Medicinal and toxicological investigation of some common NSAIDs; A computer-

aided drug design approach

Abonti Barua Tori
Faculty of Science, Department of Chemistry, University of Chittagong, Chittagong, 4331, Bangladesh.

Abstract:
Pain, fever, and inflammation are frequent conditions treated with nonsteroidal anti-inflammatory
medications (NSAIDs). Their ingestion increases the risk of several serious adverse effects, including renal,
cardiovascular, and gastrointestinal damage due to the non-selective reduction of prostaglandin
synthesis, with varying degrees of inhibition.

cyclooxygenase's part a comparative biochemical, medical, and toxicological examination has been made
in an effort to increase public knowledge regarding the use of NSAIDs. To study their physicochemical,
spectral, medicinal, pharmacological, and toxicological properties, in-silico techniques were applied. To
determine their binding affinities and interactions with the amino acid residues of the receptor protein,
molecular docking and non-bonding calculations were conducted against the human prostaglandin
synthase protein. Moreover, molecular dynamics simulation was carried out to confirm the drug-protein
stability and binding manner at the inhibiting binding point. Their geometries are supported by the
physical and chemical investigation, and the spectral data attest to the existence of crucial functional
groups at the core structure. With the exception of IBP, all of the medications are not carcinogenic
according to ADMET and PASS predictions, and nonselective NSAIDs had comparatively more side effects
than selective ones. This report can aid in gaining a deeper understanding of the medicinal and
toxicological effects of the NSAIDs that have been reported, based on the studies mentioned above.[1]

Keywords: NSAIDs, Molecular docking, and dynamics simulation ADMET, and PASS prediction Toxicology.

1.0. Introduction:

Millions of individuals get relief from symptoms by using nonsteroidal anti-inflammatory drugs (NSAIDs),
which are a broad class of medications with analgesic, antipyretic, and anti-inflammatory properties.
They are usually safe and effective for the short-term treatment of moderate to severe pain, including
musculoskeletal pain, osteoarthritis, rheumatoid arthritis, dysmenorrhea, colon cancer, breast cancer, and
prostate cancer. By inhibiting cyclooxygenase (COX) enzymes, they primarily prevent the formation of
prostanoids and prostaglandins (PGs). They are categorized as non-selective (COX-1 and COX-2) or
conventional NSAIDs and COX-2 selective inhibitors based on how they work.
Here, prostacyclin, prostaglandins, thromboxane, and other inflammatory mediators originating from the
arachidonic acid cascade route are referred to as prostanoids. Here, arachidonic acid is transformed into
prostaglandin G2 by COX enzymes. NSAIDs prevent prostaglandin production, which in turn prevents the
COX-1 and COX-2 isoenzymes from acting as active. Vane and Piper revealed two isoforms of COX enzymes
(COX-1 and COX-2) in 1976 and 1991. However, COX-3 isoenzyme, a unique COX-1 variant, has just been
found [6]. The preservation of renal function, vascular homeostasis, gastrointestinal tract defense, and
platelet aggregation are among the physiological roles of COX-1 isoenzyme. On the other hand, GI side
effects such internal bleeding, ulcers etc. COX-2 is a protein found in leukocytes, fibroblasts, macrophages,
and synovial cells. Cytokines, mitogens, growth factors, and other inflammatory mediators can activate
COX-2. Consequently, COX-2 inhibition has actions that are antipyretic, analgesic, and anti-inflammatory.
Owing to the fact that many NSAIDs are sold as over-the-counter medications, there are instances when
overuse occurs. As a result, some significant adverse effects were documented, including neurotoxicity,
hepatotoxicity, and Stevens-Johnson syndrome in addition to gastrointestinal (GI), cardiovascular (CV), and
renal damage. Upper GI tract damage symptoms include abdominal pain, perforation or bleeding, mild to
severe dyspeptic symptoms, nonspecific colitis, gastroduodenal ulcers, and, eventually, death. Non-
selective NSAIDs raise the risk of gastroduodenal ulcers by 4–8 times. Selective NSAIDs (CXB, RXB, and EXB)
cause less GI problems but greater negative CV effects than non-selective NSAIDs. Because of this elevated
risk of heart attack and stroke, the USA removed many selective (COX-2) NSAIDs off the market in 2004–
2005. Furthermore, a small amount of aspirin (ASP) has been shown in a few studies to have a
cardioprotective effect. NSAIDs may be detrimental to patients with acute viral respiratory tract infections,
particularly COVID-19, according to recent suggestions. NSAIDs have been linked to hepatoxicity, which
includes jaundice-causing hepatitis, increased serum aminotransferase, liver failure, and even death.
Lumiracoxib was taken off the market in a number of nations in 2008 because it could induce serious liver
failure. Acute tubular necrosis, renal papillary necrosis, hypokalemia, salt and water retention, acute
tubulointerstitial nephritis, decreased glomerular filtration rate, hypoaldosteronism, and other conditions
are also exampling of renal toxicity. Therefore, in order to lessen these negative effects, new NSAID
derivatives are desperately needed. Using quantum chemistry and computer-aided drug design techniques,
we presented here the comparative physical, chemical, spectral, medicinal, biological, and toxicological
impacts of various selective and non-selective NSAIDs (Fig. 1). Our study's primary goal is to list some of the
most frequent side effects of the NSAIDs listed here on a single platform in an effort to increase public
awareness of the dangerous side effects of NSAIDs and identify potential new medications that could be
used in place of NSAIDs but have less side effects.[2], [3]

