Computational strategies in discovering novel non-nucleoside inhibitors of HIV-1 RT

Article abstract of DOI:10.1021/jm049279a

A three-dimensional common feature pharmacophore model was developed using the X-ray structure of RT/non-nucleoside inhibitor (NNRTI) complexes. Starting from the pharmacophore hypothesis and the structure of the lead compound TBZ, new NNRTIs were designe

Full text of DOI:10.1021/jm049279a

J. Med. Chem. 2005, 48, 3433-3437
3433
Computational Strategies in Discovering Novel Non-nucleoside Inhibitors of
HIV-1 RT
Maria Letizia Barreca,*,‡ Angela Rao,*,‡ Laura De Luca,‡ Maria Zappala`,‡ Anna-Maria Monforte,‡
Giovanni Maga, Christophe Pannecouque,§ Jan Balzarini,§ Erik De Clercq,§ Alba Chimirri,‡ and
Pietro Monforte‡
Dipartimento Farmaco-Chimico, Universita` di Messina, Viale Annunziata, 98168 Messina, Italy, Istituto di Genetica
Molecolare IGM-CNR, via Abbiategrasso 207, 27100 Pavia, Italy, and Rega Institute for Medical Research, Katholieke
Universiteit Leuven, Minderbroedersstraat 10, B-3000 Leuven, Belgium
Received September 2, 2004
A three-dimensional common feature pharmacophore model was developed using the X-ray
structure of RT/non-nucleoside inhibitor (NNRTI) complexes. Starting from the pharmacophore
hypothesis and the structure of the lead compound TBZ, new NNRTIs were designed and
synthesized, having the benzimidazol-2-one system as a scaffold. Docking experiments showed
that these molecules docked in a position and orientation similar to that of known inhibitors.
Biological testing confirmed that our strategy was successful in searching for new leads as
NNRTIs.
Chart 1. The PDB Code of the Selected Ligand/
Enzyme Complexes, and the Names and the
2D-Structures of the Corresponding NNRTIs
Introduction
NNRTIs are an important class of anti-AIDS drugs
increasingly used in combination therapy together with
the previously approved nucleoside/nucleotide RTIs and
protease inhibitors.1 NNRTIs are a structurally diverse
group of compounds interacting with a specific allosteric
nonsubstrate binding site of HIV-1 RT, thus leading to
a noncompetitive inhibition of the enzyme. They cause
a conformational change that disrupts the catalytic site
and blocks DNA polymerase activity.2 After having
focused our researches on the identification of new
molecules with anti-RT activity,3,4 we have now used
different computational methods in attempts to gather
additional information for our rational design and to
discover new classes of NNRTIs.
We started on the X-ray crystal structures of different
ligand/enzyme complexes, showing both the bioactive
conformations of the inhibitors and their interactions
with the residues of the RT binding pocket. Such
structural data together with molecular modeling en-
hance the process of drug design. A hypothetical 3-D
pharmacophore model for NNRTIs was built by using
a combined ligand- and structure-based molecular mod-
eling approach. The postulated hypothesis allowed us
to design and synthesize a new potential class of
NNRTIs, and the antiviral activity results confirmed the
strength of our rational approach.
lographic data, into a common three-dimensional chemi-
cal feature-based pharmacophore model that could be
used to design new potentially active compounds.
Among all the available X-ray structures of wild-type
HIV-1 RT in complex with a second generation NNRTI,
we selected seven complexes based on the chemical
uniqueness of the corresponding inhibitor and the
resolution of the crystals.5-11
Moreover, since most NNRTIs are known to be
engaged in hydrogen bonding interactions with the
backbone of the amino acids Lys101 and/or Lys103, we
selected only those ligands that establish hydrogen
bonds with these RT residues. Thus, the training
set (TS) was composed of the following NNRTIs:
efavirenz, MKC442 (emivirine), HBY097, MSC204,
UC781, 739W94, and TMC120 (Chart 1).
Results and Discussion
Pharmacophore Modeling. Several cocrystals of
HIV-1 RT with structurally different nonnucleoside
inhibitors have been reported, showing that most ligands
bind in a similar “butterfly-like” conformation to a
hydrophobic pocket near the polymerase active site.
