Structure-activity relationship studies of novel benzophenones leading to the discovery of a potent, next generation HIV nonnucleoside reverse transcriptase inhibitor

Article abstract of DOI:10.1021/jm050670l

Despite the progress of the past two decades, there is still considerable need for safe, efficacious drugs that target human immunodeficiency virus (HIV). This is particularly true for the growing number of patients infected with virus resistant to currently approved HIV drugs. Our high throughput screening effort identified a benzophenone template as a potential nonnucleoside reverse transcriptase inhibitor (NNRTI). This manuscript describes our extensive exploration of the benzophenone structure-activity relationships, which culminated in the identification of several compounds with very potent inhibition of both wild type and clinically relevant NNRTI-resistant mutant strains of HIV. These potent inhibitors include 70h (GW678248), which has in vitro antiviral assay IC₅₀ values of 0.5 nM against wild-type HIV, 1 nM against the K103N mutant associated with clinical resistance to efavirenz, and 0.7 nM against the Y181C mutant associated with clinical resistance to nevirapine. Compound 70h has also demonstrated relatively low clearance in intravenous pharmacokinetic studies in three species, and it is the active component of a drug candidate which has progressed to phase 2 clinical studies.

Full text of DOI:10.1021/jm050670l

J. Med. Chem. 2006, 49, 727-739
727
Structure-Activity Relationship Studies of Novel Benzophenones Leading to the Discovery of a
Potent, Next Generation HIV Nonnucleoside Reverse Transcriptase Inhibitor
Karen R. Romines,†,* George A. Freeman,† Lee T. Schaller,† Jill R. Cowan,† Steve S. Gonzales,† Jeffrey H. Tidwell,†
Clarence W. Andrews, III,† David K. Stammers,‡ Richard J. Hazen,† Robert G. Ferris,† Steven A. Short,† Joseph H. Chan,† and
Lawrence R. Boone†
GlaxoSmithKline Inc., Research Triangle Park, North Carolina 27709, and UniVersity of Oxford, Oxford, UK
ReceiVed July 14, 2005
Despite the progress of the past two decades, there is still considerable need for safe, efficacious drugs that
target human immunodeficiency virus (HIV). This is particularly true for the growing number of patients
infected with virus resistant to currently approved HIV drugs. Our high throughput screening effort identified
a benzophenone template as a potential nonnucleoside reverse transcriptase inhibitor (NNRTI). This manuscript
describes our extensive exploration of the benzophenone structure-activity relationships, which culminated
in the identification of several compounds with very potent inhibition of both wild type and clinically relevant
NNRTI-resistant mutant strains of HIV. These potent inhibitors include 70h (GW678248), which has in
vitro antiviral assay IC50 values of 0.5 nM against wild-type HIV, 1 nM against the K103N mutant associated
with clinical resistance to efavirenz, and 0.7 nM against the Y181C mutant associated with clinical resistance
to nevirapine. Compound 70h has also demonstrated relatively low clearance in intravenous pharmacokinetic
studies in three species, and it is the active component of a drug candidate which has progressed to phase
2 clinical studies.
Introduction
The use of combination therapy, or highly active antiretroviral
therapy (HAART), for treatment of human immunodeficiency
virus (HIV) infection has dramatically impacted the disease.
Diligent use of available anti-retroviral drugs can control an
HIV infection for several years, and as a consequence, the death
rate of HIV patients in the developed world fell dramatically
in the mid to late 1990s. Nonnucleoside reverse transcriptase
inhibitors (NNRTI’s), including nevirapine and efavirenz, are
proving to be important components of HAART. In particular,
the use of efavirenz in combination with two nucleosides (e.g.,
lamivudine and zidovudine) is an excellent first-line therapy
for HIV patients, and it has been used extensively in recent
years.1
As with all antiretroviral drugs developed to date, the use of
the NNRTI’s has led to the appearance of clinical resistance to
this class of drugs. Unfortunately, the genetic barrier of all
current NNRTI’s is relatively low, and a single mutation can
Figure 1. Initial benzophenone NNRTI lead structures.
begin to reduce the susceptibility of virus to drug. Furthermore,
cross resistance between the approved NNRTI’s (efavirenz,
regimens of HIV care are complex, it’s clear that each of these
NNRTI’s is associated with a single point mutation that triggers
the development of clinical resistance. For efavirenz, it is the
K103N mutation. For nevirapine, it is the Y181C mutation.2
Consequently, these mutant viruses have been a focal point in
our search for a next generation NNRTI.
Our efforts to discover an NNRTI efficacious against resistant
viral strains originated in a high throughput screening effort
which identified the benzophenone 1 (Figure 1).6 The initial
optimization work for this template was published in 1995. The
finding that resistant viruses readily generated in the presence
of the leading analogue 2 were cross-resistant to nevaripine led
to decreased interest in further optimization of the benzophe-
nones at that time, and work in this compound series was briefly
stopped. Because of the importance of this drug class, however,
we turned our attention once more to the benzophenones in the
late 1990s, as one of several potential templates for targeting
nevirapine, and delavirdine) is quite common, and patients are
generally forced to abandon the approved NNRTI’s altogether
once they have developed resistance to one of its members.2,3
To make the situation even worse, the appearance of drug-
resistant virus is not limited to NNRTI-experienced patients.
Recent studies indicate the number of newly infected patients
with NNRTI-resistant virus is increasing.4 In one U.S. cohort
which included antiretroviral na¨ıve patients from 44 cities, 18%
had decreased phenotypic sensitivity to at least one NNRTI,
and one-third of these had decreased phenotypic sensitivity to
all currently used NNRTI’s.5
Efavirenz and nevirapine are the most widely used NNRTI’s.
Although resistance patterns in the presence of the multidrug
* Corresponding author. Phone: 919 483 4749. Fax: 919 315 2141.
E-mail: Karen.r.romines.@gsk.com.
† GlaxoSmithKline Inc.
‡ University of Oxford.
10.1021/jm050670l CCC: $33.50 © 2006 American Chemical Society
Published on Web 12/24/2005

728 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Romines et al.
Scheme 1. Synthesis of Conformationally Constrained C-Ring Analoguesa
a (a) 5 to 7: nBuLi, Et2O, -78 °C; then 3,5-diFPhCON(OMe)Me (6), -78 °C to room temperature, 42%; (b) 7 to 8: BBr3, CH2Cl2, -78 °C to room
temperature, 98%; (c) 8 to 9: BrCH2CO2Et, K2CO3, acetone, reflux, 4 h, 100%; (d) 9 to 10: LiOH, H2O, EtOH, THF, rt, 3h, 81%; (e) 10 to 11: (ClCO)2,
DMF, CH2Cl2, rt, 1 h; (f) amine, NaHCO3, acetone, H2O, rt, 2.5 h, 37-58%. IC50 values were determined using an acute infection assay in MT4 cells. Wild
type (wt) virus is the HIV-1 IIIB strain. The nevirapine resistant virus (nevR) contains the Y181C mutation.
resistant viruses. Guided by data from a panel of mutant viruses,
we were able to transform this template into a series of
compounds with potency against clinically relevant NNRTI-
resistant viruses.7 This manuscript will describe the structure-
activity relationships leading to the initial activity against mutant
viruses, as well as the additional modifications which led to a
dramatic increase in potency. We will also outline the data on
drug development characteristics which differentiated very
potent inhibitors, leading to the identification of 70h (GW678248,
see Table 4) as a promising next generation NNRTI.
template was prepared by linking the ortho position of the C-ring
to the amide of the linker region. Synthesis of this analogue,
which is illustrated in Scheme 1, was accomplished by first
linking the A- and B-rings, and then installing the linker region
and C-ring sequentially. Metal halogen exchange of a bromi-
nated phenol derivative (5), followed by quenching with a
Weinreb amide (6), allowed ready access to the A- and B-rings
of the benzophenone template (7). Deprotection to reveal the
phenol 8 then provided a handle for building the linker region
via reaction with o-bromoethyl acetate to give 9. Finally, the
C-ring was added to the molecule via amide formation between
the acid chloride 11, derived in two steps from 9, and a
commercially available amine. The o-methyl analogue 12 was
also prepared to allow direct comparison in the antiviral assays,9
and we found that the constrained analogue was less potent
against both wild type and nevirapine-resistant viruses. Although
the differences were not dramatic, there appeared to be some
advantage to using the nonconstrained amides, such as 12. This
result is consistent with the presence of an internal hydrogen
bond involving the amide NH in the enzyme-bound conforma-
tion of 12, which is precluded in 4.
A second modification of the linker region involved formation
of the hydroxylamine derivative 13 (Figure 2). This analogue
was prepared in an approach similar to that outlined in Scheme
1; an unsubstituted A-ring analogue of the acid chloride 11 was
mixed with 2-phenylhydroxylamine in acetonitrile to give 13.
This structural modification, however, did not prove to be
beneficial. The potency of 13 was significantly less than the
corresponding des-hydroxy analogue 3.