(a) (b)

Fig.1: Chemical structures of some NSAIDs (a)Diclofenac, (b)4,4’-diaminoazobenzene.

 2. Methods and materials

2.1. Geometry optimization

All the structures were collected from the online chemical structure database named PubChem. Geometry
optimization was carried out in Gaussian 09 W revision D.01 [21] software. Density functional theory (DFT),
along with B3LYP [22] hybrid functionals and Pople’s 6–31 G (d, p) basis set [23] were used to calculate
their thermochemical, and vibrational spectra (FT-IR). Further, time-dependent (TD)-DFT was employed to
investigate the ultraviolet-visible (UV–Vis) spectral analysis. Molecular orbital parameters like; HOMO-
LUMO energy gap (ΔE), chemical hardness (η), softness (S), and potentiality (μ) were calculated using the
following formulae

Gap (ΔE) = [εLUMO- εHOMO]

[εLUMO- εHOMO]
Ƞ=
2

1
S=
2Ƞ

[εLUMO + εHOMO]
Ƞ=
2

2.2. Protein preparation, docking, and interactions

The three-dimensional crystal structure of human cyclooxygenase protein (PDB ID: 5F19) was collected from the RCSB
protein data bank (PDB) at 2.04 Å resolution in pdb format [25]. The protein chain was prepared using PyMOL (Version
1.7.4) software package by removing the unwanted chain, heteroatoms, water molecules, and co-crystallized ligand
molecules. Further, energy minimization of the selected chain was performed by using Swiss_PDB viewer (Version 4.1.0)
software to remove bad contacts in this protein structure. Finally, flexible docking was performed against the human
prostaglandin synthase protein (5F19-A chain) considering the protein as a macromolecule and the drug as a ligand.
Auto Dock Vina inbuild PyRx (Version 0.8) software was used for the docking by maintaining the center grid box size of
65.2, 76.6, and 56.3 Å along x, y, and z directions, respectively. Then docked drug and protein were saved together in
pdb format to carry out the interaction calculation in BIOVIA Discovery Studio 2021.

2.3. ADMET and PASS prediction

Absorption, distribution, metabolism, excretion, and toxicity carry a significant role in pharmaceutical analysis and / or
in eco-friendly next-generation drug design. To exhibit proper and / or safe medicinal effects, a drug should be absorbed
smoothly, then distributed, and metabolized properly inside the body. Meanwhile, that drug should depart from the
body through urine, stool or others within the expected time. AdmetSAR [38], PASS [39], and Swiss_ADME online server
were utilized to predict the pharmacokinetics, biological, and drug likeness properties, respectively.
3.0. Results and discussion

3.1. Thermodynamic analysis

Thermodynamic parameters such as free energy, and enthalpy are key facts to know the possible
stability, binding energy, and spontaneity of any chemical reactions and products. The binding
interactions with the receptor are accelerated by negative free energy while slowed down by positive free
energy. The optimized structures have negative values for free energy and enthalpy, which shows that
they are all exothermic with spontaneous reactions and create relatively stable products (Table 1). Dipole
moments give a sign of enhanced polarity, which is a crucial factor in determining binding affinities.