Our goal was to transfer the knowledge of binding
requirements for NNRTIs, as deduced from crystal-
* To whom correspondence should be addressed. Phone: (++39)
090-6766464. Fax:
(++39) 090-355613. E-mail:
barreca@
pharma.unime.it; a.rao@pharma.unime.it.
‡ Universita` di Messina.
Istituto di Genetica Molecolare IGM-CNR.
§ Rega Institute for Medical Research.
The common features hypothesis generation approach
(HipHop) implemented in the program Catalyst 9.012
10.1021/jm049279a CCC: $30.25 © 2005 American Chemical Society
Published on Web 04/13/2005

3434 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 9
Table 1. Summary of the Common Feature Hypothesis Run
Brief Articles
Chart 2. Structures of TBZ and Compounds 1-3
ranking
scorec
direct hit
maskd
partial hit
maske
numbera
compositionb
Hypo1
Hypo 2
Hypo 3
Hypo 4
Hypo 5
Hypo 6
Hypo 7
Hypo 8
Hypo 9
HHDA
HHDA
HHHA
HHHA
HHDA
HHHA
HHA
56.5018
51.8973
51.1017
50.3632
48.9869
46.7207
33.1821
30.3244
26.0421
1111111
1111111
1111111
1111111
1111111
1111111
1111111
1111111
1111111
0000000
0000000
0000000
0000000
0000000
0000000
0000000
0000000
0000000
regions, one hydrogen bond acceptor and one hydrogen
bond donor), one hydrophobic group, namely HY2 and
HY3 in Figure 1, was differently located. In particular,
in the selected NNRTIs, HY3 represents a portion, such
as a chlorine atom or a methoxy group, which points
into the hydrophobic pocket formed by the hydrocarbon
chains of Val106, Phe227, Leu234, and Pro236. On the
contrary, both aliphatic and aromatic moieties can
occupy the chemical feature HY2, making hydrophobic
contacts with the side chain of Leu100, Val106, Tyr318,
Leu234, and Pro236.
HHD
HDA
a Numbers for the hypothesis are consistent with the numera-
tion as obtained by the hypothesis generation. b H: hydrophobic
group; A: hydrogen bond acceptor. D: hydrogen bond donor. c The
higher the ranking score, the less likely it is that the molecules
in the TS fit the hypothesis by a chance correlation. Best
hypotheses have highest ranks. d Direct hit mask: a TS molecule
mapped every feature in the hypothesis. 1 means yes and 0 means
no. e Partial hit mask: a TS molecule mapped all but one feature
in the hypothesis. 1 means yes and 0 means no.
Since both Hypo 1 and Hypo 2 represented plausible
pharmacophore models for NNRTIs and considering
that Hypo 3 showed all the three hydrophobic features
(HY1-HY3), Hypo 1 and Hypo 2 were merged by using
0.8 Å tolerance, leading to a final model containing three
lipophilic parts satisfying the steric and lipophilic
requirements of the enzyme pocket.
Figure 1 shows the alignment of the 3D coordinates
of the bound ligand efavirenz (the most active NNRTI
clinically approved so far) onto the 3D merged five-
feature hypothesis: the HBA and HBD are mapped by
the carboxyamide group, while the three hydrophobic
sites HYs are mapped by the cyclopropyl group (HY1),
the benzene-fused ring (HY2), and the chlorine atom
(HY3) (see Supporting Information for mapping of the
seven NNRTIs into the merged pharmacophore model).
Recently a ligand-based pharmacophore model of
NNRTIs has been reported,13 which however presents
two hydrophobic features in a different 3D arrangement.
Rational Drug Design. In previous papers we
described a series of 1-aryl-1H,3H-thiazolo[3,4-a]benz-
imidazoles which proved to be highly active as HIV-1
NNRTIs. Extensive structure-activity relationship stud-
took into account the features of these seven compounds
and found the 3D arrangement of functional groups
common to all that interact with the RT enzyme, thus
carrying out a “qualitative model”. We considered this
the most appropriate approach because the “quantita-
tive” hypothesis generation method is not suitable
seeing that, owing to the different experimental proto-
cols employed, the available biological data are not
homogeneous.