Finally, we explored the impact of reversing the amide in
the linker region. The analogue of 3 with the amide inverted
was not chemically stable, but we were able to prepare a version
with the oxygen replaced by a carbon (17) via the route
illustrated in Scheme 2. This is a different synthetic approach
than the one described above. In this case, the A-ring is installed
last. Acylation of the primary amine 14 with benzoyl chloride
was followed by a dehydrative ring closure to give intermediate
16. Ring opening, followed by a second reaction with benzoyl
chloride, produced 17. The results of an antiviral assay with 17
indicate the amide configuration is important, since compound
17 was far less potent against the wild-type virus than 3.
Results and Discussion
The original optimization of 1 indicated that the carbonyl
between the A- and B-rings was important for activity, and a
methyl substituent ortho to the amide on the C-ring gave the
molecule a measure of metabolic stability. On the C-ring, further
substitution was best placed at the para position, but no
beneficial substitution of the A-ring was identified.6
A crystal structure of the unsubstituted analogue 3 complexed
with wild-type HIV reverse transcriptase was obtained by
Stammers and colleagues,8 and this work confirmed that the
benzophenone analogue occupied the nonnucleoside binding
pocket. The X-ray crystal structure also indicated that space
could be available for meta substitution in the A-ring and para
substitution in the C-ring. Study of the structure led to the
following hypotheses: (1) Meta substitution on the A-ring would
reach into the hydrophobic region adjacent to the side chains
of Y181 and Y188 and thus would require a polar aprotic and
sterically small substituent, such as cyano (or a small nonpolar
group at the expense of solubility). (2) The para position on
the C-ring pointed directly into the exterior water through a
rather small window in the protein surface. Para substitution
could thus be used to add a water solubilizing group that could
counter the lipophilic nature of the benzophenone scaffold. (3)
The internal hydrogen bond between the benzophenone carbonyl
and the amide NH could be maintained in the benzophenone:
enzyme complex.
Modifications of the Linker Region and B-Ring. Some of
our early modifications of the benzophenone template centered
on the central portion of the molecule, the linker region and
the B-ring. A conformationally constrained version (4) of the

HIV Nonnucleoside ReVerse Transcriptase Inhibitor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 729
Scheme 3. Synthesis of B-Ring Thiophene Analoguesa
Scheme 2. Synthesis of the Reverse Amide Linker Analoguea
a (a) PhCOCl, pyridine; (b) POCl3, P2O5, xylene, 54%; (c) PhCOCl,
triethylamine, toluene, 1 drop H2O, 35%. IC50 values determined using an
acute infection assay in MT4 cells. Wild-type virus (wt) is the HIV-1 IIIb
strain.
a (a) BrCH2COBr, pyr, CH2Cl2, 69-100%; (b) ArCOCl, AlCl3, CH2Cl2,
reflux, overnight, 22-40%; (c) 21, K2CO3, acetone, reflux, 5-6 h, 22-
77%. IC50 values were determined using HIV-1 IIIB wild-type virus in an
acute infection assay in MT4 cells.
A variation of the synthetic approach of Scheme 1 allowed us
to replace the aryl B-ring with a thiophene (Scheme 3). Alum-
inum chloride-mediated coupling of the methoxy thiophene 22
and the acid chloride derivative of the A-ring gave the A/B in-
termediates 23a,b. In this case, the C-ring amide linkage was
formed before installation of the A/B-rings by reaction of the
arylamines 20a,b with bromoacetyl bromide. Subsequent alkyla-
tion of 23a,b with 21a,b produced the desired analogues in a
slightly more convergent route than the one shown above. Three
versions of the thiophene B-ring analogues were prepared (24a-
c). Again, although these compounds were active against wild-
type HIV, their potency was low relative to analogues with the
chlorophenyl B-ring.
Figure 2. N-Hydroxylated benzophenone analogue. The IC50 values
were determined using an acute infection assay in MT4 cells. Wild-
type virus (wt) is the HIV-1 IIIB strain. Neviripine resistant virus (nev-
R) contains the Y181C mutation.
SAR of the C-Ring. Since the above experimental work
indicated the B-ring and linker regions of the benzophenone
template were well optimized, we turned our attention to
modifications of the C-ring. As noted above, previous work in
this template had established the importance of an o-methyl
group and had indicated that additional substitution at the para
position could be tolerated. We chose to further explore the
potential for para substitution using a synthetic approach
analogous to that described in Scheme 1.
Varying the C-ring substitution required preparation of several
aniline derivatives, as illustrated in Schemes 4 and 5. Scheme
4 demonstrates the flexibility of starting with various para-
substituted 2-nitrotoluenes. The fluoro derivative 25 readily
yielded C-rings with secondary amines at the para position (26)
when mixed with primary amines in mild base. The para amide
derivative 30 was generated from the aniline 28. In this case,
the amine of the starting material was used to form an R,â-
unsaturated amide (29), which then gave 30 via 1,4-addition of
morpholine. Ether analogues were readily obtained either by
reaction of alcohols with the fluoro derivative 25 or from the
phenol starting material 32. In all cases, the desired C-ring
aniline was obtained by reduction of the aryl nitro group to an
amine in the final step. This routinely proceeded cleanly and in
good yields.
Figure 3. Analogues with trifluoromethyl substitution on the B-ring.
The IC50 values were determined using an acute infection assay in MT4
cells. Wild-type virus (wt) is the HIV-1 IIIB strain. Neviripine resistant
virus (nev-R) contains the Y181C mutation.
One question we had about the B-ring in the benzophenone
template was the role of the p-chloro substituent. Initiating the
route outlined in Scheme 1 with 2-bromo-1-methoxy-4-(tri-
fluromethyl)benzene,11 an analogue of compound 5, allowed
us to explore the impact of replacing the p-chloro group with a
trifluoromethyl group. The two analogues shown in Figure 3
illustrate that although the trifluoro-based compounds 18a and
19 are active against both wild type and nevirapine-resistant
viral strains, they are not as potent as the corresponding chloro-
substituted compound 18b.

730 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Romines et al.
Scheme 4. Synthesis of Aniline Intermediates for C-Ring Analoguesa
a (a) 1-(3-aminopropyl)-imidazole, NaHCO3, pyridine, 49%; (b) 26a to 27a and 30 to 31: H2, 10% Pd/C,MeOH, 80-100%; (c) ClC(O)CHdCH2, Et3N,
CH2Cl2, 0 °C; (d) 29 to 30: morpholine, EtOH, reflux, 80% for 2 steps (29 to 31); (e) 3-bromopropyl phthalimide, Cs2CO3, DMF, reflux; (f) H2NNH2-H2O,
EtOH, reflux, 85% for two steps (32 to 33); (g) 33 to 34: Me3SiNdCdO, THF, rt, 75%; (h) 34 to 35: H2, 10% Pd/C, EtOH, rt, 92%.
Scheme 5. Synthesis of the p-Sulfonamide Aniline C-Ring
(K103N and Y181C). Analogues with substitution of the
sulfonamide (42e,f) were also active in the antiviral assays, but
were not quite as potent as 42d. In general, substitution at the
para position was quite well tolerated. Most of the compounds
did not show dramatic differences in potency, even in cases
with fairly significant changes (e.g., 42a vs 42d). This is con-
sistent with the hypothesis that the para substituent reaches into
a solvent pocket when bound to the HIV reverse transcriptase.
Intermediatea
In addition to changing the substituent on the C-ring, we
explored the use of pyridine rings in this position. Synthesis of
the necessary anilines to form the C-rings was carried out as
shown in Scheme 6. For the analogues in which the pyridine
nitrogen was ortho to the sulfonamide, a chloro-substituted
pyridine (e.g., 43)12 was converted to the corresponding thiol
44, which could be oxidized to the desired sulfonamide 45.
Reduction of the aryl nitro group then provided the C-ring
aniline 46. Synthesis of the C-ring with the pyridine nitrogen
ortho to the amine required a different approach. In this case,
the pyridine 47 was sulfonylated to give an intermediate sulfonic
acid, the aniline group of which was protected to give 48. The
sulfonic acid was then converted to the desired sulfonamide 49
in a single step.
a (a) 36 to 37: BnBr, Na2CO3, H2O, CH2Cl2, reflux; (b) 37 to 38: thionyl
chloride, DMF, 0 °C, 24% for 2 steps; (c) 38 to 39: Me2NH, EtOH, 0 °C,
20%; (d) H2, 10% Pd/C, toluene, rt, 67%.
Scheme 5 exemplifies the synthesis of the p-sulfonamide-
substituted C-rings from the sulfonic acid 36. The protecting
group strategy varies according to the desired substitution of
the final product. In the case of the dimethylsulfonamide 40,
the aniline is protected with bis-benzyl groups, and the sulfonic
acid is isolated as its sodium salt. Conversion to the sulfonyl
chloride follows, and this intermediate reacts readily with
dimethylamine to produce 39. Deprotection of 39 then yields
the aniline for the coupling reaction. The monomethyl aniline
required for 42e was prepared using the same strategy, but with
an acyl protecting group.
