3-chlorotyramine acting as ligand of the D2 dopamine receptor. Molecular modeling, synthesis and D2 receptor affinity

Article abstract of DOI:10.1002/minf.201400093

We synthesized and tested 3-chlorotyramine as a ligand of the D₂ dopamine receptor. This compound displayed a similar affinity by this receptor to that previously reported for dopamine. In order to understand further the experimental results we performed a molecular modeling study of 3-chlorotyramine and structurally related compounds. By combining molecular dynamics simulations with semiempirical (PM6), ab initio and density functional theory calculations, a simple and generally applicable procedure to evaluate the binding energies of these ligands interacting with the D₂ dopamine receptors is reported here. These results provided a clear picture of the binding interactions of these compounds from both structural and energetic view points. A reduced model for the binding pocket was used. This approach allowed us to perform more accurate quantum mechanical calculations as well as to obtain a detailed electronic analysis using the Quantum Theory of Atoms in Molecules (QTAIM) technique. Molecular aspects of the binding interactions between ligands and the D₂ dopamine receptor are discussed in detail. A good correlation between the relative binding energies obtained from theoretical calculations and experimental IC50 values was obtained. These results allowed us to predict that 3-chlorotyramine possesses a significant affinity by the D₂-DR. Our theoretical predictions were experimentally corroborated when we synthesized and tested 3-chlorotyramine which displayed a similar affinity by the D₂-DR to that reported for DA.

Full text of DOI:10.1002/minf.201400093

Full Paper
www.molinf.com
DOI: 10.1002/minf.201400093
3
-Chlorotyramine Acting as Ligand of the D Dopamine
2
Receptor. Molecular Modeling, Synthesis and D Receptor
2
Affinity
Emilio Angelina,
[
a, d]
[a, b]
[c]
[a, b]
[c]
Sebastian Andujar,*
Laura Moreno, Francisco Garibotto,
Javier Pꢀrraga,
[
d]
[e]
[f]
[c]
[a, b]
Nelida Peruchena, Nuria Cabedo, Margarita Villecco, Diego Cortes, and Ricardo D. Enriz
Abstract: We synthesized and tested 3-chlorotyramine as
was used. This approach allowed us to perform more accu-
rate quantum mechanical calculations as well as to obtain
a detailed electronic analysis using the Quantum Theory of
Atoms in Molecules (QTAIM) technique. Molecular aspects
of the binding interactions between ligands and the D2
dopamine receptor are discussed in detail. A good correla-
tion between the relative binding energies obtained from
theoretical calculations and experimental IC50 values was
obtained. These results allowed us to predict that 3-chloro-
a ligand of the D dopamine receptor. This compound dis-
2
played a similar affinity by this receptor to that previously
reported for dopamine. In order to understand further the
experimental results we performed a molecular modeling
study of 3-chlorotyramine and structurally related com-
pounds. By combining molecular dynamics simulations
with semiempirical (PM6), ab initio and density functional
theory calculations, a simple and generally applicable pro-
cedure to evaluate the binding energies of these ligands in-
tyramine possesses a significant affinity by the D -DR. Our
2
teracting with the D dopamine receptors is reported here.
theoretical predictions were experimentally corroborated
when we synthesized and tested 3-chlorotyramine which
2
These results provided a clear picture of the binding inter-
actions of these compounds from both structural and ener-
getic view points. A reduced model for the binding pocket
displayed a similar affinity by the D -DR to that reported for
2
DA.
Keywords: 3-Chlorotyramine · MD simulations · QTAIM analysis · D dopamine receptor
2
1
Introduction
In the past the potential clinical usefulness of centrally
acting dopamine receptor (DR) agonists has stimulated in-
tense research on new dopaminergic agents. Thus, in gen-
eral, different efforts in medicinal chemistry and neurophar-
macology have yield substantial numbers of compounds
Particularly noteworthy is the work of Reutlinger, et al. who
have recently reported the development and application of
a computational molecular design method for obtaining
bioactive compounds with desired on- and off target bind-
ing. The authors used the molecular algorithm (MAntA)
which effectively transfers a nature-inspired optimization
[7]
[8]
[1–6]
with activity and selectivity at each of the major DRs.
[
a] E. Angelina, S. Andujar, F. Garibotto, R. D. Enriz
Exactas y Naturales y Agrimensura, Universidad Nacional del
Nordeste
Avda. Libertad 5460, (3400) Corrientes, Argentina
Instituto Multidisciplinario de Investigaciones Biologicas (IMIBIO-
SL -CONICET)
Chacabuco 915, 5700 San Luis, Argentina
phone: (54) 2652.423789
[
e] N. Cabedo
Centro de Ecologꢀa Quꢀmica Agrꢀcola-Instituto Agroforestal
Mediterrꢁneo, Universidad Politꢃcnica de Valencia, Campus de
Vera
*e-mail: saanduja@unsl.edu.ar
[
b] S. Andujar, F. Garibotto, R. D. Enriz
Facultad de Quꢀmica, Bioquꢀmica y Farmacia, Universidad
Nacional de San Luis
Edificio 6C, 46022 Valencia, EspaÇa,
[
f] M. Villecco
Chacabuco 915, 5700 San Luis, Argentina
Instituto de Quꢀmica Orgꢁnica, Facultad de Bioquꢀmica Quꢀmica y
Farmacia, Universidad Nacional de Tucumꢁn
Ayacucho 471, S. M. de Tucumꢁn, T4000INI. Argentina
[
c] L. Moreno, J. Pꢁrraga, D. Cortes
Departamento de Farmacologꢀa, Laboratorio de Fꢁrmacoquꢀmica,
Facultad de Farmacia, Universidad de Valencia
Supporting Information for this article is available on the WWW
under http://dx.doi.org/10.1002/minf.201400093.
4
6100 Burjassot, Valencia, EspaÇa
[
d] E. Angelina, N. Peruchena
Laboratorio de Estructura Molecular y Propiedades, ꢂrea de
Quꢀmica Fꢀsica, Departamento de Quꢀmica, Facultad de Ciencias
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&1
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
principle to chemistry-driven molecular design; and they
successfully designed antagonist compounds for the dopa-
mine D4 receptor demonstrating the successful application
of MAntA to dopamine receptors. However, despite the ef-
forts of these authors, in general the design and develop-
ment of DR-selective ligands remains largely empirical,
quite conservative in following molecular precedents,
somewhat unpredictable and not ready for routine applica-
tions of computer-aided drug design techniques. The com-
putational limitations and/or errors can arise in many ways
depending on the methodology: some possibilities include
errors in the structure of the host, choice of ionization
state, structure of the complex, inadequate sampling of in-
ternal degrees of freedom and the so called “ligand-recep-
tor stereo-electronic problem.
addition Claudi et al. reported that 2-chlorothyramine (com-
pound 2) is not able to discriminate between the two sub-
types of dopamine receptors (D and D ), and has seven-
1
2
[17]
fold lower affinity than dopamine for both sites.
It must be pointed out that although 3-chlorotyramine
(2-(3-chloro-4-hydroxyphenyl)ethanaminium chloride) (com-
[19]
pound 3) has been previously reported by Fuller in 1971
[20]
and Stark and Fuller in 1972; however, according to our
bibliographic information, the dopaminergic effect of com-
pound 3 has not been analyzed yet as a D -DR ligand. One
2
possible explanation for this is that due to the low activity
obtained for the analogues of this compound, compound 3
has been directly discarded as a potential ligand of interest
for the D -DR.
2
The replacement of OH group by halogen introduces sig-
nificant changes in the biological response of these ligands.
For example it is interesting to note that in the series of
benzazepines the 2,3,4,5-tetrahydro-7,8-dihidro-3-methyl-1-
phenyl-1H-3-benzazepine (compound 9) had only micromo-
In order to obtain new dopaminergic agonists, The dopa-
mine molecule (compound 1 in Figure 1) has been modi-
fied on the amino group, on the ethylamine chain, and on
[9–17]
the catechol moiety as well.
