Theor Chem Acc (2012) 131:1263

DOI 10.1007/s00214-012-1263-2

REGULAR ARTICLE

A comparative DFT study of the Schiff base formation

from acetaldehyde and butylamine, glycine
and phosphatidylethanolamine
Christian Solı́s-Calero • Joaquı́n Ortega-Castro •

Alfonso Hernández-Laguna • Francisco Muñoz

Received: 21 May 2012 / Accepted: 26 July 2012

Ó Springer-Verlag 2012

Abstract Mechanisms for the formation of the Schiff boost Schiff base formation via a neighboring catalyst
base from acetaldehyde and butylamine, glycine and effect.
phosphatidylethanolamine based on Dmol3/DFT calcula-
tions were realized. For the case of phosphatidylethanol- Keywords Monolayer model
Periodic boundary
amine, calculations were done under periodic boundary conditions
Schiff base formation
Phospholipids
conditions, in an amine-phospholipid monolayer model
with two molecules of phosphatidylethanolamine by cell.
All models contained explicit aqueous solvent. In the three 1 Introduction
cases, a neutral amino group is used to model the nucleo-
philic attack on the carbonyl group of acetaldehyde, and Schiff base (imine) formation is a very important reaction in
water molecules form hydrogen bond networks. These biological chemistry. This reaction consists of two stages,
networks were involved in the reactions by performing as the first is carbinolamine formation followed by a dehy-
proton-transfer carriers, important in some steps of reac- dration step to the formation of the Schiff base. It has been
tions, and stabilizing reaction intermediates. In all the extensively studied in various systems and processes due to
studied reactions, they take place in two steps, namely: (1) its high chemical, biological, and technological relevance
formation of a carbinolamine and (2) its dehydration to the [1–7]. One of these processes is the in vivo non-enzymatic
Schiff base, being the dehydration the rate-determining glycation that is the covalent binding of a simple reducing
step of the process, consistent with available experimental sugar to a primary amino group in a biomolecule, producing
evidence for similar reactions. The main difference a Schiff base, whose rearrangement leads to an Amadori
between the studied reactions is found in the value for product. Non-enzymatic glycation of proteins or Maillard
relative free energy for the intermediates and transition reaction is increased in diabetes mellitus due to hypergly-
states in the second step; these values are lower in the cases cemia and leads to several complications such as blindness,
of glycine and phosphatidylethanolamine in comparison heart disease, nerve damage, and kidney failure [8, 9], and
with butylamine, due the influence of their molecular the Amadori product of the glycation of phosphatidyletha-
environments. Based on the results, the aminophospholipid nolamine (PE) triggers oxidative modification in lipids via
surface environment and carboxylic group of glycine may superoxides, promotes vascular disease through their
angiogenic action on endothelial cells, and may be involved
in the development of diabetes [10, 11]. Previous experi-
C. Solı́s-Calero
J. Ortega-Castro
F. Muñoz (&)
mental and theoretical studies by our group allowed Schiff
Departament de Quı́mica, Institut d’Investigació en Ciències
de la Salut (IUNICS), Universitat de les Illes Balears, base formation mechanisms for vitamin B6 analogs and
07122 Palma de Mallorca, Spain aminophospholipids to be elucidated [12–20], and also, the
e-mail: paco.munoz@uib.es; dqufmi0@uib.es reactions of sugars and glycation target models with pyri-
doxamine have been the subject of various studies [21–24].
A. Hernández-Laguna
Instituto Andaluz de Ciencias de la Tierra (CSIC-UGR), Upon alcohol consumption, the liver enzyme alcohol
Avda. de las Palmeras 4, Armilla, 18100 Granada, Spain dehydrogenase catalyzes the oxidation of ethanol to yield