Compounds 41a-d were designed to explore the conversion
of the ether linkage in 1 to secondary amines or amides. As
illustrated in Table 1, there was little benefit to these changes
relative to the unsubstituted analogue 12. The exception is 41d,
which is potent against both the wild type and K103N mutant
viruses. Replacement of the terminal amine in 1, however, with
other groups proved to be more consistently beneficial. The
sulfonamide (42a), urea (42b), and ether (42c) substituents were
all well tolerated, and these compounds were quite potent in
the antiviral assays. The best results were achieved when the
para position was substituted with a sulfonamide (42d). This
analogue had excellent potency against both wild-type virus and
the two key mutant viruses associated with NNRTI resistance
The synthesis of the A/B-rings of the C-ring pyridine
analogues is a variation on the Scheme 1 approach (Scheme
7). In this case, the Weinreb amide derivative was made of the
B-ring (52) for reaction with the lithiated A-ring. The A-rings
were derived from bromo aryl derivatives. The bromo groups
provided the access point for all three of the A-ring substituents.
We started with 1,3,5-tribromobenzene (50) and carried out a
pair of metal-halogen exchange reactions, quenching the first
with hexachloroethane and the second with Weinreb amide 52,
to give 53a. The final bromo group acted as a cyano surrogate,
and it was converted using a palladium-mediated reaction with
sodium cyanide to provide 54a, which was readily deprotected
to 55a and then converted to the analogues shown in Table 2.
In addition to the C-ring sulfonamide intermediates prepared
in Scheme 6, two commercially available pyridine-rings were
used to prepare 56a,b. These analogues provided further
confirmation of the above observations that the p-sulfonamide
group is quite beneficial (Table 2). In the series of sulfonamide

HIV Nonnucleoside ReVerse Transcriptase Inhibitor
Table 1. SAR of Para Substituents on the C-Ringa
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 731
Scheme 6. Synthesis of Pyridine C-Ring Intermediatesa
a (a) thiourea, EtOH and then KOH, 30%; (b) 44 to 45: 1N HCl, Cl2
and then NH3/ CH2Cl2, 40%; (c) 45 to 46: H2, 10% Pd/C, EtOH; (d) H2SO4/
SO3, 160 °C, 50%; (e) SOCl2, DMF, rt, 77%; f) PCl5 and then NH4OH, rt,
45%.
Scheme 7. Synthesis of A/B-Ring Intermediatesa
a Synthesis: 41a, 41b, 41c, and 42c prepared from 25; 41d prepared
from 28; 42b prepared from 32; 42e and 42f prepared from 36. b IC50 values
were determined using HIV-1 IIIB wild-type virus in an acute infection
assay in MT4 cells. c IC50 values were determined using a HeLa MAGI
assay system. The wild-type virus was the HxB2 strain, and the K103N
and Y181C mutants were isogenic with the HxB2 virus backbone.
derivatives (57-59), placement of the pyridine nitrogen at the
X3 position (59) attenuated potency against both wild type and
K103N mutant viruses. Placement of the pyridine nitrogen at
either the X1 (57a,b) or X2 (58a,b) positions, however, gave
analogues with excellent potency against both wild type and
mutant viruses.
a (a) 50 to 51a: n-BuLi, Et2O, -78 °C; then, Cl3CCCl3, -78 °C to room
temperature, 92%; (b) 51a to 53a: n-BuLi, Et2O, -78 °C; then, 5-chloro-
N,2-dimethoxy-N-methyl benzamide (52), -78 °C to room temperature,
97%; (c) 53a to 54a: CuI, NaCN, Pd(PPh3)4, MeCN, reflux, 56%; (d) 54a
to 55a: BBr3, CH2Cl2, -78 °C to room temperature, 100%.
nitro group produced the desired aniline 64. In this case, the A/B-
ring was not converted to an acid chloride. Instead, HOBT-medi-
ated coupling was used to partner the acid 63a with the aniline
64 to give 65a. As was the case in Scheme 7, the bromo substi-
tuent on the A-ring was converted to cyano group in the final
step. In the case of the m-cyano analogue 66b, the cyano-sub-
stituted acid chloride A-ring starting material was commercially
available and the cyano group was carried through the synthesis.
In Table 3, we show representative results of the impact of
moving the A-ring substituent. Clearly, placement of a small
group at the meta position (66b) is far more favorable than
substitution of the same group at the ortho (66a) or para (66b)
position. Furthermore, substitution at the meta position offers
an advantage over the unsubstituted analogue 67, particularly
against the critical mutant viruses, K103N and Y181C. In fact,
subsequent crystallography demonstrated that tyrosine residues
181 and 188 tolerated meta substitution.8
SAR of the A-Ring. After establishing the flexibility of
C-ring substitution, we turned to the A-ring for further optimiza-
tion work. In this case, relatively small changes in structure
did prove to have a significant impact on the potency of the
analogues.
Our first step was to explore the impact of moving a small sub-
stituent around the A-ring. In this case, the A/B-rings were pre-
pared as illustrated in Scheme 8. The A-ring source was a sub-
stituted benzoyl chloride (e.g., 60a), which readily underwent
Friedel-Crafts acylation with 4-chloroanisole to produce the A/B-
ring structure 61a. The aluminum salts which catalyzed the
Friedel-Crafts reaction also removed the methoxy group, con-
veniently exposing the phenol for the subsequent alkylation step.
The C-ring for this analogue set was prepared in a fashion ana-
logous to that shown in Scheme 4. The p-fluoro group of the ni-
trotoluene 25 was displaced with thiomorpholine. Subsequent
oxidation of the thioether to a sulfoxide and reduction of the aryl

732 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Table 2. SAR of C-Ring Pyridyl Analogues
Romines et al.
potency (IC50, nM)a
compound
R1
R2
R3
R4
X1
X2
X3
wt
K103N
Y181C
V106A
56a
56b
57a
57b
58a
58b
59
F
F
Me
Cl
Me
Cl
F
F
H
H
Cl
OMe
CH
CH
N
N
N
CH
CH
N
CH
CH
CH
CH
CH
CH
N
- -
- -
2
1
1
67
72
6
3
7
- -
- -
4
3
4
- -
- -
3
7
2
CN
CN
CN
CN
CN
Me
Me
Me
Me
Me
SO2NH2
SO2NH2
SO2NH2
SO2NH2
SO2NH2
N
CH
CH
CH
N
CH
2
18
5
93
2
- -
10
- -
Me
a IC50 values were determined using a HeLa MAGI assay. Wild-type virus (wt) was the HxB2 strain. K103N, Y181C, and V106A mutants were isogenic
with the HxB2 virus backbone.
Scheme 8. Synthesis of A-Ring Monosubstituted Analoguesa
Scheme 9. Synthesis of Disubstituted A/B-Ring Intermediatesa
a (a) 2.2 equiv n-BuLi, THF, -78 °C; then, N-methyl-N,3,5-tris(methy-
loxy)benzamide (68), -78 °C to room temperature, 20%; (b) 21b, K2CO3,
acetone, reflux, 66%.
a (a) 60a to 61a: AlCl3, 4-chloroanisole, CH2Cl2, reflux, 96%; (b) 61a
to 62a: Br2CH2CO2Et, K2CO3, acetone, reflux; (c) 62a to 63a: LiOH, H2O,
EtOH, THF, rt, 98% for two steps; (d) HOBT, EDAC-HCl, 2-methyl-4-
(1-oxido-4-thiomorpholinyl)aniline (64), Et3N, 22%; (e) CuCN, DMF, 16
°C, 11%.
commercially available benzoyl chloride derivatives. The dif-
ference from Scheme 1 is the B-ring origin. In this case,
2-bromo-4-chlorophenol, rather than the anisole derivative,
provided the B-ring. When two equivalents of n-BuLi were used
in THF instead of ether, a dianion was formed by deprotonation
and metal-halogen exchange. Reaction with the Weinreb amide
then provided the A/B-ring, which did not require a deprotection
step before the alkylation to attach the linker region and C-ring.
Method B is the approach described in Scheme 7. The B-ring
source was the Weinreb amide 52, and the A-rings were formed
from metal halogen exchange of di-meta-substituted aryl
bromides. Since the C-ring was not varied in this analogue set,
it was most efficient to attach the linker to the C-ring before
alkylation of the B-rings. This is similar to the approach shown
in Scheme 3, using 21b. More efficient alkylations were
achieved with the addition of two equivalents of sodium iodide
to the reaction mixture.
Table 3. SAR of Monosubstituted A-Ring Analogues
potency (IC50, nM)a
R
wt
17
1500
4
K103N
Y181C
V106A
67
H
100
155
5
150
700
6
1400
2000
na
66a
66b
66c
o-CN
m-CN
p-CN
26
44
355
1400
Chan et al. have previously disclosed the mono-meta sub-
stituted analogue 18b,7 which is a very potent compound, but
does show some diminution of potency against a V106A mutant
virus (Table 4). Addition of a methyl group (70g) or a chloro
group (70h) at the second meta position on the A-ring had a
dramatic impact on the ability of the analogue to inhibit the
V106A strain, without sacrificing the potency in the wild type,
K103N, and Y181C viruses.
a IC50 values determined using a HeLa-MAGI assay. Wild-type virus
(wt) was the HxB2 strain. K103N, Y181 C, and V106A mutants were
isogenic with the HxB2 virus backbone.