Nine hypotheses, with the ranking scores ranging
from 56.50 to 26.04, were thus obtained as described in
the Supporting Information, and the results are pre-
sented in Table 1. As expected, our ligand-based hy-
pothesis was in agreement with the structure-based
information: in fact, the first (Hypo 1) and the second
(Hypo 2) ranked pharmacophore models contained one
hydrogen bond acceptor and one hydrogen bond donor
group that bind with the target making key interactions
(Table 1 and Figure 1). Moreover, this procedure
provided further information about the precise disposi-
tion of the hydrophobic groups. In fact, even if Hypo 1
and 2 had the same chemical features (two hydrophobic
Figure 1. Efavirenz mapped into Hypo 1 (top left), Hypo 2 (bottom left), and merged hypothesis of nonnucleoside inhibitors of
HIV-1 RT. The final pharmacophore model contains five features (HBD, hydrogen bond donor, purple; HBA, hydrogen bond acceptor,
green; HY1-3, hydrophobic regions, cyan).

Brief Articles
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 9 3435
RT, we also planned the synthesis of 5-chloro derivative
2. Finally, to prove the importance of the NH group as
hydrogen bond donor, we designed the isostere benzoxa-
zolone 3 which, lacking this functionality necessary to
interact with the carbonyl backbone of residues Lys101,
should have been less active than 1. The structures of
the designed derivatives 1-3 are reported in Chart 2.
Molecular Docking of Compounds 1-3. We sought
to validate our hypothesis by performing automated
docking studies of the designed compounds using Au-
todock 3.0.5.14 First of all, as in the docking calculations
the ligands are flexible while the protein coordinates
are fixed, test docking calculations using the seven
selected ligands were carried out to validate the docking
protocol and to confirm that the enzyme coordinates
from the efavirenz/RT complex could be used to repro-
duce the binding mode of any NNRTI. The ligands were
extracted from the corresponding RT complexes and
then docked back into the allosteric pocket of the crystal
structure 1FK9. The best Autodock-predicted conforma-
tion of all NNRTIs agreed well with the experimental
binding mode of these ligands, with root-mean square
deviations (RMSD) < 1 Å (see Supporting Information).
Later, docking of compounds 1-3 into the RT allos-
teric site generated a number of possible binding
conformations that correlated with energies. The cluster
analysis revealed a predominant ligand orientation
within the binding pocket (65, 40, and 21 conformations
in the first ranked cluster for 1, 2, and 3, respectively),
and the most energetically favorable conformation for
each ligand was chosen for further analysis.
The best docked conformation of compound 1 is shown
in Figure 3, together with the experimental position of
efavirenz and the 3D common-feature pharmacophore
model. Compound 1 interacted in a fashion similar to
known NNRTIs with RT residues of the allosteric
pocket. In fact, in agreement with the mapping of the
designed molecule into the merged hypothesis, the
carboxamide group of the molecule interacted with the
main-chain of Lys101 by hydrogen bond interactions,
while the hydrophobic features HY1-HY3 are occupied
by the p-chloroanilino and 2,6-difluorophenyl moieties.
Figure 2. 3D-common features pharmacophore model of
NNRTIs aligned to compound 1.
ies led to the discovery of TBZ (Chart 2) as lead com-
pound,4 whose binding mode into the lipophilic pocket
of the enzyme was demonstrated to be similar to other
NNRTIs. However, the fitting of TBZ into the obtained
pharmacophore model lacked the features necessary to
form hydrogen bond interaction with the Lys101 and/
or Lys103 of the enzyme. On this basis, we reasoned
that suitable chemical modifications of this molecule
might lead to new classes of NNRTI candidates. The
tricyclic system of TBZ was opened at the thiazole
nucleus level, removing the sulfur atom, which in vivo
is oxidized, giving less active or inactive metabolites.
The conversion of the 1H-benzimidazole nucleus into the
1,3-dihydro-2H-benzimidazol-2-one system could gener-
ate a class of molecules that not only fit equally well
into the allosteric site but also have the functional
groups able to act as hydrogen bond acceptor and donor.