Table 1 Molecular formula (MF), molecular weight (MW), free energy (ΔG), enthalpy (ΔH),
and dipole moment (µ) of two NSAIDs.

Name MF MW ΔG ΔH µ
4,4-diaminoazoaniline C12H12N4 212.26g/mol -679.490234 Hartree -679.490234 Hartree 7.1912
Debye
Diclofenac C14H11Cl2NO2 296.1 g/mol

3.2. Molecular orbital analysis the electronic absorption is mostly explained by one electron excitation
from HOMO to LUMO and is related to the change from the ground state to the first excited state [39]
which is also associate with the chemical reactivity/stability. A molecule’s chemical hardness, softness,
chemical potential values, and electrophilic index are all influenced by HOMO-LUMO energy. Kinetic
stability rises when the HOMO-LUMO gap widens.

Table 2 HOMO-LUMO gap, hardness (µ), and softness (S) of two NSAIDs.

Name HOMO LUMO Gap Hardness(µ) Softness(s)

4,4- -0.16379 -0.0444682 0.21061 O.105305 9.4962
diaminoazoaniline

Diclofenac
 (a)

LUMO=-0.04682
LUMO=-0.02724
Gap=0.21061 Gap=0.23362
HOMO=-0.1637
HOMO=-0.20638

(b)

Fig. 2. (b) DOS plot, and (a) HOMO-LUMO gap of Diaminoazoaniline and Diclofenac
3.3. Electrostatic potential map analysis

Molecular electrostatic potential (MEP) helps to predict the type and distributed area of charge. For the
possible degradation prediction, it’s important to find out the possible site for electrophilic and nucleophilic
attacks, where MEP was utilized. Additionally, it also aids in the understanding of biological recognition
mechanisms and hydrogen bonding interactions. The green color represents zero potential areas, blue
indicates the maximum positive area, and red represents the highest negative area. MEP concurrently
shows regions of positive, negative, and neutral electrostatic potential as well as molecule size, and shape
by means of color grading. From Fig. 4, mostly the blue regions were found on the electropositive carbon,
meanwhile, red zones were found on the electronegative chlorine.

Fig. 3. Electrostatic potential map of Diaminoazoaniline and Diclofenac.

3.4. Atomic partial charge analysis

The structural stability and chemical reactivity of a chemical bond are mostly influenced by charge
distribution. It has a significant impact on a molecule’s dipole moment polarizability and electronic
structure [43]. A variety of computational techniques can be used to determine the atomic charge. In this
work, Diclofenac and diaminoazoniline have been measured using two different procedures, NBO and
Mulliken (Fig. 5). According to the Mulliken approach, the C atoms in the ring closest to the trichloromethyl
group have negative charges, while the NBO method is willing to offer partial negativity for all C atoms.

Diclofenac Mullekin NBO

0.8
0.6
0.4
0.2
charge

0
-0.2 CL1 O1 N1 C2 C4 C6 C8 C10 C12 C14 H2 H4 H6 H8 H10
-0.4
-0.6
-0.8
-1
Atomic number

Diaminoazoniline Mullekin NBO

0.4
0.2
0
-0.2 N1 N3 C1 C3 C5 C7 C9 C11 H1 H3 H5 H7 H9 H11
-0.4
-0.6
-0.8
-1

Fig. 4. Partial charge of Diaminianiline and Diclofenac

3.5. UV-visible spectral analysis

UV-visible spectroscopy acts as a standard benchmark for the electronic absorption/desorption

measurement utilizing time-dependent density functional theory (TD-DFT), ensuring a balance between
accuracy and computation expense. Each of the two particular electronic transition states from the analogs
is shown in Table 4, and Fig. 6(b). The first electronic transition from the ground state (S0) to singlet (S1) in
this investigation governs the kinetic stability and chemical reactivity.

