed to
target sites where the free drugs exert their pharmacological effects [9,10,58,59]. Protein binding is
possibly able to impact PK and PD, driven by the changes of free drug concentrations [58,59]. However,
it has been generally accepted that the protein binding change of a victim drug shows little clinical
significance in the majority of DDIs cases [58,60].
In the following section, we describe the effects of transporter- and/or metabolic enzymes-mediated
DDIs, resulting in changes in systemic exposure or local tissue concentration of a victim drug, which
affect PD effects. In particular, since transporters and metabolic enzymes are determinant factors
causing DDIs, they are differently expressed and variously interplay together in individual tissues.
Therefore, it is difficult to generally explain the systemic exposure and tissue concentration changes
with PD effects in accordance with the change (e.g., inhibition or induction) of transporters or metabolic
enzymes. Thus, we focused on the change of systemic exposure or tissue concentration in the liver
and kidneys of a victim drug along with how their changes affect the PD effects. Several examples are
presented in Table 2.
As a major site of drug absorption, the changes in transporters in the apical and basolateral
membranes, as well as the metabolic enzymes of enterocytes determine the final amount of drug
absorbed into systemic circulation [54,57]. An increase of drug absorption happens due to a drug influx

77
Pharmaceutics 2020, 12, 417

from the intestinal lumen into enterocytes, by an increase of influx transporter or a decrease of efflux
transport in the apical membrane, a decrease of drug metabolism in enterocytes, or an increase of drug
efflux from enterocytes to blood by an increase of transporter in the basolateral membrane [37,54,57].
Recently, drug catalysis by gut microbiota has emerged as one pathway to regulate drug absorption [61].
A change in drug absorption affects the systemic exposure and intestinal concentration of a drug, and
they are generally similar in most cases. In particular, the changed intestinal concentration of a drug
can drive the alterations of efficacy and toxicity that occur in the intestine [61,62].
In the liver, the changed final hepatic clearance of a victim drug affects its systemic and local
tissue exposure depending on the contribution of the transporter and metabolic enzyme-mediated
pathways (i.e., uptake or efflux of a victim drug across the basolateral membrane of hepatocyte,
hepatic metabolism, and biliary excretion). There are two major cases. One is that the metabolic
plus biliary efflux clearance is a determinant step in drug elimination, so the basolateral uptake
becomes the rate-determining step in the hepatic clearance of the drug. In other words, the hepatic
clearance is driven mainly by uptake clearance. The inhibition of uptake transporter in the basolateral
membrane will increase plasma drug concentrations and systemic exposure of a victim drug, but will
not significantly impact the liver AUC of a drug. This may be due to the victim drug being mainly
eliminated by the liver [6], and the alteration of basolateral uptake into hepatocytes is the main factor
behind the change in the plasma concentration of a victim drug.
The second is that the inhibition of metabolic enzymes or biliary efflux transporters can increase
the liver AUC of a victim drug while minimally impacting its plasma AUC [63,64], which may be
due to the elimination pathway after uptake into hepatocytes being a determinant factor for final
hepatic clearance. In these examples, the changes of plasma and liver concentrations of the drug are
not symmetric directions. This asymmetry may lead to misinterpretations of the plasma concentration
and/or tissue concentration (e.g., the liver) of a victim drug in DDIs, and in particular, this issue is
important to explain the association of PK and PD changes in DDIs [6].
In the kidneys, the alteration of systemic exposure (or plasma concentrations) of victim drugs
in renal transporter-mediated DDIs is comparatively moderate compared with those in intestinal or
hepatic-transporter mediated DDIs. However, a change in renal clearance of drugs that are mainly
eliminated via renal routes can substantially affect the systemic levels of victim drugs, and sometimes
these changes are related to efficacy and/or toxicity changes [38,54,65]. Moreover, if the kidneys are
the pharmacological target site, a drug concentration change of a victim drug in the kidneys is a
critical factor regulating the drug efficacy [66–69]. From here, we have divided the next section into
three subsections as follows: (1) Systemic exposure change of a victim drug having a major PD effect;
(2) tissue concentration changes of a victim drug having a major PD effect; and (3) additional factors
affecting PD effects along with changes in systemic exposure or tissue concentration of a victim drug.

78
 Table 2. Examples of transporter- or metabolic enzyme-mediated DDIs.

Victim Drug Perpetrator Drug Underlying Mechanism 1 PK Change of a Victim Drug PD Change of a Victim Drug Ref.
Apixaban Ketoconazole (−) P-gp in enterocyte AUC↑ ADR↑ (bleeding risk) [70]

Dabigatran Rifampin (+) P-gp in enterocyte AUC↓ TR༸, safety༸ [71]

Digoxin Rifampin (+) P-gp in enterocyte AUC↓ TR↓ [72]

Pharmaceutics 2020, 12, 417

ADR↑(respiratory depression) by
Loperamide Quinidine (−) P-gp in enterocyte or brain AUC↑ P-gp inhibition in brain (not [73]
enterocyte)
Eltrombopag,
Rosuvastatin (−) BCRP in enterocyte AUC↑ ADR↑ (myopathy), TR↑ [43,74,75]
Fostamatinib
(+) P-gp in enterocyte AUC↓, F↓,
Clopidogrel Aspirin Platelet inhibition effect༸ [29]
(+) CYP2C9 in hepatocyte H4 (active metabolite)↑
(−) P-gp in enterocyte
AUC↑, CL↓, CLR ↓ (by
Digoxin Clarithromycin (−) CYP2D6 or 3A4 inhibition in ADR↑ (digoxin toxicity) [76]
non-glomerular renal clearance)
hepatocyte
(−) OATP1B1, 1B3, 2B1 in hepatocyte

79
Cyclosporine AUC↑, hepatic uptake༸ Muscle-related toxicity↑ [1,43,77,78]
Atorvastatin (−) CYP3A in hepatocyte

Itraconazole (−) CYP3A in hepatocyte AUC↑, hepatic uptake༸ - [62,79,80]

ADR↑
Bosentan Clarithromycin (−) OATP1B1, 1B3 in hepatocyte AUC↑ [81,82]
(cholestatic liver injury)
Cyclosporine,
Pitavastatin (−) OATP1B1, 1B3, 2B1 in hepatocyte AUC↑ ADR↑ [83,84]
rifampin
Itraconazole,
Atrovastatin, ADR↑ (myopathy, fatal
mibefradil, (−) CYP3A4 in hepatocyte AUC↑, Cmax ↑ [85]
simvastatin rhadomyolysis)
verapamil
Atrovastatin,
(−) OATP1B1, 1B3, 2B1 in hepatocyte ADR↑ (myopathy, fatal
pravastatin, Clarithromycin AUC↑, Cmax ↑ [85]
(−) CYP3A4 in hepatocyte rhadomyolysis)
simvastatin
 Table 2. Cont.

Victim Drug Perpetrator Drug Underlying Mechanism 1 PK Change of a Victim Drug PD Change of a Victim Drug Ref.

Cyclosporine (−) OATP1B1, 1B3, 2B1 in hepatocyte AUC↑, hepatic cons 2 ༸ ADR↑ [86,87]
Rosuvastatin
Gemfibrozil (−) OATP2B1 in hepatocyte AUC↑, hepatic cons 2 ༸ ADR↑ [87,88]

(−) OATP1B1 in hepatocyte

Pharmaceutics 2020, 12, 417

Simvastatin Cyclosporine AUC↑ ADR↑ (myopathy) [43]

(−) CYP3A4 in hepatocyte
Adefovir Probenecid (−) OAT1 in proximal tubule cell AUC↑, CLR ↓ ADR↑ (nephrotoxicity) [43,89]
Benzylpenicillin Probenecid (−) OAT3 in proximal tubule cell AUC↑, CLR ↓ ADR↑ [89]
(−) CYP2D6 or 3A4 inhibition in
Digoxin Quinidine hepatocyte AUC↑, CL↓, CLNR ↓, CLR ↓ ADR↑ (digoxin toxicity) [83,90]
(−) P-gp in proximal tubule cell
Trimethoprim (−) OCT2, MATE1, MATE2-K in
Lamivudine AUC↑, CLR ↓ ADR↑ (hepatotoxicity) [91,92]
/sulfamethoxazole proximal tubule cell
ADR↑ (plasma lactate↑, lactic
(−) OCT2, MATE1 in proximal tubule
Cmax ↑, AUC↑, CL/F↓, CLR ↓ acidosis especially in renal [93]

80
Trimethoprim cell
dysfunction patients)
Metformin (−) OCT2, MATE1, MATE2-K in ADR↑ (plasma lactate↑, lactic
Cmax ↑, AUC↑, CLR ↓ [94]
proximal tubule cell acidosis)
ADR↑ (plasma lactate↑, lactic
Dolutegravir (−) OCT2 in proximal tubule cell Cmax ↑, AUC↑ [95]
acidosis)
(−) Mrp2 in enterocyte; Intestinal absorption↑, AUC↑,
Pravastatin Paroxetine Lipid-lowing effect ↓ [20,21,96]
(−) Oatp2 in enterocyte/hepatocyte hepatic uptake↓, hepatic cons↓
[52,53,79,
Atorvastatin Rifampin (−) OATPs in hepatocyte AUC↑, hepatic uptake↓ Lipid-lowing effect↓
97]
Atorvastatin Metformin (−) MRP2 in hepatocyte Biliary excretion↓, hepatic cons 2 ↑ Lipid-lowing effect↑ [98]
Metformin Rifampin (+) mRNA of OCT1 in blood cells AUC↑ (probable hepatic cons 2 ↑) Glucose-lowering effect↑ [19]
Lonicera japonica
Metformin (−) MATE1 in hepatocyte AUC༸, hepatic cons 2 ↑ Glucose tolerance effect↑ [67]
extract
 Table 2. Cont.

Victim Drug Perpetrator Drug Underlying Mechanism 1 PK Change of a Victim Drug PD Change of a Victim Drug Ref.
Metformin Nuciferine (−) OCT1 and MATE1 in hepatocyte Hepatic cons 2 ↓ Glucose-lowering effect↓ [99]
AUC↑, hepatic cons 2 ↓, renal cons
Rosuvastatin Rifampin (−) OATP1B1, 1B3, 2B1 in hepatocyte ADR↑ [87]
2 ↓, hepatic biliary excretion༸
Pharmaceutics 2020, 12, 417

Houttuynia cordata (−) MATE1 in hepatocyte

Metformin AUC↑, CLR ↓, hepatic cons 2 ↑ Glucose tolerance effect↑ [66]
extract (−) OCT2 in proximal tubule cell
(−) MATE1 in hepatocyte;
AUC↑, hepatic cons 2 ↑, (biliary [51,55,56,
Metformin Cimetidine (−) MATE1, MATE2-K in proximal Glucose-lowering effect↑
excretion↓), renal cons 2 ↑, CLR ↓ 100,101]
tubule cell
(−) MATE1 in hepatocyte; AUC↑, Cmax ↑, CLR ↓, CLCR ↓, SCR
Metformin Pyrimethamine Glucose lowering effect↑ [101–103]
(−) MATE1 in proximal renal tubule 3 ↑, hepatic cons 2 ↑

The DDI cases mentioned in the main text is marked by bold style. 1 (−) and (+) present inhibition and induction, respectively. 2 Cons refers to concentration. 3 SCR refers to a serum
creatinine level.

81
Pharmaceutics 2020, 12, 417

3.1. Changed Systemic Exposure of a Victim Drug Affecting the PD Effect

First of all, it is necessary to clarify the relevance of plasma drug concentration and plasma AUC
when assessing DDIs. A plasma AUC alteration implies that the systemic exposure of a victim drug
is changed by a perpetrator drug in DDIs, which does not consider the exact time point. Since the
AUC is an integrated value from the plasma drug concentration-time curve, the AUC change is not
identical to plasma drug concentration changes at all sampling points. If the plasma drug concentration
at the specific time points affect PD effects, the plasma drug concentration with the sampling time
point or periods should be mentioned to prevent confusion of plasma drug concentrations for plasma
AUC. For example, Cmax , the highest plasma drug concentration that is generally related to toxicity,
has been used to evaluate or predict a risk of toxicity change in DDIs [24]. Assuming that plasma
drug concentrations reflect the efficacy or toxicity of a drug, the systemic exposure based on plasma
AUC from time zero to last sampling time (or time infinity) is used as a proper parameter in the
evaluation of DDIs by numerous investigators and regulatory regencies (e.g., FDA and EMA) [22–24].
The systemic exposure change based on plasma AUC is considered to trigger the drug efficacy or
toxicity, and especially, the systemic exposure reflects the duration of drug efficacy: The higher plasma
AUC results in a stronger efficacy and/or toxicity [24,25,38]. Of course, there may be an exception, such
as onset time or duration time in drug efficacy not being reflected by systemic exposure of a drug (e.g.,
decreased plasma AUC of clopidogrel is not related to a reduction in anti-coagulant activity; [29]),
which is described later in detail.
According to the guidelines for DDI studies suggested by FDA and EMA, the
AUC ratio for a victim drug in the presence and absence of a perpetrator drug
(AUCin the presence of a perpetrator /AUCin the absence of a perpetrator ) > 5 is determined as a strong PK-based
DDI occurrence, because a perpetrator drug may inhibit transporter- or metabolic enzyme-mediated
clearance of a victim drug. The AUC ratio for a victim drug in the presence and absence of a perpetrator
drug <0.5 is also considered, because a perpetrator drug acts as a strong inducer for transporter- or
metabolic enzyme-mediated elimination. Both these strong DDI cases are recommended to be labelled
as a victim drug [4,24]. Moreover, the plasma AUC represents systemic exposure of a victim drug, and
is used as a critical parameter to adjust dosage regimens to clinical levels. Since the contribution of
transporters and metabolic enzymes in the occurrence and/or severity of DDIs is different depending
on the individual drug combinations, and because there is a pronounced inter-individual variability in
the magnitude of the perpetrator effect, it is hard to predict the DDIs and it is necessary to evaluate
them case-by-case.
There are several examples where the systemic exposure of a victim drug affects its pharmacological
activity. For example, atorvastatin, a permeable drug (log D7.4 = 1.53), is a substrate of the
OATP transporter (e.g., OATP1B1, 1B3, and 2B1) and is eliminated primarily by hepatic CYP3A
metabolism [52,77,78,97]. In a case of atorvastatin co-administered with cyclosporine as a strong
OATP inhibitor and moderate CYP3A inhibitor, a 5- to 16-fold increase of plasma AUC of atorvastatin
occurs, and a reduction of the atorvastatin dose is clinically recommended, due to an increased rate of
muscle-related toxicity, such as myopathy and rhabdomyolysis [1,43,77,78]. Additional DDI studies of
atorvastatin are useful for understanding the underlying mechanism of the AUC change of atorvastatin
along with the occurrence of toxicity. Co-administered rifampin, an OATP inhibitor, leads to an increase
of plasma AUC of atorvastatin, but co-administered itraconazole, a CYP3A inhibitor, does not cause
any change of the plasma AUC of atorvastatin. This result suggests that hepatic uptake of atorvastatin
via OATPs is the rate-determining step for the hepatic clearance of atorvastatin [62,79,80], and that
the plasma concentration of atorvastatin does not reflect its liver concentration [53]. Although the
relationship between an increase of atorvastatin AUC in plasma and its pharmacological activity (e.g.,
cholesterol lowering activity or toxicity) was not mentioned in this example, it is a good example of the
contribution of transporter and metabolic enzymes to changes in plasma profiles of a victim drug that
do not affect its tissue concentration (e.g., liver) in DDIs.

82
Pharmaceutics 2020, 12, 417

Co-administration of cimetidine, a MATE1 and OCT2 inhibitor [51,55,56,100,101], with metformin

increases systemic exposure of metformin due to the reduction of renal clearance of metformin
via inhibited tubular secretion. The effect of cimetidine on the renal excretion of metformin is
time-dependent. In other words, cimetidine inhibited metformin renal secretion up to 6 h after
cimetidine co-administration, accompanied by an increased blood lactate/pyruvate ratio from 4 h after
cimetidine co-administration and reduced creatinine clearance as a dose and concentration-dependent
adverse effect of metformin [100]. This is an example of the increased systemic exposure reflecting a
toxicity risk of a victim drug in DDIs.

3.2. Changed Local Tissue Concentration of a Victim Drug Affecting the PD Effect
In contrast to the numerous DDIs evaluated by the changes of systemic exposure (i.e., plasma
drug concentrations), there are also DDIs that show pharmacological action changes caused by the
tissue concentration of a victim drug [53,62,98]. In vivo drug concentration at a target site triggering
a PD effect is compared to in vitro efficient concentration, which has been used to explain PK-PD
correlations of a victim drug in DDIs [66,98]. In vivo tissue concentration/in vitro half-maximal
inhibitory concentration for inhibiting transporter or metabolic enzymes ([I]/IC50 ) or in vivo tissue
concentration/in vitro inhibition constant for transporter-or metabolic enzyme ([I]/Ki ) of a perpetrator
drug are used to investigate whether the tissue concentration of a perpetrator drug is sufficient to
inhibit transporter-or metabolic enzyme-mediated interactions with a victim drug. When the ratio
of IC50 /Ki is over 2, the inhibitory interaction of a perpetrator drug to a victim drug can sufficiently
happen in vivo in tissue [104,105]. Especially, if there is a tissue lag time relative to peak plasma
concentration, the duration of drug efficacy or toxicity should also be considered [6,106,107].
After a drug is absorbed into blood and delivered to the tissue, the drug concentration in the tissue
is the sum of the net membrane permeation by influx and efflux across the membrane and the intrinsic
clearance by metabolism and/or excretion (e.g., biliary excretion in the liver and renal excretion in
the kidneys) [33,34,108,109]. Although membrane permeation and intrinsic clearance are variable in
individual tissues, drug influx consists of transporter-mediated uptake and passive diffusion from
blood to tissue, as well as transporter-mediated active efflux. After the sum of membrane permeation,
a drug is exposed to metabolic enzymes and/or excretion pathways within the localized tissue, which
determines the drug concentration in the tissue, as well as the plasma drug concentration as systemic
exposure of a drug.
When specific tissues are the pharmacological target sites to show drug efficacy and/or toxicity,
drug concentration changes in these tissues strongly affect the PD changes in DDIs. The tissue
concentration of a victim drug does not necessarily change in parallel to plasma AUC of a victim drug.
Thus, it is necessary to choose a tissue to evaluate DDIs based on the pharmacological mechanism and
major organ regulating drug disposition. The liver and/or kidneys are frequently involved in drug
disposition pathways, but the pharmacological target organs absolutely depend on the individual
drugs used. Time-dependent changes of drug concentrations in tissues also need to be considered,
especially when onset or duration time is an important factor for drug efficacy [38].
There are several cases showing no change of AUC of a victim drug, even though there is a change of
tissue drug concentration, which has been named a silent interaction [63,98]. In the case of atorvastatin
with metformin co-administration as a silent interaction, co-administered metformin reduces the
biliary excretion of atorvastatin by MRP2 inhibition, increasing the atorvastatin concentration in the
liver without changes in total clearance or systemic exposure of atorvastatin [98]. Shin et al. [98]
concluded that the increased atorvastatin concentration in the liver might consequently improve the
lipid-lowering effect of atorvastatin as in similar DDI studies of statins and metformin [110,111].
In the case of rosuvastatin, the inhibition of OATP uptake by rifampin leads to an increase of blood
AUC of rosuvastatin without any significant change to the hepatic AUC of rosuvastatin in rats [87].
Similarly, PBPK analysis of rosuvastatin in humans illustrated that the predicted liver concentrations of
rosuvastatin are not significantly impacted. Moreover, the OATP1B1 polymorphism affects the plasma

83
Pharmaceutics 2020, 12, 417

concentrations of rosuvastatin without affecting its PD response [112]. These cases demonstrate that
the plasma exposure changes do not reflect liver exposure changes without any relevant PD effects.
This is because the rate-determining step in the hepatic clearance of rosuvastatin is uptake clearance
via OATPs, even though rosuvastatin is also excreted into bile via MRP2/BCRP [74,75,86–88,112].
The case of metformin, an anti-diabetic agent, is an interesting example. Due to metformin’s
high pKa (most portions are positively charged at pH 7.4) and its negative logP value, the passive
diffusion of metformin through cellular membranes is minor. Therefore, transporters are pivotal for
metformin permeation through cellular membranes. In the intestine, apical OCT3 and basolateral
OCT1 primarily mediate metformin absorption. In the liver, basolateral OCT1 and OCT3 uptake
metformin from the sinusoidal blood into hepatocytes, whereas apical MATE1 is thought to efflux
metformin into the bile. In renal proximal tubule cells, OCT2, MATE1, and MATE2-K are essential for
the renal tubular secretory system of metformin [56]. The pharmacological action site of metformin
is the liver, but renal excretion as an unchanged form, is the main elimination route of metformin,
which primarily contributes to determining the systemic exposure of metformin along with intestinal
absorption. Metformin disposition changes, such as in plasma and tissue concentration with relevant
pharmacological effect alterations is determined by OCT and MATE mediated metformin transport.
This has been proven based on studies conducted of metformin administration to patients with OCTs
and MATEs genetic polymorphisms, as well as OCTs and MATEs knock-out mice [101,113–115].
With regards to DDIs of metformin, Cho et al. [19] reported that rifampin significantly enhanced
the glucose-lowering action of metformin (54.5% of AUCglucose with P = 0.020) due to the increased
mRNA level of OCT1 by rifampin probably enhancing hepatic uptake of metformin. Conversely,
co-administration of an OCT1 inhibitor, verapamil, with metformin decreased the glucose-lowering
effect of metformin without increases of systemic exposure of metformin in healthy participants [18].
In mice, co-infusion of cimetidine increases metformin concentrations in the liver and kidneys, due
to the inhibition of mMate1-mediated metformin export to biliary excretion and renal excretion,
respectively [51,55,56], suggesting that cimetidine exerts MATE1 inhibition-mediated interaction with
metformin. In another mouse study, co-administration of metformin with pyrimethamine, a MATE
inhibitor, also resulted in a ~2.5-fold increase in liver AUC of metformin compared with controls
(i.e., metformin alone administration [101,102,114,115]). These examples indicate that the inhibition
or induction of transporters primarily mediating drug disposition to pharmacological target tissue
has a strong potential to alter the efficacy of the drug, regardless of the systemic exposure, such as
plasma concentration.
Although we focused on DDIs in this review, similar phenomena have been observed in herb-drug
interactions. Han et al. [67] reported that the glucose tolerance activity of metformin was enhanced
without a change of metformin plasma concentration, because the metformin concentration in the
liver increased as a result of a reduction of mate1-mediated biliary excretion of metformin in rats
simultaneously treated with metformin and Lonicera japonica extract. Considering that renal clearance is
the main route of metformin’s elimination, only hepatic transporter-mediated interactions of metformin
will impact its hepatic concentration, and therefore, affect the PD effect without the alteration of its
plasma concentration.
Interestingly, there were several cases where the systemic exposure and tissue exposure of a
drug changed in the opposite direction (e.g., increase of systemic exposure and decrease of hepatic
exposure). For example, when paroxetine is co-administered with pravastatin in rats, paroxetine
increased the systemic exposure and decreased the liver exposure of pravastatin by the combined
effects of an increase in intestinal absorption and a decrease in hepatic uptake of pravastatin via Oatp2
inhibition as well as increased biliary excretion via Mrp2 inhibition [96]. The reduced hepatic exposure
of pravastatin had a trend to weaken the lipid-lowing effect of pravastatin in diabetic rats [20,21] in spite
of the increased systemic exposure of pravastatin. In a case of herb-drug interactions, You et al. [66]
reported that the AUC of metformin was increased due to the decrease of oct2-mediated renal excretion
of metformin, and thus, the metformin concentration in the kidneys increased due to the increase in

84
Pharmaceutics 2020, 12, 417

oct1-mediated renal uptake of metformin along with the enhancement of its glucose-lowering effect in
rats with 28-day co-treatment of metformin and Houttuynia cordata (H. Cordata) extract. Considering
that metformin’s pharmacological action site is the liver, the enhanced glucose-lowering effect is
probably due to increased hepatic uptake of metformin by the H. Cordata extract. In spite of an increase
of metformin’s systemic exposure, any toxicity of metformin identified as renal dysfunction and lactic
acidosis were not observed. Thus, the metformin-H. Cordata extract combination case can be included
as an example of a local tissue concentration change more strongly affecting the PD effect.

3.3. Additional Factors Affecting PD Effects with Changes of Systemic Exposure or Local Tissue Concentration
of a Victim Drug
Additional underlying mechanisms can also cause PD alterations in DDIs. The first case represents
how a PK change of an active metabolite can affect the PD effect in DDIs. In the case of clopidogrel
with co-administration of aspirin, the systemic exposure of clopidogrel is reduced due to the intestinal
P-gp induction lowering the bioavailability of clopidogrel, but the relative platelet inhibition effect of
clopidogrel is not changed [29]. Since co-administered aspirin increases clopidogrel metabolism via
CYP2C19 and the AUC of the active thiol metabolite, H4, of clopidogrel is consequently increased, the
reduced platelet inhibition effect due to the reduced AUC of clopidogrel might be compensated for by
an increase of H4 s AUC [29]. This example indicates that the systemic exposure of a parent drug is
not always able to explain the alteration of PD effect, and the systemic exposure of active metabolites
can also cause PD changes. Thus, the systemic exposure of a parent drug, as well as active metabolites
need to be considered together, especially when predicting pharmacological effects.
The second case is that the co-administration interval between a victim drug and a perpetrator
drug can determine the PD effect. Co-administration of nuciferine, as a potential inhibitor of OCT1
and MATE, time-dependently reduced the glucose-lowering effect of metformin and the hepatic
metformin concentration. The hepatic metformin concentration was increased until 1 h after nuciferine
administration, which subsequently enhanced the glucose-lowering effect of metformin only during the
2.5-4 h after nuciferine co-administration [99]. If the metformin concentration in the liver is measured at
longer time intervals after nuciferine co-administration, the increased hepatic metformin concentration,
due to the increased OCT1 and reduced MATE1 transport activity for metformin movement by
nuciferine would not be detected, and the alteration of the glucose-lowering effect would also be in the
same direction. This example indicates that the onset and duration time of pharmacological action is
one factor to be considered to evaluate the PK and PD changes in DDIs.

4. Challenging Experimental Approaches for Exploring the Transporter- or Metabolic Enzyme-

Mediated Ddis
It is necessary to explore the expression or activity changes of transporters- or metabolic enzymes
responsible for DDI occurrences. Even though the guidance from regulatory agents (e.g., EMA and
FDA), labeling recommendations, and the reported references have informed the DDI potential,
general in vitro and in vivo methods, as well as clinically designed approaches have been applied in
DDI evaluations [22–25,37] and elsewhere, the inconsistency between the clinical relevance and the
accumulated information (or experimental results) is still present [66,67,79,98] and elsewhere. Besides
the conventional tools (e.g., western blot analysis, q-PCR analysis, permeability test, etc.) to measure
the protein and mRNA expression or activity of transporters or metabolic enzymes in tissues, require a
guess as to the target causing PK alteration. Therefore, the proteomic quantification methods with
“total protein approach” (TPA) for transporters and metabolic enzymes in liver or intestines have been
attempted [116]. The TPA-based quantification of transporters and enzymes in human liver tissue
samples using liquid chromatography-tandem mass spectrometry can detect their protein expression
levels, which could correlate with the results from the targeted proteomic quantitation. This challenging
TPA-based quantification can identify as many proteins as possible in a small volume of the same

85
Pharmaceutics 2020, 12, 417

sample and does not require standards [116,117]. In addition, unexpected interplay of transporters or
metabolic enzymes in DDI events can be detected by TPA-based quantification approach.
As an experimental approach to investigate the tissue exposure, microdialysis technique in tissue
distribution studies has recently been introduced, making it possible to measure the drug concentration
at multiple time points in one living animal [118,119]. Animals are sacrificed at each sampling point
using the classical experimental approaches to measure drug concentration in tissues, and at least four
sampling time points are recommended to explain the tissue distribution pattern changes in DDIs,
considering Cmax and terminal half-life [90,120]. Additionally, (1) composition of a victim drug and
a perpetrator drug, and (2) the dosage regimen are suggested to be considered in designing in vivo
conventional PK experiments. A perpetrator drug can be chosen as an index substrate, inhibitor or
inducer of the transporter or metabolic enzyme responsible for the PK of a victim drug, or as a highly
recommended drug in clinical combination therapies [3,25,28]. The dosage regimens, including doses,
treatment period (i.e., single or multiple treatments), and drug-dosing schedule of a victim drug and a
perpetrator drug, can result in different DDI events [6,22–25,66,67,121].

5. Concluding Remarks
Evaluations of DDIs are a crucial issue for drug efficacy and safety. Although methodologies
to evaluate DDIs and information about mechanisms of PK-based DDIs have greatly advanced,
the incidences and severity of DDIs still remain high in clinical cases. Moreover, previously unknown
DDIs have been revealed and their mechanisms have been suggested in clinical DDI studies, but most
of them have then been confirmed by, for example, in vitro studies through a reverse translation
process. Since clinical DDI studies may not cover the numerous combinations and various factors that
cause these outcomes, it is necessary to interpret the PK and PD data to provide a comprehensive
understanding of DDIs with clinical relevance. Although there is currently no optimal way to study
DDIs, its evaluation needs to be based on interpretation of the available data about PK-based DDIs,
which can ensure the safety and maximal usefulness of DDI studies.

Funding: This research was funded by National Research Foundation of Korea (NRF) grants funded by the Korea
government (MSIT) (NRF-2016R1C1B2010849 and NRF-2018R15A2023217).
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Wiggins, B.S.; Saseen, J.J.; Page, R.L., II; Reed, B.N.; Sneed, K.; Kostis, J.B.; Lanfear, D.; Virani, S.; Morris, P.B.
Recommendations for management of clinically significant drug-drug interactions with statins and select
agents used in patients with cardiovascular disease. Circulations 2016, 134, e468–e495.
2. Dechanont, S.; Maphanta, S.; Butthum, B.; Kongkaew, C. Hospital admissions/visits associated with drug-drug
interactions: A systemic review and meta-analysis. Pharmacoepidemiol. Drug Saf. 2014, 23, 489–497. [CrossRef]
[PubMed]
3. Huang, S.M.; Strong, J.M.; Zhang, L.; Reynolds, K.S.; Nallani, S.; Temple, R.; Abraham, S.; Habet, S.A.;
Baweja, R.K.; Burchart, G.J.; et al. New era in drug interaction evaluations: US Food and Drug Administration
update on CYP enzymes, transporters, and guideline process. J. Clin. Pharmacol. 2008, 48, 662–670. [CrossRef]
[PubMed]
4. Yu, J.; Zhou, Z.; Tay-Sontheimer, J.; Levy, R.H.; Ragueneau-Majlessi, I. Risk of clinically relevant
pharmacokinetic-based drug-drug interactions with drugs approved by the U.S. Food and Drug
Administration between 2013 and 2016. Drug Metab. Dispos. 2018, 46, 835–845. [CrossRef]
5. Ito, K.; Iwatsubo, T.; Kanamitsu, S.; Ueda, K.; Suzuki, H.; Sugiyama, Y. Prediction of pharmacokinetic
alterations caused by drug-drug interactions: Metabolic interaction in the liver. Pharmacol. Rev. 1998, 50,
387–411.
6. Zhang, D.; Hop, C.E.; Patilea-Vrana, G.; Gampa, G.; Seneviratne, H.K.; Unadkat, J.D.; Kenny, J.R.; Nagapudi, K.;
Di, L.; Zhou, L.; et al. Drug concentration asymmetry in tissues and plasma for small molecule–related
therapeutic modalities. Drug Metab. Dispos. 2019, 47, 1122–1135. [CrossRef]

86
Pharmaceutics 2020, 12, 417

7. Mager, D.E.; Jusko, W.J. General pharmacokinetic model for drugs exhibiting target-mediated drug disposition.
J. Pharmacokinet. Pharmacodyn. 2001, 28, 507–532. [CrossRef]
8. Rankovic, Z. CNS drug design: Balancing physicochemical properties for optimal brain exposure. J. Med.
Chem. 2015, 58, 2584–2608. [CrossRef]
9. Smith, D.A.; Di, L.; Kerns, E.H. The effect of plasma protein binding on in vivo efficacy: Misconceptions in
drug discovery. Nat. Rev. Drug Discov. 2010, 9, 929–939. [CrossRef]
10. Di, L.; Umland, J.P.; Trapa, P.E.; Maurer, T.S. Impact of recovery on fraction unbound using equilibrium
dialysis. J. Pharm. Sci. 2012, 101, 1327–1335. [CrossRef]
11. Di, L.; Breen, C.; Chambers, R.; Eckley, S.T.; Fricke, R.; Ghosh, A.; Harradine, P.; Kalvass, J.C.; Ho, S.; Lee, C.A.;
et al. Industry perspective on contemporary protein-binding methodologies: Considerations for regulatory
drug-drug interaction and related guidelines on highly bound drugs. J. Pharm. Sci. 2017, 106, 3442–3452.
[CrossRef]
12. Gabrielsson, J.; Peletier, I.A. Pharmacokinetic steady-states highlight interesting target-mediated disposition
properties. AAPS J. 2017, 19, 772–786. [CrossRef] [PubMed]
13. Van Waterschoot, R.A.; Parrott, N.J.; Olivares-Moralres, A.; Lave, T.; Rowland, M.; Smith, D.A. Impact of
target interactions on small-molecule drug disposition: An overlooked area. Nat. Rev. Drug Discov. 2018, 17,
299–301. [CrossRef]
14. Li, R.; Kimoto, E.; Niosi, M.; Tess, D.A.; Lin, J.; Tremaine, L.M.; Di, L. A study on pharmacokinetics of
bosentan with systems modeling, part 2: Prospectively predicting systemic and liver exposure in healthy
subjects. Drug Metab. Dispos. 2018, 46, 357–366. [CrossRef] [PubMed]
15. Li, R.; Niosi, M.; Johnson, N.; Tess, D.A.; Kimoto, E.; Lin, J.; Yang, X.; Riccardi, K.A.; Ryu, S.; El-Kattan, A.F.;
et al. A study on pharmacokinetics of bosentan with systems modeling, part 1: Translating systemic plasma
concentration to liver exposure in healthy subjects. Drug Metab. Dispos. 2018, 46, 346–356. [CrossRef]
16. Tsamandouras, N.; Dickinson, G.; Guo, Y.; Hall, S.; Rostami-Hodjegan, A.; Galetin, A.; Aarons, L. Development
and application of a mechanistic pharmacokinetic model for simvastatin and its active metabolite simvastatin
acid using an integrated population PBPK approach. Pharm. Res. 2015, 32, 1864–1883. [CrossRef] [PubMed]
17. Watanabe, T.; Kusuhara, H.; Maeda, K.; Shitara, Y.; Sugiyama, Y. Physiologically based pharmacokinetic
modeling to predict transporter-mediated clearance and distribution of pravastatin in humans. J. Pharmacol.
Exp. Ther. 2009, 328, 652–662. [CrossRef]
18. Cho, S.K.; Kim, C.O.; Park, E.S.; Chung, J.Y. Verapamil decreases the glucose-lowering effect of metformin in
healthy volunteers. Br. J. Clin. Pharmacol. 2014, 78, 1426–1432. [CrossRef]
19. Cho, S.K.; Yoo, J.S.; Lee, M.G.; Lee, D.H.; Lim, L.A.; Park, K.; Park, M.S.; Chung, J.Y. Rifampin enhances
the glucose-lowering effect of metformin and increases OCT1 mRNA levels in healthy participants.
Clin. Pharmacol. Ther. 2011, 89, 416–421. [CrossRef]
20. Li, F.; Zhang, M.; Xu, D.; Liu, C.; Zhong, Z.Y.; Jia, L.L.; Hu, M.Y.; Yang, Y.; Liu, L.; Liu, X.D. Co-administration
of paroxetine and pravastatin causes deregulation of glucose homeostasis in diabetic rats via enhanced
paroxetine exposure. Acta. Pharmacol. Sin. 2014, 35, 792–805. [CrossRef]
21. Hasegawa, Y.; Kishimoto, S.; Shibatani, N.; Inotsume, N.; Takeuchi, Y.; Fukushima, S. The disposition of
pravastatin in a rat model of streptozotocin-induced diabetes and organic anion transporting polypeptide 2
and multidrug resistance associated protein 2 expression in the liver. Biol. Pharm. Bull. 2010, 33, 153–156.
[CrossRef] [PubMed]
22. European Medicines Agency. Guideline on the Investigation of Drug Interactions. 2012. Available online: https:
//www.ema.europa.eu/documents/scientific-guideline/guideline-investigation-drug-interactions-en/pdf (accessed
on 15 February 2019).
23. US Food and Drug Administration, Center for Drug Evaluation and Research. Clinical Drug Interaction
Studies-Study Design, Data Analysis, and Clinical Implications Guideline for Industry. 2017. Available
online: https://www.fda.gov/downloads/drugs/guidances/ucm292362.pdf (accessed on 27 February 2020).
24. Yu, J.; Petrie, I.D.; Levy, R.H.; Ragueneau-Majlessi, I. Mechanisms and clinical significance of
pharmacokinetic-based drug-drug interactions with drug approved by the U.S. Food and Drug Administration
in 2017. Drug Metab. Dispos. 2019, 47, 135–144. [CrossRef]
25. Tornio, A.; Filppula, A.M.; Niemi, M.; Backman, J.T. Clinical studies on drug-drug interactions involving
metabolism and transport: Methodology, pitfalls, and interpretation. Clin. Pharmacol. Ther. 2019, 105,
1345–1361. [CrossRef] [PubMed]

87
Pharmaceutics 2020, 12, 417

26. Chen, X.P.; Tan, Z.R.; Huang, Z.; Ou-Yang, D.S.; Zhou, H.H. Isozyme-specific induction of low-dose aspirin
on cytochrome P450 in healthy subjects. Clin. Pharmacol. Ther. 2003, 73, 264–271. [CrossRef] [PubMed]
27. Li, M.-P.; Tang, J.; Zhang, Z.-L.; Chen, X.-P. Induction of both P-glycoprotein and specific cytochrome P450 by
aspirin eventually does not alter the antithrombotic effect of clopidogrel. Clin. Pharmacol. Ther. 2015, 97, 324.
[CrossRef] [PubMed]
28. Morgan, P.; Brown, D.G.; Lennard, S.; Anderton, M.J.; Barrett, J.C.; Eriksson, U.; Fidock, M.; Hamren, B.;
Johnson, A.; March, R.E.; et al. Impact of a five-dimensional framework on R&D productivity at AstraZeneca.
Nat. Rev. Drug Dispos. 2018, 17, 167–181.
29. Oh, J.; Shin, D.; Lim, K.S.; Lee, S.; Jung, K.-H.; Chu, K.; Hong, K.S.; Shin, K.-H.; Cho, J.-Y.; Yoon, S.H.; et al.
Aspirin decreases systemic exposure to clopidogrel through modulation of P-glycoprotein but does not alter
its antithrombotic activity. Clin. Pharmacol. Ther. 2014, 95, 608–616. [CrossRef]
30. Poulin, P. A paradigm shift in pharmacokinetic-pharmacodynamic (PKPD) modeling: Rule of thumb for
estimating free drug level in tissue compared with plasma to guide drug design. J. Pharm. Sci. 2015, 104,
2359–2368. [CrossRef]
31. Poulin, P.; Haddad, S. Advancing prediction of tissue distribution and volume of distribution of highly
lipophilic compounds from simplified tissue composition-based models as a mechanistic animal alternative
model. J. Pharm. Sci. 2012, 101, 2250–2261. [CrossRef]
32. Rurak, C.D.; Hack, C.E.; Robinson, P.J.; Mahle, D.A.; Gearheart, M. Predicting passive and active tissue:plasma
partition coefficient: Interindividual and interspecies variability. J. Pharm. Sci. 2014, 103, 2189–2198.
[CrossRef]
33. Shirata, Y.; Hoire, T.; Sigiyama, Y. Transporters as a determinant of drug clearance and tissue distribution.
Eur. J. Pharm. Sci. 2006, 27, 425–446.
34. Shitara, Y.; Maeda, K.; Ikejiri, K.; Yoshida, K.; Horie, T.; Sugiyama, Y. Clinical significance of organic anion
transporting polypeptides (OATPs) in drug disposition: Their roles in hepatic clearance and intestinal
absorption. Biopharm. Drug Dispos. 2013, 34, 45–78. [CrossRef]
35. Bosilkovska, M.; Samer, C.; Déglon, J.; Thomas, A.; Walder, B.; Desmeules, J.; Daali, Y. Evaluation of mutual
drug-drug interaction within Geneva cocktail for cytochrome P450 phenotying using innovative dried blood
sampling method. Basic Clin. Pharmacol. Ther. 2016, 119, 284–290. [CrossRef]
36. Fuhr, U.; Hsin, C.H.; Li, X.; Jabrane, W.; Sorgel, F. Assessment of pharmacokinetic drug-drug interactions in
humans: In vivo probe substrates for drug metabolism and drug transporter revised. Annu. Rev. Pharmacol.
Toxicol. 2019, 59, 507–536. [CrossRef]
37. Yoshida, K.; Zhao, P.; Zhang, L.; Abernethy, D.R.; Rekic, D.; Reynolds, K.S.; Galetin, A.; Huang, S.M.
In vitro-ion vivo extrapolation of metabolism- and transporter-mediated drug-drug interactions-overview of
basic prediction methods. J. Pharm. Sci. 2017, 106, 2209–2213. [CrossRef]
38. Gessnet, A.; Konig, J.; Fromm, M.F. Clinical aspects of transporter-mediated drug-drug interactions.
Clin. Pharmacol. Ther. 2019, 105, 1386–1394. [CrossRef]
39. Müller, F.; Fromm, M.F. Transporter-mediated drug-drug interactions. Pharmacogenomics 2011, 12, 1017–1037.
[CrossRef]
40. Zhang, L.; Huang, S.M.; Lesko, L.J. Transporter-mediated drug-drug interactions. Clin. Pharmacol. Ther.
2011, 89, 481–484. [CrossRef]
41. Morrissey, K.M.; Stocker, S.L.; Wittwer, M.B.; Xu, L.; Giacomini, K.M. Renal transporters in drug development.
Annu. Rev. Pharmacol. Toxicol. 2013, 53, 503–529. [CrossRef]
42. Yoshida, K.; Maeda, K.; Sugiyama, Y. Hepatic and intestinal drug transporters: Prediction of pharmacokinetic
effects caused by drug-drug interactions and genetic polymorphisms. Annu. Rev. Pharmacol. Toxicol. 2013,
53, 581–612. [CrossRef]
43. König, J.; Fabian Müller, F.; Fromm, M.F. Transporters and drug-drug interactions: Important determinants
of drug disposition and effects. Pharmacol. Rev. 2013, 65, 944–966. [CrossRef]
44. Zientek, M.A.; Youdim, K. Reaction phenotyping: Advances in the experimental strategies used to characterize
the contribution of drug-metabolizing enzymes. Drug Metab. Dispos. 2015, 43, 163–181. [CrossRef]
45. Almazroo, O.A.; Miah, M.K.; Venkataramanan, R. Drug metabolism in the liver. Clin. Liver Dis. 2017, 21,
1–20. [CrossRef]
46. Mittal, B.; Tulsyan, S.; Kumar, S.; Mittal, R.D.; Agarwal, G. Cytochrome P450 in cancer susceptibility and
treatment. Adv. Clin. Chem. 2015, 71, 77–139.

88
Pharmaceutics 2020, 12, 417

47. Olsen, L.; Oostenbrink, C.; Jorgensen, F.S. Prediction of cytochrome P450 mediated metabolism. Adv. Drug
Deliv. Rev. 2015, 86, 61–71. [CrossRef]
48. Achour, B.; Barber, J.; Rostami-Hodjegan, A. Expression of hepatic drug metabolizing cytochrome p450
enzymes and their intercorrelations: A meta-analysis. Drug Metab. Dispos. 2014, 42, 1349–1356. [CrossRef]
49. Wang, Q.; Jia, R.; Ye, C.; Garcia, M.; Li, J.; Hidalgo, I.J. Glucuronidation and sulfation of 7-hydroxycoumarin
in liver matrices from human, dog, monkey, rat, and mouse. In Vitro Cell Dev. Biol. Anim. 2005, 41, 97–103.
[CrossRef]
50. Terada, T.; Hira, D. Intestinal and hepatic drug transporters: Pharmacokinetic, pathophysiological, and
pharmacogenetic roles. J. Gastroenterol. 2015, 50, 508–519. [CrossRef]
51. Ito, S.; Kusuhara, H.; Yokochi, M.; Toyoshima, J.; Inoue, K.; Yuasa, H.; Sugiyama, Y. Competitive inhibition of
the luminal efflux by multidrug and toxin extrusions, but not basolateral uptake by organic cation transporter
2, is the likely mechanism underlying the pharmacokinetic drug-drug interactions caused by cimetidine in
the kidney. J. Pharmacol. Exp. Ther. 2012, 340, 393–403. [CrossRef]
52. Kalliokoski, A.; Niemi, M. Impact of OATP transporters on pharmacokinetics. Br. J. Pharmacol. 2009, 158,
693–705.
53. Chang, J.H.; Ly, J.; Plise, E.; Zhang, X.; Messick, K.; Wright, M.; Cheong, J. Differential effects of rifampin and
ketoconazole on the blood and liver concentration of atorvastatin in wild-type and Cyp3a and Oatp1a/b
knockout mice. Drug Metab. Dispos. 2014, 42, 1067–1073.
54. Lepist, E.I.; Ray, A.S. Renal drug-drug interactions: What we have learned and where we are going. Exp. Opin.
Drug Metab. Toxicol. 2012, 8, 433–448.
55. Lepist, E.-I.; Ray, A.S. Renal transporter-mediated drug-drug interactions: Are they clinically relevant?
J. Clin. Pharmacol. 2016, 56, S73–S81.
56. Liang, X.; Giacomini, K.M. Transporters involved in metformin pharmacokinetics and treatment response.
J. Pharm. Sci. 2017, 106, 2245–2250.
57. Stieger, B.; Mahdi, Z.M.; Jager, W. Intestinal and hepatocellular transporters: Therapeutic effects and drug
interactions of herbal supplements. Annu. Rev. Pharmacol. Toxicol. 2017, 57, 399–416.
58. Benet, L.Z.; Hoener, B.A. Changes in plasma protein binding have little clinical relevance. Clin. Pharmacol.
Ther. 2002, 71, 115–121.
59. Liu, X.; Wright, M.; Hop, C.E. Rational use of plasma protein and tissue binding data in drug design. J. Med.
Chem. 2014, 57, 8238–8248.
60. Hochman, J.; Tang, C.; Prueksaritanont, T. Drug-drug interactions related to altered absorption and plasma
protein binding: Theoretical and regulatory considerations, and an industry perspective. J. Pharm. Sci. 2015,
104, 916–929.
61. Kim, D.H. Gut microbiota-mediated drug-antibiotic interactions. Drug Metab. Dispos. 2015, 43, 1581–1589.
62. Kehrer, D.F.; Sparreboom, A.; Verweij, J.; de Bruijn, P.; Nierop, C.A.; van de Schraaf, J.; Ruijgrok, E.J.; de
Jonge, M.J. Modulation of irinotecan-induced diarrhea by cotreatment with neomycin in cancer patients.
Clin. Cancer Res. 2001, 7, 1136–1141.
63. Chiba, M.; Ishii, Y.; Sugiyama, Y. Prediction of hepatic clearance in human from in vitro data for successful
drug development. AAPS J. 2009, 11, 262–276.
64. Watanabe, T.; Kusuhara, H.; Sugiyama, Y. Application of physiologically based pharmacokinetic modeling
and clearance concept to drugs showing transporter-mediated distribution and clearance in humans.
J. Pharmacokinet. Pharmacodyn. 2010, 37, 575–590.
65. Motohashi, H.; Inui, K. Multidrug and toxin extrusion family SLC47: Physiological, pharmacokinetic and
toxicokinetic importance of MATE1 and MATE2-k. Mol. Aspects Med. 2013, 34, 661–668.
66. You, B.H.; Chin, Y.-W.; Kim, H.; Choi, H.S.; Choi, Y.H. Houttuynia cordata extract increased systemic exposure
and liver concentrations of metformin through OCTs and MATEs in rats. Phytother. Res. 2018, 32, 1004–1013.
67. Han, S.Y.; Chae, H.-S.; You, B.H.; Chin, Y.-W.; Kim, H.; Choi, H.S.; Choi, Y.H. Lonicera japonica extract increases
metformin distribution in the liver without change of systemic exposed metformin in rats. Phytother. Res.
2019, 32, 1004–1013.
68. Manohar, S.; Leung, N. Cisplatin nephrotoxicity: A review of the literature. J. Nephrol. 2018, 31, 15–25.
69. Sprowl, J.A.; van Doorn, L.; Hu, S.; van Gerven, L.; de Bruijm, P.; Li, L.; Gibson, A.A.; Mathijssen, R.H.;
Sparreboom, A. Conjunctive therapy of cisplatin with the OCT2 inhibitor cimetidine: Influence on antitumor
efficacy and systemic clearance. Clin. Pharmacol. Ther. 2013, 94, 585–592.

89
Pharmaceutics 2020, 12, 417

70. Frost, C.E.; Byon, W.; Song, Y.; Wang, J.; Schuster, A.E.; Boyd, R.A.; Zhang, D.; Yu, C.; Dias, C.; Shenker, A.;
et al. Effect of ketoconazole and diltiazem on the pharmacokinetics of apixaban, an oral direct factor Xa
inhibitor. Br. J. Clin. Pharmacol. 2015, 79, 838–846.
71. Hartter, S.; Koenen-Bergmann, M.; Sharma, A.; Nehmiz, G.; Lemke, U.; Timmer, W.; Reilly, P.A. Decrease in
the oral bioavailability of dabigatran etexilate after co-medication with rifampicin. Br. J. Clin. Pharmacol.
2012, 74, 490–500.
72. Greiner, B.; Eichelbaum, M.; Fritz, P.; Kreichgauer, H.-P.; von Richter, O.; Zundler, J.; Kroemer, H.K. The role
of intestinal P-glycoprotein in the interaction of digoxin and rifampin. J. Clin. Invest. 1999, 104, 147–153.
73. Sadeque, A.J.; Wandel, C.; He, H.; Shah, S.; Wood, A.J. Increased drug delivery to the brain by P-glycoprotein
inhibition. Clin. Pharmacol. Ther. 2000, 68, 231–237.
74. Allred, A.J.; Bowe, C.J.; Park, J.W.; Peng, B.; Williams, D.D.; Wire, M.B.; Lee, E. Eltrombopag increases plasma
rosuvastatin exposure in healthy volunteers. Br. J. Clin. Pharmacol. 2011, 72, 321–329.
75. Elsby, R.; Martin, P.; Surry, D.; Sharma, P.; Fenner, K. Solitary inhibition of the breast cancer resistance protein
efflux transporter results in a clinically significant drug-drug interaction with rosuvastatin by causing up to
a 2-fold increase in statin exposure. Drug Metab. Dispos. 2016, 44, 398–408.
76. Rengelshausen, J.; Goggelmann, C.; Burhenne, J.; Riedel, K.D.; Ludwig, J.; Weiss, J.; Mikus, G.; Walter-Sack, I.;
Haefeli, W.E. Contribution of increased oral bioavailability and reduced nonglomerular renal clearance of
digoxin to the digoxin-clarithromycin interaction. Br. J. Clin. Pharmacol. 2003, 56, 32–38.
77. Asberg, A.; Hartmann, A.; Fjeldså, E.; Bergan, S.; Holdaas, H. Bilateral pharmacokinetic interaction between
cyclosporine A and atorvastatin in renal transplant recipients. Am. J. Transplant. 2001, 1, 382–386.
78. Asberg, A. Interactions between cyclosporin and lipid-lowering drugs: Implications for organ transplant
recipients. Drugs 2003, 63, 367–378.
79. Maeda, K.; Ikeda, Y.; Fujita, T.; Yoshida, K.; Azuma, Y.; Haruyama, Y.; Yamane, N.; Kumagai, Y.; Sugiyama, Y.
Identification of the rate-determining process in the hepatic clearance of atorvastatin in a clinical cassette
microdosing study. Clin. Pharmacol. Ther. 2011, 90, 575–581.
80. Mazzu, A.L.; Lasseter, K.C.; Shamblen, E.C.; Agarwal, V.; Lettieri, J.; Sundaresen, P. Itraconazole alters the
pharmacokinetics of atorvastatin to a greater extent than either cerivastatin or pravastatin. Clin. Pharmacokinet.
Ther. 2000, 68, 391–400.
81. Markert, C.; Schweizer, Y.; Hellwig, R.; Wirsching, T.; Riedel, K.D.; Burhenne, J.; Weiss, J.; Mikus, G.;
Haefeli, W.E. Clarithromycin substantially increases steady-state bosentan exposure in healthy volunteers.
Br. J. Clin. Pharmacol. 2014, 77, 141–148.
82. Fattinger, K.; Funk, C.; Pantze, M.; Weber, C.; Reichen, J.; Stieger, B.; Meier, P.J. The endothelin antagonist
bosentan inhibits the canalicular bile salt export pump: A potential mechanism for hepatic adverse reactions.
Clin. Pharmacol. Ther. 2001, 69, 223–231.
83. Preston, C.L. Stockley’s Drug Interactions, 11th ed.; Pharmaceutical Press: London, UK, 2016.
84. Prueksaritanont, T.; Chu, X.; Evers, R.; Klopfer, S.O.; Caro, L.; Kothare, P.A.; Dempsey, C.; Rasmussen, S.;
Houle, R.; Chan, G.; et al. Pitavastatin is a more sensitive and selective organic anion-transporting polypeptide
1B clinical probe than rosuvastatin. Br. J. Clin. Pharmacol. 2014, 78, 587–598.
85. Jacobson, T.A. Comparative pharmacokinetic interaction profiles of pravastatin, simvastatin, and atorvastatin
when coadministered with cytochrome P450 inhibitors. Am. J. Cardiol. 2004, 94, 1140–1146.
86. Simonson, S.G.; Raza, A.; Martin, P.D.; Mitchell, P.D.; Jarcho, J.A.; Brown, C.D.; Windass, A.S.; Schneck, D.W.
Rosuvastatin pharmacokinetics in heart transplant recipients administered an antirejection regimen including
cyclosporine. Clin. Pharmacol. Ther. 2004, 76, 167–177.
87. He, J.; Yu, Y.; Prasad, B.; Link, J.; Miyaoka, R.S.; Chen, X.; Unadkat, J.D. PET imaging of Oatp mediated
hepatobiliary transport of [(11)C] rosuvastatin in the rat. Mol. Pharm. 2014, 11, 2745–2754.
88. Schneck, D.W.; Birmingham, B.K.; Zalikowski, J.A.; Mitchell, P.D.; Wang, Y.; Martin, P.D.; Lasseter, K.C.;
Brown, C.D.; Windass, A.S.; Raza, A. The effect of gemfibrozil on the pharmacokinetics of rosuvastatin. Clin.
Pharmacol. Ther. 2004, 75, 455–463.
89. Maeda, K.; Tian, Y.; Fujita, T.; Ikeda, Y.; Kumagai, Y.; Kondo, T.; Tanabe, K.; Nakayama, H.; Horita, S.;
Kusuhara, H.; et al. Inhibitory effects of p-aminohippurate and probenecid on the renal clearance of adefovir
and benzylpenicillin as probe drugs for organic anion transporter (OAT) 1 and OAT3 in humans. Eur. J.
Pharm. Sci. 2014, 59, 94–103.

90
Pharmaceutics 2020, 12, 417

90. Leahey, E.B.; Bigger, J.T.; Butler, V.P.; Reiffel, J.A.; O’Connell, G.C.; Scaffidi, L.E.; Rottman, J.N.
Quinidine-digoxin interaction: Time course and pharmacokinetics. Am. J. Cardiol. 1981, 48, 1141–1146.
91. Moore, K.H.; Yuen, G.J.; Raasch, R.H.; Eron, J.J.; Martin, D.; Mydlow, P.K.; Hussey, E.K. Pharmacokinetics of
lamivudine administered alone and with trimethoprim-sulfamethoxazole. Clin. Pharmacol. Ther. 1996, 59,
550–558.
92. Müller, F.; Konig, J.; Hoier, E.; Mandery, K.; Fromm, M.F. Role of organic cation transporter OCT2 and
multidrug and toxin extrusion proteins MATE1 and MATE2-K for transport and drug interactions of the
antiviral lamivudine. Biochem. Pharmacol. 2013, 86, 808–815.
93. Grun, B.; Kiessling, M.K.; Burhenne, J.; Riedel, K.D.; Weiss, J.; Rauch, G.; Haefeli, W.E.; Czock, D.
Trimethoprim-metformin interaction and its genetic modulation by OCT2 and MATE1 transporters. Br. J.
Clin. Pharmacol. 2013, 76, 787–796.
94. Müller, F.; Pontones, C.A.; Renner, B.; Mieth, M.; Hoier, E.; Auge, D.; Maas, R.; Zolk, O.; Fromm, M.F.
N1-methylnicotinamide as an endogenous probe for drug interactions by renal cation transporters: Studies
on the metformin-trimethoprim interaction. Eur. J. Clin. Pharmacol. 2015, 71, 85–94.
95. Song, I.H.; Zong, J.; Borland, J.; Jerva, F.; Wynne, B.; Zamek-Gilszczynski, M.J.; Humphreys, J.E.; Bowers, G.D.;
Choukour, M. The effect of dolutegravir on the pharmacokinetics of metformin in healthy subjects. J. Acquir.
Immune. Defic. Syndr. 2016, 72, 400–407.
96. Li, F.; Xu, D.; Shu, N.; Zhong, Z.; Zhang, M.; Liu, C.; Ling, Z.; Liu, L.; Liu, X. Co-administration of paroxetine
increased the systemic exposure of pravastatin in diabetic rats due to the decrease in liver distribution.
Xenobiotica 2015, 45, 794–802.
97. Grube, M.; Köck, K.; Oswald, S.; Draber, K.; Meissner, K.; Eckel, L.; Böhm, M.; Felix, S.B.; Vogelgesang, S.;
Jedlitschky, G.; et al. Organic anion transporting polypeptide 2B1 is a high-affinity transporter for atorvastatin
and is expressed in the human heart. Clin. Pharmacol. Ther. 2006, 80, 607–620.
98. Shin, E.; Shin, N.; Oh, J.-H.; Lee, Y.-J. High-dose metformin may increase the concentration of atorvastatin in
the liver by inhibition of multidrug resistance-associated protein 2. J. Pharm. Sci. 2017, 106, 961–967.
99. Li, L.; Lei, H.; Wang, W.; Du, W.; Yuan, J.; Yuan, J.; Tu, M.; Zhou, H.; Zeng, S.; Jiang, H. Co-administration
of nuciferine reduced the concentration of metformin in liver via differential inhibition of hepatic drug
transporter OCT1 and MATE1. Biopharm. Drug Dispos. 2018, 39, 411–419.
100. Somogyi, A.; Stockley, C.; Keal, J.; Rolan, P.; Bochner, F. Reduction of metformin renal tubular secretion by
cimetidine in man. Br. J. Clin. Pharmacol. 1987, 23, 545–551.
101. Stocker, S.L.; Morrissey, K.M.; Yee, S.W.; Castro, R.A.; Xu, L.; Dahlin, A.; Ramirez, A.H.; Roden, D.M.;
Wilke, R.A.; McCarty, C.A.; et al. The effect of novel promoter variants in MATE1 and MATE2 on the
pharmacokinetics and pharmacodynamics of metformin. Clin. Pharmacol. Ther. 2013, 93, 186–194.
102. Kusuhara, H.; Ito, S.; Kumagai, Y.; Jiang, M.; Shiroshita, T.; Moriyama, Y.; Inoue, K.; Yuasa, H.; Sugiuama, Y.
Effects of a MATE protein inhibitor, pyrimethamine, on the renal elimination of metformin at oral microdose
and at therapeutic dose in healthy subjects. Clin. Pharmacol. Ther. 2011, 89, 837–844.
103. Ito, S.; Kusuhara, H.; Kuroiwa, Y.; Wu, C.; Moriyama, Y.; Inoue, K.; Kondo, T.; Yuasa, H.; Nakayama, H.;
Horita, S.; et al. Potent and specific inhibition of mMate1-mediated efflux of type 1 organic cations in the
liver and kidney by pyrimethamine. J. Pharmakos. Exp. Ther. 2010, 333, 341–350.
104. Choi, Y.H.; Lee, U.; Lee, B.K.; Lee, M.G. Pharmacokinetic interaction between itraconazole and metformin in
rats: Competitive inhibition of metabolism of each drug by each other via hepatic and intestinal CYP3A1/2.
Br. J. Pharmacol. 2010, 161, 815–829.
105. Bechmann, K.A.; Lewis, J.D. Predicting inhibitory drug–drug interactions and evaluating drug interaction
reports using inhibition constants. Ann. Pharmacother. 2005, 39, 1064–1072.
106. Ma, Y.; Khojasteh, S.C.; Hop, C.E.; Erickson, H.K.; Polson, A.; Pillow, T.H.; Yu, S.F.; Wang, H.; Dragovich, P.S.;
Zhang, D. Antibody drug conjugates differentiate uptake and DNA alkylation of pyrrolobenzodiazepines in
tumors from organs of xenograft mice. Drug Metab. Dispos. 2016, 44, 1958–1962.
107. Zhang, D.; Yu, S.F.; Ma, Y.; Xu, K.; Dragovich, P.S.; Pillow, T.H.; Liu, L.; Del Rosario, G.; He, J.; Pei, Z.;
et al. Chemical structure and concentration of intratumor catabolites determine efficacy of antibody drug
conjugates. Drug Metab. Dispos. 2016, 44, 1517–1523.
108. Yamazaki, M.; Suzuki, H.; Sugiyama, Y. Recent advances in carrier-mediated hepatic uptake and biliary
excretion of xenobiotics. Pharm. Res. 1996, 13, 497–513.

91
Pharmaceutics 2020, 12, 417

109. Shitara, Y.; Sugiyama, Y. Pharmacokinetic and pharmacodynamic alterations of 3-hydroxy-3-methylglutaryl

coenzyme A (HMG-CoA) reductase inhibitors: Drug–drug interactions and interindividual differences in
transporter and metabolic enzyme functions. Pharmacol. Ther. 2006, 112, 71–105.
110. Anitha, N.; Rao, J.; Kavimani, S.; Himabindu, V. Pharmacodynamic drug interaction of metformin with
statins in rats. J. Pharmacol. Toxicol. 2008, 3, 409–413.
111. Bjornsson, E.; Jacobsen, E.I.; Kalaitzakis, E. Hepatotoxicity associated with statins: Reports of idiosyncratic
liver injury post-marketing. J. Hepatol. 2012, 56, 374–380.
112. Rose, R.H.; Neuhoff, S.; Abduljalil, K.; Chetty, M.; Rostami-Hodjegan, A.; Jamei, M. Application of a
physiologically based pharmacokinetic model to predict OATP1B1-related variability in pharmacodynamics
of rosuvastatin. CPT Pharmacomet. Syst. Pharmacol. 2014, 3, e124.
113. Sundelin, E.; Gormsen, L.C.; Jensen, J.B.; Vendelbo, M.H.; Jakobsen, S.; Munk, O.L.; Christensen, M.;
Brøsen, K.; Frøkiaer, J.; Jessen, N. Genetic polymorphisms in organic cation transporter 1 attenuates hepatic
metformin exposure in humans. Clin. Pharmacol. Ther. 2017, 102, 841–848.
114. Jensen, J.B.; Sundelin, E.I.; Jakobsen, S.; Gormsen, L.C.; Munk, O.L.; Frokier, J.; Jessen, N. [11C]-labeled
metformin distribution in the liver and small intestine using dynamic positron emission tomography in mice
demonstrates tissue-specific transporter dependency. Diabetes 2016, 65, 1724–1730.
115. Shu, Y.; Sheardown, S.A.; Brown, C.; Owen, R.P.; Zhang, S.; Castro, R.A.; Ianculescu, A.G.; Yue, L.; Lo, J.C.;
Burchard, E.G.; et al. Effect of genetic variation in the organic cation transporter 1 (OCT1) on metformin
action. J. Clin. Invest. 2007, 117, 1422–1431.
116. Vildhede, A.; Nguyen, C.; Erickson, B.K.; Kunz, R.C.; Jones, R.; Kimoto, E. Comparison of proteomic
quantification approaches for hepatic drug transporters: Multiplexed global quantitation correlates with
targeted proteomic quantitation. Drug Metab. Dispos. 2018, 46, 692–696.
117. Couto, N.; Al-Majdoub, Z.M.; Gibson, S.; Davies, P.J.; Achour, B.; Harwood, M.D.; Carlson, G.; Barber, J.;
Rostami-Hodjegan, A.; Warhurst, G. Quantitative proteomics of clinically relevant drug-metabolizing
enzymes and drug transporters and their inter-correlations in the human small intestine. Drug Metab. Dispos.
2020, 48, 407.
118. Pena, A.; Liu, P.; Derendorf, H. Microdialysis in peripheral tissues. Adv. Drug Deliv. Rev. 2000, 45, 189–216.
119. Dahyot, C.; Marchand, S.; Bodin, M.; Debeane, B.; Mimoz, O.; Couet, W. Application of basic pharmacokinetic
concepts to analysis of microdialysis data: Illustration with imipenem muscle distribution. Clin. Pharmacokinet.
2008, 47, 181–189.
120. Tuntland, T.; Ethell, B.; Kosaka, T.; Blasco, F.; Zhang, R.X.; Jain, M.; Gould, T.; Hoffmaster, K. Implementation
of pharmacokinetic and pharmacodynamic strategies in early research phases of drug discovery and
development at Novartis Institute of Biomedical Research. Front. Pharmacol. 2014, 28, 174.
121. Guo, Y.; Chu, X.; Parrott, N.J.; Brouwer, K.L.R.; Hsu, V.; Nagar, S.; Matsson, P.; Sharma, P.; Snoeys, J.;
Sugiyama, Y.; et al. Advancing predictions of tissue and intracellular drug concentrations using in vitro,
imaging and physiologically based pharmacokinetic modeling approaches. Clin. Pharmacol. Ther. 2018, 104,
865–889.

© 2020 by the author. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

92
 pharmaceutics

Article
Effect of Rumex Acetosa Extract, a Herbal Drug, on the
Absorption of Fexofenadine
Jung Hwan Ahn † , Junhyeong Kim † , Naveed Ur Rehman, Hye-Jin Kim, Mi-Jeong Ahn
and Hye Jin Chung *
College of Pharmacy and Research Institute of Pharmaceutical Sciences, Gyeongsang National University,
Jinju 52828, Korea; anjh5803@naver.com (J.H.A.); jhk6914@naver.com (J.K.);
naveed.rehman50@gmail.com (N.U.R.); black200203@gmail.com (H.-J.K.); amj5812@gnu.ac.kr (M.-J.A.)
* Correspondence: hchung@gnu.ac.kr; Tel.: +82-55-772-2430
† These authors contributed equally to this work.

Received: 29 April 2020; Accepted: 10 June 2020; Published: 12 June 2020

Abstract: Herbal drugs are widely used for the auxiliary treatment of diseases. The pharmacokinetics
of a drug may be altered when it is coadministered with herbal drugs that can affect drug absorption.
The effects of herbal drugs on absorption must be evaluated. In this study, we investigated the effects
of Rumex acetosa (R. acetosa) extract on fexofenadine absorption. Fexofenadine was selected as a model
drug that is a substrate of P-glycoprotein (P-gp) and organic anion transporting polypeptide 1A2
(OATP1A2). Emodine—the major component of R. acetosa extract—showed P-gp inhibition in vitro
and in vivo. Uptake of fexofenadine via OATP1A2 was inhibited by R. acetosa extract in OATP1A2
transfected cells. A pharmacokinetic study showed that the area under the plasma concentration–time
curve (AUC) of fexofenadine was smaller in the R. acetosa extract coadministered group than in the
control group. R. acetosa extract also decreased aqueous solubility of fexofenadine HCl. The results of
this study suggest that R. acetosa extract could inhibit the absorption of certain drugs via intervention
in the aqueous solubility and the drug transporters. Therefore, R. acetosa extract may cause drug
interactions when coadministered with substrates of drug transporters and poorly water-soluble
drugs, although further clinical studies are needed.

Keywords: P-glycoprotein (P-gp); organic anion transporting polypeptide 1A2 (OATP1A2);

Rumex acetosa; pharmacokinetics; fexofenadine; drug interaction

1. Introduction
Oral drug administration is a preferred route, offering the advantages of convenience and
safety. Many drug interactions with foods and other drugs occur via alteration of drug absorption.
There are absorptive transporters, such as organic anion transporting polypeptide (OATP) and secretory
transporters, including P-glycoprotein (P-gp), associated with drug absorption. To improve drug
therapy, it is necessary to investigate possible interactions mediated by transporters that could alter
systemic exposure of drugs.
P-gp, belonging to the ATP binding cassette superfamily, is an ATP-dependent efflux protein
that excretes drugs out of cells [1–3]. P-gp is an important factor limiting the absorption of drugs
and plays a key role in drug distribution and resistance [3,4]. For example, P-gp overexpression
induced by a hypoxic environment in many cancers decreases the effects of chemotherapy [5,6].
Furthermore, drug–drug interactions may occur when substrates of P-gp (e.g., cimetidine, digoxin,
doxorubicin, fexofenadine, and vinblastine) are coadministered with inhibitors of P-gp (e.g., atorvastatin,
ketoconazole and quinidine) or inducers of P-gp (e.g., rifampin and clotrimazole) [7,8]. The OATP
family is also an important transporter for drug disposition. The OATP members of the solute
carrier (SLC) family, contributes to the uptake of substrates, including endogenous compounds

93
Pharmaceutics 2020, 12, 547; doi:10.3390/pharmaceutics12060547 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 547

and drugs [9,10]. Drug–drug interactions and food–drug interactions mediated by these two active
transporters—P-gp and OATP—have been reported. In addition, a study on medication use patterns
revealed that 50% of 2590 study participants had taken at least one prescription drug during the week
prior to the study, and 16% of them had taken one or more herbals/supplements [11,12]. Given that
St. John’s wort was found to increase P-gp expression [13], it is necessary to evaluate the effects of
herbal supplements on these transporters. Despite the widespread use of herbal drugs in combination
with drugs, there has been little research on the interactions between drugs and herbal medicines.
This study investigated the effects of Rumex acetosa (R. acetosa) extract on P-gp and OATP1A2 in vitro
and on fexofenadine absorption in vivo. R. acetosa, used in folk remedies for skin diseases, has been
singled out as a natural herbal medicine for its potential to be used in combination with fexofenadine [14].
R. acetosa is widely distributed in eastern Asia and decoction of this plant has been used for the treatment
of several health disorders such as fever, gastro-intestinal problems, inflammatory diseases. It is
belonging in the Polygonaceae family, known to produce many biologic metabolites [15]. Particularly,
R. acetosa is rich in anthraquinones and flavonoids that have anti-inflammatory and antiproliferative
effects [16,17]. Emodin, a major anthraquinone component of R. acetosa extract, is reported that has the
potential for P-gp mediated drug interaction [18] and has various pharmacological effects, such as
antidiabetic [19] and anticancer activities [20].
Fexofenadine, a selective histamine H1 receptor antagonist, is widely used for seasonal allergic
rhinitis and chronic idiopathic urticarial treatment [21]. There is no evidence for cardiotoxicity
associated with fexofenadine, the active metabolite of terfenadine, even though terfenadine is not used
anymore due to the risk of cardiac arrhythmia. Fexofenadine was selected as a model drug that is
a marker substrate of P-gp [22] and OATP1A2 [23]. Fexofenadine is considered a good model drug,
because only around 5% of its dose is metabolized and most of the dose is excreted into urine (11%)
and feces (80%) as the unchanged form [24,25], which means that metabolism can be excluded in
interpreting the pharmacokinetics of fexofenadine.
To date, there have been many drug interaction studies involving P-gp or OATP. However, there
have been few studies concerning drug interactions with herbal medicines involving both P-gp and
OATP1A2. Furthermore, it has been reported that the emodin acts on P-gp as an inducer [26] or an
inhibitor [18]. Our results clarify the inhibitory effect of emodin on the P-gp through in vitro and in vivo
study. In addition, our findings include the fact that R. acetosa extract could affect drug absorption via
intervention in the OATP-mediated influx and the aqueous solubility. These results indicate that the
effects of herbal medicines such as plant extracts, on drug absorption must be considered in terms of
not only efflux through P-gp, but also OATP-mediated influx and the aqueous solubility.