123
Page 2 of 12 Theor Chem Acc (2012) 131:1263

its primary metabolic product, acetaldehyde (AcH) [25]. with various types of small molecules [63–65]. As a result,
Acetaldehyde accumulates and exerts its toxic effects when some functional groups in membrane surfaces may effi-
the enzymatic pathways responsible for oxidizing alcohol ciently enhance the reaction via a neighboring catalyst
become overwhelmed. The electrophilic nature of acetal- effect; also, solvated membrane surfaces may provide a
dehyde renders it highly reactive, enabling it to react with favorable environment and lead to a faster reaction [14]. It
nucleophilic groups from proteins, lipids, DNA, and hor- is known there are differences between the proton mobility
monal biogenic amines forming adducts which may be at in bulk water and membrane/water interface [66–68], and
least in the early stages of Schiff bases [25–31]. Adducts these differences obviously influence in a reaction such as
are pathogenic, because they impair functions of proteins Schiff base formation where several proton transfers are
and lipids, promote DNA damage and mutation [29, 32, involved. The proton spreading over the membrane is
33], and increase the generation of reactive oxygen species facilitated by the hydrogen-bonded networks at the surface
(ROS) [34, 35]. Consequently, there may be interference of [69]. The membrane-buried layers of these networks can
cellular functions, in protein function, gene expression, and eventually serve as a storage/buffer for protons (proton
DNA integrity, including increased mutagenesis [34, 36– sponges) [66–70].
38], breakdown of immune tolerance, and induction of Experimental works have shown that various aldehydes
autoantibodies toward the resulting neoantigens [39–43]. and ketones can form Schiff bases with PE [71–73], such as
Upon ethanol-induced oxidative stress, more abundant glucose [74] and acetaldehyde [57]. Schiff bases formation
amounts and multiple species of adducts may be generated from amino acids has been studied extensively [75–77] and
from aldehydic products of lipid peroxidation and through particularly from glycine [6, 78–80]. Based on the above
the formation of hybrid adducts. Studies in both human studies, the mechanism of the reaction of Schiff base for-
alcoholics and experimental animals have further demon- mation has been well understood; it is known that this
strated adduct deposition in tissues including the liver, reaction is generally produced in high yields and that all
brain, gut, muscle, lungs, and heart thereby aggravating steps of these reactions are reversible [81]. However, it has
ethanol toxicity in such organs [30, 44–48]. not been analyzed the differences in the reaction in relation
Acetaldehyde adducts could be unstable or stable; in the to the different biochemical environments it could occur.
first case, they are generally Schiff bases whose stability In a previous work, we have done theoretical studies
depends on their localization in the modified target or about the chemical reactivity of on aminophospholipid
subsequent modifications. For example, hemoglobin surfaces [14, 82]; we used density functional theory (DFT)
adducts with acetaldehyde appear to be stable at 37 °C for and periodic boundary conditions (PBC’s) for the first time
up to 14 days, which means that these ‘‘stable’’ Schiff base to model a portion of the biological membrane surface with
products can serve as markers of ethanol consumption and a view to investigating its reactivity. In order to gain
explain some clinical consequences of ethanol abuse [49– insight into differences and similarities between the Schiff
51]. Stable adducts, on the other hand, are essentially base formation on aminophospholipid surface and only
irreversible products whose structures may vary, depending aqueous solvent environments, now we report a compara-
upon the particular target, but Schiff bases always serve as tive DFT study of the reaction of acetaldehyde, a bio-
intermediates of these advanced stable products and they chemical prolific reactive carbonyl compound, with
also could be used as biomarkers of pathogenic process butylamine, glycine, and PE. The primary aim of this
[52, 53]. Adducts with DNA have been reported as bio- theoretical study was to elucidate the influence of the
markers, in this case for carcinogenic process related to chemical environment on the Schiff base formation reac-
alcohol consumption such as head and neck cancer, as well tion via an H-atom-transfer mechanism and how it could
as cancer at other sites [54, 55]. In the case of phospho- explain the differences in the speed in other similar
lipids, acetaldehyde forms a Schiff base with PE and this reactions.
adduct could be reduced to the corresponding N-ethyl-
phosphatidylethanolamine [56–58].
In all these processes, the speed of formation and sta- 2 Methodology
bility of these acetaldehyde adducts depend on the chem-
ical environment in which their formation is done [59–61]. In order to make possible DFT calculus, PE surface model
In the case of non-enzymatic glycation process where was designed from the crystal structure of 1,2-Dilauroyl-
Schiff base formation is also a part, this proceeds faster in DL-phosphatidylethanolamine [83]. The PBC’s made pos-
lipids than proteins [62]. This can be ascribed to the sible to obtain a surface model of a layer of phospholipids,
chemical nature of membrane surfaces. Thus, the interfa- useful for studying theoretically, the reaction on an envi-
cial region of a membrane is known to establish electro- ronment different to aqueous solvent. The models for
static, hydrophobic, and/or hydrogen bonding interactions butylamine and glycine, due their more simple structure,

123
Theor Chem Acc (2012) 131:1263 Page 3 of 12

Fig. 1 Periodic model of Phosphatydilethanolamine surface. a Sec- phosphatidylethanolamine surface, reactive atoms are labeled, and
tion of the initial model for two phosphatidylethanolamine molecules, dotted lines represent hydrogen bonds
acetaldehyde and the water hydrogen bond network. b A sight of

were built without PBC’s, but also including water mole- obtained a good agreement with experimental values [93].
cules as explicit solvent and acetaldehyde. PBE functional has been widely used in the study of great
The PE surface model was represented using a three- variety of molecular and extended systems, having accu-
dimensionally periodic slab model. The supercell (Fig. 1a) racy for molecular systems, in the prediction of properties
contained two molecules of truncated PE, a molecule of such as ionization potentials, electron affinities, and bond
acetaldehyde, and nine water molecules as explicit solvent distances [94–97]. The maximum number of numerical
in a hydrogen bond network along the polar heads of integration mesh points available in DMol3 was chosen for
phospholipids. They were chosen as the model compound our computations, and the threshold of density matrix
to study the Schiff base formation on the amino-phospho- convergence was set to 10-6. A Fermi smearing of 0.005
lipids surface. One of the PE molecules had a neutral amine Hartree and a real-space cutoff of 4.5 Å were also used to
group intended to facilitate modeling of the nucleophilic improve the computational performance.
attack on the carbonyl group of acetaldehyde, and the other The initial models as reactants and the next models for
had a charged amine group in order to assist some steps of stationary points generated during Schiff base formation in
studied reaction acting as proton donor and acceptor. The all the cases were modeled in Materials Visualizer and
designed models for the systems with butylamine and optimized using the conjugated gradient algorithm. Tran-
glycine included an acetaldehyde molecule and 29 water sition state (TS) searches were performed with the com-
molecules. The purpose of including this number of water plete LST/QST method [98]. In this method, the linear
molecules in these molecular models was not exclusively synchronous transit (LST) maximization was performed,
to simulate a water solvation environment; rather, the water followed by an energy minimization in directions conju-
molecules were intended to act as reactive species facili- gating to the reaction pathway to obtain approximated TS.
tating several steps of studied reaction in the different The approximated TS was used to perform quadratic syn-
models. chronous transit (QST) maximization and then another
All of the calculations were performed in the frame of conjugated gradient minimization was performed. The
DFT with program package DMol3 of Accelrys, Inc. [84– cycle was repeated until a stationary point was located. The
86], using double numerical with polarization (DNP) basis obtained TS was optimized via eigenvector following
sets [86] and Perdew–Burke–Ernzerhof (PBE) generalized searching for an energy maximum along one previous
gradient approximation (GGA) exchange–correlation selected normal mode and a minimum along all other
functional [87, 88]. The DNP numerical basis set is com- nodes, using Newton–Raphson method. After this proce-
parable to Gaussian 6-31G(d, p) [89–91], minimizes the dure, one transition state was found for each reaction step.
basis set superposition error [92], and its accuracy for Each TS structure was characterized by a vibrational
describing hydrogen bond strengths has been tested, having analysis with exactly one imaginary frequency. Mulliken

123
Page 4 of 12 Theor Chem Acc (2012) 131:1263

Scheme 1 Mechanism of Schiff base formation between butylamine and acetaldehyde. Dotted lines represent hydrogen bonds (R- = butyle)

population analysis was used to understand the charge flow surface). Table 1 lists the DG values for each structure
on-group migration. involved in the process, and Fig. 2 shows the comparative
free energy profile.