Our further exploration of meta substitution patterns in the
A-ring began to focus on disubstitution. A number of examples
were prepared using two different approaches for building the
A/B-ring. Method A, a variation of the approach shown in
Scheme 1, is shown in Scheme 9. As in Scheme 1, the A-ring
was formed from a Weinreb amide, readily accessed from the
We made a number of analogues to explore the impact of
di-meta substitution, particularly on the mutant viruses. Most

HIV Nonnucleoside ReVerse Transcriptase Inhibitor
Table 4. SAR of Disubstituted A-Ring Analogues
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 733
In summary, we have developed several variants of ben-
zophenone syntheses and used them to systematically explore
the A-ring, B-ring, linker region, and C-ring sections of this
novel template. Our efforts have culminated in the discovery
of very potent analogues, including 70g and 70h, both of which
compare quite favorably to efavirenz and nevirapine in vitro
and offer potential advantages in the treatment of virus resistant
to the current NNRTIs.
Selection of 70h (GW678248). Table 4 illustrates a situation
that is not unusual in drug discovery: we have identified more
than one compound with an attractive potency profile in our
primary assays. Multiple analogues had excellent antiviral
activity against both wild type and mutant viruses, particularly
those associated with clinical resistance to efavirenz and
nevirapine. The selection of a good drug candidate, however,
does not depend on in vitro potency alone, and we obtained
some preliminary pharmacokinetic data to help distinguish
between the most potent benzophenones.
Intravenous plasma concentration vs time plots of two of the
most potent compounds are shown in Figure 4. In all three of
the species studied, 70h had a lower clearance rate and longer
elimination half-life than 70g. This is consistent with our
hypothesis that metabolism of the aryl methyl group in 70g was
blocked when that methyl was replaced by the chloro group in
70h. On the basis of the potency in antiviral assays and the
initial pharmacokinetics data, 70h was selected for further
preclinical evaluation and development.
potency (IC50, nM)a
wt K103N Y181C V106A
methodb
R1
R2
18b
42d
70a
70b
70c
70d
70e
70f
H
F
CN
CF3
0.7
1.4
1.5
2.0
370
3.3
93
6.1
3.6
140
0.9
1.0
1.1
25
1.5
2.9
13
30
10.5
500
4.2
18
13
9.4
18
1.4
3.4
5.6
1.8
8200
A
A
B
B
B
A
B
B
B
OMe OMe 16
Me
Br
Cl
Me
CF3
Br
Me
CN
CN
CN
CN
0.4
30
2.2
1.2
16
0.5
0.5
1.0
0.8
88
3.0
290
6.9
3.3
450
1.9
0.7
2.0
1.6
10300
Cl
(N)
Me
Cl
70 g
70h
70i
CF3
EFVc
NVPc
5800
a IC50 values determined using a HeLa MAGI assay. Wild-type virus
(wt) was the HxB2 strain. K103N, Y181 C, and V106A mutants were
isogenic with the HxB2 virus backbone. b Method A is shown in Scheme
9; method B is shown in Scheme 7. c EFV is efavirenz; NVP is nevirapine.
of these analogues had very good profiles against the wt,
K103N, Y181C, and V106A viruses, including the fluoro,
triflouromethyl (42d), the chloro, bromo (70d), and the chloro,
methyl (70e). Although all of these were more potent against
the V106A strain than the mono-meta substituted analogue, they
were not quite as potent as the methyl, cyano analogue 70g.
There appear to be some size restrictions to the di-meta
substitution approach. For example, the bromo, trifluromethyl
analogue 70c is significantly less potent across the panel than
its fluoro, trifluoromethyl counterpart 42d.
Use of electron-donating substituents (70a) was clearly less
favorable than electron withdrawing substituents, although
potency against the Y181C mutant virus was less affected than
potency against the other viruses in the panel. Use of neutral
substituents (methyl groups), however, was very well tolerated.
In fact, the dimethyl analogue 70b was similar in potency to
the methyl, cyano analogue 70g.
Conclusions
The benzophenone template for NNRTI inhibitors has been
transformed from a lead with moderate activity against wild
type HIV and little or no activity against clinically relevant
mutants to a very promising series with excellent in vitro
potency against both wild type and key mutant viruses, including
the Y181C mutation associated with nevirapine resistance and
the K103N mutation associated with efavirenz resistance. In
particular, 70h (GW678248) has very low IC50 values in
antiviral assays: 0.5 nM against wild-type virus, 1 nM against
K103N, and 0.7 nM against Y181C. Compound 70h has also
demonstrated relatively low clearance in intravenous pharma-
cokinetic studies in three species. On the basis of this very
promising profile, GW678248 was selected for further develop-
ment. It has been converted to a prodrug to increase its oral
bioavailability, which progressed to phase 2 clinical trials.
In addition to di-meta substitution, we also explored replace-
ment of the meta-substituted phenyl ring with a meta-substituted
pyridine (70f). In this case, although the wild type and V106A
mutants were still fairly sensitive, the K103N and Y181C
mutants were not.
Experimental Section
NMR spectra were recorded on a Varian XL200, Varian Unity
Plus 400, or Varian Unity Plus 200 MHz spectrometers. Elemental
analyses were carried out by Atlantic Microlabs, Inc., Atlanta, GA.
Figure 4. Plasma concentration-time profiles of 70g and 70h (GW678248) in rats (A), dogs (B), and cynomolgus monkeys (C) following intravenous
administration at 1 mg/kg.

734 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Romines et al.
Some mass spectral data were obtained using atmospheric pressure
chemical ionization (APCI). All purchased starting materials were
used without further purification. All solvents were reagent grades.
Unless otherwise noted, all reactions were run in oven-dried, round-
bottomed flasks under nitrogen atmosphere. Representative proce-
dures are included below for compounds 4, 6, 7-13, 15-17, 21b,
23b, 24c, 26a, 27a, 29-31, 33-35, 37-40, 44-46, 48, 49, 51a,
52, 53a, 54a, 55a, 61a, 63a, 65a, 66a, 69a, and 70h. The syntheses
of reference compounds 3,6 18b,7 42a,7 and 42d7 have been
previously published. Procedures and characterization data for all
other intermediates and final products can be found in the
Supporting Information.
3,5-Difluoro-N-methoxy-N-methylbenzamide (6): Triethy-
lamine (20 mL) was added to a solution of N,O-dimethylhydroxy-
lamine hydrochloride (4.74 g, 48.6 mmol) in CHCl3 (100 mL), and
the mixture was cooled to 0 °C. 3,5-Difluorobenzoyl chloride (6.60
g, 37.4 mmol) was added dropwise over 5 min. The reaction was
stirred at 0 °C for 35 min and then allowed to warm to room
temperature over 1 h. The reaction mixture was partitioned between
EtOAc (200 mL) and water (100 mL). The organic layer was
washed with water (100 mL) and brine, dried over MgSO4, filtered,
and concentrated in vacuo to give 6 (6.963 g, 93%): 1H NMR (400
MHz, CDCl3) δ 7.27-7.22 (m, 2H), 6.94-6.89 (m, 1H), 3.57 (s,
3H), 3.37 (s, 3H) ppm.
(5-Chloro-2-methoxyphenyl)(3,5-difluorophenyl)methanone
(7): n-BuLi (11.3 mL of 1.6 M solution in hexanes, 18.2 mmol)
was added dropwise over 4 min to a solution of 2-bromo-4-
chloroanisole (2.2 mL, 15.8 mmol) in Et2O (100 mL) cooled to
-78 °C. The resulting mixture was stirred at -78 °C for 20 min,
and then a solution of 6 (3.49 g, 17.3 mmol) in Et2O (5 mL) was
added dropwise over 5 min. The reaction mixture was stirred at
-78 °C for 1 h and then allowed to warm to room temperature
over 3.25 h. The reaction was quenched by adding water (20 mL)
and stirring for 20 min. The mixture was then partitioned between
water (100 mL) and Et2O (100 mL). The aqueous layer was
extracted with Et2O (50 mL), and the combined organic layers were
dried over MgSO4, filtered, and concentrated in vacuo to give 5.191
g of crude material. Purification by flash chromatography using
3% EtOAc/hexanes as eluant gave 7 (1.861 g, 42%) as a white
solid: 1H NMR (400 MHz, CDCl3) δ 7.44 (dd, J ) 2.5, 8.9 Hz,
1H), 7.32 (d, J ) 2.9 Hz, 1H), 7.29-7.24 (m, 2H), 7.03-6.97 (m,
1H), 6.92 (d, J ) 8.9 Hz, 1H), 3.70 (s, 3H) ppm; 13C NMR (100
MHz, CDCl3) δ 192.6, 163.0 (JCF ) 11.5, 250 Hz), 156.1, 140.6,
132.5, 129.5, 129.0, 126.2, 113.1, 112.6 (JCF ) 6.9, 19.1 Hz), 108.6
(JCF ) 25.6 Hz), 56.2 ppm.