Furthermore, a chlorine atom was introduced at position
6 of the benzimidazolone system, seeing that a p-chloro-
anilino moiety is a structural fragment present in many
potent NNRTIs such as efavirenz. As shown in Figure
2, the designed compound 1 was able to fulfill the
pharmacophore hypothesis. To verify whether the chlo-
rine position could influence the inhibitory effects on
The superimposition of the docked compound 2 and
the pharmacophore model demonstrated a different
Figure 3. AutoDock-predicted binding mode of 1 (left) and 2 (right) compared to the crystallized position of Efavirenz (yellow).
The colored spheres represented the 3D pharmacophore model developed by HipHop/Catalyst.

3436 Journal of Medicinal Chemistry, 2005, Vol. 48, No. 9
Brief Articles
Scheme 1a
to 1, as expected there was only one hydrogen bond
interaction between the carbonyl group of the ligand and
the backbone of the Lys101 residue.
On the basis of these molecular modeling results, we
designed other 1H,3H-thiazolo[3,4-a]benzimidazole de-
rivatives (i.e. 4-8) in which the 6-chlorine substitution
was maintained, whereas the 2,6-difluorophenyl moiety
was suitably modified or replaced to enhance the steric
hindrance or the lipophilicity.
Chemistry. The synthesis of compounds 1-8 was
realized according to the reaction sequence reported in
Scheme 1. The 2-nitroaniline (9 or 10) was N-alkylated
by treatment with the appropriate alkyl bromide in the
presence of potassium carbonate to give intermediates
11-17 which were reduced with Zn dust in acidic
medium. The cyclization of the amino derivatives 18-
24 with phosgene afforded target compounds 1, 2, 4-8.
Compound 3 was obtained by N-alkylation of the
commercially available 5-chloro-3H-benzoxazol-2-one
(25).
Anti-HIV Activity Assays. Compounds 1-8 were
tested in RT inhibition assays and proved to be active
as inhibitors of HIV-1 RT (Table 2). The enzymatic data
highlighted that derivative 1 turned out to be more
active than TBZ: targeting the enzyme for hydrogen
bond donor/acceptor interactions was thus a valid
strategy to increase RT inhibitory activity of the starting
compound. As shown by the inhibition profile of com-
pounds 1 and 2, the position of the halogen substituent
on the benzene ring influenced the affinity. In fact, the
presence of a chlorine substituent at C-5 led to a
decrease of potency, confirming that, as prompted by
the molecular docking experiments, the chlorine atom
at position 6 may lead to a binding conformation
resulting in improved hydrophobic contacts with the
surrounding residues in the binding site.
Compound 3 was 1 order of magnitude less active
than 1; this result was also consistent with our hypoth-
esis and clearly demonstrated the importance of the
hydrogen bond interaction of the NH group with the
carbonyl backbone of Lys101.
Considering the effect of the substituent at N-1, the
replacement of the 2,6-difluorophenyl moiety with other
aromatic or aliphatic groups negatively influences the
RT inhibitory activity. However, 2,6-dichlorophenyl (4),
4-i-propylphenyl (5), and naphthyl (6 and 7) derivatives
showed a good level of inhibitory activity whereas
cyclohexyl-substituted compound (8) is totally inactive.
This finding underlines the importance of the aromatic
moiety for this class of inhibitors.
a Reagents: (i) DMF, K2CO3, ∆, 2 h; (ii) Zn/HCl, EtOH, 80 °C,
1 h; (iii) 20% toluene solution of COCl2, 2 N HCl, ∆, 4 h.
Table 2. Anti-RT and Anti-HIV-1 Activities, Cytotoxicity, and
Selectivity Index in MT-4 Cells
compd
IC50 (µM)a
EC50 (µM)b
CC50 (µM)c
SId
1
2.86 ( 0.1
60 ( 4
0.24 ( 0.05
29.6 ( 2.1
3.72 ( 0.47
3.79 ( 1.68
2.83 ( 1.26
0.87 ( 0.16
6.35 ( 0.92
NA
>424
>1766
>14
>91
83
>147
57
>64
2
>424
3
20 ( 1.5
>382.2
4
7.50 ( 0.5
16.2 ( 0.8
12.7 ( 0.5
5.20 ( 0.5
50 ( 8
315 ( 25.9
>415
5
6
50.0 ( 2.43
>405
7
8
9.50 ( 1.72
50.0 ( 3.2
TBZ
24.57 ( 2
1.10 ( 0.32
45
a Concentration required to inhibit by 50% the in vitro RNA-
dependent DNA polymerase activity of recombinant RT. b Effective
concentration required to reduce HIV-1-induced cytopathic effect
by 50% in MT-4 cells. c Cytotoxic concentration required to reduce
MT-4 cell viability by 50%. d Selectivity index: ratio CC50/EC50
NA ) not active.