3 Results and discussion 3.1 Carbinolamine formation

The selection of PE, butylamine, and glycine-like primary The starting point for these stepwise processes are struc-
amines for studying the Schiff base formation was due to tures S1 (Schemes 1, 2, 3), where the incoming amino
their differences in the adjacent groups, and it could let us groups (N3) of the primary amines are the agent of the
evaluate the possible influence of these groups in the nucleophilic attack on the carbonyl carbons in acetalde-
reaction. In the case of PE, it is possible to evaluate hyde (C1). The amine approach starts at an N3–C1 distance
additionally the influence in the reaction of an environment of 2.70 Å in PE surface (Fig. 1), in the case of glycine this
different to aqueous solvent. PE is one of the major distance is 3.47 Å, and in butylamine is 5.75 Å. These
phospholipids of the biological membranes; in comparison differences could be attributed to the influence of the
with another phospholipids with a primary amine group, it environment around; in the case of butylamine, water
is the most simple and its reaction with acetaldehyde has molecules have more freedom for their mobility, having
been probed experimentally [56, 99, 100]. less interactions with the reactive molecules, than the case
The found structures allowed a detailed chemical path- of glycine that have a carboxylic group, and PE surface
way for the formation of a Schiff base between acetalde- where there are several groups acting as hydrogen bond
hyde and the three studied primary amines. Schemes 1, 2, donors or acceptors and stabilizing the hydrogen bonds
and 3 show the atoms directly involved in the reactions and networks.
the overall processes. In the three cases, the Schiff base The relative energy barriers for direct addition of the
formation essentially involve two steps, namely: carbinol- amino group to the carbonyl group from acetaldehyde for
amine formation (structures 1–5 for butylamine and gly- zwitterionic carbinolamine formation had values of 7.0,
cine, structures 1–7 in the case of PE surface) and its 2.4, and 2.5 kcal mol-1 for butylamine, glycine, and PE
dehydration to the Schiff base (structures 5–7 for butyl- surface, respectively. These values are comparatively low
amine and glycine, structures 7–11 in the case of PE in comparison with the obtained results by other studies.

123
Theor Chem Acc (2012) 131:1263 Page 5 of 12

Scheme 2 Mechanism of Schiff base formation between glycine and acetaldehyde. Dotted lines represent hydrogen bonds

Scheme 3 Mechanism of Schiff base formation between a phosphatidylethanolamine monolayer and acetaldehyde, using periodic boundary
conditions. Dotted lines represent hydrogen bonds

123
Page 6 of 12 Theor Chem Acc (2012) 131:1263

Table 1 Standard free energies

Step reaction Structure DG (Kcal/mol) Structure DG (Kcal/mol)
of the structures of the reaction
paths Butylamine Glycine Phosphatydilethanolamine

Carbinolamine formation S1 0.0 0.0 S1 0.0

TS2 7.0 2.4 TS2 2.5
S3 2.9 -6.2 S3 -8.5
TS4 13.3 3.8 TS4 -5.2
S5 -7.0
TS6 -3.0
Carbinolamine dehydration S5 9.3 -4.2 S7 -5.8
TS6 20.3 13.2 TS8 13.1
S7 12.5 -5.3 S9 7.0
TS10 11.6
S11 -5.7

In the reaction between dimethylamine and propanal, the 25

Butylamine
DG= barrier has value of 38.4 kcal mol-1 after ab initio Glycine
TS6
20
Phosphatydilethanolamine
gas-phase calculations, but the addition of two molecules
Δ G (Kcal/mol)
15 TS4 TS6
of dimethylamine or methanol reduces this barrier till TS8
TS10
values of 7.2 and 10.9 kcal mol-1, respectively, showing 10 TS2
the catalytic effect of the reagent or the co-catalyst for TS4
5 TS2
stabilizing the transition states [101]. The relative energy
O

barrier for the addition of pyridoxamine analog to carbonyl 0 TS2

TS6
compounds as acetaldehyde and glycolaldehyde had also -5
TS4

low values of 9.7 and 10.1 kcal mol-1 using DFT level of
-10
theory gas-phase calculations [102]. The same tendency is
appreciated for the addition of pyridoxamine to glyoxylic Carbinolamine formation Carbinolamine dehydration
-15
acid [103] and methylamine to pyridoxal [104], with Structures
relative energy barriers of 3.9 and 7.9 kcal mol-1, Fig. 2 Free energy profile for the reaction
respectively. In these cases, the catalytic effect could be
attributed to the stabilization of transition states through
structural resonance or assistance of other molecules like the model of the surface, it could act as proton donor or
water solvent or polar groups of reagents. In the case of the acceptor in the different steps of the reaction, acting as an
studied systems by us, the stabilization of transition states acid catalytic group. In the case of butylamine and glycine,
can be ascribed to the presence in the models of an explicit being alone surrounded by solvent molecules, the proton
solvent that forms hydrogen bonds with the reactants and transfers only could be done having final proton donors or
products alike, thereby facilitating addition of the amino acceptors, groups of reactive molecules. In the reaction
group to the carbonyl carbon. The water molecules can be with butylamine and glycine, the formation of a neutral
also involved in the three studied reactions by stabilizing form of carbinolamine is direct, and without a positive
zwitterions forms of the carbinolamine. However, the charged carbinolamine intermediate, having only one
zwitterionic form of the carbinolamine from butylamine is transition state (TS4). The transfer of proton from N3 to O2
less stable than formed by reaction with glycine and PE is done through hydrogen bond chains of four water mol-
(Fig. 2), and it could be due to the presence in the last cases ecules in the case of butylamine and six in glycine. It could
of charged groups, carboxylate group in glycine, and be added that in the case of glycine, some of this water
phosphate group in PE which stabilize this intermediate. molecules form hydrogen bonds with oxygen atoms of
Structure S3 is in the three cases a zwitterionic form of carboxylic group of glycine (Scheme 2), interaction absent
the carbinolamine. Atom O2, which is negatively charged, for butylamine (Scheme 1).
is explicitly solvated with water molecules via hydrogen In the case of reaction on PE surface, this part of the
bonds. It is the first point where appears more differences reaction is realized through two transition states (TS4 and
between the reactions in butylamine and glycine versus PE TS6), first, the zwitterionic form of carbinolamine is con-
surface. Due to inclusion of another amine charged PE in verted into a positive charged form by transfer of proton