Diaminoanniline
DCF

0 500 1000 1500 2000 2500 3000 3500 4000

Wavelength(cm-1)

16000 0.40
dcf
diamino
14000 0.35

12000 0.30
oscillation strength

10000
0.25
Ephsilion

8000
0.20

6000
0.15
4000
0.10
2000
0.05
0
0.00
0 200 400 600 800 1000
Wavelength(nm)

Fig. 5. (a) FT-IR, and (b) UV-vis spectra of Diaminoazoaniline and diclofenac
 Name Excited state Wave Excitation Configuration Oscillation
length energy(ev) composition Strength

DCF S0-S1(76 -> 77) 355.30nm 3.4895 ev S0-S1(76 -> 77) f=0.0021
H-L (0.70242)

H-1->L (0.68179)
S0-S2(75 -> 77 323.71nm 3.8301 eV f=0.0118

diaminoazo S0-S1(54 -> 57) 542.11 nm 2.2871 ev H-2->L (-0.25694) f=0.0837

aniline

S0-S2(56 -> 57) 345.74 nm 3.5861 eV H->L (0.25060) f=0.2506

3.6 Molecular docking and interactions calculation

The molecular docking simulation method is popularly used to predict the binding affinity and pose of
ligands at the active site of receptor proteins [29,52]. Herein, we have considered serine/threonine-
protein kinase PIM-2 receptor protein (PDB ID: 2IWI) and human sex hormone-binding globulin protein
(PDB ID: 6PYF) respectively. Where human leukemia and lymphomas have been shown to have high levels
of the serine/threonine kinase PIM2, which has been proven to positively regulate tumor cell survival and
proliferation. Its diverse ATP site makes PIM2 a promising target for the development of anticancer
agents. On the other hand, sex hormone-binding globulin (SHBG) governs the blood’s balance of free and
protein-bound androgens and estrogens as well as how rapidly those molecules get to their intended
tissues.

(a) (b)
 (a)

(b)

Fig:7. , (b) interactions, and (c) hydrogen bond surface of diaminoazoaniline and diclofenac

Table 5 Binding affinity (BA), and interaction results of Diaminoazoanilne and Diclofenac drug.

Name BA Residue Types Distance (Å)

Diaminoazoaniline -6.4 A: VAL428 Covalent Hydrogen bond 2.90785

A: VAL428 Covalent Hydrogen bond 2.35068
A: GLY1486 Covalent Hydrogen bond 2.6751

Diclofenac -7.4 A: ARG120 Covalent Hydrogen bond 2.563

A: TYR355 Covalent Hydrogen bond 2.24598
A: PHE529 Pi-Anion 4.95319
3.7. ADMET prediction

The ‘absorption, distribution, metabolism, excretion, and toxicity’ are referred to as ADMET. In
toxicological analysis, ADMET parameters play a crucial role to predict pharmacokinetics characters. From
Table 6, DDT and its analogs respond favorably to human intestinal absorption (HIA), making it possible
that no substance can be eliminated more quickly through the urine and rectal systems [57]. All of the
compounds showed promising HOB (human oral bioavailability) outcomes. Positive oral bioavailability in
humans can cause health issues.

Table 6 Predicted ADMET results of Diaminoazoanilie and Diclofenac

Name HIA BBB p-gPI hERG Carcinogen AOT

Diaminoazoan +0.9918 +09491 +0.9261 +0.9541 0.4208 0.5616

iline

Diclofenac 0.9548 0.9541 0.8254 0.9514 0.7245 0.7602

[1] M. Uzzaman et al., “Medicinal and toxicological investigation of some common NSAIDs; A
computer-aided drug design approach,” J Mol Struct, vol. 1292, Nov. 2023, doi:
10.1016/j.molstruc.2023.136124.

[2] M. F. Afrin et al., “Spectrochemical, biological, and toxicological studies of DDT, DDD, and DDE:
An in-silico approach,” Inform Med Unlocked, vol. 39, Jan. 2023, doi: 10.1016/j.imu.2023.101254.

[3] M. Uzzaman and M. N. Uddin, “Optimization of structures, biochemical properties of ketorolac

and its degradation products based on computational studies,” DARU, Journal of Pharmaceutical
Sciences, vol. 27, no. 1, pp. 71–82, Jun. 2019, doi: 10.1007/s40199-019-00243-w.