Claudi et al. reported the
[21]
synthesis and binding affinity for D and D subtypes of
lar affinity at D and D receptors. However 2,3,4,5-tetra-
1
2
1
2
DRs of 2-(4-fluoro-3-hydroxyphenyl)ethylamine (compound
hydro-7-chloro-8-hydroxy-3-methyl-1-phenyl-1H-3-benzaze-
pine (compound 10), obtained by replacing the 7-OH of 9
with a chlorine, has nanomolar affinity possessing a high
selectivity by D1 receptors. In the same way our own stud-
ies performed on a series of benzyl-tetrahydroisoquinolines
(BTHIQs) indicated that the presence of a chlorine group at
[16]
7
, Figure 1).
This compound showed about two-fold
lower affinity than dopamine for both binding sites. Previ-
ously Cardinelli et al reported low activity for other fluorine
derivatives of dopamine including 2-(3-fluoro-4-hydroxy-
[15]
phenyl)ethylamine (compound 6). Studies on aminotetra-
lins showed that 2-amino-6-chloro-7-hydroxytetralins (com-
pound 8) are weakly effective in the binding assays. In
C7 is very important for the adrenergic effect of such deriv-
[18]
[22–24]
atives (compounds 11–14).
In fact such BTHIQs pos-
sessing a chlorine group at C7 displayed the strongest af-
[23]
finity for both D and D DRs in this series. It should be
1
2
noted that the inductive effect of chlorine could influence
the acidity of the phenolic group and produce different in-
teractions at the binding pocket altering the affinity for D2-
DR. One question which arises is if compound 3, which has
not been previously studied as a D -DR ligand, have low or
2
high affinity for this receptor and in either case try to find
an explanation at sub-molecular level for such behavior.
Another question which might arise is why to evaluate
compound 3 and why now? There are different reasons to
analyze this molecule with the computational approaches
now available. On one hand it is possible to exploit recent
information obtained for dopamine with respect to its bio-
[25]
logically relevant conformation.
On the other hand is
also now possible to study with some accuracy the elec-
tronic effects introduced to the ligand when replacing an
OH by Cl and the effects of such change when the ligand is
interacting with its biological receptor. Recently we report-
ed a comprehensive conformational study of dopamine in-
teracting with the D -DR, using a combination of molecular
2
dynamics (MD) simulations, semiempirical and DFT calcula-
[
25]
tions.
In addition, a detailed electronic analysis using
[26–28]
Quantum Theory of Atoms in Molecules (QTAIM)
tech-
nique was also carried out. By using this approach it is pos-
sible to evaluate with some details the conformational and
electronic behaviours of dopamine and its halogenated de-
rivatives.
Figure 1. Structural feature of dopamine and derivatives.
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&2
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
[47]
Halogens, especially the lighter fluorine and chlorine, are
widely used substituents in medicinal chemistry. Until re-
cently, they were merely perceived as hydrophobic moiet-
ies and Lewis bases in accordance with their electronega-
tivities. Much in contrast to this perception, compounds
containing chlorine can also form directed close contacts of
the type RÀX·YÀR’, where the halogen X acts as a Lewis
acid and Y can be any electron donor moiety. This interac-
DA molecule. A further interesting finding was that addi-
tion of a 4-hydroxyl group (p-tyramine) to beta-phenyle-
thylamine (compound 4) does not enhance affinity, but ad-
dition of a 3-hydroxyl group (m-tyramine) (compound 5) is
favorable. When the 3-and 4- hydroxyls are present (the
case of dopamine), the affinity is enhanced over the effects
of the individual hydroxyl groups. These results are consis-
tent with a productive and specific interaction of the two
hydroxyl groups of DA with a network of serine residues.
[29]
tion, referred to as “halogen bonding” since 1978
is
[25]
driven by the s-hole, a positively charged region on the
hind side of X along the RÀX bond axis that is caused by
Our previous results
suggested that the region near
these serine residues may be rather mobile which is consis-
tent with a model of receptor function where the binding
of DA locks the receptor into a relatively rigid conformation
with precise interactions between ligand and receptor.
On the other hand it is well-known that the non covalent
interactions generally are weaker than the covalent ones;
such interactions are more difficult to describe properly.
However, recent advances in computational calculation of
the electron charge density make possible the proper de-
scription of the three-dimensional network of bonding and
[30,31]
an anisotropy of electron density on the halogen.
The
intriguing formal similarity between halogen bonding, RÀ
X—B, and hydrogen bonding, R–H—B, has been noted and
[32,33]
discussed in detail by Legon.
However, he also does
point out that the hydrogen bond is more likely to be non-
linear. In turn a halogen bond is a highly directional, elec-
trostatically-driven non-covalent interaction between
a region of positive electrostatic potential on the outer side
of the halogen X in a molecule RÀX and a negative site B,
such as a lone pair of a Lewis base or the p-electrons of an
unsaturated system. In a very elegant paper Politzer et al.
have introduced the “s-hole concept” allowing to explain
the empirically-observed characteristics of halogen bond-
ing: its marked directionality, its dependence upon the po-
larizability and electronegativity of the halogen atom, and
the role of the electron-withdrawing power of the R por-
[27,48–50]
non bonding interactions in biological systems
in the
context of the quantum theory of atoms in molecules
[
51]
(QTAIM). Starting with strong and moderated hydrogen
bonds or halogen bonds, moving on to weaker polar inter-
actions and ending with stacking and T-shape interactions
between aromatic rings, all of them can be evaluated by
[52]
QTAIM analysis. In fact, nowadays it is well know that the
stacking amino acids aromatic rings in proteins is evidently
much more important than it has been previously believed
and, indeed, can form one of the dominant stabilizing con-
[34]
tion of any molecule.
In a recent review on different types of proteinÀligand
[
35]
interactions relevant to medicinal chemistry, the authors
conclude that halogen bonds are a useful addition to the
arsenal of favorable interactions in molecular recognition
and can lead to significant affinity gains in some cases.
There are a number of experimental studies where the
effect of halogen substitution on binding affinity has been
[28]
tributions.
Our previous results show that the interac-
tions of the catechol OH groups of the ligand, in the differ-
ent conformations of the dopamine/D -DR complex, deter-
2
mine the decrease or increase of the electron density on
the aromatic ring of dopamine. In turn, the electronic pop-
ulation of the aromatic ring of dopamine defines its orien-
tation within the binding site and the type of interactions
that are established with the aromatic rings of the receptor.
It is evident that a description non quantum mechanical of
the problem would overlook these electronic effects that
are crucial to understand the binding modes of the ligand
within the receptor binding site.
[36–38]
systematically evaluated.
The strengths of halogen
bonds can be evaluated theoretically through quantum
chemical model calculations. It is evident that halogen
bonding is best described theoretically using high-level
[
39]
quantum chemical methods such as coupled cluster
[40,41]
(
CCSD-(T)) and perturbation theory
(MP2) calculations;
yet using these methods greatly limits the size of the
model systems amenable to computational studies. Thus,
much larger systems (like those studied here) can be treat-
We were particularly intrigued to know which is the affin-
ity of compound 3 by the D -DR and whether it was possi-
2
[
42,43]
[44]
ed using QM/MM calculations,
or using reduced model systems.
semiempirical studies
ble to explain such behavior through computational simu-
lations. Thus, first we synthesized and tested compound 3
[25]
Regarding the structural aspects of the D -DR it is well
as a D -DR ligand. Whereas this compound showed an af-
2
2
known that the binding of DA to the receptor is substan-
tially affected by multiple serine/alanine mutations. The
multiple mutations including a S193 susbstitution produce
finity comparable to that reported for dopamine in
a second stage of our study we performed MD simulations
for the complexes of compounds 1–5 with the D -DR. It
2
[45,46]
the greatest effect.