123
Theor Chem Acc (2012) 131:1263 Page 7 of 12

from a charged amine group of the another PE molecule reactive molecules. It is known interfaces between bio-
through a hydrogen bond chain of three water molecules logical membranes and water solvent environment adopt a
that facilitates protonation of the charged oxygen (O2). The dielectric constant (e) significantly lower than in the
PBC’s allow proton H14 to cleave its bond to O13 and be aqueous phase [112–115]. In some biological membranes,
transferred from one face of the unit cell to the opposite it has been determined the network of hydrogen bonds on
face in order to bond to O2 (Scheme 3). Then, the charged the surface of the PE membrane can serve as a storage
form of carbinolamine gives up a proton to the amine group mechanism in solution proton (proton sponge), allowing
of the another PE molecule that had been deprotonated in the released proton may remain for a time along the
the before point of the reaction (from N3 to N6), through a membrane surface before being dissipated in the aqueous
hydrogen bond chain of three water molecules via a con- medium of ‘‘bulk’’ of water [66, 70]. This result could also
certed transition state (TS6), forming the neutral form of explain experimental evidence that say the kinetics of lipid
carbinolamine (S7) (Scheme 3). Water has been shown to glycation is little faster than that of protein glycation [116].
take part in similar reactions in other simple systems where
the energy barrier for carbinolamine formation by proton 3.2 Dehydration
transfer via a ‘‘Grotthuss mechanism’’ was found to be
reduced if explicit water molecules were used to facilitate The next step in the reaction is dehydration of the carbi-
proton transfer [105]. Based on experimental work on other nolamine to the corresponding Schiff base, which involves
molecular systems, these protonation reactions are pH- the concerted release of the hydroxyl group from carbi-
dependent in acid–base equilibria [106–111]. nolamine and the transfer of one hydrogen from a donor
In the three studied reactions, proton transfers are done atom, in the case of glycine and butylamine from charged
through a chain of water molecules because the long dis- nitrogen N3 from the same carbinolamine through a chain
tance between the possible proton donors and acceptors, of two water molecules (Schemes 1, 2; Figs. 3, 4) forming
playing the solvation water molecules a reactant role. directly the neutral form of Schiff base. In the case of PE
Proton transfer on PE surface takes place via TS4 and TS6, surface, this happens through two transition states (TS8 and
with very low energy barriers, 3.3 and 3.4 kcal/mol, TS10): at first, a proton transfer from the protonated amino
respectively, in comparison with the energy barriers for the group, and in the second, phospholipid chain to hydroxyl
direct proton transfer in butylamine and glycine without the group O2-H14 are done, through a water molecule to give
positive carbinolamine intermediate, 10.4 and 10.1 kcal/ the leaving water molecule and the protonated form of
mol, respectively. These differences could be attributed to Schiff base the iminium ion S9 (Scheme 3; Fig. 5). Then,
the presence of polar and charged groups in the PE surface, N3 atom in intermediate S9 is deprotonated, being N6 atom
which impose limitations on the mobility of the water of the second PE molecule the final proton acceptor,
molecules on its surface, polarizing them and also the restoring in this way, its initial, charged amino group. Four

Fig. 3 The pathway for dehydration of carbinolamine molecule from reaction with butylamine. (S5) Carbinolamine; (TS6) transition state; (S7)
Neutral Schiff base

123
Page 8 of 12 Theor Chem Acc (2012) 131:1263

Fig. 4 The pathway for dehydration of carbinolamine molecule from reaction with glycine. (S5) Carbinolamine; (TS6) transition state; (S7)
Neutral Schiff base

Fig. 5 The pathway for dehydration of carbinolamine molecule from reaction with phosphatidylethanolamine. (S7) Carbinolamine; (TS8)
transition state; (S9) iminium ion product

water molecules networked by hydrogen bonds act as a and 18.9 kcal mol-1, respectively (Fig. 2). Additionally, in
bridge to facilitate the passage of protons through a con- PE surface, obtaining the neutral form of Schiff base from its
certed transition state (TS10 in Scheme 3). This step positive charged form is also subject to a free energy small
additionally causes the formation of an imine double bond barrier (4.6 kcal mol-1, Fig. 2). With the exception of het-
between C1 and N3, the distance between which is thereby erocyclic systems, the iminium ions are known to be
reduced from 1.47 (S7) to 1.29 Å (S11). unstable [117], so that the conversion of the positive Schiff
As in other molecular systems [102–104, 107], carbi- base intermediate S9 to its neutral form (S11) is a favorable
nolamine dehydration in the three cases butylamine, glycine, process. These neutral carbinolamines from glycine and PE
and PE surface is the rate-determining step in the formation (S5 and S7, respectively) have approximately similar values
of the Schiff base, with an free energy barrier of 11.0, 17.4, for their values of DG (Fig. 2).