2. Materials and Methods

2.1. Chemicals and Reagents

Fexofenadine hydrochloride and emodin were purchased from Tokyo Chemical Industry (Tokyo,
Japan). Dimethyl sulfoxide (DMSO), terfenadine, verapamil, Dulbecco’s modified Eagle’s medium
(DMEM) with high glucose, MEM non-essential amino acid solution (NEAA) and glutamine
were purchased from Sigma-Aldrich (St. Louis, MO, USA). HPLC grade acetonitrile and water
were purchased from Fisher Scientific Korea (Seoul, Korea). Emodin, emodin-8-O-β-d-glucoside,
chrysophanol, chrysophanol-8-O-β-d-glucoside, physcion and physcion-8-O-β-d-glucoside isolated
from R. acetosa were obtained from the pharmacognosy laboratory of the College of
Pharmacy at Gyeongsang National University (Jinju, Korea) [27]. Fetal bovine serum (FBS),
N-2-hydroxyethylpiperazine–N -2-ethanesulfonic acid (HEPES) and Hanks’ balanced salt solution
(HBSS) were purchased from Corning (Manassas, VA, USA). Penicillin–streptomycin, Opti-MEM and
0.25% (w/v) trypsin–EDTA were purchased from Gibco (Carlsbad, CA, USA). Phosphate buffered
saline (PBS) was purchased from Welgene (Gyeongsan, Korea). An MDR assay kit (fluorometric) was
purchased from Abcam (Cambridge, UK).

94
Pharmaceutics 2020, 12, 547

2.2. R. acetosa Extract

The R. acetosa extract was prepared by previously reported procedure [27]. Briefly, the dried
whole part of R. acetosa was extracted with 70% ethanol. The extraction was performed by the Soxhlet
extractor for 3 h at 80 ◦ C. The extract was filtered and lyophilized.
The total phenol content and total flavonoid content of R. acetosa extract were 74.5 mg GAE (gallic
acid equivalent)/g of dry weight and 180.3 μg QAE (quercetin equivalent)/g of dry weight, respectively.
The contents of anthraquinones in R. acetosa extract were determined by HPLC. The contents of
emodin, emodin-8-O-β-d-glucoside, chrysophanol, chrysophanol-8-O-β-d-glucoside, physcion and
physcion-8-O-β-d-glucoside in R. acetosa extract were 0.94 ± 0.15%, 1.29 ± 0.06%, 0.68 ± 0.09%,
0.77 ± 0.12%, 0.17 ± 0.02% and 0.41% ± 0.05% (w/w), respectively. The values were expressed as mean
± standard deviation.

2.3. Cell Culture

The Caco-2 (HTB-37™) cells were purchased from the American Type Culture Collection (ATCC,
Manassas, VA, USA). OATP1A2/SLCO1A2 transfected HEK293 cells were purchased from Corning
(New York, NY, USA). The Caco-2 cells were cultured in high glucose added DMEM with 10% FBS,
1% NEAA, 10-mM HEPES, 4-mM glutamine, 100 U/mL of penicillin and 100 μg/mL of streptomycin,
and maintained in humidified 5% CO2 at 37 ◦ C. The medium of the Caco-2 cells was replaced 2–3 times
per week.
The transfected HEK293 cells were cultured in high glucose added DMEM with 10% FBS and
1% NEAA, and maintained in 8% CO2 with low humidity at 37 ◦ C for 4 h. After incubation for 4 h,
the medium of the transfected HEK293 cells was replaced with high glucose added DMEM with 10%
FBS, 1% NEAA and 2-mM sodium butyrate, and incubated for 24 h.

2.4. Cytotoxicity Assay

The cytotoxicity of R. acetosa extract on Caco-2 cells and HEK293 cells was measured using an
EZ-Cytox cell viability assay kit (Daeil Lab Service, Seoul, Korea). The cells were cultured in DMEM
containing 10% FBS, 1% NEAA, 10-mM HEPES, 100 U/mL of penicillin and 100-μg/mL streptomycin
without phenol red. The seeding density was 3 × 104 cells/well for Caco-2 cells and 2.5 × 104 cell/well
for HEK293 cells, respectively. The Caco-2 cells were incubated for 7 days and the HEK 293 cells were
incubated for 24 h after seeding. The medium was replaced with 50 μL of new medium containing
R. acetosa extract at the concentrations of 1, 2, 5, 10, 20, 50 and 100 μg/mL achieved the 0.5% of DMSO
content. After 15 min of incubation, 5 μL of EZ-Cytox reagent (water-soluble tetrazolium) was added
to the cells, and the cells were incubated for 3 h. Cell viability was calculated as a percentage of the
absorbance at 450 nm compared to untreated cells.

2.5. P-gp Inhibition Test of Anthraquinones and R. acetosa Extract

The P-gp inhibition effect of anthraquinones from R. acetosa was evaluated via MDR assay kit
using Caco-2 cells. It was reported that verapamil has concentration-dependent inhibition effects
on absorptive and secretory transporters. Accordingly, 100-μM verapamil was used as a positive
control [28]. Caco-2 cells were cultured in 96-well plates at a density of 5 × 105 cells/mL and incubated
in humidified 5% CO2 at 37 ◦ C for 24 h. They were treated with 6 test compounds (10 μM) [18] or
R. acetosa extract in HBSS and incubated for 15 min. The concentration levels of R. acetosa extract were 5,
10, 25 and 50 μg/mL. The MDR dye-loading solution was added at a volume of 100 μL and incubated.
Fluorescence intensity was detected with a microplate reader Synerge H1 (Biotek, Winooski, VT, USA)
at a wavelength of 490 nm for the excitation and 525 nm for the emission.

95
Pharmaceutics 2020, 12, 547

2.6. Fexofenadine Uptake Test Using OATP1A2/SLCO1A2 Transfected HEK293 Cells

The seeding density of the OATP1A2 overexpressed HEK293 cells was 105 cells/well. Verapamil
was used as a positive control with a concentration of 100 μM [28]. The cultured cells were washed
twice with warmed HBSS with 5-mM MES after removing the medium, then 15-μM fexofenadine was
treated with R. acetosa extract of 10, 20 and 50 μg/mL. After 15 min of incubation in 8% CO2 with low
humidity at 37 ◦ C, they were washed twice with cold HBSS. They were gently shaken after adding
120 μL of 50-ng/mL terfenadine in 80% acetonitrile. Terfenadine was used as an internal standard.
After centrifugation at 10,000× g for 5 min, 50 μL of supernatant was mixed with 50 μL of 5-mM
ammonium formate (pH 4). The liquid chromatography-tandem mass spectrometry (LC-MS/MS) was
used to quantify the fexofenadine uptake amount [28,29].

2.7. LC-MS/MS Analysis

The chromatographic analysis was performed using an Agilent 1260 series (Agilent, Germany)
HPLC system. Chromatographic separation was achieved from the Phoroshell® column (C18,
3.0 × 50 mm, 2.7 μm). The mobile phase consisted of 5-mM ammonium formate (pH 4) in water (A)
and acetonitrile (B). A gradient method was applied at a flow rate of 0.3 mL/min and, kept on the
column temperature at 25 ◦ C. The injection volume was 2 μL. An Agilent 6460 triple-quadruple mass
spectrometer (Agilent Technologies, Singapore) with an electrospray ionization (ESI) source was used
to detect the signal. It was operated in positive ion mode on multiple reaction monitoring (MRM).
The monitored ions of fexofenadine and internal standard (terfenadine) were m/z 502→466 and m/z
472→436 [30,31], respectively. The collision energy and fragmentor of the ions were 25 V and 175 V for
fexofenadine, and 25 V and 130 V for terfenadine, respectively. The data were acquired and processed
using Mass Hunter Workstation B.06.00 software (Agilent Technologies, Singapore).

2.8. Animal Study

2.8.1. Animals
Male Sprague-Dawley rats (9 weeks, weighing 300 ± 50 g) were purchased from Koatech
(Pyeongtaek, Korea). The rats were acclimated in the Animal Laboratory (Gyeongsang National
University) under controlled condition of temperature (between 20 and 23 ◦ C) and humidity (50% ± 5%)
and allowed free access to food and water for 7 days. All rats were allowed to recover for 1 day after
cannulation into the carotid artery. The rats were fasted for 12 h with free access to water, before the
pharmacokinetic experiments.

2.8.2. Pharmacokinetic Study

The pharmacokinetic study was performed on a rat model. The dose of R. acetosa extract evaluated
was 2 g/kg, the maximum dose without the toxicity in rats (unpublished data). The selected oral dose
of emodin was 11 mg/kg that inhibited P-gp mediated efflux in rats from the reported study [32].
All test compounds—including 11 mg/kg of emodin and 2 g/kg of R. acetosa extract suspended in 0.5%
carboxy methyl cellulose (CMC)—were administered orally to rats. Same volume of 0.5% CMC was
administered to the vehicle control group rats. After 30 min, a single dose of 10 mg/kg of fexofenadine
in 10% ethanol was orally administered to each group of rats [33]. Blood samples of 120 μL were
collected from the carotid artery at each time point (0, 0.25, 0.5, 0.75, 1, 1.5, 2, 4, 6, 8, 12 and 24 h) after
oral administration of fexofenadine. The samples were then immediately centrifuged at 10,000× g and
4 ◦ C for 10 min. All plasma samples were stored at −20 ◦ C. The plasma concentrations of fexofenadine
were determined by LC-MS/MS. All experimental procedures of the animal study were approved
(GNU-170705-R0030, 5 July 2017) by the Animal Care and Use Committee of Gyeongsang National
University, Korea.

96
Pharmaceutics 2020, 12, 547

2.8.3. Sample Preparation

The method of sample preparation was a modified method of Isleyen et al. [34] for determination
of fexofenadine plasma concentration. In summary, 50 μL of 50-ng/mL terfenadine in acetonitrile
solution was added to a 50-μL aliquot of plasma, then 20 μL of aqueous 13-μM formic acid solution was
added. After vortexing, 50 μL of extraction solvent (a mixture of dichloromethane, ethyl acetate, diethyl
ether at the ratio of 30:40:30, v/v/v) was added. The sample was then vortexed for 40 s. The protein
precipitation was performed via centrifugation at 10,000× g and 4 ◦ C for 5 min. The supernatant was
cooled at −80 ◦ C for 10 min. The upper fraction of the supernatant was transferred to a polypropylene
tube and evaporated with N2 gas. After being reconstituted with 200 μL of the mobile phase initial
composition [5-mM ammonium formate (at a pH of 4): acetonitrile = 60:40], an aliquot of 2 μL was
injected into LC-MS/MS.

2.9. Physicochemical Interaction Study

To investigate the possible physicochemical interactions between drug and R. acetosa extract,
Fourier transform infrared (FT-IR) spectrum measurement and solubility test were carried out.
FT-IR spectra of fexofenadine HCl, R. acetosa extract and mixture of fexofenadine-extract (1:1)
were measured by Nicolet iS 50 FT-IR spectrometer (Thermo Scientific, Waltham, MA, USA) with
attenuated total reflectance (ATR) mode.
The change on the solubility of fexofenadine after mixing with R. acetosa extract was tested.
The method was modified previously reported method [35,36]. Briefly, 200 μg of fexofenadine and
R. acetosa extract were placed in the tube after centrifugal vacuum evaporation of solvent. The control
group has fexofenadine only, and the mixed group has both fexofenadine and the extract. A 200-μL
aliquot of the simulated intestinal fluid (SIF, pH 6.8) without enzyme [37] was added to each tube.
The tubes were then incubated in a shaking water bath at 37 ◦ C for 12 h. The concentration of
fexofenadine was 1 mg/mL, corresponding to the orally administered concentration to the rats
(10 mg/5 mL/kg-fexofenadine with 5-mL/kg extract, total 10 mL). After the incubation, the tubes were
centrifuged at 10,000× g for 10 min. The supernatant was filtered, diluted with mobile phase, and
analyzed by LC-MS/MS.

2.10. Statistical Analysis

The statistical analysis was performed using one-way analysis of variance (ANOVA) followed by a
Dunnett’s multiple comparison test. A p-value of less than 0.05 was considered statistically significant.

3. Results

3.1. Cytotoxicity Assay

Cell viability was expressed as a percentage of the absorbance value obtained from the media only
treated control group (Figure 1). Even though there were statistically significant differences in Caco-2
cell viability between control and R. acetosa treated groups at concentrations of 20, 10, 2 and 1 μg/mL,
the cell viability values were high enough to study (96.4% ± 1.3%, 95.0% ± 2.1%, 95.0% ± 1.2% and
95.5% ± 2.1% at concentrations of 20, 10, 2 and 1 μg/mL, respectively). It is suggested that there is a
negligible cytotoxic effect of R. acetosa on the Caco-2 cells at the concentration range of 1 to 100 μg/mL.
There was no significant difference on the cell viability on HEK293 cells at the concentration
ranges of 1 to 50 μg/mL. The cytotoxic effect of R. acetosa was only detected on HEK293 cells at a
concentration of 100 μg/mL with the value of 80.9% ± 11.7%. This result suggests a dose window of R.
acetosa extract for the experiment using HEK293 cells. It also indicates that the inhibitory effect of R.
acetosa on fexofenadine uptake discussed in Section 3.2 was not due to the cytotoxic effects of R. acetosa
on HEK293 cells at the concentration range tested.

97
Pharmaceutics 2020, 12, 547

 

RIFRQWURO

RIFRQWURO
 

 

 
&RQ 9HK        &RQ 9HK       
Rumex acetosa Rumex acetosa
H[WUDFWμJP/ H[WUDFWμJP/

$ %

Figure 1. Cytotoxicity of R. acetosa extract in (A) Caco-2 cells and (B) HEK293 cells (n = 6). Con—media
only treated control; Veh—vehicle treated group; *—p < 0.05 compared to media only treated
control group.

3.2. P-gp Inhibition Test of Anthraquinones and R. acetosa Extract

To determine inhibitory effect of anthraquinones on P-gp, an MDR kit was used. The accumulated
amount of fluorescent dye in the cells represented the P-gp inhibition activity. The measured
fluorescence intensity is expressed as a percentage of the fluorescence intensity in the control group
and is shown in Figure 2. The verapamil, chrysophanol-8-O-β-d-glucoside and emodin treated groups
displayed significantly different fluorescence intensities in comparison to those of the control group.
However, the chrysophanol-8-O-β-d-glucoside and emodin treated groups showed significantly higher
fluorescence intensities than the control group, with average values of 121.4% ± 2.3% and 147.2% ± 12.4%,
respectively (mean ± standard deviation). This result suggests that chrysophanol-8-O-β-d-glucoside
and emodin affect the efflux of fluorescent dye from Caco-2 cells through P-gp inhibition. This is
consistent with previous findings that emodin inhibits P-gp [18]. It is thus reasonable to suggest that
herbal drug containing chrysophanol-8-O-β-d-glucoside and emodin may also inhibit P-gp.


)OXRUHVFHQFHLQWHQVLW\

RIFRQWURO







&RQ 9HU      
&RPSRXQG
Figure 2. P-gp inhibitory effect of anthraquinones in Caco-2 cells. Cells were treated with 10-μM
anthraquinones or 100-μM verapamil (n = 3). Con—vehicle treated control; Ver—verapamil;
1—chrysophanol; 2—chrysophanol-8-O-β-d-glucoside; 3—emodin; 4—emodin-8-O-β-d-glucoside;
5—physcion; 6—physcion-8-O-β-d-glucoside; *—p < 0.05 compared to control group.

The effects of R. acetosa extract on the P-gp were also assessed using an MDR kit. The measured
fluorescence intensity is expressed as a percentage of the fluorescence intensity in the control group and
is shown in Figure 3. There was no significant difference in fluorescence intensity between the control
and the R. acetosa extract treated group. The significant inhibitory effect at the 95% confidence interval
was only detected in the verapamil group used as a positive control. Although R. acetosa extract contains

98
Pharmaceutics 2020, 12, 547

chrysophanol-8-O-β-d-glucoside and emodin at concentrations of 0.77% ± 0.12% and 0.94% ± 0.15%

(w/w), respectively [27], inhibitory effects on P-gp could not be detected from R. acetosa extract at the
concentrations tested in this assay. The concentrations of emodin and chrysophanol-8-O-β-d-glucoside
in R. acetosa extract may not be high enough to inhibit P-gp in Caco-2 cells.


)OXRUHVFHQFHLQWHQVLW\


RIFRQWURO






&RQ 9HU    
Rumex acetosaH[WUDFW
FRQFHQWUDWLRQμJP/

Figure 3. P-gp inhibition test of R. acetosa extract using an MDR kit in Caco-2 cells (n = 6). Con—vehicle
treated control; Ver—100-μM verapamil; *—p < 0.05 compared to control group.

3.3. Fexofenadine Uptake Test with OATP1A2/SLCO1A2 Transfected HEK293 Cells

The decreased fexofenadine uptake in the OATP1A2/SLCO1A2 transfected HEK293 cells
represented the inhibitory effects on OATP1A2. The accumulated amount of fexofenadine in the
OATP/SLCO1A2 transfected cells (the control group) was higher than that of untransfected cells
(untransfected control group), which means that the OATP1A2 gene was transfected and expressed
sufficiently in the control group. In addition, the fexofenadine uptake was lower in the verapamil
cotreated group than in the transfected control group. In the R. acetosa extract co-treated group,
the uptake amounts of fexofenadine in the OATP1A2 transfected HEK293 cells were significantly lower
than that in the control group (Figure 4). This result suggests that R. acetosa extract could affect the
absorption of fexofenadine through the inhibition of OATP1A2.


IH[RIHQDGLQHQJFHOO
8SWDNHDPRXQWRI









&RQ 1RY 9HU   
Rumex acetosaH[WUDFW
FRQFHQWUDWLRQμJP/

Figure 4. Inhibitory effect of R. acetosa extract on fexofenadine uptake in OATP1A2/SLCO1A2 transfected

HEK293 cells (n = 6). Con—OATP1A2/SLCO1A2 transfected control; Nov—untransfected control;
Ver—100-μM verapamil; *—p < 0.05 compared to control group.

99
Pharmaceutics 2020, 12, 547

3.4. Pharmacokinetic Study

The mean arterial plasma concentration–time profiles of fexofenadine after the pharmacokinetic
study of fexofenadine in a rat drug interaction model, are shown in Figure 5. The pharmacokinetic
parameters of fexofenadine after oral coadministration of vehicle, emodin and R. acetosa extract are
shown in Table 1. The area under the plasma concentration–time curve (AUC) values of fexofenadine
were 222.0 ± 85.5 ng·h/mL in the control group and 411.9 ± 189.1 ng·h/mL in the emodin coadministered
group. In the emodin group, the absorption of fexofenadine was significantly higher, with a larger
fexofenadine AUC and no difference in Tmax . However, the fexofenadine AUC was lower, with a value
of 132.0 ± 50.5 ng·h/mL, in the 2 g/kg R. acetosa extract coadministered group. Because fexofenadine
is characterized by limited metabolism, it is probable that the lower fexofenadine AUC is due to the
inhibitory effect of R. acetosa extract on absorption. Consequently, R. acetosa extract could inhibit the
absorption of fexofenadine. Together with the results of the in vitro assay, this suggests that R. acetosa
extract inhibits the intracellular uptake of fexofenadine via an intervention in OATP1A2.


3ODVPDFRQFHQWUDWLRQRIIH[RIHQDGLQHQJP/

9HKLFOH
(PRGLQ
5XPH[DFHWRVDH[WUDFW









      
7LPHK

Figure 5. Mean plasma concentration–time profiles of fexofenadine (ng/mL) after oral coadministration
of fexofenadine (10 mg/kg) with vehicle (•; n = 6), emodin (11 mg/ kg, ; n = 6) and R. acetosa extract
(2 g/kg, ; n = 6) to rats. Bars represent standard deviations.

Table 1. Pharmacokinetic parameters of fexofenadine after oral coadministration of fexofenadine

(10 mg/kg) with vehicle (control), emodin and R. acetosa extract to rats. Values represent means ±
standard deviations for AUC0-24 h and Cmax (ng/mL), median (range) for Tmax .

Emodin R. acetosa Extract

Control
Parameters 11 mg/kg 2 g/kg
(n = 6)
(n = 6) (n = 6)
AUC0-24 h (ng·h/mL) 222.0 ± 85.5 411.9 ± 189.1 * 132.0 ± 50.5
Cmax (ng/mL) 36.4 ± 22.8 53.4 ± 33.9 32.9 ± 28.5
Tmax (h) 0.75 (0.5–1) 0.75 (0.5–1) 0.75 (0.25–8)
AUC0-24 h , total area under the plasma concentration–time curve from time zero to 24 h; Cmax —maximum plasma
concentration; Tmax —time to reach Cmax ; *—p < 0.05 compared to vehicle only control group.

100
Pharmaceutics 2020, 12, 547

3.5. Physicochemical Interaction Study

To evaluate the possible physicochemical interactions between R. acetosa extract and fexofenadine, FT-IR
spectra of extract, fexofenadine and mixture were measured and are shown in Figure 6. The FT-IR spectrum
of fexofenadine HCl showed the characteristic absorption bands at 3291.03 (OH stretching), 2936.14 (CH
stretching), 2639.82 (OH of carboxylate), 1698.68 (CO stretching), 1448.00, 1403.11 (C=C stretching of aromatic
ring), 1167.57 (CO stretching of tertiary alcohol) and 1067.94 (CO stretching of secondary alcohol) [38,39].
According to Figure 6, the mixture of R. acetosa extract and fexofenadine HCl showed the same bands
compared to the pure fexofenadine HCl. It suggests that there is no significant physical interaction between
fexofenadine molecule and R. acetosa extract component on fexofenadine functional groups.

Figure 6. FT-IR spectra of (A) fexofenadine, (B) a mixture of fexofenadine and R. acetosa extract and
(C) R. acetosa extract.

101
Pharmaceutics 2020, 12, 547

However, there was significant difference on the solubility of fexofenadine after incubation with
the extract (Table 2). The average solubilities of fexofenadine in SIF were 1.03 ± 0.04 mg/mL and
0.83 ± 0.10 mg/mL without and with R. acetosa extract, respectively. This result indicates that R. acetosa
extract could alter the solubility of fexofenadine and lead to precipitation in gastro-intestinal tract.

Table 2. The solubility of fexofenadine HCl in simulated intestinal fluid (SIF) with and without R.
acetosa extract.

Solubility Without R. acetosa Extract(n = 3) With R. acetosa Extract(n = 3)

Fexofenadine HCl concentration (mg/mL) 1.03 ± 0.04 0.83 ± 0.10 *
*—p < 0.05 compared to without R. acetosa extract.

4. Discussion
Pharmacokinetic drug interactions involving drug absorption should be considered for optimum
drug therapy, apart from the drug interactions attributed to the oxidative metabolism via the CYP-450
system of different isozymes [40]. Ostensibly harmless natural products—such as juices, fruits,
vegetables and herbal products in the form of ayurvedic medicine—have been reported in several studies
to potentially cause many drug interactions affecting drug absorption mediated by transporters [41,42].
For example, emodin—a potential antineoplastic drug and a major component of the Rhamnus, Rumex,
Aloe, Rheum and Cassia species—has been reported to be a possible P-gp inducer [26] or an inhibitor [18].
This study evaluated the effects of R. acetosa extract on the drug transporters discussed above, as
well as its potential for drug interactions, while presenting a clear view of the interactions of emodin
with the transporter P-gp. The major six anthraquinones present in R. acetosa were shown in our
previous study [27]. A prior cytotoxicity assay was performed to establish the working range for
the extract suitable for optimal viability of the cells during the experiment. Afterwards, the effects
of these six anthraquinones on P-gp were demonstrated individually with an MDR assay kit using
Caco-2 cells. Verapamil, being an inhibitor of P-gp, served as a positive control. Only groups treated
with chrysophanol-8-O-β-d-glucoside and emodin showed higher fluorescence intensity than the
control group, with average values of 121.4% ± 2.3% and 147.2% ± 12.4%, respectively, suggesting
P-gp inhibition. This result is consistent with those obtained in a study by Min et al. [18], in which
emodin was shown to inhibit P-gp. On the other hand, the results from the P-gp inhibition test of
R. acetosa extract suggest no significant inhibition of the efflux transporter, as opposed to the emodin
and chrysophanol-8-O-β-d-glucoside, which in contrast showed significant inhibition of the P-gp
transporter when treated individually. A possible explanation is that the emodin content may not be
high enough to exert its inhibitory effect in the extract. Chemical contents of herbal plant extracts
can vary depending on various factors such as climate, harvesting seasons and extraction solvent.
The probability of inhibition of P-gp by R. acetosa extract cannot be ruled out.
OATP1A2—the uptake transporter used in our in vitro test—is widely expressed in the intestines
and serves as a major uptake mechanism for fexofenadine [43,44]. Sometimes, a substrate of P-gp—such
as this study’s selected model drug, fexofenadine—can also be a substrate for the OATP uptake
transporter [43,44], making it necessary to differentiate between the contributions of P-gp and OATP to
potential drug interactions and those of other simultaneously administered drugs that could affect
these transporters. Therefore, our in vitro studies were also performed with HEK293 cells transfected
with the polypeptide transporter OATP1A2. R. acetosa extract was found to inhibit the uptake of
fexofenadine through in vitro studies. In other words, the uptake of fexofenadine by OATP1A2 into
cells declined when R. acetosa extract was used as a co-treatment. This result suggests that R. acetosa
extract can affect the absorption of fexofenadine through the inhibition of OATP1A2.
A pharmacokinetic study was designed to verify the results of our in vitro study in view of
the observed herbal extract’s drug interactions at the uptake transporter for fexofenadine in rats.
Rat model is considered unsuitable to predict metabolic drug interaction in human [45]. However,
there is a correlation in drug intestinal permeability with both carrier-mediated absorption and passive

102
Pharmaceutics 2020, 12, 547

diffusion mechanisms between rat and human [46]. Because the property of our selected model
drug, fexofenadine, has little metabolism, it is reasonable to use the rat model for predicting the
intervention of extract on absorption. All rats were divided into 3 groups: an emodin administration
group, an R. acetosa administration group and a control group. Eleven milligrams per kilogram of
emodin, 2 g/kg of R. acetosa extract and 0.5% CMC as a control was administered orally to each group.
Fexofenadine at the dose of 10 mg/kg was given orally to each group after 30 min. The results showed
a smaller AUC of fexofenadine (132.1 ± 50.3 ng·h/mL) in the R. acetosa group in comparison to that
of the control group, in which the AUC was 222.0 ± 92.1 ng·h/mL. These results suggest decreased
absorption of fexofenadine in the rats treated with R. acetosa extract. In other words, the gut uptake
transporter OATP1A2, which is responsible for fexofenadine absorption, was inhibited, as predicted
by the in vitro results. Moreover, the alteration on the solubility of fexofenadine was also observed by
R. acetosa extract through the physicochemical interaction study. The FT-IR spectra results suggest that
there is no functional group interaction between fexofenadine and the component of R. acetosa extract.
The fexofenadine solubility in SIF changed from 1.03 ± 0.04 mg/mL to 0.83 ± 0.10 mg/mL after mixing
with the extract. It means that the solubility alteration could also be the reason for the decreased
fexofenadine AUC by R. acetosa extract because fexofenadine HCl is Biopharmaceutics Classification
System (BCS) class 3 drug with high solubility and low permeability. Drug interactions due to changes
in solubility can be avoided by adjusting the administration time. R. acetosa extract contains many
kinds of compounds, not only anthraquinones, but also flavonoids and polysaccharides [15]. They
have also the possibility of interference with the drug absorption through the intervention to the
transporters [47,48]. Particularly, one of the flavonoids of R. acetosa extract, epicatechin-3-O-gallate [49],
also has an inhibitory effect on the OATP1A2 [50]. Moreover, there was the possibility that R. acetosa
extract may change the gastric emptying time [51,52] and the pH in the gastro-intestinal tract when
coadministered with the fexofenadine. The effects of anthraquinones on OATP have been rarely
reported. Further studies are needed to elucidate the components in R. acetosa extract responsible
for inhibition of fexofenadine absorption. Meanwhile, emodin increased the AUC for fexofenadine,
possibly via the inhibitory effect on an efflux transporter of fexofenadine, P-gp [32], the effect of which
on the uptake transporter of fexofenadine has yet to be fully understood.
Given the evidence from both in vitro and in vivo studies, R. acetosa extract should be used with
caution when substrates of the drug transporters or poorly water-soluble drugs are prescribed.

5. Conclusions
The present study evaluated the effects of R. acetosa extract on 2 active transporters, P-gp and
OATP1A2 and the resulting effects on fexofenadine absorption through in vitro and in vivo studies.
The findings suggest that emodin can enhance fexofenadine absorption via an inhibitory effect on P-gp.
In addition, R. acetosa extract could decrease the absorption of fexofenadine via intervention in the
aqueous solubility and the drug transporters.

Author Contributions: Conceptualization, H.J.C.; methodology, H.J.C.; validation, J.K. and N.U.R.; formal
analysis, J.K.; investigation, J.H.A., J.K. and H.-J.K.; resources, M.-J.A. and H.J.C.; data curation, J.H.A., J.K. and
H.J.C.; writing—original draft preparation, J.K. and N.U.R.; writing—review and editing, H.J.C.; visualization,
J.K.; supervision, H.J.C. and M.-J.A.; project administration, H.J.C.; funding acquisition, H.J.C. All authors have
read and agreed to the published version of the manuscript.
Funding: This research was funded by the National Research Foundation of Korea (NRF) grant funded by the
Korean government (MSIP; Ministry of Science & ICT), Grant Number 2017R1C1B5017343.
Conflicts of Interest: The authors declare no conflict of interest.

103
Pharmaceutics 2020, 12, 547

References
1. Dean, M.; Hamon, Y.; Chimini, G. The human ATP-binding cassette (ABC) transporter superfamily. J. Lipid
Res. 2001, 42, 1007–1017. [CrossRef] [PubMed]
2. Anderle, P.; Niederer, E.; Rubas, W.; Hilgendorf, C.; Spahn-Langguth, H.; Wunderli-Allenspach, H.;
Merkle, H.P.; Langguth, P. P-Glycoprotein (P-gp) mediated efflux in Caco-2 cell monolayers: The influence
of culturing conditions and drug exposure on P-gp expression levels. J. Pharm. Sci. 1998, 87, 757–762.
[CrossRef] [PubMed]
3. Fardel, O.; Lecureur, V.; Guillouzo, A. The P-glycoprotein multidrug transporter. Gen. Pharm. 1996, 27,
1283–1291. [CrossRef]
4. Chan, L.M.; Lowes, S.; Hirst, B.H. The ABCs of drug transport in intestine and liver: Efflux proteins limiting
drug absorption and bioavailability. Eur. J. Pharm. Sci. 2004, 21, 25–51. [CrossRef]
5. Talks, K.L.; Turley, H.; Gatter, K.C.; Maxwell, P.H.; Pugh, C.W.; Ratcliffe, P.J.; Harris, A.L. The expression and
distribution of the hypoxia-inducible factors HIF-1alpha and HIF-2alpha in normal human tissues, cancers,
and tumor-associated macrophages. Am. J. Pathol 2000, 157, 411–421. [CrossRef]
6. Comerford, K.M.; Wallace, T.J.; Karhausen, J.; Louis, N.A.; Montalto, M.C.; Colgan, S.P. Hypoxia-inducible
factor-1-dependent regulation of the multidrug resistance (MDR1) gene. Cancer Res. 2002, 62, 3387–3394.
7. Kim, R.B. Drugs as P-glycoprotein substrates, inhibitors, and inducers. Drug Metab. Rev. 2002, 34, 47–54.
[CrossRef]
8. Palmeira, A.; Sousa, E.; Vasconcelos, M.H.; Pinto, M.M. Three decades of P-gp: Skimming through several
generations and scaffolds. Curr. Med. Che 2012, 19, 1946–2025. [CrossRef]
9. Shitara, Y.; Maeda, K.; Ikejiri, K.; Yoshida, K.; Horie, T.; Sugiyama, Y. Clinical significance of organic anion
transporting polypeptides (OATPs) in drug disposition: Their roles in hepatic clearance and intestinal
absorption. Biopharm. Drug Dispos. 2013, 34, 45–78. [CrossRef]
10. Yu, J.; Zhou, Z.; Tay-Sontheimer, J.; Levy, R.H.; Ragueneau-Majlessi, I. Intestinal drug interactions mediated
by OATPs: A systematic review of preclinical and clinical findings. J. Pharm. Sci. 2017, 106, 2312–2325.
[CrossRef]
11. Agbabiaka, T.B.; Spencer, N.H.; Khanom, S.; Goodman, C. Prevalence of drug-herb and drug-supplement
interactions in older adults: A cross-sectional survey. Br. J. Gen. Pr. 2018, 68, e711–e717. [CrossRef] [PubMed]
12. Kaufman, D.W.; Kelly, J.P.; Rosenberg, L.; Anderson, T.E.; Mitchell, A.A. Recent patterns of medication
use in the ambulatory adult population of the United States: The Slone survey. JAMA 2002, 287, 337–344.
[CrossRef] [PubMed]
13. Durr, D.; Stieger, B.; Kullak-Ublick, G.A.; Rentsch, K.M.; Steinert, H.C.; Meier, P.J.; Fattinger, K. St John’s
Wort induces intestinal P-glycoprotein/MDR1 and intestinal and hepatic CYP3A4. Clin. Pharm. 2000, 68,
598–604. [CrossRef] [PubMed]
14. Gescher, K.; Hensel, A.; Hafezi, W.; Derksen, A.; Kuhn, J. Oligomeric proanthocyanidins from Rumex acetosa
L. inhibit the attachment of herpes simplex virus type-1. Antivir. Res. 2011, 89, 9–18. [CrossRef] [PubMed]
15. Vasas, A.; Orban-Gyapai, O.; Hohmann, J. The genus Rumex: Review of traditional uses, phytochemistry
and pharmacology. J. Ethnopharmacol. 2015, 175, 198–228. [CrossRef] [PubMed]
16. Kucekova, Z.; Mlcek, J.; Humpolicek, P.; Rop, O.; Valasek, P.; Saha, P. Phenolic compounds from Allium
schoenoprasum, Tragopogon pratensis and Rumex acetosa and their antiproliferative effects. Molecules 2011,
16, 9207–9217. [CrossRef]
17. Bae, J.Y.; Lee, Y.S.; Han, S.Y.; Jeong, E.J.; Lee, M.K.; Kong, J.Y.; Lee, D.H.; Cho, K.J.; Lee, H.S.; Ahn, M.J.
A comparison between water and ethanol extracts of Rumex acetosa for protective effects on gastric ulcers in
mice. Biomol. Ther. (Seoul) 2012, 20, 425–430. [CrossRef] [PubMed]
18. Min, H.; Niu, M.; Zhang, W.; Yan, J.; Li, J.; Tan, X.; Li, B.; Su, M.; Di, B.; Yan, F. Emodin reverses leukemia
multidrug resistance by competitive inhibition and downregulation of P-glycoprotein. PLoS ONE 2017, 12,
e0187971. [CrossRef]
19. Feng, Y.; Huang, S.L.; Dou, W.; Zhang, S.; Chen, J.H.; Shen, Y.; Shen, J.H.; Leng, Y. Emodin, a natural product,
selectively inhibits 11beta-hydroxysteroid dehydrogenase type 1 and ameliorates metabolic disorder in
diet-induced obese mice. Br. J. Pharm. 2010, 161, 113–126. [CrossRef]
20. Hsu, S.C.; Chung, J.G. Anticancer potential of emodin. BioMedicine (Taipei) 2012, 2, 108–116. [CrossRef]

104
Pharmaceutics 2020, 12, 547

21. Simpson, K.; Jarvis, B. Fexofenadine: A review of its use in the management of seasonal allergic rhinitis and
chronic idiopathic urticaria. Drugs 2000, 59, 301–321. [CrossRef] [PubMed]
22. Tahara, H.; Kusuhara, H.; Fuse, E.; Sugiyama, Y. P-glycoprotein plays a major role in the efflux of fexofenadine
in the small intestine and blood-brain barrier, but only a limited role in its biliary excretion. Drug Metab.
Dispos. 2005, 33, 963–968. [CrossRef] [PubMed]
23. Shimizu, M.; Fuse, K.; Okudaira, K.; Nishigaki, R.; Maeda, K.; Kusuhara, H.; Sugiyama, Y. Contribution of
OATP (organic anion-transporting polypeptide) family transporters to the hepatic uptake of fexofenadine in
humans. Drug Metab. Dispos. 2005, 33, 1477–1481. [CrossRef] [PubMed]
24. Molimard, M.; Diquet, B.; Benedetti, M.S. Comparison of pharmacokinetics and metabolism of desloratadine,
fexofenadine, levocetirizine and mizolastine in humans. Fundam. Clin. Pharm. 2004, 18, 399–411. [CrossRef]
[PubMed]
25. Prescribing Information for Allegra®(fexofenadine hydrochloride). Available online: https://www.accessdata.
fda.gov/drugsatfda_docs/label/2003/20872se8-003,20625se8-010_allegra_lbl.pdf (accessed on 16 March 2020).
26. Huang, J.; Guo, L.; Tan, R.; Wei, M.; Zhang, J.; Zhao, Y.; Gong, L.; Huang, Z.; Qiu, X. Interactions between
emodin and efflux transporters on rat enterocyte by a validated ussing chamber technique. Front. Pharm.
2018, 9, 646. [CrossRef]
27. Ullah, H.M.A.; Kim, J.; Rehman, N.U.; Kim, H.J.; Ahn, M.J.; Chung, H.J. A simple and sensitive
liquid chromatography with tandem mass spectrometric method for the simultaneous determination
of anthraquinone glycosides and their aglycones in rat plasma: Application to a pharmacokinetic study of
Rumex acetosa extract. Pharmaceutics 2018, 10, 100. [CrossRef]
28. Petri, N.; Tannergren, C.; Rungstad, D.; Lennernas, H. Transport characteristics of fexofenadine in the Caco-2
cell model. Pharm. Res. 2004, 21, 1398–1404. [CrossRef]
29. Rebello, S.; Zhao, S.; Hariry, S.; Dahlke, M.; Alexander, N.; Vapurcuyan, A.; Hanna, I.; Jarugula, V. Intestinal
OATP1A2 inhibition as a potential mechanism for the effect of grapefruit juice on aliskiren pharmacokinetics
in healthy subjects. Eur. J. Clin. Pharm. 2012, 68, 697–708. [CrossRef]
30. Yamane, N.; Tozuka, Z.; Sugiyama, Y.; Tanimoto, T.; Yamazaki, A.; Kumagai, Y. Microdose clinical trial:
Quantitative determination of fexofenadine in human plasma using liquid chromatography/electrospray
ionization tandem mass spectrometry. J. Chromatogr B Anal. Technol Biomed. Life Sci 2007, 858, 118–128.
[CrossRef]
31. Bharathi, V.D.; Radharani, K.; Jagadeesh, B.; Ramulu, G.; Bhushan, I.; Naidu, A.; Mullangi, R. LC–MS–MS assay
for simultaneous quantification of fexofenadine and pseudoephedrine in human plasma. Chromatographia
2008, 67, 461–466. [CrossRef]
32. Li, X.; Hu, J.; Wang, B.; Sheng, L.; Liu, Z.; Yang, S.; Li, Y. Inhibitory effects of herbal constituents on
P-glycoprotein in vitro and in vivo: Herb-drug interactions mediated via P-gp. Toxicol. Appl. Pharm. 2014,
275, 163–175. [CrossRef] [PubMed]
33. Kamath, A.V.; Yao, M.; Zhang, Y.; Chong, S. Effect of fruit juices on the oral bioavailability of fexofenadine in
rats. J. Pharm. Sci. 2005, 94, 233–239. [CrossRef] [PubMed]
34. İşleyen, E.A.Ö.; Özden, T.; Özilhan, S.; Toptan, S. Quantitative determination of fexofenadine in human
plasma by HPLC-MS. Chromatographia 2007, 66, 109–113. [CrossRef]
35. Dai, W.G. In vitro methods to assess drug precipitation. Int. J. Pharm. 2010, 393, 1–16. [CrossRef] [PubMed]
36. Saito, Y.; Usami, T.; Katoh, M.; Nadai, M. Effects of thylakoid-rich spinach extract on the pharmacokinetics of
drugs in rats. Biol. Pharm. Bull. 2019, 42, 103–109. [CrossRef] [PubMed]
37. Stippler, E.; Kopp, S.; Dressman, J.B. Comparison of US pharmacopeia simulated intestinal fluid TS (without
pancreatin) and phosphate standard buffer pH 6.8, TS of the international pharmacopoeia with respect to
their use in in vitro dissolution testing. Dissolution Technol. 2004, 11, 6–10. [CrossRef]
38. Singh, B.; Saini, G.; Vyas, M.; Verma, S.; Thakur, S. Optimized chronomodulated dual release bilayer tablets
of fexofenadine and montelukast: Quality by design, development, and in vitro evaluation. Future J. Pharm.
Sci. 2019, 5, 5. [CrossRef]
39. Arefin, P.; Hasan, I.; Reza, M.S. Design, characterization and in vitro evaluation of HPMC K100 M CR loaded
fexofenadine HCl microspheres. Springerplus 2016, 5, 691. [CrossRef]
40. Hansten, P.D.; Levy, R.H. Role of P-glycoprotein and organic anion transporting polypeptides in drug
absorption and distribution. Clin. Drug Investig. 2001, 21, 587–596. [CrossRef]

105
Pharmaceutics 2020, 12, 547

41. Mallhi, T.H.; Sarriff, A.; Adnan, A.S.; Khan, Y.H.; Qadir, M.I.; Hamzah, A.A.; Khan, A.H. Effect of
fruit/vegetable-drug interactions on CYP450, OATP and p-glycoprotein: A systematic review. Trop. J. Pharm.
Res. 2015, 14, 1927–1935. [CrossRef]
42. Bailey, D.G. Fruit juice inhibition of uptake transport: A new type of food-drug interaction. Br. J. Clin. Pharm.
2010, 70, 645–655. [CrossRef] [PubMed]
43. Cvetkovic, M.; Leake, B.; Fromm, M.F.; Wilkinson, G.R.; Kim, R.B. OATP and P-glycoprotein transporters
mediate the cellular uptake and excretion of fexofenadine. Drug Metab. Dispos. 1999, 27, 866–871. [PubMed]
44. Niemi, M.; Kivisto, K.T.; Hofmann, U.; Schwab, M.; Eichelbaum, M.; Fromm, M.F. Fexofenadine
pharmacokinetics are associated with a polymorphism of the SLCO1B1 gene (encoding OATP1B1). Br. J.
Clin. Pharm. 2005, 59, 602–604. [CrossRef] [PubMed]
45. Jaiswal, S.; Sharma, A.; Shukla, M.; Vaghasiya, K.; Rangaraj, N.; Lal, J. Novel pre-clinical methodologies for
pharmacokinetic drug-drug interaction studies: Spotlight on "humanized" animal models. Drug Metab. Rev.
2014, 46, 475–493. [CrossRef] [PubMed]
46. Cao, X.; Gibbs, S.T.; Fang, L.; Miller, H.A.; Landowski, C.P.; Shin, H.C.; Lennernas, H.; Zhong, Y.; Amidon, G.L.;
Yu, L.X.; et al. Why is it challenging to predict intestinal drug absorption and oral bioavailability in human
using rat model. Pharm. Res. 2006, 23, 1675–1686. [CrossRef]
47. Mandery, K.; Bujok, K.; Schmidt, I.; Keiser, M.; Siegmund, W.; Balk, B.; Konig, J.; Fromm, M.F.; Glaeser, H.
Influence of the flavonoids apigenin, kaempferol, and quercetin on the function of organic anion transporting
polypeptides 1A2 and 2B1. Biochem. Pharm. 2010, 80, 1746–1753. [CrossRef]
48. Masumoto, K.; Quan, Z.; Ishiuchi, K.i.; Matsumoto, T.; Watanabe, J.; Makino, T. Drug interaction between
shoseiryuto extract or catechins and fexofenadine through organic-anion-transporting polypeptide 1A2
in vitro. Pharmacogn. Mag. 2019, 15, 304–308. [CrossRef]
49. Bicker, J.; Petereit, F.; Hensel, A. Proanthocyanidins and a phloroglucinol derivative from Rumex acetosa L.
Fitoterapia 2009, 80, 483–495. [CrossRef]
50. Roth, M.; Timmermann, B.N.; Hagenbuch, B. Interactions of green tea catechins with organic
anion-transporting polypeptides. Drug Metab. Dispos. 2011, 39, 920–926. [CrossRef]
51. Bajad, S.; Bedi, K.L.; Singla, A.K.; Johri, R.K. Piperine inhibits gastric emptying and gastrointestinal transit in
rats and mice. Planta Med. 2001, 67, 176–179. [CrossRef]
52. Hu, M.L.; Rayner, C.K.; Wu, K.L.; Chuah, S.K.; Tai, W.C.; Chou, Y.P.; Chiu, Y.C.; Chiu, K.W.; Hu, T.H. Effect of
ginger on gastric motility and symptoms of functional dyspepsia. World J. Gastroenter. 2011, 17, 105–110.
[CrossRef] [PubMed]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

106
 pharmaceutics

Article
Prevalence of Potential Drug–Drug Interaction Risk
among Chronic Kidney Disease Patients in
a Spanish Hospital
Gracia Santos-Díaz 1 , Ana María Pérez-Pico 2 , Miguel Ángel Suárez-Santisteban 1,3 ,
Vanesa García-Bernalt 3 , Raquel Mayordomo 4 and Pedro Dorado 1, *
1 Biosanitary Research Institute of Extremadura (INUBE), University of Extremadura, 06006 Badajoz, Spain;
grsantosd@unex.es (G.S.-D.); miguelangel.suarez@salud-juntaex.es (M.Á.S.-S.)
2 Department of Nursing, University of Extremadura, 10600 Plasencia, Spain; aperpic@unex.es
3 Nephrology Department, Virgen del Puerto Hospital, Servicio Extremeño de Salud, 10600 Plasencia, Spain;
vanesa.garcia@salud-juntaex.es
4 Department of Anatomy, Cellular Biology and Zoology, University of Extremadura, 10600 Plasencia, Spain;
rmayordo@unex.es
* Correspondence: pdorado@unex.es

Received: 30 June 2020; Accepted: 28 July 2020; Published: 30 July 2020

Abstract: Chronic kidney disease (CKD) is a major health problem worldwide and, in Spain, it is
present in 15.1% of individuals. CKD is frequently associated with some comorbidities and patients
need to be prescribed multiple medications. Polypharmacy increases the risk of adverse drug reactions
(ADRs). There are no published studies evaluating the prevalence of potential drug–drug interactions
(pDDIs) among CKD patients in any European country. This study was aimed to determine the
prevalence, pattern, and factors associated with pDDIs among CKD patients using a drug interactions
program. An observational cross-sectional study was carried out at Plasencia Hospital, located in
Spain. Data were collected among patients with CKD diagnoses and pDDIs were assessed by the
Lexicomp® Drug Interactions platform. Data were obtained from 112 CKD patients. A total number
of 957 prescribed medications were acknowledged, and 928 pDDIs were identified in 91% of patients.
Age and concomitant drugs were significantly associated with the number of pDDIs (p < 0.05).
According to the results, the use of programs for the determination of pDDIs (such as Lexicomp® ) is
recommended in the clinical practice of CKD patients in order to avoid serious adverse effects, as is
paying attention to contraindicated drug combinations.

Keywords: chronic kidney disease; drug–drug interactions; polypharmacy; adverse drug

reactions; Lexicomp

1. Introduction
As specified by the Kidney Disease Improving Global Outcomes (NKF KDIGO) guidelines [1],
chronic kidney disease (CKD) is defined as abnormalities of kidney structure or function, present for
more than three months, with implications for health. CKD is a general term for various heterogeneous
disorders affecting kidney structure and function with variable clinical presentations; in part, related to
the cause, severity, and rate of progression. The glomerular filtration rate (GFR) is generally accepted
as the best overall index of kidney function, and is classified into different stages (G1, G2, G3a, G3b,
G4, and G5). The diagnostic criteria of CKD are those denominated as kidney damage markers or a
threshold of GFR < 60 mL/min/1.73 m2 (GFR categories G3a–G5), or both, for more than three months.
CKD is a major health problem worldwide; in 2017, 1.2 million people died from CKD. Furthermore,
between 1990 and 2017, the global all-age mortality rate from CKD increased by 41.5% [2]. In Spain,

107
Pharmaceutics 2020, 12, 713; doi:10.3390/pharmaceutics12080713 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 713

CKD is present in 15.1% of individuals and this prevalence is more than three times higher in men
than in women (23.1% vs. 7.3%) and increases with age [3]. Diabetes and hypertension are the
main causes of CKD in all high-income and middle-income countries and in many low-income
countries [4]. Among other reasons, the prevalence of CKD is increasing worldwide due to the fact
that the prevalence of both hypertension and diabetes is also rising. Diabetes is expected to increase
by 69% in high-income countries and 20% in low-income and middle-income countries from 2010 to
2030 [5]. Regarding hypertension, it is predicted to increase by 60% from 2000 to 2025 [6]. Additionally,
CKD is also associated with other comorbidities such as dyslipidemia, hyperuricemia, or cardiovascular
disease [7], and patients need to be prescribed multiple medications.
Polypharmacy is usually defined as the concomitant prescription of five or more medications [8]
and it is a major risk factor of drug–drug interactions, which increases with the number of prescribed
drugs leading to 100% with eight or more medications [9]. The elderly are at risk for polypharmacy,
and this fact increases the risk of adverse drug reactions (ADRs) from 13 to 58% with two and five
medications, respectively. Seven or more medications increase the risk of ADRs to 82% [9].
Previous studies have evaluated the prevalence and severity of potential drug–drug interactions
(pDDIs) using different drug–drug interaction programs among CKD patients from Brazil [10,11],
India [12,13], Pakistan [14], Palestine [15], and Nigeria [16–19]; however, there are no published studies
evaluating the prevalence of pDDIs among CKD patients in any European country.
Lexicomp® (Wolters Kluwer Clinical Drug Information) is considered one of the best performing
drug–drug interaction programs and it was reported to be highly sensitive and specific (around 90–100%).
It focuses on the depth and duplication of information and it is a resource of choices for locating the
mechanism of a drug–drug interaction [20–22].
This study was aimed to determine the prevalence, pattern, and factors associated with potential
drug–drug interaction among CKD patients attending a hospital nephrology department using the
drug interaction program Lexicomp® .

2. Materials and Methods

2.1. Subjects
An observational cross-sectional study was carried out at Virgen del Puerto Hospital in Plasencia
(Cáceres, Spain). All participants were patients attended by the nephrology department, and were
invited to participate in the study. The inclusion criteria were: patients with CKD diagnosis, over the
age of 18, and having signed an informed consent form. Data were collected during 2019 and included:
age, gender, list of medications at the time of last clinic visit, comorbidities, and serum creatinine.
The study was performed in accordance with the principles of the Declaration of Helsinki of
1975, revised in 2013, and approved by the Clinical Research Ethics Committee, Cáceres (reference:
MASR/2016), and the Bioethics and Biosecurity Committee, University of Extremadura
(reference: 64/2016).
The serum creatinine value was used to calculate eGFR (estimated glomerular filtration rate
in mL/min/1.73 m2 ) and patients were classified following the criteria of the KDIGO Guideline 1
into different CKD stages: G1, eGFR ≥ 90; G2, eGFR 60–89; G3a, eGFR 45–59; G3b, eGFR 30–44;
G4, eGFR 15–29; G5, eGFR < 15.

2.2. Methods
The electronic drug–drug interactions (DDIs) checking platform Lexicomp® was used to evaluate
patient medication regimens for pDDIs. The Lexicomp® (Wolters Kluwer Health Inc. Riverwoods, IL,
USA) database system provides accurate information about the risk, type, mechanism, and pattern
of distribution of pDDIs. It also gives recommendations on how to prevent and manage DDIs if
they occur. This software identifies and classifies pDDIs into five types according to the degree
of clinical significance. Type A: no known interaction, Type B: minor or mild interaction, Type C:

108
Pharmaceutics 2020, 12, 713

moderate or significant interaction, Type D: major or serious interaction, and Type X: contraindication
or avoid combination.

2.3. Statistical Analysis

Descriptive statistics were used, and results were presented as percentages and frequencies.
ANOVA Kruskal–Wallis test or Mann–Whitney t-test analyses were performed to evaluate the effect of
covariates on the incidence of pDDIs. A p-value of less than 0.05 was considered statistically significant.
All statistical analyses were performed using IBM SPSS v.22 (SPSS Inc., Chicago, IL, USA).

3. Results

3.1. Clinical and Demographic and Characteristics of Patients

Data were obtained from 112 CKD patients, 69 (61.6%) females and 43 (38.4%) males. The mean
age of this study population was 77.1 ± 10.4 years: 11 patients (10.0%) were between 30 and 60 years,
8 (7.1%) were between 61 and 70 years, 44 (39.3%) were between 71 and 80 years, and 49 (43.7%) were
older than 80 years (Table 1).

Table 1. Characteristics of the study population (n = 112).

Characteristics N (%) or Mean ± SD

Female 69 (61.6)
Male 43 (38.4)
Mean age (years) 77.1 ± 10.4
Age group (years)
30–60 11 (10.0)
61–70 8 (7.1)
71–80 44 (39.3)
>80 49 (43.7)
CKD stage
1 10 (8.9)
2 15 (13.4)
3a 34 (30.3)
3b 33 (29.5)
4 15 (13.4)
5 5 (4.5)
Comorbidities 91 (81.2)
Hypertension 52 (46.4)
Diabetes mellitus 25 (22.3)
Dyslipidemia/hypercholesterolemia 33 (29.5)
Anemia 13 (11.6)
Hyperuricemia 11 (9.8)
Mean prescribed drugs per patient 8.6 ± 3.4
Number of prescribed drugs
≤5 22 (19.6)
6–10 59 (53.2)
11–15 29 (25.9)
≥16 2 (1.8)

The most common comorbid conditions (Table 1) were hypertension in 52 patients (46.4%),
diabetes mellitus in 25 (22.3%), dyslipidemia in 33 (29.5%), anemia in 13 (11.6%), and hyperuricemia in
11 (9.8%).

3.2. Prevalence and Pattern of Potential Drug–Drug Interactions

A total number of 957 prescribed medications were identified. The minimum number of
prescribed medications per patient was 1, the maximum was 17, and the mean number was

109
Pharmaceutics 2020, 12, 713

8.6 ± 3.4 medications. Only one patient was not taking any medication. The most commonly prescribed
medications were omeprazole (30.6%), acetaminophen (30.6%), salicylic acid (26.1%), bisoprolol (25.2%),
furosemide (22.5%), and allopurinol (21.6%).
Among 111 individuals 928 pDDIs were identified, and 67 (60.3%) patients showed 1–10 pDDIs,
while 34 (30.6%) presented more than 10. Only 10 patients (9%) did not have any interaction (Table 2).

Table 2. Frequency of potential drug–drug interactions (pDDIs) per patient (n = 111 *).

Number of pDDIs N %
None 10 9.0
1–5 40 36.0
6–10 27 24.3
11–15 15 13.5
16–20 11 9.9
21–25 4 3.6
>25 4 3.6
* One patient had not taken any drugs.

According to the Lexicomp® severity classification, 11 (1.2%) pDDIs were Type A (no known
interaction), 84 (9.1%) were Type B (mild severity), 717 (77.3%) were Type C (moderate severity),
106 (11.4%) were Type D (major severity), and 10 (1.1%) were Type X (avoid drug combination) (Table 3).

Table 3. Severity of potential drug–drug interactions (pDDIs; n = 928) among studied chronic kidney
disease (CKD) patients.

Severity of pDDIs N %
Type A (No known interaction) 11 1.2
Type B (mild severity) 84 9.1
Type C (moderate severity) 717 77.3
Type D (major severity) 106 11.4
Type X (avoid drug combination) 10 1.1

Table 4 shows the most frequent pDDIs by severity group: levothyroxine + omeprazole with
9 cases in the Type B group (10.7%), acenocoumarol + omeprazole with 11 cases in Type C (1.5%),
and acenocoumarol + allopurinol with 8 cases in Type D (7.5%).

Table 4. Most frequent potential drug–drug interactions (pDDIs) by severity group.

Severity of pDDI N PDDIs Frequency (%)

Type B (mild severity) 84 Levothyroxine + Omeprazole 10.7
Levothyroxine + Furosemide 9.5
Acenocoumarol + Spironolactone 7.1
Type C (moderate severity) 717 Acenocoumarol + Omeprazole 1.5
Ferrous Sulfate + Omeprazole 1.3
Metformin + Acetylsalicylic Acid 1.3
Type D (major severity) 106 Acenocoumarol + Allopurinol 7.5
Levothyroxine + Ferrous Sulfate 4.7
Tramadol + Trazodone 3.8

In addition, Type X (avoid drug combination) pDDIs were found in 10 CKD patients (Table 5).

110
Pharmaceutics 2020, 12, 713

Table 5. Potential drug–drug interactions Type X (avoid drug combination) found in the studied
CKD patients.

Drug–Drug Interaction Potential Adverse Effects Severity Reliability Rating

Aclidinium may enhance the anticholinergic
Amitriptyline + Aclidinium Major Fair
effect of Anticholinergic Agents
Alpha1-Blockers may enhance the
Doxazosin + Dutasteride and Tamsulosin antihypertensive effect of Major Fair
other Alpha1-Blockers
Alpha1-Blockers may enhance the
Dutasteride And Tamsulosin + Tamsulosin antihypertensive effect of Major Fair
other Alpha1-Blockers
Anticholinergic Agents may diminish the
Levosulpiride + Solifenacin Major Fair
therapeutic effect of Levosulpiride
Beta-Blockers (Nonselective) may diminish the
Aclidinium and Formoterol + Propranolol Major Fair
bronchodilatory effect of Beta2-Agonists
Dexketoprofen may enhance the adverse/toxic
Metamizole + Dexketoprofen effect of Nonsteroidal Major Fair
Anti-Inflammatory Agents
Tramadol and Acetaminophen Opioids (Mixed Agonist/Antagonist) may
Major Good
+ Buprenorphine diminish the analgesic effect of Opioid Agonists
Rivastigmine may enhance the bradycardic
Atenolol + Rivastigmine Moderate Fair
effect of Beta-Blockers
Rivastigmine may enhance the bradycardic
Bisoprolol + Rivastigmine Moderate Fair
effect of Beta-Blockers
Salicylates may enhance the adverse/toxic effect
of Dexketoprofen. Dexketoprofen may
Dexketoprofen + Acetylsalicylic Acid diminish the therapeutic effect of Salicylates. Major Fair
Salicylates may decrease the serum
concentration of Dexketoprofen

It was also observed that some drugs were present in a large number of pDDIs such as
hydrochlorothiazide (15%), acetylsalicylic acid (10%), or furosemide (9%). The most frequent drugs
present in pDDIs in the study group are shown in Figure 1.

Figure 1. Frequency of main drugs with potential drug–drug interactions (n = 928).

111
Pharmaceutics 2020, 12, 713

3.3. Factors Associated with Potential Drug–Drug Interactions on CKD Patients

Age and concomitant drugs were significantly associated with the number of pDDIs (p < 0.05;
Table 6). In contrast, demographic and clinical variables, such as gender, CKD stage, or the number of
chronic comorbid diseases were not significantly associated with the number of pDDIs (Table 6).

Table 6. Potential drug–drug interactions (pDDIs) among 111 * CKD patients according to demographic
and clinical variables groups.

Median
Variable Number p–Value 1
(25%–75% Percentile)
Gender 0.5768 2
Female 68 7 (2–12.7)
Male 43 7 (2–11)
Age 0.0191
30–60 11 2 (0–5)
61–70 8 3.5 (0.5–7.7)
71–80 44 8 (3.2–16)
>80 48 8 (3–11)
CKD stage ** 0.4930
G1 10 3.5 (0.7–12.7)
G2 15 8 (3–12)
G3a 34 5 (1–9.2)
G3b 33 8 (3–13)
G4 14 4 (3–13)
G5 5 11 (4–17.5)
Concomitant drugs <0.0001
≤5 21 1 (0–2)
6–10 59 6 (4–9)
11–15 29 16 (11–20)
>15 2 26 (16–36)
Chronic comorbid disease 0.0611
0 20 6.5 (3.2–11)
1 21 7 (2–12.5)
2 20 6.5 (3–10.7)
3 10 12 (7.2–20.5)
4 17 7 (1–17.5)
≥5 23 4 (1–7)
* One patient had not taken any drugs; ** according to the classification of chronic kidney disease from Kidney
Disease: Improving Global Outcomes (KDIGO); 1 ANOVA Kruskal–Wallis test; 2 Mann–Whitney t-test.

4. Discussion

4.1. Frequency and Severity of Potential Drug–Drug Interactions

Table 8 shows previous reports in which the prevalence and severity of pDDIs has been evaluated
on CKD patients. It is remarkable that there are two studies published in 2017 that were carried out in
different hospitals in Nigeria, with different numbers of individuals, but their results are practically
identical despite using different analysis tools [17,18].

112
 Table 7. Previous studies in which prevalence and severity of potential drug–drug interactions has been evaluated on CKD patients.

Number of CKD Patients on Number of Drugs Number of Patients

Years Software for Most Frequent Drug Combinations with
Study Patients Country Stage 5 or per Patient with Potential DDIs
(Mean ± SD) Potential DDI Potential DDIs (%)
(% Female) Hemodialysis (%) (Mean ± SD) Contraindicated (%)
Ascorbid Acid + Cyanocobalamine (12.4%)
Clonidine + Metoprolol (3.8%)
Rama et al. [12]. 205 (25.8%) 48.6 ± 16.2 India 68.5% Micromedex 12.1 ± 6.3 0 (0.0%)
Amlodipine + Metoprolol (3.4%)
Insulin + Metoprolol (2.9%)
Pharmaceutics 2020, 12, 713

Furosemide + Aspirin (7.8%)

Enalapril + Furosemide (5.9%)
Marquito et al. [10]. 558 (45.3%) n.s. Brazil 6.6% Micromedex 5.6 ± 3.2 5 (0.9%)
Captopril + Furosemide (5.1%)
Enalapril + Losartan (3.7%)
Calcium Carbonate + Atenolol (8.0%)
Calcium Carbonate + Ferrous Sulfate (8.0%)
Sgnaolin et al. [11]. 65 (50.8%) 59.1±14.7 Brazil 100% Micromedex 6.3 ± 3.1 2 (3.1%)
Calcium Carbonate + Ticlopidine (6.3%)
Enalapril + Eritropoietin (4.5%)
Sodium Bicarbonate + Ferrous Sulfate (8.9%)
Medscape Drug Calcium Carbonate + Ferrous Sulfate (5.5%)
Hegde et al. [13]. 120 (45%) 58.5 ± 8.4 India n.s. 9.4 ± 3.9 0 (0.0%)
interaction checker Aspirin + Carvedilol (5.5%)
Sodium Bicarbonate + Allopurinol (5.5%)
Calcium Carbonate + Amlodipine (12.3%)
Al-Ramahi et al. Calcium Carbonate + Aspirin (8.2%)
275 (45.1%) 50.7 ± 15.9 Palestina 100% LexiComp 7.9 ± 2.4 2 (0.7%)
[15]. Aspirin + Furosemide (7.9%)

113
Aspirin + Enoxaparin (4.3%)
Calcium Carbonate + Ferrous Sulfate (8.4%)
Olumuyiwa et al. Folic Acid + Furosemide (3.4%)
123 (33.3%) 53.8 ± 16.0 Nigeria 69.9% Lexi-Interact database 10.1 ± 4.0 1 (0.8%)
[18]. Calcium Carbonate + Calcidol (3.2%)
Vitamin E + Ferrous Sulfate (3.0%)
Calcium Carbonate + Ferrous Sulfate (9.9%)
Medscape Drug Folic Acid + Furosemide (3.4%)
Fasipe et al. [17]. 123 (48.8%) 53.8 ± 16.0 Nigeria 69.9% 10.3 ± 3.9 1 (0.8%)
interaction checker Calcium Carbonate + Calcidol (3.2%)
Vitamin E + Ferrous Sulfate (3.0%)
 Table 8. Previous studies in which prevalence and severity of potential drug–drug interactions has been evaluated on CKD patients.

Number of CKD Patients on Number of Drugs Number of Patients

Years Software for Most Frequent Drug Combinations with
Study Patients Country Stage 5 or per Patient with Potential DDIs
(Mean ± SD) Potential DDI Potential DDIs (%)
(% Female) Hemodialysis (%) (Mean ± SD) Contraindicated (%)
Ferrous Sulfate + Omeprazole (5.8%)
Micromedex Calcium/Vitamin D + Ciprofloxacin (4.8%)
Saleem et al. [14]. 209 (39.2%) 38.3 ± 16.8 Pakistan 74.2% n.s. 28 (13.4)
Drug-Reax Captopril + Furosemide (4.1%)
Calcium Gluconate + Ceftriaxone (3.6%)
Pharmaceutics 2020, 12, 713

Lisinopril + Furosemide (9.1%)

Medscape Drug
Furosemide + Calcium Carbonate (7.2%)
Adibe et al. [16]. 169 (52.1%) 51.0 ± 14.9 Nigeria 28.4% interaction 6.1 ± 2.0 0 (0.0%)
Calcium Carbonate + Lisinipril (6.1%)
checker
Aspirin + Furosemide (4.6%)
Calcium Carbonate + Ferrous Sulfate (45.8%)
Omnio drug
Okoro and Farate Lisinopril + Furosemide (7.7%)
201 (66%) 49.5 ± 14.5 Nigeria 69.2% interaction 5.8 ± 1.5 5 (2.5%)
[19]. Captopril + Furosemide (6.6%)
checker
Captopril + Spironolactone (6.6%)
Acenocoumarol + Omeprazole (1.1%)
Ferrous Sulfate + Omeprazole (1.0%)
Present study 111 (61.3%) 77.1 ± 10.4 Spain 4.5% LexiComp 8.6 ± 3.4 10 (9.0%)
Metformin + Aspirin (1.0%)
Levothyroxine + Omeprazole (1.0%)
n.s.: not studied.

114
Pharmaceutics 2020, 12, 713

As mentioned, most of the pDDIs were Type C (moderate severity), and 12.5% were Type D
(major severity) and Type X (avoid drug combination). These data are similar to those observed in
a previous study in Palestinian patients [15]. Other studies highlighted differences in the frequency
of pDDIs types (Table 8). Among other causes, the variability in the reported pDDIs could also be a
consequence of using different screening platforms to analyze potential drug interactions. In our case,
we used Lexicomp® , which classifies pDDIs in different levels of severity. However, other software
(Micromedex Drug-Reax, Medscape Drug interaction checker, etc.) for similar drug combinations
perform dissimilar categorizations, or find different pDDIs.
Even though the majority of pDDIs reported in this study were Type C, it is necessary to closely
monitor patients in order to identify adverse events. Moreover, major severity drug interactions and
avoided drug combinations present a high risk to the health of patients and, consequently, physicians
or clinical pharmacists must analyze, detect, and early prevent pDDIs.
In the present study, the majority of the patients were in CKD stage 3, and only 4.5% of the total
were in CKD stage 5 or hemodialysis. This result is comparable with another study from Brazil [10]
in which patients in CKD stage 5 represented 6.6% of the total sample. However, in the remaining
previous studies, most of the patients were in stage 5 or hemodialysis (Table 8). This could affect the
number and type of prescribed treatment and, therefore, the pDDI. The present study did not only
focus on patients on hemodialysis, but on all patients with CKD.

4.2. Factors Associated with Potential Drug–Drug Interactions

Regarding comorbid conditions, these appeared in 91 patients (81.2%), and the most frequent
were hypertension (44.6%), dyslipidemia/hypercholesterolemia (28.6%), and diabetes mellitus (22.3%).
The prevalence of hypertension and diabetes in previous studies [10,13–15] were higher than
those found in the present study. In addition, each country could have implemented different clinical
guidelines for similar disease conditions, which results in the prescription of different drugs and
thereby other pDDIs. The selected hospital is a reference hospital with a nephrology unit similar and
representative of most hospitals in the country. The percentage of patients with chronic kidney disease
is also similar to those other nephrology units in Spain.
In the present study, the mean age of patients was higher (77.1 ± 10.4 years) than in the rest of the
studies, which reported mean age data from 38.3 ± 16.8 to 59.1 ± 14.7 years (Table 8). Polypharmacy
prevalence increases with advancing age [23], and hence also pDDIs. Furthermore, people aged 80 and
over are still much more likely to have DDIs [24]. Therefore, the main reasons for the differences found
in the present study, comparing to most of the previous studies, could be the different CKD stages,
age, or the country of patients. These factors could affect the number and type of prescribed drug
treatment and, therefore, the number and severity of pDDIs. On the other hand, the use of different
software to evaluate the pDDIs in the reported studies (Table 8) could lead to differences of the severity
of pDDIs. Furthermore, many of the analyzed drugs that appear in Lexicomp® do not appear in some
other databases.
This could be one of the reasons for variability in the pattern of frequency and severity of pDDIs
observed in the present study compared to previous studies.

5. Conclusions
The frequency and severity of pDDIs could be affected by the type and number of drugs per
patient, which, at the same time, could be influenced by comorbidities and age. On the other hand,
the advancement of CKD increases the risk of a major cardiac event and the possibility of hospitalization,
which increases the number of medications [25,26].
It should be noted that in CKD patients, the association of medications is sometimes inevitable,
and according to the present results, the use of programs for determination of pDDIs (such as
Lexicomp® ) are recommended in clinical practice for CKD patients in order to avoid serious adverse
effects, paying attention to the contraindicated drug combinations. Therefore, as a way to classify and

115
Pharmaceutics 2020, 12, 713

identify pDDIs according to interaction risk, severity, and reliability, it would be convenient to consider
and evaluate pDDIs in clinical practice in order to avoid or prevent some avoidable adverse effects.

Author Contributions: Conceptualization, G.S.-D., P.D., M.Á.S.-S.; methodology, G.S.-D. and P.D.; software,
G.S.-D. and P.D.; investigation, G.S.-D., A.M.P.-P., M.Á.S.-S., V.G.-B., R.M. and P.D.; writing—original draft
preparation, G.S.-D. and P.D.; writing—review and editing, G.S.-D., A.M.P.-P., M.Á.S.-S., V.G.-B., R.M. and P.D.;
supervision, P.D.; project administration, P.D. All authors have read and agreed to the published version of
the manuscript.
Funding: This work is supported by Junta de Extremadura and European Regional Development Fund (FEDER)
(grant IB16138; V Plan Regional de I+D+i).
Acknowledgments: The authors thank all patients who kindly participated in the study, as well as the clinical
assistance of Laura Piquero Calleja and Anika Tyszkiewiez. This work was supported.
Conflicts of Interest: The authors declare no conflicts of interest.