(aq), and the mixture was then partitioned between EtOAc (50 mL)
and water (50 mL). The aqueous layer was extracted with EtOAc
(35 mL), and the combined organic layers were dried over MgSO4,
filtered, and concentrated in vacuo to give 10 (1.691 g, 81%): MS
(ES-) m/z 325 (M - H); 1H NMR (400 MHz, DSMO-d6) δ 7.54-
7.50 (m, 3H), 7.48 (dd, J ) 2.7, 9.1 Hz, 1H), 7.36 (d, J ) 2.7 Hz,
1H), 6.93 (d, J ) 9.1 Hz, 1H), 4.27 (s, 2H) ppm; 13C NMR (100
MHz, DSMO-d6) δ 192.9, 162.9 (JCF ) 12, 248 Hz), 155.7, 140.4
(JCF ) 8 Hz), 132.4, 129.1, 129.0, 124.6, 115.7, 113.3 (JCF ) 26
Hz), 109.5 (JCF ) 26 Hz), 67.6 ppm.
[4-Chloro-2-(3,5-difluorobenzoyl)phenoxy]acetyl chloride (11):
Oxalyl chloride (2.7 mL of 2.0 M solution in CH2Cl2, 5.44 mmol)
was added dropwise over 14 min to a solution of 10 (0.611 g, 2.04
mmol) in DMF (0.10 mL) and CH2Cl2 (12 mL). The resulting
mixture was stirred at room temperature for 2.25 h then concentrated
in vacuo. The crude product was used immediately without further
purification or characterization.
{5-Chloro-2-[2-(2,3-dihydro-1H-indol-1-yl)-2-oxoethoxy]phe-
nyl}(3,5-difluorophenyl)methanone (4): A solution of 11 (0.51
mmol) in acetone (2 mL) was added dropwise over 2 min to a
mixture of indoline (0.05 mL, 0.44 mmol), acetone (5 mL), sodium
bicarbonate (0.384 g, 8.97 mmol), and water (0.5 mL), and the
resulting mixture was stirred at room temperature for 2.5 h. The
mixture was then poured into EtOAc (50 mL) and water (25 mL).
The aqueous layer was separated and extracted with EtOAc (25
mL). The combined organic layers were dried over MgSO4, filtered
in concentrated in vacuo to yield 0.278 g of a yellow oil. Purification
by flash chromatography using 25% EtOAc/hexanes as eluant,
followed by crystallization from CH2Cl2/hexanes gave 4 (0.069 g,
1
37%) as white crystals: mp 158-160 °C; H NMR (400 MHz,
CDCl3) δ 8.14 (d, J ) 8.2 Hz, 1H), 7.44-7.39 (m, 4H), 7.22-
7.18 (m, 2H), 7.07-6.97 (m, 3H), 4.70 (s, 2H), 3.98 (t, J ) 8.2
Hz, 2H), 3.18 (t, J ) 8.2 Hz, 2H) ppm; 13C NMR (100 MHz,
CDCl3) δ 190.1, 162.5, 160.8 (JCF ) 251.7 Hz), 152.4, 140.4, 138.1,
130.3, 128.9, 127.4, 127.3, 125.7, 125.0, 122.7, 122.6, 115.2, 112.6,
110.7 (JCF ) 26.7 Hz), 106.7 (JCF ) 25.1 Hz), 66.8, 44.9, 26.4
ppm; IR (KBr) 1676, 1657 cm-1; MS (ES+) m/z 428 (M + H);
Anal. (C23H16NO3ClF2 + 0.28 equiv H2O) C, H, N.
2-[4-Chloro-2-(3,5-difluorobenzoyl)phenoxy]-N-(2-methylphe-
nyl)acetamide (12): Prepared in a method analogous to that
described for 4 above from acid chloride 11 (0.50 mmol) and
o-toluidine (0.05 mL, 0.43 mmol). Purification by flash chroma-
tography using 10% EtOAc/hexanes as eluant gave 12 (0.121 g,
58%) as a white solid: 1H NMR (400 MHz, CDCl3) δ 8.30 (br s,
1H); 7.71 (d, J ) 8.6 Hz, 1H), 7.53 (dd, J ) 2.5, 8.6 Hz, 1H), 7.36
(d, J ) 2.5 Hz, 1H), 7.34-7.31 (m, 2H), 7.22-7.17 (m, 2H), 7.09
(app t, J ) 7.1 Hz, 1H), 7.05-7.01 (m, 2H), 4.77 (s, 2H), 2.18 (s,
3H) ppm; 13C NMR (100 MHz, CDCl3) δ 191.6, 165.4, 163.1 (JCF
) 12, 252 Hz), 154.3, 139.9 (JCF ) 7.6 Hz), 134.7, 133.4, 130.8,
130.5, 130.3, 128.1, 127.3, 126.8, 126.0, 123.7, 114.8, 113.1 (JCF
) 7.6, 19 Hz), 109.4 (JCF ) 26 Hz), 68.1, 17.7 ppm; MS (ES+)
m/z 416 (M + H), 438 (M + Na); Anal. (C22H16NO3ClF2) C, H,
N.
(5-Chloro-2-hydroxyphenyl)(3,5-difluorophenyl)methanone (8):
BBr3 (9.8 mL of 1.0 M solution in CH2Cl2, 9.80 mmol) was added
dropwise over 20 min to a solution of 7 (1.847 g, 6.53 mmol) in
CH2Cl2 (65 mL) at -78 °C. The resulting mixture was stirred at
-78 °C for 35 min and then allowed to warm to room temperature
over 1.5 h. The reaction mixture was poured onto ice, stirred for
30 min, and extracted with CH2Cl2 (50 mL). The organic layer
was washed with brine, dried over MgSO4, filtered, and concen-
trated in vacuo to give 8 (1.725 g, 98%): MS (ES-) m/z 267 (M -
2-(2-Benzoyl-4-chlorophenoxy)-N-hydroxy-N-phenylaceta-
mide (13): 2-(2-Benzoyl-4-chlorophenoxy)acetyl chloride6 (0.1 g,
32 mmol) was dissolved in dry MeCN (2 mL), and a solution of
2-phenylhydroxylamine13 (0.35 g, 32 mmol) in MeCN (2 mL) was
added. The resulting mixture was stirred at room temperature for
3 h and then filtered and concentrated in vacuo. Silica gel
chromatography (elution with 25% EtOAc/hexanes) gave 13 (0.062
1
H); H NMR (400 MHz, CDCl3) δ 11.58 (s, 1H), 7.49-7.47 (m,
2H), 7.19-7.1 (m, 2H), 7.06-7.03 (m, 2H) ppm.
Ethyl [4-chloro-2-(3,5-difluorobenzoyl)phenoxy]acetate (9):
A mixture of 8 (1.72 g, 6.40 mmol), K2CO3 (3.98 g, 28.8 mmol),
and ethyl bromoacetate (0.82 mL, 7.36 mmol) in acetone (45 mL)
was heated to reflux for 4.33 h. The reaction mixture was then
partitioned between EtOAc (130 mL) and water (100 mL). The
aqueous layer was extracted with EtOAc (30 mL). The combined
organic layers were then dried over MgSO4, filtered, and concen-
trated in vacuo to give 9 (2.321 g, 100%): 1H NMR (400 MHz,
DSMO-d6) δ 7.43-7.34 (m, 4H), 7.02-6.99 (m, 1H), 6.77 (d, J )
8.5 Hz, 1H), 4.52 (s, 2H), 4.18 (q, J ) 7.1 Hz, 2H), 1.22 (t, J )
7.1 Hz, 3H) ppm.
1
g, 50%): MS (APCI+) m/z 404 (M + Na); H NMR (400 MHz,
CDCl3) δ 9.85 (s, 1 H), 7.85-7.0 (m, 13 H), 4.95 (s, 2H) ppm;
HRMS (ES+) m/z Calcd. 382.08406; Exp. 382.08409 (M + H).
N-[2-(4-Chlorophenyl)ethyl]benzamide (15): Benzoyl chloride
(3.9 mL, 33.7 mmol) was added in one portion to a solution of
2-(4-chlorophenyl)ethanamine (14) (5 g, 32.1 mmol) in pyridine
(50 mL) at 0 °C. The reaction mixture was then allowed to warm
to room temperature over 1.67 h and then poured into 200 mL of
water/ice mixture. Product 15 was then collected by filtration,
[4-Chloro-2-(3,5-difluorobenzoyl)phenoxy]acetic acid (10): A
suspension of 9 (6.40 mmol), LiOH-H2O (0.671 g, 16.0 mmol),
EtOH (5 mL), water (5 mL), and THF (20 mL) was stirred at room
temperature for 2.75 h. The pH was adjusted to 5 using 1 N HCl
1
washed with four portions of water, and dried in vacuo. H NMR
(400 MHz, CDCl3) δ 7.68-7.66 (m, 2H), 7.47 (d, J ) 7.2 Hz,

HIV Nonnucleoside ReVerse Transcriptase Inhibitor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 735
1H), 7.38-7.42 (m, 2H), 7.28 (d, J ) 8.2 Hz, 2H), 7.16 (d, J )
8.2 Hz, 2H), 6.09 (br s, 1H), 3.71-3.65 (m, 2H), 2.90 (t, J ) 6.8
Hz, 2H) ppm; 13C NMR (100 MHz, CDCl3) δ 168.6, 138.5, 135.6,
133.4, 132.6, 131.2, 129.9, 129.7, 127.9, 42.2, 36.2 ppm.