.
binding mode compared to 1 and less efficient hydro-
phobic/hydrogen-bond interactions with the surrounding
residues in the allosteric site (Figure 3). In particular,
it was surprising to note that this molecule did not map
the feature HY1, lacking any hydrophobic contact with
the region of the enzyme formed by Pro95, Leu100,
Tyr181, Tyr188, Trp229, and Leu234. Finally, even if
compound 3 docked in a position and orientation similar
Compounds 1-8 were also evaluated for inhibition
of the cytopathic effects of HIV-1 (IIIB) in MT-4 cells.
Compound-induced cytotoxicity was also measured in
MT-4 cells in parallel with the antiviral activity (Table
2). The results of this test provided the confirmation
that our rational design project worked since compound
1 was more potent than TBZ in preventing the cyto-
Table 3. Anti-HIV-1 Activity of 1 and TBZ against Mutant HIV-1 Strains in CEM Cells
EC50 (µM)a
compd
HIV-1IIIB
L100I
K103N
>20
15.6 ( 3.2
E138K
Y181C
>20
13.8 ( 4.2
Y188H
>20
4.16 ( 1.12
1
TBZ
0.88 ( 0.51
1.39 ( 0.56
2.0 ( 1.4
3.12 ( 0.94
2.9 ( 2.2
1.73 ( 1.02
a Effective concentration to protect CEM cells against the cytopathogenicity of HIV-induced by 50%.

Brief Articles
Journal of Medicinal Chemistry, 2005, Vol. 48, No. 9 3437
dro-2H-benzimidazol-2-one (4) from 20 (75 mg, 0.25 mmol);
6-Chloro-1-(4-isopropylbenzyl)-1,3-dihydro-2H-benzimi-
dazol-2-one (5) from 21 (69 mg, 0.25 mmol); 6-Chloro-1-(1-
naphthyl)methyl-1,3-dihydro-2H-benzimidazol-2-one (6)
from 22 (70 mg, 0.25 mmol); 6-Chloro-1-(2-naphthyl)-
methyl-1,3-dihydro-2H-benzimidazol-2-one (7) from 23 (70
mg, 0.25 mmol); 6-Chloro-1-(cyclohexylmethyl)-1,3-dihy-
dro-2H-benzimidazol-2-one (8) from 24 (60 mg, 0.25 mmol).
5-Chloro-3-(2,6-difluorobenzyl)-3H-benzoxazol-2-one (3).
Compound 3 was synthesized with a procedure similar to that
reported for 11, starting from chlorzoxazone (25) (170 mg, 1
mmol) and 2,6-difluorobenzyl bromide (310 mg, 1.5 mmol).
pathic effect of HIV-1 and was minimally toxic to MT-4
cells, resulting in a remarkably high selectivity index.
Compound 1 was also tested in CEM cell cultures
against an extensive panel of mutant virus strains
containing in their RT a single mutation that is char-
acteristic of the HIV-NNRTI resistance profile. The
compound generally kept activity against L100I and
E138K RT HIV-1, but lost antiviral activity against
K103N, Y181C, and Y188H RT HIV-1 strains (Table 3).
Conclusions. A combined ligand- and structure-
based molecular modeling approach led to the design
of potential NNRTIs structurally related to our previ-
ously reported 1H,3H-thiazolo[3,4-a]benzimidazoles. A
simple synthetic route of novel 1,3-dihydro-2H-benz-
imidazol-2-ones has been developed. Biological results
showed compound 1 as a starting point for lead opti-
mization strategy, which is already ongoing.
Acknowledgment. Financial support for this re-
search by Fondo Ateneo di Ricerca (2002, Messina,
Italy), MIUR (COFIN2002, Roma, Italy), and the Eu-
ropean Commission (QLK2-CT-2000-00 291) is grate-
fully acknowledged.
Supporting Information Available: Additional experi-
mental data are available free of charge via Internet at http://
pubs.acs.org.