References
1. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO clinical practice guideline
for the evaluation and management of chronic kidney disease. Kidney Int. Suppl. 2013, 3, 1–150.
2. GBD Chronic Kidney Disease Collaboration. Global, regional, and national burden of chronic kidney disease,
1990-2017: A systematic analysis for the Global Burden of Disease Study 2017. Lancet 2020, 395, 709–733.
[CrossRef]
3. Gorostidi, M.; Sánchez-Martínez, M.; Ruilope, L.M.; Graciani, A.; de la Cruz, J.J.; Santamaría, R.; del Pino, M.;
Guallar-Castillón, P.; de Álvaro, F.; Rodríguez-Artalejo, F.; et al. Prevalencia de enfermedad renal crónica
en España: Impacto de la acumulación de factores de riesgo cardiovascular. Nefrologia 2018, 38, 606–615.
[CrossRef]
4. Webster, A.; Nagler, E.; Morton, R.; Masson, P. Chronic Kidney Disease. Lancet 2017, 389, 1238–1252.
[CrossRef]
5. Shaw, J.E.; Sicree, R.A.; Zimmet, P.Z. Global estimates of the prevalence of diabetes for 2010 and 2030.
Diabetes Res. Clin. Pract. 2010, 87, 4–14. [CrossRef]
6. Kearney, P.; Whelton, M.; Reynolds, K.; Muntner, P.; Whelton, P.; He, J. Global burden of hypertension:
Analysis of worldwide data. Lancet 2005, 365, 217–223. [CrossRef]
7. Major, R.; Cheng, M.; Grant, R.; Shantikumar, S.; Xu, G.; Oozeerally, I.; Brunskill, N.; Gray, L. Cardiovascular
disease risk factors in chronic kidney disease: A systematic review and meta-analysis. PLoS ONE 2018,
13, e0192895. [CrossRef] [PubMed]
8. Masnoon, N.; Shakib, S.; Kalisch-Ellett, L.; Caughey, G. What is polypharmacy? A systematic review of
definitions. BMC Geriatr. 2017, 17, 230. [CrossRef] [PubMed]
9. Shah, B.; Hajjar, E. Polypharmacy, Adverse Drug Reactions, and Geriatric Syndromes. Clin. Geriatr. Med.
2012, 28, 173–186. [CrossRef] [PubMed]
10. Marquito, A.; Fernandes, N.; Colugnati, F.; Paula, R. Identifying potential drug interactions in chronic kidney
disease patients. J. Bras. Nefrol. 2014, 36, 26–34. [CrossRef]
11. Sgnaolin, V.; Sgnaolin, V.; Engroff, P.; De Carli, G.; Prado Lima Figueiredo, A. Avaliação dos medicamentos
utilizados e possíveis interações medicamentosas em doentes renais crônicos. Sci. Med. (Porto Alegre) 2014,
24, 329–335. [CrossRef]
12. Rama, M.; Viswanathan, G.; Acharya, L.; Attur, R.; Reddy, P.; Raghavan, S. Assessment of drug-drug
interactions among renal failure patients of nephrology ward in a south Indian tertiary care hospital. Indian J.
Pharm. Sci. 2012, 74, 63–68. [PubMed]
13. Hegde, S.; Udaykumar, P.; Manjuprasad, M.S. Potential drug interactions in chronic kidney disease patients-A
cross sectional study. Int. J. Recent Trends Sci. Technol. 2015, 16, 56–60.
14. Saleem, A.; Masood, I.; Khan, T. Clinical relevancy and determinants of potential drug-drug interactions in
chronic kidney disease patients: Results from a retrospective analysis. Integr. Pharm. Res. Pract. 2017, 6,
71–77. [CrossRef] [PubMed]

116
Pharmaceutics 2020, 12, 713

15. Al-Ramahi, R.; Raddad, A.; Rashed, A.; Bsharat, A.; Abu-Ghazaleh, D.; Yasin, E.; Shehab, O. Evaluation
of potential drug-drug interactions among Palestinian hemodialysis patients. BMC Nephrol. 2016, 17, 96.
[CrossRef] [PubMed]
16. Adibe, M.O.; Ewelum, P.C.; Amorha, K.C. Evaluation of drug-drug interactions among patients with chronic
kidney disease in a South-Eastern Nigeria tertiary hospital: A retrospective study. Pan Afr. Med. J. 2017,
28, 199. [CrossRef]
17. Fasipe, O.J.; Akhideno, P.E.; Nwaiwu, O.; Adelosoye, A.A. Assessment of prescribed medications and pattern
of distribution for potential drug–drug interactions among chronic kidney disease patients attending the
Nephrology Clinic of Lagos University Teaching Hospital in Sub-Saharan West Africa. Clin. Pharmacol. 2017,
9, 125–132. [CrossRef] [PubMed]
18. Olumuyiwa, J.F.; Akinwumi, A.A.; Ademola, O.A.; Oluwole, B.A.; Ibiene, E.O. Prevalence and pattern
of potential drug-drug interactions among chronic kidney disease patients in south-western Nigeria.
Niger. Postgrad. Med. J. 2017, 24, 88–92.
19. Okoro, R.; Farate, V. Evaluation of potential drug–drug interactions among patients with chronic kidney
disease in northeastern Nigeria. Afr. J. Nephrol. 2019, 22, 77–81.
20. Roblek, T.; Vaupotic, T.; Mrhar, A.; Lainscak, M. Drug-drug interaction software in clinical practice:
A systematic review. Eur. J. Clin. Pharmacol. 2015, 1, 131–142. [CrossRef]
21. Kheshti, R.; Aalipour, M.; Namazi, S. A comparison of five common drug-drug interaction software programs
regarding accuracy and comprehensiveness. J. Res. Pharm. Pract. 2016, 5, 257–263. [PubMed]
22. Patel, R.I.; Beckett, R.D. Evaluation of resources for analyzing drug interactions. J. Med. Libr. Assoc. 2016,
104, 290–295. [CrossRef] [PubMed]
23. Fulton, M.; Allen, E. Polypharmacy in the elderly: A literature review. J. Am. Acad. Nurse Pract. 2005, 17,
123–132. [CrossRef] [PubMed]
24. Guthrie, B.; Makubate, B.; Hernandez-Santiago, V.; Dreischulte, T. The rising tide of polypharmacy and
drug-drug interactions: Population database analysis 1995–2010. BMC Med. 2015, 13, 74. [CrossRef]
[PubMed]
25. Go, A.S.; Chertow, G.M.; Fan, D.; McCulloch, C.E.; Hsu, C.Y. Chronic kidney disease and the risks of death,
cardiovascular events, and hospitalization. N. Engl. J. Med. 2004, 351, 1296–1305. [CrossRef] [PubMed]
26. Parikh, N.I.; Hwang, S.J.; Larson, M.G.; Levy, D.; Fox, C.S. Chronic kidney disease as a predictor of
cardiovascular disease (from the Framingham Heart Study). Am. J. Cardiol. 2008, 102, 47–53. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

117
 pharmaceutics

Article
Subset Analysis for Screening Drug–Drug Interaction
Signal Using Pharmacovigilance Database
Yoshihiro Noguchi 1, *, Tomoya Tachi 1 and Hitomi Teramachi 1,2, *
1 Laboratory of Clinical Pharmacy, Gifu Pharmaceutical University, 1-25-4, Daigakunishi, Gifu-shi,
Gifu 501-1196, Japan; tachi@gifu-pu.ac.jp
2 Laboratory of Community Healthcare Pharmacy, Gifu Pharmaceutical University, Daigakunishi,
Gifu-shi, Gifu 501-1196, Japan
* Correspondence: noguchiy@gifu-pu.ac.jp (Y.N.); teramachih@gifu-pu.ac.jp (H.T.);
Tel.: +81-58-230-8100 (Y.N. & H.T.)

Received: 13 July 2020; Accepted: 10 August 2020; Published: 12 August 2020

Abstract: Many patients require multi-drug combinations, and adverse event profiles reflect not
only the effects of individual drugs but also drug–drug interactions. Although there are several
algorithms for detecting drug–drug interaction signals, a simple analysis model is required for early
detection of adverse events. Recently, there have been reports of detecting signals of drug–drug
interactions using subset analysis, but appropriate detection criterion may not have been used. In this
study, we presented and verified an appropriate criterion. The data source used was the Japanese
Adverse Drug Event Report (JADER) database; “hypothetical” true data were generated through a
combination of signals detected by three detection algorithms. The accuracy of the signal detection
of the analytic model under investigation was verified using indicators used in machine learning.
The newly proposed subset analysis confirmed that the signal detection was improved, compared
with signal detection in the previous subset analysis, on the basis of the indicators of Accuracy (0.584
to 0.809), Precision (= Positive predictive value; PPV) (0.302 to 0.596), Specificity (0.583 to 0.878), Youden’s
index (0.170 to 0.465), F-measure (0.399 to 0.592), and Negative predictive value (NPV) (0.821 to 0.874).
The previous subset analysis detected many false drug–drug interaction signals. Although the newly
proposed subset analysis provides slightly lower detection accuracy for drug–drug interaction signals
compared to signals compared to the Ω shrinkage measure model, the criteria used in the newly
subset analysis significantly reduced the amount of falsely detected signals found in the previous
subset analysis.

Keywords: subset analysis; signal detection algorithms; drug-drug interaction; spontaneous

reporting systems

1. Introduction
Drug-induced adverse events (AEs) caused by individual drugs and drug combinations not
only hinder treatment but also cause new health hazards. To alleviate this problem, AEs caused by
individual drug candidates are closely monitored and investigated during the drug development and
approval process [1]. Pre-marketing randomized clinical trials are performed under certain conditions
associated with age, gender, and co-morbidities, and some AEs may not be detected. In particular,
in pre-marketing randomized clinical trials, patients on combination therapy are usually excluded
because the focus is to establish the safety and efficacy of single drugs and not to investigate drug–drug
interactions [2]. However, in the real world, many patients suffer from a variety of co-morbidities
and use a number of drugs to treat them. The concomitant use of two or more drugs increases the
risk of AEs due to drug–drug interactions; the proportion of such AEs is estimated to be up to 30% of
unexpected AEs [3].

119
Pharmaceutics 2020, 12, 762; doi:10.3390/pharmaceutics12080762 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 762

Therefore, in order to use drugs appropriately in the real world, it is important to understand,
in advance, the AEs caused by drug–drug interactions. Post-marketing analysis of AE reports could
significantly contribute to the discovery of AEs caused by single drugs or drug–drug interactions that
could not be detected before marketing.
For the safety surveillance of drugs, AE reports collected post-marketing are maintained by
regulatory agencies as a spontaneous reporting system. There are several algorithms for detecting
adverse event signals using the spontaneous reporting system [4]. Of these, the algorithms commonly
used for quantitative signal detection include the proportional reporting ratio (PRR) [5], the reporting
odds ratio (ROR) [6], the Bayesian confidence propagation neural network (BCPNN) [7], and the
empirical Bayesian geometric mean (EBGM) [8].
Additionally, multiple statistical algorithms have been proposed for detecting drug–drug
interaction signals [9,10]. However, calculation of the PRR, similar to the risk ratio, and the ROR,
similar to the odds ratio, is simple, but that of other algorithms (particularly the algorithm for detecting
drug–drug interaction signals) is very complicated.
Therefore, in order to detect the drug–drug interaction signals between drug D1 and drug D2 ,
the subset analysis that detects the signal of drug D1 using the ROR, which is easy to calculate in a
subset of patient groups, is often reported [11–13].
Of previous studies, several [11,12] have used animal experiments and/or pharmacological data
to ensure signal reliability, but the signals obtained with this analysis model are not strictly drug–drug
interaction signals; they only showed the effect of drug combinations for the following two reasons [14]:

1. The subset analysis used in this study detects signals from the target AE when the patient group
using drug D1 takes drug D2 . In all patient groups, when the signal value of the target AE is large
for drug D2 , the signal is detected regardless of whether the patient group is using drug D1 .
2. Target AE signal intensities when a patient group using drug D1 takes drug D2 vs. that when a
patient group using drug D2 takes drug D1 do not necessarily match. In other words, the value
to be adopted as the target AE signal value when drug D1 and drug D2 are used concomitantly
has not been fixed (i.e., no clear detection criteria have been defined for detecting drug–drug
interaction signals).

On the other hand, because the ROR, often used in subset analysis, is easy to calculate, if these
shortcomings are improved and the appropriate detection criterion can be set, it might lead to early
detection of AEs caused by drug–drug interactions.
In this study, we proposed a new detection criterion for the subset analysis (the newly proposed
subset analysis) and verified the detection power using the spontaneous reporting system.

2. Materials and Methods

The design of this study is based on a previous paper that discussed trends in methods to detect
the signals of AEs caused by individual drugs [15] or drug–drug interactions [14].

2.1. Data Sources

The validation dataset was created from the Japanese Adverse Drug Event Report database
(JADER), using data from the first quarter of 2004 to the fourth quarter of 2015. The JADER consists of
four comma-separated values (csv) files as data tables: DEMO.csv (patient information), DRUG.csv
(medicine information), HIST.csv (patient past history), and REAC.csv (AE event information).
This study used 374,327 cases registered in the verification dataset.
However, the Japanese authority, the Pharmaceuticals and Medical Devices Agency (PMDA),
which owns these data, does not permit sharing the data directly. Therefore, we do not own the
JADER. It can be accessed directly here: [http://www.info.pmda.go.jp/fukusayoudb/CsvDownload.jsp]
(in Japanese only).

120
Pharmaceutics 2020, 12, 762

2.2. Definitions of Adverse Drug Events

The drugs targeted for the survey are all registered and classified as “suspect drugs” in the
validation dataset. The AEs in JADER are based on the preferred terms (PTs) in the Medical
Dictionary for Regulatory Activities Japanese version (MedDRA/J). The AE targeted for this study was
Stevens–Johnson syndrome (SJS), which was extracted from the dataset using the PT in MedDRA/J.
However, the choice of target adverse events is the same as in previous similar studies [14,15], and
there was no medical or pharmacological reason for this choice.

2.3. “Hypothetical” True Data of Adverse Events for Comparative Verification

The signals obtained from the spontaneous reporting system including the JADER used in this
study included unknown AEs that were also detected, which needs to be verified in order to confirm
they were true AEs. Moreover, the information provided by the regulatory authorities, of course,
does not include unknown AEs. That is, there are no “real” true data for every AE. Therefore, we set
“hypothetical” true data because we cannot use “real” true data for validation in this study.
To verify the power of the subset analysis, we prepared “hypothetical” true data of AEs. To generate
“hypothetical” true data, we excluded the Ω shrinkage measure model [16] that detected the most
conservative signal and the combination risk ratio model [17], which would not detect a signal with a
small number of reports, from the five detection algorithms. That is, this study used the combination
of signals detected by three algorithms (the additive model [18], the multiplicative model [18], and the
chi-square statistical model [19]) as “hypothetical” true data.

2.4. Statistical Models and Criteria

2.4.1. Subset Analysis

To detect the signal for the interaction between drug D1 and drug D2 , we created subsets of the
patient group using drug D1 (or the patient group using drug D2 ) (Table 1).

Table 1. The 4 × 2 contingency table for signal detection: AE: adverse event; n: the number of
reports (e.g., n+++ : the number of all reports, n111 : the number of drug D1 - and drug D2 -induced target
AE reports).

Target AE Other AEs Total

Concomitant use of drug D1 and drug D2 n111 n110 n11+
only drug D1 n101 n100 n10+
only drug D2 n011 n010 n01+
Neither drug D1 or drug D2 n001 n000 n00+
Total n++1 n++0 n+++

The following equations (Equations (1) and (2)) were used to calculate the ROR and 95% confidence
interval (95% CI) of the target AE caused by drug D1 (or drug D2 ) from the generated subset, respectively.
For the signal of a patient group on drug D1 that takes drug D2 , the number of each report can be
expressed as follows: N11 = n111 , N10 = n110 , N01 = n101 , N00 = n100 . On the other hand, for the signal of
a patient group on drug D2 that takes drug D1 , the number of each report can be expressed as follows:
N11 = n111 , N10 = n110 , N01 = n011 , N00 = n010 .

N11 /N00
ROR = (1)
N01 /N10

ln (ROR) ±1.96 N11 + N10 + N01 + N00
1 1 1 1
ROR (95% CI) = e (2)

121
Pharmaceutics 2020, 12, 762

In previous studies [11–13], if the signal for drug D2 was detected in the subset of a patient group
using drug D1 or if the signal for drug D1 was detected in the subset of a patient group using drug D2 ,
this signal was considered the drug–drug interaction signal. The criterion that a signal only needs to
be detected from a subset of either patient group is ambiguous, highlighting the two shortcomings
mentioned earlier. Therefore, for the newly proposed subset analysis, a case was redefined as the
drug–drug interaction signal if a signal was detected in both subsets of a patient group using drug D1
and a patient group using drug D2 .

2.4.2. Ω Shrinkage Measure Model

The Ω shrinkage measure model [16] is based on a measure calculated as the ratio of the observed
reporting ratio of the AE associated with the combination of two drugs and its expected value; this
model is used by the Uppsala Monitoring Center (UMC) and the World Health Organization (WHO)
Collaborating Centre for International Drug Monitoring for signal analysis of drug–drug interactions
(Table 1, Equations (3)–(7)).
n + 0.5
Ω = log2 111 (3)
E111 + 0.5
where n111 is the reported number of AEs caused by the combination of two drugs, and E111 is the
expected value of AEs caused by the combination of two drugs.
φ(0.975) is 97.5% of the standard normal distribution and Ω025 > 0 is used as a threshold to screen
for signals under the combination of two drugs (Equation (4)).

φ(0.975)
Ω025 = Ω − √ (4)
log(2) n111

To calculate E111 , we used the following Equations (5)–(7).

n001 n n n
f00 = , f10 = 101 , f01 = 011 , f11 = 111 (5)
n00+ n10+ n01+ n11+

1
g11 = 1 −     (6)
f f10 f00 f01 f00
max 1−00f ,
00 1− f10 + max 1− f00 , 1− f01 − 1− f00 +1

When f 10 < f 00 (which denotes no risk of AE caused by drug D1 ), the most sensible estimator g11
= max (f 00 , f 01 ) is yielded and vice versa when f 01 < f 00 .
Norén et al. re-expressed the observed and expected RRR f 11 and g11 in terms of the observed
number of reports n111 and expected numbers of reports E111 = g11 × n11+ , respectively:

f11 n111 /n11+ n

= = 111 (7)
g11 E111 /n11+ E111

2.5. Evaluation of Detection Models

2.5.1. Using Evaluations of Classification in Machine Learning

The evaluation indicators that we set were Accuracy (Table 2, Equation (8)), Precision (Positive
predictive value; PPV) (Table 2, Equation (9)), Recall (Sensitivity) (Table 2, Equation (10)), Specificity
(Table 2, Equation (11)), Youden’s index (Table 2, Equation (12)), F-measure (Table 2, Equation (13)), and
Negative predictive value (NPV) (Table 2, Equation (14)).

TP + TN
Accuracy = (8)
TP + FP + TN + FN

TP
Precision (Positive predictive value; PPV ) = (9)
TP + FP

122
Pharmaceutics 2020, 12, 762

TP
Recall (Sensitivity) = (10)
TP + FN
TN
Speci f icity = (11)
FP + TN
Youden s index = Sensitivity + Speci f icity − 1 (12)
2 × Recall × Precision
F − measure = (13)
Recall + Precision
TN
Negative predictive value (NPV ) = (14)
TN + FN

Table 2. Agreement between the criterion A and the “hypothetical” true data. AE: adverse event, TP:
True Positive, FP: False Positive, FN: False Negative, TN: True Negative.

“Hypothetical” True Data

AE non-AEs
signal TP FP
analysis model
Non-signal FN TN

2.5.2. Cohen’s Kappa Coefficient

The commonality of the signals detected by each statistical model was evaluated using Cohen’s
kappa coefficient (κ), proportionate agreement for positive rating (Ppositive ), and proportionate agreement
for negative rating (Pnegative ), as reported in a previous study [14,15]. In this study, we investigated the
similarities with Ω shrinkage measure model for the previous/newly proposed subset analysis.

2.6. Analysis Software

The analysis software in this study used Visual Mining Studio (NTT DATA Mathematical
Systems Inc., Shinjuku-ku, Tokyo, Japan) version 8.4 and Microsoft Excel 2019 (Microsoft Corp.,
Redmond, WA, USA).

3. Results

3.1. Evaluations of Classification in Machine Learning

Among all 374,327 cases analyzed, there were 3924 drug D1 –drug D2 –SJS combinations. Of these,
923 combinations were detected by all three algorithms—the additive model [18], the multiplicative
model [18], and the chi-square statistics model [19]. In this study, these combinations were treated as
“hypothetical” true data.
The evaluation of the analysis model is shown in Tables 3 and 4.

Table 3. The number of True positive, False positive, True negative, and False negative.

Analysis Model TP FP TN FN
Previous subset analysis 542 1251 1750 381
Newly proposed subset analysis 542 367 2634 381
Ω shrinkage measure model 538 174 2827 385
TP: True positive, FP: False positive, TN: True negative, FN: False negative.

123
Pharmaceutics 2020, 12, 762

Table 4. Evaluation of detected drug–drug interaction signals.

Precision Recall Youden’s

Analysis Model Accuracy Specificity F-Measure NPV
(PPV) (Sensitivity) Index
Previous subset analysis 0.584 0.302 0.587 0.583 0.170 0.399 0.821
Newly proposed subset analysis 0.809 0.596 0.587 0.878 0.465 0.592 0.874
Ω shrinkage measure model 0.858 0.756 0.583 0.942 0.525 0.658 0.880
PPV: Positive predictive value, NPV: Negative predictive value.

Table 3 shows the number of True positive (TP), False Positive (FP), True Negative (TN), and False
Negative (FN).
A total of 1793 combinations were detected by the previous subset analysis (True positive: 542,
False positive: 1251). On the other hand, the newly proposed subset analysis detected 909 combinations
of signals (True positive: 542, False positive: 367) (Table 3).
The detection accuracy shown in Table 4 was calculated from the values shown in Table 3.
In addition, the newly proposed subset analysis confirmed that the signal detection was improved
with respect to the indicators of Accuracy (0.584 to 0.809), Precision (PPV) (0.302 to 0.596), Specificity
(0.583 to 0.878), Youden’s index (0.170 to 0.465), F-measure (0.399 to 0.592), and NPV (0.821 to 0.874) as
compared with the signal detection in the previous subset analysis (Table 3).
The values of each indicator of the Ω shrinkage measure model were Accuracy (0.858), Precision
(PPV) (0.756), Recall (Sensitivity) (0.583), Specificity (0.942), Youden’s index (0.525), F-measure (0.658), and
NPV (0.880) (Table 4).

3.2. Cohen’s Kappa Coefficient

The similarity between the detection results of the Ω shrinkage measure model and that of the
newly proposed subset analysis was κ (95% CI): 0.375 (0.355–0.395), Ppositive : 0.502, and Pnegative : 0.870.
The similarity was κ (95% CI): 0.355 (0.327–0.384), Ppositive : 0.678, and Pnegative : 0.674 when targeting
three or more reports (Table 5).

Table 5. The similarity between the Ω shrinkage measure model and subset analysis.

All Case n111 ≥ 3

Analysis Model κ κ
Ppositive Pnegative Ppositive Pnegative
(95% CI) (95% CI)
0.088 −0.120
Previous subset analysis 0.325 0.684 0.556 0.296
(0.071–0.105) (−0.151–0.088)
0.375 0.355
Newly proposed subset analysis 0.502 0.870 0.678 0.674
(0.355–0.395) (0.327–0.384)
n111 : targeting three or more reports, κ: Cohen’s kappa coefficient, Ppositive : proportionate agreement for positive rating,
Pnegative : proportionate agreement for negative rating.

4. Discussions
In this study, we evaluated the accuracy of drug–drug interaction signals for the newly proposed
subset analysis that modified two shortcomings of the previous subset analysis on the basis of data
from the spontaneous reporting system.
There were 3924 pairs of drug D1 –drug D2 –SJS in the spontaneous reporting system, JADER. There
are several known combinations of drugs that onset SJS by drug–drug interactions [20]. On the other
hand, there are some combinations that have not yet been reported. Recently, we used the Ω shrinkage
measure model to report potential drug combinations for the onset of SJS in concomitant use with
antiepileptic drugs [21]. Not all AEs have been identified and there are still many unknown AEs.
Unfortunately, unknown AE data do not exist anywhere in the world; there were no “real” true data for
AEs. Therefore, to verify the accuracy of the subset analysis, we needed to prepare “hypothetical” true
data of AEs. A previous comparative study [14] of five algorithms for detecting drug–drug interaction
signals revealed that the Ω shrinkage measure model [16] detected the most conservative signal, while

124
Pharmaceutics 2020, 12, 762

the combination risk ratio model [17] did not detect any interaction signal in less than three reports
due to the detection criterion. Therefore, of the five algorithms, we used the combination of signals
detected by the three algorithms (the additive model, the multiplicative model, and the chi-square
statistical model) as “hypothetical” true data.
Among the previous subset analysis, the newly proposed subset analysis, and the Ω shrinkage
measure model, most signals were detected by the previous subset analysis with 1793 pairs (45.7%
of the total combinations, Accuracy: 0.584, Precision (PPV): 0.302, Recall (Sensitivity): 0.587, Specificity:
0.583, Youden’s index: 0.170, F-measure: 0.399, and NPV: 0.821), followed by the newly proposed
subset analysis with 909 pairs (23.2% of the total combinations, Accuracy: 0.809, Precision (PPV): 0.596,
Recall (Sensitivity): 0.587, Specificity: 0.878, Youden’s index: 0.465, F-measure: 0.592, and NPV: 0.874).
In contrast, the Ω shrinkage measure model detected the fewest signals with 712 pairs (18.1% of the
total combinations, Accuracy: 0.858, Precision (PPV): 0.756, Recall (Sensitivity): 0.583, Specificity: 0.942,
Youden’s index: 0.525, F-measure: 0.658, and NPV: 0.880) (Table 2, Table 4).
This result indicates that the accuracy of signal detection has been greatly improved in the
newly proposed subset analysis with a simple modification of the previous subset analysis. However,
the newly proposed subset analysis exhibited slightly lower power and accuracy for detecting the
drug–drug interaction signals compared to the Ω shrinkage measure model.
Verification by the number of reports showed that when the number of reports (N11 ; n111 ) < 2,
the accuracy (Youden’s index, F-measure) of signal detection was higher in the newly proposed subset
analysis than in the Ω shrinkage measure model (Youden’s index: the newly proposed subset analysis
(0.337) vs. the Ω shrinkage measure model (0.174), F-measure: the newly proposed subset analysis
(0.448) vs. the Ω shrinkage measure model (0.298)).
However, as the number of reports increased, the Ω shrinkage measure model became more
accurate (Youden’s index: the newly proposed subset analysis (0.465) vs. the Ω shrinkage measure
model (0.525), F-measure: the newly proposed subset analysis (0.592) vs. the Ω shrinkage measure
model (0.658)) (Table 4).
Additionally, the True positive values for the previous subset analysis and the newly proposed
subset analysis were the same (Table 3). Since all signals obtained by the newly proposed subset
analysis were included in the previous subset analysis, this result indicates that the detection criterion
of the previous subset analysis was loose and that the data contained false positives.
The similarity between the newly proposed subset analysis and the Ω shrinkage measure model
was κ (95% CI): 0.375 (0.355–0.395), Ppositive : 0.502, and Pnegative : 0.870. On the other hand, the similarity
between the previously subset analysis and the Ω shrinkage measure model was κ (95% CI): 0.088
(0.071–0.105), Ppositive : 0.325, and Pnegative : 0.684. Thus, the newly proposed subset analysis w
more similar to the Ω shrinkage measure model than the previously subset analysis. However,
the similarity of the newly proposed subset analysis and the Ω shrinkage measure model is not very
high. Additionally, when the number of reports (N11 ; n111 ) was ≥3, no significant change was observed
in the similarity between the Ω shrinkage measure model and the newly proposed subset analysis.
Despite not being similar to the Ω shrinkage model, the newly subset analysis showed a high degree
of accuracy. This result suggests that the newly subset analysis may be detecting signals that the Ω
shrinkage model has failed to detect.
This study has the following three limitations. First, unfortunately, unknown AE data do not exist
anywhere in the world [14]. Therefore, there were no “real” true data for AEs. Thus, for the purpose
of verification, it was necessary to set “hypothetical” true data for AEs instead of “real” true data.
Therefore, of the five algorithms for detecting drug–drug interaction signals, we used the combination
of signals detected by the three algorithms (the additive model, the multiplicative model, and the
chi-square statistical model) as “hypothetical” true data in this study. In other words, the hypothetical
true data consisted of statistically based drug D1 –drug D2 –AE combinations, not pharmacologically
based combinations.

125
Pharmaceutics 2020, 12, 762

Second, usually it is important to compare detection trends using all AEs recorded in the validation
dataset created on the basis of a spontaneous reporting system; however, it takes an extremely long time
to calculate signal values for all combinations of drug–drug interactions. Such a study design is not
realistic. Therefore, this study targeted SJS, the same AE used in previous comparative studies [14,15];
if different reference sets were used, the possibility of obtaining different performance characteristics
might not be ruled out. There are fewer enrolled cases than in the global dataset because JADER is
limited to cases in Japan. However, the signal detection is based on a comparison between the ratio of
reported cases (N) to expected values (E). Therefore, differences in the number of cases enrolled in the
spontaneous reporting system had only a very small statistical impact in this study. Recently, validation
of the number of cases enrolled in the spontaneous reporting system has also been reported by Caster
et al. [22]. Moreover, differences in the way regulatory authorities think may result in a different
tendency to register AEs to the spontaneous reporting system. For example, the Food and Drug
Administration Adverse Events Reporting System (FAERS) in the United States has also registered
reports from non-medical professionals, but JADER has not registered reports from patients until
recently. It is unknown how the differences in registration tendencies affect the results of this study.
Finally, neither the general algorithms for detecting drug–drug interaction signals nor the proposed
subset analysis in this study were antagonistic; only signals of synergistic interactions were detected [10].

5. Conclusions
In recent years, the need for safety signal screening has been demanded, not only for single
drugs but also for drug–drug interactions. Although several methods for detecting signals of drug
interaction have been reported, it is difficult to say that these methods are used because many of them
are complicated in calculation. Therefore, there were several cases [11–13] where subset analysis using
the algorithm for detecting signals of single drugs (e.g., ROR [6]) was used for signal detection of
drug–drug interactions before its validity was verified.
This study showed that there were many false positives in the existing subset analysis, albeit
under limited conditions. Additionally, very simple modifications of the detection criteria were made
to solve two problems associated with the previous subset analysis for exploring recently reported
drug–drug interaction signals. This modification helped to reduce falsely detected signals found in the
previous subset analysis.
Moreover, the newly proposed subset analysis is more similar to the Ω shrinkage measure model
than the previous subset analysis, but the similarity with the Ω shrinkage measure model is not as
high. However, the newly proposed subset analysis showed that although the detection accuracy
of the drug–drug interaction signal was slightly lower than that of the Ω shrinkage measure model,
the detection accuracy was sufficient. This result may also indicate the possibility of detecting signals
that cannot be detected by the Ω shrinkage measure model.

Author Contributions: Conceptualization, Y.N.; funding acquisition, Y.N.; investigation, Y.N.; methodology, Y.N.
and T.T.; project administration, Y.N. and H.T.; supervision, Y.N.; validation, Y.N. and T.T.; visualization, Y.N.;
writing—original draft, Y.N.; writing—review and editing, Y.N., T.T., and H.T. All authors have read and agreed
to the published version of the manuscript.
Funding: This research was funded by JSPS KAKENHI grant number 19K20731.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Berlin, J.A.; Glasser, S.C.; Ellenberg, S.S. Adverse event detection in drug development: Recommendations
and obligations beyond phase 3. Am. J. Public Health 2008, 98, 1366–1371. [CrossRef] [PubMed]
2. Noguchi, Y.; Ueno, A.; Otsubo, M.; Katsuno, H.; Sugita, I.; Kanematsu, Y.; Yoshida, A.; Esaki, H.; Tachi, T.;
Teramachi, H. A New Search Method Using Association Rule Mining for Drug-Drug Interaction Based on
Spontaneous Report System. Front. Pharmacol. 2018, 9, 197. [CrossRef] [PubMed]

126
Pharmaceutics 2020, 12, 762

3. Iyer, S.V.; Harpaz, R.; LePendu, P.; Bauer-Mehren, A.; Shah, N.H. Mining clinical text for signals of adverse
drug-drug interactions. J. Am. Med. Inf. Assoc. 2013, 21, 353–362. [CrossRef] [PubMed]
4. Suling, M.; Pigeot, I. Signal Detection and Monitoring Based on Longitudinal Healthcare Data. Pharmaceutics
2012, 4, 607–640. [CrossRef] [PubMed]
5. Evans, S.J.; Waller, P.C.; Davis, S. Use of proportional reporting ratios (PRRs) for signal generation from
spontaneous adverse drug reaction reports. Pharmacoepidemiol. Drug Saf. 2001, 10, 483–486. [CrossRef]
[PubMed]
6. Rothman, K.J.; Lanes, S.; Sacks, S.T. The reporting odds ratio and its advantages over the proportional
reporting ratio. Pharmacoepidemiol. Drug Saf. 2004, 13, 519–523. [CrossRef]
7. Bate, A.; Lindquist, M.; Edwards, I.R.; Olsson, S.; Orre, R.; Lansner, A.; De Freitas, R.M. A Bayesian neural
network method for adverse drug reaction signal generation. Eur. J. Clin. Pharmacol. 1998, 54, 315–321.
[CrossRef]
8. DuMouchel, W. Bayesian Data Mining in Large Frequency Tables, with an Application to the FDA Spontaneous
Reporting System. Am. Stat. 1999, 53, 177–190. [CrossRef]
9. Vilar, S.; Friedman, C.; Hripcsak, G. Detection of drug-drug interactions through data mining studies using
clinical sources, scientific literature and social media. Brief. Bioinform. 2018, 19, 863–877. [CrossRef]
10. Noguchi, Y.; Tachi, T.; Teramachi, H. Review of Statistical Methodologies for Detecting Drug-Drug Interactions
Using Spontaneous Reporting Systems. Front. Pharmacol. 2019, 10, 1319. [CrossRef]
11. Nagashima, T.; Shirakawa, H.; Nakagawa, T.; Kaneko, S. Prevention of antipsychotic-induced hyperglycaemia
by vitamin D: A data mining prediction followed by experimental exploration of the molecular mechanism.
Sci. Rep. 2016, 6, 26375. [CrossRef] [PubMed]
12. Uno, T.; Wada, K.; Hosomi, K.; Matsuda, S.; Ikura, M.M.; Takenaka, H.; Terakawa, N.; Oita, A.; Yokoyama, S.;
Kawase, A.; et al. Drug interactions between tacrolimus and clotrimazole troche: A data mining approach
followed by a pharmacokinetic study. Eur. J. Clin. Pharmacol. 2020, 76, 117–125. [CrossRef] [PubMed]
13. Sanagawa, A.; Hotta, Y.; Kondo, M.; Nishikawa, R.; Tohkin, M.; Kimura, K. Tumor lysis syndrome associated
with bortezomib: A post-hoc analysis after signal detection using the US Food and Drug Administration
Adverse Event Reporting System. Anti-Cancer Drugs 2020, 31, 183–189. [CrossRef] [PubMed]
14. Noguchi, Y.; Tachi, T.; Teramachi, H. Comparison of signal detection algorithms based on frequency statistical
model for drug-drug interaction using spontaneous reporting systems. Pharm. Res. 2020, 37, 86. [CrossRef]
15. Kubota, K.; Koide, D.; Hirai, T. Comparison of data mining methodologies using Japanese spontaneous
reports. Pharmacoepidemiol. Drug Saf. 2004, 13, 387–394. [CrossRef]
16. Norén, G.N.; Sundberg, R.; Bate, A.; Edwards, I.R. A statistical methodology for drug-drug interaction
surveillance. Stat. Med. 2008, 27, 3057–3070. [CrossRef]
17. Susuta, Y.; Takahashi, Y. Safety risk evaluation methodology in detecting the medicine concomitant use risk
which might cause critical drug rash. Jpn. J. Pharmacoepidemiol. 2014, 19, 39–49. [CrossRef]
18. Thakrar, B.T.; Grundschober, S.B.; Doessegger, L. Detecting signals of drug-drug interactions in a spontaneous
reports database. Br. J. Clin. Pharmacol. 2007, 64, 489–495. [CrossRef]
19. Gosho, M.; Maruo, K.; Tada, K.; Hirakawa, A. Utilization of chi-square statistics for screening adverse
drug-drug interactions in spontaneous reporting systems. Eur. J. Clin. Pharmacol. 2017, 73779–73786.
[CrossRef]
20. Cheng, F.J.; Syu, F.K.; Lee, K.H.; Chen, F.C.; Wu, C.H.; Chen, C.C. Correlation between drug-drug
interaction-induced Stevens-Johnson syndrome and related deaths in Taiwan. J. Food Drug Anal. 2016, 24,
427–432. [CrossRef]
21. Noguchi, Y.; Takaoka, M.; Hayashi, T.; Tachi, T.; Teramachi, H. Antiepileptic combination therapy with
Stevens-Johnson syndrome and toxic epidermal necrolysis: Analysis of a Japanese pharmacovigilance
database. Epilepsia 2020. [CrossRef] [PubMed]
22. Caster, O.; Aoki, Y.; Gattepaille, L.M.; Grundmark, B. Disproportionality Analysis for Pharmacovigilance
Signal Detection in Small Databases or Subsets: Recommendations for Limiting False-Positive Associations.
Drug Saf. 2020, 43, 479–487. [CrossRef] [PubMed]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

127
 pharmaceutics

Review
Mechanisms of CYP450 Inhibition: Understanding
Drug-Drug Interactions Due to Mechanism-Based
Inhibition in Clinical Practice
Malavika Deodhar 1 , Sweilem B Al Rihani 1 , Meghan J. Arwood 1 , Lucy Darakjian 1 ,
Pamela Dow 1 , Jacques Turgeon 1,2 and Veronique Michaud 1,2, *
1 Tabula Rasa HealthCare Precision Pharmacotherapy Research and Development Institute,
Orlando, FL 32827, USA; mdeodhar@trhc.com (M.D.); srihani@trhc.com (S.B.A.R.);
marwood@trhc.com (M.J.A.); ldarakjian@trhc.com (L.D.); pdow@trhc.com (P.D.); jturgeon@trhc.com (J.T.)
2 Faculty of Pharmacy, Université de Montréal, Montreal, QC H3C 3J7, Canada
* Correspondence: vmichaud@trhc.com; Tel.: +1-856-938-8697

Received: 5 August 2020; Accepted: 31 August 2020; Published: 4 September 2020

Abstract: In an ageing society, polypharmacy has become a major public health and economic issue.
Overuse of medications, especially in patients with chronic diseases, carries major health risks.
One common consequence of polypharmacy is the increased emergence of adverse drug events,
mainly from drug–drug interactions. The majority of currently available drugs are metabolized by
CYP450 enzymes. Interactions due to shared CYP450-mediated metabolic pathways for two or more
drugs are frequent, especially through reversible or irreversible CYP450 inhibition. The magnitude
of these interactions depends on several factors, including varying affinity and concentration of
substrates, time delay between the administration of the drugs, and mechanisms of CYP450 inhibition.
Various types of CYP450 inhibition (competitive, non-competitive, mechanism-based) have been
observed clinically, and interactions of these types require a distinct clinical management strategy.
This review focuses on mechanism-based inhibition, which occurs when a substrate forms a reactive
intermediate, creating a stable enzyme–intermediate complex that irreversibly reduces enzyme
activity. This type of inhibition can cause interactions with drugs such as omeprazole, paroxetine,
macrolide antibiotics, or mirabegron. A good understanding of mechanism-based inhibition and
proper clinical management is needed by clinicians when such drugs are prescribed. It is important
to recognize mechanism-based inhibition since it cannot be prevented by separating the time of
administration of the interacting drugs. Here, we provide a comprehensive overview of the different
types of mechanism-based inhibition, along with illustrative examples of how mechanism-based
inhibition might affect prescribing and clinical behaviors.

Keywords: drug–drug interactions; mechanism-based inhibition; competitive inhibition;

non-competitive inhibition; substrate; inhibitor; cytochromes P450

1. Introduction
In clinical practice, the need for the use of multiple drugs is common, as patients often present
with numerous chronic diseases. To improve commodity and drug adherence, several medications
are often administered concomitantly. Although this may represent a preferred clinical strategy, the
administration of two or more drugs at overlapping times increases the likelihood of drug–drug
interactions [1,2]. As the risk of drug–drug interactions increases, the risk of debilitating, even fatal,
adverse drug events also increases [3]. From a pharmacokinetics standpoint, drug–drug interactions
occur when one drug—the perpetrator drug—alters the disposition of another co-administered
drug—the victim drug.

129
Pharmaceutics 2020, 12, 846; doi:10.3390/pharmaceutics12090846 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 846

Inhibition of cytochrome P450 (CYP450) enzymes is the most common mechanism leading to
drug–drug interactions [4]. CYP450 inhibition can be categorized as reversible (including competitive
and non-competitive inhibition) or irreversible (or quasi-irreversible), such as mechanism-based
inhibition. Each type of interaction involves a distinct clinical management strategy. This is why a
comprehensive understanding of mechanisms of CYP450-mediated metabolism inhibition is needed to
prevent or mitigate these harmful drug interactions. This review will focus on the CYP450 enzymatic
system with a special look at one specific type of CYP450 inhibition, namely mechanism-based
inhibition; clinical cases involving mechanism-based inhibitors will be discussed in this context.

2. Basic Concepts of Enzyme, Substrate, and Inhibitor

For clarity, as we address the various types of drug metabolism inhibition, the concepts of active
(orthosteric) and allosteric sites, substrates, and inhibitors need to first be reviewed.

2.1. Active Site or Orthosteric Site of the Enzyme

The active or orthosteric site is a physical space or pocket within the protein structure of an enzyme
where a molecule can bind and from where a catalytic reaction takes place to convert the molecule into
a metabolite (addition of a hydroxyl moiety, removal of alkyl moieties, etc.). (Figure 1) For CYP450
isoforms, binding to the active site is independent of the NADPH-P450 oxidase reactions; however,
the chemical reaction leading to the formation of the metabolite will employ electrons originated
from NADPH.

Figure 1. Illustration of an enzyme with its active binding site for drug transformation and allosteric
binding site (or regulatory site).

2.2. Allosteric Site

The allosteric site is a physical space or pocket within the protein structure of an enzyme spatially
separated from the active site. (Figure 1) The allosteric site allows molecules to modulate enzyme
activity. These molecules can be allosteric activators or allosteric inhibitors, depending on how they
influence enzyme activity. Drugs can bind to this site and change the three-dimensional structure of the
enzyme (conformational change). Allosteric inhibitors may render the active site no longer accessible
for substrate binding or make the site unable to catalyze reactions. It is widely known that almost all
cases of non-competitive inhibition are caused by allosteric regulation (see discussion in Section 3).

2.3. Substrates
Substrates are drugs that bind to the active site of an enzyme and are transformed into metabolites
while being present in this active site. The biotransformation process of a drug may involve
multiple enzymes leading to various metabolites; each metabolic route relies on specific characteristics.

130
Pharmaceutics 2020, 12, 846

The strength of attraction between an enzyme and a substrate is measured as the “binding affinity”.
A substrate can exhibit varying binding affinity for an active site depending on their chemical structure
and physical properties. Based on their binding affinity for a specific enzyme, substrates can be
classified into weak, intermediate, and strong affinity substrates. Advanced clinical decision support
systems (such as MedWise™) depict the various degrees of affinity by different colors: light yellow
(weak), dark yellow (intermediate), and orange (strong affinity).

2.4. Binding Affinity

Binding affinity to an enzyme is measured by the Km , i.e., the concentration at which 50% of the
maximum metabolic reaction (Vmax ) occurs; the lower the Km , the greater the affinity. The intrinsic
clearance measures the ability of an organ to clear unbound drug when there are no limitations to
blood flow and binding considerations. The intrinsic clearance (CLint ) of a substrate is defined by:

Vmax
CLint =
(Km + [S])

where [S] is the substrate concentration. In most clinical situations, liver enzymes are rarely saturated
so that, generally, the substrate concentration is much smaller than the Km and the equation can be
simplified to:
Vmax
CL int ≈
Km
The binding affinity of a substrate can be modified by the presence of other molecules (in many
drug–drug interaction situations, the Km is increased for the victim drug such that its CLint is decreased).

2.5. Inhibitors
Drugs defined as inhibitors bind either to the active site or to an allosteric site of the enzyme.
However, they can also bind to both; in these cases, the process is called “mixed inhibition” and can
often be more potent than simple competitive or non-competitive inhibition. Inhibitors can be either
substrates or non-substrates of the enzyme. As mentioned previously, non-substrate inhibitors typically
bind to an allosteric site of the enzyme. If the inhibitor is a substrate transformed by the enzyme, the
substrate itself or its metabolites could contribute to the inhibition mechanism. For example, studies
on the inhibitory potency of gemfibrozil indicated that gemfibrozil is a potent inhibitor of CYP2C9
in vitro, but that it is a more potent inhibitor of CYP2C8 than CYP2C9 in vivo [5–7]. This observation
is substantiated by several clinical reports of interactions between gemfibrozil and CYP2C8 substrates
including cerivastatin, repaglinide, and glitazones [8–11]. The mechanism of this clinical interaction is
explained by the formation of the major metabolite of gemfibrozil, gemfibrozil 1-O-β-glucuronide,
which was found to potently inhibit CYP2C8 [10,12].

3. Mechanism of CYP450 Inhibition

Drug interactions due to drug metabolism inhibition are frequent since (1) CYP450-mediated
metabolism is the major route of elimination for a large number of drugs, and (2) multiple
drugs can compete for the same CYP450 active site. Mechanisms of CYP450 inhibition can be
categorized as reversible (including competitive or non-competitive) or irreversible/quasi-irreversible
(mechanism-based inhibition).

3.1. Reversible CYP450 Inhibition

Reversible inhibition is a result of rapid association and dissociation between the substrate
drugs and the enzyme and can be categorized as competitive or non-competitive. A third category,
uncompetitive inhibitor, also considered as a reversible inhibition type, is a very rare phenomenon and

131
Pharmaceutics 2020, 12, 846

will not be considered in this current review; this type of inhibitor binds only the enzyme–substrate
complex, leading to a dead-end complex.

3.1.1. Competitive Inhibition

The ability of a single CYP450 isoform to metabolize multiple substrates is responsible for several
drug interactions associated with reversible competitive inhibition. Competitive inhibition occurs when
two substrates compete for the same active site—such as the prosthetic heme iron or substrate-binding
region—of CYP450s. The competition is a function of the respective affinities of the two substrates
for the binding site and their concentrations in the proximity of the enzyme. First, the most clinically
relevant situation will be discussed.

Two Substrates with Different Affinities Administered Concomitantly

This situation is often encountered in clinical practice. Under a competitive inhibition condition,
a substrate with strong affinity (acting as a perpetrator) can displace a weaker substrate (behaving as a
victim) from the active site (Figure 2), thus increasing the Km of the victim drug (decreased affinity)
and reducing the extent of its breakdown (decrease in its CLint ) over a period of time.

Figure 2. Illustration of reversible competitive inhibition where ligand A (orange) is a substrate with
strong affinity and ligand B (yellow) is a substrate with a weaker affinity for a specific enzyme (purple).
As long as the concentrations of the two substrates are comparable, the stronger affinity substrate with
higher binding affinity will be preferred at the active site of the enzyme resulting in an accumulation of
ligand B.

For an active drug, the decrease in the CLint of one of its metabolic pathways can lead to a decrease
in the total clearance of the drug (capacity for eliminating the drug) and can result in increased plasma
concentrations, potentially precipitating adverse effects. However, for prodrugs, this interaction
can instead result in a decrease in the formation of the active metabolite, reducing drug efficacy.
The magnitude of changes observed in the overall disposition of the victim drug (increase in its plasma
levels) will be a function of the relative contribution of the inhibited metabolic pathway to the clearance
of this drug. For example, if 15% of a drug is excreted unchanged in urine—35% by enzyme 1 and 50%
by enzyme 2—a 50% decrease in the total CL of the victim drug is expected if enzyme 2 is inhibited:

CL = CLrenal + CLmetabolic = 0.15 + 0.85

132
Pharmaceutics 2020, 12, 846

and,
CLmetabolic = CLenzyme 1 + CLenzyme 2

or,
0.85 = 0.35 + 0.5

under conditions of enzyme 2 inhibition (whether it is reversible or irreversible),

CLmetabolic = CLenzyme 1 + CLenzyme 2

or,
0.35 = 0.35 + 0.0

and,
CL = CLrenal + CLmetabolic = 0.15 + 0.35 = 0.5

Since,
CL = Dose/AUC0-∞

Under steady-state conditions, the area under the drug concentration curve (AUC) measured over
a dosing interval (τ) is equal to AUC0–∞ . Since the average concentration over a dosing interval (Cav )
at steady state can be estimated by AUC0-τ /τ, the equation could be rearranged in a simpler manner
to yield:
CL = Dose/(Cav × τ)

A 50% decrease in CL will be associated with a doubling in the average plasma concentrations of
the victim drug.
According to a competitive inhibition mechanism, every substrate of an enzyme is a potential
perpetrator drug (inhibitor) towards another substrate metabolized by the same enzyme. Competitive
inhibition is almost immediate and the degree of inhibition of the enzyme does not change with time if
the concentration of the two substrates remains the same.
In an example illustrating this scenario, theophylline (weak CYP1A2 substrate) is largely
metabolized by CYP1A2 by binding to its active site. (Figure 3) If that active site is occupied by a
stronger substrate like duloxetine (moderate CYP1A2 affinity substrate), breakdown of theophylline
will be reduced (↓CLint ), leading to increased plasma levels of theophylline and possibly side effects
(e.g., headache, nausea, vomiting). To minimize competitive inhibition, two competing substrates
should be administered with as much time apart as possible.

Figure 3. CYP450 metabolic pathways involved in the metabolism of duloxetine and theophylline and
their respective affinities are depicted. Competitive inhibition between duloxetine and theophylline
will be expected at CYP1A2. Duloxetine acts as the perpetrator drug over theophylline, the victim drug.

133
Pharmaceutics 2020, 12, 846

Two Substrates with Largely Different Concentrations

As mentioned previously, the competitive inhibition process is sensitive to substrate concentrations.
If concentrations of the weaker affinity substrate are much higher than concentrations of the stronger
affinity substrate, the weaker affinity substrate can displace the stronger affinity substrate and overcome
the enzyme inhibition, which is why this type of inhibition is deemed reversible. (Figure 4) The greater
the difference there is between the affinity of the weaker affinity substrate and the stronger affinity
substrate, the more the concentration of the weaker affinity substrate needs to be increased to displace
the stronger affinity substrate. This situation can be observed clinically when very high concentrations
of a weak affinity substrate are present in the intestine or liver (high micromolar concentrations),
following its oral administration, while concentrations of another higher affinity substrate have long
been absorbed and distributed to various tissues leading to plasma concentrations in the low nanomolar
range. In this case, the extent of victim drug inhibition would be minimal. A direct application of this
principle is to alleviate the degree of inhibition by separating the time of administration of the two
competing drugs.

Figure 4. Illustration of reversible competitive inhibition where ligand A (orange) is a substrate with
strong affinity and ligand B (yellow) is a substrate with weaker affinity for a specific enzyme (purple).
When the concentrations of the weaker affinity substrate are sufficiently high, it can outcompete the
stronger affinity substrate for the active site of the enzyme.

3.1.2. Non-Competitive Inhibition

The non-competitive inhibitor does not generally have any structural resemblance to the substrate
as it binds to an allosteric site. The non-competitive inhibitor will cause a conformational change in the
structure of the active site such that the active site loses its affinity for the substrate. (Figure 5) Thus,
there is no direct competition between the inhibitor and the substrate at the active site. This type of
inhibition is often long lasting and cannot be overcome by increasing substrate concentrations. Under
these conditions, a decrease in the CLint of the substrate due to a decrease in its Vmax is observed. Similar
to competitive inhibitors, non-competitive inhibitors also have an almost immediate effect. As long as
the concentration of the inhibitor is not changed, the amount of inhibition will not increase over time.
This type of inhibition does not require the involvement of NADPH as a cofactor, i.e., the inhibitor is
not metabolized by the enzyme, but merely sits in an allosteric site. Other non-competitive inhibition
conditions may involve CYPb5 and/or CYP450 oxidoreductase as these factors have been shown to
modulate CYP450 activities, at least in in vitro systems [13,14]. Separating the time of dosing will not
alleviate non-competitive inhibition. Fluvoxamine (CYP2C19) and terbinafine (CYP2D6) are some
common examples of non-competitive inhibitors at other CYP isoforms [15–17].

134
Pharmaceutics 2020, 12, 846

Figure 5. Illustration of reversible non-competitive inhibition. An inhibitor (red) binds to an allosteric

site on the enzyme and causes conformational changes that prevent a substrate (orange) from binding
to the active site. Over time, as the inhibitor is flushed out, the conformation of the enzyme can return
to normal and substrate (orange) can bind to the active site again.

3.1.3. Mixed inhibition

In the case of mixed inhibition, both competitive and non-competitive inhibition occur. Mixed
inhibitors can simultaneously bind to both the heme iron atom (at the active site) and lipophilic regions
of the protein (allosteric site). Mixed inhibitors are usually more potent inhibitors than competitive or
non-competitive inhibitors. Ketoconazole and fluconazole, both imidazole antifungals, exhibit potent
mixed reversible inhibition of CYP3As. However, fluconazole is a weaker mixed reversible inhibitor
compared to ketoconazole, mainly due to its lower lipophilicity (less binding to an allosteric site).
For a CYP3A substrate like midazolam, concomitant use of non-competitive or mixed CYP3A
inhibitors will reduce its transformation to α-hydroxy midazolam, increasing midazolam plasma levels
and augmenting the risk of adverse drug events [18].

3.2. Irreversible CYP450 Inhibition

Several clinically important pharmacokinetic drug interactions result from a decrease in the
metabolic clearance of a substrate due to CYP450 irreversible inhibition. Mechanism-based inhibition
is a condition often encountered with irreversible CYP450 inhibitors.

3.2.1. Mechanism-Based Inhibition

Mechanism-based inhibition can be irreversible or quasi-irreversible. It generally derives from
the activation of a substrate drug by a CYP450 isoform into a reactive metabolite, which binds to
the enzyme heme prosthetic site (part of the active site), resulting in irreversible long-lasting loss of
enzyme activity (decrease in Vmax ). (Figure 6) Several drugs undergo metabolic activation by a specific
CYP450 isoform to produce inhibitory intermediate metabolites, which can form stable intermediate
complexes. As a result, the CYP450 isoform is sequestered in an inactive state. Even though the
reactive intermediate metabolite plays a key role in the mechanism-based inactivation of the CYP450
isoform, in many instances, the exact reactive metabolite involved in this phenomenon is unknown.
In the case of quasi-irreversible inhibition, the metabolites form very stable complexes with
the heme prosthetic site (metabolite–intermediate complex), so that the enzyme is sequestered in a
functionally inactive state. This phenomenon is called quasi-irreversible since, in theory, this complex
can be disrupted. In the case of irreversible inhibition, the metabolites covalently bind to the heme
prosthetic site or the protein part of the CYP450, leading to irreversible inactivation [19,20].

135
Pharmaceutics 2020, 12, 846

Figure 6. Illustration of mechanism-based inhibition. The mechanism-based inhibitor (orange) binds

to the active site as a substrate. During the normal process of metabolism, it forms either stable
intermediate–enzyme complexes or reactive electrophilic species that can lock up or destroy the enzyme,
and new enzyme synthesis is required to restore the enzymatic activity.

Hence, mechanism-based inhibition is active site mediated, and the allosteric site is not involved.
In contrast to reversible inhibition mechanisms, mechanism-based inhibition is time dependent and
NADPH dependent. This means that the enzyme has to start breaking down the substrate in order for
inhibition to proceed. As more drug molecules are metabolized, more complexes are stably formed in
the active sites, increasing inhibition over time before it reaches a plateau. Mechanism-based inhibition
is therefore also saturable. New enzyme formation is necessary to restore activity: the relationship
between the amount of intermediate complex formed and the speed of new enzyme synthesis dictate
the equilibrium and extent of enzyme inhibition.
Mechanism-based inhibitors can be classified into two categories: metabolic–intermediate complex
formation inhibitors and protein and/or heme alkylation inhibitors.

Metabolic–Intermediate Complex Formation (or Alternate Substrate Inhibition)

Such a condition occurs when a stable intermediate metabolite formed during the normal metabolic
cycle forms covalent bonds at the active site. This stable intermediate–enzyme complex is not easily
broken by increasing substrate concentration. Since the enzyme structure remains otherwise unchanged,
theoretically this reaction is reversible with time. However, in in vivo conditions, with this metabolic
intermediate complex being excessively stable, the metabolic intermediate cannot be displaced and the
enzyme remains inaccessible for metabolism so the reaction seems irreversible.
An example of alternate substrate inhibition is observed with paroxetine as its methoxy diene
carbon moiety was found to be responsible for the formation of covalent bonds at the active site of
CYP2D6 [21,22]. Another example of this type of inhibition was observed with clarithromycin when
the nitrosoalkene intermediate generated by N-demethylation forms covalent bonds with the active
site of CYP3A4 [23].

Protein and/or Heme Alkylation (or Suicide Inhibition)

This situation takes place when a latent highly reactive (generally electrophilic) intermediate is
formed in the catalysis process. The reactive intermediate forms covalent bonds (strong irreversible
bonds) with the enzyme in a step that is not part of the normal metabolic pathway. This process can
change the conformational structure of the enzyme significantly—it can even destroy the enzyme in
some cases—making it functionally unviable. For example, inhibition of CYP2C19 by esomeprazole
was found to be mediated by crosslinking the heme and apoprotein moieties in the enzyme, changing
its conformational structure [24].
It is important to note that since mechanism-based inhibitors are substrates of the enzyme,
they can also cause acute competitive inhibition when co-administered with other sensitive substrates.

136
Pharmaceutics 2020, 12, 846

The difference between competitive inhibition and mechanism-based inhibition is that as the time
period of exposure to mechanism-based inhibitors increases, the degree of inhibition also increases.
(Table 1)

Table 1. Summary of major pharmacokinetic characteristics of various drug inhibition models.

Inhibitor Type
Characteristics Non-Competitive
Competitive Mechanism-Based
Non-Mechanism Based
Metabolism required No No Yes
Active site mediated Yes No Yes
Time dependent No No Yes
Substrate concentration dependent Yes No Yes
Km (victim drug) ↑ ↔ ↔
Vmax (victim drug) ↔ ↓ ↓
CLint (victim drug) ↓ ↓ ↓

4. Clinical Cases

4.1. The Case of Omeprazole and Clopidogrel

Proton pump inhibitors (PPIs) are commonly prescribed along with antiplatelet drugs like
clopidogrel to reduce the incidence of gastric bleeding during treatment with antiplatelet therapy [25,26].
Omeprazole has long been one of the most widely used PPIs [27]. Omeprazole is a known
strong affinity substrate of CYP2C19, leading to the formation of its hydroxy and desmethyl
metabolites. The antiplatelet drug clopidogrel is sequentially activated by CYP450 isoforms, including
CYP2C19, into its active metabolite (H4) [28]. When omeprazole is co-administered with clopidogrel,
omeprazole acts as a strong affinity substrate of CYP2C19 (perpetrator), whereas clopidogrel is a
weaker sensitive substrate (victim). (Figure 7) Multiple in vitro studies have reported a potential
pharmacokinetic interaction between omeprazole and clopidogrel [29–32]. A clinical study conducted
by Angiolillo et al. demonstrated that plasma levels of clopidogrel’s active metabolite H4, and
consequently the platelet aggregation induced by adenosine diphosphate, were decreased when
omeprazole and clopidogrel were administered concomitantly [30]. Various other clinical studies
have demonstrated that co-administration of omeprazole and clopidogrel diminishes the antiplatelet
activity of clopidogrel [31,32].

Figure 7. CYP450 metabolic pathways involved in the metabolism of clopidogrel and omeprazole,
and their respective affinities are depicted. Competitive inhibition will be expected at CYP3A4 and
mechanism-based inhibition at the CYP2C19 enzymatic level. Clopidogrel is the victim drug and
omeprazole acts as the perpetrator drug.

137
Pharmaceutics 2020, 12, 846

Since omeprazole is a strong affinity substrate for CYP2C19, an “immediate” competitive inhibition
is expected between these two drugs. Since competitive inhibition was expected, separating time of
administration was considered a logical mitigation strategy to avoid or alleviate the extent of the drug
interaction [29]. Others have suggested that increasing the dose of clopidogrel might compensate for
the diminished formation of the active metabolite [29]. These recommendations to separate the time of
administration of the two drugs or to increase the dose of the victim drug (clopidogrel) come from
a sound rationale and have been proven to be efficacious in mitigating drug interactions associated
with competitive inhibition. However, it has been shown that following chronic administration,
separating the time of administration does not alleviate the reduction in clopidogrel active metabolite
(H4) caused by omeprazole [30,33]. This is due to the fact that omeprazole is not only a competitive
inhibitor, but also a mechanism-based inhibitor of CYP2C19, which results in a gradual increase in
irreversible inhibition of the CYP2C19 enzyme, to a point where clopidogrel activation and its clinical
efficacy are significantly impaired. From these observations, the FDA warns that separating the time of
administration between these two substrates will not alleviate this interaction [34].
Multiple studies have been conducted to determine the clinical impact of the potential reduced
antiplatelet efficacy resulting from this interaction. In two retrospective studies looking at interactions
between clopidogrel and PPIs and the effects on clinical outcomes, it was reported either that PPIs
were associated with increased cardiac adverse events in acute coronary syndrome patients, or that
cardiac adverse events were less common in PPI non-users [35,36]. Short-term mortality odds ratios
also favored PPI non-users, but no significant differences were observed in long-term mortality [35].
Though a wide range of PPIs were reviewed, omeprazole and esomeprazole remained the most widely
prescribed when all studies were combined. A similar study was conducted by Mahabaleshwarkar
et al., which found that PPIs were slightly, but significantly, associated with all-cause mortality [37].
The odds ratio of adverse cardiac events and all-cause mortality for omeprazole in particular was 1.23.
In another retrospective cohort study, clopidogrel use post discharge for acute coronary syndrome
hospitalizations was studied [38]. Concurrent clopidogrel and PPI use was associated with an increased
risk of death or rehospitalization; among patients prescribed a PPI, 60% were on omeprazole [38].
PPI plus clopidogrel use also remained significantly associated with recurrent acute coronary syndrome
and revascularization procedures [38]. Another study evaluated the association between various
PPIs (all PPIs combined) and individual PPI agents with clopidogrel use and increased risk of
hospitalization [39]. There was no significant association between any PPI and increased risk of
rehospitalization with clopidogrel, but this association was significant with omeprazole [39].
Several studies also report no change in the frequency of cardiovascular adverse events with
omeprazole administration during clopidogrel treatment [26,40–42]. Dosing regimens in these studies
suggest that the extent of interaction between clopidogrel (75 mg vs. 600 mg) and omeprazole (20 mg
vs. 80 mg) may be dose dependent. If an alternative PPI has to be considered, in vitro studies using
human liver microsomes have confirmed that PPIs like rabeprazole, lansoprazole, dexlansoprazole,
and pantoprazole do not show evidence of mechanism-based inhibition [29]. Clinical studies have
also reported that effects of lansoprazole and pantoprazole on clopidogrel antiplatelet activity are
not as potent as omeprazole [43–45]. When clopidogrel and PPI coadministration is necessary,
switching to a PPI other than omeprazole or esomeprazole may be considered, in light of the evidence
presented herein.

4.2. The Case of Paroxetine

The antidepressant paroxetine is a known substrate of CYP2D6, but also a potent mechanism-based
inhibitor of this enzyme. As such, paroxetine is expected to inhibit its own metabolism over time.
This has been illustrated in two clinical studies, where the Cmax and AUC of paroxetine were increased
5.2- and 7-fold after 2 weeks of paroxetine administration, respectively [46,47]. In the second study,
Laine et al. revealed that ultra-rapid metabolizers of CYP2D6 were converted to extensive or poor
metabolizers with chronic paroxetine use [47]. The FDA-approved label states that paroxetine takes

138
Pharmaceutics 2020, 12, 846

nearly 10 days to achieve steady-state concentrations even though the drug has a reported elimination
half-life of 21 h; so, within 4–5 days, steady-state levels should be reached under normal circumstances.
The label also states that saturation of CYP2D6 contributes to the non-linear pharmacokinetics of
paroxetine. Thus, it may be assumed that inhibition of its own mechanism contributes to achieving
later than expected steady-state levels and clinical efficacy [48].
As paroxetine is a strong CYP2D6 affinity substrate, it can exhibit acute competitive inhibition
when co-administered with sensitive substrates like nortriptyline. (Figure 8) Though separating the
time of administration may seem appropriate initially, the effects of mechanism-based inhibition over
time should be factored into medication risk management; dosage adjustment or substitution of the
victim substrate may be necessary. The extent of this interaction also depends on when paroxetine
and the victim drug are added to the regimen. If paroxetine is newly added to a victim drug that
has already reached steady-state plasma levels, the extent of inhibition and plasma concentrations of
the victim drug will increase over time until reaching a new steady state. However, if steady-state
levels of paroxetine are already achieved before adding another sensitive substrate, the extent of
inhibition will be maximum at initiation and will remain stable, since saturation of enzyme inhibition
is already established.

Figure 8. CYP450 metabolic pathways involved in the metabolism of nortriptyline and paroxetine
and their respective affinities for the isoform are depicted. Mechanism-based inhibition at CYP2D6
enzymatic level will be expected. Nortriptyline is the victim drug and paroxetine acts as the perpetrator
drug for the CYP2D6 elimination pathway.

4.3. The Case of Erythromycin

Another example that illustrates the concept of saturability in mechanism-based inhibition is
observed with the commonly used macrolide antibiotics erythromycin (Figure 9) or clarithromycin.
These drugs are known to cause mechanism-based inhibition of CYP3A4. Clinical studies have
demonstrated saturability (i.e., degree of inhibition reaches a maximum value) of enzyme inhibition
using CYP3A4-sensitive substrates like alfentanil or midazolam with or without erythromycin
pretreatment (treatment or control, respectively). In a study with healthy males after single or multiple
oral dose(s) of erythromycin 500 mg, the effects on alfentanil pharmacokinetics were measured [49].
A 25% increase of alfentanil half-life was observed following a single erythromycin dose compared
to control. After a 7-day pretreatment with erythromycin, the half-life of alfentanil was further
increased by 25% (up to 56% compared to control), suggesting an increase in inhibition with time [49].
Similar effects were seen on clearance. Although a direct association of this potential drug interaction
on clinical outcomes has not been systematically reported, two case reports suggest that erythromycin
pretreatment may cause prolonged respiratory depression when alfentanil is administered compared
to patients who did not receive erythromycin [50,51].

139
Pharmaceutics 2020, 12, 846

Figure 9. CYP450 metabolic pathways involved in the metabolism of erythromycin and midazolam
and their respective affinities for the isoform are depicted. Mechanism-based inhibition at the CYP3A4
enzymatic level will be expected. Midazolam is the victim drug and erythromycin acts as the
perpetrator drug.

In another study with 12 healthy volunteers, the effects of erythromycin on midazolam metabolism
were studied [52]. It was observed that the AUC of midazolam increased 2.3-fold after 2 days of
erythromycin pretreatment, compared to control. Following 4 days of pretreatment, midazolam’s
AUC increase was 3.38-fold. After a 7-day pretreatment, a similar increase (3.38-fold) was observed,
indicating an increase in inhibition with repeated administration of erythromycin and that a plateau
effect had been reached after 4 days of exposure [52].

4.4. The Case of Mirabegron

A widely prescribed drug in the treatment of overactive bladder, mirabegron, also displays
characteristics of mechanism-based inhibition for CYP2D6. (Figure 10) The difference in the degree of
metabolism inhibition between a competitive inhibitor and a mechanism-based inhibitor is perceived
when they are compared using the same victim drug. One study investigated the victim drug
desipramine, with the potential competitive inhibitor duloxetine and the mechanism-based inhibitor
mirabegron. In this study, duloxetine (moderate CYP2D6 affinity substrate) 30 mg twice a day
was administered for 10 days, after which desipramine (weak CYP2D6 affinity substrate) 50 mg was
administered as a single dose. Here, duloxetine would act as a competitive inhibitor and desipramine as
a victim drug. Accordingly, a 1.2-fold increase in AUC and 0.6-fold increase in Cmax of desipramine was
observed [53]. In a similar study design, desipramine was administered with or without mirabegron
pretreatment. First, desipramine was administered alone and, after a washout period, mirabegron
(100 mg) was administered for 13 days. On the fourteenth day, mirabegron 100 mg and desipramine
50 mg were co-administered. A 3.41-fold increase in the desipramine AUC was observed with
mirabegron pretreatment compared to control [54]. This increase in the AUC of desipramine was much
larger than the increase observed with duloxetine, a “purely” potential competitive inhibitor. It is
important to note that in the short term, mechanism-based inhibitors can act as competitive inhibitors
if the other drug has lower affinity for the metabolizing enzyme. Duloxetine and mirabegron are
both substrates of CYP2D6 with moderate affinity, exhibiting potential competitive inhibition over
desipramine; therefore, following a single dose of each drug, a similar level of CYP2D6 inhibition is
expected towards the CYP2D6 victim drug. (Figure 10) The higher inhibition observed with multiple
doses of mirabegron versus single dose of duloxetine is explained by the mechanism-based inhibition
observed over time. Similar effects of chronic mirabegron administration on metoprolol (a weak
CYP2D6 substrate) pharmacokinetics are also reported [54]. A 3.3-fold increase in the metoprolol
AUC was observed following a pretreatment of 5 days with mirabegron in CYP2D6 normal (i.e.,
previously called “extensive”) metabolizer subjects [55]. The coadministration of quinidine (a potent

140
Pharmaceutics 2020, 12, 846

CYP2D6 inhibitor) with metoprolol was associated with a similar magnitude of increase in metoprolol’s
AUC after a single dose [56]. The intensity of drug–drug interaction through “purely” competitive
inhibition is expected to be lower between substrates compared to quinidine’s inhibition; therefore,
the mechanism-based inhibition property of mirabegron can explain why the magnitude of drug–drug
interaction observed between mirabegron and metoprolol is similar to that observed between quinidine
and metoprolol.

Figure 10. CYP450 metabolic pathways involved in the metabolism of desipramine, duloxetine,
metoprolol, mirabegron, and their respective affinities for the isoform are depicted. Competitive
inhibition will be expected at CYP2D6 between duloxetine (perpetrator; CYP2D6 substrate with
higher affinity) and either desipramine or metoprolol (victim drugs; both CYP2D6 substrates with
weaker affinity). Mechanism-based inhibition at CYP2D6 will be expected between mirabegron and
desipramine, metoprolol, or duloxetine.

In addition to the case examples discussed above, Figure 11 provides a list of CYP450
mechanism-based inhibitors, along with the enzyme inhibited and relevant CYP450 pathways involved
in their metabolism. This list is not exhaustive, but provides a quick reference for commonly
used medications.
In addition to drug–drug interactions, high variability in terms of CYP450 expression and/or
activities can be explained by genetic polymorphisms in genes encoding specific isoforms (such as
CYP2C9, CYP2C19, and CYP2D6). This variability on CYP450 expression/activities translates
into intersubject variability in drug disposition and drug response. Often, the impact of genetic
polymorphisms and drug–drug interactions on CYP450s have been studied separately. However,
an interaction exists between these factors. Genetic polymorphisms could also contribute to variability
observed in the magnitude of drug–drug interactions observed between two drugs. So, genetic
polymorphisms in drug-metabolizing enzymes can affect the occurrence of phenoconversion induced
by drug inhibitors. As reported by Storelli et al., differences in CYP2D6 inhibition observed in vitro
with paroxetine (mechanism-based inhibitor) or duloxetine (competitive inhibitor) across CYP2D6
genotypes were not related to their inhibition parameters but likely due to a differential level of
functional enzymes as a function of the CYP2D6 genotype [57,58].

141
Pharmaceutics 2020, 12, 846

Figure 11. List of commonly prescribed medications producing mechanism-based inhibition.