It has been previously reported that
should be noted that these compounds were carefully se-
lected in function of their substituents at the catecholic
portion. The selected compounds were: compounds 4 and
5 possessing only one hydroxyl group at para and meta po-
sition, respectively and compounds 2 and 3 possessing
chlorine and hydroxyl groups at para and meta positions,
the effects of the multiple mutations were not additive,
with the single serine mutation having relatively larger ef-
fects, which is indicative of a very precise network of hydro-
gen bonds between the TMV (transmembrane spanning
region) serine residues and the catechol hydroxyls of the
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&3
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
respectively. We also include in our comparative study dop-
amine (compound 1), possessing the catecholic ring. The
next step was to construct a reduced model for the binding
site which allowed us to perform more accurate quantum
mechanical calculations. The fourth step was to simulate
the molecular interactions between the different ligands
and purified by semi-preparative HPLC using a Tracer Excel
120 ODS-B C18 column, 5 mm (25.0ꢂ1 cm), and MeOH/H O
2
in 1% HOAc (20 :80) as mobile phase with a flow of 2 mL/
min. The 3-chloro-4-hydroxy-b-fenilethylamine (7 mg,
0.041 mmol, 75% yield) was isolated with a retention time
1
(Rt) of 14.5 min as a white solid. H NMR* (300 MHz, CDCl )
3
and the D -DR using a QTAIM analysis; being the principal
d 7.20 (s, 1H, H-2), 6.9 (s, 1H, H-6), 6.8 (s, 1H, H-5), 3.2 (m,
2
+
goal of such calculations try to obtain a detailed descrip-
tion of the molecular interactions which stabilize and desta-
bilize the different complexes. It is clear that a detailed
analysis of the such interactions would be of paramount
importance to determine the intricacies of the network of
2H, CH -a), 2.8 (m, 2H, CH -b); ESMS m/z (%) 172 [M+1]
2
2
(39), 155 (100).
2.3 Binding Experiments
serine residues located at the binding pocket of the D -DR.
These experiments were performed on striatal membranes.
Each striatum was homogenized in 2 mL ice-cold Tris-HCl
buffer (50 mM, pH 7.4 at 228C) with a Polytron (4 s, maxi-
mal scale) and immediately diluted with Tris buffer. The ho-
mogenate was centrifuged on four times at 20000g for
2
Finally, we performed a comparative analysis among the
different complexes and the conclusions are put forward in
the last section.
1
0 min at 48C with resuspension in the same volume of
Tris buffer between centrifugations. The final pellet was re-
suspended in Tris buffer containing 120 mM NaCl, 5 mM
KCl, 1 mM CaCl2, 1 mM MgCl2 and 0.1% ascorbic acid (Tris-
ions buffer), and the suspension was treated as described
above. A 200 mL aliquot of freshly prepared membrane sus-
pension (200 mg of striatal protein) was incubated for 1 h
at 258C with 200 mL of Tris buffer containing [3H] raclopr-
ide (0.5 nM, final concentration) and 400 mL of Tris-ions
buffer containing the drug under investigation. Non-specif-
ic binding was determined in the presence of 50 mM apo-
morphine and represented around 5–7% of the total bind-
ing. In both cases, incubations were stopped by addition of
3 mL of ice-cold buffer (Tris-Mg buffer or Tris-ions buffer, as
appropriate) followed by rapid filtration through Whatman
GF/B filters. Tubes were rinsed with 3 mL of ice-cold buffer,
and filters were washed with 3ꢂ3 mL ice-cold buffer. After
the filters had been dried, radioactivity was counted in
4 mL BCS scintillation liquid at an efficiency of 45%. Filter
blanks corresponded to approximately 0.5% of total bind-
ing and were not modified by drugs.
2
Methods
2
.1 Chromatographic and Spectroscopic Analysis
The reaction was monitored by analytical TLC with silicagel
0 F254 (Merck 5554). The residue was purified through
6
silica gel CÀ18 (SPE, Alltech, 100 mg/1.5 mL) column chro-
matography. Isolation and purification was carried out by
a Waters HPLC system with a 600 pump and both a 2996
Photodiode Array Detector (PDA) and ELSD 2420 Detector
1
(
Milford, MA). H, NMR spectra was recorded on a Bruker AV
3
00 MHz instrument (Rheinstetten, Germany). Chemical
shifts (d) are reported in ppm for a solution of the com-
pound in CDCl and the coupling constants (J) values are
3
given in Hz. High resolution ESIMS (electrospray) data were
carried out on a Micromass Q-TOF Micro coupled with
a HPLC Waters Alliance 2695 (Milford, MA). The instrument
was calibrated by using a PEG mixture from 200 to 1000
MW (resolution specification 5000 FWHM, deviation
<
5 ppm RMS in the presence of a known lock mass). The
HCl salts of the synthesized compounds were prepared
from the corresponding base with 5% HCl in MeOH. N-(4-
benzyloxy)-3-chlorophenethyl)benzamide was prepared by
standard methods from 3-chloro-4-methoxybenzaldehyde
2
.4 Molecular Modeling
A 3D model of the human D -DR was used for the MD sim-
ulations. This model is based on the homology model from
2
[22]
and benzoyl chloride.
the crystallized D -DR, b adrenoreceptor and A2a adeno-
3
2
[49,53,54]
sine receptor as templates.
Molecular docking simulations were performed with Au-
2
.2 Synthesis of 3-Chlorothyramine (3)
[55]
A solution of the N-(4-benzyloxy)-3-chlorophenethyl)benza-
mide (1, 20 mg, 0.055 mmol) in a mixture of 2.5 N HCl/
HOAc (4 :1 mL) was heated for 42 h at 1008C. Then, the sol-
vent was removed under reduced pressure and the residue
toDock4 program using rotatable bonds in the ligands
and flexible side chains in selected amino acids of the fifth
transmembrane domain due to its implication in the forma-
[47]
tion of binding pocket.
Assignation of atomic partial
partitioned between H O/EtOAc. The phases were separat-
charges (Gasteiger), rotable bonds as well as merging of
non-polar hydrogens were performed with AutoDock Tools
1.5.2. The ligands were then docked inside a cubic GRID
box (70ꢂ70ꢂ70 ꢃ, grid points separated by 0.375 ꢃ) cen-
tered at the midpoint between the D114 and S194 alfa car-
2
ed and the acid aqueous layer evaporated under reduced
pressure to give a residue which was purified by column
chromatography CÀ18 (SPE, Alltech, 100 mg/1.5 mL) and
[
55]
eluted in a gradient from 100% H O to 100% MeOH. The
2
first fraction eluted with H O/MeOH (9 :1) was evaporated
bons (both residues conserved at the D -DR putative bind-
2
2
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&4
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
ing site). This docking simulation was achieved under the
hybrid Lamarckian genetic algorithm with an initial popula-
tion of 100 randomly placed individuals and a maximum
points r where D1(r)=0. Critical points are classified ac-
cording to their type (w,s) by stating their rank, w, and sig-
nature, s. The rank is equal to the number of nonzero ei-
c
7
number of energy evaluations set at 1ꢂ10 . All of the other
genvalues of the Hessian matrix of 1(r) at r , while the sig-
c
[
55]
parameters were maintained at their default settings. The
resulting docked orientations within a root mean square
deviation (RMSD) of 0.5 ꢃ were clustered together. From
the lowest energy cluster, the best ranked conformation for
each compound was then soaked in boxes of explicit water
nature is the algebraic sum of the signs of the eigenvalues
of this matrix. Critical points of (3, À1) and (3, +1) type de-
scribe saddle points, while the (3, À3) is a maximum and
(3, +3) is a minimum in the field. Among these critical
points, the (3, À1) or bond critical points are the most rele-
vant ones since they are found between any two atoms
linked by a chemical bond. The determination of all the
bond critical points and the corresponding bond paths
connecting these point with bonded nuclei, were per-
[56]
using the TIP3P model and subjected to MD simulation.
All MD simulations were performed with the Amber soft-
[57,58]
ware package (All-atoms force field FF99SB)
using peri-
odic boundary conditions and cubic simulation cells. The
[59,60]
[66]
particle mesh Ewald method (PME)
was applied using
formed with the AIMAII software.
a grid spacing of 1.2 ꢃ , a spline interpolation order of 4
and a real space direct sum cutoff of 10 ꢃ. The SHAKE algo-
rithm was applied allowing for an integration time step of
3
Results and Discussion
2
fs. MD simulations were carried out at 300 K target tem-
perature and extended to 10 ns overall simulation time. Po-
sitional restraints were applied to all the backbone alpha
carbons of the receptor except those of the transmem-
brane 5 (TM5) domain that were left free. The NPT ensem-
ble was employed using Berendsen coupling to a baro/
thermostat (target pressure 1 atm, relaxation time 0.1 ps).