Experimental Section
Chemistry: 5-Chloro-1-(2,6-difluorobenzyl)-2-nitro-
aniline (11). To a stirred solution of 5-chloro-2-nitroaniline
(10) (346 mg, 2 mmol) in DMF (10 mL) were added 2,6-
difluorobenzyl bromide (621 mg, 3 mmol) and anhydrous
potassium carbonate (1.382 mg, 10 mmol). The reaction
mixture was refluxed for 2 h, cooled, filtered and, after addition
of water (60 mL), extracted with chloroform (2 × 50 mL). After
removal of the solvent under reduced pressure, the residue
was powdered by treatment with diethyl ether and recrystal-
lized from ethanol. 4-Chloro-1-(2,6-difluorobenzyl)-2-ni-
troaniline (12). With a similar procedure, 12 was prepared
from 4-chloro-2-nitroaniline (9) (346 mg, 2 mmol) and 2,6-
difluorobenzyl bromide (621 mg, 3 mmol). With a similar
procedure and starting from 10, we have prepared: 5-chloro-
1-(2,6-dichlorobenzyl)-2-nitroaniline (13) with 2,6-dichlo-
robenzyl bromide (720 mg, 3 mmol); 5-chloro-1-(4-isopro-
pylbenzyl)-2-nitroaniline (14) with 4-i-propylbenzyl bromide
(640 mg, 3 mmol); 5-chloro-1-(1-naphthyl)methyl-2-nitro-
aniline (15)with (1-naphthyl)methyl bromide (663 mg, 3
mmol); 5-chloro-1-(2-naphthyl)methyl-2-nitroaniline (16)
with (2-naphthyl)methyl bromide (663 mg, 3 mmol); 5-chloro-
1-(cyclohexylmethyl)-2-nitroaniline (17) with cyclohexyl-
methyl bromide (531 mg, 3 mmol).
References
(1) De Clercq, E. Antiviral drugs in current clinical use. J. Clin.
Virol. 2004, 30, 115-133.
(2) Balzarini, J. Current status of the nonnucleoside reverse tran-
scriptase inhibitors of human immunodeficiency virus type 1.
Curr. Top Med. Chem. 2004, 4, 921-944.
(3) Barreca, M. L.; Balzarini, J.; Chimirri, A.; De Clercq, E.; De Luca,
L.; et al. Design, synthesis, structure-activity relationships, and
molecular modeling studies of 2,3-diaryl-1,3-thiazolidin-4-ones
as potent anti-HIV agents. J. Med. Chem. 2002, 45, 5410-5413.
(4) Chimirri, A.; Grasso, S.; Molica, C.; Monforte, A. M.; Monforte,
P.; et al. Structural features and anti-human immunodeficiency
virus (HIV) activity of the isomers of 1-(2′,6′-difluorophenyl)-
1H,3H-thiazolo[3,4-a]benzimidazole, a potent nonnucleoside HIV-1
reverse transcriptase inhibitor. Antiviral Chem. Chemother.
1997, 8, 363-370.
(5) Ren, J.; Milton, J.; Weaver, K. L.; Short, S. A.; Stuart, D. I.; et
al. Structural basis for the resilience of efavirenz (DMP-266) to
drug resistance mutations in HIV-1 reverse transcriptase.
Structure Fold Des. 2000, 8, 1089-1094.
(6) Hopkins, A. L.; Ren, J.; Esnouf, R. M.; Willcox, B. E.; Jones, E.
Y.; et al. Complexes of HIV-1 reverse transcriptase with inhibi-
tors of the HEPT series reveal conformational changes relevant
to the design of potent nonnucleoside inhibitors. J. Med. Chem.
1996, 39, 1589-1600.