5. Conclusions
Polypharmacy in many cases is deemed to be required and elderly patients are particularly
prone to this phenomenon. Aging is associated with the presence of multiple independent chronic
diseases and is almost always accompanied by multiple drug regimens. Polypharmacy has been
associated with many adverse clinical outcomes, such as drug–drug interactions, leading to adverse
drug events. Among these, mechanism-based inhibitor–victim drug combinations like clopidogrel
and omeprazole are commonly prescribed. Drugs for overactive bladder like mirabegron and tricyclic
antidepressants like paroxetine are also commonly prescribed in elderly populations. Polypharmacy
is not necessarily synonymous with inappropriate treatment, but in several situations, it can lead to
significant drug–drug interactions especially in the presence of mechanism-based inhibitor drugs,

142
Pharmaceutics 2020, 12, 846

as described in the current review. In these cases, polypharmacy might cause problems like blunted
efficacy of clopidogrel due to the co-administration of omeprazole, or increased toxicity of other drugs
co-administered with paroxetine or mirabegron. Clinicians must be able to recognize and intervene
appropriately based on the mechanism of these interactions.
As highlighted in this current review, mechanism-based inhibition can cause severe clinical
interactions. The magnitude of these interactions and their impacts depend on the duration of the
mechanism-based inhibitor exposure and the exact time-point when the victim drug is introduced
into the drug regimen. Without a comprehensive understanding of this mechanism, healthcare
professionals might find it difficult to diagnose and mitigate these interactions. Separating the time
of administration of interacting drugs cannot mitigate an interaction caused by mechanism-based
inhibition. A careful titration of dose, monitoring of clinical effects, or switching to an alternate drug
with a different metabolic pathway may become necessary to avoid such interactions. Advanced
clinical decision support systems that consider and distinguish competitive versus mechanism-based
inhibition drug–drug interactions would help identify potential interactions mediated by enzyme
inhibition. Using such tools can help pharmacists quickly identify and mitigate drug interactions,
thus helping reduce preventable drug interactions.

Author Contributions: Conceptualization, M.D. and V.M.; methodology, M.D., J.T., and V.M.; resources, M.D., P.D.,
J.T., and V.M.; writing—original draft preparation, M.D., S.B.A.R., M.J.A., L.D., P.D., J.T., and V.M.; writing—review
and editing, M.D., P.D., J.T., and V.M.; visualization, M.D.; supervision, J.T. and V.M.; project administration, P.D.,
J.T., and V.M. All authors have read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Acknowledgments: The authors recognize the contribution of Ernesto Lucio for his design of the included figures.
The authors would like to thank Dana Filippoli for her comprehensive review and comments pertaining to the
contents of this manuscript.
Conflicts of Interest: Malavika Deodhar, Sweilem Al Rihani, Meghan Arwood, Lucy Darakjian, Pamela Dow,
Jacques Turgeon, and Veronique Michaud are all employees and shareholders of Tabula Rasa HealthCare.
The company, TRHC, had no role in the design of the study; in the collection, analyses, or interpretation of data; in
the writing of the manuscript, or in the decision to publish the results.

References
1. Doan, J.; Zakrzewski-Jakubiak, H.; Roy, J.; Turgeon, J.; Tannenbaum, C. Prevalence and risk of potential
cytochrome P450-mediated drug-drug interactions in older hospitalized patients with polypharmacy.
Ann. Pharmacother. 2013, 47, 324–332. [CrossRef] [PubMed]
2. Mallet, L.; Spinewine, A.; Huang, A. The challenge of managing drug interactions in elderly people. Lancet
2007, 370, 185–191. [CrossRef]
3. Bankes, D.L.; Jin, H.; Finnel, S.; Michaud, V.; Knowlton, C.H.; Turgeon, J.; Stein, A. Association of a Novel
Medication Risk Score with Adverse Drug Events and Other Pertinent Outcomes Among Participants of the
Programs of All-Inclusive Care for the Elderly. Pharmacy 2020, 8, 87. [CrossRef] [PubMed]
4. Lynch, T.; Price, A. The effect of cytochrome P450 metabolism on drug response, interactions, and adverse
effects. Am. Fam. Physician 2007, 76, 391–396. [PubMed]
5. Wen, X.; Wang, J.S.; Backman, J.T.; Kivistö, K.T.; Neuvonen, P.J. Gemfibrozil is a potent inhibitor of human
cytochrome P450 2C9. Drug Metab. Dispos. 2001, 29, 1359–1361.
6. Shitara, Y.; Hirano, M.; Sato, H.; Sugiyama, Y. Gemfibrozil and its glucuronide inhibit the organic anion
transporting polypeptide 2 (OATP2/OATP1B1:SLC21A6)-mediated hepatic uptake and CYP2C8-mediated
metabolism of cerivastatin: Analysis of the mechanism of the clinically relevant drug-drug interaction
between cerivastatin and gemfibrozil. J. Pharmacol. Exp. Ther. 2004, 311, 228–236.
7. Lilja, J.J.; Backman, J.T.; Neuvonen, P.J. Effect of gemfibrozil on the pharmacokinetics and pharmacodynamics
of racemic warfarin in healthy subjects. Br. J. Clin. Pharmacol. 2005, 59, 433–439. [CrossRef]
8. Backman, J.T.; Kyrklund, C.; Neuvonen, M.; Neuvonen, P.J. Gemfibrozil greatly increases plasma
concentrations of cerivastatin. Clin. Pharmacol. Ther. 2002, 72, 685–691. [CrossRef]

143
Pharmaceutics 2020, 12, 846

9. Varma, M.V.S.; Lin, J.; Bi, Y.-A.; Kimoto, E.; Rodrigues, A.D. Quantitative Rationalization of Gemfibrozil
Drug Interactions: Consideration of Transporters-Enzyme Interplay and the Role of Circulating Metabolite
Gemfibrozil 1-O-β-Glucuronide. Drug Metab. Dispos. 2015, 43, 1108–1118. [CrossRef]
10. Niemi, M.; Backman, J.T.; Neuvonen, M.; Neuvonen, P.J. Effects of gemfibrozil, itraconazole, and their
combination on the pharmacokinetics and pharmacodynamics of repaglinide: Potentially hazardous
interaction between gemfibrozil and repaglinide. Diabetologia 2003, 46, 347–351. [CrossRef]
11. Jaakkola, T.; Backman, J.T.; Neuvonen, M.; Neuvonen, P.J. Effects of gemfibrozil, itraconazole, and their
combination on the pharmacokinetics of pioglitazone. Clin. Pharmacol. Ther. 2005, 77, 404–414. [CrossRef]
[PubMed]
12. Ogilvie, B.W.; Zhang, D.; Li, W.; Rodrigues, A.D.; Gipson, A.E.; Holsapple, J.; Toren, P.; Parkinson, A.
Glucuronidation converts gemfibrozil to a potent, metabolism-dependent inhibitor of CYP2C8: Implications
for drug-drug interactions. Drug Metab. Dispos. 2006, 34, 191–197. [CrossRef] [PubMed]
13. Zhang, H.; Gao, N.; Liu, T.; Fang, Y.; Qi, B.; Wen, Q.; Zhou, J.; Jia, L.; Qiao, H. Effect of Cytochrome b5 Content
on the Activity of Polymorphic CYP1A2, 2B6, and 2E1 in Human Liver Microsomes. PLoS ONE 2015, 10,
e0128547. [CrossRef] [PubMed]
14. Bart, A.G.; Scott, E.E. Structural and functional effects of cytochrome b(5) interactions with human cytochrome
P450 enzymes. J. Biol. Chem. 2017, 292, 20818–20833. [CrossRef] [PubMed]
15. Abdel-Rahman, S.M.; Gotschall, R.R.; Kauffman, R.E.; Leeder, J.S.; Kearns, G.L. Investigation of terbinafine
as a CYP2D6 inhibitor in vivo. Clin. Pharmacol. Ther. 1999, 65, 465–472. [CrossRef]
16. Vickers, A.E.M.; Sinclair, J.R.; Zollinger, M.; Heitz, F.; Glänzel, U.; Johanson, L.; Fischer, V. Multiple
Cytochrome P-450s Involved in the Metabolism of Terbinafine Suggest a Limited Potential for Drug-Drug
Interactions. Drug Metab. Dispos. 1999, 27, 1029–1038. [PubMed]
17. Rasmussen, B.B.; Nielsen, T.L.; Brøsen, K. Fluvoxamine inhibits the CYP2C19-catalysed metabolism of
proguanil in vitro. Eur. J. Clin. Pharmacol. 1998, 54, 735–740. [CrossRef] [PubMed]
18. Lam, Y.W.; Alfaro, C.L.; Ereshefsky, L.; Miller, M. Pharmacokinetic and pharmacodynamic interactions of
oral midazolam with ketoconazole, fluoxetine, fluvoxamine, and nefazodone. J. Clin. Pharmacol. 2003, 43,
1274–1282. [CrossRef]
19. Drolet, B.; Khalifa, M.; Daleau, P.; Hamelin, B.A.; Turgeon, J. Block of the rapid component of the delayed
rectifier potassium current by the prokinetic agent cisapride underlies drug-related lengthening of the QT
interval. Circulation 1998, 97, 204–210. [CrossRef]
20. Naritomi, Y.; Teramura, Y.; Terashita, S.; Kagayama, A. Utility of microtiter plate assays for human cytochrome
P450 inhibition studies in drug discovery: Application of simple method for detecting quasi-irreversible and
irreversible inhibitors. Drug Metab. Pharmacokinet. 2004, 19, 55–61. [CrossRef]
21. Uttamsingh, V.; Gallegos, R.; Liu, J.F.; Harbeson, S.L.; Bridson, G.W.; Cheng, C.; Wells, D.S.; Graham, P.B.;
Zelle, R.; Tung, R. Altering Metabolic Profiles of Drugs by Precision Deuteration: Reducing Mechanism-Based
Inhibition of CYP2D6 by Paroxetine. J. Pharmacol. Exp. Ther. 2015, 354, 43–54. [CrossRef] [PubMed]
22. Bertelsen, K.M.; Venkatakrishnan, K.; Von Moltke, L.L.; Obach, R.S.; Greenblatt, D.J. Apparent
mechanism-based inhibition of human CYP2D6 in vitro by paroxetine: Comparison with fluoxetine and
quinidine. Drug Metab. Dispos. 2003, 31, 289–293. [CrossRef] [PubMed]
23. Kouladjian, L.; Chen, T.F.; Gnjidic, D.; Hilmer, S.N. Education and Assessment of Pharmacists on the Use of
the Drug Burden Index in Older Adults Using a Continuing Professional Development Education Method.
Am. J. Pharm. Educ. 2016, 80, 63. [CrossRef] [PubMed]
24. Ogilvie, B.W. An In Vitro Investigation into the Mechanism of the Clinically Relevant Drug-Drug Interaction
between Omeprazole or Esomeprazole and Clopidogrel. Ph.D. Thesis, University of Kansas, Lawrence, KS,
USA, 23 April 2015.
25. Ray, W.A.; Murray, K.T.; Griffin, M.R.; Chung, C.P.; Smalley, W.E.; Hall, K.; Daugherty, J.R.; Kaltenbach, L.A.;
Stein, C.M. Outcomes with concurrent use of clopidogrel and proton-pump inhibitors: A cohort study.
Ann. Int. Med. 2010, 152, 337–345. [CrossRef]
26. Bhatt, D.L.; Cryer, B.L.; Contant, C.F.; Cohen, M.; Lanas, A.; Schnitzer, T.J.; Shook, T.L.; Lapuerta, P.;
Goldsmith, M.A.; Laine, L.; et al. Clopidogrel with or without omeprazole in coronary artery disease. N. Engl.
J. Med. 2010, 363, 1909–1917. [CrossRef]
27. Sidney, M.; Wolfe, M.D. Proton Pump Inhibitors: Dangerous and Habit-Forming Heartburn Drugs; Citizen, P., Ed.;
The Citizen: Washington, DC, USA, 2011; Volume 27, p. 9.

144
Pharmaceutics 2020, 12, 846

28. Karaźniewicz-Łada, M.; Danielak, D.; Burchardt, P.; Kruszyna, L.; Komosa, A.; Lesiak, M.; Główka, F.
Clinical pharmacokinetics of clopidogrel and its metabolites in patients with cardiovascular diseases.
Clin. Pharmacokinet. 2014, 53, 155–164. [CrossRef]
29. Zvyaga, T.; Chang, S.-Y.; Chen, C.; Yang, Z.; Vuppugalla, R.; Hurley, J.; Thorndike, D.; Wagner, A.;
Chimalakonda, A.; Rodrigues, A.D. Evaluation of Six Proton Pump Inhibitors As Inhibitors of Various
Human Cytochromes P450: Focus on Cytochrome P450 2C19. Drug Metab. Dispos. 2012, 40, 1698–1711.
[CrossRef]
30. Angiolillo, D.J.; Gibson, C.M.; Cheng, S.; Ollier, C.; Nicolas, O.; Bergougnan, L.; Perrin, L.; LaCreta, F.P.;
Hurbin, F.; Dubar, M. Differential Effects of Omeprazole and Pantoprazole on the Pharmacodynamics
and Pharmacokinetics of Clopidogrel in Healthy Subjects: Randomized, Placebo-Controlled, Crossover
Comparison Studies. Clin. Pharmacol. Ther. 2011, 89, 65–74. [CrossRef]
31. Furtado, R.H.M.; Giugliano, R.P.; Strunz, C.M.C.; Filho, C.C.; Ramires, J.A.F.; Filho, R.K.; Neto, P.A.L.;
Pereira, A.C.; Rocha, T.R.; Freire, B.T.; et al. Drug Interaction Between Clopidogrel and Ranitidine or
Omeprazole in Stable Coronary Artery Disease: A Double-Blind, Double Dummy, Randomized Study. Am. J.
Cardiovasc. Drugs 2016, 16, 275–284. [CrossRef]
32. Gilard, M.; Arnaud, B.; Cornily, J.-C.; Le Gal, G.; Lacut, K.; Le Calvez, G.; Mansourati, J.; Mottier, D.;
Abgrall, J.-F.; Boschat, J. Influence of Omeprazole on the Antiplatelet Action of Clopidogrel Associated With
Aspirin: The Randomized, Double-Blind OCLA (Omeprazole CLopidogrel Aspirin) Study. J. Am. Coll.
Cardiol. 2008, 51, 256–260. [CrossRef]
33. Frelinger, A.L., III; Lee, R.D.; Mulford, D.J.; Wu, J.; Nudurupati, S.; Nigam, A.; Brooks, J.K.; Bhatt, D.L.;
Michelson, A.D. A randomized, 2-period, crossover design study to assess the effects of dexlansoprazole,
lansoprazole, esomeprazole, and omeprazole on the steady-state pharmacokinetics and pharmacodynamics
of clopidogrel in healthy volunteers. J. Am. Coll. Cardiol. 2012, 59, 1304–1311. [CrossRef] [PubMed]
34. U.S. Food and Drug Administration. Information for Healthcare Professionals: Update to the
Labeling of Clopidogrel Bisulfate (Marketed as Plavix) to Alert Healthcare Professionals about
a Drug Interaction with Omeprazole (Marketed as Prilosec and Prilosec OTC). Available online:
http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/
DrugSafetyInformationforHeathcareProfessionals/ucm190787 (accessed on 25 July 2020).
35. Bundhun, P.K.; Teeluck, A.R.; Bhurtu, A.; Huang, W.-Q. Is the concomitant use of clopidogrel and Proton Pump
Inhibitors still associated with increased adverse cardiovascular outcomes following coronary angioplasty?
A systematic review and meta-analysis of recently published studies (2012–2016). BMC Cardiovasc. Disord.
2017, 17, 3. [CrossRef] [PubMed]
36. Bhurke, S.M.; Martin, B.C.; Li, C.; Franks, A.M.; Bursac, Z.; Said, Q. Effect of the clopidogrel-proton pump
inhibitor drug interaction on adverse cardiovascular events in patients with acute coronary syndrome.
Pharmacotherapy 2012, 32, 809–818. [CrossRef] [PubMed]
37. Mahabaleshwarkar, R.K.; Yang, Y.; Datar, M.V.; Bentley, J.P.; Strum, M.W.; Banahan, B.F.; Null, K.D. Risk of
adverse cardiovascular outcomes and all-cause mortality associated with concomitant use of clopidogrel and
proton pump inhibitors in elderly patients. Curr. Med. Res. Opin. 2013, 29, 315–323. [CrossRef] [PubMed]
38. Ho, P.M.; Maddox, T.M.; Wang, L.; Fihn, S.D.; Jesse, R.L.; Peterson, E.D.; Rumsfeld, J.S. Risk of Adverse
Outcomes Associated With Concomitant Use of Clopidogrel and Proton Pump Inhibitors Following Acute
Coronary Syndrome. JAMA 2009, 301, 937–944. [CrossRef] [PubMed]
39. Lin, C.F.; Shen, L.J.; Wu, F.L.; Bai, C.H.; Gau, C.S. Cardiovascular outcomes associated with concomitant use
of clopidogrel and proton pump inhibitors in patients with acute coronary syndrome in Taiwan. Br. J. Clin.
Pharmacol. 2012, 74, 824–834. [CrossRef] [PubMed]
40. Vaduganathan, M.; Cannon, C.P.; Cryer, B.L.; Liu, Y.; Hsieh, W.H.; Doros, G.; Cohen, M.; Lanas, A.;
Schnitzer, T.J.; Shook, T.L.; et al. Efficacy and Safety of Proton-Pump Inhibitors in High-Risk Cardiovascular
Subsets of the COGENT Trial. Am. J. Med. 2016, 129, 1002–1005. [CrossRef]
41. Sherwood, M.W.; Melloni, C.; Jones, W.S.; Washam, J.B.; Hasselblad, V.; Dolor, R.J. Individual Proton
Pump Inhibitors and Outcomes in Patients with Coronary Artery Disease on Dual Antiplatelet Therapy:
A Systematic Review. J. Am. Heart Assoc. 2015, 4, e002245. [CrossRef]

145
Pharmaceutics 2020, 12, 846

42. Yano, H.; Tsukahara, K.; Morita, S.; Endo, T.; Sugano, T.; Hibi, K.; Himeno, H.; Fukui, K.; Umemura, S.;
Kimura, K. Influence of omeprazole and famotidine on the antiplatelet effects of clopidogrel in addition to
aspirin in patients with acute coronary syndromes: A prospective, randomized, multicenter study. Circ. J.
2012, 76, 2673–2680. [CrossRef]
43. Lau, W.C.; Gurbel, P.A. The drug–drug interaction between proton pump inhibitors and clopidogrel. Can.
Med. Assoc. J. 2009, 180, 699–700. [CrossRef]
44. Juurlink, D.N.; Gomes, T.; Ko, D.T.; Szmitko, P.E.; Austin, P.C.; Tu, J.V.; Henry, D.A.; Kopp, A.; Mamdani, M.M.
A population-based study of the drug interaction between proton pump inhibitors and clopidogrel. Can. Med.
Assoc. J. 2009, 180, 713–718. [CrossRef] [PubMed]
45. Zhang, J.R.; Wang, D.Q.; Du, J.; Qu, G.S.; Du, J.L.; Deng, S.B.; Liu, Y.J.; Cai, J.X.; She, Q. Efficacy of
Clopidogrel and Clinical Outcome When Clopidogrel Is Coadministered With Atorvastatin and Lansoprazole:
A Prospective, Randomized, Controlled Trial. Medicine 2015, 94, e2262. [CrossRef] [PubMed]
46. Chen, R.; Shen, K.; Hu, P. Single- and multiple-dose pharmacokinetics and tolerability of a paroxetine
controlled-release tablet in healthy Chinese subjects. Int. J. Clin. Pharmacol. Ther. 2017, 55, 231–236.
[CrossRef] [PubMed]
47. Laine, K.; Tybring, G.; Härtter, S.; Andersson, R.N.K.; Svensson, J.-O.; Widén, J.; Bertilsson, L. Inhibition of
cytochrome P4502D6 activity with paroxetine normalizes the ultrarapid metabolizer phenotype as measured
by nortriptyline pharmacokinetics and the debrisoquin test. Clin. Pharmacol. Ther. 2001, 70, 327–335.
[CrossRef]
48. U.S. Food and Drug Administration. Paxil [Drug Label]. Available online: https://www.accessdata.fda.gov/
drugsatfda_docs/label/2008/020031s060,020936s037,020710s024lbl.pdf (accessed on 31 July 2020).
49. Bartkowski, R.R.; Goldberg, M.E.; Larijani, G.E.; Boerner, T. Inhibition of alfentanil metabolism by
erythromycin. Clin. Pharmacol. Ther. 1989, 46, 99–102. [CrossRef]
50. Bartkowski, R.R.; McDonnell, T.E. Prolonged Alfentanil Effect Following Erythromycin Administration.
Anesthesiology 1990, 73, 566–567. [CrossRef]
51. Yate, P.M.; Short, S.M.; Sebel, P.S.; Thomas, D.; Morton, J. Comparison of infusions of alfentanil or pethidine
for sedation of ventilated patients on THE ITU. Br. J. Anaesth. 1986, 58, 1091–1099. [CrossRef]
52. Okudaira, T.; Kotegawa, T.; Imai, H.; Tsutsumi, K.; Nakano, S.; Ohashi, K. Effect of the Treatment Period With
Erythromycin on Cytochrome P450 3A Activity in Humans. J. Clin. Pharmacol. 2007, 47, 871–876. [CrossRef]
53. Patroneva, A.; Connolly, S.M.; Fatato, P.; Pedersen, R.; Jiang, Q.; Paul, J.; Guico-Pabia, C.; Isler, J.A.;
Burczynski, M.E.; Nichols, A.I. An assessment of drug-drug interactions: The effect of desvenlafaxine and
duloxetine on the pharmacokinetics of the CYP2D6 probe desipramine in healthy subjects. Drug Metab.
Dispos. 2008, 36, 2484–2491. [CrossRef]
54. U.S. Food and Drug Administration. Consultation for NDA 202611. In Clinical Pharmacology and
Biopharmaceutics Review(s), Center for Drug Evaluation and Research Division of Cardiovascular and Renal
Products; U.S. Food and Drug Administration: Pharr, TX, USA, 2012; p. 218.
55. Krauwinkel, W.; Dickinson, J.; Schaddelee, M.; Meijer, J.; Tretter, R.; van de Wetering, J.; Strabach, G.;
van Gelderen, M. The effect of mirabegron, a potent and selective β3-adrenoceptor agonist, on the
pharmacokinetics of CYP2D6 substrates desipramine and metoprolol. Eur. J. Drug Metab. Pharmacokinet.
2014, 39, 43–52. [CrossRef]
56. Bramer, S.L.; Suri, A. Inhibition of CYP2D6 by Quinidine and its Effects on the Metabolism of Cilostazol.
Clin. Pharmacokinet. 1999, 37, 41–51. [CrossRef] [PubMed]
57. Storelli, F.; Matthey, A.; Lenglet, S.; Thomas, A.; Desmeules, J.; Daali, Y. Impact of CYP2D6 Functional Allelic
Variations on Phenoconversion and Drug-Drug Interactions. Clin. Pharmacol. Ther. 2018, 104, 148–157.
[CrossRef] [PubMed]
58. Storelli, F.; Desmeules, J.; Daali, Y. Genotype-sensitive reversible and time-dependent CYP2D6 inhibition in
human liver microsomes. Basic Clin. Pharmacol. Toxicol. 2019, 124, 170–180. [CrossRef] [PubMed]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

146
 pharmaceutics

Review
Role of OATP1B1 and OATP1B3 in Drug-Drug
Interactions Mediated by Tyrosine Kinase Inhibitors
Dominique A. Garrison † , Zahra Talebi † , Eric D. Eisenmann, Alex Sparreboom *
and Sharyn D. Baker *
Division of Pharmaceutics and Pharmacology, College of Pharmacy, The Ohio State University, Columbus,
OH 43210, USA; garrison.220@osu.edu (D.A.G.); talebi.9@osu.edu (Z.T.); eisenmann.11@osu.edu (E.D.E.)
* Correspondence: sparreboom.1@osu.edu (A.S.); baker.2480@osu.edu (S.D.B.);
Tel.: +1-614-685-6014 (A.S.); +1-614-685-6016 (S.D.B.)
† These authors contributed equally to this work.

Received: 15 August 2020; Accepted: 2 September 2020; Published: 9 September 2020

Abstract: Failure to recognize important features of a drug’s pharmacokinetic characteristics is

a key cause of inappropriate dose and schedule selection, and can lead to reduced efficacy and
increased rate of adverse drug reactions requiring medical intervention. As oral chemotherapeutic
agents, tyrosine kinase inhibitors (TKIs) are particularly prone to cause drug-drug interactions as
many drugs in this class are known or suspected to potently inhibit the hepatic uptake transporters
OATP1B1 and OATP1B3. In this article, we provide a comprehensive overview of the published
literature and publicly-available regulatory documents in this rapidly emerging field. Our findings
indicate that, while many TKIs can potentially inhibit the function of OATP1B1 and/or OATP1B3 and
cause clinically-relevant drug-drug interactions, there are many inconsistencies between regulatory
documents and the published literature. Potential explanations for these discrepant observations
are provided in order to assist prescribing clinicians in designing safe and effective polypharmacy
regimens, and to provide researchers with insights into refining experimental strategies to further
predict and define the translational significance of TKI-mediated drug-drug interactions.

Keywords: OATP1B1; OATP1B3; tyrosine kinase inhibitors; drug-drug interactions

1. Introduction
The economic burden of drug-related morbidity and mortality as a result of non-optimized
medication therapy is estimated to be more than 16% of total US health care annual expenditures [1].
Overlooking major pharmacokinetic characteristics of a drug is one of the key players in inappropriate
pharmaceutical dosing, which can lead to reduced efficacy and an increased rate of adverse drug
reactions (ADRs) requiring medical intervention [2]. Pharmacokinetic drug-drug interactions (DDIs)
can be responsible for about half of all DDIs depending on the patient group [3,4]. Furthermore,
these DDIs have the potential to cause very pronounced (several hundred-fold) and abrupt changes
in concentration and effect of the victim drug, depending on the start and stop of the causative
(perpetrator) comedication and on fluctuations of its concentration during therapy [2,5,6].
Different components in absorption, distribution, metabolism and excretion can affect the
overall pharmacokinetic profile of drugs. For agents that primarily undergo hepatic elimination,
transport-mediated mechanisms of hepatocellular uptake can have a particularly significant clinical
impact on pharmacotherapy; thus, this field of research has gained increased attention in recent
years [7]. The organic anion transporting polypeptides OATP1B1 and OATP1B3 are examples of such
transporters that can facilitate the uptake of a diverse array of xenobiotics, including many anticancer
drugs, into the liver in advance of metabolism, and that are sensitive to inhibition by other medicines
given concurrently.

147
Pharmaceutics 2020, 12, 856; doi:10.3390/pharmaceutics12090856 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 856

Two of the most commonly acknowledged risk factors of DDIs are polypharmacy and advanced
age [2,8–10]. Consistent with this notion, cancer patients are particularly at high risk for the occurrence
of potentially harmful DDIs, since they often take a large number of medications concomitantly,
which tends to increase as their disease progresses, and because the majority of cancer diagnoses
happens in older ages [10,11]. Indeed, prior investigations have demonstrated that as many as
30% of cancer patients receiving chemotherapeutic treatment are at a risk for DDIs [12,13]. As the
number of new treatment options in oncology continues to grow, DDIs are increasingly recognized
as significant health hazards that can negatively influence treatment outcomes. These issues are
particularly concerning given the increasing use orally-administered chemotherapeutic agents. While
such drugs offer advantages in terms of patient preference, the convenience of use, reduced healthcare
resource utilization, the possibility to achieve sustained drug exposure associated with the need for
chronic use without requiring prolonged drug infusions, and may improve the overall quality of life,
recent studies have suggested that the use of such agents increases the risk of potentially serious
DDIs with commonly used outpatient medications [14]. In addition, unsupervised administration of
other medications as well as their possibly prolonged use has been advanced as concerns with oral
chemotherapy drugs, which could potentiate DDIs that may remain unanticipated. Although recent
studies have suggested that the prevalence of DDIs with oral chemotherapy drugs is as high as 50%
with nearly 20% potentially increasing toxicity, the clinical impact of DDIs involving oral chemotherapy
remains largely unstudied [10].
In this article, we provide an overview of this field of research in relation to tyrosine kinase
inhibitors (TKIs), a rapidly expanding group of orally-administered drugs commonly used in the
treatment of solid tumors and hematological malignancies, with particular emphasis on OATP1B1-
and OATP1B3-related mechanisms. In addition to reviewing existing published data, we aimed to
identify potential knowledge gaps that could help improve our understanding of the clinical impact of
DDIs mediated through this mechanism.

2. Tyrosine Kinase Inhibitors (TKIs)

Since the US Food and Drug Administration (FDA) approval of the first TKI, imatinib, in 2001 for
the treatment of chronic myeloid leukemia (CML), almost 50 additional TKIs have been approved for
the treatment of various cancers, and many more are currently being developed and evaluated [15,16].
Protein tyrosine kinases (PTKs) are enzymes that catalyze the transfer of a gamma phosphate group
from adenosine triphosphate (ATP) to a tyrosine residue on a protein. The phosphorylation of PTKs
leads to the downstream activation of signal transduction pathways that are important in the regulation
of cell growth, differentiation, and a series of other physiological and biochemical processes involved
in cell survival and migration. Dysregulation of PTK function results in proliferation disorders, with
those most notably being cancers [17–19]. Because of their importance in signal transduction, many
PTKs have been the target of therapeutic intervention with the use of small-molecule TKIs. As a
result, TKIs function by competing with ATP for the ATP-binding pocket of PTKs, thus reducing the
downstream signaling cascade and provide useful targeted strategies in oncogenic treatment [20,21].
While TKIs have revolutionized anticancer therapy, some challenges have also risen in the use of
these agents. Unlike conventional cytotoxic agents that are given intravenously, TKIs are administered
orally and daily for prolonged periods [22]. As mentioned before, while this is more convenient, this
also increases their susceptibility to unpredictable patterns of oral absorption and causes both wide
inter-individual pharmacokinetic variability and potential for DDIs with co-administered agents [23–25].
Most TKIs are highly prone to cause DDIs [26], as patients receiving these agents are often subsequently
treated for concomitant diseases, and because polypharmacy is highly prevalent [25]. Comorbid
conditions such as hypertension, chronic obstructive pulmonary disease, diabetes, cardiovascular
disease, congestive heart failure, and peripheral vascular disease are frequently reported in the
population of cancer patients [27], and this further increases the risk for potential DDIs. Indeed, a recent
study indicated that 97.1% of patients receiving treatment with TKIs were using at least one other

148
Pharmaceutics 2020, 12, 856

drug simultaneously, with a median of 4 concurrent medications, and 47.4% experienced at least one
potential TKI-mediated DDI [28]. In another study, 44.7% of the potential DDIs identified involving
TKIs were considered severe [29]. Interestingly, most available data in this field have investigated
TKIs as victims in DDIs [30–33], and conclusive information on their role as perpetrators in DDIs is
generally lacking.

3. Organic Anion Transporting Polypeptides (OATPs)

The vast majority of orally-administered TKIs are eliminated from the body by enzyme-mediated
metabolism, which occurs predominantly in the liver, followed by biliary or urinary excretion of the
metabolites. These processes require drugs to cross the selectively permeable biological membrane of
hepatocytes and are dependent, at least in part, on interaction with membrane transporters. These
include the organic anion transporting polypeptides (OATPs), a family of influx transporters expressed
in various tissues, including the liver [34–36]. Experimental studies with TKIs have predominantly
evaluated transport by the liver-specific transporters OATP1B1 and OATP1B3, which are encoded
by the SLCO1B1 and SLCO1B3 genes [37], respectively. Moreover, it has also been shown that some
TKIs can additionally act as inhibitors of the transporters for which they are substrates [38]. Inhibition
of OATPs can lead to defective elimination, result in sudden increases in plasma concentration and
area under the curve (AUC) for drugs that are substrates of these transporters [36], and ultimately
increase the risk of therapy-related side effects. Known substrates of OATP1B1 and OATP1B3 include
statins, repaglinide, olmesartan, enalapril, valsartan, several xenobiotic glucuronide metabolites, as
well as a host of cytotoxic chemotherapeutic agents, including the taxanes paclitaxel and docetaxel,
the platinum-based drug cisplatin, and methotrexate. As hypertension and diabetes are among the
prevalent comorbidities in cancer patients, many xenobiotic OATP1B1 and OATP1B3 substrate drugs
are likely to be co-administrated with OATP-inhibitory TKIs, and therefore, clinically significant
toxicities such as rhabdomyolysis, hyperkalemia, and hypoglycemia can be anticipated [39–41].

4. Regulatory Guidance Documents

As more and more DDIs involving uptake transporters have been reported in recent years, so
have regulatory agencies such as the FDA and the European Medicines Agency (EMA) put increasing
emphasis on investigating each new drug entity for their potential to induce/inhibit such transporters.
It should be noted that both the “EMA Guideline on the Investigation of Drug Interactions” and “FDA
guidance for In Vitro Drug Interaction Studies—Cytochrome P450 Enzyme- and Transporter-Mediated
Drug Interactions” recognize the fact that the field of transporter interaction assessments is still rapidly
evolving and therefore the recommendations offered are relatively flexible and advocate the use of a
variety of methods. However, some specifications have been proposed as a means to ensure that the
in vitro models have optimal prediction potential for transporter-mediated interactions:

- Both the FDA and EMA documents suggest that the sponsor should conduct in vitro studies to
evaluate whether an investigational drug is an inhibitor of OATP1B1 and/or OATP1B3.
- Both documents recommend using an appropriate, predictive in vitro models, such as human
hepatocytes or mammalian cells engineered to overexpress transporters of interest (e.g., CHO,
HEK293, MDCK) to explore potential transporter interactions.
- Different concentrations of the investigational drug on the transport of a specific substrate should
be investigated, such that at least 3 and 4 concentrations should be tested, according to EMA and
FDA guidance documents, respectively, and values for the inhibition constant (Ki) should be
obtained, with known inhibitors present as controls.
- According to EMA, Ki values that are lower than a concentration representing 25-fold the unbound
hepatic inlet concentration after oral administration warrant the conduct of an in vivo DDI study
with the use of a prototypical probe substrate. The most recent FDA guidance, which aligns with
the EMA, uses unbound concentrations of the investigational drug, not the total drug, for the

149
Pharmaceutics 2020, 12, 856

calculation of R values with the formula R = 1 + ((fu,p × Iin,max)/ IC50 ) where fu,p is the unbound
fraction in plasma, IC50 is the half-maximal inhibitory concentration and Iin, max is the estimated
maximum plasma inhibitor concentration at the inlet to the liver. An R-value ≥ 1.1 suggests that
the drug has the potential to inhibit OATP1B1 and/or OATP1B3 in vivo.
- The 2017 version of the FDA guidance on in vitro assessment of DDIs requires a strategy employing
a 30-min preincubation with the inhibitor before the addition of substrate. Although this design
is recommended as it may lead to changes in the observed IC50 values, the latest version of the
guidance does not specify an exact duration of the preincubation conditions.
- The FDA guidance also mentions that the observed degree of inhibition by a particular agent can
be dependent on the substrate used in the experiment, and therefore it has been suggested that
substrates more likely to be used in clinical studies, or substrates that usually generate lower IC50
values for known inhibitors should be chosen in in vitro investigations to avoid underestimation
of effects in vivo.

5. Identification and Retrieval of Relevant Data

Acquisition of the data for this article was compiled independently up to and including June 2020
by various members of the Division of Pharmaceutics and Pharmacology at the Ohio State University
with specific expertise in drug transporters (D.A.G.), pharmacy (Z.T.), and cancer pharmacology
(E.D.E.), and subsequently reviewed by members with expertise in pharmacokinetics (A.S.) and TKIs
(S.D.B.). Data on FDA-approved TKIs was extracted from the full prescribing information as provided
by the respective drug manufacturers. A search was subsequently conducted using publicly-available,
unpublished databases from the FDA and EMA guidance documents for industry to further collect
information on OATP1B1 and OATP1B3 inhibition studies previously conducted for each of the TKIs
(Figure 1). It should be noted that although published studies have indicated that certain TKIs such as
erlotinib are inhibitors of 2B1 and can cause DDIs, this was considered beyond the scope of the present
article since regulatory guidance documents lack information on this transporter [42].

Figure 1. Applied methods for the acquisition of relevant data on TKI-related interactions with
OATP1B1 and OATP1B3.

All DDI data included for consideration focused exclusively on the TKIs as inhibitors of the
transporter (the perpetrator) of interest. The selection of relevant literature articles for inclusion was
performed based on predefined inclusion/exclusion criteria, where eligible articles included either
peer-reviewed publications, meeting abstracts, and previously published reviews. As a primary
search module, PubMed (National Library of Medicine) was utilized to identify potentially relevant
publications using the following MeSH terms in the search strategy: [“TKI of interest”] AND [OATP1B1]
or [“TKI of interest”] AND [OATP1B3]. Google Scholar was consecutively consulted to ensure no
published article of relevance to this literature review was omitted. Three authors (D.A.G., Z.T.,
and E.D.E.) independently reviewed the collected data for eligibility and accuracy. In our analysis,

150
Pharmaceutics 2020, 12, 856

concordant outcomes were defined as those for which the prescribing information, documentation from
the FDA and/or EMA, and all the retrieved published literature on a specific TKI were in agreement
that the TKI was either an inhibitor or not an inhibitor of OATP1B1 and/or OATP1B3. Outcomes were
considered discordant outcomes if the identified reports on a particular TKI regarding its inhibitory
properties towards OATP1B1 and/or OATP1B3 were conflicting. All data of relevance was tabulated to
highlight such discrepancies (see below).

6. Effects of TKIs on the Function of OATP1B1 and OATP1B3

A descriptive summary of the main findings resulting from surveying the available prescribing
information (PIs), and FDA and EMA guidance documents are shown in Table 1. The PIs showed
that of the 48 FDA-approved TKIs evaluated, 7 (15%) are claimed to be inhibitors of OATP1B1 and
5 (10%) are inhibitors of OATP1B3. In addition, it is reported that of those 48 TKIs, 22 (48%) and 21
(44%) are reported in the PIs to not be inhibitors of OATP1B1 or OATP1B3, respectively. However,
it is of note that the PIs for 19 (40%) of the TKIs do not mention whether or not drug interactions with
OATP1B1 are of concern, and 22 (46%) do not mention that information for OATP1B3. As shown in
Table 1, some inconsistencies were observed for some TKIs between what is reported in the regulatory
guidance. Many of the differences can be accounted for by differences in cutoff for IC50 values (shown
in Supplementary Materials Tables S1–S9).

Table 1. Comparison of regulatory guidance documents on OATP1B inhibition by FDA-approved TKIs.

Kinase
TKI Disease Indication OATP1B1 OATP1B3
Target
PI FDA EMA PI FDA EMA
Bacritinib Rheumatoid Arthritis JAK No No No Yes Yes No
Ceritinib Metastatic Non-Small Cell Lung Cancer ALK No Yes No No Yes No
ALK,
Crizotinib Metastatic Non-Small Cell Lung Cancer No Yes - No Yes -
ROS1
Laroctrectinib Solid Tumors NTKR No No Yes No No No
Differentiated Thyroid Cancer, Renal Cell
Lenvatinib VEGFR No Yes Yes No No No
Carcinoma, Hepatocellular Carcinoma
Anaplastic Lymphoma Positive Metastatic
Lorlatinib ALK No No Yes No No Yes
Non-Small Cell Lung Cancer
Acute Myeloid Leukemia, Aggressive
Systemic Mastocytosis, Associated
Midostaurin FLT3 Yes Yes Yes - Yes No
Hematological Neoplasm, Mast Cell
Leukemia
Osimertinib Metastatic Non-Small Cell Lung Cancer EGFR No No Yes No No Yes
“Yes” indicates a TKI as an OATP1B1/3 inhibitor provided by the prescribing information, FDA documents, or EMA
documents. “No” indicates a TKI is not an inhibitor of OATP1B1/3 inhibitor provided by the prescribing information,
FDA documents, or EMA documents. Sources: PI, FDA, EMA documents provided on public databases, details of
the links can be found in the Supplementary Materials. Access date: May 2020.

Next, we conducted a literature search on published data addressing OATP1B1 or OATP1B3

inhibition by different TKIs. In vitro, in vivo, and clinical data were extracted. The details of the
articles were inserted into tables (shown in Supplementary Materials Tables S1–S9) [43–47] For alectinib,
avapritinib, baracitinib, binimetinib, brigatinib, cobimetinib, dacomitinib, encorafenib, erdafitnib,
fedratinib, gilteritinib, ibrutinib, laroctrectinib, lorlatinib, midostaurin, pexidartinib, ponatinib,
trametinib, and zanbrutinib no published reports were found. In data collected for 17 TKIs, the results
of the published data were largely inconsistent in that some of the published results for a given TKI
identified the TKI as an inhibitor of OATP1B1 or OATP1B3, while other sources identified it expressly
as a non-inhibitor. It should be noted that different transfected cell lines (Flp-In T-Rex293, HEK293,
MDCK-II, CHO, SF9, or HepaRG) and different substrates were used in the various studies. The latter
included estradiol-17b-d-glucuronide (E2G), 8-(2-(fluoresceinyl)-aminoethylthio)-adenosine-3 ,5 -cyclic

151
Pharmaceutics 2020, 12, 856

monophosphate (8FcA), fluorescein (FL), 2 ,7 -dichlorofluorescein (DCF), valsartan, atorvastatin, SN-38,
Na-Fluo, fluvastatin, estrone-3-sulfate (E1S) for OATP1B1 and taurocholic acid (TCA), cholecystokinin
octapeptide (CCK-8) for OATP1B3. Furthermore, the preincubation time, the method of detection,
the data analysis metric (percent inhibition or IC50 ), and even the concentration of the TKI were found
to vary among the published reports. The details of these methodological differences are summarized
in Table 2.
Data from clinical and in vivo studies were also collected and reviewed for this article, the results
of which can be seen in the supplements. Very few studies have directly investigated the role of
OATPs in TKI pharmacokinetics with different methodologies, however the results from available
studies seem to be consistent with regulatory data. Since the main scope of this review is to focus on
discrepancies between published data and FDA and EMA guidelines, their results were not further
explored here. Moreover, as OATP1B1 and OATP1B3 substrates used in the retrieved data have complex
pharmacokinetic profiles involving drug-metabolizing enzymes and other transporters, the results
of such case reports should be carefully analyzed to decide on the importance of each part of the
pathway [48–59].

6.1. Omissions
In numerous studies, TKIs have been indicated as victims in DDIs while considerably less is
known about their role as perpetrators via transporter inhibition [32,60–64]. In this context, it is
noteworthy that transporter inhibition studies are not required by regulatory agencies for approval,
but rather recommended to evaluate DDI potential [38].
Currently, there are 20 FDA approved TKIs for which the PI does not contain any information on
their inhibitory effects on OATP1B1 and/or OATP1B3, and this is the case for both agents approved
long ago as well as those that were approved more recently. The transport interactions of some of these
omitted drugs have been examined by academic investigators as reported in the published literature,
and it seems prudent that this information is captured and included in the future in individual PIs
and regulatory databases alike. Interestingly, we found that some of the PIs address DDIs that are
plausibly attributable to OATPs but this is not always consistently acknowledged due to inconclusive
mechanistic insights. For example, dasatinib can dramatically increase plasma levels of the dual
OATP1B1 and CYP3A4 substrate, simvastatin, and the individual contribution of each one of these
pharmacokinetic components to the DDI is not clearly defined. On the other hand, for many TKIs, no
data were found in the published literature on their potential to inhibit OATP1B1 and/or OATP1B3.

6.2. Discrepancies
The discrepancies observed during our evaluation can be categorized into two groups:
discrepancies between the information provided by EMA and FDA, and discrepancies between
different published articles. Table 1 summarizes the cases where data provided by FDA and EMA data
were not congruent in terms of reported OATP-inhibitory properties of TKIs. Specific discrepancies
of interest are highlighted below. Authors do acknowledge that reporting an IC50 , even when it is
relatively low, does not guarantee a significant clinical impact, unless special formulas are implemented,
therefore the inconsistencies reported here, address instances where the guidance is not followed and
the reported IC50 is not further explored:

• The PI and FDA guidance documents for baricitinib report the agent as an OATP1B3 inhibitor,
whereas the EMA documents claim that it is not an inhibitor of this transporter. The existence of
this discrepancy is not explained or discussed in any of the regulatory materials.
• For ceritinib, the PI and EMA state that based on in vitro data, the TKI is unlikely to inhibit
OATP1B1 and OATP1B3 at clinically-relevant concentrations. However, the FDA guidance
document for ceritinib reports that ceritinib inhibits OATP1B1 and OATP1B3 by 31.8% and 24.1%,
respectively, and that because the R-value is <1.25, an in vivo study was considered unnecessary.

152
Pharmaceutics 2020, 12, 856

However, the FDA guidance on DDI potential states that a drug has the potential to inhibit
OATP1B1 or OATP1B3 in vivo if the R-value is >1.1
• The PI for crizotinib reports the TKI as not an inhibitor of OATP1B1 or OATP1B3, but the FDA
guidance reports that crizotinib demonstrated a weak, concentration-dependent inhibitory effect
on pravastatin, an OATP1B1 substrate, and rosuvastatin, an OATP1B3 substrate uptake, with IC50
values of 48 μM and 44 μM, respectively.
• The PI and FDA documents state that OATP1B1 and OATP1B3 are not inhibited by laroctrectinib,
although the EMA materials state that there are inhibitory effects of laroctrectinib on OATP1B1
with an IC50 of 48 μM.
• For lenvatinib, the PI states that there is no potential to inhibit OATP1B1 in vivo, whereas in the
FDA guidance it is concluded that lenvatinib inhibited OATP1B1 with an IC50 of 7.29 μM.
• The PI and FDA report that lorlatinib does not inhibit OATP1B1 and OATP1B3,
while the EMA claims that this TKI has the potential to inhibit these transporters at
clinically-relevant concentrations.
• For osimertinib, the PI and FDA information state that is no observed inhibition of OATP1B1
and OATP1B3, whereas the EMA claims that osimertinib inhibits transport by OATP1B1 and
OATP1B3 albeit at concentrations that are unlikely to result in a clinically-significant DDI.

Some of the potential explanations for these discrepancies are similar to those responsible for
the apparent discrepant data between different published articles and are discussed in more detail
below. However, some interesting points might explain the inconsistencies in regulatory data, such
as the equations used to establish whether a clinical evaluation is indeed necessary for the drug or
not. While EMA suggests calculating (R = 1 + Iu,in,max/Ki or IC50 ) ≥ 1.04, FDA uses a different
equation and different cutoff criteria (R = 1 + Iu,in,max/Ki or IC50 ) ≥ 1.1). This latter equation has been
suggested in the latest FDA draft guidance, although prior versions of this document have proposed
alternative criteria for consideration. It has also been suggested recently that, while most of the
proposed equations and criteria hold merit, they are different in terms of their potential to ultimately
arrive at false positive and false negative predictions. In particular, it has been suggested that the
equation applied in the EMA guidance has a lower positive predictive value than the one proposed in
the current FDA guidance, which offers arguably more dependable predictions [65]. When comparing
data from these two regulatory agencies, this aspect should be taken into consideration, along with
different manners of data reporting (either with or without calculation of the R-value), and variation in
reported IC50 values that could be due to differences in the applied methods.
The results of our comparative literature survey also show that there are instances of substantial
inconsistency between reports in the published literature as well as between published studies and
publicly-available data reported by manufacturers. Since all this collective work is ultimately aimed at
improving clinical decision making, it is pertinent to establish an unequivocal, dependable approach
to data interpretation. The following are some of the elements that can potentially contribute to the
reported inconsistencies:

- Inhibitor concentration: A large number of the published articles have relied on the use of a single
concentration of TKI, although regulatory guidance documents specifically recommend the need
to perform experiments with at least 3–4 different concentrations, in order to more rigorously
evaluate potential inhibitory properties. This is exemplified by a recent study involving the
TKIs afatinib, nintedanib, lenvatinib, and ceritinib in which diverse degrees of inhibition were
observed depending on the concentration (up to 30 μM), and where some concentrations would
even increase transport function [66]. As TKIs tend to get concentrated in the liver and can
potentially increase intracellular levels that are much higher than concurrent levels in plasma [2],
the selection of relevant concentration ranges to be used in in vitro uptake studies requires
careful consideration.

153
Pharmaceutics 2020, 12, 856

- Data reporting: Several studies have only reported results as percent inhibition relative to control,
while more quantitative measures (IC50 or inhibition constant) might be more informative and
offer increased predictive value. According to regulatory guidance documents, certain equations
could be utilized to predict if the observed degree of inhibition has potential clinical relevance.
However, such strategies are rarely implemented and reported studies often fail to include
positive and negative control inhibitors into the experimental design, which is recommended
in the regulatory guidance documents. These issues complicate the interpretation of data and
can result in discrepant views on extrapolating from in vitro studies to the clinical situation,
as reported for ruxolitinib or crizotinib, where experimental data would suggest statistically
significant but not clinically relevant degrees of inhibition [67–69].
- Substrate selection: Since substrate-dependent inhibition by xenobiotics, including TKIs,
has been well documented and is acknowledged expressly in the FDA guidance document,
the degree to which findings obtained with one particular substrate can be extrapolated to
other conditions is uncertain, and potentially accounts for several reported inconsistencies.
Substrate-dependent inhibition has been previously reported when comparing inhibitory
properties in OATP1B1-overexpressed models comparing the substrates fluorescein (FL),
2 ,7 -dichlorofluorescein (DCF), atorvastatin, SN-38, and valsartan, as well as in a recent study
comparing E2G and 8Fc-A [66], where some TKIs such as lapatinib, pazopanib, and nintedanib
show inhibitory effects with some but not all test substrates. The difference between the results
for different substrates is occasionally quite substantial; for example, ceritinib can cause 50%
inhibition of OATP1B1 function when using FL, DCF, atorvastatin, or SN-38 as test substrates, but
causes an apparent increase (by 50%) in OATP1B1-mediated transport of valsartan. Similar results
have been reported for nintedanib, which stimulated the OATP1B1-mediated uptake of FL and
valsartan, while inhibiting that of DCF and SN-38 (by 70%). One strategy recommended by the
FDA to prevent the creation of such apparent, internally conflicting results is to advocate the use
of test substrates in the in vitro model system that is predicted to generate the lowest IC50 value, or
alternatively, to use the most clinically-relevant substrate. While this is a generally useful approach,
several published examples highlight the limitations associated with this strategy. For example,
Koide et al., have demonstrated that the use of DCF as a model substrate generates the lowest
IC50 values for most but not necessarily all substrate-inhibitor combinations [66,68] and that TKIs
with known OATP1B1-inhibitory properties, such as pazopanib, fail to affect transport function
when using the clinically relevant substrates atorvastatin and valsartan [70–73]. The reported
differences in inhibitory properties of TKIs toward the function of transporters such as OATP1B1
as a function of the test substrate used in in vitro studies can directly impact calculated R-values,
and influence the reliability of DDI predictions and the clinical decision-making process, especially
for weak-to-moderate inhibitors [74].
- Incubation conditions: Several studies have demonstrated that the mechanism by which TKIs
inhibit the function of OATP1B1 and/or OATP1B1 can be time-dependent [75], for example in the
case of pazopanib, where preincubation times are inversely correlated with the degree of transport
inhibition such as that longer preincubation times result in lower IC50 estimates [72]. The FDA
guidance recommends the inclusion of a preincubation condition, in addition to simultaneous
incubation of inhibitor and substrate, to ensure that optimal prediction values can be derived
from in vitro experiments. Despite this recommendation, most of the published literature fails to
provide specific detail on the design of the reported experiments where the preincubation condition
is either not considered or not defined. Although the original FDA guidance recommendation was
to include preincubation times of up to 30 min in the experimental study design, recent studies
have demonstrated that more prolonged times, for example, one hour in the case of dasatinib or
even up to three hours for other compounds, may be required to obtained reliable results [45,76].
Proper consideration of this aspect is especially relevant for a class of agents such as TKIs as they
are generally administrated daily for prolonged periods, and may cause transporter inhibition

154
Pharmaceutics 2020, 12, 856

predominantly through an indirect, kinase-mediated mechanism involving post-translational

events that affect tyrosine phosphorylation. This suggests that a comprehensive evaluation of
TKI-transporter inhibition studies require careful consideration and optimization of preincubation
times in order to derive translationally useful DDI predictions.
- Cell line selection: Although regulatory guidance documents do not currently expressly specify
any particular cell-based model system for standardized use in in vitro transporter studies,
prior findings have supported the notion that the choice of cell lines used for transfection can
influence conclusions about inhibitory properties of xenobiotics. Indeed, McFeely et al. have
argued that the selection of cell lines as one of the most important factors contributing to
variability in observed OATP-mediated transport inhibition when using in vitro models [75].
In addition to intrinsic differences between commonly used cell lines that may be linked with
differential baseline expression of other transport mechanisms of putative relevance and artificial
compensatory dysregulation of other transporters in overexpressed models, factors such cell origin
(e.g., mammalian vs amphibian), cell passage number, cell culture conditions, and maintenance
procedures, seeding density, media composition (e.g., presence of binding proteins), and duration
of time that cells are in culture (e.g., expression drifting), which are often not clearly documented,
could further affect the outcome of each study [77].
- Other contributing variables: In addition to the considerations outlined above as well as in Table 2
and Supplementary Materials Tables S1–S9, several other factors can contribute to variation in the
reported transport inhibition data. These include the use of non-standardized software when
calculating kinetic parameters such as IC50 or Ki, and the implementation of varying methods
in quantifying levels of substrate drugs used in the transport assays [77,78]. An example of
the latter would the use of an LC-MS/MS-based method to measure the intracellular levels of
unchanged substrate drugs, whereas more commonly studies would employ the use of fluorescent
substrates of radiolabeled substrates that would be analyzed for total fluorescence or total
radioactivity, respectively, and thus would simultaneously measure the total of the parent drug
and metabolite(s) formed intracellularly. This is an important methodological difference as
certain compounds can undergo rapid enzyme-mediated metabolism once inside cells to form
metabolites that may easily escape detection and result in underestimating the actual extent
of uptake. Furthermore, even the use of identical protocols in different locales can influence
the outcome of particular experimental studies as a result of uncontrollable factors such as
interlaboratory differences, as has been documented extensively before for P-glycoprotein IC50
determinations [77]. It should also be pointed out that inconsistencies, as reported here for
inhibition of OATP1B1- and OATP1B3-mediated transport, are relatively common and have
previously been documented for models involving several other drug-metabolizing enzymes and
transporters with a putative relevance in predicting clinically relevant DDIs [77,79,80].

Table 2. Inconsistencies in reporting OATP1B inhibition by TKIs in published literature.

1B1 1B3 Reported Pre-Incubation
TKI Reported Values Model Substrate References
Inhibitor Inhibitor Values (mins)
>60% inhibition at Flp-In 0.1 mM (3H)
Yes 15 FDA: No [68,70,81]
10 μM T-Rex293/OATP1B1*1A (E2G)
109 ± 5%
function
121 ± 6% function
remaining 300 nM E3S
remaining after HEK293/OATP1B1 or
Bosutinib No No after UNK (1B1) or 2 nM
incubation with 3
incubation CCK-8 (1B3)
10 μM
with
10 μM
>25% 10 μM on E2G
Yes HEK293/OATP1B1 15
E2G, >50% on 8Fc-A 8Fc-A

155
Pharmaceutics 2020, 12, 856

Table 2. Cont.
1B1 1B3 Reported Pre-Incubation
TKI Reported Values Model Substrate References
Inhibitor Inhibitor Values (mins)
OATP1B1:
MDCK-II cell 2 μM; E2G
No >15 μM No >10 μM UNK EMA: No [66,82]
monolayers OATP1B3:
2 μM CCK

Cabozantinib 59% inhibition at

Yes HEK/OATP1B1 3 μM FL
30 μM
61% inhibition at
Yes HEK/OATP1B1 1 μM DCF
30 μM
74% inhibition at
Yes HEK/OATP1B1 1 μM Valsartan
30 μM
50% inhibition at FDA: Yes, PI: No
Yes HEK/OATP1B1 10 3 μM FL
30 μM [66]
50% inhibition at
Yes HEK/OATP1B1 10 1 μM DCF
30 μM
50% inhibition at 0.5 μM
Ceritinib Yes HEK/OATP1B1 10
30 μM atorvastatin
50% inhibition at
Yes HEK/OATP1B1 10 1 μM SN-38
30 μM
150% stimulation at
No HEK/OATP1B1 10 1 μM valsartan
30 μM
11nM (3H]E3S
[1B1)
50nM
(3H)TCA (1B3)
0.5 μM FDA: Yes, PI: No
No No HEK/OATP1B1 or 1B3
Crizotinib fluvastatin [67,68]
(1B1)
2 μM
fluvastatin (;
1B1)
>25% inhibition at E2G
Yes HEK293/OATP1B1 15
10 μM 8Fc-A
0.25 μCi/mL
(3H)ES (for
CHO/OATP-1B1 and OATP-1B1) or
No No UNK NI [68,70,81,83]
-1B3 (3H)CCK-8
(for
OATP-1B3)
Erlotinib >60% decrease at Flp-In 0.1 mM (3H)
Yes 15
10 μM T-Rex293/OATP1B1*1A (E2G)
104 ± 5% function 50%
300 nM E3S
remaining after inhibition HEK293/OATP1B1 or
No Yes UNK (1B1) or 2 nM
incubation with at 3
CCK-8 (1B3)
10 μM 1.19 μM
>25% inhibtion at
E2G
Yes 10 μM on E2G, >50% HEK293/OATP1B1 15
8Fc-A
inhibition on 8Fc-A
>70% decrease with Flp-In 0.1 mM (3H)
Yes 15 NI [67,68,70,81]
10 μM T-Rex293/OATP1B1*1A E2G
50% inhibition at 18.8 ± HEK293/OATP1B1,
Yes Yes fluvastatin
17.2 ± 1.47 μM, 2.74 mM OATP1B3,
EC50
value of HEK293/ OATP1B1,
Inducer (3H)TCA
14.1 ± 4.6 OATP1B3,
Gefitinib mM
78 ± 3%
function
105 ± 3% function
remaining 300 nM E3S
remaining after HEK293/OATP1B1 or
No No after UNK (1B1) or 2 nM
incubation with 3
incubation CCK-8 (1B3)
10 μM
with
10 μM
>25% inhibition at
E2G
Yes 10 μM on E2G, >75% HEK293/OATP1B1 15
8Fc-A
inhibition on 8Fc-A
~20% inhibition at Flp-In 0.1 mM (3H)
Yes 15 NI [47,68,70]
10 μM T-Rex29/OATP1B1 E2G
Imatinib
No Sf9 /OATP1b1 5 1 μM Na-Fluo
>25% inhibition at E2G
Yes HEK293/OATP1B1 15
10 μM on both 8Fc-A

156
Pharmaceutics 2020, 12, 856

Table 2. Cont.
1B1 1B3 Reported Pre-Incubation
TKI Reported Values Model Substrate References
Inhibitor Inhibitor Values (mins)
>70% inhibition at
Yes HEK293/OATP1B1 15 E2G8Fc-A YES [68,70,81,84,85]
10 μM
>70% inhibition at Flp-In T-Rex29/ 0.1 mM (3H)
Yes 15
10 μM OATP1B1 E2G
Yes,
CHO/ OATP-1B1 or
No slight UNK fluro-methotrexate
-1B3
inhibition
Lapatinib 50% inhibition at
Yes CHO-OATP1B1 15–30 (3H) E2G
4.0 μM (Sd:2.1)
98 ± 16%
function
123 ± 13% function
remaining 300 nM E3S
remaining after HEK293/OATP1B1 or
No No after UNK (1B1) or 2 nM
incubation with 3
incubation CCK-8 (1B3)
10 μM
with
10 μM
No HEK/OATP1B1 10 3 μM FL EMA:No [66,81]
No HEK/OATP1B1 10 1 μM DCF
0.5 μM
No HEK/OATP1B1 10
atorvastatin
30% inhibition at
Neratinib Yes HEK/OATP1B1 10 1 μM SN-38
30 μM
No HEK/OATP1B1 10 1 μM valsartan
123 ± 13% function 50%
300 nM E3S
remaining after inhibition HEK293/OATP1B1 or
No Yes UNK (1B1) or 2nM
incubation with at 18.13 3
CCK-8 (1B3)
10 μM ± 1.21
NI
No HEK/OATP1B1 10 3 μM FL
[66,68,70,83,86,87]
~50% inhibition at
Yes HEK/OATP1B1 10 1 μM DCF
30 μM
~50% inhibition at 0.5 μM
Yes HEK/OATP1B1 10
30 μM atorvastatin
~50% inhibition at
Yes HEK/OATP1B1 10 1 μM SN-38
30 μM
~50% inhibition at
Yes HEK/OATP1B1 10 1 μM valsartan
30 μM
>95% inhibition at Flp-In T-Rex29/ 0.1 mM (3H)
Yes 10
10 μM OATP1B1 E2G
Nilotinib 100 ± 3%
function
remaining 300 nM E3S
110 ± 7% HEK293/OATP1B1 or
No No after UNK (1B1) or 2nM
stimulation at 10 μM 3
incubation CCK-8 (1B3)
with
10 μM
5–40 μM 8Fc-A
>80% inhibition at
Yes HEK/OATP1B1 or
0–20 μM
2 μM E2G
50% inhibition at HEK293/OATP1B1 or E2G or
Yes Yes
1.3 μM 3 8FcA
>50% at 10 μM, IC50 : E2G or
Yes HEK293/OATP1B1 15
~1 μM 8FcA
0.25 μCi/mL
(3H)ES (for
50% inhibition at CHO/ OATP-1B1 or OATP-1B1) or
Yes No
2.78 ± 1.13 μM -1B3 (3H)CCK-8
(for
OATP-1B3)
312% stimulation at
No HEK/OATP1B1 3 μM FL No [66]
30 μM
74% inhibition at
Yes HEK/OATP1B1 1 μM DCF
Nintedanib 30 μM
133% stimulation at
No HEK/OATP1B1 1 μM Valsartan
30 μM
78% inhibition at
Yes HEK/OATP1B1 1 μM SN-38
30 μM

157
Pharmaceutics 2020, 12, 856

Table 2. Cont.
1B1 1B3 Reported Pre-Incubation
TKI Reported Values Model Substrate References
Inhibitor Inhibitor Values (mins)
120% stimulation at
No HEK/OATP1B1 10 3 μM FL Yes [66,68,70–73,83]
30 μM
No HEK/OATP1B1 10 1 μM DCF
0.5 μM
No HEK/OATP1B1 10
atorvastatin
No HEK/OATP1B1 10 1 μM SN-38
No HEK/OATP1B1 10 1 μM valsartan
0.25 μCi/mL of
(3H)ES (for
50% inhibition 3.89 CHO/ OATP-1B1 or OATP-1B1) or
Yes No
± 1.21 μM -1B3 (3H)CCK-8
(for
OATP-1B3)
>50% inhibition E2G or
Yes with 8Fc-A, >90% HEK293/OATP1B1 15 (1B1),8FcA
inhibition with E2G (1B1, 1B3)
Pazopanib
50% inhibition at
Yes CHO-OATP1B1 15–30 (3H)-EG
0.79 μM
No HEK293/OATP1B1 SN-38
>95% inhibition at Flp-In 0.1 μM (3H)
Yes 15
10 μM T-Rex29/OATP1B1 E2G
(3H) E1S and
IC50 E1S: 1.42 ± 0.23,
Yes HEK293/OATP1B1 0 (3H)
IC50 E2G: 13.5 ± 6.0
E2G
IC50 E1S:0.594 ± (3H) (E1S) and
Yes 0.030 IC50 E2G: 7.25 HEK293/OATP1B2 1 (3H)
± 0.53 E2G
IC50 E1S: 0.374 ± (3H) E1S and
Yes 0.074, IC50 E2G: 2.58 HEK293/OATP1B4 30 (3H)
± 0.77 E2G
IC50 E1S: 0.530 ± (3H) E1S and
Yes 0.022, IC50 E2G:2.03 HEK293/OATP1B5 60 (3H)
± 0.71 E2G
0.5 μM FDA: No
No 30% stimulation HEK293/OATP1B6 10
Atorvastatin [66,68,70,88]
estrone-3-sulfate
50% inhibition at
Regorafenib Yes No HEK293/OATP1B1/1B3 2 (1B1)/taurocholic
~10 μM
acid (1B3)
>50% inhibition at Flp-In T-Rex29/ 0.1mM (3H)
Yes 15
10 μM OATP1B1 E2G
Yes >50% inhibition HEK293/OATP1B1 15 E2G, 8FcA
>25% inhibition at
Yes HEK293/OATP1B1 15 E2G, 8FcA No [68–70]
10 μM on 8Fc-A

Ruxolitinib No HepaRG 4 nM E3S

~20% inhibition at Flp-In T-Rex29/ 0.1mM (3H)
Yes 15
10 μM OATP1B1 E2G
>75% at 10 μM on
Yes HEK293/OATP1B1 15 E2G, 8FcA NI [66,70,81]
both
>90% inhibition at Flp-In (3H) E2G 0.1
Yes 15
10 μM T-Rex293/OATP1B1*1A mM
50% inhibition at Flp-In T-Rex293/ 0.1 mM (3H)
Yes 15
69.6 μM OATP1B1*1A docetaxel

Sorafenib No HEK/OATP1B1 10 3 μM FL
No HEK/OATP1B1 10 1 μM DCF
0.5 μM
No HEK/OATP1B1 10
atorvastatin
No HEK/OATP1B1 10 1 μM SN-38
No HEK/OATP1B1 10 1 μM valsartan
68 ± 0.5%
function
96 ± 7% function
remaining 300nM E3S
remaining after HEK293/OATP1B1 or
No Yes after UNK (1B1) or 2nM
incubation with 3
incubation CCK-8 (1B3)
10 μM
with
10 μM

158
Pharmaceutics 2020, 12, 856

Table 2. Cont.
1B1 1B3 Reported Pre-Incubation
TKI Reported Values Model Substrate References
Inhibitor Inhibitor Values (mins)
>25% decrease at Flp-In 0.1 mM (3H)
Yes 15 NI [68,70,84]
10 μM T-Rex293/OATP1B1*1A E2G
101 ±
10%
Sunitinib 109 ± 10% function function
300nM E3S
remaining after remaining HEK293/OATP1B1
No No UNK (1B1) or 2nM
incubation with after or 3
CCK-8 (1B3)
10 μM incubation
with
10 μM
Yes >25% inhibition HEK293/OATP1B1 15 E2G, 8FcA
>25% inhibition at Flp-In T-Rex293/ 0.1 mM (3H)
Yes 15 NI [53,68,70,83,84]
10 μM OATP1B1*1A E2G
0.25 μCi/mL of
Vandetanib 50% (3H)ES (for
inhibition CHO/ OATP-1B1 or OATP-1B1)
No Yes
at 18.13 -1B3 or (3H)CCK-8
± 1.21 (for
OATP-1B3)
71± 5%
function
110 ± 6% function
remaining 300nM E3S
remaining after HEK293/OATP1B1
No Yes after UNK (1B1) or 2nM
incubation with or 3
incubation CCK-8 (1B3)
10 μM
with
10 μM
>25% inhibition at
Yes HEK293/OATP1B1 15 E2G, 8FcA
10 μM

UNK indicates not mentioned in the study/Unknown. NI is not indicated

7. Conclusions
The development and use of TKIs as molecular targeted therapies for the treatment of a diverse
array of malignant diseases continues to rapidly increase, and 50 of such agents have now been
approved for human use. However, polypharmacy regimens commonly applied in oncology with these
TKIs creates a high risk for the occurrence of clinically-relevant DDIs. Although the extent to which
such DDIs are influenced by the ability of many TKIs to impact the function of transporter-mediated
uptake mechanisms in hepatocytes remains relatively poorly studied, data have accumulated in recent
years highlighting that TKIs can act as perpetrators in DDIs by inhibiting OATP1B1 and/or OATP1B3.
Many of these recent observations have been made with the use of transfected cell-based in vitro
models, and a summary of this available evidence has identified substantial methodological differences
between various studies and has highlighted several important limitations in the chosen approaches
that have generated incongruent reports. Given that these in vitro studies are the most frequently
employed nonclinical tool in aiding decision making for patient care, it is pertinent that regulatory
guidance documents and available published literature provide consistent and corresponding results.
To further improve consistency in the outcome of transporter-mediated DDI studies involving TKIs,
specific recommendations are offered that may assist investigators in the design of future studies in
order to provide unequivocal data pertaining to the inhibitory potential of both established as well as
investigational TKIs that could be rationally used to further refine the predictive ability of DDIs and
ultimately optimize the outcome of treatment in patients with cancer.