123
Theor Chem Acc (2012) 131:1263 Page 9 of 12

There are other systems for Schiff base formation where amino group in butylamine, glycine, and PE surface. In the
the barriers for dehydration step are higher than the three cases, the water molecules take part in the reaction by
obtained by us. In the reaction between dimethylamine and performing proton transfer and stabilizing the reactants and
propanal, the dehydration step has a DG= barrier from intermediates.
57.5 kcal mol-1 after ab initio gas-phase calculations As can be seen in Fig. 2 and Table 1, the intermediate
[101]. The assistance of other molecules was necessary, molecules and Schiff base products from butylamine are
like two methanol molecules, in order to reduce this DG= less stable than products from glycine or PE. This result
barrier to 32.6 kcal mol-1, showing the importance of the could be attributed to the experimentally probed instability
assistance of other molecules for making possible this step of aliphatic imines, in comparison with imines with other
of the reaction. However, in certain systems, the free substituents, as consequence of it, aliphatic imines were not
energy barrier for the dehydration step could be slightly possible to isolate, having been studied less than other
lower than the obtained barriers in the reactions between imines [120, 121]. A comparative experimental study of
acetaldehyde and glycine or PE. This may be a result of the reactivity of different amines in their reaction benzalde-
additional assistance in these systems, provided by other hyde as carbonyl compound showed the increasing of
chemical groups from the reagent molecules. A DFT study reaction equilibrium constants from \10 M-1 in aromatic
in the gas phase of the irreversible transamination between amines to around 103 M-1 for aliphatic amines [122]. This
glyoxylic acid and pyridoxamine analog showed a value of fact could be related to the high values for relative free
16.9 kcal mol-1 for relative energy barrier for dehydration energies of intermediates from butylamine reaction, in
step, and it has revealed that a carboxylic group in the comparison with glycine and PE (Fig. 2; Table 1).
amino acid acts as a proton donor facilitating water elim- Differences in whole processes could be also attributed
ination and also that a phenol group in pyridoxamine to the influence of intramolecular and intermolecular
analog helps stabilize the system [103]. A DFT study of the groups that could participate catalytically in the reactions.
Schiff base formation between a pyridoxamine analog and It has been shown glycine carboxylic group can even
acetaldehyde or glycolaldehyde in the gas phase provided participate in an intramolecular general acid catalysis
relative energy barriers from 10 to 15 kcal mol-1, [123]. For the reaction on PE surfaces, there is a general
depending of used correlation functional for calculus. acid catalysis, but intermolecular and mediated by charged
Inclusion of solvent effects through CPCM implicit solvent amine group of other PE molecule, making the process
method also reduced slightly the energy barriers. In this longer but more effective, reducing free energy barriers for
system, a phenolic hydroxyl group was found to act as a carbinolamine formation in its neutral form. Catalysis by
proton donor to the carbinolamine hydroxyl group in order general acids has been also reported for other aldehyde
to produce the leaving water molecule [102]. The intra- amine reactions [110, 111, 124, 125], and the influence of
molecular assistance for the dehydration step has been also the environment of phospholipids surfaces in the proton
determined experimentally for the case of the reaction transfers has been probed experimentally by other studies
between a cyclohexene-1–carboxaldehyde and glycine or [126, 127]. Additionally, weak interactions as Van der
aspartic acid in aqueous solution [118]. In this work, it was Waals forces between reactive molecules and water-PE
found an important acceleration of the reaction with these surface contribute to stabilize the intermediates and prod-
two amino acids in comparison with the reaction with ucts of the reactions and could also influence in the orga-
aliphatic amines, it was attributed to intramolecular general nization of interfacial water molecules.
base catalysis of water attack by the internal carboxyl In PE surfaces, each PE molecule also possesses a
groups, having also determined that this behavior is phosphate group that may play a role in this reaction. We
exceedingly efficient in a relatively nonpolar solvent probed a phosphate group as proton acceptor; without
mixture [119]. obtaining any stable species, probably due its too low pKa,
In the three theoretically studied reactions for Schiff the experimental value for which in PE is 0.5 [128].
base formation, water plays a prominent role in all proton However, phosphate anion might enhance Schiff base
transfers, acting as bridge along which protons are trans- formation via another way, a neighboring catalyst effect; in
ferred through water molecules networked by hydrogen fact, we found it to form hydrogen bonds with water
bonds. Water can influence the reaction barrier by elec- molecules in the network connecting donor and acceptor
trostatic stabilization of ionic transition structures and other protons, and amino groups of PE (Fig. 1), in different steps
reaction intermediates, formation of a strong hydrogen of the studied mechanism. Phosphate groups could facili-
bond, and acting as a proton-transfer carrier. Moreover, the tate accumulation of H2O on the membrane surface and
whole reaction mechanism is governed to a great extent by raising local concentrations as a result (as found in previ-
the network of hydrogen bonds in the different intermedi- ous studies, negatively charged phosphate groups are
ates formed upon condensation of acetaldehyde with the tightly solvated by an average of four water molecules