8.06 (d, J ) 5.6 Hz, 1H), 7.72-7.52 (m, 3H), 7.50-7.40 (m, 3H),
7.26 (br s, 2H), 7.17 (d, J ) 5.6 Hz, 1H), 4.89 (s, 2H), 2.23 (s,
3H) ppm; 13C NMR (100 MHz, CDCl3) δ 185.0, 166.7, 166.6, 162.5
(JCF ) 13, 248 Hz), 160.1, 142.5, 139.1 (JCF ) 12 Hz), 136.3,
132.0, 131.8, 128.4, 124.5, 124.4, 120.0, 119.0, 112.5 (JCF ) 27
Hz), 107.8 (JCF ) 26 Hz), 18.3, 14.4 ppm; MS (ES+) m/z 467 (M
+ H), 489 (M + Na); Anal. (C20H16N2O5F2S2) C, H, N.
N-[3-(1H-Imidazol-1-yl)propyl]-3-methyl-4-nitroaniline (26a):
A mixture of 5-fluoro-2-nitrotoluene (0.24 mL, 2.0 mmol), 1-(3-
aminopropyl)-imidazole (0.41 mL, 3.4 mmol), and sodium bicar-
bonate (0.302 g, 3.6 mmol) in pyridine (5 mL) and water (0.5 mL)
was heated to reflux for 3 h. The reaction mixture was then
partitioned between water (50 mL) and EtOAc (50 mL). The organic
layer was concentrated to give a yellow solid, which was purified
by crystallization from EtOAc/hexanes to give 26a (0.255 g,
49%): 1H NMR (400 MHz, CDCl3) δ 7.92 (d, J ) 9.2 Hz, 1H),
7.60 (s, 1H), 7.16 (s, 1H), 7.08 (t, J ) 5.5 Hz, 1H), 6.87 (s, 1H),
6.47 (dd, J ) 2.4, 9.2 Hz, 1H), 6.40 (d, J ) 2.4 Hz, 1H), 4.04-
3.98 (m, 2H), 3.06-3.01 (m, 2H), 2.47 (s, 3H), 1.98-1.91 (m,
2H) ppm; 13C NMR (100 MHz, CDCl3) δ 153.9, 138.0, 136.9,
129.2, 128.7, 120.0, 114.1, 110.3, 44.2, 30.5, 22.9 ppm.
7-Chloro-1-phenyl-3,4-dihydroisoquinoline (16): Solution of
15 (1.18 g, 4.5 mmol) in p-xylene (25 mL) warmed to reflux to
remove water and then removed from heat. POCl3 (10 g) and P2O5
(5 g) were added, and the mixture was heated at reflux for 6 h.
The xylene was decanted, and the solid residue was dissolved in
water. The aqueous layer was extracted with three portions of ether.
The combined organic layers were then concentrated in vacuo. Silica
gel chromatography (elution with 20-33% EtOAc/hexanes) gave
1
16 (0.5 g, 54%): MS (AP+) m/z 242 (M + H); H NMR (300
MHz, CDCl3) δ 7.60-7.57 (m, 2H), 7.48-7.43 (m, 3H), 7.37 (dd,
J ) 2.1, 8.0 Hz, 1H), 7.26-7.21 (m, 2H), 3.88-3.83 (m, 2H),
2.80-2.75 (m, 2H) ppm.
N-[2-(2-Benzoyl-4-chlorophenyl)ethyl]benzamide (17): A mix-
ture of 16 (0.1 g, 0.41 mmol), benzoyl chloride (0.2 mL, 1.7 mmol),
and triethylamine (0.1 mL, 0.72 mmol) in toluene (1 mL) and a
few drops of water was stirred at room temperature for 2 h. The
reaction mixture was washed with brine, dried over MgSO4, filtered,
and concentrated in vacuo. Silica gel chromatography (elution with
20-50% EtOAc/hexanes) gave 17 (0.053 g, 35%): 1H NMR (300
MHz, CDCl3) δ 7.95-7.92 (m, 2H), 7.86-7.69 (m, 4H), 7.57 (d,
J ) 2.1 Hz, 1H), 7.50 (d, J ) 8.2 Hz, 1H), 7.32-7.29 (m, 5H),
4.22-4.17 (m, 2H), 3.24-3.19 (m, 2H) ppm; 13C NMR (100 MHz,
CDCl3) δ 195.8, 165.6, 137.8, 135.5, 132.4, 132.2, 130.4, 129.9,
129.3, 129.2, 128.7, 126.8, 126.7, 126.5, 125.1, 40.1, 29.3 ppm;
Anal. (C22H18NO2Cl + 0.2 equiv H2O) C, H, N.
N-[4-(Aminosulfonyl)-2-methylphenyl]-2-bromoacetamide
(21b): A solution of 4-amino-3-methylbenzenesulfonamide (20b)7
(5.0 g, 26.85 mmol) and pyridine (2.4 mL, 29.53 mmol) in CHCl3
(150 mL) was cooled to 0 °C in an ice bath. Bromoacetyl bromide
(2.6 mL, 29.53 mmol) was added dropwise over 20 min, and the
resulting mixture was allowed to warm to room temperature
overnight. The reaction mixture was then poured into water (150
mL) and extracted with two 100-mL portions of CH2Cl2. The
aqueous layer was then filtered to give crude product as a beige
solid. The crude material was suspended in 1 N HCl and then
filtered and washed with CH2Cl2, MeOH, and hexane to give 21b
(5.705 g, 69%): 1H NMR (400 MHz, DMSO-d6) δ 9.84 (s, 1H),
7.66-7.56 (m, 3H), 7.23 (br s, 2H), 4.09 (s, 2H), 2.24 (s, 3H) ppm;
13C NMR (100 MHz, DMSO-d6) δ 166.0, 141.2. 139.4, 132.3,
128.4, 124.9, 124.4, 30.5, 18.4 ppm; Anal. (C9H11N2O3BrS) C, H,
N.
N4-[3-(1H-Imidazol-1-yl)propyl]-2-methyl-1,4-benzenedi-
amine (27a): A mixture of 26a (0.233 g, 0.90 mmol) and 10%
Pd/C in MeOH (20 mL) was stirred at room temperature under 52
psi of hydrogen gas for 1 h. The reaction mixture was filtered
through Celite and concentrated in vacuo to give 27a (0.166,
80%): 1H NMR (400 MHz, CDCl3) δ 7.48 (s, 1H), 7.07 (s, 1H),
6.92 (s, 1H), 6.58 (d, J ) 8.2 Hz, 1H), 6.40 (d, J ) 2.6 Hz, 1H),
6.36 (dd, J ) 2.6, 8.2 Hz, 1H), 4.08 (t, J ) 6.9 Hz, 2H), 3.49-
3.48 (m, 1H), 3.26 (br s, 2 H), 3.08-3.05 (m, 2H), 2.13 (s, 3H),
2.08-2.02 (m, 2H) ppm; 13C NMR (100 MHz, CDCl3) δ 140.9,
137.2, 136.6, 129.7, 124.3, 118.9, 116.6, 116.4, 112.3, 445, 41.9,
31.1, 17.7 ppm.
N-(3-Methyl-4-nitrophenyl)acrylamide (29): A mixture of
3-methyl-4-nitroaniline (28) (1.052 g, 6.91 mmol) and triethylamine
(1.16 mL, 8.29 mmol) in CH2Cl2 (20 mL) was cooled to 0 °C and
acryloyl chloride (0.62 mL, 7.61 mmol) was added dropwise over
5 min. The resulting mixture was stirred 1.5 h at 0 °C and then
diluted with CH2Cl2 (35 mL), washed with brine, dried over MgSO4,
filtered, and concentrated in vacuo to give 29 (1.941 g), which was
used without further purification: MS (ES-) m/z 205.3 (M - H);
1H NMR (400 MHz, CDCl3) δ 8.52 (br s, 1H), 8.01 (d, J ) 9.0
Hz, 1H), 7.75 (d, J ) 2.3 Hz, 1H), 7.65 (dd, J ) 2.3, 9.0 Hz, 1H),
6.49-6.40 (m, 2H), 5.78 (dd, J ) 3.4, 8.1 Hz, 1H), 2.60 (s, 3H)
ppm.
N-(3-Methyl-4-nitrophenyl)-3-(4-morpholinyl)propanamide
(30): A mixture of 29 (6.91 mmol) and morpholine (0.63 mL, 7.26
mmol) in EtOH (25 mL) was warmed to reflux for 2.3 h. The
reaction mixture was then concentrated in vacuo, suspended in
EtOAc, and filtered. The filtrate was concentrated in vacuo,
dissolved in EtOAc, and allowed to crystallize. The crystalline
impurity was removed by filtration, and the filtrate was concentrated
in vacuo to give 30 (1.767 g, 87%): 1H NMR (400 Mz, CDCl3) δ
11.24 (br s, 1H), 8.03 (d, J ) 8.9 Hz, 1H), 7.54 (d, J ) 2.3 Hz,
1H), 7.43 (dd, J ) 2.3, 8.9 Hz, 1H), 3.84-3.82 (m, 4H), 2.76-
2.73 (m, 2H), 2.64 (br s, 4H), 2.62 (s, 3H), 2.58-2.55 (m, 2H)
ppm.