Supplementary Materials: The following are available online at http://www.mdpi.com/1999-4923/12/9/856/s1,

Table S1: Prescribing Information, Table S2. FDA guidance for TKI interactions with OATP1B1 and OATP1B3,
Table S3. EMA guidance for TKI interactions with OATP1B1 and OATP1B3, Table S4. Comparison of PI, FDA,
and EMA regulatory documents, Table S5. In vitro studies for TKIs as perpetrators in OATP-mediated DDIs,
Table S6. In vivo studies for TKIs as perpetrators in OATP-mediated DDIs Table S7. Literature on clinical studies
focused on TKIs in potential DDIs Table S8. Comparison of the PI, FDA, and EMA guidance documents to the
literature for OATP1B1 inhibition and OATP1B3 inhibition, Table S9. TKIs for which no relevant literature was
found for OATP-mediated inhibition by TKIs.
Author Contributions: Each of the authors have read and approved this version of the manuscript and agrees to
be held accountable for the accuracy and integrity of all included content. Each named author has substantially

159
Pharmaceutics 2020, 12, 856

contributed to conducting the underlying research and drafting this manuscript. Individual contributions are
illustrated in the following statements: Conceptualization, D.A.G., Z.T., A.S. and S.D.B.; methodology, D.A.G., Z.T.
and E.D.E.; investigation, D.A.G., Z.T. and E.D.E.; interpretation of data, D.A.G. and Z.T.; writing—Original draft
preparation, D.A.G. and Z.T.; writing—Review and editing, D.A.G., Z.T., E.D.E., A.S. and S.D.B.; supervision, A.S.
and S.D.B.; funding acquisition, E.D.E., A.S. and S.D.B. All authors have read and agreed to the published version
of the manuscript.
Funding: The work was supported in part by the National Institutes of Health (Grants R01CA215802 (to A.S.) and
R01CA138744 (to S.D.B.)), by the OSU Comprehensive Cancer Center Pelotonia foundation (A.S. and S.D.B.) and
by the Pelotonia Fellowship Program (E.D.E.). The content is solely the responsibility of the authors and does not
necessarily represent the official views of the funding agencies.
Conflicts of Interest: The authors declare no conflict of interest

References
1. Watanabe, J.H.; McInnis, T.; Hirsch, J.D. Cost of Prescription Drug–Related Morbidity and Mortality.
Ann. Pharmacother. 2018, 52, 829–837. [CrossRef] [PubMed]
2. Fuhr, U.; Hsin, C.-H.; Li, X.; Jabrane, W.; Sörgel, F. Assessment of Pharmacokinetic Drug–Drug Interactions
in Humans: In Vivo Probe Substrates for Drug Metabolism and Drug Transport Revisited. Annu. Rev.
Pharmacol. Toxicol. 2019, 59, 507–536. [CrossRef]
3. Magro, L.; Moretti, U.; Leone, R. Epidemiology and characteristics of adverse drug reactions caused by
drug–drug interactions. Expert Opin. Drug Saf. 2011, 11, 83–94. [CrossRef]
4. Subramanian, A.; Adhimoolam, M.; Kannan, S. Study of drug–Drug interactions among the hypertensive
patients in a tertiary care teaching hospital. Perspect. Clin. Res. 2018, 9, 9–14. [CrossRef] [PubMed]
5. Backman, J.T.; Kivistö, K.T.; Olkkola, K.T.; Neuvonen, P.J. The area under the plasma concentration-time
curve for oral midazolam is 400-fold larger during treatment with itraconazole than with rifampicin. Eur. J.
Clin. Pharmacol. 1998, 54, 53–58. [CrossRef] [PubMed]
6. de Jong, J.; Skee, D.; Murphy, J.; Sukbuntherng, J.; Hellemans, P.; Smit, J.; de Vries, R.; Jiao, J.J.; Snoeys, J.;
Mannaert, E. Effect of CYP3A perpetrators on ibrutinib exposure in healthy participants. Pharmacol. Res.
Perspect. 2015, 3, e00156. [CrossRef]
7. Gessner, A.; König, J.; Fromm, M.F. Clinical Aspects of Transporter-Mediated Drug–Drug Interactions. Clin.
Pharmacol. Ther. 2019, 105, 1386–1394. [CrossRef]
8. Riechelmann, R.P.; del Giglio, A. Drug interactions in oncology: How common are they? Ann. Oncol. 2009,
20, 1907–1912. [CrossRef]
9. Riechelmann, R.; Girardi, D. Drug interactions in cancer patients: A hidden risk? J. Res. Pharm. Pr. 2016,
5, 77–78. [CrossRef]
10. Solomon, J.M.; Ajewole, V.B.; Schneider, A.M.; Sharma, M.; Bernicker, E.H. Evaluation of the prescribing
patterns, adverse effects, and drug interactions of oral chemotherapy agents in an outpatient cancer center. J.
Oncol. Pharm. Pract. 2019, 25, 1564–1569. [CrossRef]
11. Howlader, N.; Mariotto, A.B.; Besson, C.; Suneja, G.; Robien, K.; Younes, N.; Engels, E.A. Cancer-specific
mortality, cure fraction, and noncancer causes of death among diffuse large B-cell lymphoma patients in the
immunochemotherapy era. Cancer 2017, 123, 3326–3334. [CrossRef] [PubMed]
12. Riechelmann, R.P.; Zimmermann, C.; Chin, S.N.; Wang, L.; O’Carroll, A.; Zarinehbaf, S.; Krzyzanowska, M.K.
Potential Drug Interactions in Cancer Patients Receiving Supportive Care Exclusively. J. Pain Symptom
Manag. 2008, 35, 535–543. [CrossRef] [PubMed]
13. van Leeuwen, R.; Brundel, D.H.S.; Neef, C.; van Gelder, T.; Mathijssen, R.H.J.; Burger, D.M.; Jansman, F.G.A.
Prevalence of potential drug–drug interactions in cancer patients treated with oral anticancer drugs. Br. J.
Cancer 2013, 108, 1071–1078. [CrossRef] [PubMed]
14. Bartel, S. Safe practices and financial considerations in using oral chemotherapeutic agents. Am. J. Health
Pharm. 2007, 64, S8–S14. [CrossRef]
15. Dagher, R.; Cohen, M.; Williams, G.; Rothmann, M.; Gobburu, J.; Robbie, G.; Rahman, A.; Chen, G.; Staten, A.;
Griebel, D.; et al. Approval summary: Imatinib mesylate in the treatment of metastatic and/or unresectable
malignant gastrointestinal stromal tumors. Clin. Cancer Res. 2002, 8, 3034–3038. [PubMed]
16. Roskoski, R. Properties of FDA-approved small molecule protein kinase inhibitors. Pharmacol. Res. 2019,
144, 19–50. [CrossRef]

160
Pharmaceutics 2020, 12, 856

17. Lemmon, M.A.; Schlessinger, J. Cell Signaling by Receptor Tyrosine Kinases. Cell 2010, 141, 1117–1134.
[CrossRef]
18. Hubbard, S.R.; Till, J.H. Protein Tyrosine Kinase Structure and Function. Annu. Rev. Biochem. 2000,
69, 373–398. [CrossRef]
19. Robinson, D.R.; Wu, Y.-M.; Lin, S.-F. The protein tyrosine kinase family of the human genome. Oncogene
2000, 19, 5548–5557. [CrossRef]
20. Arora, A.; Scholar, E.M. Role of Tyrosine Kinase Inhibitors in Cancer Therapy. J. Pharmacol. Exp. Ther. 2005,
315, 971–979. [CrossRef]
21. Shawver, L.K.; Slamon, D.; Ullrich, A. Smart drugs: Tyrosine kinase inhibitors in cancer therapy. Cancer Cell
2002, 1, 117–123. [CrossRef]
22. Jeong, W.; Doroshow, J.H.; Kummar, S. United States Food and Drug Administration approved oral kinase
inhibitors for the treatment of malignancies. Curr. Probl. Cancer 2013, 37, 110–144. [CrossRef] [PubMed]
23. Herviou, P.; Thivat, E.; Richard, D.; Roche, L.; Dohou, J.; Pouget, M.; Eschalier, A.; Durando, X.; Authier, N.
Therapeutic drug monitoring and tyrosine kinase inhibitors. Oncol. Lett. 2016, 12, 1223–1232. [CrossRef]
[PubMed]
24. Haouala, A.; Widmer, N.; Duchosal, M.A.; Montemurro, M.; Buclin, T.; Decosterd, L.A. Drug interactions
with the tyrosine kinase inhibitors imatinib, dasatinib, and nilotinib. Blood 2011, 117, e75–e87. [CrossRef]
[PubMed]
25. Iurlo, A.; Nobili, A.; Latagliata, R.; Bucelli, C.; Castagnetti, F.; Breccia, M.; Abruzzese, E.; Cattaneo, D.;
Fava, C.; Ferrero, D.; et al. Imatinib and polypharmacy in very old patients with chronic myeloid leukemia:
Effects on response rate, toxicity and outcome. Oncotarget 2016, 7, 80083–80090. [CrossRef]
26. Hussaarts, K.; Veerman, G.D.M.; Jansman, F.G.A.; van Gelder, T.; Mathijssen, R.H.J.; van Leeuwen, R.W.F.
Clinically relevant drug interactions with multikinase inhibitors: A review. Ther. Adv. Med Oncol. 2019, 11,
1758835918818347. [CrossRef]
27. Fowler, H.; Belot, A.; Ellis, L.; Maringe, C.; Luque-Fernandez, M.A.; Njagi, E.N.; Navani, N.; Sarfati, D.;
Rachet, B. Comorbidity prevalence among cancer patients: A population-based cohort study of four cancers.
BMC Cancer 2020, 20, 2–15. [CrossRef]
28. Ergun, Y.; Ozdemir, N.Y.; Toptas, S.; Kurtipek, A.; Eren, T.; Yazici, O.; Sendur, M.A.N.; Akinci, B.; Ucar, G.;
Oksuzoglu, B.; et al. Drug-drug interactions in patients using tyrosine kinase inhibitors: A multicenter
retrospective study. J. Buon 2019, 24, 1719–1726.
29. Keller, K.L.; Franquiz, M.J.; Duffy, A.P.; Trovato, J.A. Drug–drug interactions in patients receiving tyrosine
kinase inhibitors. J. Oncol. Pharm. Pract. 2016, 24, 110–115. [CrossRef]
30. da Silva, C.; Honeywell, R.J.; Dekker, H.; Peters, G.J. Physicochemical properties of novel protein kinase
inhibitors in relation to their substrate specificity for drug transporters. Expert Opin. Drug Metab. Toxicol.
2015, 11, 703–717. [CrossRef]
31. Herbrink, M.; Nuijen, B.; Schellens, J.H.; Beijnen, J.H. Variability in bioavailability of small molecular tyrosine
kinase inhibitors. Cancer Treat. Rev. 2015, 41, 412–422. [CrossRef] [PubMed]
32. van Leeuwen, R.; van Gelder, T.; Mathijssen, R.; Jansman, F.G.A. Drug–drug interactions with tyrosine-kinase
inhibitors: A clinical perspective. Lancet Oncol. 2014, 15, e315–e326. [CrossRef]
33. Schulte, R.R.; Ho, R.H. Organic Anion Transporting Polypeptides: Emerging Roles in Cancer Pharmacology.
Mol. Pharmacol. 2019, 95, 490–506. [CrossRef]
34. Ho, R.H.; Kim, R.B. Transporters and drug therapy: Implications for drug disposition and disease. Clin.
Pharmacol. Ther. 2005, 78, 260–277. [CrossRef] [PubMed]
35. Kalliokoski, A.; Niemi, M. Impact of OATP transporters on pharmacokinetics. Br. J. Pharmacol. 2009,
158, 693–705. [CrossRef]
36. Shitara, Y. Clinical Importance of OATP1B1 and OATP1B3 in DrugDrug Interactions. Drug Metab.
Pharmacokinet. 2011, 26, 220–227. [CrossRef]
37. Zimmerman, E.I.; Hu, S.; Roberts, J.L.; Gibson, A.A.; Orwick, S.J.; Li, L.; Sparreboom, A.; Baker, S. Contribution
of OATP1B1 and OATP1B3 to the disposition of sorafenib and sorafenib-glucuronide. Clin. Cancer Res. 2013,
19, 1458–1466. [CrossRef]
38. Teo, Y.L.; Ho, H.K.; Chan, A. Risk of tyrosine kinase inhibitors-induced hepatotoxicity in cancer patients: A
meta-analysis. Cancer Treat. Rev. 2013, 39, 199–206. [CrossRef]

161
Pharmaceutics 2020, 12, 856

39. Kellick, K. Organic Ion Transporters and Statin Drug Interactions. Curr. Atheroscler. Rep. 2017, 19, 65.
[CrossRef]
40. Alam, K.; Crowe, A.; Wang, X.; Zhang, P.; Ding, K.; Li, L.; Yue, W. Regulation of Organic Anion Transporting
Polypeptides (OATP) 1B1- and OATP1B3-Mediated Transport: An Updated Review in the Context of
OATP-Mediated Drug-Drug Interactions. Int. J. Mol. Sci. 2018, 19, 855. [CrossRef]
41. Lancaster, C.S.; Sprowl, J.A.; Walker, A.L.; Hu, S.; Gibson, A.A.; Sparreboom, A. Modulation of OATP1B-type
transporter function alters cellular uptake and disposition of platinum chemotherapeutics. Mol. Cancer Ther.
2013, 12, 1537–1544. [CrossRef] [PubMed]
42. Chen, M.; Hu, S.; Li, Y.; Gibson, A.A.; Fu, Q.; Baker, S.D.; Sparreboom, A. Role of Oatp2b1 in Drug Absorption
and Drug-Drug Interactions. Drug Metab. Dispos. 2020, 48, 420–426. [CrossRef] [PubMed]
43. Podoll, T.; Pearson, P.G.; Evarts, J.; Ingallinera, T.; Sun, H.; Byard, S.; Fretland, A.J.; Slatter, J.G. Abstract 13:
Structure elucidation, metabolism, and drug interaction potential of ACP-5862, an active, major, circulating
metabolite of acalabrutinib. Cancer Res. 2019, 79. [CrossRef]
44. Ellens, H.; Johnson, M.; Lawrence, S.K.; Chen, L.; Richards-Peterson, L.E.; Watson, C. Prediction of the
Transporter-Mediated Drug-Drug Interaction Potential of Dabrafenib and Its Major Circulating Metabolites.
Drug Metab. Dispos. 2017, 45, 646–656. [CrossRef] [PubMed]
45. Pahwa, S.; Alam, K.; Crowe, A.; Farasyn, T.; Neuhoff, S.; Hatley, O.; Ding, K.; Yue, W. Pretreatment With
Rifampicin and Tyrosine Kinase Inhibitor Dasatinib Potentiates the Inhibitory Effects Toward OATP1B1- and
OATP1B3-Mediated Transport. J. Pharm. Sci. 2017, 106, 2123–2135. [CrossRef] [PubMed]
46. Elsby, R.; Martin, P.; Surry, D.; Sharma, P.; Fenner, K. Solitary Inhibition of the Breast Cancer Resistance
Protein Efflux Transporter Results in a Clinically Significant Drug-Drug Interaction with Rosuvastatin by
Causing up to a 2-Fold Increase in Statin Exposure. Drug Metab. Dispos. 2015, 44, 398–408. [CrossRef]
47. Patik, I.; Kovacsics, D.; Német, O.; Gera, M.; Várady, G.; Stieger, B.; Hagenbuch, B.; Szakács, G.;
Özvegy-Laczka, C. Functional expression of the 11 human Organic Anion Transporting Polypeptides
in insect cells reveals that sodium fluorescein is a general OATP substrate. Biochem. Pharmacol. 2015,
98, 649–658. [CrossRef]
48. Bergman, E.; Hedeland, M.; Bondesson, U.; Lennernäs, H. The effect of acute administration of rifampicin
and imatinib on the enterohepatic transport of rosuvastatinin vivo. Xenobiotica 2010, 40, 558–568. [CrossRef]
49. Nakamura, Y.; Hirokawa, Y.; Kitamura, S.; Yamasaki, W.; Arihiro, K.; Tatsugami, F.; Iida, M.; Kakizawa, H.;
Date, S.; Awai, K. Effect of lapatinib on hepatic parenchymal enhancement on gadoxetate disodium
(EOB)-enhanced MRI scans of the rat liver. Jpn. J. Radiol. 2013, 31, 386–392. [CrossRef]
50. Martin, P.D.; Gillen, M.; Ritter, J.; Mathews, D.; Brealey, C.; Surry, D.; Oliver, S.; Holmes, V.; Severin, P.;
Elsby, R. Effects of Fostamatinib on the Pharmacokinetics of Oral Contraceptive, Warfarin, and the Statins
Rosuvastatin and Simvastatin: Results From Phase I Clinical Studies. Drugs R&D 2016, 16, 93–107. [CrossRef]
51. Harvey, R.D.; Aransay, N.R.; Isambert, N.; Lee, J.-S.; Arkenau, T.; Vansteenkiste, J.; Dickinson, P.A.; Bui, K.;
Weilert, D.; So, K.; et al. Effect of multiple-dose osimertinib on the pharmacokinetics of simvastatin and
rosuvastatin. Br. J. Clin. Pharmacol. 2018, 84, 2877–2888. [CrossRef]
52. Vishwanathan, K.; Cantarini, M.; So, K.; Masson, E.; Fetterolf, J.; Ramalingam, S.S.; Harvey, R.D. Impact
of Disease and Treatment Response in Drug–Drug Interaction Studies: Osimertinib and Simvastatin in
Advanced Non-Small Cell Lung Cancer. Clin. Transl. Sci. 2019, 13, 41–46. [CrossRef] [PubMed]
53. Calvo, E.; Lee, J.-S.; Kim, S.-W.; Moreno, V.; Carpeno, J.D.; Weilert, D.; Laus, G.; Mann, H.; Vishwanathan, K.
Modulation of Fexofenadine Pharmacokinetics by Osimertinib in Patients With Advanced EGFR-Mutated
Non–Small Cell Lung Cancer. J. Clin. Pharmacol. 2019, 59, 1099–1109. [CrossRef] [PubMed]
54. Reddy, V.P.; Walker, M.; Sharma, P.; Ballard, P.; Vishwanathan, K. Development, Verification, and Prediction
of Osimertinib Drug-Drug Interactions Using PBPK Modeling Approach to Inform Drug Label. CPT
Pharmacomet. Syst. Pharmacol. 2018, 7, 321–330. [CrossRef] [PubMed]
55. Komatsu, H.; Enomoto, M.; Shiraishi, H.; Morita, Y.; Hashimoto, D.; Nakayama, S.; Funakoshi, S.; Hirano, S.;
Terada, Y.; Miyamura, M.; et al. Severe hypoglycemia caused by a small dose of repaglinide and concurrent
use of nilotinib and febuxostat in a patient with type 2 diabetes. Diabetol. Int. 2020, 1–5. [CrossRef]
56. Kendra, K.; Plummer, R.; Salgia, R.; O’Brien, M.E.R.; Paul, E.M.; Suttle, A.B.; Compton, N.; Xu, C.-F.;
Ottesen, L.H.; Villalona-Calero, M.A. A Multicenter Phase I Study of Pazopanib in Combination with
Paclitaxel in First-Line Treatment of Patients with Advanced Solid Tumors. Mol. Cancer Ther. 2014,
14, 461–469. [CrossRef]

162
Pharmaceutics 2020, 12, 856

57. Poje, D.K.; Božina, N.; Šimičević, L.; Žabić, I. Severe hyperglycaemia following pazopanib treatment: The
role of drug-drug-gene interactions in a patient with metastatic renal cell carcinoma—A case report. J. Clin.
Pharm. Ther. 2020, 45, 628–631. [CrossRef]
58. Hamberg, P.; Mathijssen, R.H.J.; de Bruijn, P.; Leonowens, C.; van der Biessen, D.; Eskens, F.A.L.M.; Sleijfer, S.;
Verweij, J.; de Jonge, M.J.A. Impact of pazopanib on docetaxel exposure: Results of a phase I combination
study with two different docetaxel schedules. Cancer Chemother. Pharmacol. 2014, 75, 365–371. [CrossRef]
59. Xu, C.-F.; Xue, Z.; Bing, N.; King, K.S.; McCann, L.A.; de Souza, P.L.; Goodman, V.L.; Spraggs, C.F.;
Mooser, V.E.; Pandite, L.N. Concomitant use of pazopanib and simvastatin increases the risk of transaminase
elevations in patients with cancer. Ann. Oncol. 2012, 23, 2470–2471. [CrossRef]
60. Mandery, K.; Glaeser, H.; Fromm, M.F. Interaction of innovative small molecule drugs used for cancer
therapy with drug transporters. Br. J. Pharmacol. 2011, 165, 345–362. [CrossRef]
61. Lawrence, S.K.; Nguyen, D.; Bowen, C.; Richards-Peterson, L.; Skordos, K.W. The Metabolic Drug-Drug
Interaction Profile of Dabrafenib: In Vitro Investigations and Quantitative Extrapolation of the P450-Mediated
DDI Risk. Drug Metab. Dispos. 2014, 42, 1180–1190. [CrossRef] [PubMed]
62. Filppula, A.; Neuvonen, P.J.; Backman, J.T. In Vitro Assessment of Time-Dependent Inhibitory Effects on
CYP2C8 and CYP3A Activity by Fourteen Protein Kinase Inhibitors. Drug Metab. Dispos. 2014, 42, 1202–1209.
[CrossRef]
63. Grenader, T.; Gipps, M.; Shavit, L.; Gabizon, A. Significant drug interaction: Phenytoin toxicity due to
erlotinib. Lung Cancer 2007, 57, 404–406. [CrossRef] [PubMed]
64. Kuhn, E.L.; Lévêque, D.; Lioure, B.; Gourieux, B.; Bilbault, P. Adverse event potentially due to an interaction
between ibrutinib and verapamil: A case report. J. Clin. Pharm. Ther. 2016, 41, 104–105. [CrossRef]
65. Vaidyanathan, J.; Yoshida, K.; Arya, V.; Zhang, L. Comparing Various In Vitro Prediction Criteria to Assess
the Potential of a New Molecular Entity to Inhibit Organic Anion Transporting Polypeptide 1B1. J. Clin.
Pharmacol. 2016, 56, S59–S72. [CrossRef] [PubMed]
66. Koide, H.; Tsujimoto, M.; Takeuchi, A.; Tanaka, M.; Ikegami, Y.; Tagami, M.; Abe, S.; Hashimoto, M.;
Minegaki, T.; Nishiguchi, K. Substrate-dependent effects of molecular-targeted anticancer agents on activity
of organic anion transporting polypeptide 1B1. Xenobiotica 2017, 48, 1059–1071. [CrossRef]
67. Sato, T.; Ito, H.; Hirata, A.; Abe, T.; Mano, N.; Yamaguchi, H. Interactions of crizotinib and gefitinib with
organic anion-transporting polypeptides (OATP)1B1, OATP1B3 and OATP2B1: Gefitinib shows contradictory
interaction with OATP1B3. Xenobiotica 2017, 48, 73–78. [CrossRef]
68. Leblanc, A.F.; Sprowl, J.A.; Alberti, P.; Chiorazzi, A.; Arnold, W.D.; Gibson, A.A.; Hong, K.W.; Pioso, M.S.;
Chen, M.; Huang, K.M.; et al. OATP1B2 deficiency protects against paclitaxel-induced neurotoxicity. J. Clin.
Investig. 2018, 128, 816–825. [CrossRef]
69. Febvre-James, M.; Bruyère, A.; Le Vée, M.; Fardel, O. The JAK1/2 Inhibitor Ruxolitinib Reverses
Interleukin-6-Mediated Suppression of Drug-Detoxifying Proteins in Cultured Human Hepatocytes. Drug
Metab. Dispos. 2018, 46, 131–140. [CrossRef]
70. Hu, S.; Mathijssen, R.H.J.; de Bruijn, P.; Baker, S.; Sparreboom, A. Inhibition of OATP1B1 by tyrosine kinase
inhibitors: In vitro–in vivo correlations. Br. J. Cancer 2014, 110, 894–898. [CrossRef]
71. Iwase, M.; Fujita, K.-I.; Nishimura, Y.; Seba, N.; Masuo, Y.; Ishida, H.; Kato, Y.; Kiuchi, Y. Pazopanib interacts
with irinotecan by inhibiting UGT1A1-mediated glucuronidation, but not OATP1B1-mediated hepatic uptake,
of an active metabolite SN-38. Cancer Chemother. Pharmacol. 2019, 83, 993–998. [CrossRef] [PubMed]
72. Taguchi, T.; Masuo, Y.; Sakai, Y.; Kato, Y. Short-lasting inhibition of hepatic uptake transporter OATP1B1 by
tyrosine kinase inhibitor pazopanib. Drug Metab. Pharmacokinet. 2019, 34, 372–379. [CrossRef] [PubMed]
73. Xu, C.F.; Reck, B.H.; Xue, Z.; Huang, L.; Baker, K.L.; Chen, M.; Chen, E.P.; Ellens, H.E.; Mooser, V.E.;
Cardon, L.R.; et al. Pazopanib-induced hyperbilirubinemia is associated with Gilbert’s syndrome UGT1A1
polymorphism. Br. J. Cancer 2010, 102, 1371–1377. [CrossRef] [PubMed]
74. Izumi, S.; Nozaki, Y.; Maeda, K.; Komori, T.; Takenaka, O.; Kusuhara, H.; Sugiyama, Y. Investigation of the
Impact of Substrate Selection on In Vitro Organic Anion Transporting Polypeptide 1B1 Inhibition Profiles for
the Prediction of Drug-Drug Interactions. Drug Metab. Dispos. 2014, 43, 235–247. [CrossRef]
75. McFeely, S.J.; Ritchie, T.K.; Ragueneau-Majlessi, I. Variability in In Vitro OATP1B1/1B3 Inhibition Data: Impact of
Incubation Conditions on Variability and Subsequent Drug Interaction Predictions. Clin. Transl. Sci. 2020, 13, 47–52.
[CrossRef]

163
Pharmaceutics 2020, 12, 856

76. Tátrai, P.; Schweigler, P.; Poller, B.; Domange, N.; de Wilde, R.; Hanna, I.; Gaborik, Z.; Huth, F. A Systematic
In Vitro Investigation of the Inhibitor Preincubation Effect on Multiple Classes of Clinically Relevant
Transporters. Drug Metab. Dispos. 2019, 47, 768–778. [CrossRef]
77. Bentz, J.; O’Connor, M.P.; Bednarczyk, D.; Coleman, J.; Lee, C.; Palm, J.; Pak, Y.A.; Perloff, E.S.; Reyner, E.;
Balimane, P.; et al. Variability in P-Glycoprotein Inhibitory Potency (IC50) Using Various in Vitro Experimental
Systems: Implications for Universal Digoxin Drug-Drug Interaction Risk Assessment Decision Criteria. Drug
Metab. Dispos. 2013, 41, 1347–1366. [CrossRef]
78. Volpe, D.A.; Hamed, S.S.; Zhang, L.K. Use of Different Parameters and Equations for Calculation of IC50
Values in Efflux Assays: Potential Sources of Variability in IC50 Determination. AAPS J. 2013, 16, 172–180.
[CrossRef]
79. Greenblatt, D.J.; Venkatakrishnan, K.; Harmatz, J.S.; Parent, S.J.; von Moltke, L.L. Sources of variability in
ketoconazole inhibition of human cytochrome P450 3Ain vitro. Xenobiotica 2010, 40, 713–720. [CrossRef]
80. Huang, K.M.; Uddin, M.E.; Digiacomo, D.; Lustberg, M.B.; Hu, S.; Sparreboom, A. Role of SLC transporters
in toxicity induced by anticancer drugs. Expert Opin. Drug Metab. Toxicol. 2020, 16, 493–506. [CrossRef]
81. Johnston, R.A.; Rawling, T.; Chan, T.; Zhou, F.; Murray, M. Selective Inhibition of Human Solute Carrier
Transporters by Multikinase Inhibitors. Drug Metab. Dispos. 2014, 42, 1851–1857. [CrossRef] [PubMed]
82. Lacy, S.; Hsu, B.; Miles, D.; Aftab, D.; Wang, R.; Nguyen, L. Metabolism and Disposition of Cabozantinib in
Healthy Male Volunteers and Pharmacologic Characterization of Its Major Metabolites. Drug Metab. Dispos.
2015, 43, 1190–1207. [CrossRef]
83. Khurana, V.; Minocha, M.; Pal, D.; Mitra, A.K. Inhibition of OATP-1B1 and OATP-1B3 by tyrosine kinase
inhibitors. Drug Metab. Drug Interact. 2014, 29, 249–259. [CrossRef] [PubMed]
84. Kotsampasakou, E.; Brenner, S.; Jaeger, W.; Ecker, G.F. Identification of Novel Inhibitors of Organic Anion
Transporting Polypeptides 1B1 and 1B3 (OATP1B1 and OATP1B3) Using a Consensus Vote of Six Classification
Models. Mol. Pharm. 2015, 12, 4395–4404. [CrossRef] [PubMed]
85. Polli, J.W.; Humphreys, J.E.; Harmon, K.A.; Castellino, S.; O’mara, M.J.; Olson, K.L.;
John-Williams, L.S.; Koch, K.M.; Serabjit-Singh, C.J. The role of efflux and uptake transporters
in [N-{3-chloro-4-[(3-fluorobenzyl)oxy]phenyl}-6-[5-({[2-(methylsulfonyl)ethyl]amino}methyl)-2-furyl]-4-
quinazolinamine (GW572016, lapatinib) disposition and drug interactions. Drug Metab. Dispos. 2008,
36, 695–701. [CrossRef] [PubMed]
86. Hayden, E.R. Phosphorylation and function of OATP1B1 with tyrosine kinase inhibitors. FASEB J. 2020, 34, 1.
[CrossRef]
87. Sprowl, J.A.; Chen, M.; Gibson, A.A.; Pasquariello, K.Z.; Sparreboom, A.; Hu, S. Characterization of OATP1B1
and OATP1B3 inhibition by Nilotinib. FASEB J. 2019, 33, 506–507.
88. Ohya, H.; Shibayama, Y.; Ogura, J.; Narumi, K.; Kobayashi, M.; Iseki, K. Regorafenib Is Transported by the
Organic Anion Transporter 1B1 and the Multidrug Resistance Protein 2. Biol. Pharm. Bull. 2015, 38, 582–586.
[CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

164
 pharmaceutics

Article
BCS Class IV Oral Drugs and Absorption Windows:
Regional-Dependent Intestinal Permeability
of Furosemide
Milica Markovic 1 , Moran Zur 1 , Inna Ragatsky 1 , Sandra Cvijić 2 and Arik Dahan 1, *
1 Department of Clinical Pharmacology, School of Pharmacy, Faculty of Health Sciences,
Ben-Gurion University of the Negev, Beer-Sheva 8410501, Israel; milica@post.bgu.ac.il (M.M.);
moranfa@post.bgu.ac.il (M.Z.); inna.ragatsky@gmail.com (I.R.)
2 Department of Pharmaceutical Technology and Cosmetology, Faculty of Pharmacy, University of Belgrade,
Vojvode Stepe 450, 11221 Belgrade, Serbia; sandra.cvijic@pharmacy.bg.ac.rs
* Correspondence: arikd@bgu.ac.il; Tel.: +972-8-647-9483; Fax: +972-8-647-9303

Received: 16 November 2020; Accepted: 30 November 2020; Published: 2 December 2020

Abstract: Biopharmaceutical classification system (BCS) class IV drugs (low-solubility low-permeability)

are generally poor drug candidates, yet, ~5% of oral drugs on the market belong to this class.
While solubility is often predictable, intestinal permeability is rather complicated and highly
dependent on many biochemical/physiological parameters. In this work, we investigated the
solubility/permeability of BCS class IV drug, furosemide, considering the complexity of the entire
small intestine (SI). Furosemide solubility, physicochemical properties, and intestinal permeability
were thoroughly investigated in-vitro and in-vivo throughout the SI. In addition, advanced in-silico
simulations (GastroPlus® ) were used to elucidate furosemide regional-dependent absorption pattern.
Metoprolol was used as the low/high permeability class boundary. Furosemide was found to be a
low-solubility compound. Log D of furosemide at the three pH values 6.5, 7.0, and 7.5 (representing
the conditions throughout the SI) showed a downward trend. Similarly, segmental-dependent
in-vivo intestinal permeability was revealed; as the intestinal region becomes progressively distal,
and the pH gradually increases, the permeability of furosemide significantly decreased. The opposite
trend was evident for metoprolol. Theoretical physicochemical analysis based on ionization, pKa ,
and partitioning predicted the same trend and confirmed the experimental results. Computational
simulations clearly showed the effect of furosemide’s regional-dependent permeability on its
absorption, as well as the critical role of the drug’s absorption window on the overall bioavailability.
The data reveals the absorption window of furosemide in the proximal SI, allowing adequate
absorption and consequent effect, despite its class IV characteristics. Nevertheless, this absorption
window so early on in the SI rules out the suitability of controlled-release furosemide formulations,
as confirmed by the in-silico results. The potential link between segmental-dependent intestinal
permeability and adequate oral absorption of BCS Class IV drugs may aid to develop challenging
drugs as successful oral products.

Keywords: BCS class IV drugs; segmental-dependent intestinal permeability; intestinal absorption; oral
drug delivery; biopharmaceutics; physiologically-based pharmacokinetic (PBPK) modeling; furosemide

1. Introduction
The biopharmaceutical classification system (BCS) developed by Amidon et al. revealed that
the solubility/dissolution of the drug and its intestinal permeability are the two key factors that
dictate drug absorption following oral administration [1,2]. Drug solubility in the gastrointestinal
milieu may change in different intestinal segments, e.g., due to pH changes, in a fairly predictable

165
Pharmaceutics 2020, 12, 1175; doi:10.3390/pharmaceutics12121175 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 1175

manner; depending on the pKa, the solubility of acidic drugs may increase as the luminal pH rises
in more distal regions of the small intestine, and vice versa for basic drugs [3–5]. On the other hand,
time- and segmental-dependent intestinal permeability is more complicated and harder to predict [1].
Mechanisms contributing to segmental-dependent permeability throughout the gastrointestinal tract
(GIT) include different morphology along the GIT, variable intestinal mucosal cell differentiation,
changes in the drug concentration (in case of carrier-mediated transport), modulation of tight junction
permeability, and luminal contents and properties, e.g., pH, transporter expression, variability in the
structure/composition of the intestinal membrane itself, and more [6–11].
The four BCS classes highlight the limiting factors of the absorption process: (1) Class I,
high-solubility high-permeability drugs, indicate the easier and straightforward development process,
and complete absorption is expected; (2) Class II, low-solubility high-permeability drugs, indicate
that a solubility/dissolution limitation is expected; (3) Class III, high-solubility low-permeability
drugs, indicate that the intestinal absorption of this class of drugs will be limited by the permeability
rate; and (4) Class IV, low-solubility low-permeability drugs [12]. Since Class IV drugs suffer from
inadequate solubility and permeability, they have very poor oral bioavailability and are inclined to
exhibit very large inter- and intrasubject variability. Therefore, unless the drug dose is very low,
they are generally poor oral drug candidates. Yet, according to some estimates, ~5% of the world’s top
oral drugs belong to this class [13–15]. In some cases, this is due to the absorption window, which is
often critical for the success or failure of a certain drug. In order to gather information about the drug
absorption window, extensive work and thorough analysis of luminal conditions and drug absorption
is needed, within different locations throughout the GIT. Here, we present such analysis for BCS class
IV drug, furosemide [16].
Furosemide is a powerful loop diuretic and is indicated for treating edematous conditions
associated with heart, renal, and hepatic failure, as well as for the treatment of hypertension [17,18].
Drug therapy with furosemide is often complex, due to apparent erratic oral systemic availability and
unpredictable responses to an administered dose [19]. Even though furosemide is a class IV drug, it is
a very common and widely prescribed drug on the market.
In this work, we aimed to investigate the reason for apparent success of furosemide as a marketed
product, despite its poor biopharmaceutical properties, and classification as BCS class IV drug, in order
to allow development of future class IV compounds. We posit that segmental-dependent permeability
of furosemide may contribute to its absorption complexity and provide a certain absorption window in
which the drug has suitable permeability and, hence, gets absorbed. For this reason, we investigated
the in-vivo intestinal permeability of furosemide throughout different segments of the small intestine.
Solubility studies, as well as theoretical physicochemical analysis of furosemide and advanced modern
in-silico GastroPlus® simulations, were performed, in order to elucidate the mechanistic reasons behind
the experimental results. Furosemide data were compared to the β-blocker metoprolol, the Food and
Drug Administration (FDA) reference drug for the low/high permeability class boundary. Overall,
this experimental setup allowed us to reveal important insights on the performance of furosemide,
despite its unfavorable drug-like properties, and discuss extrapolation of these insights to other BCS
class IV drug candidates.

2. Methods

2.1. Materials
Furosemide, metoprolol, phenol red, potassium chloride, potassium phosphate monobasic,
potassium phosphate dibasic, sodium chloride, acetic acid, maleic acid, n-octanol, and trifluoroacetic
acid (TFA) were all purchased from Sigma Chemical Co. (St. Louis, MO, USA). Acetonitrile and
water, ultra-performance liquid chromatography (UPLC) grade were purchased from Merck KGaA,
Darmstadt, Germany. Remaining chemicals were of analytical reagent grade.

166
Pharmaceutics 2020, 12, 1175

2.2. Solubility Studies

The pH-dependent solubility studies were performed using the shake flask method, as previously
reported [20–23]. The equilibrium solubility of furosemide was determined at both 37 ◦ C and at room
temperature (25 ◦ C), in phosphate buffer pH 7.5, acetate buffer pH 4.0, and maleate buffer pH 1.0.
Surplus quantity of furosemide was introduced to glass vials holding buffer solutions with different
pH; the pH of those solutions was measured following drug addition to the buffers and, consequently,
placed in the shaking incubator (100 rpm) at 37 ◦ C. The vials were centrifuged (10,000 rpm, 10 min),
and the supernatant was instantly analyzed by UPLC. The dose number for furosemide was calculated
using the established equation: D0 = M/V 0 /Cs ; M being the highest single-unit dose strength of
furosemide (taken as 80 mg [24]), V 0 is the initial volume of water (250 mL), and Cs is the solubility at
each pH; the drug is considered highly soluble if the D0 < 1.

2.3. Evaluation of Octanol-Buffer Partition Coefficients (Log D)

Furosemide and metoprolol experimental octanol-buffer partition coefficients (Log D) were
studied at pH 6.5, 7.0, and 7.5 using the shake-flask method [8,11]. Drug solutions in octanol-saturated
phosphate buffer (pH 6.5, 7.0, 7.5) were equilibrated at 37 ◦ C for 48 h. The octanol and water phase
were divided via centrifugation, and the drug content in the water phase was quantified using UPLC;
the furosemide/metoprolol concentration in the octanol phase was determined by mass balance.

2.4. Physicochemical Analysis

The theoretical fraction extracted into octanol (fe ) was calculated using the following equation [25,26].

fu P
fe = ,
1 + fu P

in which P represents the octanol-water partition coefficient of the unionized drug form, and fu is the
fraction unionized of the drug at a certain pH. Experimental Log P values were taken from the literature
for both furosemide (2.29) [27] and metoprolol (2.19) [28]. The fu versus pH was plotted according to
the Henderson-Hasselbalch equation, using the pKa literature values: 9.68 for metoprolol [29] and
3.8 for furosemide [24].

2.5. Rat Single-Pass Intestinal Perfusion

Effective permeability coefficient (Peff ) of furosemide versus metoprolol in various intestinal
segments was assessed using the single-pass rat intestinal perfusion (SPIP) in-vivo model. The murine
studies were completed according to the approved protocol by Ben-Gurion University of the Negev
Animal Use and Care Committee (Protocol IL-08-01-2015). The animals (male Wistar rats weighing
230–260 g, Harlan, Israel) were housed and handled according to Ben-Gurion University of the Negev
Unit for Laboratory Animal Medicine Guidelines. All animals were fasted overnight (12–18 h) with free
access to water; rats were randomly allocated to different experimental groups. The intestinal perfusion
study was performed according to the previous reports [7,9,30–32]. Animals were anesthetized via
intramuscular injection of 1 mL/kg ketamine-xylazine solution (9%:1%) and placed on a heated (37 ◦ C)
surface (Harvard Apparatus Inc., Holliston, MA, USA); the rat abdomen was uncovered via a midline
incision (~3 cm). Permeability (Peff ) was measured in proximal jejunum (starting 2 cm lower from the
ligament of Treitz), mid-small intestine (SI) segment (isolated between the end of the upper and the
beginning of the lower segments), and distal segment of the ileum (ending 2 cm above the cecum)
accounting for the complexity of the entire SI [7]. Intestinal segments were cannulated on both ends
and perfused with drug-free buffer. Working solutions containing furosemide (320 μg/mL), metoprolol
(400 μg/mL), and phenol red (a non-absorbable marker for water flux measurements) were prepared
with potassium phosphate monobasic and sodium phosphate dibasic, to achieve pH of 6.5, 7.0 and
7.5; osmolarity (290 mOsm/L) and ionic strength in all buffers was maintained throughout the study.

167
Pharmaceutics 2020, 12, 1175

Drug solutions were incubated in a 37 ◦ C water bath. Steady-state environment was ensured by
perfusing the drug-containing buffer for 1 h, followed by additional 1 h of perfusion, during which
sampling was done every 10 min. The pH of the collected samples was measured in the outlet sample
to verify that there was no pH change throughout the perfusion. All samples were assayed by UPLC.
The length of each perfused intestinal segment was measured in the end of the experiment. The effective
permeability (Peff ; cm/s) through the rat SI wall was calculated according to the following equation:

−Qln (C out /C in )

Peff = ,
2πRL

in which Q is the perfusion buffer flow rate (0.2 mL/min); C out /C in is the ratio of the outlet/inlet drug
concentration adjusted for water transport; R is the radius of the intestinal segment (conventionally used
as 0.2 cm); and L is the exact length of the perfused SI segment as was measured at the experiment
endpoint [7,33,34].

2.6. Analytical Methods

Concentration of furosemide and metoprolol was evaluated using an UPLC instrument Waters
Acquity UPLC H-Class (Milford, MA, USA), with a photodiode array detector and Empower software.
Furosemide and metoprolol were separated on Acquity UPLC XTerra C18 3.5 μm 4.6 mm × 250 mm
column (Waters Co., Milford, MA, USA). Gradient mobile phase, going from 70:30% to 90:10% v/v 0.1%
trifluoroacetic acid in water/acetonitrile, respectively, on a flow rate of 1 mL/min (25 ◦ C). The inter- and
intraday coefficients of variation were < 1.0% and 0.5%, respectively.

2.7. Statistics
Solubility studies were performed in four replicates; Log D studies were performed in six
replicates, whereas animal perfusion studies were n = 4. Values are expressed as means ± standard
deviation (SD). To determine statistically significant differences among the experimental groups,
a 2-tailed nonparametric Mann–Whitney U test for 2-group comparison was used; p < 0.05 was
termed significant.

2.8. In-Silico Simulations

Computer simulations of furosemide absorption and concomitant plasma concentrations following
oral administration in humans were conducted using GastroPlusTM software package (v. 9.7.0009, 2019,
Simulations Plus Inc., Lancaster, CA, USA). The required input data regarding drug physicochemical
and pharmacokinetic properties were experimentally determined, taken from literature or in-silico
predicted. Human permeability values throughout the SI were calculated from the experimental
rat single-pass intestinal perfusion data, using the software integrated “permeability converter”.
Drug disposition was best described by three-compartmental pharmacokinetic model, whereas the
relevant parameters (clearance (CL), volume of distribution (Vd) and distribution constants between
central and peripheral compartments) were estimated using PKPlus software module, based on
the in-vivo plasma concentration data for an intravenous (i.v.) bolus dose [35]. The application of
three-compartmental model to describe furosemide pharmacokinetics has already been reported in
literature [36,37]. Graphical data from literature were digitized using DigIt™ program (version 1.0.4,
2001–2008, Simulations Plus, Inc., Lancaster, CA, USA). Physiological parameters were the software
default values representing fasted state physiology of a healthy human representative.
The software simulates drug absorption from the GIT using the integrated Advanced Compartment
Absorption and Transit (ACAT) GIT model that consists of nine compartments (stomach, duodenum,
two segments of jejunum, three segments of ileum, caecum, and ascendant colon). These compartments
are linked in series, and the amount of drug dissolved and absorbed from each compartment is
calculated by the system of differential equations. More details on the ACAT model can be found

168
Pharmaceutics 2020, 12, 1175

in the literature [38,39]. Regarding the fact that furosemide is a poorly-soluble drug, the model
accounted for the effect of bile salt on drug solubility and diffusion coefficient. Drug dissolution rate
under physiological conditions was predicted using the software default Johnson dissolution equation
(based on modified Nernst-Bruner equation) [40].
The validity of the model (i.e., the selection of input values) was validated by comparison of the
prediction results (bioavailability (F), maximum plasma concentration (Cmax ), time to reach Cmax (tmax ),
and area under the plasma concentration-time curve (AUC0–∞ )) with published data from the in-vivo
studies for peroral (p.o.) drug administration. Percent prediction error (%PE) between the predicted
and mean in-vivo observed data from a clinical study was calculated using the following equation:

(Observed value − Predicted value) × 100

%PE = .
Observed value
In the next step, the generated model was used to mechanistically interpret furosemide regional
absorption pattern, and to estimate the outcomes for various hypothetical drug dissolution scenarios
(illustrating drug dissolution from immediate-release (IR) and controlled-release (CR) oral formulations).
In the last case, hypothetical dissolution profiles were used as additional inputs to describe drug
release rate in-vivo, and the selected dosage form was “CR dispersed” to allow input of the tabulated
dissolution data.

3. Results
The solubility values obtained for furosemide at 37 ◦ C and at room temperature (25 ◦ C) are
summarized in Table 1, as well as the corresponding dose number (D0 ). Furosemide showed
pH-dependent solubility, in accordance with its acidic nature. It can be seen that, while, at pH 7.5,
furosemide has suitable solubility (as evident by D0 lower than 1), at the lower pH values, 1.0 and 4.0,
it is poorly soluble. When taking 80 mg as the highest dose strength, although D0 < 1 was obtained at
pH 7.5, at pH 1.0 and 4.0, the D0 is higher than 1; hence, furosemide was found to be a low-solubility
compound according to the BCS.

Table 1. Furosemide solubility values (μg/mL) at the tree pH values 1.0, 4.0, and 7.5, at 37 ◦ C
(upper panel), and at room temperature (25 ◦ C; lower panel), as well as the corresponding dose number
(D0 ) calculated for an 80-mg dose. Data presented as mean ± SD; n = 6.

At 37 ◦ C
pH Solubility (μg/mL) Corresponding D0
1 19.4 ± 3.7 16.5
4 65.5 ± 9.0 4.8
7.5 8340.1 ± 81.6 0.04
At 25 ◦ C
pH Solubility (μg/mL) Corresponding D0
1 40.3 ± 16.2 7.9
4 56.7 ± 12.2 5.6
7.5 8550.6 ± 149.4 0.04

Octanol-buffer partition coefficient values of furosemide and metoprolol at the three pH values
6.5, 7.0, and 7.5 (representing the conditions throughout the small intestine) are presented in Figure 1.
Both drugs presented a clear pH-dependent Log D values across the studied pH range, with opposite
trends; while furosemide’s partitioning decreases as the pH rises, metoprolol shows higher partitioning
into octanol at higher pH (metoprolol is the acceptable reference drug for the low/high permeability
class boundary). In addition, furosemide’s Log D at pH 6.5 was higher than that of metoprolol at the
same pH; this is a surprising finding since Log D may sometimes be used as a surrogate for passive

169
Pharmaceutics 2020, 12, 1175

permeability. Indeed, at higher pH values (7.0 and 7.5), metoprolol Log D increases, while furosemide
decreases, and metoprolol Log D becomes higher than furosemide.

Figure 1. The octanol-buffer partition coefficients, Log D, for furosemide and metoprolol at the three
pH values 6.5, 7.0, and 7.5. Data are presented as the mean ± S.D.; n = 6 in each experimental group.

Furosemide and metoprolol physicochemical properties are presented in Table 2. Figure 2 presents
furosemide versus metoprolol theoretical fraction unionized (fu ) and fraction extracted into octanol (fe )
as a function of pH. The plots have a standard sigmoidal shape, with opposite trends for furosemide
vs. metoprolol. The fe vs. pH plot follows the same pattern to the fu plot, only with a shift to the
right (higher pH values) for acidic drug (furosemide), and to the left (lower pH values) for basic
(metoprolol) drugs. The shift magnitude in both cases equals Log(P − 1) at the midpoint of the fe and
fu curves [25,26]. The experimental drug octanol-buffer partitioning at the three pH values (6.5, 7.0,
and 7.5) are illustrated in Figure 2, as well, and it can be seen that they were in excellent agreement
with the theoretical plots.

Table 2. Physicochemical parameters and chemical structure of furosemide and metoprolol.

Drug Chemical Structure pKa Log P PSA

Furosemide 3.8 2.3 127.7

Metoprolol 9.7 2.2 53.2

170
Pharmaceutics 2020, 12, 1175

Figure 2. The theoretical fraction unionized (fu ) and fraction extracted into octanol (fe ) plots as a
function of pH for furosemide and metoprolol, as well as experimental buffer-octanol partitioning of
the drugs in the three pH values 6.5, 7.0, and 7.5 (n = 5).

The effective permeability coefficient (Peff , cm/sec) values of furosemide and metoprolol determined
using the single-pass intestinal perfusion (SPIP) rat model, in three intestinal segments, namely proximal
jejunum (pH 6.5), mid small intestine (pH 7.0), and distal ileum (pH 7.5), are presented in Figure 3.
It can be seen that significant regional-dependent permeability of furosemide throughout the small
intestine was evident: the permeability of furosemide gradually decreases, while the permeability of
metoprolol gradually increases, as the SI segments become more distal.

Figure 3. Effective permeability values (Peff ; cm/s) obtained for furosemide and metoprolol after in-situ
single pass perfusion to the rat proximal jejunum at pH 6.5, mid-small intestine at pH 7.0, and to the
distal ileum at pH 7.5. Mean ± S.D.; n = 4 in each experimental group; ** p < 0.01, *** p < 0.001.

The input data regarding drug physicochemical and pharmacokinetic properties, used for in-silico
simulations, are presented in Table 3. The simulated furosemide plasma concentration profile following
p.o. administration is depicted in Figure 4, along with the mean profiles observed in the in-vivo
studies. In addition, the observed and model predicted pharmacokinetic parameters are compared in
Table 4. The presented data demonstrate that the generated model adequately describes furosemide
absorption and disposition. The course of the predicted plasma profile fairly resembles the observed

171
Pharmaceutics 2020, 12, 1175

data. However, certain variations are observed between the mean in-vivo data from different studies
referring to the same drug dose (Figure 4, Table 4). Indeed, it has been reported that furosemide
oral absorption is highly variable between individuals, e.g., Cmax varied three-fold, and tmax varied
five-fold [36,37,41]; moreover, individual AUC values for 40 mg furosemide oral dose varied between
1.57 and 3.76 μg·h/mL (more than two-fold) [36,37,41], and even larger AUC values were observed
in another study with the same drug dose (2.23–6.10 μg·h/mL) [42], indicating that, regardless of the
high PE(%) values in Table 4, the model predicted value of 3.66 μg·h/mL is not an overestimate of
the extent of drug absorption. In addition, extensive intrasubject variability was observed for orally
dosed furosemide, and these variations were attributed to the absorption process (i.e., day to day
variations in physiological factors) since the repeated i.v. doses showed only marginal intrasubject
variability [36,37,41]. Considering pronounced inter- and intraindividual variability in furosemide
oral absorption, the simulated profile can be seen as a reasonable estimate (Figure 4). Moreover,
the predicted fraction of oral drug absorption (cc. 52%) is in accordance with the values reported in the
literature [36,37].

Table 3. The selected input parameters for furosemide absorption GastroPlus® simulation.

Parameter Value Source

Molecular weight (g/mol) 330.75 /
Log D (pH 7.5) −1.0818
19.4 (pH 1.0) experimental values
Solubility at 37 ◦ C (μg/mL) 65.5 (pH 4.0)
8340.1 (pH 7.5)
pKa (acid) 3.8 [24]
0.4043 × 10−4 (duodenum, jejunum) values converted using
Human effective permeability, Peff (cm/s) 0.2246 × 10−4 (ileum 1 and 2) GastroPlus™ integrated
0.1392 × 10−4 (ileum 3, caecum, colon) “permeability converter” based on
experimental rat perfusion data
GastroPlus™ calculated value
Diffusion coefficient (cm2 /s) 0.7289 × 10−5
(based on molecular weight)
Mean precipitation time (s) 900
Particle density (g/mL) 1.2
GastroPlus™ default values
Particle radius (μm) 25
Blood/plasma concentration ratio 1
Plasma fraction unbound (%) 1 [24]
Clearance, CL (L/h/kg) 0.121
Volume of distribution, Vd (L/kg) 0.043
Distribution constant k12 (1/h) 0.964 calculated using GastroPlus™
PKPlus module, based on the
Distribution constant k21 (1/h) 1.614 i.v. data [35]
Distribution constant k13 (1/h) 0.925
Distribution constant k32 (1/h) 0.708
Regional pH in the GIT 1.3; 6.0; 6.2; 6.4; 6.6; 6.9; 7.4; 6.4; 6.8
GastroPlus™ default values for
46.56; 40.54; 150.00; 119.30; 91.71; 68.88;
Regional volume of fluid in the GIT (mL) stomach, duodenum, jejunum 1,
48.57; 46.44; 49.21
jejunum 2, ileum 1, ileum 2, ileum
0.25; 0.26; 0.93; 0.74; 0.58; 0.42; 0.29; 4.13; 3, caecum, and ascendant colon
Regional transit time in the GIT (h)
12.38

172
Pharmaceutics 2020, 12, 1175

Figure 4. GastroPlus® simulated (line) versus mean observed (markers) plasma concentration profiles
following p.o. administration of furosemide. Mean observed values represent 40 mg immediate-release
(IR) tablet profile I [43] and 40 mg IR tablet profile II [37].

Table 4. Comparison between GastroPlus® simulated and in-vivo observed furosemide

pharmacokinetic parameters following p.o. drug administration.

40 mg p.o. Dose
Parameter In-Vivo I a In-Vivo II b Predicted PE(%) I PE(%) II
Cmax (μg/mL) 0.61 0.75 0.71 −17.14 5.54
tmax (h) 1.5 1.12 1.36 9.33 −22.22
AUC0→∞ (μg·h/mL) 2.13 2.44 3.66 −71.25 −50.06
AUC0→24 h (μg·h/mL) 2.11 2.33 2.52 −19.25 −8.15
F (%) NA NA 52.2 NA NA
a Refers to the mean plasma profile from [43] (40 mg IR tablet); b refers to the mean plasma profile from [37]
(40 mg IR tablet); NA, not available/not applicable.

The predicted furosemide dissolution and absorption profiles following an IR oral formulation
(IR tablet) are illustrated in Figure 5. The generated profiles clearly indicate that drug permeability is
the limiting factor for absorption under fasted state GIT conditions. Namely, although furosemide is a
low-solubility drug, due to ionization at the elevated pH conditions in the proximal SI, drug dissolution
from an IR formulation is expected to be fast (>85% in 30 min). Therefore, furosemide absorption
from an IR formulation is mainly governed by poor permeability. The predicted regional-dependent
absorption distribution (Figure 6) further highlights the role of furosemide segmental absorption on
the overall drug bioavailability. As implied by the regional-dependent permeability data, but also
considering the surface area available for absorption, furosemide absorption predominantly happens
in the proximal parts of the SI (76.6% of the total amount absorbed into the enterocytes), and only a
minor fraction of drug (23.2% of the total amount absorbed into the enterocytes) passes into systemic
circulation through mid and distal GIT regions.

173
Pharmaceutics 2020, 12, 1175

Figure 5. GastroPlus® simulated dissolution and absorption profiles following p.o. administration of
40 mg furosemide dose (dissolution profile was simulated using the software default Johnson equation).

Figure 6. GastroPlus® simulated regional absorption of furosemide following p.o. administration of

40 mg drug dose (the simulated values refer to the fraction of drug dose that entered into the enterocytes).

The prediction results corresponding to various dissolution scenarios are presented in Figure 7b–d
and Table 5. According to the simulated data, furosemide release rate from an oral formulation highly
impacts the concomitant absorption process, whereas prolonged drug release rate leads to marked
delay in the rate and extent of drug absorption. The estimated pharmacokinetic parameters (Table 5)
indicate that furosemide bioavailability would show more than a 10-fold decrease in case the complete
drug dissolution is achieved within 24 h in comparison to 15 min. A similar trend is observed for
Cmax and AUC values (17.75- and 17.38-fold decrease, respectively), while tmax would be prolonged
(about two-fold). It is interesting to note that tmax increases with decrease in drug dissolution up to
some point, but further decrease in drug dissolution (e.g., 85% in more than 6 h) would not cause
additional delay in peak plasma concentration. This is because, after cc. 2 h, the drug leaves proximal
parts of the intestine, where majority of furosemide absorption takes place, and, later on, in mid and
especially distal intestine, only a small fraction of drug can be absorbed, as illustrated in Figure 7d.

174
Pharmaceutics 2020, 12, 1175

Figure 7. GastroPlus® simulated furosemide dissolution profiles (a); and (b) the corresponding
simulated plasma profiles; (c) absorption profiles; and (d) regional absorption distribution.

Table 5. GastroPlus® predicted pharmacokinetic parameters for different furosemide virtual dissolution
profiles from 40 mg p.o. dosage forms.

Dissolution Cmax (μg/mL) tmax (h) AUC0→ ∞ (μg·h/mL) F (%)

85% in 15 min 0.71 1.36 3.65 51.91
85% in 1 h 0.64 1.76 3.71 46.35
85% in 6 h 0.15 2.80 0.80 16.64
85% in 8 h 0.11 2.80 0.61 12.73
85% in 12 h 0.08 2.80 0.41 8.65
85% in 24 h 0.04 2.80 0.21 4.36

4. Discussion
BCS class IV drugs (e.g., sulfamethoxazole, ritonavir, paclitaxel, and furosemide) exhibit
numerous unfavorable characteristics (low solubility and permeability, high presystemic metabolism,
efflux transport), which make their oral drug delivery challenging. In addition to this, class IV drugs
often demonstrate inter/intra-subject variability. Indeed, following oral administration, the absorption
and bioavailability of furosemide are highly variable (37–51%) [35,41]. It has been suggested that
this variability is highly depend on the absorption process [41], which in turn is dependent on
drug aqueous solubility and intestinal permeability following oral administration [1,44]. It has also
been hypothesized that variable gastric/intestinal first-pass metabolism can be a factor in causing
incomplete and irregular furosemide absorption in humans [45]. Despite the unfavorable class IV
drug characteristics, furosemide was shown to be exceptionally useful and successful marketed drug
product for the treatment of edema [17]. For this reason, we decided to investigate furosemide’s
solubility and in-vivo regional-dependent permeability throughout the GIT, as main parameters that
guide absorption of oral drugs.

175
Pharmaceutics 2020, 12, 1175

It was shown that a correlation between human Peff in the jejunum and physicochemical parameters
advocates that there is a high pH-dependent influence on the passive intestinal permeability in-vivo [46].
Indeed, furosemide in-vivo permeability data demonstrate a downward trend towards the distal
intestinal segments as the pH gradually increases, a trend that can be expected for acidic drugs,
since the pH in the intestinal lumen gradually increases towards distal SI regions (Figure 3). Many BCS
class IV drugs are substrates for efflux transporters [47]. There is some evidence that furosemide might
be a substrate for efflux transporters [48,49]; thus, such permeability trend could also be influenced
by the P-glycoprotein (P-gp) transporter in which expression levels are increased from proximal to
distal SI segments [6,50–52]. Since metoprolol’s intestinal permeability is passive and does not involve
carrier-mediated absorption, it exhibited pH-dependent intestinal permeability, with reverse tendency
compared to furosemide; as a basic drug, metoprolol showed upward increase in permeability towards
distal SI segments with rising pH values (Figure 3). At any point throughout the SI, furosemide
exhibited significantly lower permeability than the benchmark (metoprolol’s jejunum permeability),
which confirms its BCS low-permeability classification and incomplete absorption. Despite the fact
that furosemide is a low-permeability drug, the higher permeability in the proximal intestinal regions
provides a window for furosemide absorption, and we posit that this is one of the main reasons for
furosemide’s sufficient bioavailability and success as a marketed drug. Theoretical fu and fe as a function
of pH were found to be in excellent correlation to these in-vivo data. In addition, in-silico modeling
indicated that furosemide dissolution from an IR formulation would be fairly complete before the drug
leaves proximal SI (Figure 5), although the drug is generally classified as low-soluble, enabling timely
delivery of the dissolved drug to the distinct absorption site. Complete furosemide dissolution under
physiological conditions is also confirmed by the experimental solubility results (Table 1).
Furosemide Log D studies showed higher partition coefficient in comparison to metoprolol at
pH 6.5, whereas, in the in-vivo intestinal perfusion experiment, furosemide showed significantly
lower jejunum permeability than metoprolol (Figure 1). A possible reason for this difference in the
partitioning and in-vivo permeability can be the polar surface area (PSA) of both drugs [53]. A sigmoidal
relationship between the fraction absorbed following oral administration and the dynamic polar surface
area was reported in the past [54–56]. It was shown that orally administered drugs with large PSA
(>120) are hardly absorbed by the passive transcellular route, while drugs with a small PSA (<60) are
almost completely absorbed [55,56]. This is in agreement with our results, as furosemide has much
higher PSA (127.7) than metoprolol (53.2) [54,55]. Another reason for the difference in the partitioning
and in-vivo permeability may be the presence of active efflux transport involved in the intestinal
permeability. The influence of efflux transport at pH 6.5 (proximal intestinal segments) could decrease
furosemide’s permeability in-vivo, which was not accounted for in the octanol partitioning studies.
The Log P value of furosemide (2.3) is in the close proximity to that of metoprolol (2.2), pointing to
high permeability (Table 2). However, the Log P calculation is based on the unionized drug fraction,
and, since furosemide has acidic nature it is likely that, once it passes the acidic stomach environment,
it will mostly be in ionized form (the pH throughout the GIT varies from 5.9–6.3 in the proximal SI
to 7.4–7.8 in distal SI segments; pH in the colon is fluctuating between pH 5–8 [57]); therefore the
high furosemide Log P is not in correspondence with permeability in-vivo. Thus, we posit that no
single parameter can be used for measuring the drug absorption process, but rather, a combination
of physicochemical parameters and in-vitro and in-vivo findings, as well as careful consideration of
inclusion criteria prior to making decisions. Despite the high Log P value for furosemide, it was indeed
confirmed that furosemide is a BCS class IV drug, based on both the solubility data (Table 1) and the
intestinal permeability (Figure 3).
Suitable formulation is the main approach to create an efficacious drug product for the
administration of BCS class IV drugs [47]. Absorption windows in the proximal intestinal segments
can restrict the oral drug bioavailability and can be a significant limitation for the development of
CR drug formulation. The underlying reasons are mechanistically explained by our in-silico results
(Figure 7). As mentioned, furosemide permeability results revealed acceptable permeability in the

176
Pharmaceutics 2020, 12, 1175

proximal segments of the SI, which is presumably the reason why furosemide has appropriate drug
bioavailability, despite being a BCS class IV drug. However, since CR products release the drug over
12–24 h, mostly in the colon, (transit time throughout the small intestine is 3–4 h [58]), the fact that
furosemide is mainly absorbed from proximal SI segments, (with decreased permeability at distant GIT
segments) prevents the formulation of furosemide as a CR product, as shown previously [21,59,60].
However, we believe that formulations based on gastro-retentive dosage forms (GRDF) can be shown
as prosperous for furosemide [61]. There are several similar examples in the literature where absorption
window occurs in the upper GI, and this has been used to create GDRF formulations to improve the
drug absorption, such as riboflavin [62] and levodopa [59,63].
Several types of bariatric surgeries (specifically Roux-en-Y gastric bypass and mini bypass) result
in bypassing the upper SI. In cases where the absorption window is indeed in this upper SI region,
the absorption following the bariatric surgery can be hampered vastly, since the actual segment
responsible for the majority of absorption is bypassed [64–66].

5. Conclusions
Regional-dependent permeability throughout the small intestine was evident for furosemide.
The permeability of furosemide gradually decreases throughout the small intestine as a function of the
pH change in the intestinal lumen. However, at any point throughout the small intestine, furosemide
exhibited significantly lower permeability than the benchmark of metoprolol s permeability in the
jejunum, which may explain the incomplete absorption of the drug. We propose that, for a drug to be
classified as BCS low-permeability, its intestinal permeability should not match/exceed the low/high
class benchmark anywhere throughout the intestinal tract, as well as is not restricted necessarily to the
jejunum. Nevertheless, low-permeable drugs should not be treated as ‘unfavorable’ by default; instead,
therapeutic potential and suitable formulation strategies should be considered on a case-by-case basis,
taking into account the overall results of in-vitro, in-vivo, and in-silico testing, throughout the entire
gastrointestinal tract.

Author Contributions: M.M., M.Z., I.R., S.C. and A.D. worked on conceptualization, methodology, investigation,
analyzed the data, and outlined the manuscript. S.C. worked on software investigation. Writing: S.C., A.D. and
M.M. prepared the original draft of the article, and M.Z. and I.R. contributed to the writing-review and editing of
the full version. All authors have read and agreed to the published version of the manuscript.
Funding: This work received no external funding.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Amidon, G.L.; Lennernäs, H.; Shah, V.P.; Crison, J.R. A theoretical basis for a biopharmaceutic drug
classification: The correlation of in vitro drug product dissolution and in vivo bioavailability. Pharm. Res.
1995, 12, 413–420. [CrossRef]
2. Dahan, A.; Miller, J.M.; Amidon, G.L. Prediction of solubility and permeability class membership: Provisional
BCS classification of the world’s top oral drugs. AAPS J. 2009, 11, 740–746. [CrossRef]
3. Dahan, A.; Beig, A.; Lindley, D.; Miller, J.M. The solubility-permeability interplay and oral drug formulation
design: Two heads are better than one. Adv. Drug Deliv. Rev. 2016, 101, 99–107. [CrossRef]
4. Dahan, A.; Miller, J.M. The solubility-permeability interplay and its implications in formulation design and
development for poorly soluble drugs. AAPS J. 2012, 14, 244–251. [CrossRef]
5. Miller, J.M.; Beig, A.; Carr, R.A.; Webster, G.K.; Dahan, A. The solubility-permeability interplay when using
cosolvents for solubilization: Revising the way we use solubility-enabling formulations. Mol. Pharm. 2012, 9,
581–590. [CrossRef]
6. Dahan, A.; Amidon, G.L. Segmental dependent transport of low permeability compounds along the small
intestine due to P-Glycoprotein: The role of efflux transport in the oral absorption of BCS class III drugs.
Mol. Pharm. 2009, 6, 19–28. [CrossRef]

177
Pharmaceutics 2020, 12, 1175

7. Dahan, A.; West, B.T.; Amidon, G.L. Segmental-dependent membrane permeability along the intestine
following oral drug administration: Evaluation of a triple single-pass intestinal perfusion (TSPIP) approach
in the rat. Eur. J. Pharm. Sci. 2009, 36, 320–329. [CrossRef]
8. Fairstein, M.; Swissa, R.; Dahan, A. Regional-dependent intestinal permeability and BCS classification:
Elucidation of pH-related complexity in rats using pseudoephedrine. AAPS J. 2013, 15, 589–597. [CrossRef]
9. Lozoya-Agullo, I.; Zur, M.; Beig, A.; Fine, N.; Cohen, Y.; Gonzalez-Alvarez, M.; Merino-Sanjuan, M.;
Gonzalez-Alvarez, I.; Bermejo, M.; Dahan, A. Segmental-dependent permeability throughout the small
intestine following oral drug administration: Single-pass vs. Doluisio approach to in-situ rat perfusion.
Int. J. Pharm. 2016, 515, 201–208. [CrossRef]
10. Markovic, M.; Zur, M.; Dahan, A.; Cvijić, S. Biopharmaceutical characterization of rebamipide: The role of
mucus binding in regional-dependent intestinal permeability. Eur. J. Pharm. Sci. 2020, 152, 105440. [CrossRef]
11. Zur, M.; Hanson, A.S.; Dahan, A. The complexity of intestinal permeability: Assigning the correct BCS
classification through careful data interpretation. Eur. J. Pharm. Sci. 2014, 61, 11–17. [CrossRef]
12. U.S. Department of Health and Human Services, Food and Drug Administration; Center for Drug Evaluation
and Research (CDER). Waiver of In-Vivo Bioavailability and Bioequivalence Studies for Immediate-Release Solid
Oral Dosage Forms Based on a Biopharmaceutics Classification System; Guidance for Industry; Center for Drug
Evaluation and Research (CDER): Silver Spring, MD, USA, 2017.
13. Dahan, A.; Wolk, O.; Kim, Y.H.; Ramachandran, C.; Crippen, G.M.; Takagi, T.; Bermejo, M.; Amidon, G.L.
Purely in silico BCS classification: Science based quality standards for the world’s drugs. Mol. Pharm. 2013,
10, 4378–4390. [CrossRef]
14. Takagi, T.; Ramachandran, C.; Bermejo, M.; Yamashita, S.; Yu, L.X.; Amidon, G.L. A provisional
biopharmaceutical classification of the top 200 oral drug products in the United States, Great Britain,
Spain, and Japan. Mol. Pharm. 2006, 3, 631–643. [CrossRef]
15. Wolk, O.; Agbaria, R.; Dahan, A. Provisional in-silico biopharmaceutics classification (BCS) to guide oral
drug product development. Drug Des. Dev. Ther. 2014, 8, 1563–1575. [CrossRef]
16. Lindenberg, M.; Kopp, S.; Dressman, J.B. Classification of orally administered drugs on the World Health
Organization model list of essential medicines according to the biopharmaceutics classification system. Eur. J.
Pharm. Biopharm. 2004, 58, 265–278. [CrossRef]
17. Furosemide Tablets, United States Pharmacopeia Label. Available online: https://www.accessdata.fda.gov/
drugsatfda_docs/label/2016/018487s043lbl.pdf (accessed on 11 July 2020).
18. Ellison, D.H.; Felker, G.M. Diuretic Treatment in Heart Failure. N. Engl. J. Med. 2017, 377, 1964–1975. [CrossRef]
19. Hammarlund-Udenaes, M.; Benet, L.Z. Furosemide pharmacokinetics and pharmacodynamics in health and
disease—An update. J. Pharmacokinet. Biopharm. 1989, 17, 1–46. [CrossRef]
20. Dahan, A.; Wolk, O.; Zur, M.; Amidon, G.L.; Abrahamsson, B.; Cristofoletti, R.; Groot, D.W.; Kopp, S.;
Langguth, P.; Polli, J.E.; et al. Biowaiver monographs for immediate-release solid oral dosage forms: Codeine
phosphate. J. Pharm. Sci. 2014, 103, 1592–1600. [CrossRef]
21. Markovic, M.; Zur, M.; Fine-Shamir, N.; Haimov, E.; González-Álvarez, I.; Dahan, A. Segmental-dependent
solubility and permeability as key factors guiding controlled release drug product development. Pharmaceutics
2020, 12, 295. [CrossRef]
22. Zur, M.; Cohen, N.; Agbaria, R.; Dahan, A. The biopharmaceutics of successful controlled release drug
product: Segmental-dependent permeability of glipizide vs. metoprolol throughout the intestinal tract.
Int. J. Pharm. 2015, 489, 304–310. [CrossRef]
23. Zur, M.; Gasparini, M.; Wolk, O.; Amidon, G.L.; Dahan, A. The low/high BCS permeability class boundary:
Physicochemical comparison of metoprolol and labetalol. Mol. Pharm. 2014, 11, 1707–1714. [CrossRef]
24. Granero, G.E.; Longhi, M.R.; Mora, M.J.; Junginger, H.E.; Midha, K.K.; Shah, V.P.; Stavchansky, S.;
Dressman, J.B.; Barends, D.M. Biowaiver monographs for immediate release solid oral dosage forms:
Furosemide. J. Pharm. Sci. 2010, 99, 2544–2556. [CrossRef] [PubMed]
25. Wagner, J.G.; Sedman, A.J. Quantitaton of rate of gastrointestinal and buccal absorption of acidic and basic
drugs based on extraction theory. J. Pharmacokinet. Biopharm. 1973, 1, 23–50. [CrossRef]
26. Winne, D. Shift of pH-absorption curves. J. Pharmacokinet. Biopharm. 1977, 5, 53–94. [CrossRef]
27. Berthod, A.; Carda-Broch, S.; Garcia-Alvarez-Coque, M.C. Hydrophobicity of ionizable compounds.
A theoretical study and measurements of diuretic octanol−water partition coefficients by countercurrent
chromatography. Anal. Chem. 1999, 71, 879–888. [CrossRef]

178
Pharmaceutics 2020, 12, 1175

28. Henchoz, Y.; Guillarme, D.; Martel, S.; Rudaz, S.; Veuthey, J.L.; Carrupt, P.A. Fast log P determination
by ultra-high-pressure liquid chromatography coupled with UV and mass spectrometry detections.
Anal. Bioanal. Chem. 2009, 394, 1919–1930. [CrossRef]
29. Teksin, Z.S.; Hom, K.; Balakrishnan, A.; Polli, J.E. Ion pair-mediated transport of metoprolol across a three
lipid-component PAMPA system. J. Control. Release Off. J. Control. Release Soc. 2006, 116, 50–57. [CrossRef]
30. Lozoya-Agullo, I.; Gonzalez-Alvarez, I.; Zur, M.; Fine-Shamir, N.; Cohen, Y.; Markovic, M.; Garrigues, T.M.;
Dahan, A.; Gonzalez-Alvarez, M.; Merino-Sanjuán, M.; et al. Closed-loop doluisio (colon, small intestine)
and single-pass intestinal perfusion (colon, jejunum) in rat—Biophysical model and predictions based on
Caco-2. Pharm. Res. 2017, 35, 2. [CrossRef]
31. Lozoya-Agullo, I.; Zur, M.; Fine-Shamir, N.; Markovic, M.; Cohen, Y.; Porat, D.; Gonzalez-Alvarez, I.;
Gonzalez-Alvarez, M.; Merino-Sanjuan, M.; Bermejo, M.; et al. Investigating drug absorption from the colon:
Single-pass vs. Doluisio approaches to in-situ rat large-intestinal perfusion. Int. J. Pharm. 2017, 527, 135–141.
[CrossRef]
32. Lozoya-Agullo, I.; Zur, M.; Wolk, O.; Beig, A.; Gonzalez-Alvarez, I.; Gonzalez-Alvarez, M.;
Merino-Sanjuan, M.; Bermejo, M.; Dahan, A. In-situ intestinal rat perfusions for human Fabs prediction
and BCS permeability class determination: Investigation of the single-pass vs. the Doluisio experimental
approaches. Int. J. Pharm. 2015, 480, 1–7. [CrossRef]
33. Dahan, A.; Miller, J.M.; Hilfinger, J.M.; Yamashita, S.; Yu, L.X.; Lennernas, H.; Amidon, G.L. High-permeability
criterion for BCS classification: Segmental/pH dependent permeability considerations. Mol. Pharm. 2010,
7, 1827–1834. [CrossRef]
34. Wolk, O.; Markovic, M.; Porat, D.; Fine-Shamir, N.; Zur, M.; Beig, A.; Dahan, A. Segmental-dependent
intestinal drug permeability: Development and model validation of in silico predictions guided by in vivo
permeability values. J. Pharm. Sci. 2019, 108, 316–325. [CrossRef]
35. Kelly, M.R.; Cutler, R.E.; Forrey, A.W.; Kimpel, B.M. Pharmacokinetics of orally administered furosemide.
Clin. Pharmacol. Ther. 1974, 15, 178–186. [CrossRef]
36. Benet, L.Z. Pharmacokinetics/pharmacodynamics of furosemide in man: A review. J. Pharmacokinet. Biopharm.
1979, 7, 1–27. [CrossRef]
37. Hammarlund, M.M.; Paalzow, L.K.; Odlind, B. Pharmacokinetics of furosemide in man after intravenous and
oral administration. Application of moment analysis. Eur. J. Clin. Pharmacol. 1984, 26, 197–207. [CrossRef]
38. Agoram, B.; Woltosz, W.S.; Bolger, M.B. Predicting the impact of physiological and biochemical processes on
oral drug bioavailability. Adv. Drug Deliv. Rev. 2001, 50, S41–S67. [CrossRef]
39. Lin, L.; Wong, H. Predicting oral drug absorption: Mini review on physiologically-based pharmacokinetic
models. Pharmaceutics 2017, 9, 41. [CrossRef]
40. Lu, A.T.K.; Frisella, M.E.; Johnson, K.C. Dissolution modeling: Factors affecting the dissolution rates of
polydisperse powders. Pharm. Res. 1993, 10, 1308–1314. [CrossRef]
41. Grahnén, A.; Hammarlund, M.; Lundqvist, T. Implications of intraindividual variability in bioavailability
studies of furosemide. Eur. J. Clin. Pharmacol. 1984, 27, 595–602. [CrossRef]
42. Waller, E.S.; Hamilton, S.F.; Massarella, J.W.; Sharanevych, M.A.; Smith, R.V.; Yakatan, G.J.; Doluisio, J.T.
Disposition and absolute bioavailability of furosemide in healthy males. J. Pharm. Sci. 1982, 71, 1105–1108.
[CrossRef]
43. Beermann, B.; Midskov, C. Reduced bioavailability and effect of furosemide given with food. Eur. J.
Clin. Pharmacol. 1986, 29, 725–727. [CrossRef]
44. Dahan, A.; Lennernäs, H.; Amidon, G.L. The fraction dose absorbed, in humans, and high jejunal human
permeability relationship. Mol. Pharm. 2012, 9, 1847–1851. [CrossRef]
45. Lee, M.G.; Chiou, W.L. Evaluation of potential causes for the incomplete bioavailability of furosemide:
Gastric first-pass metabolism. J. Pharm. Biopharm. 1983, 11, 623–640. [CrossRef]
46. Dahlgren, D.; Lennernas, H. Intestinal permeability and drug absorption: Predictive experimental,
computational and in vivo approaches. Pharmaceutics 2019, 11, 411. [CrossRef]
47. Ghadi, R.; Dand, N. BCS class IV drugs: Highly notorious candidates for formulation development. J. Control.
Release Off. J. Control. Release Soc. 2017, 248, 71–95. [CrossRef]
48. Flanagan, S.D.; Cummins, C.L.; Susanto, M.; Liu, X.; Takahashi, L.H.; Benet, L.Z. Comparison of furosemide
and vinblastine secretion from cell lines overexpressing multidrug resistance protein (P-glycoprotein) and
multidrug resistance-associated proteins (MRP1 and MRP2). Pharmacology 2002, 64, 126–134. [CrossRef]

179
Pharmaceutics 2020, 12, 1175

49. Takahashi, M.; Washio, T.; Suzuki, N.; Igeta, K.; Fujii, Y.; Hayashi, M.; Shirasaka, Y.; Yamashita, S. Characterization
of gastrointestinal drug absorption in cynomolgus monkeys. Mol. Pharm. 2008, 5, 340–348. [CrossRef]
50. Cao, X.; Yu, L.X.; Barbaciru, C.; Landowski, C.P.; Shin, H.C.; Gibbs, S.; Miller, H.A.; Amidon, G.L.; Sun, D.
Permeability dominates in vivo intestinal absorption of P-gp substrate with high solubility and high
permeability. Mol. Pharm. 2005, 2, 329–340. [CrossRef]
51. Englund, G.; Rorsman, F.; Rönnblom, A.; Karlbom, U.; Lazorova, L.; Gråsjö, J.; Kindmark, A.; Artursson, P.
Regional levels of drug transporters along the human intestinal tract: Co-expression of ABC and SLC
transporters and comparison with Caco-2 cells. Eur. J. Pharm. Sci. 2006, 29, 269–277. [CrossRef]
52. Zimmermann, C.; Gutmann, H.; Hruz, P.; Gutzwiller, J.-P.; Beglinger, C.; Drewe, J. Mapping of multidrug
resistance gene 1 and multidrug resistance-associated protein isoform 1 to 5 mRNA expression along the
human intestinal tract. Drug Metab. Dispos. 2005, 33, 219. [CrossRef]
53. Winiwarter, S.; Bonham, N.M.; Ax, F.; Hallberg, A.; Lennernäs, H.; Karlén, A. Correlation of human jejunal
permeability (in vivo) of drugs with experimentally and theoretically derived parameters. A multivariate
data analysis approach. J. Med. Chem. 1998, 41, 4939–4949. [CrossRef]
54. Clark, D.E. Rapid calculation of polar molecular surface area and its application to the prediction of transport
phenomena. 1. Prediction of intestinal absorption. J. Pharm. Sci. 1999, 88, 807–814. [CrossRef]
55. Palm, K.; Luthman, K.; Ungell, A.-L.; Strandlund, G.; Beigi, F.; Lundahl, P.; Artursson, P. Evaluation of
dynamic polar molecular surface area as predictor of drug absorption: Comparison with other computational
and experimental predictors. J. Med. Chem. 1998, 41, 5382–5392. [CrossRef]
56. Ertl, P.; Rohde, B.; Selzer, P. Fast calculation of molecular polar surface area as a sum of fragment-based
contributions and its application to the prediction of drug transport properties. J. Med. Chem. 2000, 43,
3714–3717. [CrossRef]
57. Koziolek, M.; Grimm, M.; Becker, D.; Iordanov, V.; Zou, H.; Shimizu, J.; Wanke, C.; Garbacz, G.; Weitschies, W.
Investigation of pH and temperature profiles in the GI tract of fasted human subjects using the intellicap(® )
system. J. Pharm. Sci. 2015, 104, 2855–2863. [CrossRef]
58. Davis, S.S.; Hardy, J.G.; Fara, J.W. Transit of pharmaceutical dosage forms through the small intestine.
Gut 1986, 27, 886–892. [CrossRef]
59. Streubel, A.; Siepmann, J.; Bodmeier, R. Drug delivery to the upper small intestine window using
gastroretentive technologies. Curr. Opin. Pharmacol. 2006, 6, 501–508. [CrossRef]
60. Clear, N.J.; Milton, A.; Humphrey, M.; Henry, B.T.; Wulff, M.; Nichols, D.J.; Anziano, R.J.; Wilding, I.
Evaluation of the Intelisite capsule to deliver theophylline and frusemide tablets to the small intestine and
colon. Eur. J. Pharm. Sci. Off. J. Eur. Fed. Pharm. Sci. 2001, 13, 375–384. [CrossRef]
61. Darandale, S.S.; Vavia, P.R. Design of a gastroretentive mucoadhesive dosage form of furosemide for
controlled release. Acta Pharm. Sin. B 2012, 2, 509–517. [CrossRef]
62. Kagan, L.; Lapidot, N.; Afargan, M.; Kirmayer, D.; Moor, E.; Mardor, Y.; Friedman, M.; Hoffman, A.
Gastroretentive accordion pill: Enhancement of riboflavin bioavailability in humans. J. Control. Release Off. J.
Control. Release Soc. 2006, 113, 208–215. [CrossRef]
63. Klausner, E.A.; Eyal, S.; Lavy, E.; Friedman, M.; Hoffman, A. Novel levodopa gastroretentive dosage form:
In-vivo evaluation in dogs. J. Control. Release 2003, 88, 117–126. [CrossRef]
64. Israel, S.; Elinav, H.; Elazary, R.; Porat, D.; Gibori, R.; Dahan, A.; Azran, C.; Horwitz, E. Case report of increased
exposure to antiretrovirals following sleeve gastrectomy. Antimicrob. Agents Chemother. 2020, 64. [CrossRef]
65. Porat, D.; Dahan, A. Medication management after bariatric surgery: Providing optimal patient care.
J. Clin. Med. 2020, 9, 1511. [CrossRef]
66. Porat, D.; Markovic, M.; Zur, M.; Fine-Shamir, N.; Azran, C.; Shaked, G.; Czeiger, D.; Vaynshtein, J.;
Replyanski, I.; Sebbag, G.; et al. Increased paracetamol bioavailability after sleeve gastrectomy: A crossover
pre- vs. post-operative clinical trial. J. Clin. Med. 2019, 8, 1949. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

180
 pharmaceutics

Review
Drug–Drug Interactions Involving Intestinal and
Hepatic CYP1A Enzymes
Florian Klomp 1 , Christoph Wenzel 2 , Marek Drozdzik 3 and Stefan Oswald 1, *
1 Institute of Pharmacology and Toxicology, Rostock University Medical Center, 18057 Rostock, Germany;
florian.klomp@uni-rostock.de
2 Department of Pharmacology, Center of Drug Absorption and Transport, University Medicine Greifswald,
17487 Greifswald, Germany; Christoph.Wenzel@med.uni-greifswald.de
3 Department of Experimental and Clinical Pharmacology, Pomeranian Medical University,
70-111 Szczecin, Poland; marek.drozdzik@pum.edu.pl
* Correspondence: stefan.oswald@med.uni-rostock.de; Tel.: +49-381-494-5894

Received: 9 November 2020; Accepted: 8 December 2020; Published: 11 December 2020

Abstract: Cytochrome P450 (CYP) 1A enzymes are considerably expressed in the human intestine
and liver and involved in the biotransformation of about 10% of marketed drugs. Despite this
doubtless clinical relevance, CYP1A1 and CYP1A2 are still somewhat underestimated in terms of
unwanted side effects and drug–drug interactions of their respective substrates. In contrast to this,
many frequently prescribed drugs that are subjected to extensive CYP1A-mediated metabolism
show a narrow therapeutic index and serious adverse drug reactions. Consequently, those drugs are
vulnerable to any kind of inhibition or induction in the expression and function of CYP1A. However,
available in vitro data are not necessarily predictive for the occurrence of clinically relevant drug–drug
interactions. Thus, this review aims to provide an up-to-date summary on the expression, regulation,
function, and drug–drug interactions of CYP1A enzymes in humans.