As was pointed out in the introduction it is curious that
until now it has not been reported the affinity for the D2-
DR of compound 3 which is relatively a simple molecule, at
least from a structural point of view. One possible reason is
that in that moment when the structurally related com-
pounds were studied, this compound was discarded with-
out an exhaustive analysis. It is also fair to note that at the
time in which the analogues of compound 3 were reported
it was not possible to perform theoretical calculations with
a relatively detailed electronic analysis as has been done in
this work. Anyway, the first thing we did was to synthesize
[57]
Post MD analysis was carried out with program PTRAJ.
2.5 MM-GBSA Free Energy and Per-Reside Decomposition
Analysis
The MM-GBSA protocol was applied to calculate the rela-
tive binding energies of the complexes by taking snapshots
at 10 ps time intervals from the last 1000 ps of the MD tra-
jectory.
compound 3 and then test its affinity for the D -DR.
2
3.1 Synthesis and Binding Affinity of 3-Chlorotyramine
(
Compound 3)
In order to determine the residues involved in the com-
plexes intermolecular interactions, a per-reside free energy
decomposition analysis using the mm_pbsa program in
Compound 3 has been previously obtained by the Chung’s
[
67]
method, which consists in the oxidative chlorination of
phenols with HCl and m-chloro perbenzoic acid (m-CPBA)
in N,N-dimethyl formamide, and also using an enzymatic
[61,62]
AMBER
was performed.
[68]
procedure. In the course of our work on the synthesis of
2.6 Quantum Mechanics Calculations and Topological Study
[22]
1
-substituted-1,2,3,4-tetrahydroisoquinolines,
the availa-
of the Electron Charge Density Distribution
bility of the N-(4-(benzyloxy)-3-chlorophenethyl)benzamide
(1) allowed us to obtain compound 3 by hydrolysis of the
benzamide bond. Initially and unsuccessfully, various basic
and acid conditions excessively strong for the hydrolysis of
the benzamide (1) were assayed. Finally, we used a hydroly-
Reduced 3D models systems representing the D -DR bind-
2
ing pocket in the five molecular complexes 1/D -DR, 2/D -
2
2
DR, 3/D -DR, 4/D -DR and 5/D -DR were constructed from
2
2
2
the coordinates of the potential energy minimum during
the molecular dynamics simulation trajectory. Then, geo-
metrical optimizations of the reduced model systems were
carried out at semiempirical level employing a PM6
[69]
sis reaction in milder conditions, a mixture of 2.5 N HCl/
HOAc (4 :1) at reflux for 42 h, which afforded the 3-chlori-
nated dopamine analogue (compound 3) in good yield
(Scheme 1).
[63]
method with halogen bonding correction (PM6-D2X) in-
[64]
cluded in the MOPAC program. The optimized models
were then used as input for the calculation of the electron
charge density topological analysis. Single point calcula-
In order to minimize the experimental errors due to the
methodology employed we used a very similar experimen-
[70]
tal protocol to that reported in reference.
Our experi-
[
65]
tions were computed with Gaussian 03
a hybrid PBE functional and 6-31G(d) as basis set.
employing
mental measurements indicated that compounds 3 possess
a Ki value of 0.146 mM. It should be noted that this com-
The topological properties of a scalar field such as 1(r)
pound displayed a significant affinity by the D -DR; In fact
2
are summarized in terms of their critical points, i.e., the
this compound displayed an affinity very similar to that re-
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&5
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Scheme 1. Scheme of synthesis of compound 3.
[
70]
ported for the endogen ligand DA (Ki=0.52 mM). This
result was somewhat surprising to us since we expected
a lower affinity for compound 3 relative to dopamine. After
all the affinity reported for compound 2 is significantly
lower than that of dopamine. From these results we per-
formed a molecular modeling study in which we analyzed
exhaustively the molecular interactions in order to better
understand these experimental results.
compound 3. To try to better understand these results we
compared the results obtained for 1/D -DR complex with
those complexes obtained for compounds 2–5.
Figures 2 and 3 show the complexes obtained for com-
pounds 3 and 2, respectively, superimposed to complex 1/
2
[17]
D -DR for comparison. Figures 1S and 2S (electronic sup-
2
porting material) show complexes 5/D -DR and 4/D -DR, re-
2
2
spectively.
The superposition of the complexes 1/D -DR and 3/D -
2
2
DR (see Figure 2) shows that both structures are very simi-
lar of the orientation of the ligand in the receptor binding
site. In both complexes the ligand adopts extended confor-
mation (torsional angle f1=1718 and 1778 respectively;
see Figure 1 for f1 definition) and the aromatic ring of
compound 3 overlaps almost perfectly with the catecholic
ring of 1, with the difference that the first one is rotated
1808 with respect to dopamine ring. In contrast, the super-
imposition of complexes 2/D -DR and 1/D -DR (see
3
.2 Molecular Dynamic Simulations
Comparing the MD trajectories obtained for the different
complexes, interesting general conclusions might be ob-
tained. Consistent with previous experimental
[45–47]
and the-
[24,25]
oretical
results, our simulations indicate the importance
of the negatively charged aspartate 114 for the binding of
these ligands. It is well-known that a highly conserved as-
partic acid (D114) in transmembrane 3 (TM3) is important
2
2
for the binding of both agonists and antagonists to the D -
Figure 3) reveals a marked deviation of compound 2 orien-
tation with respect to DA, with the first one displaying
a smaller value of the f1 torsional angle (f1=978).
Moreover, by comparison of chlorine atom binding mode
in both chlorinated analogues one can see that in complex
of compound 2 the chlorine atom is located in an hydro-
phobic environment composed of residues I184, F389,
V190 and CÀH bonds from H393 whereas in complex of
compound 3 the same atom establishes interactions with
the more polar residues S193 and S194.
2
[24,71,72]
DR.
In the present study, all the compounds simulat-
ed were docked into the receptor with the protonated
amino group near D114. After 10 ns of MD simulations, the
ligands had moved slightly but in a different form com-
pared with the initial position. However, the strong interac-
tion with D114 was maintained for all the complexes (see
figures 2–5) which support that D114 is the anchoring
point for ligands with a protonated amino group.
In the next step of our study we evaluated the relative
free energy (D(DGbind)) obtained for the different complexes
(
table 1, second column). The results obtained for com-
3
.3 Conformational Change in the Transmembrane 5 (TM5)
pounds 4 and 5 were as expected since both compounds
showed a lower affinity for the D -DR than compound 1.
Domain
2
Regarding compounds 2 and 3 and from the relative bind-
ing energies obtained in our MD simulations, one can ap-
preciate that replacing the OH in meta (m-OH) by a chlorine
Figures 4A and 4B give a different spatial view of the same
complexes shown in Figures 2 and 3. One can see in these
figures that in complexes 1/D -DR and 3/D -DR (see Fig-
2
2
atom increases the affinity for the D -DR in comparison to
ure 4A) the side-chain of S197 from transmembrane 5
(TM5) domain interacts with T119 from transmembrane 3
2
compound 1, indicated by the lowest value of the relative
free energy obtained for compound 3. In contrast replacing
the OH in para position (p-OH) displays the opposite effect,
ie decreases the affinity for the receptor (5.57 kcal/mol
above the value obtained for 3). These results are at least
qualitatively in agreement with the experimental results
(TM3) domain. In contrast, in 2/D -DR complex (see Fig-
2
ure 4B) this interaction has been disrupted during the MD
simulation and the side chain of S197 associates with the
backbone of residue S193 from the same TM5, causing dis-
tortion of the backbone interactions of TM5 and therefore
the change in the structure of this transmembrane domain.
It is worth noting that the change experienced by TM5
[17,70,73]
previously reported for compounds 1, 2, 4 and 5
as
well as with the experimental results reported here for
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&6
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 2. Spatial view of complex 3/D -DR (cyan). Complex 1/D -DR (in gray) is overimposed for comparison.
2
2
Figure 3. Spatial view of complex 2/D -DR (orange). Complex 1/D -DR (in gray) is overimposed for comparison.