123
Page 10 of 12 Theor Chem Acc (2012) 131:1263

each) [129]. Together other polar and charged groups of PE 8. Ramasamy R, Yan SF, Schmidt AM (2011) Ann N Y Acad Sci
surface, it is also able to polarize water bonds through 1243:88–102
9. Busch M, Franke S, Rüster C, Wolf G (2010) Eur J Clin Invest
interaction with them [130], which may facilitate the role 40:742–755
of solvation water molecules as bridges for the proton 10. Miyazawa T, Oak JH, Nakagawa K (2005) Ann N Y Acad Sci
exchange between donor and acceptor protons in the 1043:280–283
reaction, and exerting a passive catalytic effect by stabi- 11. Oak JH, Nakagawa K, Oikawa S, Miyazawa T (2003) FEBS Lett
555:419–423
lizing charge in various reaction intermediates through 12. Vilanova B, Gallardo JM, Caldés C, Adrover M, Ortega-Castro
direct electrostatic interactions with the positively charged J, Muñoz F, Donoso J (2012) J Phys Chem A 116:1897–1905
groups produced in the different reaction steps. The Schiff 13. Caldés C, Vilanova B, Adrover M, Muñoz F, Donoso J (2011)
base formation reaction between on a model PE surface Bioorg Med Chem 19:4536–4543
14. Solı́s-Calero C, Ortega-Castro J, Muñoz F (2010) J Phys Chem
and acetaldehyde provides a simple means for illustrating B 114:15879–15885
the catalytic potential of phospholipid groups in cell 15. Salvà A, Donoso J, Frau J, Muñoz F (2002) Int J Quant Chem
membranes and their solvating water molecules to enhance 89:48–56
a reactions that happen in their surface via a neighboring 16. Salvà A, Donoso J, Frau J, Muñoz F (2002) J Mol Struc-
Theochem 577:229–238
catalyst effect. The designed model of PE surface could be 17. Vázquez MA, Donoso J, Muñoz F, Garcı́a Blanco F, Garcia del
used for studying other reactions on PE surface at the DFT Vado MA, Echevarrı́a G (1991) J Chem Soc Perkin Trans
level. Considering the chemical environments where these 2:1143–1147
reactions proceed, reaction mechanisms could be explained 18. Vázquez MA, Donoso J, Muñoz F, Garcı́a Blanco F (1991) J
Chem Soc Perkin Trans 2:275–281
in terms of polarization and electronic transfer effects that 19. Vázquez MA, Muñoz F, Donoso J, Garcı́a Blanco F (1990) Int J
are turned on and off along the reaction coordinate [131]. Chem Kin 22:905–914
Despite the advantage of the reaction on PE surface, this 20. Vázquez MA, Muñoz F, Donoso J, Garcı́a Blanco F, Garcia del
study also showed the efficiency of acetaldehyde as car- Vado MA, Echevarrı́a G (1990) J Mol Catal 59:137–145
21. Adrover M, Vilanova B, Muñoz F, Donoso J (2009) Bioorg
bonyl compound for Schiff base formation, reacting with Chem 37:26–32
different molecules with a free amine group. Our results 22. Adrover M, Vilanova B, Muñoz F, Donoso J (2008) Ann N Y
are also agreed with the extent promiscuity of acetaldehyde Acad Sci 1126:235–240
to react with a great deal of kind of molecules with amine 23. Adrover M, Vilanova B, Muñoz F, Donoso J (2007) Int J Chem
Kinet 39:154–167
groups in biological systems. The produced adducts, when 24. Adrover M, Vilanova B, Muñoz F, Donoso J (2005) Chem
acetaldehyde modify proteins, lipids, or nuclei acids, have Biodivers 2:964–975
been implicated not only in carcinogenic process and 25. Murrey SJ, Brecher AS (2010) Digest Diseas Sci 55:21–27
processes related to tobacco and ethanol abuse [132, 133], 26. Tong M, Longato L, Nguyen QG, Chen WC, Spaisman A, de la
Monte SM (2011) Oxid Med Cell Longev ID 213286
but also in the pathogenesis of vascular disease and aging 27. Seitz HK, Becker P (2007) Alcohol Res Health 30:38–41
[134]. 28. Brooks PJ, Theruvathu JA (2005) Alcohol 35:187–193
29. Tuma DJ, Casey CA (2003) Alcohol Res Health 27:285–290
Acknowledgments This work was funded by the Spanish Govern- 30. Tuma DJ (2002) Free Radic Biol Med 32:303–308
ment in the framework of Project CTQ2008-02207/BQU. One of us 31. Yu HS, Oyama T, Isse T, Kitagawa K, Pham TT, Tanaka M,
(C. S-C) wishes to acknowledge MAE-AECI fellowship from the Kawamoto T (2010) Chem Biol Interact 188:367–375
Spanish Ministry of Foreign Affairs and Cooperation. The authors are 32. Abraham J, Balbo S, Crabb D, Brooks PJ (2011) Alcohol Clin
grateful to Centro de Cálculo de Computación de Galicia (CESGA), Exp Res 35:2113–2120
and the Centro de Cálculo de Computación de Cataluña (CESCA), for 33. Rulten SL, Hodder E, Ripley TL, Stephens DN, Mayne LV
access to their computational facilities. (2008) Alcohol Clin Exp Res 32:1186–1196
34. Boffetta P, Hashibe M (2006) Lancet Oncol 7:149–156
35. Goodlett CR, Horn KH, Zhou FC (2005) Exp Biol Med
230:394–406
References 36. Kharbanda KK, Todero SL, Shubert KA, Sorrell MF, Tuma DJ
(2001) Alcohol 25:123–128
1. Gokhale MY, Kearney WR, Kirsch LE (2009) AAPS Pharm- 37. Niemelä O (2001) Free Radic Biol Med 31:1533–1538
SciTech 10:317–328 38. Tuma DJ, Thiele GM, Xu D, Klassen LW, Sorrell MF (1996)
2. Sztanke K, Pasternak K (2003) Ann Univ Mariae Curie Sklo- Hepatology 23:872–880
dowska Med 58:159–162 39. Latvala J, Melkko J, Parkkila S, Järvi K, Makkonen K, Niemelä
3. Martins SIFS, Van Boekel MAJS (2005) Food Chem 92:437– O (2001) Alcohol Clin Exp Res 25:1648–1653
448 40. Rolla R, Vay D, Mottaran E, Parodi M, Traverso N, Aricó S,
4. Knol JJ, Van Loon WAM, Linssen JPH, Ruck AL, Van Boekel Sartori M, Bellomo G, Klassen LW, Thiele GM, Tuma DJ,
M, Voragen AGJ (2005) J Agric Food Chem 53:6133–6139 Albano E (2000) Hepatology 31:878–884
5. Shipar MAH (2004) J Mol Struc-Theochem 710:45–50 41. Niemelä O (1993) Scand J Clin Lab Invest 213:45–54
6. Vázquez MA, Echevarrı́a G, Muñoz F, Donoso J, Garcı́a Blanco 42. Xu D, Thiele GM, Beckenhauer JL, Klassen LW, Sorrell MF,
F (1989) J Chem Soc Perkin Trans 2:1617–1622 Tuma DJ (1998) Gastroenterology 115:686–692
7. Donoso J, Muñoz F, Garcia del Vado MA, Echevarrı́a G, Garcı́a- 43. Trudell JR, Ardies CM, Green CE, Allen K (1991) Alcohol Clin
Espantaleón A, Garcı́a Blanco F (1986) Biochem J 238:137–144 Exp Res 15:295–299