Keywords: cytochrome P450; CYP1A1; CYP1A2; drug–drug interaction; expression; metabolism;

regulation

1. Introduction
The oral bioavailability of many drugs is determined by first-pass metabolism taking place in
human gut and liver. In this regard, a considerable fraction of the administered dose is presystemically
eliminated by intestinal and hepatic phase I and/or phase II drug metabolism Consequently, only a
minor fraction of the administered dose reaches the central compartment and in turn the site of action.
Thus, alterations of the aforementioned presystemic metabolism in terms of inhibition or induction
of the involved metabolizing enzymes may result in two unwanted clinical scenarios: (1) increased
drug exposure as caused by enzyme inhibition with an increased risk of side effects up to drug-related
toxicity, and (2) subtherapeutic drug levels due to induction of the respective metabolizing enzymes,
which may threaten the therapeutic drug effects [1,2].
During the last decades, it was clearly demonstrated that major cytochrome P450 (CYP) enzymes
such as CYP3A4, CYP2C9/19, and CYP2D6 play a major role in first-pass metabolism of drugs [3–5].
Here, extensive pharmacokinetic and pharmacogenetic studies have been conducted and identified
these enzymes as crucial determinants in the pharmacokinetics and, in turn, for efficacy and safety of
their substrates [6–9]. However, beside these major enzymes, the information about other CYPs that
are considerably expressed in the human intestine or liver and significantly involved in the metabolism
of frequently used drugs is much more limited. Examples for these somewhat “under-investigated”
enzymes are CYP1A1 and CYP1A2, which are involved in the metabolism of about 10% of the drugs
on the market [10,11]. Despite their clinical relevance, considerably fewer studies related to human

181
Pharmaceutics 2020, 12, 1201; doi:10.3390/pharmaceutics12121201 www.mdpi.com/journal/pharmaceutics
Pharmaceutics 2020, 12, 1201

pharmacokinetics and drug–drug interactions compared to the above-mentioned major enzymes have
so far been published. For example, Medline search (via PubMed® ) on “human pharmacokinetics” and
certain enzymes listed 6379 entries for CYP3A4, 2902 for CYP2C9/19, and 2794 for CYP2D6, but “only”
734 and 1749 have been found for CYP1A1 and CYP1A2, respectively (assessed 22 October, 2020). Thus,
the aim of this mini review article is to provide an up-to-date overview about the current knowledge
on the expression, regulation, and clinically relevant drug–drug interactions of CYP1A1 and CYP1A2
in humans.

2. Expression
CYP1A1 and CYP1A2 belong to the CYP1 gene family; they are highly conserved and located
on chromosome 15 [10,12]. CYP1A2 is constitutively expressed at high levels in human liver,
whereas CYP1A1 was shown to be expressed at markedly lower levels in the organ but is also found
in extrahepatic tissues including lung, intestine, prostate, kidney and placenta [11,13–19]. However,
data on the protein abundance of hepatic CYP1A1 are conflicting; some studies demonstrated its
absence in human liver, while others could clearly quantify the enzymatic protein [14,16,18,20–23].
In particular, recent mass spectrometry-based studies have demonstrated a substantial abundance of
CYP1A1 in the human liver ranging from 0.5 to 9 pmol/mg microsomal protein [18,23].
In general, the available data on CYP1A1/2 expression are somewhat limited. There are sporadic
studies on intestinal or hepatic CYP1A1 and/or CYP1A2 gene expression or protein abundance that
are partly inconsistent and conflicting (Table 1). To overcome this limitation counteracting reliable
conclusions, especially on the role of intestinal and hepatic CYP1A1, comprehensive information about
gene and protein expression of CYP1A1 and CYP1A2 in intestinal and hepatic tissues from the same
individuals are needed. So far, only one study from our group is available that included intestinal
and hepatic tissue samples in parallel from organ donors, in order to overcome the known issue of
high inter-subject variability in the expression of metabolizing enzymes [24]. However, this study only
considered CYP1A2, but not CYP1A1.

Table 1. Overview of available data on mRNA expression and protein abundance of cytochrome
P450 (CYP) 1A1 and CYP1A2 in the human intestine and liver (+, gene/protein expression was
shown; -, not investigated n.d., not detectable; PTC, proteomics; WB, Western blot). Data are ranked in
chronological order (publication date).

Liver Small Intestine

CYP1A1 CYP1A2 CYP1A1 CYP1A2
Protein Protein Protein Protein
Gene Gene Gene Gene Reference
(Method) (Method) (Method) (Method)
n.d.
- - + (WB) - - - - [20]
(WB)
+ - + + (WB) - - - - [13]
- + (WB) - - - - - - [25]
- - - - + + (WB) n.d. - [26]
- - - - - + (WB) - - [27]
+ - + - + - n.d. - [15]
n.d.
+ + + (WB) - - - - [21]
(WB)
- + (WB) - + (WB) - - - - [14]
- + (WB) - - - + (WB) - - [16]
+ - + - + - n.d. - [17]
- - + + (PTC) - - - - [28]
- + (PTCs) - + (PTC) - - - - [23]
- - - + (PTC) - - - n.d. (PTC) [29]
- - - - - + (PTC) - +, traces (PTC) [30]
- - + + (PTC) - - n.d. n.d. (PTC) [24]
- - - + (PTC) - - - - [31]
- - - - + - n.d. - [32]
n.d. n.d.
- + (PTC) - - n.d. (PTC) [22]
(PTC) (PTC)
- + (PTC) - + (PTC) - - - - [18]

182
Pharmaceutics 2020, 12, 1201

Compared to CYP3A4 and CYP2C9, which are the most abundant intestinal CYP enzymes [16,24],
CYP1A1 expression is rather little in the intestine and highly variable as well [16,30] (Figure 1).
Paine et al. were able to detect CYP1A1 enzyme in only three of 31 investigated human jejunal
samples at a range of 3.6 to 7.7 pmol/mg (Western blotting), while Miyauchi et al. using the
targeted proteomics approach found CYP1A1 in 15 out of 28 analyzed human small intestinal
samples (range: 0.07–2.3 pmol/mg) [16,30]. In parallel, also Shrivas et al. found CYP1A1 only
in three out of 32 human liver microsomal samples using global proteomics [23]. Those findings
indicate a substantial inter-subject variability in CYP1A1 protein that is most likely attributed to the
individual lifestyle, including exposure to different diets or smoking, which were already shown
to be important determinants of variability in the expression as well as the metabolic function of
CYP1A enzymes [13,21,27,33]. In addition, experimental conditions may have partly contributed to
the observed variability.

   
 

 

 

 
& $

& $


& 

& (
<3 
& &

<3 

& '

& $

& $


& 

& (
& %

& &

<3 

& '
$
%

& &

$
& &
&





&







<3

<3

<3
<3

<3

<3

<3
<3

<3

<3

<3
<3

<3

<3

<3
&

&

&
SURWHLQDEXQGDQFHSPROPJ

 
UHODWLYHJHQHH[SUHVVLRQ









$

$

$

$
&

'

&

'
&

&
<3

<3

<3

<3
<3

<3

<3

<3
<3

<3
&

&

&

&
&

&

&

&
&

&

Figure 1. Comparative intestinal and hepatic gene expression and protein abundance of CYP1A1 and
CYP1A2 (mean ± SD). Relative gene expression and absolute protein abundance of clinically relevant
CYP450 enzymes in the human liver is presented in section (A,B) (data taken from Drozdzik et al. 2019
(A) and Achour et al. 2014 (B)). Sections C and D show relative gene expression and absolute protein
abundance of clinically relevant CYP450 enzymes in the human small intestine observed in 30 ((C),
own unpublished data) and 28 human jejunal tissue samples ((D), Miyauchi et al. 2016).

In contrast to the mentioned controversies on CYP1A1, CYP1A2 is doubtlessly expressed at

considerable levels in the human liver, but not in the intestine (Table 1). A meta-analysis on the
expression of 15 hepatic drug-metabolizing enzymes revealed that CYP1A2 contributes about 10%
to all CYP enzymes [34]. After CYP3A4, CYP2E1, and CYP2C9, CYP1A2 showed the fourth highest
protein abundance of all investigated hepatic CYP enzymes. These data have been confirmed by
more recent mass spectrometry-based studies [24,31,35–37]. The mean hepatic microsomal protein
abundance ranged from 14 to 35 pmol/mg, as determined by targeted proteomics [18,22,28,31,36,38].
Former data as derived from immunostaining studies determined considerably higher abundance,
up to 65 pmol/mg [39,40], which might be interpreted with caution considering the issue of unspecific
binding of antibodies. As a general feature of CYP1A1 and CYP1A2 expression, one must consider
a substantial inter-subject variability, which seems to be caused by a complex interplay of genetic,
epigenetics, and environmental factors [11,41].

183
Pharmaceutics 2020, 12, 1201

3. Regulation

3.1. Transcriptional Regulation

The promoter region of CYP1A1 and CYP1A2 genes contains several Aryl hydrocarbon receptor
(AhR) response elements [42,43], which, after binding of respective compounds, initiate coordinated
transcription of both genes. Consequently, both genes are highly inducible by AhR ligands
such as polycyclic aromatic hydrocarbons (PAHs), dioxins and numerous xenobiotics [44].
Examples for well-established experimental inducers are 2,3,7,8-tetrachlorodibenzo-p-dioxon (TCDD),
3-methlycholanthrene and β-naphthoflavone [45,46]. In addition, omeprazole was shown to be a
potent model inducer for CYP1A1 and CYP1A2 [47–49]. There is evidence that the induction of
CYP1A1 by AhR is stronger than that of CYP1A2 [44,49,50]. This is also in agreement with the fact
that tissues possessing high expression of CYP1A1 (i.e., lung, placenta, intestine, urinary bladder)
also show high expression levels of AhR [51]. Table 2 provides an overview of in vitro, in vivo,
and ex vivo data on CYP1A1/1A2 induction by clinically relevant drugs. In this regard, albendazole,
carbamazepine, omeprazole, lansoprazole, primaquine, and rosiglitazone were shown to be strong
inducers of both, expression and metabolic activity of CYP1A1 and CYP1A2 [48–50,52]. As discussed
later, these findings do not necessarily translate for all compounds to clinically relevant drug–drug
interactions (e.g., for omeprazole). Relevant exogenous sources of AhR activators are charcoal grilled
food, tobacco smoking as well as other natural sources including broccoli or fish oil supplementation,
which strikingly induce endogenous eicosanoids [11,53,54]. Consequently, some of these environmental
factors may contribute to the substantial variability in the expression and function of both CYP1A
isoenzymes. Associated with this—tobacco smoking, especially, was shown to have significant effects on
the pharmacokinetics and actions of many CYP1A substrates (see paragraph “Drug–Drug Interactions”).

Table 2. Impact of clinically relevant drugs, smoking, and diet on the induction of CYP1A1/1A2 mRNA,
protein, and activity.

Drug Object of Investigation Induction Effect Reference

↑CYP1A1 (32-fold) and CYP1A2
Albendazole (5–30 μM) HepG2 cells mRNA (5.6-fold); ↑EROD-activity [49,50]
(4-fold)
Carbamazepine (7–183 μM) HepaRG cells ↑CYP1A2 mRNA (10-fold) [52]
↑hepatic CYP1A2 activity (CBT,
Carbamazepine Pediatric patients [55]
2.2-fold)
↑CYP1A2 mRNA (26-fold);
Lansoprazole (50 μM) Human hepatocytes ↑CYP1A2 protein (32-fold); [48]
↑EROD activity (32-fold)
Omeprazole (0.03–3 μM) Human hepatocytes ↑CYP1A1 mRNA (37-fold) [47]
↑CYP1A2 mRNA (12-fold);
Omeprazole (50 μM) Human hepatocytes ↑CYP1A2 protein (4.6-fold); [48]
↑EROD activity (39-fold)
Omeprazole (25 μM) HepG2 cells ↑CYP1A1 and CYP1A2 mRNA [49]
Human duodenal ↑CYP1A1 protein; ↑EROD activity
Omeprazole (treatment) [56]
biopsies (2.2-fold)
↑CYP1A2 protein (3.4-fold);
Omeprazole (20 mg SID, 4 d) Human liver biopsies [57]
↑EROD activity (6-fold)
Phenobarbital (100–250 μM) Human hepatocytes ↑EROD activity (1.9-fold) [46]
↑CYP1A2 mRNA (1.5-fold);
Phenobarbital (1 mM) Human hepatocytes ↑CYP1A2 protein (1.8-fold); ↑POD [58]
activity (3.1-fold)
↑CYP1A1 (~7-fold) and CYP1A2
Primaquine (10–30 μM) HepG2 cells (~3-fold) mRNA; ↑EROD-activity [49,50]
(7.5-fold)

184
Pharmaceutics 2020, 12, 1201

Table 2. Cont.

Drug Object of Investigation Induction Effect Reference

↑CYP1A1 (~9-fold) and CYP1A2
Quinine (30 μM) HepG2 cells mRNA (2.4-fold); ↑EROD-activity [50]
(5.5-fold)
↑CYP1A1 (2.2-fold) and CYP1A2
Rifampicin (10/33 μM) Human hepatocytes [47,59–61]
mRNA (2.2-fold)
Rifampicin (20/50 μM) Human hepatocytes ↑EROD-activity (2.3-fold) [46]
±CYP1A2 mRNA (0.8-fold);
Ritonavir (0.1–5 μM) Human hepatocytes ±CYP1A2 protein (1.0-fold); [58]
↑POD activity (1.6-fold)
↑CYP1A2 mRNA (4-fold); ↑POD
Ritonavir (1–25 μM) Human hepatocytes [61]
activity (2-fold)
↑CYP1A2 mRNA (11-fold);
Rosiglitazone (10 μM) Human hepatocytes ↑CYP1A2 protein (7-fold); ↑EROD [48]
activity (37-fold)
Smoking Human liver biopsies ↑EROD activity (3.3-fold) [62]
Human duodenal
Smoking (3–30/d), 7d ↑EROD activity (4.2-fold) [56]
biopsies
Human duodenal ↑CYP1A1 protein, ↑hepatic
Chargrilled meat diet (7 d) [33]
biopsies CYP1A2 activity (CBT, 1.9-fold)
↑, increase; ±, unchanged; CBT, caffeine breath test; EROD, 7-ethoxyresorufin O-deethylase; POD,
phenacetin O-deethylation.

As already described for genes regulated by other nuclear receptors (e.g., ABCB1 by
pregnane-X-receptor (PXR)), a partial transactivation of human CYP1A by nuclear receptors other than
AhR is possible. In this regard, CYP1A1 and 1A2 were also shown to be induced upon activation of the
human constitutive androstane receptor (CAR) [63]. This explains considerable induction of CYP1A
enzymes by typical CAR ligands, such as carbamazepine, phenobarbital and phenytoin. On the other
side, the relevance of PXR in the regulation of CYP1A seems to be negligible as shown in vitro [47,59,60]
and in vivo [64].

3.2. Impact of Gender, Age, and Diseases

In addition to the described transcriptional regulation, also several nongenetic factors seem to
influence CYP1A2 expression and function. For example, protein abundance and metabolic CYP1A2
activity for different substrates was shown to be considerably lower in woman than in men [31,65–67].
However, as smoking and the intake of oral contraceptives (inhibitors of CYP1A function) represent
substantial confounders of CYP1A2 expression and function, those data have to be interpreted with
caution and need further verification. Moreover, in analyzing potential gender differences in the
pharmacokinetics of CYP1A substrates, dose-adjustment was shown to be essential as demonstrated
for tizanidine [68]. There is also evidence that CYP1A2 activity is significantly higher in younger
(<20 years) than in older people (>20–60 years and >60 years) [65].
Disease-related changes have been also reported for CYP1A. Here, CYP1A2 expression in liver
dysfunction and cholestasis was found to be decreased [40,69]. Other studies failed to confirm
those differences in vivo [53,54] and in human liver tissue at both, mRNA and protein levels [40,70].
More recent mRNA expression data demonstrated that the expression of CYP1A2 was decreased by
about 90% in hepatocellular carcinoma livers, 80% in alcoholic cirrhosis, and 65% in severe cirrhosis [71].
In parallel, analysis of liver biopsy samples of patients with chronic hepatitis C revealed significantly
lower gene expression levels of CYP1A1 and CYP1A2 [72,73]. These data have been recently confirmed
by a targeted proteomic analysis [74]. Likewise, nonalcoholic fatty liver disease (NAFLD) was
associated with decreased mRNA, protein amount, and functional activity of microsomal CYP1A2
compared to healthy liver tissue [75].

185
Pharmaceutics 2020, 12, 1201

3.3. Genetics and Epigenetics

The large inter-individual variability in the elimination of drugs undergoing CYP1A2 metabolism
has been attributed to genetic and environmental factors [11,76,77]. In this regard, Rasmussen and
colleagues demonstrated in a large study in 378 mono- and dizygotic twins for the caffeine metabolic
ratio (a surrogate for CYP1A2 activity) a strong overall heritability of 0.72 [78].
Common polymorphisms in the CYP1 gene have been found to be only of limited relevance for
human drug metabolism. However, considering the involvement of CYP1A enzymes in bioactivation
of procarcinogens, many studies investigated certain single nucleotide polymorphisms in association
to various types of cancer [10,11]. The Pharmacogene Variation Consortium website (www.pharmvar.
org) lists 15 alleles for CYP1A1 [79]. Of the most frequent variants m1 to m4, only the common
non-synonymous variant CYP1A1*2C (rs1048943, 2454A>G, Ile462Val), which has a global minor allele
frequency of about 12%, was shown to be associated with substantially modified enzymatic activity,
i.e., 6- to 12-fold higher for its substrates 17β-estradiol and estrone [80]. This variant was associated
with an increased risk for lung cancer in Chinese and breast, and prostate cancer in Caucasians [10,11].
Several alleles, namely 24, have been also reported for CYP1A2 [79], of which only the most
established will be briefly mentioned here. The CYP1A2*6 variant was shown to result in a nonfunctional
protein [81]. However, considering the rare occurrence of this and other variants [82], they are expected
to be of limited clinical relevance. The CYP1A2*1C was associated with reduced CYP1A2 induction by
cigarette smoking in Japanese [83]. On the contrary, the CYP1A2*1F variant (–163C>A) was linked
with enhanced enzyme inducibility in Caucasian smokers [53,84] and heavy coffee drinkers [85].
Interestingly, carriers of the combined genotype CYP1A2*1C/*1F were not inducible by the AhR
ligand omeprazole [86]. Both variants were described to increase the susceptibility to certain cancers.
Despite the described multiplicity of CYP1A2 polymorphisms, clear gene dose relationships by
comparing common SNPs to the respective protein abundance or metabolic phenotype could not
be demonstrated yet. Thus, so far no single SNP or haplotype in the CYP1A2 gene seems to be
predictive [41]. In this regard, a multivariate linear modeling by Klein et al. revealed that genetic
polymorphisms contribute about 35% of hepatic CYP1A2 activity variation, whereas some 40% of the
variation were explained by nongenetic factors together [40].
However, the clinical impact of genetic variation in terms of susceptibility factors for cancer or
pharmacokinetics, efficacy and safety of certain CYP1A substrates is not systematically covered here
but was excellently summarized by others [10,11,87–90].
Finally, there is also evidence for an epigenetic regulation of CYP1A2 expression as concluded
from the observation that the extent of DNA methylation of a CpG island close to the translation start
site was inversely correlated to the hepatic CYP1A2 mRNA expression [53,54]. Recent studies point
also to an involvement of certain microRNAs in the expression and induction of CYP1A2 [91,92].

4. Metabolic Function, Substrates, and Inhibitors

4.1. Metabolic Features

Considering that CYP1A2 shares about 80% amino acid sequence identity with CYP1A1, it is
not surprising that the substrate specificities of these enzymes often overlap, owing to a CYP1
family–specific distortion of the F helix in the area of the substrate binding cavity, which produces
bending of the helix and results in the formation of an enclosed and planar substrate binding site
observed in both CYP1A1 and CYP1A2 [93]. It has been demonstrated that commonly used probe
drugs for CYP1A2 such as caffeine, theophylline, phenacetin, propranolol, and 7-ethoxyresorufin are
metabolized by both CYP1A isoenzymes [94,95]. Despite this considerable similarity, CYP1A1 shows a
preference for planar aromatic hydrocarbons (e.g., naphthalene, PAHs), while CYP1A2 prefers aromatic
amines and heterocyclic compounds (e.g., 2-naphthylamine, xanthines) (Table 3). The metabolic feature
of CYP1A1 in combination with its expression pattern in tissues potentially exposed to high amounts
of PAHs (e.g., the lung via tobacco smoke, the intestine via charbroiled food) makes it plausible that

186
Pharmaceutics 2020, 12, 1201

increased expression and function of CYP1A1 may result in higher formation rates of potentially
carcinogenic metabolites. In this regard, benzo[a]pyrene and other procarcinogens (e.g., arylarenes,
nitroarenes, arylamines) are bioactivated by CYP1A1 to reactive and carcinogenic intermediates such
as epoxides which may cause DNA damage and in long term malignancies. In the same manner,
CYP1A2 is involved in the bioactivation of heterocyclic aromatic amines (HAAs) originating from cook
muscle meats such as beef, pork, or fish to carcinogenic hydroxylamines. Thus, it can be assumed that
induction of CYP1A1/1A2 in smokers by inhaling frequently high amounts of PAHs may contribute to
strikingly increased risk for lung cancer [96]. However, the toxicological impact of both isoenzymes on
the bioactivation of carinogenes from environmental compounds is beyond the scope of this article but
summarized elsewhere [11,97].

Table 3. Overview for clinically relevant drugs undergoing significant CYP1A2-mediated metabolism (≥25%).

Contribution of
Substrate Drug Class Metabolic Reaction Reference
CYP1A2 (Other CYPs)
Aminopyrine Analgesic drug N-demethylation 40–50% (CYP2C8/2C19) [98]
melatonin receptor agonist hydroxylation and
Agomelatine 90% (10% CYP2C9) [99]
(antidepressant) demethylation
Caffeine CNS stimulant N-demethylation >95% [94,95]
N-demethylation and
Clozapine Atypical antipsychotic drug 40–55% (CYP3A4/2C19) [100]
N-oxidation
Dacarbazine Anticancer drug N-demethylation 20–40% (CYP1A1/2E1) [101]
4-, 5- and 6-hydroxylation
Duloxetine Antidepressant 30–40% (CYP2D6/2C9) [102]
major extent substrate
Flutamide Non-steroidal antiandrogen 2-Hydroxylation ~25% (CYP3A4/2C19) [103]
Disease-modifying
Leflunomide N-O bond cleavage 40–55% [104]
anti-inflammatory drug
6-hydroxylation and
Melatonin Pineal hormone 40–60% (CYP1A1/1B1) [105]
O-demethylation
8-hydroxylation and
Mirtazapine Antidepressant 30–50% (CYP3A4/2D6) [106]
N-demethylation
Nabumetone NSAID aliphatic hydroxylation 30–40% (CYP2C9) [107]
N-demethylation and
Olanzapine Atypical antipsychotic drug 30–40% (CYP2D6) [108]
7-hydroxylation
O-deethylation and
Phenacetin Analgesic drug 86% [94]
C-hydroxylation
N-demethylation and
Promazine Antipsychotic drug 30–45% (CYP2C19/3A4) [109]
5-sulfoxidation
N-deisopropylation, and 4-
Propranolol β-Blocker 30–50% (CYP2D6) [110]
and 5-hydroxylation
Melatonin receptor agonist
Ramelteon Aliphatic hydroxylation ~50% (CYP2C19/3A4) [111]
(hypnotic)
N-dealkylation and
Rasagiline Antiparkinson drug >50% [112]
hydroxylation
Antiglutamate agent
Riluzole N-hydroxylation ~80% [113]
(treatment of ALS)
N-depropylation and
Ropinirole Antiparkinson drug 30–45% [114]
hydroxylation (major)
Ropivacaine Local anesthetic drug 3-, 4-hydroxylation 50–65% (CYP3A4) [115]
cholinesterase inhibitor 1-, 2-, 4- and
Tacrine 50–65% [116]
(Alzheimer’s disease) 7-Hydroxylation
Bronchodilator
Theophylline N-demethylation 90–95% [117]
(Asthma/COPD)
Tizanidine Muscle relaxant Hydroxylation 80–95% [118,119]
N-demethylation and
Verpamil Calcium channel blocker 20–30% (CYP2C8/3A4) [120]
N-dealkylation
Selective 5-HT1B/1D N-demethylation and
Zolmitriptan 30–40% [121]
(treatment of migraine) O-demethlyation
5-HT, 5-hydroxy tryptamine; ALS, Amyotrophic lateral sclerosis; CNS, central nervous system; COPD, Chronic
obstructive pulmonary disease; NSAID, non-steroidal anti-inflammatory drug.

187
Pharmaceutics 2020, 12, 1201

4.2. Substrates
Under the consideration that CYP1A1 is markedly lower expressed in the human liver than
CYP1A2 and is also considered to be of extrahepatic relevance, its impact on the metabolic clearance
of drugs was formerly assumed to be negligible [10,11]. In contrast to this conclusion, recent studies
have clearly verified CYP1A1 protein abundance in human intestine and liver, which challenges the
former paradigm of the pharmacokinetically irrelevant CYP1A1 [18,23,30]. Moreover, 15–20 years
ago, several studies convincingly demonstrated high metabolic CYP1A1 activity of intestinal and
hepatic microsomal fractions [14,27,122]. Associated to this, riociguat (guanylate cyclase stimulator
used for the treatment of pulmonary hypertension) and granisetron (5-HT3 receptor antagonist for the
treatment of nausea and vomiting following chemotherapy or radiotherapy) were shown to be highly
and specifically biotransformed by CYP1A1 [18,122]. In addition, the tyrosine kinase inhibitors axitinib,
erlotinib, gefitinib, and ningetinib as well as the toll-like receptor agonist imiquimod and conivaptan
(inhibitor of the antidiuretic hormone) have been reported as substrates of CYP1A1 [18,123,124]. Thus,
one has to conclude that CYP1A1 should be considered as an additional potentially relevant clearance
pathway for some drugs. However, in the past, the metabolic stability of a drug was in most cases
studied by using human liver microsomes or recombinant CYP1A2, but not for both isoforms of
CYP1A, as done in very recent studies [18,123]. Consequently, the individual contribution of CYP1A1
to the metabolism of established CYP1A2 substrates as summarized in Table 3 remains uncertain,
and asks for additional research efforts. However, even today, these kind of head-to-head comparisons
of CYP1A1 and CYP1A2 in drug metabolism are challenging because established manufactures of
life science consumables (e.g., Thermo Fisher Scientific and Corning) do not provide microsomal
preparations of recombinant CYP1A1, but almost exclusively CYP1A2.
It was estimated by analyzing the metabolic pathways of about 250 frequently used
drugs, that CYP1A2 is involved in the biotransformation of about 10% of drugs on the
market [10]. CYP1A2-typical biotransformation reactions include N-demethylation of caffeine to
1,7-dimethylxanthine (paraxanthine), N-demethylation of clozapine, O-deethylation of phenacetin,
and N-demethylation as well as 8-hydroxylation of theophylline. In particular, caffeine and phenacetin
were frequently used as probe compounds in vitro and for phenotype determination in vivo [11,125].
Due to its high abundance in the human liver, CYP1A2 plays an important role in the metabolism of
many clinically important drugs, including antipsychotics (clozapine, olanzapine), antidepressants
(duloxetine, agomelatine, mirtazapine), cardiovascular drugs (propranolol, verapamil), non-steroidal
anti-inflammatory drugs (NSAID) (phenacetin), the Alzheimer’s disease drug tacrine, a cholinesterase
inhibitor, the muscle relaxant tizanidine, antiparkinson drugs (rasagilin, ropinirol), and the
methylxanthines caffeine, and theophylline [10,11]. Over 100 clinically used drugs have been described
to be substrates of CYP1A2 [11]. However, many compounds are subjected to complex metabolism by
several CYP enzymes so that the allover contribution of CYP1A2 is limited (~5–20%) and dominated by
other pathways. Examples for drugs that are frequently and somewhat misleadingly labelled as typical
CYP1A2 substrates are acetaminophen, amitriptyline, bupivacaine, carbamazepine, estradiol, fluvoxamine,
haloperidol, imipramine, lidocaine, mianserin, naproxen, ondansetrone, triamterene, warfarin,
and zolpidem. Although, CYP1A2 contributes to their metabolism, relevant drug-drug interactions
(DDIs) cannot be expected as other metabolic pathways take over in the case of CYP1A2 inhibition.
Thus, Table 3 summarizes only drugs whose systemic clearance is assumed to be >25% dependent
on CYP1A2 metabolism based on the in vitro phenotyping studies and human pharmacokinetic
data, as also suggested by the current Food and Drug Administration (FDA) guidance of drug–drug
interactions (https://www.fda.gov/regulatory-information/search-fda-guidance-documents/vitro-drug-
interaction-studies-cytochrome-p450-enzyme-and-transporter-mediated-drug-interactions). Similar to
CYP3A4, CYP1A2 is a rather low affinity but high capacity metabolic enzyme. Thus, only very high
concentrations of respective substrates are able to cause competitive inhibition (e.g., by extremely high
doses of caffeine).

188
Pharmaceutics 2020, 12, 1201

Endogenous substrates of CYP1A include arachidonic acid, bilirubin, prostaglandins, estrogens,

melatonin and retinoic acid [11,126].

4.3. Inhibitors
Established inhibitors of CYP1A function include 7-hydroxyflavone and α-naphthoflavone that
have been extensively used in vitro [11,14,18,94]. Ketoconazole, a potent inhibitor of CYP3A4 and
P-glycoprotein, was also shown to inhibit CYP1A1 in a significant manner [27]. Again, we must
state that there are so far insufficient data on the specific inhibitory properties of established CYP1A2
inhibitors on the function of CYP1A1. Considering the similarity in terms of sequence and function,
one may hypothesize again a substantial overlap between both isoenzymes. An exception of this
conclusion is furafylline, a methylxanthine, which was demonstrated to inhibit specifically CYP1A2
but not CYP1A1 [94]. Thus, it serves as an in vitro tool to distinguish between the metabolic activites
of both isoenzymes in microsomal studies.
Typical inhibitors of CYP1A2 are rather small molecules, which are often heterocyclic or
halogenated. Drugs resulting in potent competitive but reversible inhibition of CYP1A2 include
fluoroquinolones such as ciprofloxacin and enoxacin, selective serotonin reuptake inhibiting (SSRI)
antidepressants fluvoxamine and fluoxetin, the azole antimycotics ketoconazole, and clotrimazole,
as well as estrogens (oral contraceptives). Some drugs (e.g., amiodarone, carbamazepine, duloxetine,
isoniazid, resveratrol, and rofecoxib) were described to be mechanism-based inhibitors [11], i.e.,
they cause irreversible inhibition of CYP1A enzymes, which requires de novo synthesis of the
respective proteins, which, in turn, results in long-lasting enzyme inhibition. Table 4 provides an
overview of clinically used drugs that were identified as potent inhibitors of CYP1A. As it can be
seen from the given inhibitory potential of each compound as assessed in vitro (Ki or IC50 values),
compounds with high inhibition potency, such as ciprofloxacin (Ki, 144 nM for CYP1A2), fluxoxamine
(Ki, 11-40 nM for CYP1A2), or ketoconazole (Ki, 40 nM for CYP1A1) are especially expected to cause
clinically relevant drug–drug interactions [27,127,128].

Table 4. Overview of clinically relevant drugs with inhibitory properties on CYP1A1/1A2.

Inhibitory Effect
Drug Drug Class In Vitro System Reference
(Isoenzyme)
5HT3 -receptor
Alosetron 1 antagonist (irritable HLM IC50 = 2 μM (CYP1A2) [129]
bowel syndrome)
Amiodarone 1 Antiarrhythmic drug HLM IC50 = 86 μM [130]
Artemesinin 1 Antimalaria drug HLM Ki = 0.43 μM (CYP1A2) [131]
Carbamazepine
2 Anticonvulsant HLM n.d. (CYP1A2) [132]
Celecoxib 1 COX-2 inhibitor HLM Ki = 25.4 μM (CYP1A2) [127]
Antibiotic HLM 70.4% (CYP1A2) [133]
Ciprofloxacin 1
(fluoroquinolone) HLM Ki = 144 nM (CYP1A2) [127]
Cimetidine 1 H2 -receptor antagonist HLM Ki = 200 μM (CYP1A2) [134]
human
Clotrimazole 1 Antifungal agent Ki = 7.9 μM (CYP1A2) [135]
lymphoblast cells
Hormone
Desogestrel 1 HLM Ki = 39.4 μM (CYP1A2) [127]
(oral contraceptive)
Duloxetine 2 Antidepressant (SSRI) HLM n.d. (CYP1A2) [136]
1 Antibiotic
Enoxacin HLM 74.9% (CYP1A2) [133]
(fluoroquinolone)
Ethinyl Hormone
HLM Ki = 10.6 μM (CYP1A2) [127]
estradiol 1 (oral contraceptive)
Fluoxetine 1 Antidepressant (SSRI) HLM Ki = 4.4 μM (CYP1A2) [137]
Ki = 33 μM (CYP1A1) [128]
Fluvoxamine 1 Antidepressant (SSRI) HLM Ki = 40 nM (CYP1A2) [128]
Ki = 11 nM (CYP1A2) [127]

189
Pharmaceutics 2020, 12, 1201

Table 4. Cont.

Inhibitory Effect
Drug Drug Class In Vitro System Reference
(Isoenzyme)
Isoniazid 2 Antibiotic HLM Ki = 285 μM (CYP1A2) [138]
1 HIM Ki = 40 nM (CYP1A1) [27]
Ketoconazole Antifungal agent
HLM IC50 = 0.33 μM CYP1A2) [139]
human
Miconazole 1 Antifungal agent Ki = 3.2 μM (CYP1A2) [135]
lymphoblast cells
n.d. (CYP1A1)
Nifedipine 1 Calcium channel blocker HLM [94]
n.d. (CYP1A2)
Antibiotic
Norfloxacin 1 HLM 55.7% (CYP1A2) [133]
(fluoroquinolone)
Paroxetine 1 Antidepressant (SSRI) HLM Ki = 5.5 μM (CYP1A2) [137]
Propafenone 1 Antiarrhythmic drug HLM IC50 = 29 μM [130]
n.d. (CYP1A1)
Propranolol Beta blocker HLM [94]
n.d. (CYP1A2)
IC50 = 23 μM (CYP1A1) [140]
Resveratrol 2 Natural compound HLM
Ki = 2.2 μM (CYP1A2) [141]
Amyotrophic lateral
Riluzole 1 HLM Ki = 12.1 μM (CYP1A2) [113]
sclerosis drug
Rofecoxib 2 COX-2 inhibitor HLM Ki = 6.2 μM (CYP1A2) [127]
Sertraline 1 Antidepressant (SSRI) HLM Ki = 8.8 μM (CYP1A2) [137]
human
Sulconazole Antifungal agent Ki = 0.4 μM (CYP1A2) [135]
lymphoblast cells
Antifungal/antiparasitic
Thiabendazol 2 HLM Ki = 1.54 μM (CYP1A2) [131]
agent
human
Tioconazole 1 Antifungal agent Ki = 0.4 μM (CYP1A2) [135]
lymphoblast cells
Tolfenamic acid
1 NSAID HLM Ki= 1.4 μM (CYP1A2) [127]
1Competitive (reversible) inhibitor; 2 mechanism-based (irreversible) inhibitor; 5-HT, 5-hydroxy tryptamine; COX,
cyclooxygenase; HLM, human liver microsomes; IC50 , half maximal inhibitory concentration; Ki, inhibition constant;
NSAID, nonsteroidal anti-inflammatory drug; SSRI, selective serotonin reuptake inhibitor.

5. Drug–Drug Interactions
Under consideration of the high number of frequently prescribed drugs that were described to be
substrates (Table 3), inhibitors (Table 4), or inducers (Table 2) of human CYP1A1/A2, several unwanted
drug–drug interactions can be assumed in the case of combined administration.

5.1. Inhibition Studies

In this regard, the most pronounced interactions have been described for the combination of CYP1A
substrates with potent inhibitors including ciprofloxacin, fluvoxamine, ethinyl estradiol, and rofecoxib.
Their combination with established CYP1A substrates resulted in clinically relevant interactions
increasing the systemic drug exposure of caffeine, clozapine, mirtazapine, olanzapine and theophylline
by 1.5 to 3-fold [106,131,142–144]. For agomelatine, ramelteon, tracrine, and tizanidine much more
dramatic increases of serum area under the concentration-time curve (AUC) by 10–190-fold have
been observed [118,119,145,146], which is expected to cause drug-related side effects and even toxicity.
For example, the elevation of plasma levels of clozapine by ciprofloxacine resulted in rhabdomyolysis,
delirium, and death during combination in psychotic patients [147,148].
The reasons for these dramatic interactions might be due to extensive metabolism by CYP1A
enzymes and/or a high volume of distribution of the victim drug (e.g., 168 and 349 l for tizanidine and
tacrine). In order to estimate the in vivo potential of a certain CYP1A inhibitor (Table 4) of an in vitro
function, to cause clinically relevant interactions, focusing on the observed inhibitory potential (Ki,
IC50 value) alone is not sufficient, but additional pharmacokinetic aspect of the perpetrator compounds
must be considered as well. For sufficient inhibitory potential in vivo, a perpetrator drug needs to

190
Pharmaceutics 2020, 12, 1201

generate free unbound concentrations (fraction unbound, fu) around or above the observed Ki/IC50
value and needs to be present in the systemic circulation for several hours to cause substantial metabolic
inhibition as determined by an elimination half-life of several hours. Consequently, fluvoxamine and
ciprofloxacin that are characterized by rather low-to-medium protein binding (fu, 0.23 for fluvoxamine
and fu, 0.8 for ciprofloxacin), but high serum levels as caused by their comparatively high administered
doses (50–100 mg for fluvoxamine, 100–750 mg for ciprofloxacin) and medium to long terminal
half-lives (4–7 h for ciprofloxacin, 17–22 h for fluvoxamine), cause that both drugs are strong inhibitors
of CYP1A2 in vivo, and cause many clinically relevant drug–drug interactions.
This scenario is not true for other drugs mentioned in Table 4. For example, although the NSAIDs
celecoxib and tolfenamic acid demonstrated a considerable inhibition of CYP1A2 in human liver
microsomes (HLM) with a Ki values of 25 μM and 1.4 μM [127], they did not show clinically relevant
interactions, most likely due to their high protein binding of ~98% and 99.7%, respectively. As a
conclusion, drugs undergoing substantial CYP1A1/2 metabolism should be combined with caution
together with the perpetrator drugs mentioned in Table 5. If possible, dose escalation combined with
therapeutic drug monitoring should be used for CYP1A2 drugs with a narrow therapeutic index such
as theophylline, clozapine or tizanidine. Whether the mentioned in vivo inhibitors of CYP1A2 may
also cause clinically relevant interactions with CYP1A1 substrates remains uncertain and requires
further studies.

Table 5. Overview of clinically relevant interaction as caused by inhibition of CYP1A1/1A2 enzymes.

Perpetrator
Substrate (Victim Drug) PK Change Reference
(Inhibitor)
Product
Agomelatine Fluvoxamine AUC ↑ 60-fold
information
Caffeine AUC ↑ 1.6-fold
Thiabendazol (500 mg, SD) [131]
(137 mg, SD) t1/2 ↑ 2.4-fold
Clozapine
Fluvoxamine (50–100 mg, SID, MD) CSS ↑ 5-10-fold [149]
(50–700 mg)
Clozapine
Fluvoxamine (50 mg, SID, MD) CSS ↑ 3-fold [150]
(2.5–3.0 mg/kg)
Clozapine
Fluvoxamine (50 mg, SID, MD) CSS ↑ 2.2-fold [142]
(200–350 mg)
Clozapine
Ciprofloxacin (250 mg BID, 7 d) CSS ↑ 1.3-fold [151]
(150–400 mg)
Duloxetine AUC ↑ 5.6-fold
Fluvoxamine (100 mg SID, 16 d) [102]
(60 mg, SD) CMAX ↑ 2.4-fold
Melatonin AUC ↑ 17-fold
Fluvoxamine (50 mg, SD) [105]
(5 mg, SD) CMAX ↑ 12-fold
Mirtazapin
Fluvoxamine (50–100 mg, SID, MD) CSS ↑ 1.3-fold [106]
(15–30 mg)
Olanzapine AUC ↑ 1.5-fold
Fluvoxamine (100 mg, SID, 14 d) [143]
(10 mg, SD) CMAX ↑ 1.6-fold
Propranolol
Fluvoxamine (100 mg) CMAX ↑ 5-fold [152]
(160 mg, SID)
Product
Ramelteon AUC ↑ 190-fold
Fluvoxamine (100 mg BID, 3 d) information
(16 mg, SD) CMAX ↑ 70-fold
[111]
Ropivacaine
Ciprofloxacin (500 mg BID, 2.5 d) CL ↓ 31% [153]
(0.6 mg/kg, iv)
Tacrine AUC ↑ 8.3-fold
Fluvoxamine (100 mg SID, 6 d) [145]
(40 mg, SD) CMAX ↑ 5.6-fold
Theophylline AUC ↑ 2.4-fold
Fluvoxamine (75 mg, SD) [144]
(250 mg, SD) t1/2 ↑ 2.5-fold

191
Pharmaceutics 2020, 12, 1201

Table 5. Cont.

Perpetrator
Substrate (Victim Drug) PK Change Reference
(Inhibitor)
Theophylline CL ↓ 19%
Ciprofloxacin (500 mg BID, 3 d) [154]
(3.4 mg/kg, SD) t1/2 ↑ 26%
Tizanidine AUC ↑ 13.6-fold
Rofecoxib (25 mg SID, 4d) [146]
(4 mg, SD) CMAX ↑ 6.1-fold
Tizanidine AUC ↑ 10-fold
Ciprofloxacin (500 mg BID, 3 d) [118]
(4 mg, SD) CMAX ↑ 7-fold
Tizanidine AUC ↑ 33-fold
Fluvoxamine (100 mg SID, 4d) [119]
(4 mg, SD) CMAX ↑ 12-fold
Tizanidine Ethinyl estradiol 20–30 μg, gestodene AUC ↑ 3.9-fold
[155]
(4 mg, SD) 75 μg CMAX ↑ 3-fold
Ropivacaine
Ciprofloxacin (500 mg BID, 2.5 d) CL ↓ 31% [153]
(0.6 mg/kg, iv)
↑, increase; ↓, decrease; AUC, area under the concentration-time curve; BID, twice daily; CL, clearance; Cmax,
maximum serum concentration; Css, trough serum concentrations at steady-state; d, days; MD, multiple doses; PK,
pharmacokinetic; SID, once daily; SD, single dose; t1/2, elimination half-life.

5.2. Induction Studies

On the other side, carbamazepine, lansoprazole, omeprazole, phenobarbital, primaquine,
and rosiglitazone were reported to be potent inducers of CYP1A1/1A2 by binding to AhR or CAR
receptor as briefly described above [44,46–49] (Table 2), while the effects of prototypical PXR activators
such as rifampicin, ritonavir and St. John’s wort are rather negligible [47,58–60,156]. Of these drugs,
omeprazole was one of the most potent and most extensively investigated inducer in vitro and
in vivo, resulting in several-fold induction of the gene expression, protein abundance and metabolic
function of CYP1A1/2. However, significant effects on the pharmacokinetics and efficacy of CYP1A
substrates have not been observed yet. Well-established substrates, including caffeine, phenacetin,
theophylline, or propranolol did not show any changes in their pharmacokinetics in the presence of
omeprazole [157–160]. Thus, one might conclude that the interaction potential of omeprazole and
other proton pump inhibitors for clinically relevant DDIs might be very limited although there are
also data from a case report indicating slight increase in CYP1A2 metabolism [161]. An explanation
could be found in the relatively low peak concentrations of omeprazole (0.7–4.6 μM) in the systemic
circulation compared to the inductive in vitro concentrations (25–50 μM) and its short half-life of
0.5–1 h (Regardh et al. 1990). In contrast to this, treatment with carbamazepine considerably induced
clozapine metabolism, leading to significantly lower serum level in schizophrenic patients [149].
Carbamazepine was furthermore shown to induce hepatic caffeine metabolism as well as the systemic
clearance of olanzapine and mirtazapine in a significant manner [55,162,163]. Thus, it can be stated
that CYP1A2 substrates should not be combined with carbamazepine or dose-adjustment should be
taken into account.
However, estimations on potential drug interactions using in vitro data on induction properties
alone can be misleading. An example for that phenomenon is ritonavir, a HIV protease inhibitor.
Although it showed no (or only weak) induction of CYP1A2 mRNA and activity in human
hepatocytes [58,61], the pharmacokinetics of caffeine and olanzapine was significantly affected,
i.e., AUC was reduced by 75% and 53% [164,165]. To overcome decreased drug efficacy due to the
considerable changes in the pharmacokinetics, Jacobs et al. (2014) proposed that doubling the dose
of olanzapine as a successful strategy in the case of co-medication with ritonavir [166]. The same
disconnection between in vitro and in vivo effects could be observed for rifampicin, which has not
been shown to be an AhR ligand and demonstrated also only a weak induction of CYP1A2 expression
and metabolic function in human hepatocytes [46,47,59,61]. Accordingly, rifampicin premedication
for 5–15 days reduced serum AUC of caffeine and tizanidine by 50–60% [64,164]. The reasons for this
surprising finding might rely on nuclear receptor cross-talk or insensitivity of the respective in vitro on

192
Pharmaceutics 2020, 12, 1201

nuclear receptor activation [167]. Table 6 provides an overview about clinically relevant interactions of
CYP1A substrates caused by enzyme induction.

Table 6. Overview of clinically relevant interaction as caused by induction of CYP1A1/1A2 enzymes.

Substrate (Victim Drug) Perpetrator PK Change Reference

Antipyrine
Smoking CL ↑ 46% [62]
(20 mg/kg, SD)
Caffeine Carbamazepine
CL ↑ 27–47% [163]
(100 mg, SD) (400 mg, SID, 14 d)
Caffeine Lopinavir (400 mg)/Ritonavir (100
MR ↑ 43% [168]
(2 mg/mg, SD) mg), BID, 14 d
Caffeine AUC ↓ 60%
Rifampicin (400 mg BID, 14 d) [164]
(200 mg, SD) CL ↑ 214%
Caffeine AUC ↓ 75%
Ritonavir (400 mg BID, 14 d) [164]
(200 mg, SD) CL ↑ 290%
Clozapine CSS ↓ 50% [169]
Smoking (7–>20/d)
(150–900 mg) CSS ↓ 40% [170]
Clozapine Omeprazole
CSS ↓ 42–45% [161]
(325 mg) (40–60 mg, MD)
Clozapine Carbamazepine CSS ↓ 50% [149]
Duloxetine
Smoking CSS ↓ 53% [171]
(86–102 mg, MD)
Estradiol
Smoking CSS ↓ 43% [172]
(1–2 mg)
Mirtazapine Carbamazepine AUC ↓ 63%
[173]
(15–30 mg, SID, 7 d) (200-400 mg, BID, 21 d) Cmax ↓ 44%
Mirtazapine
Smoking CSS ↓ 41% [174]
(30 mg, SID, 28 d)
CL ↑ 46%
Olanzapine Carbamazepine [162]
t1/2 ↓ 20%
Olanzapine AUC ↓ 48%
Rifampicin, 600 mg, SID, 7 d [175]
(10 mg, SD) Cmax ↓ 11%
Ritonavir AUC ↓ 53%
Olanzapine [165]
(300–500 mg BID, 3–5 d) CMAX ↓ 40%
Smoking (light, 1–4/d) AUC ↓ 45% [176]
Olanzapine Smoking (medium, >5) AUC ↓ 68% [176]
Smoking (heavy, 7–>20) CSS ↓ 67% [169]
CL ↑ 58–100%
Theophylline Smoking [177]
t1/2 ↓ 63%
Tizanidine Rifampicin, 500 mg, SID, 5 d AUC ↓ 53% [64]
Tizanidine Smoking AUC ↓ 33% [68]
Verapamil Smoking AUC ↓ 20% [178]
↑, increase; ↓, decrease; AUC, area under the concentration-time curve; BID, twice daily; CL, clearance; Cmax ,
maximum serum concentration; Css, trough serum concentrations at steady-state; d, days; MD, multiple doses; MR,
metabolic ratio; PK, pharmacokinetic; SID, once daily; SD, single dose; t1/2 , elimination half-life.

Although it was shown in vitro experiments that CYP1A1 can be inhibited and induced by several
compounds (Tables 2 and 4), there are to the best of our knowledge no clinical drug–drug interactions
that can be attributed by specific CYP1A1 inhibition or induction. However, considering the overlap in
substrate recognition, inhibitors, and inducers one might speculate similar interactions as described for
CYP1A2 substrates (Tables 5 and 6). Accordingly, relevant DDIs have been estimated for CYP1A1 [179].
Nevertheless, given the low expression levels of CYP1A1 in human intestine and liver (if even),
the extent of these interactions is expected to be much lower than those caused by inhibition or
induction of hepatic CYP1A2.

193
Pharmaceutics 2020, 12, 1201

5.3. Impact of Smoking and Diet

Finally, smoking can have a profound effect on the pharmacokinetics and efficacy of several
CYP1A1/2 substrates, which is comparable to potent inducing drugs, as summarized in Table 6. In all
summarized examples, the systemic drug exposure of CYP1A substrates was significantly decreased
in smokers compared to nonsmokers by 30–70% (Figure 2). Thus, smokers require higher doses than
nonsmokers. A questionable benefit might be that smokers show also less adverse drug reactions than
nonsmokers [177,180]. However, this is so far only well established in neuropsychopharmacology,
i.e., treatment with antipsychotics and antidepressants. Here, individual dose adjustment is routinely
performed in dependence on therapeutic drug monitoring [181].

^ĐĞŶĂƌŝŽϭ͗EŽŶƐŵŽŬĞƌ ^ĐĞŶĂƌŝŽϮ͗ ^ŵŽŬĞƌ

dŽdžŝĐ ƌĂŶŐĞ
ƉůĂƐŵĂ ĐŽŶĐĞŶƚƌĂƚŝŽŶ

dŚĞƌĂƉĞƵƚŝĐ
ƌĂŶŐĞ

^ƚĂƌƚƐŵŽŬŝŶŐ ^ƚŽƉ ƐŵŽŬŝŶŐ ;LzWϭϭͬϮͿ

^ƚĂƌƚŽĨ ;KzWϭϭͬϮͿ ^ƚĂƌƚŽĨ ;ŚŝŐŚĞƌĚŽƐĞƐƌĞƋƵŝƌĞĚƚŽƌĞĂĐŚ
ƚŚĞƌĂƉLJ ƚŚĞƌĂƉLJ ƚŚĞƌĂƉĞƵƚŝĐƌĂŶŐĞ͕ŚŝŐŚzWϭϭͬϮͿ

ƚŝŵĞ;ĚͿ
Figure 2. Schematic overview of potential interaction scenarios of tobacco smoking. In scenario 1,
a nonsmoker reaches the steady state conditions of a certain CYP1A1/2 substrate after 5–6 half-lives
using standard doses. After start smoking, CYP1A enzymes are significantly induced in intestine
and liver resulting in increased drug clearance and decreasing plasma levels of the respective drug.
In scenario 2, a smoker, who has already substantially higher expression and metabolic activity of
CYP1A1/2, requires significantly higher doses to reach the therapeutic range. After stopping smoking,
CYP1A1/2 will gradually return to the native expression levels, while the daily drug doses are not
adjusted, which results in markedly increased and potentially toxic plasma concentrations.

It was shown that tobacco consumption induces CYP1A2 activity in a dose-dependent manner;
smoking of daily 1–5, 6–10 and >10 cigarettes increases CYP1A2 activity 1.2-, 1.5- and 1.7-fold [182].
The maximum induction effect is already reached after smoking about 10 cigarettes daily, which abates
after about three days of stopping smoking [182,183]. In particular, the latter effect may cause safety
issues in the case of treatment with highly CYP1A-metabolized drugs with serious side effects, such as
clozapine, olanzapine, tacrine, theophylline, or tizanidine. In this case, systemic drug exposure will
substantially increase due to decreasing metabolic capacity, but unchanged high doses associated
with an augmented risk for side effects, or even drug-related toxicity (Figure 2). Associated with this,
cases of agranulocytosis and seizures have been reported for clozapine [181,184]. Because nicotine
alone does not possess any inductive effects on CYP1A, the same risk is true in case of using e-cigarettes
and other ways of nicotine substitution [185]. This should be considered by adjusting the appropriate
dose, especially in case of changes in smoking habit (Figure 2).
Although a chargrilled meat diet was shown to significantly induce intestinal CYP1A1 protein
as well as the metabolic activity of hepatic CYP1A2, as concluded from the caffeine breath test [33],
altered pharmacokinetics of tacrine and caffeine could not be observed in a respective clinical study [186].
Some in vivo findings suggest also a potential in vivo inducing effects of broccoli [187] and another
brassica vegetable, kale [188], on CYP1A2 mediated metabolism of caffeine. The brassica vegetable
CYP1A2 induction is most probably mediated by 3,3 -diindolylmethane (DIM), a condensation product

194
Pharmaceutics 2020, 12, 1201

of indole-3-carbinol being a metabolite of the indole glucosinolate glucobrassicin. DIM has been shown
to induce CYP1A2 in cultured human liver slices [189]. However, there is a lack of information about
brassica vegetables interaction with clinically relevant drugs.

6. Summary and Conclusions

CYP1A1 and CYP1A2 are expressed in human intestine and liver. However, their inter-
subject expression and function is highly variable as most likely caused by genetic, epigenetic,
environmental factors (smoking, diet) and diseases. Considering the high number of drugs that have
been identified as substrates, inhibitors, or inducers of CYP1A enzymes, many clinically relevant
interactions have been reported and can be expected for other substrates. Thus, respective combinations
should be avoided or appropriate dose adjustment is recommended in case of victim drugs with
a narrow therapeutic index. In general, there is a substantial lack of data regarding CYP1A1 and
its distinct role in the pharmacokinetics of drugs. However, from today’s perspective, its allover
contribution to serious drug–drug interactions seems to be limited considering its low expression
levels and the potential functional compensation by CYP1A2. On the other hand, CYP1A2 has to be
considered as one of the big five hepatic drug metabolizing enzymes (along with CYP3A4, CYP2C9,
CYP2C19, CYP2D6, CYP2E1), which is of high clinical relevance in terms of inter-subject variability of
drug efficacy and safety, as well as drug–drug interactions.