2
2
domain the complex 2/D -DR reflects the importance of
gether by CÀH·p interaction type (some of these interac-
2
preserving the flexibility of the backbone of TM5 during
tions are shown in the molecular graphs later) forming
[
53,74]
the simulation, as suggested by other authors.
a conserved structure that has been observed also in D -DR
2
complexes with alkaloids possessing a tetrahidroprotober-
[49]
berine structure. This conserved structure forms specific
interactions with the ligand, contributing, together with
the principal interaction D114 to its anchoring in the
proper orientation within the binding site. On the other
3
Site
.4 Possible Role of the Aromatic Residues of the Binding
From Figures 4A and 4B one also can see how the spatial
distribution of the aromatic residues of the binding pocket
of complexes 2/D -DR and 3/D -DR differs from that of
hand, in complex 2/D -DR the side chains of the phenylala-
2
nine triad show a marked deviation in its rings spatial distri-
2
2
complex 1/D -DR. Figure 4A shows that the side chains of
bution relative to 1/D -DR (see Figure 4B). This alteration in
2
2
F389, F390 and F198 overlap almost perfectly in complexes
the phenylalanine triad structure is a consequence of the
1
/D -DR and 3/D -DR. These three residues are linked to-
distortion that TM5 undergoes in the 2/D -DR complex,
2
2
2
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&7
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 4. Similar spatial view to that shown in Figure 2 (4A) and 3 (4B) but in this case displaying the spatial distribution of the aromatic
residues of the binding site.
which produces a displacement of the F198 (residue that
belongs to TM5) so that it can no longer be associated
with F190. Since phenylalanine triad produces specific in-
teractions with the ligand, the alteration of its structure
necessarily affects the binding mode of the ligand into the
binding site, being this another factor that helps to explain
the change in orientation of compound 2 within the bind-
ing site with respect to compounds 1 and 3.
reduced model systems using combined semiempirical and
DFT calculations.
3.5 Constructing the Reduced Models for the Binding Site
Figure 5 shows the ligand-residue interaction spectra calcu-
lated by the per residue free energy decomposition, which
suggests that the interaction spectra of compounds 1–5
At this stage of our work we consider the trend predict-
ed for the MD simulations as certainly significant but, on
the other hand the approximations involved in this ap-
proach compels us to go beyond the classical treatment of
the interactions to confirm our results. It should be noted
that we are dealing with relatively weak interactions and
therefore MD simulations might underestimate such inter-
actions. Thus, in the next step of our study we constructed
with D -DR are closely related and reflects their similar
2
binding modes. As shown in Figure 5, the residues D114
and V115 are those with the greatest contribution to the in-
teraction energy, this is true not only for compounds 1–5,
but also for other compounds with dopaminergic activity
[49]
previously reported.
Since the salt bridge with D114 is the strongest interac-
tion established in the binding site it is considered a guide-
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&8
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 5. Histograms of interaction energies partitioned for D -DR amino acids when complexed with compound 1 (a), compound 3 (b),
2
compound 2 (c), compound 5 (d) and compound 4 (e). The x-axis denotes the residue number of D -DR, and the y-axis shows the interac-
2
tion energy between the compound and the specific residue. Negative and positive values represent favorable or unfavorable binding, re-
spectively.
line interaction for ligand anchoring into the D -DR binding
lizing and destabilizing interactions showing non negligible
contribution in the per residue energy decomposition spec-
tra. Thus, residues D114, V115, M116, C118, T119, I184, F189,
V190, V191, Y192, S193, S194, I195, V196, S197, F198, W386,
F389, F390, H393, Y408, T412 and Y416 were included in
2
site. With respect to V115, this amino acid forms several CÀ
H·p interactions with the aromatic ring of dopamine. These
interactions are not shown in the molecular graphs dis-
played below for easy viewing of the molecular interactions
of meta-OH/Cl and para-OH/Cl. Moreover, among the serine
residues of the binding site, S197 is the residue that has
the largest contribution to the interaction energy in com-
plexes 1/D -DR and 2/D -DR, being higher in the first one,
the reduced model for the binding pocket of D -DR and
2
therefore a final number of 23 amino acids were included
in our model. A spatial view of this reduced model is
shown in Figure 6.
2
2
while S193 presents the most important contribution in the
complex 3/D -DR. Finally, the residue decomposition analy-
2
3
.6 Quantum Mechanics Calculations and Topological
sis also shows a significant contribution of one of the resi-
dues of the phenylalanine triad, F389, its contribution
being greater in 1/D -DR and 3/D -DR than in 2/D -DR.
Analysis of the Electron Density
The starting geometries for each complex were obtained
from the coordinates of the potential energy minimum
during the simulation time. PM6-D2X optimizations were
performed by considering the mentioned 23 residues from
the receptor binding pocket. Next, DFT (PBE/6-31G(d)
2
2
2
From these results we considered prudent to include in
the reduced model not just those amino acids involved in
the most relevant molecular interactions displayed in the
different spectra, but also all the residues involved in stabi-
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&9
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 6. Spatial view of compound 3 (magenta)/D -DR interaction. Magnification of the receptor active site at the right. The names of the
2
residues involved in the main interactions are written in the figure.
single point calculations were carried out for each of the
PM6-D2X optimized complexes.
Figures 8, 9 and 10 show the molecular graph of electron
density obtained for compounds 1, 3 and 2 in the D -DR
2
binding site, respectively.
3.7 Evaluating the Molecular Interactions for the Different
Complexes
On the basis of the results obtained from the MD simula-
tions we focused the quantum mechanical analysis on dop-
amine (compound 1) and its two chlorinated analogues
3
.8 Catechol Interactions
The molecular graph of Figure 8 shows the most important
interactions of compound 1 with the residues of the bind-
ing site. In particular, it is observed that OH in the meta po-
sition (m-OH) establishes a strong OÀH·S hydrogen bond
(
compounds 2 and 3).
Figure 7 gives the values of 1 summation (S1 ) corre-
b
b
sponding to all the intermolecular interactions (Figure 7A)
and only those interactions involving substituent in the
meta (m-OH/Cl) and para positions (p-OH/Cl) (Figure 7B) of
(HB) with C118 (1 =0.0202 au), other weaker HBs of the
b
type CÀH·O with F390 (1 =0.0023 au), S197 (1 =0.0065 au)
b
b
and V115 (1 =0.0032 au) and two O·O contacts with T119
compounds 1–3 at the D -DR binding site.
b
2
(
1 =0.0061 au) and S197 (1 =0.0072 au). While in position
b b
In accordance with the relative free energy data (see
para the OH behaves as a proton donor against S197 (1 =
Table 1) the S1 values corresponding to all the intermolec-
b
b
0
.0212 au), and as proton acceptor against CÀH bonds of
ular interactions (Figure 7A) shows that compounds 3 and
F390 (1 =0.0034 au). Furthermore, like m-OH, p-OH also
2
are more strongly and more weakly anchored to the D2-
b
establishes a contact of the type O·O with backbone of
S193 (1 =0.0027 au).
DR binding site than compound 1, respectively.
b
It is important to highlight the three O·O interactions in
Table 1. Relative free energy (D(DGbind)) obtained in kcal/mol for
the five complexes studied here. IC50 experimental values are given
in the second column.
the 1/D -DR complex. Even more interesting, the oxygen
2
atoms of S197 and T119 that are engaged to the oxygen in
meta position (m-O) are connected each other through
a strong OH·O hydrogen bond. Thus, the three oxygen
atoms are connected directly or indirectly by bond paths to
give a topological ring (see Figure 7). This kind of O·O con-
Complex
D(DGbind
)
IC50 (mM)
Reference
1
2
3
4
5
/D -DR
3.10
5.57
0.00
4.84
4.00
0.52
26.31
0.15
>300
26.5
[41]
[9]
This work
[42]
[42]
2
/D -DR
2
/D -DR
[25,49]
2
tacts has been already described in previous reports
/D -DR
2
and in such sense we believe they could be involved in the
ligand/receptor recognition process.
/D -DR
2
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
10&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 7. S1 values obtained for all the intermolecular interactions (7A) and only those involving substituents in meta (m-OH/Cl) and posi-
b
tion (p-OH/Cl) (7B) of compounds 1–3 at the D -DR binding site.
2
3.9 Interactions of the Chlorine Atom
the dopamine O·O interactions with carbonyl and hydroxyl
groups of the biological receptor.