2-Amino-5-chloro-1-(2,6-difluorobenzyl)aniline (18). The
mixture of 11 (179 mg, 0.6 mmol) in 3 mL of HCl and 4 mL of
EtOH was stirred vigorously, then zinc dust (1.32 g, 20 mmol)
was added in several portions at room temperature. After this
addition was completed, the reaction mixture was heated on
a water bath for 1 h, cooled, made alkaline with 2 N NaOH,
and then extracted with ethyl acetate. The extracted was
washed with water, dried over Na2SO4, and evaporated. The
residue was crystallized from ethanol. With a similar proce-
dure, we have prepared: 2-amino-4-chloro-1-(2,6-difluo-
robenzyl)-aniline (19) from 12 (149 mg, 0.5 mmol); 2-amino-
5-chloro-1-(2,6-dichlorobenzyl)-aniline (20) from 13 (166
mg, 0.5 mmol); 2-amino-5-chloro-1-(4-isopropylbenzyl)-
aniline (21) from 14 (145 mg, 0.5 mmol); 2-amino-5-chloro-
1-(1-naphthyl)methyl-aniline (22) from 15 (155 mg, 0.5
mmol); 2-amino-5-chloro-1-(2-naphthyl)methyl-aniline (23)
from 16 (155 mg, 0.5 mmol); 2-amino-5-chloro-1-(cyclo-
hexylmethyl)-aniline (24) from 17 (135 mg, 0.5 mmol).
6-Chloro-1-(2,6-difluorobenzyl)-1,3-dihydro-2H-benz-
imidazol-2-one (1). To a solution of 18 (67 mg, 0.25 mmol)
in 2 N HCl (4 mL) was added an excess of a 20% toluene
solution of phosgene (1 mL), and the resulting mixture was
heated for 4 h. After cooling, the reaction mixture was
neutralized with 2 N NaOH, extracted with ethyl acetate,
washed with water, and evaporated under reduced pressure.
The residue was crystallized from ethyl acetate. With a similar
procedure, we have prepared: 5-Chloro-1-(2,6-difluoroben-
zyl)-1,3-dihydro-2H-benzimidazol-2-one (2) from 19 (67
mg, 0.25 mmol); 6-Chloro-1-(2,6-dichlorobenzyl)-1,3-dihy-
(7) Hsiou, Y.; Das, K.; Ding, J.; Clark, A. D., Jr.; Kleim, J. P.; et al.
Structures of Tyr188Leu mutant and wild-type HIV-1 reverse
transcriptase complexed with the nonnucleoside inhibitor HBY
097: inhibitor flexibility is a useful design feature for reducing
drug resistance. J. Mol. Biol. 1998, 284, 313-323.
(8) Hogberg, M.; Sahlberg, C.; Engelhardt, P.; Noreen, R.; Kangas-
metsa, J.; et al. Urea-PETT compounds as a new class of HIV-1
reverse transcriptase inhibitors. 3. Synthesis and further struc-
ture-activity relationship studies of PETT analogues. J. Med.
Chem. 2000, 43, 304.
(9) Ren, J.; Esnouf, R. M.; Hopkins, A. L.; Warren, J.; Balzarini, J.;
et al. Crystal structures of HIV-1 reverse transcriptase in
complex with carboxanilide derivatives. Biochemistry 1998, 37,
14394-14403.
(10) Chan, J. H.; Hong, J. S.; Hunter, R. N., 3rd; Orr, G. F.; Cowan,
J. R.; et al. 2-Amino-6-arylsulfonylbenzonitriles as nonnucleoside
reverse transcriptase inhibitors of HIV-1. J. Med. Chem. 2001,
44, 1866-1882.
(11) Das, K.; Clark, A. D., Jr.; Lewi, P. J.; Heeres, J.; De Jonge, M.
R.; et al. Roles of conformational and positional adaptability in
structure-based design of TMC125-R165335 (etravirine) and
related nonnucleoside reverse transcriptase inhibitors that are
highly potent and effective against wild-type and drug-resistant
HIV-1 variants. J. Med. Chem. 2004, 47, 2550-2560.
(12) Catalyst 4.9, Accelrys Inc., San Diego, http://www.accelrys.com.
(13) Patel, Y.; Gillet, V. J.; Bravi, G.; Leach, A. R. A comparison of
the pharmacophore identification programs: Catalyst, DISCO
and GASP. J. Comput.-Aided Mol. Des. 2002, 16, 653-681.
(14) Morris, G. M.; Goodsell, D. S.; Halliday, R. S.; Huey, R.; Hart,
W. E.; et al. Automated Docking Using a Lamarckian Genetic
Algorithm and Empirical Binding Free Energy Function. J.
Comput. Chem. 1998, 19, 1639-1662.
JM049279A