Author Contributions: Conceptualization, investigation, writing—original draft preparation, and writing—review

and editing, F.K., C.W., M.D., and S.O. All authors have read and agreed to the published version of the manuscript.
Funding: The research was funded by the German federal ministry of economic affairs and energy (ZIM,
project number: 16KN077223) and by an institutional grants of the Institute of Pharmacology and Toxicology,
Rostock University Medical Center, Rostock, Germany.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Tanaka, E. Clinically important pharmacokinetic drug-drug interactions: Role of cytochrome P450 enzymes.
J. Clin. Pharm. Ther. 1998, 23, 403–416. [CrossRef] [PubMed]
2. Palleria, C.; Di Paolo, A.; Giofrè, C.; Caglioti, C.; Leuzzi, G.; Siniscalchi, A.; de Sarro, G.; Gallelli, L.
Pharmacokinetic drug-drug interaction and their implication in clinical management. J. Res. Med. Sci. 2013,
18, 601–610. [PubMed]
3. Paine, M.F.; Shen, D.D.; Kunze, K.L.; Perkins, J.D.; Marsh, C.L.; McVicar, J.P.; Barr, D.M.; Gillies, B.S.;
Thummel, K.E. First-pass metabolism of midazolam by the human intestine. Clin. Pharmacol. Ther. 1996,
60, 14–24. [CrossRef]
4. Thummel, K.E.; O’Shea, D.; Paine, M.F.; Shen, D.D.; Kunze, K.L.; Perkins, J.D.; Wilkinson, G.R. Oral
first-pass elimination of midazolam involves both gastrointestinal and hepatic CYP3A-mediated metabolism.
Clin. Pharmacol. Ther. 1996, 59, 491–502. [CrossRef]
5. Galetin, A.; Houston, J.B. Intestinal and hepatic metabolic activity of five cytochrome P450 enzymes: Impact
on prediction of first-pass metabolism. J. Pharmacol. Exp. Ther. 2006, 318, 1220–1229. [CrossRef]
6. Eichelbaum, M.; Ingelman-Sundberg, M.; Evans, W.E. Pharmacogenomics and individualized drug therapy.
Annu. Rev. Med. 2006, 57, 119–137. [CrossRef]
7. Evans, W.E.; Relling, M.V. Moving towards individualized medicine with pharmacogenomics. Nature 2004,
429, 464–468. [CrossRef]
8. Dresser, G.K.; Spence, J.D.; Bailey, D.G. Pharmacokinetic-pharmacodynamic consequences and clinical
relevance of cytochrome P450 3A4 inhibition. Clin. Pharmacokinet. 2000, 38, 41–57. [CrossRef]
9. Bahar, M.A.; Setiawan, D.; Hak, E.; Wilffert, B. Pharmacogenetics of drug-drug interaction and
drug-drug-gene interaction: A systematic review on CYP2C9, CYP2C19 and CYP2D6. Pharmacogenomics
2017, 18, 701–739. [CrossRef]
10. Zanger, U.M.; Schwab, M. Cytochrome P450 enzymes in drug metabolism: Regulation of gene expression,
enzyme activities, and impact of genetic variation. Pharmacol. Ther. 2013, 138, 103–141. [CrossRef]

195
Pharmaceutics 2020, 12, 1201

11. Zhou, S.-F.; Wang, B.; Yang, L.-P.; Liu, J.-P. Structure, function, regulation and polymorphism and the clinical
significance of human cytochrome P450 1A2. Drug Metab. Rev. 2010, 42, 268–354. [CrossRef] [PubMed]
12. Murray, G.I.; Melvin, W.T.; Greenlee, W.F.; Burke, M.D. Regulation, function, and tissue-specific expression
of cytochrome P450 CYP1B1. Annu. Rev. Pharmacol. Toxicol. 2001, 41, 297–316. [CrossRef] [PubMed]
13. Schweikl, H.; Taylor, J.A.; Kitareewan, S.; Linko, P.; Nagorney, D.; Goldstein, J.A. Expression of CYP1A1 and
CYP1A2 genes in human liver. Pharmacogenetics 1993, 3, 239–249. [CrossRef] [PubMed]
14. Stiborová, M.; Martínek, V.; Rýdlová, H.; Koblas, T.; Hodek, P. Expression of cytochrome P450 1A1 and its
contribution to oxidation of a potential human carcinogen 1-phenylazo-2-naphthol (Sudan I) in human livers.
Cancer Lett. 2005, 220, 145–154. [CrossRef] [PubMed]
15. Nishimura, M.; Yaguti, H.; Yoshitsugu, H.; Naito, S.; Satoh, T. Tissue distribution of mRNA expression
of human cytochrome P450 isoforms assessed by high-sensitivity real-time reverse transcription PCR.
Yakugaku Zasshi 2003, 123, 369–375. [CrossRef] [PubMed]
16. Paine, M.F.; Hart, H.L.; Ludington, S.S.; Haining, R.L.; Rettie, A.E.; Zeldin, D.C. The human intestinal
cytochrome P450 “pie”. Drug Metab. Dispos. 2006, 34, 880–886. [CrossRef]
17. Bièche, I.; Narjoz, C.; Asselah, T.; Vacher, S.; Marcellin, P.; Lidereau, R.; Beaune, P.; de Waziers, I.
Reverse transcriptase-PCR quantification of mRNA levels from cytochrome (CYP)1, CYP2 and CYP3
families in 22 different human tissues. Pharmacogenet. Genom. 2007, 17, 731–742. [CrossRef]
18. Lang, D.; Radtke, M.; Bairlein, M. Highly Variable Expression of CYP1A1 in Human Liver and
Impact on Pharmacokinetics of Riociguat and Granisetron in Humans. Chem. Res. Toxicol. 2019,
32, 1115–1122. [CrossRef]
19. Ding, X.; Kaminsky, L.S. Human extrahepatic cytochromes P450: Function in xenobiotic metabolism and
tissue-selective chemical toxicity in the respiratory and gastrointestinal tracts. Annu. Rev. Pharmacol. Toxicol.
2003, 43, 149–173. [CrossRef]
20. Murray, B.P.; Edwards, R.J.; Murray, S.; Singleton, A.M.; Davies, D.S.; Boobis, A.R. Human hepatic CYP1A1
and CYP1A2 content, determined with specific anti-peptide antibodies, correlates with the mutagenic
activation of PhIP. Carcinogenesis 1993, 14, 585–592. [CrossRef]
21. Chang, T.K.H.; Chen, J.; Pillay, V.; Ho, J.-Y.; Bandiera, S.M. Real-time polymerase chain reaction analysis of
CYP1B1 gene expression in human liver. Toxicol. Sci. 2003, 71, 11–19. [CrossRef] [PubMed]
22. Grangeon, A.; Clermont, V.; Barama, A.; Gaudette, F.; Turgeon, J.; Michaud, V. Development and validation
of an absolute protein assay for the simultaneous quantification of fourteen CYP450s in human microsomes
by HPLC-MS/MS-based targeted proteomics. J. Pharm. Biomed. Anal. 2019, 173, 96–107. [CrossRef] [PubMed]
23. Shrivas, K.; Mindaye, S.T.; Getie-Kebtie, M.; Alterman, M.A. Mass spectrometry-based proteomic analysis of
human liver cytochrome(s) P450. Toxicol. Appl. Pharmacol. 2013, 267, 125–136. [CrossRef] [PubMed]
24. Drozdzik, M.; Busch, D.; Lapczuk, J.; Müller, J.; Ostrowski, M.; Kurzawski, M.; Oswald, S. Protein Abundance
of Clinically Relevant Drug-Metabolizing Enzymes in the Human Liver and Intestine: A Comparative
Analysis in Paired Tissue Specimens. Clin. Pharmacol. Ther. 2018, 104, 515–524. [CrossRef]
25. Drahushuk, A.T.; McGarrigle, B.P.; Larsen, K.E.; Stegeman, J.J.; Olson, J.R. Detection of CYP1A1 protein
in human liver and induction by TCDD in precision-cut liver slices incubated in dynamic organ culture.
Carcinogenesis 1998, 19, 1361–1368. [CrossRef]
26. Zhang, Q.Y.; Dunbar, D.; Ostrowska, A.; Zeisloft, S.; Yang, J.; Kaminsky, L.S. Characterization of human
small intestinal cytochromes P-450. Drug Metab. Dispos. 1999, 27, 804–809.
27. Paine, M.F.; Schmiedlin-Ren, P.; Watkins, P.B. Cytochrome P-450 1A1 expression in human small bowel:
Interindividual variation and inhibition by ketoconazole. Drug Metab. Dispos. 1999, 27, 360–364.
28. Ohtsuki, S.; Schaefer, O.; Kawakami, H.; Inoue, T.; Liehner, S.; Saito, A.; Ishiguro, N.; Kishimoto, W.;
Ludwig-Schwellinger, E.; Ebner, T.; et al. Simultaneous absolute protein quantification of transporters,
cytochromes P450, and UDP-glucuronosyltransferases as a novel approach for the characterization of
individual human liver: Comparison with mRNA levels and activities. Drug Metab. Dispos. 2012,
40, 83–92. [CrossRef]
29. Nakamura, K.; Hirayama-Kurogi, M.; Ito, S.; Kuno, T.; Yoneyama, T.; Obuchi, W.; Terasaki, T.; Ohtsuki, S.
Large-scale multiplex absolute protein quantification of drug-metabolizing enzymes and transporters
in human intestine, liver, and kidney microsomes by SWATH-MS: Comparison with MRM/SRM and
HR-MRM/PRM. Proteomics 2016, 16, 2106–2117. [CrossRef]

196
Pharmaceutics 2020, 12, 1201

30. Miyauchi, E.; Tachikawa, M.; Declèves, X.; Uchida, Y.; Bouillot, J.-L.; Poitou, C.; Oppert, J.-M.; Mouly, S.;
Bergmann, J.-F.; Terasaki, T.; et al. Quantitative Atlas of Cytochrome P450, UDP-Glucuronosyltransferase,
and Transporter Proteins in Jejunum of Morbidly Obese Subjects. Mol. Pharm. 2016, 13, 2631–2640. [CrossRef]
31. Couto, N.; Al-Majdoub, Z.M.; Achour, B.; Wright, P.C.; Rostami-Hodjegan, A.; Barber, J. Quantification
of Proteins Involved in Drug Metabolism and Disposition in the Human Liver Using Label-Free Global
Proteomics. Mol. Pharm. 2019, 16, 632–647. [CrossRef] [PubMed]
32. Clermont, V.; Grangeon, A.; Barama, A.; Turgeon, J.; Lallier, M.; Malaise, J.; Michaud, V. Activity and mRNA
expression levels of selected cytochromes P450 in various sections of the human small intestine. Br. J.
Clin. Pharmacol. 2019, 85, 1367–1377. [CrossRef]
33. Fontana, R.J.; Lown, K.S.; Paine, M.F.; Fortlage, L.; Santella, R.M.; Felton, J.S.; Knize, M.G.; Greenberg, A.;
Watkins, P.B. Effects of a chargrilled meat diet on expression of CYP3A, CYP1A, and P-glycoprotein levels in
healthy volunteers. Gastroenterology 1999, 117, 89–98. [CrossRef]
34. Achour, B.; Barber, J.; Rostami-Hodjegan, A. Expression of hepatic drug-metabolizing cytochrome p450
enzymes and their intercorrelations: A meta-analysis. Drug Metab. Dispos. 2014, 42, 1349–1356.
[CrossRef] [PubMed]
35. Vildhede, A.; Wiśniewski, J.R.; Norén, A.; Karlgren, M.; Artursson, P. Comparative Proteomic Analysis
of Human Liver Tissue and Isolated Hepatocytes with a Focus on Proteins Determining Drug Exposure.
J. Proteome Res. 2015, 14, 3305–3314. [CrossRef]
36. Achour, B.; Al Feteisi, H.; Lanucara, F.; Rostami-Hodjegan, A.; Barber, J. Global Proteomic Analysis of Human
Liver Microsomes: Rapid Characterization and Quantification of Hepatic Drug-Metabolizing Enzymes.
Drug Metab. Dispos. 2017, 45, 666–675. [CrossRef]
37. Achour, B.; Russell, M.R.; Barber, J.; Rostami-Hodjegan, A. Simultaneous quantification of the abundance
of several cytochrome P450 and uridine 5 -diphospho-glucuronosyltransferase enzymes in human liver
microsomes using multiplexed targeted proteomics. Drug Metab. Dispos. 2014, 42, 500–510. [CrossRef]
38. Kawakami, H.; Ohtsuki, S.; Kamiie, J.; Suzuki, T.; Abe, T.; Terasaki, T. Simultaneous absolute quantification
of 11 cytochrome P450 isoforms in human liver microsomes by liquid chromatography tandem mass
spectrometry with in silico target peptide selection. J. Pharm. Sci. 2011, 100, 341–352. [CrossRef]
39. Shimada, T.; Yamazaki, H.; Mimura, M.; Inui, Y.; Guengerich, F.P. Interindividual variations in human liver
cytochrome P-450 enzymes involved in the oxidation of drugs, carcinogens and toxic chemicals: Studies
with liver microsomes of 30 Japanese and 30 Caucasians. J. Pharmacol. Exp. Ther. 1994, 270, 414–423.
40. Klein, K.; Winter, S.; Turpeinen, M.; Schwab, M.; Zanger, U.M. Pathway-Targeted Pharmacogenomics of
CYP1A2 in Human Liver. Front. Pharmacol. 2010, 1, 129. [CrossRef]
41. Jiang, Z.; Dragin, N.; Jorge-Nebert, L.F.; Martin, M.V.; Guengerich, F.P.; Aklillu, E.; Ingelman-Sundberg, M.;
Hammons, G.J.; Lyn-Cook, B.D.; Kadlubar, F.F.; et al. Search for an association between the human CYP1A2
genotype and CYP1A2 metabolic phenotype. Pharmacogenet. Genom. 2006, 16, 359–367. [CrossRef]
42. Ueda, R.; Iketaki, H.; Nagata, K.; Kimura, S.; Gonzalez, F.J.; Kusano, K.; Yoshimura, T.; Yamazoe, Y. A common
regulatory region functions bidirectionally in transcriptional activation of the human CYP1A1 and CYP1A2
genes. Mol. Pharmacol. 2006, 69, 1924–1930. [CrossRef]
43. Jorge-Nebert, L.F.; Jiang, Z.; Chakraborty, R.; Watson, J.; Jin, L.; McGarvey, S.T.; Deka, R.; Nebert, D.W.
Analysis of human CYP1A1 and CYP1A2 genes and their shared bidirectional promoter in eight world
populations. Hum. Mutat. 2010, 31, 27–40. [CrossRef] [PubMed]
44. Nebert, D.W.; Dalton, T.P.; Okey, A.B.; Gonzalez, F.J. Role of aryl hydrocarbon receptor-mediated induction
of the CYP1 enzymes in environmental toxicity and cancer. J. Biol. Chem. 2004, 279, 23847–23850.
[CrossRef] [PubMed]
45. Edwards, R.J.; Price, R.J.; Watts, P.S.; Renwick, A.B.; Tredger, J.M.; Boobis, A.R.; Lake, B.G. Induction of
cytochrome P450 enzymes in cultured precision-cut human liver slices. Drug Metab. Dispos. 2003, 31, 282–288.
[CrossRef] [PubMed]
46. Madan, A.; Graham, R.A.; Carroll, K.M.; Mudra, D.R.; Burton, L.A.; Krueger, L.A.; Downey, A.D.;
Czerwinski, M.; Forster, J.; Ribadeneira, M.D.; et al. Effects of prototypical microsomal enzyme inducers
on cytochrome P450 expression in cultured human hepatocytes. Drug Metab. Dispos. 2003, 31, 421–431.
[CrossRef] [PubMed]

197
Pharmaceutics 2020, 12, 1201

47. Moscovitz, J.E.; Kalgutkar, A.S.; Nulick, K.; Johnson, N.; Lin, Z.; Goosen, T.C.; Weng, Y. Establishing
Transcriptional Signatures to Differentiate PXR-, CAR-, and AhR-Mediated Regulation of Drug Metabolism and
Transport Genes in Cryopreserved Human Hepatocytes. J. Pharmacol. Exp. Ther. 2018, 365, 262–271. [CrossRef]
48. Roymans, D.; van Looveren, C.; Leone, A.; Parker, J.B.; McMillian, M.; Johnson, M.D.; Koganti, A.; Gilissen, R.;
Silber, P.; Mannens, G.; et al. Determination of cytochrome P450 1A2 and cytochrome P450 3A4 induction in
cryopreserved human hepatocytes. Biochem. Pharmacol. 2004, 67, 427–437. [CrossRef]
49. Yoshinari, K.; Ueda, R.; Kusano, K.; Yoshimura, T.; Nagata, K.; Yamazoe, Y. Omeprazole transactivates
human CYP1A1 and CYP1A2 expression through the common regulatory region containing multiple
xenobiotic-responsive elements. Biochem. Pharmacol. 2008, 76, 139–145. [CrossRef]
50. Bapiro, T.E.; Andersson, T.B.; Otter, C.; Hasler, J.A.; Masimirembwa, C.M. Cytochrome P450 1A1/2 induction
by antiparasitic drugs: Dose-dependent increase in ethoxyresorufin O-deethylase activity and mRNA caused
by quinine, primaquine and albendazole in HepG2 cells. Eur. J. Clin. Pharmacol. 2002, 58, 537–542. [CrossRef]
51. Dolwick, K.M.; Swanson, H.I.; Bradfield, C.A. In vitro analysis of Ah receptor domains involved in
ligand-activated DNA recognition. Proc. Natl. Acad. Sci. USA 1993, 90, 8566–8570. [CrossRef] [PubMed]
52. Sugiyama, I.; Murayama, N.; Kuroki, A.; Kota, J.; Iwano, S.; Yamazaki, H.; Hirota, T. Evaluation of cytochrome
P450 inductions by anti-epileptic drug oxcarbazepine, 10-hydroxyoxcarbazepine, and carbamazepine using
human hepatocytes and HepaRG cells. Xenobiotica 2016, 46, 765–774. [CrossRef]
53. Ghotbi, R.; Christensen, M.; Roh, H.-K.; Ingelman-Sundberg, M.; Aklillu, E.; Bertilsson, L. Comparisons of
CYP1A2 genetic polymorphisms, enzyme activity and the genotype-phenotype relationship in Swedes and
Koreans. Eur. J. Clin. Pharmacol. 2007, 63, 537–546. [CrossRef] [PubMed]
54. Dobrinas, M.; Cornuz, J.; Oneda, B.; Kohler Serra, M.; Puhl, M.; Eap, C.B. Impact of smoking, smoking
cessation, and genetic polymorphisms on CYP1A2 activity and inducibility. Clin. Pharmacol. Ther. 2011,
90, 117–125. [CrossRef] [PubMed]
55. Yoshinari, K.; Yoda, N.; Toriyabe, T.; Yamazoe, Y. Constitutive androstane receptor transcriptionally activates
human CYP1A1 and CYP1A2 genes through a common regulatory element in the 5 -flanking region.
Biochem. Pharmacol. 2010, 79, 261–269. [CrossRef] [PubMed]
56. Feidt, D.M.; Klein, K.; Hofmann, U.; Riedmaier, S.; Knobeloch, D.; Thasler, W.E.; Weiss, T.S.; Schwab, M.;
Zanger, U.M. Profiling induction of cytochrome p450 enzyme activity by statins using a new liquid
chromatography-tandem mass spectrometry cocktail assay in human hepatocytes. Drug Metab. Dispos. 2010,
38, 1589–1597. [CrossRef] [PubMed]
57. Rae, J.M.; Johnson, M.D.; Lippman, M.E.; Flockhart, D.A. Rifampin is a selective, pleiotropic inducer of
drug metabolism genes in human hepatocytes: Studies with cDNA and oligonucleotide expression arrays.
J. Pharmacol. Exp. Ther. 2001, 299, 849–857. [PubMed]
58. Backman, J.T.; Granfors, M.T.; Neuvonen, P.J. Rifampicin is only a weak inducer of CYP1A2-mediated
presystemic and systemic metabolism: Studies with tizanidine and caffeine. Eur. J. Clin. Pharmacol. 2006,
62, 451–461. [CrossRef] [PubMed]
59. Parkinson, A.; Mudra, D.R.; Johnson, C.; Dwyer, A.; Carroll, K.M. The effects of gender, age, ethnicity,
and liver cirrhosis on cytochrome P450 enzyme activity in human liver microsomes and inducibility in
cultured human hepatocytes. Toxicol. Appl. Pharmacol. 2004, 199, 193–209. [CrossRef] [PubMed]
60. Relling, M.V.; Lin, J.S.; Ayers, G.D.; Evans, W.E. Racial and gender differences in N-acetyltransferase,
xanthine oxidase, and CYP1A2 activities. Clin. Pharmacol. Ther. 1992, 52, 643–658. [CrossRef]
61. Ou-Yang, D.S.; Huang, S.L.; Wang, W.; Xie, H.G.; Xu, Z.H.; Shu, Y.; Zhou, H.H. Phenotypic polymorphism
and gender-related differences of CYP1A2 activity in a Chinese population. Br. J. Clin. Pharmacol. 2000,
49, 145–151. [CrossRef] [PubMed]
62. Backman, J.T.; Schröder, M.T.; Neuvonen, P.J. Effects of gender and moderate smoking on the pharmacokinetics
and effects of the CYP1A2 substrate tizanidine. Eur. J. Clin. Pharmacol. 2008, 64, 17–24. [CrossRef] [PubMed]
63. Orlando, R.; Padrini, R.; Perazzi, M.; de Martin, S.; Piccoli, P.; Palatini, P. Liver dysfunction markedly
decreases the inhibition of cytochrome P450 1A2-mediated theophylline metabolism by fluvoxamine.
Clin. Pharmacol. Ther. 2006, 79, 489–499. [CrossRef] [PubMed]
64. Zhang, Y.; Klein, K.; Sugathan, A.; Nassery, N.; Dombkowski, A.; Zanger, U.M.; Waxman, D.J. Transcriptional
profiling of human liver identifies sex-biased genes associated with polygenic dyslipidemia and coronary
artery disease. PLoS ONE 2011, 6, e23506. [CrossRef] [PubMed]

198
Pharmaceutics 2020, 12, 1201

65. Chen, H.; Shen, Z.-Y.; Xu, W.; Fan, T.-Y.; Li, J.; Lu, Y.-F.; Cheng, M.-L.; Liu, J. Expression of P450 and
nuclear receptors in normal and end-stage Chinese livers. World J. Gastroenterol. 2014, 20, 8681–8690.
[CrossRef] [PubMed]
66. Nakai, K.; Tanaka, H.; Hanada, K.; Ogata, H.; Suzuki, F.; Kumada, H.; Miyajima, A.; Ishida, S.; Sunouchi, M.;
Habano, W.; et al. Decreased expression of cytochromes P450 1A2, 2E1, and 3A4 and drug transporters
Na+-taurocholate-cotransporting polypeptide, organic cation transporter 1, and organic anion-transporting
peptide-C correlates with the progression of liver fibrosis in chronic hepatitis C patients. Drug Metab. Dispos.
2008, 36, 1786–1793. [CrossRef] [PubMed]
67. Hanada, K.; Nakai, K.; Tanaka, H.; Suzuki, F.; Kumada, H.; Ohno, Y.; Ozawa, S.; Ogata, H. Effect of nuclear
receptor downregulation on hepatic expression of cytochrome P450 and transporters in chronic hepatitis C
in association with fibrosis development. Drug Metab. Pharmacokinet. 2012, 27, 301–306. [CrossRef]
68. Prasad, B.; Bhatt, D.K.; Johnson, K.; Chapa, R.; Chu, X.; Salphati, L.; Xiao, G.; Lee, C.; Hop, C.E.C.A.;
Mathias, A.; et al. Abundance of Phase 1 and 2 Drug-Metabolizing Enzymes in Alcoholic and
Hepatitis C Cirrhotic Livers: A Quantitative Targeted Proteomics Study. Drug Metab. Dispos. 2018,
46, 943–952. [CrossRef]
69. Fisher, C.D.; Lickteig, A.J.; Augustine, L.M.; Ranger-Moore, J.; Jackson, J.P.; Ferguson, S.S.; Cherrington, N.J.
Hepatic cytochrome P450 enzyme alterations in humans with progressive stages of nonalcoholic fatty liver
disease. Drug Metab. Dispos. 2009, 37, 2087–2094. [CrossRef]
70. Parker, A.C.; Pritchard, P.; Preston, T.; Choonara, I. Induction of CYP1A2 activity by carbamazepine in
children using the caffeine breath test. Br. J. Clin. Pharmacol. 1998, 45, 176–178. [CrossRef]
71. Buchthal, J.; Grund, K.E.; Buchmann, A.; Schrenk, D.; Beaune, P.; Bock, K.W. Induction of cytochrome P4501A
by smoking or omeprazole in comparison with UDP-glucuronosyltransferase in biopsies of human duodenal
mucosa. Eur. J. Clin. Pharmacol. 1995, 47, 431–435. [CrossRef]
72. Diaz, D.; Fabrev, I.; Daujat, M.; Aubert, B.S.; Bories, P.; Michel, H.; Maurel, P. Omeprazole is an aryl
hydrocarbon-like inducer of human hepatic cytochrome P450. Gastroenterology 1990, 99, 737–747. [CrossRef]
73. Halladay, J.S.; Wong, S.; Khojasteh, S.C.; Grepper, S. An ‘all-inclusive’ 96-well cytochrome P450 induction
method: Measuring enzyme activity, mRNA levels, protein levels, and cytotoxicity from one well using
cryopreserved human hepatocytes. J. Pharmacol. Toxicol. Methods 2012, 66, 270–275. [CrossRef] [PubMed]
74. Dixit, V.; Hariparsad, N.; Li, F.; Desai, P.; Thummel, K.E.; Unadkat, J.D. Cytochrome P450 enzymes and
transporters induced by anti-human immunodeficiency virus protease inhibitors in human hepatocytes:
Implications for predicting clinical drug interactions. Drug Metab. Dispos. 2007, 35, 1853–1859.
[CrossRef] [PubMed]
75. Pelkonen, O.; Pasanen, M.; Kuha, H.; Gachalyi, B.; Kairaluoma, M.; Sotaniemi, E.A.; Park, S.S.; Friedman, F.K.;
Gelboin, H.V. The effect of cigarette smoking on 7-ethoxyresorufin O-deethylase and other monooxygenase
activities in human liver: Analyses with monoclonal antibodies. Br. J. Clin. Pharmacol. 1986, 22, 125–134.
[CrossRef] [PubMed]
76. Gunes, A.; Dahl, M.-L. Variation in CYP1A2 activity and its clinical implications: Influence of environmental
factors and genetic polymorphisms. Pharmacogenomics 2008, 9, 625–637. [CrossRef] [PubMed]
77. Ingelman-Sundberg, M.; Sim, S.C.; Gomez, A.; Rodriguez-Antona, C. Influence of cytochrome
P450 polymorphisms on drug therapies: Pharmacogenetic, pharmacoepigenetic and clinical aspects.
Pharmacol. Ther. 2007, 116, 496–526. [CrossRef] [PubMed]
78. Rasmussen, B.B.; Brix, T.H.; Kyvik, K.O.; Brøsen, K. The interindividual differences in the 3-demthylation
of caffeine alias CYP1A2 is determined by both genetic and environmental factors. Pharmacogenetics 2002,
12, 473–478. [CrossRef]
79. Zanger, U.M.; Klein, K.; Thomas, M.; Rieger, J.K.; Tremmel, R.; Kandel, B.A.; Klein, M.; Magdy, T. Genetics,
epigenetics, and regulation of drug-metabolizing cytochrome p450 enzymes. Clin. Pharmacol. Ther. 2014,
95, 258–261. [CrossRef]
80. Kisselev, P.; Schunck, W.-H.; Roots, I.; Schwarz, D. Association of CYP1A1 polymorphisms with differential
metabolic activation of 17beta-estradiol and estrone. Cancer Res. 2005, 65, 2972–2978. [CrossRef]
81. Zhou, H.; Josephy, P.D.; Kim, D.; Guengerich, F.P. Functional characterization of four allelic variants of
human cytochrome P450 1A2. Arch. Biochem. Biophys. 2004, 422, 23–30. [CrossRef] [PubMed]

199
Pharmaceutics 2020, 12, 1201

82. Palma, B.B.; Silva E Sousa, M.; Vosmeer, C.R.; Lastdrager, J.; Rueff, J.; Vermeulen, N.P.E.; Kranendonk, M.
Functional characterization of eight human cytochrome P450 1A2 gene variants by recombinant protein
expression. Pharm. J. 2010, 10, 478–488. [CrossRef] [PubMed]
83. Nakajima, M.; Yokoi, T.; Mizutani, M.; Kinoshita, M.; Funayama, M.; Kamataki, T. Genetic polymorphism in
the 5 -flanking region of human CYP1A2 gene: Effect on the CYP1A2 inducibility in humans. J. Biochem.
1999, 125, 803–808. [CrossRef] [PubMed]
84. Sachse, C.; Brockmöller, J.; Bauer, S.; Roots, I. Functional significance of a C–A polymorphism in intron
1 of the cytochrome P450 CYP1A2 gene tested with caffeine. Br. J. Clin. Pharmacol. 1999, 47, 445–449.
[CrossRef] [PubMed]
85. Djordjevic, N.; Ghotbi, R.; Jankovic, S.; Aklillu, E. Induction of CYP1A2 by heavy coffee consumption is
associated with the CYP1A2 -163CA polymorphism. Eur. J. Clin. Pharmacol. 2010, 66, 697–703. [CrossRef]
86. Han, X.-M.; Ouyang, D.-S.; Chen, X.-P.; Shu, Y.; Jiang, C.-H.; Tan, Z.-R.; Zhou, H.-H. Inducibility of CYP1A2
by omeprazole in vivo related to the genetic polymorphism of CYP1A2. Br. J. Clin. Pharmacol. 2002,
54, 540–543. [CrossRef]
87. Sergentanis, T.N.; Economopoulos, K.P. Four polymorphisms in cytochrome P450 1A1 (CYP1A1) gene and
breast cancer risk: A meta-analysis. Breast Cancer Res. Treat. 2010, 122, 459–469. [CrossRef]
88. Yao, L.; Yu, X.; Yu, L. Lack of significant association between CYP1A1 T3801C polymorphism and breast
cancer risk: A meta-analysis involving 25,087 subjects. Breast Cancer Res. Treat. 2010, 122, 503–507. [CrossRef]
89. Cui, X.; Lu, X.; Hiura, M.; Omori, H.; Miyazaki, W.; Katoh, T. Association of genotypes of
carcinogen-metabolizing enzymes and smoking status with bladder cancer in a Japanese population.
Environ. Health Prev. Med. 2013, 18, 136–142. [CrossRef]
90. Wang, H.; Zhang, Z.; Han, S.; Lu, Y.; Feng, F.; Yuan, J. CYP1A2 rs762551 polymorphism contributes to cancer
susceptibility: A meta-analysis from 19 case-control studies. BMC Cancer 2012, 12, 528. [CrossRef]
91. Chen, Y.; Zeng, L.; Wang, Y.; Tolleson, W.H.; Knox, B.; Chen, S.; Ren, Z.; Guo, L.; Mei, N.; Qian, F.; et al.
The expression, induction and pharmacological activity of CYP1A2 are post-transcriptionally regulated by
microRNA hsa-miR-132-5p. Biochem. Pharmacol. 2017, 145, 178–191. [CrossRef] [PubMed]
92. Gill, P.; Bhattacharyya, S.; McCullough, S.; Letzig, L.; Mishra, P.J.; Luo, C.; Dweep, H.; James, L. MicroRNA
regulation of CYP 1A2, CYP3A4 and CYP2E1 expression in acetaminophen toxicity. Sci. Rep. 2017, 7, 12331.
[CrossRef] [PubMed]
93. Sansen, S.; Yano, J.K.; Reynald, R.L.; Schoch, G.A.; Griffin, K.J.; Stout, C.D.; Johnson, E.F. Adaptations for the
oxidation of polycyclic aromatic hydrocarbons exhibited by the structure of human P450 1A2. J. Biol. Chem.
2007, 282, 14348–14355. [CrossRef] [PubMed]
94. Tassaneeyakul, W.; Birkett, D.J.; Veronese, M.E.; McManus, M.E.; Tukey, R.H.; Quattrochi, L.C.; Gelboin, H.V.;
Miners, J.O. Specificity of substrate and inhibitor probes for human cytochromes P450 1A1 and 1A2.
J. Pharmacol. Exp. Ther. 1993, 265, 401–407. [PubMed]
95. Tassaneeyakul, W.; Mohamed, Z.; Birkett, D.J.; McManus, M.E.; Veronese, M.E.; Tukey, R.H.; Quattrochi, L.C.;
Gonzalez, F.J.; Miners, J.O. Caffeine as a probe for human cytochromes P450: Validation using
cDNA-expression, immunoinhibition and microsomal kinetic and inhibitor techniques. Pharmacogenetics
1992, 2, 173–183. [CrossRef] [PubMed]
96. Ma, Q.; Lu, A.Y.H. CYP1A induction and human risk assessment: An evolving tale of in vitro and in vivo
studies. Drug Metab. Dispos. 2007, 35, 1009–1016. [CrossRef]
97. Androutsopoulos, V.P.; Tsatsakis, A.M.; Spandidos, D.A. Cytochrome P450 CYP1A1: Wider roles in cancer
progression and prevention. BMC Cancer 2009, 9, 187. [CrossRef]
98. Nakamura, H.; Ariyoshi, N.; Okada, K.; Nakasa, H.; Nakazawa, K.; Kitada, M. CYP1A1 is a major enzyme
responsible for the metabolism of granisetron in human liver microsomes. Curr. Drug Metab. 2005,
6, 469–480. [CrossRef]
99. Mescher, M.; Tigges, J.; Rolfes, K.M.; Shen, A.L.; Yee, J.S.; Vogeley, C.; Krutmann, J.; Bradfield, C.A.; Lang, D.;
Haarmann-Stemmann, T. The Toll-like receptor agonist imiquimod is metabolized by aryl hydrocarbon
receptor-regulated cytochrome P450 enzymes in human keratinocytes and mouse liver. Arch. Toxicol. 2019,
93, 1917–1926. [CrossRef]
100. Liu, L.; Wang, Q.; Xie, C.; Xi, N.; Guo, Z.; Li, M.; Hou, X.; Xie, N.; Sun, M.; Li, J.; et al. Drug interaction of
ningetinib and gefitinib involving CYP1A1 and efflux transporters in non-small cell lung cancer patients.
Br. J. Clin. Pharmacol. 2020. [CrossRef]

200
Pharmaceutics 2020, 12, 1201

101. Niwa, T.; Sato, R.; Yabusaki, Y.; Ishibashi, F.; Katagiri, M. Contribution of human hepatic cytochrome
P450s and steroidogenic CYP17 to the N-demethylation of aminopyrine. Xenobiotica 1999, 29, 187–193.
[CrossRef] [PubMed]
102. Norman, T.R.; Olver, J.S. Agomelatine for depression: Expanding the horizons? Expert Opin. Pharmacother.
2019, 20, 647–656. [CrossRef] [PubMed]
103. Bertilsson, L.; Carrillo, J.A.; Dahl, M.L.; Llerena, A.; Alm, C.; Bondesson, U.; Lindström, L.; La Rodriguez
de Rubia, I.; Ramos, S.; BENITEZ, J. Clozapine disposition covaries with CYP1A2 activity determined by a
caffeine test. Br. J. Clin. Pharmacol. 1994, 38, 471–473. [CrossRef] [PubMed]
104. Long, L.; Dolan, M.E. Role of cytochrome P450 isoenzymes in metabolism of O(6)-benzylguanine:
Implications for dacarbazine activation. Clin. Cancer Res. 2001, 7, 4239–4244.
105. Lobo, E.D.; Bergstrom, R.F.; Reddy, S.; Quinlan, T.; Chappell, J.; Hong, Q.; Ring, B.; Knadler, M.P. In vitro
and in vivo evaluations of cytochrome P450 1A2 interactions with duloxetine. Clin. Pharmacokinet. 2008,
47, 191–202. [CrossRef] [PubMed]
106. Shet, M.S.; McPhaul, M.; Fisher, C.W.; Stallings, N.R.; Estabrook, R.W. Metabolism of the antiandrogenic
drug (Flutamide) by human CYP1A2. Drug Metab. Dispos. 1997, 25, 1298–1303.
107. Kalgutkar, A.S.; Nguyen, H.T.; Vaz, A.D.N.; Doan, A.; Dalvie, D.K.; McLeod, D.G.; Murray, J.C. In vitro
metabolism studies on the isoxazole ring scission in the anti-inflammatory agent lefluonomide to its
active alpha-cyanoenol metabolite A771726: Mechanistic similarities with the cytochrome P450-catalyzed
dehydration of aldoximes. Drug Metab. Dispos. 2003, 31, 1240–1250. [CrossRef]
108. Härtter, S.; Grözinger, M.; Weigmann, H.; Röschke, J.; Hiemke, C. Increased bioavailability of oral melatonin
after fluvoxamine coadministration. Clin. Pharmacol. Ther. 2000, 67, 1–6. [CrossRef]
109. Anttila, A.K.; Rasanen, L.; Leinonen, E.V. Fluvoxamine augmentation increases serum mirtazapine
concentrations three- to fourfold. Ann. Pharmacother. 2001, 35, 1221–1223. [CrossRef]
110. Turpeinen, M.; Hofmann, U.; Klein, K.; Mürdter, T.; Schwab, M.; Zanger, U.M. A predominate role of CYP1A2
for the metabolism of nabumetone to the active metabolite, 6-methoxy-2-naphthylacetic acid, in human liver
microsomes. Drug Metab. Dispos. 2009, 37, 1017–1024. [CrossRef]
111. Ring, B.J.; Catlow, J.; Lindsay, T.J.; Gillespie, T.; Roskos, L.K.; Cerimele, B.J.; Swanson, S.P.; Hamman, M.A.;
Wrighton, S.A. Identification of the human cytochromes P450 responsible for the in vitro formation of
the major oxidative metabolites of the antipsychotic agent olanzapine. J. Pharmacol. Exp. Ther. 1996,
276, 658–666. [PubMed]
112. Wójcikowski, J.; Pichard-Garcia, L.; Maurel, P.; Daniel, W.A. Contribution of human cytochrome p-450
isoforms to the metabolism of the simplest phenothiazine neuroleptic promazine. Br. J. Pharmacol. 2003,
138, 1465–1474. [CrossRef]
113. Masubuchi, Y.; Hosokawa, S.; Horie, T.; Suzuki, T.; Ohmori, S.; Kitada, M.; Narimatsu, S. Cytochrome
P450 isozymes involved in propranolol metabolism in human liver microsomes. The role of CYP2D6 as
ring-hydroxylase and CYP1A2 as N-desisopropylase. Drug Metab. Dispos. 1994, 22, 909–915. [PubMed]
114. Obach, R.S.; Ryder, T.F. Metabolism of ramelteon in human liver microsomes and correlation with the effect of
fluvoxamine on ramelteon pharmacokinetics. Drug Metab. Dispos. 2010, 38, 1381–1391. [CrossRef] [PubMed]
115. Guay, D.R.P. Rasagiline (TVP-1012): A new selective monoamine oxidase inhibitor for Parkinson’s disease.
Am. J. Geriatr. Pharmacother. 2006, 4, 330–346. [CrossRef] [PubMed]
116. Sanderink, G.J.; Bournique, B.; Stevens, J.; Petry, M.; Martinet, M. Involvement of human CYP1A isoenzymes
in the metabolism and drug interactions of riluzole in vitro. J. Pharmacol. Exp. Ther. 1997, 282, 1465–1472.
117. Kaye, C.M.; Nicholls, B. Clinical pharmacokinetics of ropinirole. Clin. Pharmacokinet. 2000, 39, 243–254. [CrossRef]
118. Oda, Y.; Furuichi, K.; Tanaka, K.; Hiroi, T.; Imaoka, S.; Asada, A.; Fujimori, M.; Funae, Y. Metabolism of
a new local anesthetic, ropivacaine, by human hepatic cytochrome P450. Anesthesiology 1995, 82, 214–220.
[CrossRef] [PubMed]
119. Spaldin, V.; Madden, S.; Pool, W.F.; Woolf, T.F.; Park, B.K. The effect of enzyme inhibition on the metabolism and
activation of tacrine by human liver microsomes. Br. J. Clin. Pharmacol. 1994, 38, 15–22. [CrossRef] [PubMed]
120. Ha, H.R.; Chen, J.; Freiburghaus, A.U.; Follath, F. Metabolism of theophylline by cDNA-expressed human
cytochromes P-450. Br. J. Clin. Pharmacol. 1995, 39, 321–326. [CrossRef] [PubMed]
121. Granfors, M.T.; Backman, J.T.; Neuvonen, M.; Neuvonen, P.J. Ciprofloxacin greatly increases concentrations
and hypotensive effect of tizanidine by inhibiting its cytochrome P450 1A2-mediated presystemic metabolism.
Clin. Pharmacol. Ther. 2004, 76, 598–606. [CrossRef] [PubMed]

201
Pharmaceutics 2020, 12, 1201

122. Granfors, M.T.; Backman, J.T.; Neuvonen, M.; Ahonen, J.; Neuvonen, P.J. Fluvoxamine drastically increases
concentrations and effects of tizanidine: A potentially hazardous interaction. Clin. Pharmacol. Ther. 2004,
75, 331–341. [CrossRef] [PubMed]
123. Kroemer, H.K.; Gautier, J.C.; Beaune, P.; Henderson, C.; Wolf, C.R.; Eichelbaum, M. Identification of P450
enzymes involved in metabolism of verapamil in humans. Naunyn Schmiedebergs Arch. Pharmacol. 1993,
348, 332–337. [CrossRef] [PubMed]
124. Wild, M.J.; McKillop, D.; Butters, C.J. Determination of the human cytochrome P450 isoforms involved in the
metabolism of zolmitriptan. Xenobiotica 1999, 29, 847–857. [CrossRef]
125. Fuhr, U.; Jetter, A.; Kirchheiner, J. Appropriate phenotyping procedures for drug metabolizing enzymes and
transporters in humans and their simultaneous use in the “cocktail” approach. Clin. Pharmacol. Ther. 2007,
81, 270–283. [CrossRef]
126. Nebert, D.W.; Dalton, T.P. The role of cytochrome P450 enzymes in endogenous signalling pathways and
environmental carcinogenesis. Nat. Reviews Cancer 2006, 6, 947–960. [CrossRef]
127. Karjalainen, M.J.; Neuvonen, P.J.; Backman, J.T. In vitro inhibition of CYP1A2 by model inhibitors,
anti-inflammatory analgesics and female sex steroids: Predictability of in vivo interactions. Basic Clin.
Pharmacol. Toxicol. 2008, 103, 157–165. [CrossRef]
128. Pastrakuljic, A.; Tang, B.K.; Roberts, E.A.; Kalow, W. Distinction of CYP1A1 and CYP1A2 activity by selective
inhibition using fluvoxamine and isosafrole. Biochem. Pharmacol. 1997, 53, 531–538. [CrossRef]
129. Somers, G.I.; Harris, A.J.; Bayliss, M.K.; Houston, J.B. The metabolism of the 5HT3 antagonists ondansetron,
alosetron and GR87442 I: A comparison of in vitro and in vivo metabolism and in vitro enzyme kinetics
in rat, dog and human hepatocytes, microsomes and recombinant human enzymes. Xenobiotica 2007,
37, 832–854. [CrossRef]
130. Kobayashi, K.; Nakajima, M.; Chiba, K.; Yamamoto, T.; Tani, M.; Ishizaki, T.; Kuroiwa, Y. Inhibitory effects of
antiarrhythmic drugs on phenacetin O-deethylation catalysed by human CYP1A2. Br. J. Clin. Pharmacol.
1998, 45, 361–368. [CrossRef]
131. Bapiro, T.E.; Sayi, J.; Hasler, J.A.; Jande, M.; Rimoy, G.; Masselle, A.; Masimirembwa, C.M. Artemisinin
and thiabendazole are potent inhibitors of cytochrome P450 1A2 (CYP1A2) activity in humans. Eur. J.
Clin. Pharmacol. 2005, 61, 755–761. [CrossRef] [PubMed]
132. Masubuchi, Y.; Nakano, T.; Ose, A.; Horie, T. Differential selectivity in carbamazepine-induced inactivation
of cytochrome P450 enzymes in rat and human liver. Arch. Toxicol. 2001, 75, 538–543. [CrossRef] [PubMed]
133. Fuhr, U.; Anders, E.M.; Mahr, G.; Sörgel, F.; Staib, A.H. Inhibitory potency of quinolone antibacterial agents
against cytochrome P450IA2 activity in vivo and in vitro. Antimicrob. Agents Chemother. 1992, 36, 942–948.
[CrossRef] [PubMed]
134. Martinez, C.; Albet, C.; Agundez, J.; Herrero, E.; Carrillo, J.; Marquez, M.; Benitez, J.; Ortiz, J. Comparative
in vitro and in vivo inhibition of cytochrome P450 CYP1A2, CYP2D6, and CYP3A by H -receptor antagonists.
Clin. Pharmacol. Ther. 1999, 65, 369–376. [CrossRef]
135. Zhang, W.; Ramamoorthy, Y.; Kilicarslan, T.; Nolte, H.; Tyndale, R.F.; Sellers, E.M. Inhibition of cytochromes
P450 by antifungal imidazole derivatives. Drug Metab. Dispos. 2002, 30, 314–318. [CrossRef] [PubMed]
136. Paris, B.L.; Ogilvie, B.W.; Scheinkoenig, J.A.; Ndikum-Moffor, F.; Gibson, R.; Parkinson, A. In vitro inhibition
and induction of human liver cytochrome p450 enzymes by milnacipran. Drug Metab. Dispos. 2009,
37, 2045–2054. [CrossRef] [PubMed]
137. von Moltke, L.L.; Greenblatt, D.J.; Schmider, J.; Duan, S.X.; Wright, C.E.; Harmatz, J.S.; Shader, R.I. Midazolam
hydroxylation by human liver microsomes in vitro: Inhibition by fluoxetine, norfluoxetine, and by azole
antifungal agents. J. Clin. Pharmacol. 1996, 36, 783–791. [CrossRef]
138. Wen, X.; Wang, J.-S.; Neuvonen, P.J.; Backman, J.T. Isoniazid is a mechanism-based inhibitor of cytochrome
P450 1A2, 2A6, 2C19 and 3A4 isoforms in human liver microsomes. Eur. J. Clin. Pharmacol. 2002,
57, 799–804. [CrossRef]
139. Sai, Y.; Dai, R.; Yang, T.J.; Krausz, K.W.; Gonzalez, F.J.; Gelboin, H.V.; Shou, M. Assessment of specificity of
eight chemical inhibitors using cDNA-expressed cytochromes P450. Xenobiotica 2000, 30, 327–343. [CrossRef]
140. Chun, Y.J.; Kim, M.Y.; Guengerich, F.P. Resveratrol is a selective human cytochrome P450 1A1 inhibitor.
Biochem. Biophys. Res. Commun. 1999, 262, 20–24. [CrossRef]

202
Pharmaceutics 2020, 12, 1201

141. Chang, T.K.; Chen, J.; Lee, W.B. Differential inhibition and inactivation of human CYP1 enzymes by
trans-resveratrol: Evidence for mechanism-based inactivation of CYP1A2. J. Pharmacol. Exp. Ther. 2001,
299, 874–882. [PubMed]
142. Augustin, M.; Schoretsanitis, G.; Pfeifer, P.; Gründer, G.; Liebe, C.; Paulzen, M. Effect of fluvoxamine
augmentation and smoking on clozapine serum concentrations. Schizophr. Res. 2019, 210, 143–148.
[CrossRef] [PubMed]
143. Chiu, C.-C.; Lane, H.-Y.; Huang, M.-C.; Liu, H.-C.; Jann, M.W.; Hon, Y.-Y.; Chang, W.-H.; Lu, M.-L.
Dose-dependent alternations in the pharmacokinetics of olanzapine during coadministration of fluvoxamine
in patients with schizophrenia. J. Clin. Pharmacol. 2004, 44, 1385–1390. [CrossRef] [PubMed]
144. Yao, C.; Kunze, K.L.; Kharasch, E.D.; Wang, Y.; Trager, W.F.; Ragueneau, I.; Levy, R.H.
Fluvoxamine-theophylline interaction: Gap between in vitro and in vivo inhibition constants toward
cytochrome P4501A2. Clin. Pharmacol. Ther. 2001, 70, 415–424. [CrossRef]
145. Becquemont, L.; Ragueneau, I.; Le Bot, M.A.; Riche, C.; Funck-Brentano, C.; Jaillon, P. Influence of the
CYP1A2 inhibitor fluvoxamine on tacrine pharmacokinetics in humans. Clin. Pharmacol. Ther. 1997,
61, 619–627. [CrossRef]
146. Backman, J.T.; Karjalainen, M.J.; Neuvonen, M.; Laitila, J.; Neuvonen, P.J. Rofecoxib is a potent inhibitor of
cytochrome P450 1A2: Studies with tizanidine and caffeine in healthy subjects. Br. J. Clin. Pharmacol. 2006,
62, 345–357. [CrossRef]
147. Meyer, J.M.; Proctor, G.; Cummings, M.A.; Dardashti, L.J.; Stahl, S.M. Ciprofloxacin and Clozapine:
A Potentially Fatal but Underappreciated Interaction. Case Rep. Psychiatry 2016, 2016, 5606098. [CrossRef]
148. Brouwers, E.E.M.; Söhne, M.; Kuipers, S.; van Gorp, E.C.M.; Schellens, J.H.M.; Koks, C.H.W.; Beijnen, J.H.;
Huitema, A.D.R. Ciprofloxacin strongly inhibits clozapine metabolism: Two case reports. Clin. Drug Investig.
2009, 29, 59–63. [CrossRef]
149. Jerling, M.; Lindström, L.; Bondesson, U.; Bertilsson, L. Fluvoxamine inhibition and carbamazepine induction
of the metabolism of clozapine: Evidence from a therapeutic drug monitoring service. Ther. Drug Monit.
1994, 16, 368–374. [CrossRef]
150. Wetzel, H.; Anghelescu, I.; Szegedi, A.; Wiesner, J.; Weigmann, H.; Härter, S.; Hiemke, C. Pharmacokinetic
interactions of clozapine with selective serotonin reuptake inhibitors: Differential effects of fluvoxamine and
paroxetine in a prospective study. J. Clin. Psychopharmacol. 1998, 18, 2–9. [CrossRef]
151. Raaska, K.; Neuvonen, P.J. Ciprofloxacin increases serum clozapine and N-desmethylclozapine: A study in
patients with schizophrenia. Eur. J. Clin. Pharmacol. 2000, 56, 585–589. [CrossRef] [PubMed]
152. Perucca, E.; Gatti, G.; Spina, E. Clinical pharmacokinetics of fluvoxamine. Clin. Pharmacokinet. 1994,
27, 175–190. [CrossRef] [PubMed]
153. Jokinen, M.J.; Olkkola, K.T.; Ahonen, J.; Neuvonen, P.J. Effect of ciprofloxacin on the pharmacokinetics of
ropivacaine. Eur. J. Clin. Pharmacol. 2003, 58, 653–657. [CrossRef] [PubMed]
154. Batty, K.T.; Davis, T.M.; Ilett, K.F.; Dusci, L.J.; Langton, S.R. The effect of ciprofloxacin on theophylline
pharmacokinetics in healthy subjects. Br. J. Clin. Pharmacol. 1995, 39, 305–311. [CrossRef]
155. Granfors, M.T.; Backman, J.T.; Laitila, J.; Neuvonen, P.J. Oral contraceptives containing ethinyl estradiol and
gestodene markedly increase plasma concentrations and effects of tizanidine by inhibiting cytochrome P450
1A2. Clin. Pharmacol. Ther. 2005, 78, 400–411. [CrossRef]
156. Komoroski, B.J.; Zhang, S.; Cai, H.; Hutzler, J.M.; Frye, R.; Tracy, T.S.; Strom, S.C.; Lehmann, T.; Ang, C.Y.W.;
Cui, Y.Y.; et al. Induction and inhibition of cytochromes P450 by the St. John’s wort constituent hyperforin in
human hepatocyte cultures. Drug Metab. Dispos. 2004, 32, 512–518. [CrossRef]
157. Andersson, T.; Bergstrand, R.; Cederberg, C.; Eriksson, S.; Lagerström, P.-O.; Skånberg, I. Omeprazole
treatment does not affect the metabolism of caffeine. Gastroenterology 1991, 101, 943–947. [CrossRef]
158. Rizzo, N.; Padoin, C.; Palombo, S.; Scherrmann, J.M.; Girre, C. Omeprazole and lansoprazole are not
inducers of cytochrome P4501A2 under conventional therapeutic conditions. Eur. J. Clin. Pharmacol. 1996,
49, 491–495. [CrossRef]
159. Dilger, K.; Zheng, Z.; Klotz, U. Lack of drug interaction between omeprazole, lansoprazole, pantoprazole
and theophylline. Br. J. Clin. Pharmacol. 1999, 48, 438–444. [CrossRef]
160. Henry, D.; Brent, P.; Whyte, I.; Mihaly, G.; Devenish-Meares, S. Propranolol steady-state pharmacokinetics
are unaltered by omeprazole. Eur. J. Clin. Pharmacol. 1987, 33, 369–373. [CrossRef]

203
Pharmaceutics 2020, 12, 1201

161. Frick, A.; Kopitz, J.; Bergemann, N. Omeprazole reduces clozapine plasma concentrations. A case report.
Pharmacopsychiatry 2003, 36, 121–123. [CrossRef] [PubMed]
162. Lucas, R.A.; Gilfillan, D.J.; Bergstrom, R.F. A pharmacokinetic interaction between carbamazepine
and olanzapine: Observations on possible mechanism. Eur. J. Clin. Pharmacol. 1998, 54, 639–643.
[CrossRef] [PubMed]
163. Magnusson, M.O.; Dahl, M.-L.; Cederberg, J.; Karlsson, M.O.; Sandström, R. Pharmacodynamics of
carbamazepine-mediated induction of CYP3A4, CYP1A2, and Pgp as assessed by probe substrates midazolam,
caffeine, and digoxin. Clin. Pharmacol. Ther. 2008, 84, 52–62. [CrossRef] [PubMed]
164. Kirby, B.J.; Collier, A.C.; Kharasch, E.D.; Dixit, V.; Desai, P.; Whittington, D.; Thummel, K.E.; Unadkat, J.D.
Complex drug interactions of HIV protease inhibitors 2: In vivo induction and in vitro to in vivo correlation
of induction of cytochrome P450 1A2, 2B6, and 2C9 by ritonavir or nelfinavir. Drug Metab. Dispos. 2011,
39, 2329–2337. [CrossRef] [PubMed]
165. Penzak, S.R.; Hon, Y.Y.; Lawhorn, W.D.; Shirley, K.L.; Spratlin, V.; Jann, M.W. Influence of ritonavir
on olanzapine pharmacokinetics in healthy volunteers. J. Clin. Psychopharmacol. 2002, 22, 366–370.
[CrossRef] [PubMed]
166. Jacobs, B.S.; Colbers, A.P.H.; Velthoven-Graafland, K.; Schouwenberg, B.J.J.W.; Burger, D.M. Effect of
fosamprenavir/ritonavir on the pharmacokinetics of single-dose olanzapine in healthy volunteers. Int. J.
Antimicrob. Agents 2014, 44, 173–177. [CrossRef]
167. Pascussi, J.-M.; Gerbal-Chaloin, S.; Duret, C.; Daujat-Chavanieu, M.; Vilarem, M.-J.; Maurel, P. The tangle of
nuclear receptors that controls xenobiotic metabolism and transport: Crosstalk and consequences. Annu. Rev.
Pharmacol. Toxicol. 2008, 48, 1–32. [CrossRef]
168. Yeh, R.F.; Gaver, V.E.; Patterson, K.B.; Rezk, N.L.; Baxter-Meheux, F.; Blake, M.J.; Eron, J.J.; Klein, C.E.;
Rublein, J.C.; Kashuba, A.D.M. Lopinavir/ritonavir induces the hepatic activity of cytochrome P450 enzymes
CYP2C9, CYP2C19, and CYP1A2 but inhibits the hepatic and intestinal activity of CYP3A as measured by a
phenotyping drug cocktail in healthy volunteers. J. Acquir. Immune Defic. Syndr. 2006, 42, 52–60. [CrossRef]
169. Haslemo, T.; Eikeseth, P.H.; Tanum, L.; Molden, E.; Refsum, H. The effect of variable cigarette consumption
on the interaction with clozapine and olanzapine. Eur. J. Clin. Pharmacol. 2006, 62, 1049–1053. [CrossRef]
170. Seppälä, N.H.; Leinonen, E.V.; Lehtonen, M.L.; Kivistö, K.T. Clozapine serum concentrations are lower in
smoking than in non-smoking schizophrenic patients. Pharmacol. Toxicol. 1999, 85, 244–246. [CrossRef]
171. Fric, M.; Pfuhlmann, B.; Laux, G.; Riederer, P.; Distler, G.; Artmann, S.; Wohlschläger, M.; Liebmann, M.;
Deckert, J. The influence of smoking on the serum level of duloxetine. Pharmacopsychiatry 2008, 41, 151–155.
[CrossRef] [PubMed]
172. Cassidenti, D.L.; Vijod, A.G.; Vijod, M.A.; Stanczyk, F.Z.; Lobo, R.A. Short-term effects of smoking on the
pharmacokinetic profiles of micronized estradiol in postmenopausal women. Am. J. Obstet. Gynecol. 1990,
163, 1953–1960. [CrossRef]
173. Sitsen, J.; Maris, F.; Timmer, C. Drug-drug interaction studies with mirtazapine and carbamazepine in healthy
male subjects. Eur. J. Drug Metab. Pharmacokinet. 2001, 26, 109–121. [CrossRef] [PubMed]
174. Lind, A.-B.; Reis, M.; Bengtsson, F.; Jonzier-Perey, M.; Powell Golay, K.; Ahlner, J.; Baumann, P.; Dahl, M.-L.
Steady-state concentrations of mirtazapine, N-desmethylmirtazapine, 8-hydroxymirtazapine and their
enantiomers in relation to cytochrome P450 2D6 genotype, age and smoking behaviour. Clin. Pharmacokinet.
2009, 48, 63–70. [CrossRef] [PubMed]
175. Sun, L.; McDonnell, D.; Yu, M.; Kumar, V.; Moltke, L. von. A Phase I Open-Label Study to Evaluate the
Effects of Rifampin on the Pharmacokinetics of Olanzapine and Samidorphan Administered in Combination
in Healthy Human Subjects. Clin. Drug Investig. 2019, 39, 477–484. [CrossRef]
176. Wu, T.-H.; Chiu, C.-C.; Shen, W.W.; Lin, F.-W.; Wang, L.-H.; Chen, H.-Y.; Lu, M.-L. Pharmacokinetics of
olanzapine in Chinese male schizophrenic patients with various smoking behaviors. Prog. Neuropsychopharmacol.
Biol. Psychiatry 2008, 32, 1889–1893. [CrossRef] [PubMed]
177. Zevin, S.; Benowitz, N.L. Drug interactions with tobacco smoking. An update. Clin. Pharmacokinet. 1999,
36, 425–438. [CrossRef]
178. Fuhr, U.; Woodcock, B.G.; Siewert, M. Verapamil and drug metabolism by the cytochrome P450 isoform
CYP1A2. Eur. J. Clin. Pharmacol. 1992, 42, 463–464. [CrossRef]

204
Pharmaceutics 2020, 12, 1201

179. Jungmann, N.A.; Lang, D.; Saleh, S.; van der Mey, D.; Gerisch, M. In vitro-in vivo correlation of the drug-drug
interaction potential of antiretroviral HIV treatment regimens on CYP1A1 substrate riociguat. Expert Opin.
Drug Metab. Toxicol. 2019, 15, 975–984. [CrossRef]
180. Kroon, L.A. Drug interactions with smoking. Am. J. Health Syst. Pharm. 2007, 64, 1917–1921. [CrossRef]
181. Hiemke, C.; Bergemann, N.; Clement, H.W.; Conca, A.; Deckert, J.; Domschke, K.; Eckermann, G.;
Egberts, K.; Gerlach, M.; Greiner, C.; et al. Consensus Guidelines for Therapeutic Drug Monitoring
in Neuropsychopharmacology: Update 2017. Pharmacopsychiatry 2018, 51, 9–62. [CrossRef] [PubMed]
182. Faber, M.S.; Fuhr, U. Time response of cytochrome P450 1A2 activity on cessation of heavy smoking.
Clin. Pharmacol. Ther. 2004, 76, 178–184. [CrossRef] [PubMed]
183. Faber, M.S.; Jetter, A.; Fuhr, U. Assessment of CYP1A2 activity in clinical practice: Why, how, and when?
Basic Clin. Pharmacol. Toxicol. 2005, 97, 125–134. [CrossRef]
184. Chochol, M.D.; Kataria, L.; O’Rourke, M.C.; Lamotte, G. Clozapine-Associated Myoclonus and Stuttering
Secondary to Smoking Cessation and Drug Interaction: A Case Report. J. Clin. Psychopharmacol. 2019,
39, 275–277. [CrossRef] [PubMed]
185. Kocar, T.; Freudenmann, R.W.; Spitzer, M.; Graf, H. Switching From Tobacco Smoking to Electronic Cigarettes
and the Impact on Clozapine Levels. J. Clin. Psychopharmacol. 2018, 38, 528–529. [CrossRef]
186. Larsen, J.T.; Brøsen, K. Consumption of charcoal-broiled meat as an experimental tool for discerning
CYP1A2-mediated drug metabolism in vivo. Basic Clin. Pharmacol. Toxicol. 2005, 97, 141–148.
[CrossRef] [PubMed]
187. Hakooz, N.; Hamdan, I. Effects of dietary broccoli on human in vivo caffeine metabolism: A pilot study on a
group of Jordanian volunteers. Curr. Drug Metab. 2007, 8, 9–15. [CrossRef]
188. Charron, C.S.; Novotny, J.A.; Jeffery, E.H.; Kramer, M.; Ross, S.A.; Seifried, H.E. Consumption of baby kale
increased cytochrome P450 1A2 (CYP1A2) activity and influenced bilirubin metabolism in a randomized
clinical trial. J. Funct. Foods 2020, 64, 103624. [CrossRef]
189. Lake, B.G.; Tredger, J.M.; Renwick, A.B.; Barton, P.T.; Price, R.J. 3,3 -Diindolylmethane induces CYP1A2 in
cultured precision-cut human liver slices. Xenobiotica 1998, 28, 803–811. [CrossRef]

Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

205
 pharmaceutics

Article
Oral Proteasomal Inhibitors Ixazomib, Oprozomib, and
Delanzomib Upregulate the Function of Organic Anion
Transporter 3 (OAT3): Implications in OAT3-Mediated
Drug-Drug Interactions
Yunzhou Fan, Zhengxuan Liang, Jinghui Zhang and Guofeng You *

Department of Pharmaceutics, Rutgers, The State University of New Jersey, 160 Frelinghuysen Road, Piscataway,
NJ 08854, USA; yunzhou.fan@rutgers.edu (Y.F.); zhengxuan.liang@rutgers.edu (Z.L.);
jinghui.zhang@rutgers.edu (J.Z.)
* Correspondence: gyou@pharmacy.rutgers.edu; Tel.: +1-848-445-6349

Abstract: Organic anion transporter 3 (OAT3) is mainly expressed at the basolateral membrane of
kidney proximal tubules, and is involved in the renal elimination of various kinds of important drugs,
potentially affecting drug efficacy or toxicity. Our laboratory previously reported that ubiquitin
modification of OAT3 triggers the endocytosis of OAT3 from the plasma membrane to intracellular
endosomes, followed by degradation. Oral anticancer drugs ixazomib, oprozomib, and delanzomib,
as proteasomal inhibitors, target the ubiquitin–proteasome system in clinics. Therefore, this study
investigated the effects of ixazomib, oprozomib, and delanzomib on the expression and transport
activity of OAT3 and elucidated the underlying mechanisms. We showed that all three drugs sig-
Citation: Fan, Y.; Liang, Z.; Zhang, J.;
nificantly increased the accumulation of ubiquitinated OAT3, which was consistent with decreased
You, G. Oral Proteasomal Inhibitors
intracellular 20S proteasomal activity; stimulated OAT3-mediated transport of estrone sulfate and
Ixazomib, Oprozomib, and
Delanzomib Upregulate the Function
p-aminohippuric acid; and increased OAT3 surface expression. The enhanced transport activity and
of Organic Anion Transporter 3 OAT3 expression following drug treatment resulted from an increase in maximum transport velocity
(OAT3): Implications in of OAT3 without altering the substrate binding affinity, and from a decreased OAT3 degradation.
OAT3-Mediated Drug-Drug Together, our study discovered a novel role of anticancer agents ixazomib, oprozomib, and delan-
Interactions. Pharmaceutics 2021, 13, zomib in upregulating OAT3 function, unveiled the proteasome as a promising target for OAT3
314. https://doi.org/10.3390/ regulation, and provided implication of OAT3-mediated drug–drug interactions, which should be
pharmaceutics13030314 warned against during combination therapies with proteasome inhibitor drugs.

Academic Editor: Dong Hyun Kim Keywords: drug transporter; ubiquitination; ixazomib; regulation

Received: 31 July 2020

Accepted: 24 February 2021
Published: 28 February 2021
1. Introduction
Publisher’s Note: MDPI stays neutral
Organic anion transporter 3 (OAT3), which is encoded by the SLC22A8 gene, is pri-
with regard to jurisdictional claims in
marily expressed at the basolateral membrane of kidney proximal tubules, and actively
published maps and institutional affil- translocates corresponding substrates from the blood into renal tubule epithelial cells.
iations. Those substrates are then effluxed out of the apical membrane into urine by other trans-
porters [1–3]. OAT3 is involved in the renal elimination of various kinds of important
clinical drugs from the kidney, such as anticancer agents (e.g., methotrexate), antivirals (e.g.,
tenofovir, valacyclovir), antibiotics (e.g., benzylpenicillin, cefotaxime), antihypertensives
Copyright: © 2021 by the authors.
(e.g., furosemide, sitagliptin), H2 receptor antagonists (e.g., cimetidine, famotidine), and
Licensee MDPI, Basel, Switzerland.
nonsteroidal anti-inflammatory drugs (e.g., ketoprofen, ibuprofen) [4–6]. Therefore, the
This article is an open access article
renal OAT3 function is a critical determinant in drug clearance out of the body, and in the
distributed under the terms and pharmacokinetic and pharmacodynamic properties of drugs, which ultimately affect the
conditions of the Creative Commons drugs’ efficacy and systemic or renal toxicity.
Attribution (CC BY) license (https:// Combination therapies by coadministration of different drugs are often used for treat-
creativecommons.org/licenses/by/ ment of a single or multiple diseases. If one drug is an inhibitor, substrate, or inducer of
4.0/). OAT3, it will inhibit uncompetitively or competitively, or stimulate the renal transport and

207

Pharmaceutics 2021, 13, 314. https://doi.org/10.3390/pharmaceutics13030314 https://www.mdpi.com/journal/pharmaceutics

Pharmaceutics 2021, 13, 314

excretion of other clinical substrates, cause potential drug-drug interactions (DDIs), and
sequent intra- and interindividual variation in clinical response to drugs [6,7]. Transporter-
mediated clinical DDIs have attracted the attention of academic, industrial, and regulatory
agencies. OAT3-mediated DDIs abundantly exist between imipenem-cilastatin, piperacillin-
tazobactam, bezafibrate-mizoribine, puerarin-methotrexate, benzylpenicillin–acyclovir, etc.,
and markedly alter the pharmacokinetic parameters of affected drugs [8–12]. Through
inhibition of OAT1/3, probenecid, wedelolactone, and wogonin prevented the kidney accu-
mulation of aristolochic acid and related nephropathy, apigenin- or cilastatin-ameliorated
imipenem, or diclofenac-induced nephrotoxicity [13–17].
The transporter expression and function may be modulated by certain drugs, phy-
tomedicines, or xenobiotics, resulting in altered disposition of clinical substances, which
is an indirect manner of obtaining transporter-mediated DDIs, in contrast to direct inter-
action with the transporter-like inhibitors or substrates [4]. For example, administration
of 1α,25-dihydroxyvitamin D3, mercuric chloride, or methotrexate decreased OAT3 ex-
pression in crude or basolateral membranes of rat kidneys; while the renal expression
was increased in normal rats by ochratoxin A treatment, in diabetic rats by insulin or
atorvastatin plus insulin treatment, or in obese rats by prebiotic Lactobacillus paracasei HII01
or xylooligosaccharide treatment [18–25].
OAT3 expression and activity can be regulated through posttranslational modifica-
tions, including phosphorylation, ubiquitination, and SUMOylation [26–28]. As ubiqui-
tination of OAT3 is an initiating process that triggers the internalization of OAT3 from
the plasma membrane to intracellular endosomes, it is a critical molecular mechanism for
OAT3 regulation [29,30]. Our lab demonstrated that activation of protein kinase C (PKC)
could enhance OAT3 ubiquitination, and accelerate OAT3 internalization and subsequent
degradation [27]. The transport activity and quantity of OAT3 on the plasma membrane
were then reduced. Since proteasome inhibition can affect ubiquitination of targeted pro-
teins and degradation, modulation of proteasome activity could potentially interfere with
the physiological function of transporters. Proteasome inhibitors have shown to influ-
ence the copper transporter 1, Na+ /H+ exchanger-3, ATP-binding cassette transporters A1
(ABCA1) and ABCG1, organic-anion-transporting polypeptide (OATP) 1B3, metal trans-
porter ZIP14, and OAT1 [31–36]. However, it is not clear whether OAT3 can be regulated by
controlling proteasome activity. Ixazomib, oprozomib, and delanzomib are oral proteasome
inhibitors that target the ubiquitin–proteasome system for multiple myeloma treatment. In
the present study, we investigated the influence of ixazomib, oprozomib, and delanzomib
on OAT3 expression and transport activity, and elucidated the underlying mechanisms.

2. Materials and Methods

2.1. Materials
COS-7 cells and HEK293 cells were purchased from ATCC (Manassas, VA, USA). [3 H]-
labeled estrone sulfate (ES) and [3 H]-labeled p-aminohippuric acid (PAH) were ordered
from PerkinElmer (Waltham, MA, USA). Mouse anti-Myc antibody (9E10) was purchased
from Roche (Indianapolis, IN, USA). Mouse anti-E-Cadherin antibody was from Abcam
(Cambridge, MA, USA). Streptavidin agarose resin, protein G agarose, and Sulfo-NHS-SS-
biotin were ordered from Thermo Scientific (Rockford, IL, USA). The 20S proteasome assay
kit was ordered from Cayman Chemical Company (Ann Arbor, MI, USA). Mouse anti-β-
actin antibody, normal mouse IgG, and mouse anti-ubiquitin antibody were obtained from
Santa Cruz Biotechnology (Dallas, TX, USA). Ixazomib, oprozomib, and delanzomib were
purchased from Selleck Chemicals (Houston, TX, USA). Probenecid, lactacystin, epoxomicin
and all other reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA).

2.2. Cell Culture

Parental COS-7 and parental HEK293 cells were cultured in Dulbecco’s modified
Eagle’s medium (DMEM) (Corning, Tewksbury, MA, USA) supplemented with 10% fe-
tal bovine serum (Gibco, Grand Island, NY, USA) at 37 ◦ C in 5% CO2 . Human OAT3-

208
Pharmaceutics 2021, 13, 314

expressing (hOAT3) COS-7 cells and hOAT3-expressing HEK293 cells were established in
our group [37,38]. The hOAT3 cells were cultured in DMEM supplemented with 10% fetal
bovine serum and 0.2 mg/mL G418 sulfate (Gibco, Grand Island, NY, USA).

2.3. Transport Measurement

The transport activity was assayed using the method published by our lab [30]. Cells
per well were incubated in uptake solution of [3 H]ES (250 nM) or [3 H]PAH (20 μM) in
phosphate-buffered saline (PBS)/Ca2+ /Mg2+ (PBS/CM) for 3 min. After discarding the
uptake solution, the cells were washed twice with cold PBS, then lysed in NaOH solution
(0.2 N) and neutralized by adding HCl solution (0.2 N). The amount of ES or PAH uptake
was assayed using a Beckman LS 6500 liquid scintillation counter.

2.4. 20S Proteasome Activity Assay

After incubation with ixazomib, oprozomib, delanzomib, or lactacystin for 6 h, hOAT3
cells were washed once with assay buffer (200 μL) and solubilized in lysis buffer (100 μL).
Then, the supernatant (90 μL) was removed to a black 96-well plate, and incubated with
SUC-LLVY-AMC solution (10 μL) for 1 h at 37 ◦ C. Fluorescence intensity per well (excita-
tion = 360 nm, emission = 480 nm) was assayed using a Molecular Devices Spectramax M3
microplate reader.

2.5. Cell-Surface Biotinylation

Cell surface hOAT3 expression was assayed using the procedures introduced by our
group [39]. The hOAT3 cells were labeled with sulfo-NHS-SS-biotin solution (0.5 mg/mL
in PBS/CM) on ice, with slow shaking for two continuous 20 min. After discarding the
biotin solution, the cells were washed once with glycine solution (100 mM in PBS/CM)
and incubated with glycine solution for 20 min to completely quench the unbound sulfo-
NHS-SS-biotin. The cells were then lysed in lysis buffer consisted of 10 mM Tris-HCl,
pH 7.5, 150 mM NaCl, 1 mM EDTA, 0.1% SDS, 1% Triton X-100, and 1% proteinase inhibitor
cocktail. The cell lysates were centrifuged at 16,000× g at 4 ◦ C, and the supernatant was
then mixed with streptavidin agarose resin (40 μL) to separate the cell surface proteins. The
hOAT3 at the cell surface was detected by immunoblotting using the anti-Myc antibody.

2.6. Immunoprecipitation
The hOAT3 ubiquitination was investigated using the method published by our
group [39]. The hOAT3 cells were lysed in lysis buffer consisted of 50 mM Tris-HCl,
pH 8.0, 150 mM NaCl, 1% Triton X-100, 10% glycerol, 5 mM EDTA, 1 mM NaF, 20 mM
N-ethylmaleimide, and 1% of proteinase inhibitor cocktail. Cell lysates were precleared
with protein G agarose to decrease nonspecific binding at 4 ◦ C for 2 h. Anti-Myc antibody
was mixed with protein G agarose (30 μL) and incubated at 4 ◦ C for 2 h. The precleared
protein was then added to antibody-bound protein G agarose suspension and mixed with
end-over-end rotation at 4 ◦ C overnight. Proteins coupled to protein G agarose were
released with urea buffer containing β-mecaptoethanol and detected by immunoblotting
using the anti-ubiquitin antibody.

2.7. Degradation Assay of OAT3

The hOAT3 degradation was investigated using the method utilized in our group [38].
The hOAT3 cells were first labeled with sulfo-NHS-SS-biotin, then the biotinylated cells
were treated with vehicle, ixazomib, oprozomib, or delanzomib at 37 ◦ C for 0, 3, and 6 h.
Then the cells were collected, and the undegraded cell surface hOAT3 was isolated and
detected following the procedures in Section 2.5.

2.8. Electrophoresis and Immunoblotting

The electrophoresis and immunoblotting experiments were carried out using the
method published by our group [30]. Protein samples were loaded on 7.5% precast

209
Pharmaceutics 2021, 13, 314

polyacrylamide gels and transferred onto polyvinylidene difluoride membranes. The

immunoblot membranes were blocked with 5% nonfat dry milk in PBS-0.05% tween 20 for
1 h, and incubated with primary antibodies at 4 ◦ C overnight, followed by incubation of
horseradish peroxidase-conjugated secondary antibodies. The protein bands were visual-
ized using a SuperSignal West Dura Extended Duration Substrate kit (Thermo Scientific,
Rockford, IL, USA), and corresponding densities were analyzed using the FluorChem
8000 imaging system (Alpha Innotech Corp., San Leandro, CA, USA).

2.9. Data Analysis

One-way ANOVA or two-way ANOVA Tukey’s test was utilized for statistical analysis
among multiple groups by using GraphPad Prism software (GraphPad Software Inc.,
San Diego, CA, USA). Each experiment was repeated at least three times. A p value less
than 0.05 was statistically significant, and a p value more than 0.05 was not statistically
significant (ns).

3. Results
3.1. Effects of Ixazomib, Oprozomib, and Delanzomib on the Ubiquitination of OAT3
Ixazomib, oprozomib, and delanzomib, as proteasome inhibitors, target the ubiquitin-
proteasome system for cancer therapy. First, we investigated their effects on the intracellular
ubiquitination of OAT3 in OAT3-expressing COS-7 cells. OAT3-expressing cells were
treated with ixazomib, oprozomib, or delanzomib for 6 h, then harvested and lysed.
OAT3 was pulled down from cell lysate by anti-Myc antibody (Myc tag was fused onto
OAT3, enabling immunodetection), followed by immunoblotting (IB) using anti-ubiquitin
antibody to probe the ubiquitinated OAT3. The results (Figure 1) showed that incubating
cells with ixazomib, oprozomib, or delanzomib resulted in a significant accumulation
of ubiquitinated OAT3, which was not because of the difference in immunoprecipitated
OAT3, since there were similar quantities of OAT3 pulled down from all samples. Further
study showed that like lactacystin, a classical proteasome inhibitor, ixazomib, oprozomib,
and delanzomib inhibited the 20S proteasome activity by 50% (95% confidence interval
(CI): 46% to 54%), 87% (95% CI: 83% to 91%), and 61% (95% CI: 57% to 65%), respectively,
after 6 h of treatment (Figure 2). Therefore, the accumulation of ubiquitinated OAT3 was
attributed to the decreased proteasome activity, suggesting that ubiquitinated OAT3 can be
modulated by interfering the ubiquitin–proteasome system in the cell model that we used.

Figure 1. Effect of proteasomal inhibitors ixazomib, oprozomib, or delanzomib on the accumulation

of ubiquitinated OAT3. Top panel: OAT3-expressing COS-7 cells were treated with ixazomib (30 nM),
oprozomib (200 nM), or delanzomib (30 nM) for 6 h. Treated cells were then lysed, and OAT3 was
immunoprecipitated with anti-Myc antibody or mouse IgG (as negative control, lane 5), followed
by IB with anti-Ub. Bottom panel: The same immunoblot from the top panel was reprobed with
anti-Myc antibody to determine the amount of OAT3 immunoprecipitated.

210
Pharmaceutics 2021, 13, 314

Figure 2. Effect of ixazomib, oprozomib, or delanzomib on the 20S proteasome activity. OAT3-
expressing COS-7 cells were treated with lactacystin (10 μM), a classical proteasome inhibitor as
positive control, ixazomib (30 nM), oprozomib (200 nM), or delanzomib (30 nM) for 6 h. The 20S
proteasome activity of cells was then performed. The 20S proteasome activity was expressed as the
percentage of control cells from three independent experiments. Values are means ± S.D. (n = 3).
* p < 0.05.