N-(4-Amino-3-methylphenyl)-3-(4-morpholinyl)propana-
mide (31): A mixture of 30 (0.202 g, 0.69 mmol) and 10% Pd/C
(0.018 g) in MeOH (10 mL) was stirred at room temperature under
53 psi of hydrogen gas for 2.17 h. The reaction mixture was then
filtered through Celite and concentrated in vacuo to give 31 (0.192
g, quant.): 1H NMR (400 MHz, CDCl3) δ 10.44 (br s, 1H), 7.38
(d, J ) 2.5 Hz, 1H), 7.27 (dd, J ) 2.5, 8.3 Hz, 1H), 6.76 (d, J )
8.3 Hz, 1H), 3.97-3.92 (m, 4H), 2.91-2.83 (m, 2H), 2.77-2.72
(m, 4H), 2.66-2.62 (m, 2H), 2.25 (s, 3H) ppm.
(3,5-Difluorophenyl)(3-hydroxy-2-thienyl)methanone (23b): A
mixture of 3-methoxythiophene (22) (1.14 g, 10 mmol), AlCl3 (2.70
g, 20.2 mmol), and 3,5-difluorobenzoyl chloride (1.18 mL, 10
mmol) in CH2Cl2 (50 mL) was heated to reflux for 20 h and then
stirred an additional 27 h at room temperature. The reaction mixture
was then poured over ice and stirred for 40 min, and the aqueous
layer was separated and extracted with CH2Cl2 (20 mL). The
combined organic layers were dried over MgSO4, filtered, and
concentrated in vacuo to give a dark brown solid (1.214 g).
Purification by flash chromatography using 2% EtOAc/hexanes
gave 23b (0.518 g, 22%) as a yellow solid: 1H NMR (400 MHz,
CDCl3) δ 12.04 (s, 1H), 7.57 (d, J ) 5.5 Hz, 1H), 7.43 (dd, J )
2.1, 7.3 Hz, 2H), 7.02-6.97 (m, 1H), 6.84 (d, J ) 5.5 Hz, 1H)
ppm; 13C NMR (100 MHz, CDCl3) δ 188.2, 169.7, 163.1 (JCF
)
12.2, 252 Hz), 141.2, 135.9, 120.2, 111.9, 111.6 (JCF ) 7.6, 19.1
Hz), 107.9 (JCF ) 25.2 Hz) ppm.
N-[4-(Aminosulfonyl)-2-methylphenyl]-2-{[2-(3,5-difluoroben-
zoyl)-3-thienyl]oxy}acetamide (24c): A mixture of 23b (0.192 g,
0.80 mmol), 21b (0.252 g, 0.82 mmol), and K2CO3 (0.498 g, 3.6
mmol) in acetone (10 mL) was warmed to reflux for 6 h. The
reaction mixture was then poured into water (30 mL) and extracted
with two 30-mL portions of EtOAc. The combined organic layers
were dried over MgSO4, filtered, and concentrated in vacuo to give
0.272 g of crude material. Purification by flash chromatography
using 35-50% EtOAc/hexanes as eluant gave 24c as a yellow solid
3-(3-Methyl-4-nitrophenoxy)propylamine (33): 3-Methyl-4-
nitrophenol (32) (5.0 g, 33 mmol), 3-bromopropyl phthalimide (8.8
g, 33 mmol), and Cs2CO3 (16.1 g, 5.0 mmol) in DMF (60 mL)
were heated to 55 °C for 2 h and then cooled to room temperature.
The mixture was poured into a mixture of Et2O and water, and the
1
(0.103 g, 28%): H NMR (400 MHz, CDCl3) δ 9.47 (br s, 1H),

736 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Romines et al.
resulting mixture was filtered, washed with water and Et2O, and
dried in a vacuum oven at 45 °C for 16 h to give 2-[3-(3-methyl-
4-nitrophenoxy)propyl]-1H-isoindole-1,3(2H)-dione (94) (9.5 g,
85%) as a tan solid: 1H NMR (300 MHz, DMSO-d6) δ 8.05-8.01
(m, 1H), 7.90-7.84 (m, 4H), 6.88-6.85 (m, 2H), 4.16 (t, J ) 5.8
Hz, 2H), 3.79 (t, J ) 6.6 Hz, 2H), 2.51 (s, 3H), 2.15-2.07 (m,
2H) ppm. A solution of 94 (3.0 g, 8.8 mmol) and hydrazine hydrate
(1.6 mL, 53 mmol) in EtOH (50 mL) was heated to reflux for 4 h
and then stirred at room temperature for 48 h. The reaction mixture
was then filtered and concentrated in vacuo. The resulting solid
was washed with CH2Cl2 and then dissolved in EtOAc. The EtOAc
solution was washed with water, dried over MgSO4, filtered, and
concentrated in vacuo to give 33 (1.1 g, 59%) as a yellow oil: 1H
NMR (400 MHz, DMSO-d6) δ 7.99 (d, J ) 9.2 Hz, 1H), 6.97 (d,
J ) 2.8 Hz, 1H), 6.91 (dd, J ) 2.8, 9.1 Hz, 1H), 4.10 (t, J ) 6.4
Hz, 2H), 2.63 (t, J ) 6.7 Hz, 2H), 2.49 (s, 3H), 1.77-1.71 (m,
2H) ppm; 13C NMR (100 MHz, DMSO-d6) δ 163.1, 142.1, 137.1,
128.0, 118.5, 113.5, 67.1, 38.8, 33.1, 21.4 ppm.
N-[3-(3-Methyl-4-nitrophenoxy)propyl]urea (34): A solution
of 33 (0.3 g, 1.43 mmol) and trimethylsilyl isocyanate (0.21 mL,
0.18 mmol) in THF (5 mL) was stirred at room temperature for 3
h. The reaction mixture was then quenched with water (1 mL) and
concentrated in vacuo. The resulting residue was washed with a
mixture of EtOAc and Et2O, filtered, and dried to afford 34 (0.273
g, 75%) as a pale yellow solid: 1H NMR (400 MHz, DMSO-d6) δ
8.00 (d, J ) 9.2 Hz, 1H), 6.97 (d, J ) 2.5 Hz, 1H), 6.90 (dd, J )
2.8, 9.1 Hz, 1H), 5.97 (t, J ) 5.6 Hz, 2H), 5.35 (br s, 2H), 4.04 (t,
J ) 6.4 Hz, 2H), 3.08-3.03 (m, 2H), 2.44 (s, 3H), 1.81-1.74 (m,
2H) ppm.
N-[3-(4-Amino-3-methylphenoxy)propyl]urea (35): Compound
35 (0.045 g, 92%) was prepared from 34 according to the procedure
described above for 27a: 1H NMR (400 MHz, DMSO-d6) δ 6.51
(s, 1H), 6.48-6.43 (m, 2H), 5.92 (t, J ) 5.6 Hz, 1H), 5.33 (br s,
2H), 4.30 (br s, 2H), 3.75 (t, J ) 6.3 Hz, 2H), 3.08-3.03 (m, 2H),
1.96 (s, 3H), 1.70-1.64 (m, 2H) ppm.
Sodium 4-(dibenzylamino)-3-methylbenzenesulfonate (37):
Mixture of 2-aminotoluene-5-sulfonic acid (10 g, 53 mmol), Na2-
CO3 (22.3 g, 210 mmol), benzyl bromide (14.3 mL, 120 mmol),
water (120 mL), and CH2Cl2 (120 mL) was heated to reflux for 72
h. Reaction was quenched with EtOH, and the mixture was
concentrated in vacuo to give 27.4 g of crude 37, which was used
without further purification.
4-(Dibenzylamino)-3-methylbenzenesulfonyl chloride (38): A
solution of 37 (20.6 g, 53 mmol) in DMF (200 mL) was cooled to
0 °C, and thionyl chloride (19.0 g, 160 mmol) was added dropwise.
After stirring for 2 h at room temperature, the reaction mixture
was poured into ice, stirred for 30 min, and extracted with EtOAc.
The organic layers were then dried over Na2SO4, filtered, and
concentrated in vacuo to give 38 (5.0 g, 24%), which was used
immediately without further purification or characterization.
4-(Dibenzylamino)-N,N,3-trimethylbenzenesulfonamide (39):
Intermediate 38 (5.0 g, 13 mmol) was added slowly to a solution
of dimethylamine (11.6 mL of 5.6M solution in EtOH, 65 mmol)
at 0 °C, and the resulting mixture was stirred for 30 min at 0 °C.