123
Theor Chem Acc (2012) 131:1263 Page 11 of 12

44. McCaskill ML, Kharbanda KK, Tuma DJ, Reynolds JD, 77. Feeney RE, Blankenhorn G, Dixon HB (1975) Adv Protein
DeVasure JM, Sisson JH, Wyatt TA (2011) Alcohol Clin Exp Chem 29:135–203
Res 35:1106–1113 78. Mitra J, Metzler DE (1988) Biochim Biophys Acta 965:93–96
45. Setshedi M, Wands JR, Monte SM (2010) Oxid Med Cell 79. Jirousová J, Davı́dek J (1975) Zeits Lebens 157:269–276
Longev 3:178–185 80. Eichhorn GL, Marchand D (1956) J Am Chem Soc 78:2688–
46. Niemelä O (2007) Novartis Found Symp 285:183–192 2691
47. Nakamura K, Iwahashi K, Furukawa A, Ameno K, Kinoshita H, 81. Borisova NE, Reshetova MD, Ustynyuk YA (2007) Chem Rev
Ijiri I, Sekine Y, Suzuki K, Iwata Y, Minabe Y, Mori N (2003) 107:46–79
Arch Toxicol 77:591–593 82. Solı́s-Calero C, Ortega-Castro J, Muñoz F (2011) J Phys Chem
48. Thiele GM, Worrall S, Tuma DJ, Klassen LW, Wyatt TA, C 115:22945–22953
Nagata N (2001) Alcohol Clin Exp Res 25:218S–224S 83. Elder M, Hitchcock P, Mason R, Shipley GG (1977) Proc R Soc
49. Stevens VJ, Fantl WJ, Newman CB, Sims RV, Cerami A, Pet- London A 354:157–170
erson CM (1981) J Clin Invest 67:361–369 84. Delley B (2000) J Chem Phys 113:7756–7764
50. Niemelä O, Israel Y (1992) Lab Invest 67:246–252 85. Delley B (1996) J Phys Chem 100:6107–6110
51. Braun KP, Pavlovich JG, Jones DR, Peterson CM (1997) 86. Delley B (1990) J Chem Phys 92:508–517
Alcohol Clin Exp Res 21:40–43 87. Perdew JP, Burke K, Ernzerhof M (1996) Phys Rev Letters
52. Duryee MJ, Klassen LW, Schaffert CS, Tuma DJ, Hunter CD, 77:3865–3868
Garvin RP, Anderson DR, Thiele GM (2010) Free Radic Biol 88. Perdew JP, Chevary JA, Vosko SH, Jackson KA, Pederson MR,
Med 49:1480–1486 Singh DJ, Fiolhais C (1992) Phys Rev B 46:6671–6687
53. Tuma DJ, Hoffman T, Sorrell MF (1991) Alcohol Alcoholism 89. Lee CS, Hwang TS, Wang Y, Peng SM, Hwang CS (1996) J
1:271–276 Phys Chem 100:2934–2941
54. Balbo S, Meng L, Bliss RL, Jensen JA, Hatsukami DK, Hecht 90. Wang ZG, Zeng QD, Luan YB, Wu XJ, Wan LJ, Wang C, Lee
SS (2012) Cancer Epidemiol Biomarkers Prev 21:601–608 GU, Yin SX, Yang JL, Bai CL (2003) J Phys Chem B
55. Seitz HK, Stickel F (2010) Genes Nutr 5:121–128 107:13384–13388
56. Trudell JR, Ardies CM, Anderson WR (1990) Mol Pharmacol 91. Lin TT, Zhang WD, Huang JC, He CB (2005) J Phys Chem B
38:587–593 109:13755–13760
57. Kenney WC (1984) Alcohol Clin Exp Res 8:551–555 92. Matsuzawa N, Seto J, Dixon DA (1997) J Phys Chem A
58. Kenney WC (1982) Alcohol Clin Exp Res 6:412–415 101:9391–9398
59. Fowles LF, Beck E, Worrall S, Shanley BC, de Jersey J (1996) 93. Andzelm J, Govind N, Fitzgerald G, Maiti A (2003) Int J
Biochem Pharmacol 51:1259–1267 Quantum Chem 91:467–473
60. Braun KP, Cody RB, Jones DR, Peterson CM (1995) J Biol 94. Xu X, Goddard WA (2004) J Chem Phys 121:4068–4082
Chem 270:11263–11266 95. Fabiano E, Constantin LA, Della Sala F (2010) Phys Rev B
61. Gross MD, Hays R, Gapstur SM, Chaussee M, Potter JD (1994) 82:113104
Alcohol Alcoholism 29:31–41 96. del Campo JM, Gázquez JL, Trickey SB, Vela A (2012) J Chem
62. Higuchi O, Nakagawa K, Tsuzuki T, Suzuki T, Oikawa S, Mi- Phys 136:104108
yazawa T (2006) J Lipid Res 47:964–974 97. Ernzerhof M, Scuseria GE (1999) J Chem Phys 110:5029–5035
63. Lukacova V, Peng M, Fanucci G, Tandlich R, Hinderliter A, 98. Halgren TA, Lipscomb WN (1977) Chem Phys Letters
Maity B, Manivannan E, Cook GR, Balaz S (2007) J Biomol 49:225–232
Screen 12:186–202 99. Murzyn K, Róg T, Pasenkiewicz-Gierula M (2005) Biophys J
64. Pohle W, Gauger DR, Bohl M, Mrazkova E, Hobza P (2004) 88:1091–1103
Biopolymers 74:27–31 100. McIntosh TJ (1996) Chem Phys Lipids 81:117–131
65. Barry JA, Gawrisch K (1994) Biochemistry 33:8082–8088 101. Patil MP, Sunoj RB (2007) J Org Chem 72:8202–8215
66. Mulkidjanian AY, Heberle J, Cherepanov DA (2006) Biochim 102. Ortega-Castro J, Adrover M, Frau J, Salvà A, Donoso J, Muñoz
Biophys Acta 1757:913–930 F (2010) J Phys Chem A 114:4634–4640
67. Grudinin S, Büldt G, Gordeliy V, Baumgaertner A (2005) 103. Liao RZ, Ding WJ, Yu JG, Fang WH, Liu RZ (2008) J Comput
Biophys J 88:3252–3261 Chem 29:1919–1929
68. Kühlbrandt W (2000) Nature 406:569–570 104. Salvà A, Donoso J, Frau J, Muñoz F (2003) J Phys Chem A
69. Mathias G, Marx D (2007) Proc Natl Acad Sci USA 104: 107:9409–9414
6980–6985 105. Hall NE, Smith BJ (1998) J Phys Chem A 102:4930–4938
70. Heberle J, Riesle J, Thiedemann G, Oesterhelt D, Dencher NA 106. Gokhale MY, Kirsch LE (2009) J Pharm Sci 98:4616–4628
(1994) Nature 370:379–382 107. Gokhale MY, Kirsch LE (2009) J Pharm Sci 98:4639–4649
71. Bach D, Wachtel E, Miller IR (2009) Chem Phys Lipids 157: 108. Baymak MS, Zuman P (2007) Tetrahedron 63:5450–5454
51–55 109. Muangsiri W, Kearney WR, Teesch LM, Kirsch LE (2005) Int J
72. Wachtel E, Bach D, Epand RF, Tishbee A, Epand RM (2006) Pharm 289:133–150
Biochemistry 45:1345–1351 110. Cordes EH, Jencks WP (1962) J Am Chem Soc 84:832–837
73. Fishkin NE, Sparrow JR, Allikmets R, Nakanishi K (2005) Proc 111. Cordes EH, Jencks WP (1962) J Am Chem Soc 84:4319–4328
Natl Acad Sci USA 102:7091–7096 112. Feniouk BA, Cherepanov DA, Junge W, Mulkidjanian AY
74. Oak JH, Nakagawa K, Miyazawa T (2002) J Lipid Res 43: (1999) FEBS Lett 445:409–414
523–529 113. Cherepanov DA, Feniouk BA, Junge W, Mulkidjanian AY
75. Bouifraden S, Drouot C, Hadrami M, Guenoun F, Lecointe L, (2003) Biophys J 85:1307–1316
Mai N, Paris M, Pothion C, Sadoune M, Sauvagnat B, Amblard 114. Adelroth P, Brzezinski P (2004) Biochim Biophys Acta
M, Aubagnac JL, Calmes M, Chevallet P, Daunis J, Enjalbal C, 1655:102–115
Fehrentz JA, Lamaty F, Lavergne JP, Lazaro R, Rolland V, 115. Nachliel E, Gutman M, Kiryati S, Dencher NA (1996) Proc Natl
Roumestant ML, Viallefont P, Vidal Y, Martinez J (1999) Acad Sci USA 93:10747–10752
Amino Acids 16:345–379 116. Levi V, Villamil Giraldo AM, Castello PR, Rossi JP, González
76. O’Donnell JP (1982) Drug Metab Rev 13:123–159 Flecha FL (2008) Biochem J 416:145–152