As can be seen in Figure 9, the chlorine atom of compound
(in meta position, m-Cl) establishes six interactions with
Going back to Figure 7B, the 1 summation (S1 ) corre-
sponding to interactions of m-Cl and p-Cl shows that the
chlorine atom in compound 2 is more strongly anchored to
b
b
3
the binding site of D -DR. Three CÀH·Cl interactions with
2
residues I184, V190 and H393 (S1 =0.0149 au), two inter-
the D -DR binding pocket than the same atom in com-
b
2
actions of the O·Cl type with S193 (1 =0.0128 au) and
pound 3. Thus, based in these results, it is fair to say that
b
S194 (1 =0.0013 au) and a strong intramolecular (OÀ)H·Cl
the chlorine atom prefers to be in an hydrophobic environ-
b
HB interaction with the OH in para position (1 =
ment as in complex 2/D -DR where only C–H·Cl interac-
b
2
0
.0240 au).
tions are formed rather than a more polar one, as in com-
Moreover, in complex of compound 2, the chlorine atom
plex 3/D -DR where two Cl·O interactions are established.
2
is more “buried” in the hydrophobic region (see Figure 10)
On the basis of this results, the next question one might
forming only CÀH·Cl interactions (a total of six interactions
ask is what then makes compound 3 be more strongly
with residues I184, F189, V190 and H393 (S1 =0.0326 au)).
anchored to the D -DR binding site than compound 2? This
b
2
Note that in this complex the chlorine atom is not forming
an intramolecular HB with m-OH as in the case of com-
pound 3.
issue is explored in the next section.
3.10 Effects of the Chlorine Substitutions on the Interactions
The two Cl·O interactions that are only established in
of the OH Groups
complex 3/D -DR could explain the favorable binding be-
2
havior observed for compound 3. In other words, the Cl·O
Figure 7B shows that when the OH in meta position of
compound 1 is replaced by a chlorine atom as in com-
in complex 3/D -DR might be mimicking the behavior of
2
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
11
&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 8. Molecular graph of compound 1 at the D -DR binding site. Large spheres represent attractors or nuclear critical points (3, À3) at-
2
tributed to the atomic nuclei. Lines connecting the nuclei are bond paths and small spheres on them are bond critical points (3, À1).
Figure 9. Molecular graph of compound 3 at the D -DR binding site. Large spheres represent attractors or nuclear critical points (3, À3) at-
2
tributed to the atomic nuclei. Lines connecting the nuclei are bond paths and small spheres on them are bond critical points (3, À1).
pound 3, it cannot be anchored to the binding site with
the same force as the m-OH of compound 1. However, the
introduction of Cl at meta position improves the interac-
tions of p-OH with respect to the same group in 1. Thus,
both groups together, m-Cl and p-OH are anchored to the
binding site with almost the same strength as the corre-
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
12&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
Figure 10. Molecular graph of compound 2 at the D -DR binding site. Large spheres represent attractors or nuclear critical points (3, À3)
2
attributed to the atomic nuclei. Lines connecting the nuclei are bond paths and small spheres on them are bond critical points (3, À1).
sponding catecholic hydroxyls of 1. Thus, the presence of
the chlorine atom in the meta position modifies the bind-
ing site (throught the Cl·O interactions with serine resi-
dues) in order to strengthen interaction of p-OH.
On the other hand, when p-OH of compound 1 is re-
placed by chlorine as in compound 2, the halogen provides
stronger interactions than the corresponding OH group.
However, the introduction of chlorine significantly weakens
the interactions of m-OH in comparison to the same group
hydroxyl group to interact with the serine residues from
the receptor binding pocket. Furthermore, it was shown
that the binding mode of compound 2 to D -DR deter-
2
mines a distortion of TM5 that in turn disturbs specific in-
teractions between phenylalanine resides of the binding
pocket and the ligand (Figures 4A,B).
3.11 Scope and Limitations of the Methodology Used Here
in 1/D -DR complex, so that both groups together bind
much more weakly to the binding site that the correspond-
ing OH groups of compound 1.
Let us now make some comments on the scope and limita-
tions that might have the application of calculations used
in this work.
2
As discussed above, in complex 2/D -DR the chlorine
atom is more “buried” in the hydrophobic pocket formed
With respect to the applicability to other biological sys-
tems, It is important to note that studies using molecular
techniques QTAIM applied to large systems of biological in-
terest are relatively new and therefore there are few studies
previously reported in the literature. However, it is impor-
tant to note that these simulations have been used suc-
2
by residues I48, F154 and V190 that in the 3/D -DR com-
2
plex. This determines that m-OH be poorly positioned to in-
teract with the serines of the binding site.
Summarizing the results of the molecular modeling sec-
tion, it was shown that the binding mode of compound 3
[48]
cessfully in inhibitors of dihydrofolate reductase (DHFR)
[75,76]
in the D -DR binding pocket resemble the DA binding
as well as with BACE 1 inhibitors.
Although the types of
2
mode (see Figure 2) whereas in compound 2, the anchor-
ing of the chorine atom in an hydrophobic pocket deter-
mines a different binding mode for this compound (see
Figure 3). While the chlorine atom in compound 2 is more
receptors studied so far are very few, at least it is interest-
ing to note that it is possible to do this kind of simulations
successfully in different types of receptors. Now we are
doing studies on different types of ligand-receptor com-
plexes with different types of interactions (from very strong
to very weak ones). It is clear that the results obtained in
such study will give us a better picture about the scope of
applicability of the simulations used here.
strongly anchored in the D -DR binding pocket than the
2
same atom in compound 3 (Figure 7B), the overall binding
of the last compound is stronger than the first one (Table 1
and Figure 7A) in part due to the better placement of the
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
13&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
With respect to the applicability of these simulations to
structurally unrelated chemotypes. We have recently re-
ported some works in which we have studied compounds
By combining MD simulations with semiempirical and
DFT calculations, a simple and generally applicable proce-
dure to evaluate the binding energies of ligands interacting
[25,49,50]
structurally different with this type of approach.
In
with the D -DR has been reported here. Thus, our theoreti-
2
a study of DHFR inhibitors it was possible to include in the
correlation two new series of compounds possessing signif-
icant structural differences with respect to the known clas-
cal and experimental results contribute to the understand-
ing of the non covalent interactions in the context of the
ligand - receptor binding event in a two-way manner, by
providing a detailed topological description of the interac-
tion network of the ligand in the receptor binding pocket
and by showing the convenience of going beyond the con-
cept of pair-wise interactions in order to “see” the electron-
ic effects within the intricate biological environment. Un-
doubtedly the results presented here show the importance
of performing these studies as comprehensive as possible
when hydrogen bonds are involved and even more if there
are halogen atoms involved in these interactions. Thus, we
believe our results may be helpful in the structural identifi-
cation and understanding of the minimum structural re-
quirements for these molecules and can provide a guide in
[48]
sical and non-classical inhibitors. In the particular case of
D -DR in previous works we have reported the analysis of
2
[24]
BTHIQs compounds as well as of protoberberine deriva-
[49]
tives by using this type of simulations. While the structur-
al variability is still very meager, the results obtained so far
are very promising for this kind of approach using reduced
models and QTAIM calculations. Thus, it appears very rea-
sonable that these simulations might be successfully ap-
plied to compounds with different structural chemotypes.
Importantly, in all the above cases, these studies including
quantum mechanical calculations and QTAIM studies have
significantly improved correlations between experimental
and theoretical results in a quantitative mode which indi-
cates how valuable is this protocol for prioritizing certain
chemotypes.
the design of new ligands for the D receptor of dopamine.
2
A somewhat negative aspect also needs to be highlight-
ed. The protocol used is far from to be considered a post-
docking routine method in its current form. However, con-
sidering that the type of calculations performed are not
particularly sophisticated, it would be possible to design
a protocol developing the various steps in a more system-
atic and friendly way. We are working on it.
List of the Abbreviations
QTAIM
DR
Quantum theory of atoms in molecules
Dopamine receptor
DA
Dopamine
MD
Molecular dynamics
BTHIQs Benzyl-tetrahydroisoquinolines
Acknowledgements
4
Conclusions
Grants from Universidad Nacional de San Luis (UNSL), partial-
ly supported this work. This research was also supported
by PIP 095 CONICET. R. D. Enriz; N. M. Peruchena, F. M. Gari-
botto, E. L. Angelina and S. A. Andujar are members of the
Consejo Nacional de Investigaciones Cientꢀficas y Tꢃcnicas
(CONICET-Argentina) staff.