3.2. Cis-Effect of Ixazomib, Oprozomib, or Delanzomib on OAT3-Mediated Uptake of Estrone Sulfate

As OAT3 has multispecificity toward multiple substrates, we investigated whether
ixazomib, oprozomib, and delanzomib are inhibitors or inducers of OAT3 by performing a
cis-inhibition assay. Estrone sulfate (ES) is a prototypical OAT3 substrate, and probenecid
is a well-recognized competitive inhibitor of OAT3 [2,40]. We measured 3 min of uptake
of [3 H]ES (250 nM) into OAT3-expressing cells with or without probenecid, ixazomib,
oprozomib, or delanzomib existing in the ES solution. The results (Figure 3) showed that
probenecid inhibited OAT3-mediated transport of [3 H]ES by 40% (95% CI: 34% to 46%),
while ixazomib, oprozomib, and delanzomib did not have any effects, indicating that
ixazomib, oprozomib, and delanzomib are not inhibitors or inducers of OAT3. There-
fore, ixazomib, oprozomib, and delanzomib did not affect OAT3 function through direct
interaction with the transporter.

Figure 3. Cis-effect of ixazomib, oprozomib, or delanzomib on OAT3-mediated uptake of [3 H]ES.

The uptake of [3 H]ES (250 nM) in the presence of ixazomib (1 μM), oprozomib (1 μM), delanzomib
(1 μM), or probenecid (5 μM) for 3 min was measured in OAT3-expressing COS-7 cells. Each data
point represented only carrier-mediated transport after subtraction of values from parental cells.
Uptake activity was expressed as the percentage of uptake measured in control cells from three
independent experiments. Values are means ± S.D. (n = 3). * p < 0.05; ns = not statistically significant.

211
Pharmaceutics 2021, 13, 314

3.3. Effects of Ixazomib, Oprozomib, or Delanzomib on OAT3-Mediated Uptake of Estrone Sulfate

or P-Aminohippuric Acid
Since ixazomib, oprozomib, and delanzomib can increase OAT3 ubiquitination, we
further investigated their effect on the transport activity. OAT3-expressing cells were treated
with ixazomib, oprozomib, or delanzomib for 6 h, then OAT3-mediated uptake of ES was
measured. The results (Figure 4A–C) showed that ixazomib, oprozomib, and delanzomib
all induced a dose-dependent stimulation of ES uptake in OAT3-expressing COS-7 cells.
The OAT3 transport activity was stimulated by 72% (95% CI: 46% to 97%), 45% (95% CI:
33% to 56%), and 48% (95% CI: 31% to 64%) at 30 nM ixazomib, 200 nM oprozomib, and
30 nM delanzomib, respectively. Consistently, 6 h of treatment with classical proteasome
inhibitors lactacystin or epoxomicin stimulated the uptake of ES (Figure 4D). Besides, p-
aminohippuric acid (PAH) is another OAT3 substrate [41,42]. Like ES, our result (Figure 5)
showed that all three proteasome inhibitor drugs also significantly stimulated PAH uptake
in a substrate-independent manner. Similar stimulative effects also existed in OAT3-
expressing HEK293 cells, excluding cell-specific effects of proteasome inhibitors (Figure 6).
Further study showed that 10~40 nM ixazomib induced a dose-dependent inhibition of
proteasome activity (Figure 7A), and there was a strongly association between transport
activity and proteasomal activity (correlation coefficient was 0.98, Figure 7B). We selected
the concentration of 30 nM ixazomib, 200 nM oprozomib, and 30 nM delanzomib for the
following mechanisms study.

Figure 4. Effect of ixazomib, oprozomib, delanzomib, or classical proteasome inhibitors on OAT3

activity. OAT3-expressing COS-7 cells were treated with ixazomib (A), oprozomib (B), delanzomib
(C), or classical proteasome inhibitors lactacystin or epoxomicin (D) at indicated concentrations for
6 h. The uptake of [3 H]ES (250 nM) for 3 minu was then performed. Each data point represented
only carrier-mediated transport after subtraction of values from parental cells. Uptake activity was
expressed as the percentage of uptake measured in control cells from three independent experiments.
Values are means ± S.D. (n = 3). * p < 0.05; ns = not statistically significant.

212
Pharmaceutics 2021, 13, 314

Figure 5. Effect of ixazomib, oprozomib, or delanzomib on OAT3-mediated transport of p-aminohippuric

acid. OAT3-expressing COS-7 cells were treated with ixazomib (30 nM), oprozomib (200 nM), or delan-
zomib (30 nM) for 6 h. The uptake of [3 H]PAH (20 μM) for 3 min was then performed. Each data point
represented only carrier-mediated transport after subtraction of values from parental cells. Uptake activity
was expressed as the percentage of uptake measured in control cells from three independent experiments.
Values are means ± S.D. (n = 3). * p < 0.05.

Figure 6. Effect of ixazomib, oprozomib, or delanzomib on OAT3 activity in OAT3-expressing HEK293

cells. OAT3-expressing HEK293 cells were treated with ixazomib (30 nM), oprozomib (200 nM), or
delanzomib (30 nM) for 6 h. The uptake of [3 H]ES (250 nM) for 3 min was then performed. Each data
point represented only carrier-mediated transport after subtraction of values from parental cells. Uptake
activity was expressed as the percentage of uptake measured in control cells from three independent
experiments. Values are means ± S.D. (n = 3). * p < 0.05.

3.4. Kinetic Analysis of the Effects of Ixazomib, Oprozomib, or Delanzomib on OAT3-Mediated

Uptake of Estrone Sulfate
To examine the mechanism of ixazomib-, oprozomib-, and delanzomib-induced stim-
ulation of OAT3 activity, we determined [3 H]ES uptake at a series of concentrations
(0.3~10 μM). Eadie–Hofstee analyses of the derived data (Figure 8) showed that incubation
of ixazomib (Figure 8A), oprozomib (Figure 8B), or delanzomib (Figure 8C) resulted in an in-
creased maximum transport velocity Vmax (128 ± 3 pmol·mg−1 ·3 min−1 with untreated cells
and 176 ± 7 pmol·mg−1 ·3 min−1 in the presence of ixazomib; 130 ± 2 pmol·mg−1 ·3 min−1
with untreated cells and 223 ± 4 pmol·mg−1 ·3 min−1 in the presence of oprozomib;
128 ± 11 pmol·mg−1 ·3 min−1 with untreated cells and 175 ± 11 pmol·mg−1 ·3 min−1
in the presence of delanzomib), with no significant change of substrate-binding-affinity
Km for ES (4.2 ± 0.3 μM with untreated cells and 4.6 ± 0.4 μM in the presence of ixa-
zomib; 4.6 ± 0.1 μM with untreated cells and 6.0 ± 0.2 μM in the presence of oprozomib;
4.3 ± 0.9 μM with untreated cells and 4.9 ± 0.7 μM in the presence of delanzomib). These
results indicated that stimulated activity of ixazomib, oprozomib, and delanzomib resulted
from an increase of the transport rate of OAT3, and not from an enhanced affinity at the
substrate-binding site.

213
Pharmaceutics 2021, 13, 314

Figure 7. Dose-effect of ixazomib on the 20S proteasome activity. (A) OAT3-expressing COS-7 cells
were treated with ixazomib at indicated concentrations for 6 h. The 20S proteasome activity of cells
was then performed. The 20S proteasome activity was expressed as the percentage of control cells
from three independent experiments. Values are means ± S.D. (n = 3). * p < 0.05. (B) Correlation
analysis was performed between transport activity from Figure 4A and proteasomal activity from
Figure 7A after ixazomib treatment.

Figure 8. Effect of ixazomib, oprozomib, or delanzomib on the kinetics of hOAT3-mediated estrone

sulfate transport. OAT3-expressing COS-7 cells were treated with 30 nM ixazomib (A), 200 nM
oprozomib (B), or 30 nM delanzomib (C) for 6 h, and initial uptake (3 min) of [3 H]ES was measured
at the concentration of 0.3~10 μM. The data represent uptake into hOAT3-expressing cells minus
uptake into mock cells (parental COS-7 cells). Values are means ± S.D. (n = 3). V = velocity;
S = substrate concentration.

214
Pharmaceutics 2021, 13, 314

3.5. Effect of Ixazomib, Oprozomib, or Delanzomib on OAT3 Expression

As ixazomib, oprozomib, and delanzomib did not alter the binding affinity of OAT3,
the increase of transport activity may mainly result from the increased expression on the
plasma membrane. OAT3-expressing cells were treated with ixazomib, oprozomib, or
delanzomib for 6 h, and OAT3 expression on the plasma membrane and in the whole cell
lysates were investigated. The result showed that treatment with ixazomib, oprozomib, or
delanzomib all caused an increase of OAT3 expression on the cell membrane (Figure 9A,B)
and in the whole cell lysate (Figure 9C,D), which was not because of the overall interference
in cellular proteins, as there were similar quantities of membrane fraction marker E-
Cadherin (Figure 9A) and whole cellular fraction marker β-actin (Figure 9C) in all samples.

Figure 9. Effect of ixazomib, oprozomib, or delanzomib on OAT3 expression. (A) Top panel: OAT3-
expressing COS-7 cells were treated with ixazomib (30 nM), oprozomib (200 nM), or delanzomib
(30 nM) for 6 h. Cell-surface biotinylation was performed. Biotinylated (cell surface) proteins were
separated with using streptavidin agarose resin and analyzed by IB with an anti-Myc antibody.
Bottom panel: The same blot from the top panel was reprobed with an anti-E-Cadherin antibody.
E-Cadherin is an integral membrane protein marker. (B) Densitometry plot of results from (A),
top panel, as well as from other experiments. Values are means ± S.D. (n = 3). * p < 0.05. (C) Top
panel: OAT3-expressing COS-7 cells were treated with ixazomib (30 nM), oprozomib (200 nM), or
delanzomib (30 nM) for 6 h. Cells were then lysed, followed by IB with anti-Myc antibody. Bottom
panel: The same blot from the top panel was reprobed with an anti-β-actin antibody. β-actin is a
cellular protein marker. (D) Densitometry plot of results from (C), top panel, as well as from other
experiments. Values are means ± S.D. (n = 3). * p < 0.05.

3.6. Effect of Ixazomib, Oprozomib, and Delanzomib on OAT3 Degradation

The ubiquitin–proteasome pathway ultimately affects the degradation of targeted pro-
teins, therefore the degradation of cell-membrane OAT3 was investigated by biotinylation
and isolation of cell-surface proteins. OAT3-expressing cells were first labeled with sulfo-
NHS-SS-biotin on all membrane proteins at 4 ◦ C, then biotinylated cells were incubated
with ixazomib, oprozomib, or delanzomib for 3 and 6 h at 37 ◦ C. At the time points, those
cells were harvested and lysed, and cell-membrane proteins were enriched in streptavidin
agarose beads, followed by immunoblotting detection of OAT3 using anti-Myc antibody.
215
Pharmaceutics 2021, 13, 314

The results (Figure 10) revealed that compared to control, the degradation of cell membrane
OAT3 was reduced markedly after 6 h incubation of the three drugs, and without effect
at 3 h, indicating that ixazomib, oprozomib, and delanzomib chronically enhanced the
stability of membrane OAT3.

Figure 10. Effect of ixazomib, oprozomib, or delanzomib on OAT3 stability. (A) OAT3-expressing
COS-7 cells were biotinylated with membrane-impermeable biotinylation reagent sulfo-NHS-SS-
biotin. Labeled cells were then treated with ixazomib (30 nM), oprozomib (200 nM), or delanzomib
(30 nM) at 37 ◦ C for 3 and 6 h, respectively. Treated cells were lysed, and cell-surface proteins were iso-
lated using streptavidin-agarose resin, followed by IB with anti-Myc antibody. (B) Densitometry plot
of results from (A), as well as from other experiments. The amount of undegraded cell-surface hOAT3
was expressed as % of total initial cell-surface hOAT3 pool. Values are means ± S.D. (n = 3). * p < 0.05;
ns = not statistically significant. Two-way ANOVA Tukey’s test was applied for statistical analysis.

4. Discussion
OAT3 function is predominantly dependent on the amount located on the plasma
membrane, which can be regulated by mitogen-activated protein kinase (MAPK), protein
kinase A (PKA), PKC signaling pathways [43–45]. Ubiquitination is a significant post-
translational mechanism for OAT3 regulation. Our previous study had demonstrated the
essential role of Nedd4-2 (a ubiquitin ligase) in the ubiquitination, surface expression, and
transport activity of OAT3 [27]. Serum- and glucocorticoid-inducible kinases 1 (sgk1), PKC,
janus tyrosine kinase 2 (JAK2) regulated OAT3 through Nedd4-2, which showed Nedd4-2
is molecular target for OAT3 regulation [27,37,38,46]. In this study, we further discovered
proteasome was a novel target for regulation of OAT3 and stimulating OAT3 function can
be achieved through inhibiting proteasomal activity.
COS-7 and HEK293 cells lacking in endogenous OATs were commonly utilized as
heterologous expression systems for OATs. Both cell lines were broadly selected for study
the regulation and mechanisms of the cloned OATs and other drug transporters in kidney
with several advantages [13,47–49]. Expression of exogenous OAT3 will allow us to study
the transport characteristics of OAT3 without being disturbed by other OATs. They are
originated from the kidney and have the proteasome activity and signaling pathways
involved in OAT3 regulation. COS-7 cells and HEK293 cells used in our studies will
provide the research basis for the upcoming work focusing on validating whether primary
epithelia possess the similar mechanisms.
Ixazomib is an FDA-approved anticancer drug, while oprozomib and delanzomib
are in phases of clinical trials. All of them are administered orally, and preferentially
bind reversibly (ixazomib and delanzomib) or irreversibly (oprozomib) and inhibit the
chymotrypsin-like activity of the 20S proteasome in various tissues and organs. There were
reports that ixazomib inhibited the proteasome activity in the whole blood and tumor; opro-
zomib could inhibit the proteasome activity in the blood, peripheral blood mononuclear
cells, liver, kidney, and adrenal glands; and delanzomib inhibited the proteasome activity
in blood mononuclear cells, kidney, and spleen [50–55]. Ixazomib prevented antibody-
mediated rejection in kidney transplantation and treated patients with metastatic kidney

216
Pharmaceutics 2021, 13, 314

cancer [56,57]. Delanzomib can ameliorate lupus nephritis in mice [55]. These results sug-
gested that proteasomal inhibitors can be used to treat kidney diseases, through proteasome
inhibition-mediated reduction in aberrant cytokines and antibodies, or downregulation of
nuclear factor kappa B-dependent gene expression and resulted tumor growth [58,59].
Ixazomib, oprozomib, or delanzomib treatment substantially increased the accumula-
tion of ubiquitinated OAT3 (Figure 1), which was consistent with decreased 20S proteaso-
mal activity in cell lysate in OAT3-expressing cells (Figure 2), stimulated OAT3-mediated
transport of estrone sulfate and p-aminohippuric acid (Figures 4–6), and increased OAT3
membrane expression (Figure 9). The enhanced transport activity of OAT3 following drug
pretreatment resulted from an increase in maximum transport velocity without altering the
binding affinity of the transporter (Figure 8). Ubiquitinated OAT3 exhibited the molecular
mass above 180 kDa, ~100 kDa more than OAT3 (~80 kDa). As ubiquitin is an 8-kDa
polypeptide, OAT3 may be modified by poly- or multiubiquitination (Figure 1).
The OAT3 function was chronically stimulated with 6 h of treatment with ixazomib,
oprozomib, or delanzomib. As the alteration of trafficking processes, including internal-
ization or recycling of OATs, can be reflected in function change during acute regulation
(such as 0.5 h), we can exclude the reduced internalization and increased recycling that
are the underlying mechanisms for those drugs [27–30,39]. With further exploring, the
degradation of OAT3 was decelerated by ixazomib, oprozomib, or delanzomib (Figure 10).
Our results showed they inhibited the 20S proteasome activity (Figure 2), and there was
a negative association between proteasomal activity and transport activity at 10–40 nM
ixazomib (Figure 7B). Together, ixazomib-, oprozomib-, and delanzomib-upregulated
OAT3 function was mainly through suppression of proteasome activity and decelerated
degradation of OAT3.
The concentrations of ixazomib (10–40 nM), oprozomib (50–400 nM), and delanzomib
(10–50 nM) used in our study are in the clinically therapeutic range. After once-weekly
oral dosing of 2.23 mg/m2 for 3 weeks in combination therapy with lenalidomide and
dexamethasone, the mean maximum plasma concentration (Cmax ) of ixazomib in multi-
ple myeloma patients at day 1 and day 15 was 22.3 ng/mL (61.7 nM) and 31.4 ng/mL
(87.0 nM), respectively [60]. For oprozomib, after 2 consecutive days weekly oral dosing at
210 mg/day for 4 weeks plus pomalidomide and dexamethasone in relapsed/refractory
multiple myeloma patients, the mean Cmax of oprozomib at day 1 and day 8 was 744 ng/mL
(1.4 μM) and 1030 ng/mL (1.9 μM), respectively [61]. Until now, there were only reports
about intravenous pharmacokinetic data of delanzomib in human. After 2 days weekly
intravenous dosing 0.4–1.8 mg/m2 for 2 weeks in patients with solid tumors and multiple
myeloma, the mean Cmax of delanzomib on day 1 was 88.4–557.3 ng/mL (0.2–1.3 μM) [54].
Ixazomib and delanzomib have a long terminal plasma half-life of 3.6–11.3 days and
62.0 ± 43.5 h, respectively [54,62]. Though oprozomib has a short plasma half-life of about
1 h resulting from rapid systemic clearance, the recovery of proteasome activity in tissues
needed a longer time of 24~72 h due to irreversible binding [63,64]. Therefore, the inductive
effects of ixazomib, oprozomib, and delanzomib on drug elimination and DDIs potentially
exist, though they are administered once or twice weekly. The in vitro regulation and
related mechanisms in cell models were reported in this study, and further in vivo study
in Sprague Dawley rats by oral ixazomib will be performed to further explore the roles
of ixazomib in proteasome activity, OTA3 ubiquitination, drug uptake in kidney slices,
membrane and total expression in kidney, and renal clearance of drugs by kidney in our lab.
Ixazomib, oprozomib, or delanzomib are all indicated in combination with dexamethasone,
a synthetic glucocorticoid for the treatment of patients with multiple myeloma [61,65,66]. Our
previous study showed dexamethasone stimulates OAT3 transport activity and membrane
expression through the serum- and glucocorticoid-inducible kinases 1 signaling pathway,
suggesting the stimulatory effect on OAT3 may be further magnified using ixazomib,
oprozomib, and delanzomib in combination with dexamethasone [37].
Ixazomib is a low-affinity substrate of P-glycoprotein (P-gp); is not a substrate of
breast cancer resistance protein (BCRP), multidrug resistance protein 2 (MRP2), or hep-

217
Pharmaceutics 2021, 13, 314

atic OATPs; and is not an inhibitor of P-gp, BCRP, MRP2, OATP1B1, OATP1B3, organic
cation transporter 2 (OCT2), OAT1, OAT3, multidrug and toxin extruder 1 (MATE1), or
MATE2-K. Therefore, the manufacturer claimed that ixazomib is not expected to cause
transporter-mediated drug–drug interactions [67]. Consistent with this, our study found
that ixazomib is not an inhibitor of OAT3 (Figure 3). However, although ixazomib did
not cause DDIs through direct interaction (inhibiting or competing) with the transporters,
our study showed that ixazomib can upregulate OAT3 activity through induced mem-
brane expression, which may affect the disposition of other drugs in an indirect manner of
transporter-mediated DDIs. Besides, potential DDIs may be occurred by direct OATs in-
duction. There were reports that ursolic acid and ciprofloxacin stimulated OAT1-mediated
p-aminohippuric acid uptake, and 1,5-dicaffeoylquinic acid and 18β-glycyrrhetinic acid
stimulated hOAT4-mediated estrone sulfate uptake [68,69].
Proteasome inhibition drugs are now well utilized for cancer treatment. In contrast,
impaired proteasome function and related elevation of toxic intracellular protein or aggre-
gates are involved in neurodegenerative disorders (e.g., Parkinson’s disease, Alzheimer’s
disease) and cardiac dysfunctions, and enhancement of proteasome activity may also be a
promising therapeutic strategy for those diseases [70–72]. PD169316, pyrazolones and chlor-
promazine as small molecules, were found to be proteasome activators [70,73,74]. It would
be interesting to study whether proteasomal activators can regulate the OAT3 function.
Our findings that oral proteasome inhibitors ixazomib, oprozomib, and delanzomib
can increase OAT3 transport activity have important physiological implications. First, it can
accelerate the drugs clearance from body, resulting in reduced plasma concentration and
therapeutic efficacy of drugs. We can also use this mechanism for noninvasive detoxification
in the event of drug overdoses. Second, it may enhance the entering and distribution
of drugs in proximal tubular cells, leading to potential nephrotoxicity. Those points
should attract the attention of physicians and pharmacists for rational use of medicines
and irrational drug combinations, and avoiding potential drug–drug interactions. Third,
bilateral ureteral obstruction (BUO), a common clinical disease, impaired renal elimination
of drugs partly resulted from reduced cell-surface expression of OAT3 [75]. Proteasome
inhibition may provide a potential strategy to reverse BUO or other kidney-disease-induced
downregulation of OAT3. Last, it also can promote renal clearance of toxins, metabolites,
signaling molecules, nutrients, and other substances as OAT3 substrates, and maintain
homeostasis within the body.

5. Conclusions
Our studies showed for the first time that anticancer drugs ixazomib, oprozomib, and
delanzomib had a critical role in upregulating OAT3 transport activity and expression,
suggesting their potential impact on the OAT3-mediated drug disposition and clinical
drug–drug interactions during combination therapies of proteasome inhibitor drugs and
other types of drugs (Figure 11).

Figure 11. Oral proteasomal inhibitors ixazomib, oprozomib, and delanzomib upregulate the trans-
port activity and expression of OAT3. Ub = ubiquitin; DDIs = drug–drug interactions.

218
Pharmaceutics 2021, 13, 314

Author Contributions: Conceptualization, Y.F. and G.Y.; methodology, Y.F.; software, Y.F.; vali-
dation, Y.F.; formal analysis, Y.F.; investigation, Y.F., Z.L., J.Z.; resources, Y.F.; data curation, Y.F.;
writing—original draft preparation, Y.F.; writing—review and editing, G.Y.; supervision, G.Y.; project
administration, G.Y.; funding acquisition, G.Y. All authors have read and agreed to the published
version of the manuscript.
Funding: This research was funded by grants (to Guofeng You) from National Institute of General
Medical Sciences (R01-GM079123 and R01-GM127788).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. You, G. Structure, function, and regulation of renal organic anion transporters. Med. Res. Rev. 2002, 22, 602–616. [CrossRef] [PubMed]
2. Nigam, S.K.; Bush, K.T.; Martovetsky, G.; Ahn, S.Y.; Liu, H.C.; Richard, E.; Bhatnagar, V.; Wu, W. The organic anion transporter
(OAT) family: A systems biology perspective. Physiol. Rev. 2015, 95, 83–123. [CrossRef]
3. Wang, L.; Sweet, D.H. Renal organic anion transporters (SLC22 family): Expression, regulation, roles in toxicity, and impact on
injury and disease. AAPS J. 2013, 15, 53–69. [CrossRef]
4. Liang, Y.; Li, S.; Chen, L. The physiological role of drug transporters. Protein Cell 2015, 6, 334–350. [CrossRef] [PubMed]
5. Burckhardt, G. Drug transport by Organic Anion Transporters (OATs). Pharmacol. Ther. 2012, 136, 106–130. [CrossRef]
6. Nigam, S.K. What do drug transporters really do? Nat. Rev. Drug Discov. 2015, 14, 29–44. [CrossRef]
7. The International Transporter Consortium; Giacomini, K.M.; Huang, S.M.; Tweedie, D.J.; Benet, L.Z.; Brouwer, K.L.; Chu, X.;
Dahlin, A.; Evers, R.; Fischer, V.; et al. Membrane transporters in drug development. Nat. Rev. Drug Discov. 2010,
9, 215–236. [CrossRef]
8. Zhu, Y.; Huo, X.; Wang, C.; Meng, Q.; Liu, Z.; Sun, H.; Tan, A.; Ma, X.; Peng, J.; Liu, K. Organic anion transporters also mediate
the drug-drug interaction between imipenem and cilastatin. Asian J. Pharm. Sci. 2020, 15, 252–263. [CrossRef] [PubMed]
9. Wen, S.; Wang, C.; Duan, Y.; Huo, X.; Meng, Q.; Liu, Z.; Yang, S.; Zhu, Y.; Sun, H.; Ma, X.; et al. OAT1 and OAT3 also mediate the
drug-drug interaction between piperacillin and tazobactam. Int. J. Pharm. 2018, 537, 172–182. [CrossRef]
10. Feng, Y.; Wang, C.; Liu, Q.; Meng, Q.; Huo, X.; Liu, Z.; Sun, P.; Yang, X.; Sun, H.; Qin, J.; et al. Bezafibrate-mizoribine interaction:
Involvement of organic anion transporters OAT1 and OAT3 in rats. Eur. J. Pharm. Sci. 2016, 81, 119–128. [CrossRef] [PubMed]
11. Liu, Q.; Wang, C.; Meng, Q.; Huo, X.; Sun, H.; Peng, J.; Ma, X.; Sun, P.; Liu, K. MDR1 and OAT1/OAT3 mediate the drug-drug
interaction between puerarin and methotrexate. Pharm. Res. 2014, 31, 1120–1132. [CrossRef]
12. Ye, J.; Liu, Q.; Wang, C.; Meng, Q.; Sun, H.; Peng, J.; Ma, X.; Liu, K. Benzylpenicillin inhibits the renal excretion of acyclovir by
OAT1 and OAT3. Pharmacol. Rep. 2013, 65, 505–512. [CrossRef]
13. Xue, X.; Gong, L.K.; Maeda, K.; Luan, Y.; Qi, X.M.; Sugiyama, Y.; Ren, J. Critical role of organic anion transporters 1 and 3 in
kidney accumulation and toxicity of aristolochic acid I. Mol. Pharm. 2011, 8, 2183–2192. [CrossRef]
14. Li, C.; Wang, X.; Bi, Y.; Yu, H.; Wei, J.; Zhang, Y.; Han, L.; Zhang, Y. Potent Inhibitors of Organic Anion Transporters 1 and 3 from
natural compounds and their protective effect on aristolochic acid nephropathy. Toxicol. Sci. 2020, 175, 279–291. [CrossRef]
15. Huo, X.; Meng, Q.; Wang, C.; Zhu, Y.; Liu, Z.; Ma, X.; Ma, X.; Peng, J.; Sun, H.; Liu, K. Cilastatin protects against imipenem-induced
nephrotoxicity via inhibition of renal organic anion transporters (OATs). Acta Pharm. Sin. B 2019, 9, 986–996. [CrossRef]
16. Huo, X.; Meng, Q.; Wang, C.; Wu, J.; Wang, C.; Zhu, Y.; Ma, X.; Sun, H.; Liu, K. Protective effect of cilastatin against diclofenac-
induced nephrotoxicity through interaction with diclofenac acyl glucuronide via organic anion transporters. Br. J. Pharmacol.
2020, 177, 1933–1948. [CrossRef] [PubMed]
17. Huo, X.; Meng, Q.; Wang, C.; Wu, J.; Zhu, Y.; Sun, P.; Ma, X.; Sun, H.; Liu, K. Targeting renal OATs to develop renal protective
agent from traditional Chinese medicines: Protective effect of Apigenin against Imipenem-induced nephrotoxicity. Phytother. Res.
2020. [CrossRef] [PubMed]
18. Miao, Q.; Liu, Q.; Wang, C.; Meng, Q.; Guo, X.; Sun, H.; Peng, J.; Ma, X.; Kaku, T.; Liu, K. Inhibitory effect of 1alpha,25-
dihydroxyvitamin D(3) on excretion of JBP485 via organic anion transporters in rats. Eur. J. Pharm. Sci. 2013, 48, 351–359. [CrossRef]
19. Di Giusto, G.; Anzai, N.; Ruiz, M.L.; Endou, H.; Torres, A.M. Expression and function of Oat1 and Oat3 in rat kidney exposed to
mercuric chloride. Arch. Toxicol. 2009, 83, 887–897. [CrossRef]
20. Shibayama, Y.; Ushinohama, K.; Ikeda, R.; Yoshikawa, Y.; Motoya, T.; Takeda, Y.; Yamada, K. Effect of methotrexate treatment on
expression levels of multidrug resistance protein 2, breast cancer resistance protein and organic anion transporters Oat1, Oat2
and Oat3 in rats. Cancer Sci. 2006, 97, 1260–1266. [CrossRef] [PubMed]
21. Zlender, V.; Breljak, D.; Ljubojevic, M.; Flajs, D.; Balen, D.; Brzica, H.; Domijan, A.M.; Peraica, M.; Fuchs, R.; Anzai, N.; et al. Low
doses of ochratoxin A upregulate the protein expression of organic anion transporters Oat1, Oat2, Oat3 and Oat5 in rat kidney
cortex. Toxicol. Appl. Pharmacol. 2009, 239, 284–296. [CrossRef] [PubMed]

219
Pharmaceutics 2021, 13, 314

22. Phatchawan, A.; Chutima, S.; Varanuj, C.; Anusorn, L. Decreased renal organic anion transporter 3 expression in type 1 diabetic
rats. Am. J. Med. Sci. 2014, 347, 221–227. [CrossRef] [PubMed]
23. Thongnak, L.; Pongchaidecha, A.; Jaikumkao, K.; Chatsudthipong, V.; Chattipakorn, N.; Lungkaphin, A. The additive effects
of atorvastatin and insulin on renal function and renal organic anion transporter 3 function in diabetic rats. Sci. Rep. 2017,
7, 13532. [CrossRef]
24. Wanchai, K.; Yasom, S.; Tunapong, W.; Chunchai, T.; Eaimworawuthikul, S.; Thiennimitr, P.; Chaiyasut, C.; Pongchaidecha, A.;
Chatsudthipong, V.; Chattipakorn, S.; et al. Probiotic Lactobacillus paracasei HII01 protects rats against obese-insulin resistance-
induced kidney injury and impaired renal organic anion transporter 3 function. Clin. Sci. (Lond.) 2018, 132, 1545–1563.
[CrossRef] [PubMed]
25. Wanchai, K.; Yasom, S.; Tunapong, W.; Chunchai, T.; Thiennimitr, P.; Chaiyasut, C.; Pongchaidecha, A.; Chatsudthipong, V.;
Chattipakorn, S.; Chattipakorn, N.; et al. Prebiotic prevents impaired kidney and renal Oat3 functions in obese rats. J. Endocrinol.
2018, 237, 29–42. [CrossRef] [PubMed]
26. Zhang, J.; Yu, Z.; You, G. Insulin-like growth factor 1 modulates the phosphorylation, expression, and activity of organic anion
transporter 3 through protein kinase A signaling pathway. Acta Pharm. Sin. B 2020, 10, 186–194. [CrossRef] [PubMed]
27. Xu, D.; Wang, H.; You, G. An Essential Role of Nedd4-2 in the Ubiquitination, expression, and function of organic anion
Transporter-3. Mol. Pharm. 2016, 13, 621–630. [CrossRef]
28. Wang, H.; Zhang, J.; You, G. Activation of protein kinase a stimulates SUMOylation, expression, and transport activity of organic
anion transporter 3. AAPS J. 2019, 21, 30. [CrossRef] [PubMed]
29. Zhang, Q.; Suh, W.; Pan, Z.; You, G. Short-term and long-term effects of protein kinase C on the trafficking and stability of human
organic anion transporter 3. Int. J. Biochem. Mol. Biol. 2012, 3, 242–249.
30. Zhang, Q.; Hong, M.; Duan, P.; Pan, Z.; Ma, J.; You, G. Organic anion transporter OAT1 undergoes constitutive and protein kinase
C-regulated trafficking through a dynamin- and clathrin-dependent pathway. J. Biol. Chem. 2008, 283, 32570–32579. [CrossRef]
31. Jandial, D.D.; Farshchi-Heydari, S.; Larson, C.A.; Elliott, G.I.; Wrasidlo, W.J.; Howell, S.B. Enhanced delivery of cisplatin to
intraperitoneal ovarian carcinomas mediated by the effects of bortezomib on the human copper transporter 1. Clin. Cancer Res.
2009, 15, 553–560. [CrossRef]
32. Hu, M.C.; Di Sole, F.; Zhang, J.; McLeroy, P.; Moe, O.W. Chronic regulation of the renal Na(+)/H(+) exchanger NHE3 by dopamine:
Translational and posttranslational mechanisms. Am. J. Physiol. Renal. Physiol. 2013, 304, F1169–F1180. [CrossRef] [PubMed]
33. Ogura, M.; Ayaori, M.; Terao, Y.; Hisada, T.; Iizuka, M.; Takiguchi, S.; Uto-Kondo, H.; Yakushiji, E.; Nakaya, K.; Sasaki, M.; et al.
Proteasomal inhibition promotes ATP-binding cassette transporter A1 (ABCA1) and ABCG1 expression and cholesterol efflux
from macrophages in vitro and in vivo. Arterioscler. Thromb. Vasc. Biol. 2011, 31, 1980–1987. [CrossRef] [PubMed]
34. Alam, K.; Farasyn, T.; Crowe, A.; Ding, K.; Yue, W. Treatment with proteasome inhibitor bortezomib decreases organic anion
transporting polypeptide (OATP) 1B3-mediated transport in a substrate-dependent manner. PLoS ONE 2017, 12, e0186924.
[CrossRef] [PubMed]
35. Zhao, N.; Zhang, A.S.; Worthen, C.; Knutson, M.D.; Enns, C.A. An iron-regulated and glycosylation-dependent proteasomal
degradation pathway for the plasma membrane metal transporter ZIP14. Proc. Natl. Acad. Sci. USA 2014, 111, 9175–9180.
[CrossRef] [PubMed]
36. Fan, Y.; You, G. Proteasome Inhibitors Bortezomib and carfilzomib stimulate the transport activity of human organic anion
transporter 1. Mol. Pharmacol. 2020, 97, 384–391. [CrossRef]
37. Wang, H.; Liu, C.; You, G. The activity of organic anion transporter-3: Role of dexamethasone. J. Pharmacol. Sci. 2018, 136, 79–85. [CrossRef]
38. Zhang, J.; Liu, C.; You, G. AG490, a JAK2-specific inhibitor, downregulates the expression and activity of organic anion
transporter-3. J. Pharmacol. Sci. 2018, 136, 142–148. [CrossRef]
39. Zhang, Q.; Li, S.; Patterson, C.; You, G. Lysine 48-linked polyubiquitination of organic anion transporter-1 is essential for its
protein kinase C-regulated endocytosis. Mol. Pharmacol. 2013, 83, 217–224. [CrossRef]
40. Wang, C.; Wang, C.; Liu, Q.; Meng, Q.; Cang, J.; Sun, H.; Peng, J.; Ma, X.; Huo, X.; Liu, K. Aspirin and probenecid inhibit
organic anion transporter 3-mediated renal uptake of cilostazol and probenecid induces metabolism of cilostazol in the rat.
Drug Metab. Dispos. 2014, 42, 996–1007. [CrossRef]
41. Antonescu, I.E.; Karlgren, M.; Pedersen, M.L.; Simoff, I.; Bergstrom, C.A.S.; Neuhoff, S.; Artursson, P.; Steffansen, B.; Nielsen, C.U.
Acamprosate is a substrate of the human organic anion transporter (OAT) 1 without OAT3 Inhibitory properties: Implications for
renal acamprosate secretion and drug-drug interactions. Pharmaceutics 2020, 12, 390. [CrossRef] [PubMed]
42. Sweet, D.H.; Miller, D.S.; Pritchard, J.B.; Fujiwara, Y.; Beier, D.R.; Nigam, S.K. Impaired organic anion transport in kidney and
choroid plexus of organic anion transporter 3 (Oat3 (Slc22a8)) knockout mice. J. Biol. Chem. 2002, 277, 26934–26943. [CrossRef]
43. Barros, S.A.; Srimaroeng, C.; Perry, J.L.; Walden, R.; Dembla-Rajpal, N.; Sweet, D.H.; Pritchard, J.B. Activation of protein kinase
Czeta increases OAT1 (SLC22A6)- and OAT3 (SLC22A8)-mediated transport. J. Biol. Chem. 2009, 284, 2672–2679. [CrossRef]
44. Soodvilai, S.; Chatsudthipong, V.; Evans, K.K.; Wright, S.H.; Dantzler, W.H. Acute regulation of OAT3-mediated estrone sulfate
transport in isolated rabbit renal proximal tubules. Am. J. Physiol. Renal. Physiol. 2004, 287, F1021–F1029. [CrossRef]
45. Soodvilai, S.; Wright, S.H.; Dantzler, W.H.; Chatsudthipong, V. Involvement of tyrosine kinase and PI3K in the regulation
of OAT3-mediated estrone sulfate transport in isolated rabbit renal proximal tubules. Am. J. Physiol. Renal. Physiol. 2005,
289, F1057–F1064. [CrossRef]

220
Pharmaceutics 2021, 13, 314

46. Wang, H.; You, G. SGK1/Nedd4-2 signaling pathway regulates the activity of human organic anion transporters 3.
Biopharm. Drug Dispos. 2017, 38, 449–457. [CrossRef]
47. El-Sheikh, A.A.; Greupink, R.; Wortelboer, H.M.; van den Heuvel, J.J.; Schreurs, M.; Koenderink, J.B.; Masereeuw, R.; Russel, F.G.
Interaction of immunosuppressive drugs with human organic anion transporter (OAT) 1 and OAT3, and multidrug resistance-
associated protein (MRP) 2 and MRP4. Transl. Res. 2013, 162, 398–409. [CrossRef]
48. Bhardwaj, R.K.; Herrera-Ruiz, D.; Eltoukhy, N.; Saad, M.; Knipp, G.T. The functional evaluation of human peptide/histidine
transporter 1 (hPHT1) in transiently transfected COS-7 cells. Eur. J. Pharm. Sci. 2006, 27, 533–542. [CrossRef]
49. Goyal, S.; Vanden Heuvel, G.; Aronson, P.S. Renal expression of novel Na+/H+ exchanger isoform NHE8. Am. J. Physiol.
Renal. Physiol. 2003, 284, F467–F473. [CrossRef] [PubMed]
50. Gupta, N.; Hanley, M.J.; Xia, C.; Labotka, R.; Harvey, R.D.; Venkatakrishnan, K. Clinical Pharmacology of Ixazomib: The first oral
proteasome inhibitor. Clin. Pharm. 2019, 58, 431–449. [CrossRef] [PubMed]
51. FDA. Pharmacology Review(s) for NINLARO® (Ixazomib). Available online: https://www.accessdata.fda.gov/drugsatfda_
docs/nda/2015/208462Orig1s000PharmR.pdf (accessed on 29 July 2020).
52. Infante, J.R.; Mendelson, D.S.; Burris, H.A., 3rd; Bendell, J.C.; Tolcher, A.W.; Gordon, M.S.; Gillenwater, H.H.; Arastu-Kapur, S.;
Wong, H.L.; Papadopoulos, K.P. A first-in-human dose-escalation study of the oral proteasome inhibitor oprozomib in patients
with advanced solid tumors. Investig. New Drugs 2016, 34, 216–224. [CrossRef]
53. Muchamuel, T.; Basler, M.; Aujay, M.A.; Suzuki, E.; Kalim, K.W.; Lauer, C.; Sylvain, C.; Ring, E.R.; Shields, J.; Jiang, J.; et al.
A selective inhibitor of the immunoproteasome subunit LMP7 blocks cytokine production and attenuates progression of
experimental arthritis. Nat. Med. 2009, 15, 781–787. [CrossRef]
54. Gallerani, E.; Zucchetti, M.; Brunelli, D.; Marangon, E.; Noberasco, C.; Hess, D.; Delmonte, A.; Martinelli, G.; Bohm, S.;
Driessen, C.; et al. A first in human phase I study of the proteasome inhibitor CEP-18770 in patients with advanced solid tumours
and multiple myeloma. Eur. J. Cancer 2013, 49, 290–296. [CrossRef]
55. Seavey, M.M.; Lu, L.D.; Stump, K.L.; Wallace, N.H.; Ruggeri, B.A. Novel, orally active, proteasome inhibitor, delanzomib (CEP-
18770), ameliorates disease symptoms and glomerulonephritis in two preclinical mouse models of SLE. Int. Immunopharmacol.
2012, 12, 257–270. [CrossRef]
56. Reese, S.R.; Wilson, N.A.; Huang, G.; Redfield, R.R., 3rd; Zhong, W.; Djamali, A. Calcineurin Inhibitor Minimization with
Ixazomib, an investigational proteasome inhibitor, for the prevention of antibody mediated rejection in a preclinical model.
Transplantation 2015, 99, 1785–1795. [CrossRef] [PubMed]
57. Msaouel, P.; Carugo, A.; Genovese, G. Targeting proteostasis and autophagy in SMARCB1-deficient malignancies: Where next?
Oncotarget 2019, 10, 3979–3981. [CrossRef]
58. Adams, J. The proteasome: A suitable antineoplastic target. Nat. Rev. Cancer 2004, 4, 349–360. [CrossRef] [PubMed]
59. Neubert, K.; Meister, S.; Moser, K.; Weisel, F.; Maseda, D.; Amann, K.; Wiethe, C.; Winkler, T.H.; Kalden, J.R.; Manz, R.A.; et al.
The proteasome inhibitor bortezomib depletes plasma cells and protects mice with lupus-like disease from nephritis. Nat. Med.
2008, 14, 748–755. [CrossRef] [PubMed]
60. FDA. Clinical Pharmacology Biopharmaceutics Review(s) for NINLARO® (Ixazomib). Available online: https://www.accessdata.
fda.gov/drugsatfda_docs/nda/2015/208462Orig1s000ClinPharmR.pdf (accessed on 29 July 2020).
61. Shah, J.; Usmani, S.; Stadtmauer, E.A.; Rifkin, R.M.; Berenson, J.R.; Berdeja, J.G.; Lyons, R.M.; Klippel, Z.; Chang, Y.L.; Niesvizky, R.
Oprozomib, pomalidomide, and Dexamethasone in Patients With Relapsed and/or Refractory Multiple Myeloma. Clin. Lymphoma
Myeloma Leuk. 2019, 19, 570–578.e571. [CrossRef] [PubMed]
62. Kumar, S.K.; Bensinger, W.I.; Zimmerman, T.M.; Reeder, C.B.; Berenson, J.R.; Berg, D.; Hui, A.M.; Gupta, N.; Di Bacco, A.;
Yu, J.; et al. Phase 1 study of weekly dosing with the investigational oral proteasome inhibitor ixazomib in relapsed/refractory
multiple myeloma. Blood 2014, 124, 1047–1055. [CrossRef] [PubMed]
63. Zhou, H.J.; Aujay, M.A.; Bennett, M.K.; Dajee, M.; Demo, S.D.; Fang, Y.; Ho, M.N.; Jiang, J.; Kirk, C.J.; Laidig, G.J.; et al. Design
and synthesis of an orally bioavailable and selective peptide epoxyketone proteasome inhibitor (PR-047). J. Med. Chem. 2009,
52, 3028–3038. [CrossRef] [PubMed]
64. Ou, Y.; Xu, Y.; Gore, L.; Harvey, R.D.; Mita, A.; Papadopoulos, K.P.; Wang, Z.; Cutler, R.E., Jr.; Pinchasik, D.E.; Tsimberidou, A.M.
Physiologically-based pharmacokinetic modelling to predict oprozomib CYP3A drug-drug interaction potential in patients with
advanced malignancies. Br. J. Clin. Pharmacol. 2019, 85, 530–539. [CrossRef] [PubMed]
65. Moreau, P.; Masszi, T.; Grzasko, N.; Bahlis, N.J.; Hansson, M.; Pour, L.; Sandhu, I.; Ganly, P.; Baker, B.W.; Jackson, S.R.; et al. Oral
Ixazomib, Lenalidomide, and Dexamethasone for Multiple Myeloma. N. Engl. J. Med. 2016, 374, 1621–1634. [CrossRef]
66. Sanchez, E.; Li, M.; Li, J.; Wang, C.; Chen, H.; Jones-Bolin, S.; Hunter, K.; Ruggeri, B.; Berenson, J.R. CEP-18770 (delan-
zomib) in combination with dexamethasone and lenalidomide inhibits the growth of multiple myeloma. Leuk. Res. 2012,
36, 1422–1427. [CrossRef]
67. FDA. Label Revision for NINLARO® (Ixazomib). Available online: https://www.accessdata.fda.gov/drugsatfda_docs/label/20
20/208462s006lbl.pdf (accessed on 29 July 2020).
68. Vanwert, A.L.; Srimaroeng, C.; Sweet, D.H. Organic anion transporter 3 (oat3/slc22a8) interacts with carboxyfluoroquinolones,
and deletion increases systemic exposure to ciprofloxacin. Mol. Pharmacol. 2008, 74, 122–131. [CrossRef] [PubMed]
69. Wang, L.; Sweet, D.H. Interaction of natural dietary and herbal anionic compounds and flavonoids with human organic anion
transporters 1 (SLC22A6), 3 (SLC22A8), and 4 (SLC22A11). Evid. Based Complement Altern. Med. 2013, 2013, 612527. [CrossRef]

221
Pharmaceutics 2021, 13, 314

70. Leestemaker, Y.; de Jong, A.; Witting, K.F.; Penning, R.; Schuurman, K.; Rodenko, B.; Zaal, E.A.; van de Kooij, B.; Laufer, S.;
Heck, A.J.R.; et al. Proteasome activation by small molecules. Cell Chem. Biol. 2017, 24, 725–736.e727. [CrossRef] [PubMed]
71. Kors, S.; Geijtenbeek, K.; Reits, E.; Schipper-Krom, S. Regulation of proteasome activity by (Post-)transcriptional mechanisms.
Front. Mol. Biosci. 2019, 6, 48. [CrossRef] [PubMed]
72. Njomen, E.; Tepe, J.J. Proteasome activation as a new therapeutic approach to target proteotoxic disorders. J. Med. Chem. 2019,
62, 6469–6481. [CrossRef]
73. Trippier, P.C.; Zhao, K.T.; Fox, S.G.; Schiefer, I.T.; Benmohamed, R.; Moran, J.; Kirsch, D.R.; Morimoto, R.I.; Silverman, R.B.
Proteasome activation is a mechanism for pyrazolone small molecules displaying therapeutic potential in amyotrophic lateral
sclerosis. ACS Chem. Neurosci. 2014, 5, 823–829. [CrossRef]
74. Jones, C.L.; Njomen, E.; Sjogren, B.; Dexheimer, T.S.; Tepe, J.J. Small molecule enhancement of 20S proteasome activity targets
intrinsically disordered proteins. ACS Chem. Biol. 2017, 12, 2240–2247. [CrossRef] [PubMed]
75. Villar, S.R.; Brandoni, A.; Anzai, N.; Endou, H.; Torres, A.M. Altered expression of rat renal cortical OAT1 and OAT3 in response
to bilateral ureteral obstruction. Kidney Int. 2005, 68, 2704–2713. [CrossRef] [PubMed]

222
 pharmaceutics

Article
In Silico Prediction of Drug–Drug Interactions Mediated by
Cytochrome P450 Isoforms
Alexander V. Dmitriev *, Anastassia V. Rudik, Dmitry A. Karasev, Pavel V. Pogodin, Alexey A. Lagunin, Dmitry
A. Filimonov and Vladimir V. Poroikov

Laboratory of Structure-Function Based Drug Design, Department of Bioinformatics, Institute of Biomedical

Chemistry, Pogodinskaya Str. 10, bldg. 8, 119121 Moscow, Russia; rudik_anastassia@mail.ru (A.V.R.);
dmitry.karasev@ibmc.msk.ru (D.A.K.); pogodinpv@gmail.com (P.V.P.); alexey.lagunin@ibmc.msk.ru (A.A.L.);
dmitry.filimonov@ibmc.msk.ru (D.A.F.); vladimir.poroikov@ibmc.msk.ru (V.V.P.)
* Correspondence: a.v.dmitriev@mail.ru; Tel.: +7-499-246-3029

Abstract: Drug–drug interactions (DDIs) can cause drug toxicities, reduced pharmacological effects,
and adverse drug reactions. Studies aiming to determine the possible DDIs for an investigational
drug are part of the drug discovery and development process and include an assessment of the DDIs
potential mediated by inhibition or induction of the most important drug-metabolizing cytochrome
P450 isoforms. Our study was dedicated to creating a computer model for prediction of the DDIs
mediated by the seven most important P450 cytochromes: CYP1A2, CYP2B6, CYP2C19, CYP2C8,
CYP2C9, CYP2D6, and CYP3A4. For the creation of structure–activity relationship (SAR) models
that predict metabolism-mediated DDIs for pairs of molecules, we applied the Prediction of Activity
Spectra for Substances (PASS) software and Pairs of Substances Multilevel Neighborhoods of Atoms
(PoSMNA) descriptors calculated based on structural formulas. About 2500 records on DDIs medi-
Citation: Dmitriev, A.V.; Rudik, A.V.;
Karasev, D.A.; Pogodin, P.V.; Lagunin,
ated by these cytochromes were used as a training set. Prediction can be carried out both for known
A.A.; Filimonov, D.A.; Poroikov, V.V. drugs and for new, not-yet-synthesized substances. The average accuracy of the prediction of DDIs
In Silico Prediction of Drug–Drug mediated by various isoforms of cytochrome P450 estimated by leave-one-out cross-validation (LOO
Interactions Mediated by Cytochrome CV) procedures was about 0.92. The SAR models created are publicly available as a web resource
P450 Isoforms. Pharmaceutics 2021, 13, and provide predictions of DDIs mediated by the most important cytochromes P450.
538. https://doi.org/10.3390/
pharmaceutics13040538 Keywords: drug interaction; DDI; computational prediction; in silico; QSAR; drug metabolism;
ADME; pharmacokinetics; CYP; polypharmacy; metabolic DDI; P450; 1A2; 2B6; 2C19; 2C8; 2C9;
Academic Editors: Dong Hyun Kim
2D6; 3A4
and Sangkyu Lee

Received: 16 February 2021

Accepted: 8 April 2021
1. Introduction
Published: 13 April 2021
For the treatment of complex disorders, patients often take multiple medications at
Publisher’s Note: MDPI stays neutral
the same time, which potentially cause drug–drug interactions (DDIs). Usually, DDIs are
with regard to jurisdictional claims in
divided into three types: pharmaceutical, pharmacodynamic, and pharmacokinetic [1].
published maps and institutional affil- Pharmaceutical DDIs may appear due to physical or chemical interactions, for example,
iations. when drugs are mixed in a syringe before infusion, and such DDIs are rare. Pharmaco-
dynamic DDIs may occur when a pair or more co-administered drugs act on the same
physiological system or target. Pharmacokinetic DDIs are very common and occur when
one of the drugs (“violator” or “precipitant” drug) affects the absorption, distribution,
metabolism, or excretion of another drug (“victim” or “object” drug). Such DDIs provoke
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
an increase or a decrease in the exposure of an object drug and lead to a change in drug
This article is an open access article
pharmacological action. In this study, we focused on the pharmacokinetic DDIs at the
distributed under the terms and metabolism level (biotransformation), the so-called “metabolic DDIs.”
conditions of the Creative Commons The most common drug-metabolizing enzymes (DMEs) in the first phase of xenobiotic
Attribution (CC BY) license (https:// metabolism in the human body are several isoforms of the cytochrome P450 superfam-
creativecommons.org/licenses/by/ ily. The U.S. Department of Health and Human Services Food and Drug Administra-
4.0/). tion Center for Drug Evaluation and Research (FDA CDER) requires determining which

223

Pharmaceutics 2021, 13, 538. https://doi.org/10.3390/pharmaceutics13040538 https://www.mdpi.com/journal/pharmaceutics

Pharmaceutics 2021, 13, 538

drug-metabolizing enzymes (CYP3A, CYP2D6, CYP2C19, CYP2C9, CYP2C8, CYP2B6, or

CYP1A2) metabolize the investigational drug during in vitro studies of metabolic DDIs
estimates [2].
In silico methods can help prioritize drug discovery efforts by guiding, but not replac-
ing, in vitro and in vivo experiments. Previously, we presented a comprehensive review
of the methods for predicting the DDIs related to the inhibition or induction of DMEs [3].
Most of such in silico methods predict DDIs indirectly. A recently presented machine
learning (ML) method used different molecular fingerprints to classify compounds as
inhibitors or noninhibitors of five major cytochrome P450 isoenzymes [4]. Ligand-based
and structure-based methods dealing with substrates, inhibitors [5,6], and inducers [7] of
particular DMEs. Results of prediction could help to determine possible DDIs. However,
such conclusions are not sufficiently reliable, as the pairs of substances that are substrates
and inhibitors (or inducers) of DMEs may not exhibit DDIs. On the other hand, DDIs
have often occurred between substances that could act as substrates, inducers, and in-
hibitors (that may act by various inhibition mechanisms); for example, this is a widespread
case for cytochrome P450 CYP3A4 [8]. At best, a pair of potentially exhibiting DDI sub-
stances should be considered together during prediction as the whole entity. However,
previously developed ligand-based and structure-based computational methods did not
consider two substances in pairs simultaneously. Direct DDIs estimation methods for
the pairs of substances include structure resemblance and functional similarities methods
and literature-based DDIs prediction methods [9–14]. These methods deal with the pairs
of substances but require information about the pharmacokinetics and pharmacodynam-
ics [9,14], interaction profile, target and side-effects [10,13], and the phenotypic, therapeutic,
chemical, and genomic properties [11] of substances or medical records [12]. It is clear that
for new, not-yet-synthesized, and virtual substances, such information does not exist. The
results of predictions of this group of methods [9–14] have often been presented as data sets
containing a bulk conglomerate of information about potential DDIs predicted between
the existing drugs. Such examples include 430,128 [10], 145,068 [13], and over 250,000 [14]
records of unknown potential DDIs in the sets of predicted results. However, this bulk of
information concerning drug pairs is provided without assessment of the possibility of
DDIs manifestation.
The current study aimed to create the computational structure–activity relationship
(SAR) models to predict metabolic DDIs mediated by CYP1A2, CYP2B6, CYP2C19, CYP2C8,
CYP2C9, CYP2D6, or CYP3A4. We have previously developed models for DDIs sever-
ity prediction [15,16] that used the PASS (Prediction of Activity Spectra for Substances)
program and PoSMNA (Pairs of Substances Multilevel Neighborhoods of Atoms) substruc-
tural descriptors. These models were able to predict the classes of DDIs severity for pairs of
molecules according to OpeRational ClassificAtion (ORCA). In the current study, we used
the same methods and descriptors but implemented them to predict whether two molecules
would manifest metabolic DDIs mediated by the seven cytochromes mentioned above.
Due to the limited possibilities of creating an appropriate training set, the stereochemical
features of molecules were not taken into account by our descriptors. In addition, in the cur-
rent realization of the method, DDI predictions were obtained in qualitative mode (“YES”
or “NO”). Unlike other ligand-based and structure-based methods [4–8], our approach
operated with two substances in pairs at once. This is reasonable for the DDI phenomenon,
in which two substances interact simultaneously. It gives a direct indication of DDIs for
the pairs of molecules without suggestions of the role of particular compounds, which
is not always obvious (without consideration of inhibition or induction of a particular
enzyme). In contrast to structure resemblance, functional similarities, and literature-based
DDIs prediction methods [9–14], our prediction method uses only structural formulas
of compounds; it does not require any information about their biological activity. This
means that our method can be applied for not-yet-investigated, new, and virtual substances.
Moreover, our method provides a probabilistic assessment of possible DDIs and evaluates
the possibility of DDIs manifestation for predicted pairs.

224
Pharmaceutics 2021, 13, 538

2. Materials and Methods

2.1. Information on DDIs and Training Set Creation
We used DDIs data collected from two sources of information. The first source was
DrugBank Version 4.1 (University of Alberta and The Metabolomics Innovation Centre,
Edmonton, AB, Canada) [17] that contains information about interactions derived from
public drug databases. The second source of DDIs data was the Fujitsu ADME Database
(Chemistry & Life Science Group, Fujitsu, Tokyo, Japan) [18].
The final data set includes information from both sources. It was used to create
the training set containing information about 2345 pairs of single-component organic
compounds that interacted due to CYP1A2, CYP2B6, CYP2C19, CYP2C8, CYP2C9, CYP2D6,
or CYP3A4. The detailed information is presented in Table 1.
Table 1. The number of drug–drug interactions (DDIs) mediated by various isoforms of cytochrome
P450 in the training set.

Isoforms of Cytochrome P450 Number of DDIs

CYP1A2 132
CYP2B6 27
CYP2C19 80
CYP2C8 55
CYP2C9 204
CYP2D6 231
CYP3A4 1616

It is well known that CYP3A4 is the major isoform of human cytochrome P450 involved
in drug metabolism and pharmacokinetic DDIs. As we can see from Table 1, the number of
DDIs associated with CYP3A4 in the training set is twice as high as the number of pairs for
the remaining six cytochromes. It fully reflects the real situation and illustrates that the
training set is representative.

2.2. PASS
The PASS software (Laboratory of Structure-Function Based Drug Design, Institute
of Biomedical Chemistry, Moscow, Russia) [19] is based on the advanced naïve Bayes
classifier and predicts the profiles of biological activity for drug-like compounds. The
PASS algorithm creates a classification model of structure–activity relationships based on
the training set with structures and known biological activities of known pharmaceutical
agents. The PASS prediction results are presented as a ranked list of various biological
activities with calculated probabilities Pa (“to be active”) and Pi (“to be inactive”). The most
probable activities are those predicted with the maximum value ΔP = Pa − Pi . Currently,
PASS predicts more than 8000 types of biological activities, including pharmacological
effects, mechanisms of action, influences on gene expression, toxic and adverse effects, and
interactions with metabolic enzymes and transporters. Biological activities for particular
molecules in the PASS program are represented qualitatively as “active” or “inactive.”
The structural formulae of drug-like organic compounds are described by Multilevel
Neighborhoods of Atoms (MNA) descriptors.
The prediction of DDIs occurring due to interactions with various cytochrome P450
isoforms is similar to the prediction of biological activity using the PASS software. For DDIs
prediction mediated by cytochrome P450 isoforms, the input data are represented by the
pairs of structural formulas of studied drug-like compounds. The prediction results for each
pair of compounds are presented by the probabilities Pa and Pi lists, which estimate DDIs
that may occur due to interactions with CYP1A2, CYP2B6, CYP2C19, CYP2C8, CYP2C9,
CYP2D6, and CYP3A4.

225
Pharmaceutics 2021, 13, 538

2.3. Pairs of Substances Multilevel Neighborhoods of Atoms Descriptors

To describe the structures of drug pairs, we used PoSMNA descriptors instead of the
MNA descriptors applied in the standard PASS software version [19]. PoSMNA descriptors
can be used to predict various phenomena, e.g., synergistic effects of two drugs or the
prediction of DDIs. Initially, we developed and used PoSMNA descriptors to predict
DDIs severity [15,16]. The set of PoSMNA descriptors is the direct product of a combi-
nation of two sets of MNA descriptors for each molecule in the DDI pair as {a,b,c, . . . }
× {d,e,f, . . . } = {ad,ae,af, . . . , bd,be,bf, . . . , cd,ce,cf, . . . }. MNA/2 (second level of MNA
descriptors) for non-hydrogen heavy atoms is used for PoSMNA creation. The MNA
descriptors are ordered lexicographically for each pair of compounds, for example, from
string “C(C(CCC)C(CC-H)C(CC-H)) C(C(CCC)C(CC-H)O(CC))” to “-O(-C(-C-C-O)) -O(-
C(-C-O-O))” (see the examples of PoSMNA descriptors for warfarin and naproxen in
Figure 1).

Figure 1. Representation of the warfarin and naproxen molecules by Pairs of Substances Multilevel Neighborhoods of
Atoms (PoSMNA) descriptors.

To create the models for DDIs prediction, PoSMNA descriptors were generated for all
pairs of compounds with known DDIs mediated by CYP1A2, CYP2B6, CYP2C19, CYP2C8,
CYP2C9, CYP2D6, or CYP3A4 isoforms of cytochrome P450 in the training set.

3. Results
To evaluate the DDIs prediction accuracy, the IAP (Invariant Accuracy of Prediction)
values were calculated using leave-one-out cross-validation procedures (LOO CV). The
IAP criterion is numerically equivalent to the AUC ROC (Area Under Curve of the Receiver

226
Pharmaceutics 2021, 13, 538

Operating Characteristic) [19]. The IAP value is a sample estimate of the probability
randomly selected from an independent test set that will correctly classify positive and
negative examples. The accuracy of the prediction of DDIs caused by different isoforms of
cytochrome P450 is presented in Table 2.

Table 2. Accuracy of the DDIs prediction.

Isoforms of Cytochrome P450 IAP

Interaction CYP1A2 0.95
Interaction CYP2B6 0.91
Interaction CYP2C19 0.82
Interaction CYP2C8 0.98
Interaction CYP2C9 0.95
Interaction CYP2D6 0.90
Interaction CYP3A4 0.93
Average 0.92

The developed models showed good accuracy varying from 0.82 (for CYP2C19 DDIs)
to 0.98 (for CYP2C8 DDIs) with an average IAP of about 0.92. It is essential that the accuracy
for DDIs mediated by CYP3A4 is high (0.93) because interactions on the level CYP3A4
can cause severe DDIs that must be detected and avoided during the investigation of new
drugs. Thus, the accuracy of SAR models is adequate to use this method for practical tasks
of drug discovery and development.
The models created are freely available via the Internet on the Way2Drug.com web por-
tal on the DDIs web-service [20] that allows for the prediction of various DDIs parameters
and does not require registration or log-in. The combinations of warfarin taken regularly
(widely used anticoagulant with narrow therapeutic index) with various nonsteroidal anti-
inflammatory drugs (NSAIDs) are common and can increase the risk of gastrointestinal
bleeding [21]. As an example for illustrating the web-service analysis, the potential DDI
for a pair of warfarin and naproxen (one of the commonly used NSAIDs) was predicted
(see Figure 2).

Figure 2. DDI prediction for warfarin and naproxen performed using the web service [20].

The results of the prediction displayed in the block “Prediction of DDIs mediated
by P450 (PASS double mol) (7 CYP)” show that the maximum ΔP value (0.364) was
calculated for cytochrome P450 CYP2C9 (see Table 3). Therefore, the DDI for warfarin and
227
Pharmaceutics 2021, 13, 538

naproxen is most likely to occur at the level of biotransformation carried out by cytochrome
P450 CYP2C9.

Table 3. DDI prediction for warfarin and naproxen at the level of cytochrome P450 isoforms.

Isoforms of Cytochrome P450 Pa Pi ΔP

Interaction CYP2C9 0.382 0.018 0.364
Interaction CYP2C19 0.174 0.346 −0.172
Interaction CYP2C8 0.017 0.313 −0.296
Interaction CYP2D6 0.057 0.406 −0.349
Interaction CYP2B6 0.04 0.553 −0.513
Interaction CYP1A2 0.011 0.617 −0.606
Interaction CYP3A4 0.004 0.739 −0.735

Negative ΔP values for the other six isoforms of cytochrome P450 indicate that these
enzymes are not involved in DDIs at the level of warfarin and naproxen biotransformation.

4. Discussion
Because of polypharmacy, when several drugs are taken simultaneously, the phe-
nomenon of metabolic DDIs may appear. DDIs manifest in the mutual influence of drugs
on their biotransformation, its slowdown, or acceleration, and leads to a change in the
pharmacological action of drugs.
To avoid drug withdrawal from the market due to DDIs, pharmaceutical companies
perform in vitro and in vivo studies. Physiologically based pharmacokinetic (PBPK) mod-
eling is the in silico method of DDIs prediction that has already proved its applicability in
the drug discovery and development process. It is clear that in silico methods will be used
more intensively to reduce investigation costs [3].
The main problem we consider is the study and use of the relationship of chemical
compound structure and the phenomenon of metabolic DDIs mediated by the seven
isoforms of cytochrome P450 most involved in drug metabolism. The models created
can be applied for virtual and not-yet-synthesized molecules using only their structural
formulas. The implementation of PoSMNA descriptors and the PASS program algorithm
for DDIs prediction at the level of cytochromes P450 makes it possible to consider a pair
of molecules interacting as one entity without specifying the roles (substrate, inhibitor
or inducer, “object” or “precipitant” drug) of particular substances in the DDI process.
Such an approach is unique and has already been used to create models for DDIs severity
prediction [15,16]. However, when predicting the DDIs severity without taking into account
concrete pharmacokinetic or pharmacodynamic DDIs mechanisms, the accuracy of the
prediction was not high enough, as compared to that obtained in the current study that
considers only pharmacokinetic DDIs mediated by the seven cytochrome P450 isoforms
(0.84 for three classes and 0.75 for five classes of severity vs. 0.92 for DDIs prediction
mediated by cytochrome P450 isoforms). Such a lower accuracy may be explained by the
unclear separation of DDIs of these severity classes among themselves and the cases of
DDIs in neighboring classes in the training set and by neglecting the DDIs mechanisms. In
this study, the average accuracy of DDIs prediction at the level of cytochrome P450 isoforms
is higher (0.92) due to the structural specificity of substances from the pairs that interact at
a particular level of the cytochrome P450 isoform. Further research should combine the
prediction of DDIs severity at the level of a particular metabolic enzyme. To achieve this
goal, it is necessary to expand, improve, and refine the training sets.

Author Contributions: Conceptualization, A.V.D.; methodology, D.A.F. and A.V.R., software D.A.F.
and A.V.R.; investigation, D.A.K., P.V.P., A.V.R. and A.V.D.; resources, D.A.K., P.V.P., A.V.R. and
A.V.D.; data curation, D.A.K., P.V.P., A.V.R. and A.V.D. writing—original draft preparation, A.V.D.
and V.V.P.; writing—review and editing, A.V.D. and V.V.P.; supervision, A.V.D., D.A.F. and V.V.P.;
project administration, A.V.D., A.A.L., D.A.F. and V.V.P. All authors have read and agreed to the
published version of the manuscript.
228
Pharmaceutics 2021, 13, 538

Funding: This research was supported by the Russian Science Foundation (Grant No. 17-75-20250).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Data are contained within the article.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kennedy, C.; Brewer, L.; Williams, D. Drug Interactions. Medicine 2016, 44, 422–426. [CrossRef]
2. In Vitro Drug Interaction Studies—Cytochrome P450 Enzyme- and Transporter-Mediated Drug Interactions Guidance for Industry.
Available online: https://www.fda.gov/media/134582/download (accessed on 27 October 2020).
3. Dmitriev, A.V.; Lagunin, A.A.; Karasev, D.A.; Rudik, A.V.; Pogodin, P.V.; Filimonov, D.A.; Poroikov, V.V. Prediction of Drug-Drug
Interactions Related to Inhibition or Induction of Drug-Metabolizing Enzymes. Curr. Top. Med. Chem. 2019, 19, 319–336.
[CrossRef] [PubMed]
4. Banerjee, P.; Dunkel, M.; Kemmler, E.; Preissner, R. SuperCYPsPred—A web server for the prediction of cytochrome activity.
Nucleic Acids Res. 2020, 48, W580–W585. [CrossRef] [PubMed]
5. Hochleitner, J.; Akram, M.; Ueberall, M.; Davis, R.A.; Waltenberger, B.; Stuppner, H.; Sturm, S.; Ueberall, F.; Gostner, J.M.;
Schuster, D. A Combinatorial Approach for the Discovery of Cytochrome P450 2D6 Inhibitors from Nature. Sci. Rep. 2017, 7, 8071.
[CrossRef] [PubMed]
6. Kaserer, T.; Höferl, M.; Müller, K.; Elmer, S.; Ganzera, M.; Jäger, W.; Schuster, D. In Silico Predictions of Drug—Drug Interactions
Caused by CYP1A2, 2C9 and 3A4 Inhibition—A Comparative Study of Virtual Screening Performance. Mol. Inform. 2015, 34,
431–457. [CrossRef] [PubMed]
7. Torimoto-Katori, N.; Huang, R.; Kato, H.; Ohashi, R.; Xia, M. In Silico Prediction of hPXR Activators Using Structure-Based
Pharmacophore Modeling. J. Pharm. Sci. 2017, 106, 1752–1759. [CrossRef] [PubMed]
8. Fahmi, O.A.; Maurer, T.S.; Kish, M.; Cardenas, E.; Boldt, S.; Nettleton, D. A combined model for predicting CYP3A4 clinical net
drug-drug interaction based on CYP3A4 inhibition, inactivation, and induction determined in vitro. Drug. Metab. Dispos. 2008,
36, 1698–1708. [CrossRef] [PubMed]
9. Takeda, T.; Hao, M.; Cheng, T.; Bryant, S.H.; Wang, Y. Predicting drug-drug interactions through drug structural similarities and
interaction networks incorporating pharmacokinetics and pharmacodynamics knowledge. J. Cheminform. 2017, 9, 16. [CrossRef]
[PubMed]
10. Vilar, S.; Uriarte, E.; Santana, L.; Lorberbaum, T.; Hripcsak, G.; Friedman, C.; Tatonetti, N.P. Similarity-based modeling in
large-scale prediction of drug-drug interactions. Nat. Protoc. 2014, 9, 2147–2163. [CrossRef] [PubMed]
11. Cheng, F.; Zhao, Z. Machine learning-based prediction of drug-drug interactions by integrating drug phenotypic, therapeutic,
chemical, and genomic properties. J. Am. Med. Inf. Assoc. 2014, 21, e278–e286. [CrossRef] [PubMed]
12. Duke, J.D.; Han, X.; Wang, Z.; Subhadarshini, A.; Karnik, S.D.; Li, X.; Hall, S.D.; Jin, Y.; Callaghan, J.T.; Overhage, M.J.; et al.
Literature based drug interaction prediction with clinical assessment using electronic medical records: Novel myopathy associated
drug interactions. PLoS Comput. Biol. 2012, 8, e1002614. [CrossRef] [PubMed]
13. Zhang, P.; Wang, F.; Hu, J.; Sorrentino, R. Label Propagation Prediction of Drug-Drug Interactions Based on Clinical Side Effects.
Sci. Rep. 2015, 5, 12339. [CrossRef]
14. Ferdousi, R.; Safdari, R.; Omidi, Y. Computational prediction of drug-drug interactions based on drugs functional similarities.
J. Biomed. Inf. 2017, 70, 54–64. [CrossRef] [PubMed]
15. Dmitriev, A.V.; Filimonov, D.A.; Rudik, A.V.; Pogodin, P.V.; Karasev, D.A.; Lagunin, A.A.; Poroikov, V.V. Drug-drug interaction
prediction using PASS. SAR QSAR Environ. Res. 2019, 30, 655–664. [CrossRef]
16. Dmitriev, A.; Filimonov, D.; Lagunin, A.; Karasev, D.; Pogodin, P.; Rudik, A.; Poroikov, V. Prediction of Severity of Drug-Drug
Interactions Caused by Enzyme Inhibition and Activation. Molecules 2019, 24, 3955. [CrossRef] [PubMed]
17. DrugBank. Available online: https://go.drugbank.com/ (accessed on 27 October 2020).
18. ADME Database. Available online: https://www.fujitsu.com/jp/group/kyushu/en/solutions/industry/lifescience/
admedatabase/ (accessed on 27 October 2020).
19. Poroikov, V.V.; Filimonov, D.A.; Borodina, Y.V.; Lagunin, A.A.; Kos, A. Robustness of Biological Activity Spectra Predicting
by Computer Program PASS for Noncongeneric Sets of Chemical Compounds. J. Chem. Inf. Comput. Sci. 2000, 40, 1349–1355.
[CrossRef] [PubMed]
20. DDI-Pred: Web-Service for Drug-Drug Interaction Prediction. Available online: http://way2drug.com/ddi/ (accessed on 27
October 2020).
21. Hansten, P.D.; Horn, J.R. The Top 100 Drug Interactions A Guide to Patient Management, 2017th ed.; H&H Publications, LLP: Freeland,
WA, USA, 2017; p. 10.

229
 MDPI
St. Alban-Anlage 66
4052 Basel
Switzerland
Tel. +41 61 683 77 34
Fax +41 61 302 89 18
www.mdpi.com

Pharmaceutics Editorial Office

E-mail: pharmaceutics@mdpi.com
www.mdpi.com/journal/pharmaceutics
MDPI
St. Alban-Anlage 66
4052 Basel
Switzerland
Tel: +41 61 683 77 34
Fax: +41 61 302 89 18
www.mdpi.com ISBN 978-3-0365-2036-0