123
Page 12 of 12 Theor Chem Acc (2012) 131:1263

117. Cervinka O (1969) Enamines, synthesis, structure and reactions. 127. Nagle JF, Tristram-Nagle S (1983) J Membrane Biol 74:1–14
Marcel Dekker, New York 128. Moncelli MR, Becucci L, Guidelli R (1994) Biophys J 66:1969–
118. Kayser RH, Pollack RM (1977) J Am Chem Soc 99:3379–3387 1980
119. Pollack RM, Brault M (1976) J Am Chem Soc 98:247–248 129. Tobias DJ (2001) Curr Opin Struct Biol 11:253–261
120. de Carvalho AF, Pilo-Veloso D, Nelson DL (1996) J Braz Chem 130. Tobias DJ (1999) In: Bellissent-Funel MC (ed) Hydration pro-
Soc 7:225–232 cesses in biology: theoretical and experimental approaches. IOS
121. Raman A, Christou M, Gorrod JW (1987) Eur J Drug Metab Press, Amsterdam
Pharmacokinet 12:279–283 131. Flores-Morales P, Gutiérrez-Oliva S, Silva E, Toro-Labbé A
122. Godoy-Alcántar C, Yatsimirsky AK, Lehn JM (2005) J Phys (2010) J Mol Struc-Theochem 943:121–126
Org Chem 18:979–985 132. Hwang PH, Lian L, Zavras AI (2012) Med Hypotheses 78:
123. Liao RZ, Ding WJ, Yu JG, Fang WH, Liu RZ (2007) J Phys 197–202
Chem A 111:3184–3190 133. Warnakulasuriya S, Parkkila S, Nagao T, Preedy VR, Pasanen
124. Sayer JM, Pinsky B, Schonbrunn A, Washtein W (1974) J Am M, Koivisto H, Niemelä O (2008) J Oral Pathol Med 37:157–
Chem Soc 96:7998–8009 165
125. Rosenberg S, Silver SM, Sayer JM, Jencks WP (1974) J Am 134. Hazen SL, Heller J, Hsu FF, d’Avignon A, Heinecke JW (1999)
Chem Soc 96:7986–7997 Chem Res Toxicol 12:19–27
126. Teissié J, Prats M, Soucaille P, Tocanne JF (1985) Proc Natl
Acad Sci USA 82:3217–3221

123

View publication stats