In this paper we performed a molecular modeling study of
-chlorotyramine and analogues. The theoretical and exper-
3
imental results reported here allowed us to reach at least
two interesting conclusions. On one side we are reporting
that 3-chlorotyramine has a D receptor affinity that it is
2
comparable to that of the endogenous ligand (dopamine).
This result is very interesting because until now it has been
reported that all halogenated dopamine derivatives have
a significantly lower affinity for both D and D receptors.
References
1
2
On the other hand it should be noted that our theoretical
study (including a comprehensive analysis of the molecular
interactions) has been able to explain such a significant af-
[1] L. Bettinetti, S. Lçber, S. Hꢄbner, P. Gmeiner, J. Comb. Chem.
2
005, 7, 309–316.
[2] J. W. Kebabian, F. I. Tarazi, N. S. Kula, R. J. Baldessarini, Drug
Discov. Today 1997, 2, 333–340.
finity of 3-chlorotyramine for the D receptor. However, it is
2
[3] R. J. Baldessarini, in The Dopamine Receptors (Eds: K. A. Neve,
R. L. Neve), Humana Press, Totowa, NJ, 1997, 457–498.
important to note that in order to evaluate in detail the
various molecular interactions of ligands at its site of
action, it is necessary to use reduced models systems
which allow to perform quantum mechanical calculations
and QTAIM type studies. While these calculations require
more time than traditional or standard simulations, appa-
rently this effort is justified because it allows a deeper and
more appropriate description of the ligand-receptor inter-
actions.
[4] R. B. Mailman, D. E. Nichols, A. Tropsha, in The Dopamine Re-
ceptors (Eds: K. A. Neve, R. L. Neve), Humana Press, Totowa,
NJ, 1997, 105–133.
[
[
[
5] J. P. Kelleher, F. Centorrino, M. J. Albert, R. J. Baldessarini, CNS
Drugs. 2002, 16, 249–261.
6] M. Rowley, L. J. Bristow, P. H. Hutson, J. Med. Chem. 2001, 44,
4
77–501.
7] M. Reutlinger, T. Rodrigues, P. Schneider, G. Schneider, Angew.
Chem. Int. Ed. 2014, 53, 1–6.
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
14&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
[
8] a) J. A. Hiss, M. Hartenfeller, G. Schneider, Curr. Pharm. Des.
Hartmann, D. W. Banner, W. Haap, F. Diederich, Angew. Chem.,
2010, 16, 1656–1665; b) J. A. Hiss, M. Reutlinger, C. P. Koch,
Int. Ed. 2011, 50(1), 314–318.
A. M. Perna, P. Schneider, T. Rodrigues, S. Haller, G. Folkers, L.
Weber, R. B. Baleeiro, P. Walden, P. Wrede, G. Schneider, Future
Med. Chem. 2014, 6, 267–280.
[37] L. A. Hardegger, B. Kuhn, B. Spinnler, L. Anselm, R. Ecabert, M.
Stihle, B. Gsell, R. Thoma, J. Diez, J. Benz, J. M. Plancher, G.
Hartmann, Y. Isshiki, K. Morikami, N. Shimma, W. Haap, D. W.
Banner, F. Diederich, ChemMedChem. 2011, 6(11), 2048–2054.
[38] R. Wilcken, X. Liu, M. O. Zimmermann, T. J. Rutherford, A. R.
Fersht, A. C. Joerger, F. M. Boeckler, J. Am. Chem. Soc. 2012,
134(15), 6810–6818.
[
9] G. Gallagher, P. J. Lavanchy, J. W. Wilson, J. P. Hieble, R. M. Mari-
nis, J. Med. Chem. 1985, 28, 1533–1536.
10] K. L. Kirk, C. R. Creveling, Med. Res. Reu. 1984, 4, 189–220.
11] R. M. De Marinis, G. Gallagher, R. F. Hall, R. G. Franz, C. Webster,
W. F. Huffmann, M. S. Schwartz, C. Kaiser, S. T. Ross, J. W.
Wilson, P. Hieble, J. Med. Chem. 1986, 29, 939–947.
[
[
[39] H. G. Kꢄmmel, Biography of the Coupled Cluster Method; World
Scientific, Singapore. 2002.
[
12] J. R. McCarthy, J. McCowan, M. B. Zimmerman, M. A. Wenger,
L. W. Emmert, J. Med. Chem. 1986, 29, 1586–1590.
[
[
40] C. Møller, M. S. Plesset, Phys. Rev. 1934, 46(7), 618.
41] J. A. Pople, J. S. Binkley, R. Seeger, Int. J. Quantum Chem. 1976,
[
13] J. Weinstock, D. E. Gaitanopouloe, O. D. Stringer, R. G. Franz,
J. P. Hieble, L. B. Kinter, W. A. Maan, K. E. Flaim, G. Gessner, J.
Med. Chem. 1987, 30, 1166–1176.
1
–19.
42] Y. Lu, Y. Wang, W. Zhu, Phys. Chem. Chem. Phys. 2010, 12(18),
543–4551.
[
[
[
[
[
[
[
4
[
[
[
[
[
14] L. Nedelec, C. Dumont, C. Obrlander, D. Frechet, J. Laurent,
J. R. Bohier, Eur. J. Med. Chem. 1978, 13, 553–563.
15] M. Cardellini, G. M. Cingolani, F. Claudi, V. Perlini, W. Balduini,
F. Cattabeni, Farmaco Ed. Sci. 1988, 43, 49–59.
16] F. Claudi, M. Cardelhi, G. M. Ciolani, A. Piergentili, G. Peruzzi,
W. Balduini, SynWi, J. Med. Chem. 1990, 33, 2408–2412.
17] F. Claudi, G. Giorgioni, A. Di Stefano, M. P. Abbracchio, A. M.
Paoletti, W. Balduinio, J. Med. Chem. 1992, 35, 4408–4414.
18] J. Weinstock, D. E. Gaitanopoulos, O. Hye-Ja, F. R. Pfeiffer, C. B.
Karaeh, J. W. Venslavsky, H. M. Sarau, K. E. Flaim, J. P. Hieble, C.
Kaiser, J. Med. Chem. 1986, 29, 1613–1627.
43] Y. Lu, T. Shi, Y. Wang, H. Yang, X. Yan, X. Luo, H. Jiang, W. Zhu,
J. Med. Chem. 2009, 52(9), 2854–2862.
44] P. Dobes, J. Rezꢀ c´ , J. Fanfrlꢅk, M. Otyepka, P. Hobza, J. Phys.
Chem. B 2011, 115(26), 8581–8589.
45] C. Coley, R. Woodward, A. M. Johansson, P. G. Strange, L. H.
Naylor, J. Neurochem. 2000, 74(1), 358–366.
46] R. Woodward, C. Coley, S. Daniell, L. H. Naylor, P. G. Strange, J.
Neurochem. 1996, 66(1), 394–402.
47] B. A. Cox, R. A. Henningsen, A. Spanoyannis, R. L. Neve, K. A.
Neve, J. Neurochem. 1992, 59, 627–635.
48] R. D. Tosso, S. A. Andujar, L. Gutierrez, E. Angelina, R. Rodrꢅ-
guez, M. Nogueras, H. Baldoni, F. D. Suvire, J. Cobo, R. D. Enriz,
J. Chem. Inf. Model. 2013, 53(8), 2018–2032.
[
19] R. W. Fuller, J. Mills, M. M. Marsh, J. Med. Chem. 1971, 14, 322–
325.
[
[
20] P. Stark, W. Fuller, Neurophormacology 1972, 11, 261–272.
21] K. M. O‘Boyle, J. L. Waddington, Eur. J. Pharmacol. 1985, 115,
[
[
49] J. Pꢀrraga, N. Cabedo, S. Andujar, L. Piqueras, L. Moreno, A.
Galꢀn, E. Angelina, R. D. Enriz, M. D. Ivorra, M. J. Sanz, D.
Cortes, Eur. J. Med. Chem. 2013, 68, 150–166.
50] E. Angelina, S. A. Andujar, R. D. Tosso, R. D. Enriz, N. Peruchena,
Non-covalent interactions in receptor-ligand complexes. A study
based on the electron charge density. CLAFCO Congress, Foz
Iguazu (Brasil), 2013.
291–295.
[
[
[
[
22] I. Berenguer, N. El Aouad, S. Andujar, V. Romero, F. Suvire, T.
Freret, A. Bermejo, M. D. Ivorra, R. D. Enriz, M. Boulouard, N.
Cabedo, D. Cortes, Bioorg. Med. Chem. 2009, 17, 4968–4980.
23] N. El Aouad, I. Berenguer, V. Romero, P. Marꢅn, A. Serrano, S.
Andujar, F. Suvire, A. Bermejo, M. D. Ivorra, R. D. Enriz, N.
Cabedo, D. Cortes, Eur. J. Med. Chem, 2009, 44, 4616–4621.
24] S. Andujar, F. Suvire, I. Berenguer, N. Cabedo, P. Marꢅn, L.
Moreno, M. D. Ivorra, D. Cortes, R. D. Enriz, J. Mol. Model. 2012,
[51] R. Bader, Atoms in Molecules: A Quantum Theory, Oxford Uni-
versity Press, New York, 1990.
[
[
52] S. J. Grabowski, J. Phys. Chem. A 2012, 116, 1838–1846.
53] M. A. Soriano-Ursffla, J. O. Ocampo-López, K. Ocampo-Mendo-
za, J. G. Trujillo-Ferrara, J. Correa-Basurto, Comput. Biol. Med.
18, 419–431.
25] S. A. Andujar, R. D. Tosso, F. D. Suvire, E. Angelina, N. Peruche-
2
011, 41, 537–445.
na, N. Cabedo, D. Cortes, R. D. Enriz, J. Chem. Inf. Model. 2012,
[54] M. Y. Kalani, N. Vaidehi, S. E. Hall, R. J. Trabanino, P. L. Freddoli-
52(1), 99–112.
rd
no, M. A. Kalani, W. B. Floriano, V. W. Kam, W. A. Goddard 3 ,
Proc. Natl. Acad. Sci. 2004, 101, 3815–3820.
[
[
26] U. Koch, P. L. Popelier, J. Phys. Chem. 1995, 99, 9747–9754.
27] C. F. Matta, N. Castillo, R. J. Boyd, J. Phys. Chem. B. 2005, 110,
[
[
[
55] G. M. Morris, D. S. Goodsell, R. S. Halliday, R. Huey, E. Hart, R. K.
Belew, A. J. Olson, J. Comp. Chem. 1998, 19, 1639–1662.
56] W. L. Jorgensen, J. Chandrasekhar, J. D. Madura, R. W. Impey,
M. L. Klein, J. Chem. Phys. 1983, 79, 926–935.
57] D. A. Case, T. E. Cheatham, T. Darden, H. Gohlke, R. Luo, K. M.
Merz Jr., A. Onufriev, C. Simmerling, B. Wang, R. J. Woods, J.
Comput. Chem. 2005, 26, 1668–1688.
5
63–578.
[
28] R. A. Mosquera, M. J. G. Moa, L. Estꢆvez, M. Mandado, A. M.
GraÇa, in Quantum Biochemistry (Ed: C. F. Matta), Wiley-VCH,
Weinheim, Germany, 2010.
[
[
[
29] J. M. Dumas, H. Peurichard, M. Gomel, J. Chem. Res. 1978, 2,
5
4–55.
30] T. Clark, M. Hennemann, J. S. Murray, P. Politzer, J. Mol. Model.
007, 13(2), 291–296.
[
58] D. A. Case, T. A. Darden, T. E. Cheatham III, C. L. Simmerling, J.
Wang, R. A. Duke, R. Luo, R. C. Walker, W. Zhang, K. M. Merz,
AMBER12, University of California, San Francisco. 2012.
2
31] T. Clark, s-Holes, Comput. Mol. Sci. 2012, DOI: 10.1002/
wcms.1113.
[
[
[
32] A. Legon, Phys. Chem. Chem. Phys. 2010, 12, 7736–7747.
33] A. C. Legon, Angew. Chem., Int. Ed. 1999, 38, 2686–2714.
34] P. Politzer, J. S. Murray, T. Clark, Phys. Chem. Chem. Phys. 2010,
[59] T. Darden, D. York, L. Pedersen, J. Chem. Phys. 1993, 98,
10089–10092.
[60] U. Essmann, L. Perera, M. L. Berkowitz, T. Darden, H. Lee, L. G.
Pedersen, J. Chem. Phys. 1995, 103, 8577–8593.
[61] T. Hou, N. Li, Y. Li, W. Wang, J. Proteome Res. 2012, 11, 2982–
95.
1
2, 7748–7757.
35] C. Bissantz, B. Kuhn, M. Stahl, J. Med. Chem. 2010, 53(14),
061–5084.
[
[
5
36] L. A. Hardegger, B. Kuhn, B. Spinnler, L. Anselm, R. Ecabert, M.
Stihle, B. Gsell, R. Thoma, J. Diez, J. Benz, J. M. Plancher, G.
[62] H. Gohlke, C. Kiel, D. A. Case, J. Mol. Biol. 2003, 330, 891–913.
ˇ
[63] J. Rezꢀ cˇ , P. Hobza, Chem. Phys. Lett. 2011, 506, 286–289.
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
15&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
[
64] J. J. P. Stewart, MOPAC2009 Stewart Computational Chemistry,
[73] S. L. Payne, A. M. Johansson, P. G. Strange, J. Neurochem. 2002,
82(5), 1106–1117.
Colorado Springs, CO, USA, 2008.
[
65] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A.
Robb, J. R. Cheeseman, J. A. Montgomery Jr., T. Vreven, K. N.
Kudin, J. C. Burant, Gaussian 03, Revision C05,Gaussian, Inc.,
Wallingford, CT, 2004.
[74] V. Katritch, K. A. Reynolds, V. Cherezov, M. A. Hanson, C. B.
Roth, M. Yeager, R. Abagyan, J. Mol. Recognit. 2009, 22, 307–
318.
[75] L. J. Gutierrez, E. Angelina, N. Peruchena, H. A. Baldóni, R. D.
Enriz, Small-size peptides acting as inhibitors of the BACE1-exo-
site. A molecular modeling study using MD simulations and QM
calculations, 10th Congress of the World Association of Theoreti-
cal and Computational Chemists (WATOC), Santiago, Chile,
[
[
[
[
[
[
[
66] T. Keith, AIMA11, 12.11.09, Gristmill Software, Overland Park,
KS, USA, 2012.
67] K. H. Chung, H. J. Kim, H. R. Kim, E. K. Ryu, Synth. Commun.
1990, 20, 2991–2997.
68] H. Nakazawa, K. Sano, K. Matsuda, K. Mitsugi, Biosci. Biotech.
Biochem. 1993, 57, 1210–1211.
69] M. Alꢅas, M. P. López, C. Cativiela, Tetrahedron 2004, 60, 885–
2
014. http://watoc2014.com/
[
76] L. J. Gutierrez, Search and molecular modeling study of the exo-
site of BACE1, PhD Thesis, National University of San Luis (Ar-
gentina), 2013.
8
91.
70] L. B. Kozell, K. A. Neve, Mol. Pharmacol. 1997, 52(6), 1137–
149.
71] M. M. Teeter, M. F. Froimowitz, B. Stec, C. J. Durand, J. Med.
Chem. 1994, 37, 2874–2888.
1
Received: July 2, 2014
Accepted: September 4, 2014
Published online: && &&, 0000
72] A. Mansour, F. Meng, J. H. Meador-Woodruff, L. P. Taylor, O. Civ-
elli, H. Akil, Eur J Pharmacol. 1992, 227, 205–214.
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
16&
ÞÞ
These are not the final page numbers!

Full Paper
www.molinf.com
FULL PAPERS
E. Angelina, S. Andujar,* L. Moreno,
F. Garibotto, J. Pꢁrraga, N. Peruchena,
N. Cabedo, M. Villecco, D. Cortes,
R. D. Enriz
&& – &&
3
-Chlorotyramine Acting as Ligand
of the D Dopamine Receptor.
2
Molecular Modeling, Synthesis and
D Receptor Affinity
2
ꢁ
0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Mol. Inf. 0000, 00, 1 – 17
&
17&
ÞÞ
These are not the final page numbers!