The reaction mixture was then poured into EtOAc and water, and
the organic layer was separated, washed with brine, dried over
MgSO4, filtered, and concentrated in vacuo to give 39 (1.0 g,
20%): 1H NMR (400 MHz, DMSO-d6) δ 7.47 (d, J ) 2.2 Hz,
1H), 7.31-7.14 (m, 11H), 7.05 (d, J ) 8.6 Hz, 1H), 4.13 (s, 4H),
2.48 (s, 6H), 2.44 (s, 3H) ppm.
4-Amino-N,N,3-trimethylbenzenesulfonamide (40): A mixture
of 39 (0.330 g, 0.85 mmol), toluene (10 mL), and 10% Pd/C (0.050
g) was stirred under 40 psi of H2 gas for 6 h. The reaction mixture
was then filtered through Celite, and the filtrate was washed with
satd NaHCO3 (aq) and water. The organic layer was separated, dried
over MgSO4, filtered, and concentrated in vacuo to give 40 (0.120
g, 67%): 1H NMR (400 MHz, DMSO-d6) δ 77.21-7.18 (m, 2H),
6.64 (d, J ) 8.0 Hz, 1H), 5.74 (br s, 2H), 2.44 (s, 6H), 2.04 (s,
3H) ppm.
(30 mL), and the mixture was heated to reflux for 7 h. A solution
of KOH (0.2 g, 3.6 mmol) in water (1 mL) was added to the reaction
mixture, which was then heated for an additional 1 h. The reaction
was diluted with 1 N NaOH (25 mL) and extracted with three 25-
mL portions of CH2Cl2. The pH of the aqueous layer was then
adjusted to 4 with concentrated HCl to precipitate 44 (0.124 g,
1
30%): LCMS (APCI+) m/z 171 (M + H); H NMR (400 MHz,
DMSO-d6) δ 7.92 (d, J ) 9.5 Hz, 1H, 7.14 (d, J ) 9.5 Hz, 1H),
2.67 (s, 3H) ppm.
6-Methyl-5-nitro-2-pyridinesulfonamide (45): A mixture of 44
(0.115 g, 0.67 mmol) in 1 N HCl (5 mL) was cooled to 5 °C.
Chlorine gas was passed through the mixture for 30 min, and the
reaction was then stirred an additional 15 min. The reaction mixture
was extracted with two 5-mL portions of CH2Cl2. The combined
organic layers were cooled to 0 °C and NH3 (l) was dripped into
the mixture for 15 min via condensation of NH3 (g) on a -78 °C
coldfinger. The reaction was allowed to warm to room temperature
and stir overnight. The crude mixture was concentrated in vacuo,
and the residue was dissolved in EtOAc (15 mL), washed with
NaHCO3 (15 mL), dried over MgSO4, filtered, and concentrated
in vacuo to give 45 (0.06 g, 40%): GCMS (CI+) m/z 218 (M +
1
H); H NMR (400 MHz, DMSO-d6) δ 8.60 (d, J ) 8.4 Hz, 1H),
7.94 (d, J ) 8.5 Hz, 1H), 7.71 (br s, 2H), 2.75 (s, 3H) ppm.
5-Amino-6-methyl-2-pyridinesulfonamide (46): A mixture of
45 (0.06 g, 0.27 mmol) and 10% Pd/C (0.011 g) in EtOH (10 mL)
was stirred under H2 (g) overnight. The reaction mixture was filtered
and concentrated in vacuo to give 46, which was used without
further characterization or purification.
6-Amino-5-methyl-3-pyridinesulfonic acid (71): 3-Methyl-2-
pyridineamine (1 mL, 1 mmol) was heated to 160 °C in 20% fuming
sulfuric acid (2 mL) for 20 h. The reaction was then allowed to
cool to room temperature, and ice (ca. 10 mL) was added. Filtration
of the precipitate yielded 71 (0.094 g, 50%): 1H NMR (300 MHz,
1
DMSO-d6) δ 13 (br s, 1H), 7.88 (s, 4H), 2.14 (s, 3H) ppm; H
NMR (300 MHz, DMSO-d6 with D2O) δ 7.88 (s, 2H), 7.87 (s,
2H), 2.13 (s, 3H) ppm; LCMS (ES+) m+1/z 189.
6-{[(E)-(Dimethylamino)methylidene]amino}-5-methyl-3-py-
ridinesulfonic acid (48): Thionyl chloride (0.5 mL, 6.8 mmol) was
added to a mixture of 71 (0.9 g, 4.8 mmol) and DMF (30 mL), and
the resulting mixture was stirred at room temperature for 40 min.
The reaction mixture was filtered to obtain a solid, which was
washed with hexane to give 48 (0.85 g, 77%): 1H NMR (400 MHz,
DMSO-d6) δ 8.12 (d, J ) 2.2 Hz, 1H), 7.50 (s, 1H), 7.00 (s, 2H),
6.47 (s, 2H), 2.02 (s, 3H) ppm; LCMS (ES+) m/z 244 (M + H).
6-Amino-5-methyl-3-pyridinesulfonamide (49): A mixture of
48 (1.0 g, 4.1 mmol) and PCl5 (0.85 g, 4.1 mmol) was heated to
130 °C for 1.5 h. The resultant POCl3 was removed under high
vacuum. Concentrated NH4OH (25 mL) was carefully added at
room temperature, and the resulting mixture was heated to reflux
for 4 h and then allowed to stand at room temperature for 60 h.
Filtration gave 49 (0.35 g, 45%): 1H NMR (400 MHz, DMSO-d6)
δ 8.12 (d, J ) 2.4 Hz, 1H), 7.50 (d, J ) 2.3 Hz, 1H), 7.00 (br s,
2H), 6.47 (br s, 2H), 2.02 (s, 3H) ppm; LCMS (ES+) m/z 188 (M
+ H).
1,3-Dibromo-5-chlorobenzene (51a): A solution of 1,3,5-
tribromobenzene (50) (1.57 g, 5 mmol) in Et2O (25 mL) was cooled
to -78 °C, and n-BuLi (3.4 mL of 1.6 M solution in hexanes, 5.5
mmol) was added dropwise over 5 min. The reaction mixture was
stirred for 10 min, and then a mixture of Cl3CCCl3 (1.18 g, 5 mmol)
in Et2O (10 mL) was added dropwise over 3 min. The reaction
mixture was stirred for 15 min at -78 °C and then allowed to warm
to room temperature over 1 h. The reaction mixture was poured
into water (50 mL) and extracted with two 50-mL portions of
EtOAc. The combined organic layers were then dried over MgSO4,
filtered, and concentrated in vacuo to give 51a as a yellow solid
(1.24 g, 92%): 1H NMR (300 MHz, CDCl3) δ 7.57 (t, J ) 1.6 Hz,
1H), 7.47 (d, J ) 1.6 Hz, 2H) ppm; 13C NMR (100 MHz, CDCl3)
δ 136.0, 132.8, 130.6, 123.3 ppm; Anal. (C6H3Br2Cl) C, H, N.
5-Chloro-N,2-dimethoxy-N-methylbenzamide (52): Oxalyl
chloride (25 mL of 2 M solution in CH2Cl2, 50 mmol) was added
dropwise over 30 min to a mixture of 5-chloro-2-methoxybenzoic
6-Methyl-5-nitro-2-pyridinethiol (44): Thiourea (0.2 g, 2.6
mmol) was added to a solution of 43 (0.4 g, 2.4 mmol) in ethanol

HIV Nonnucleoside ReVerse Transcriptase Inhibitor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 737
acid (3.73 g, 20 mmol) and 1 drop of DMF in CH2Cl2 (100 mL).
The reaction mixture was stirred at room temperature for 21 h and
then concentrated to yield 5-chloro-2-methoxybenzoyl chloride (4.1
g, quant), which was used immediately. A mixture of (MeO)-
MeNH-HCl (2.54 g, 26 mmol) and triethylamine (10 mL) in CHCl3
(40 mL) was stirred at room temperature for 10 min and then cooled
to 0 °C. A solution of 5-chloro-2-methoxybenzoyl chloride (4.1 g,
20 mmol) in CHCl3 (20 mL) was added dropwise over 30 min,
and the resulting mixture was stirred at 0 °C for 30 min and then
allowed to warm to room temperature over 2 h. The reaction mixture
was then poured into water (100 mL) and extracted with CH2Cl2
(100 mL). The organic layer was washed with water (100 mL) and
brine, dried over MgSO4, filtered, and concentrated in vacuo to
give 52 (4.265 g, 93%): 1H NMR (400 MHz, CDCl3) δ 7.27 (dd,
J ) 2.6, 8.8 Hz, 1H), 7.20 (s, 1H), 6.82 (d, J ) 8.8 Hz, 1H), 3.79
(s, 3H), 3.46 (br s, 3H), 3.29 (br s, 3H) ppm.
of 7 to 9: 1H NMR (400 MHz, DMSO-d6) δ 12.96 (br s, 1H),
7.67-7.64 (m, 1H), 7.59 (dd, J ) 2.8, 9.0 Hz, 1H), 7.49 (d, J )
2.8 Hz, 1H), 7.43-7.38 (m, 3H), 7.07 (d, J ) 9.0 Hz, 1H), 4.49 (s,
2H) ppm.
2-({2-[(2-Bromophenyl)carbonyl]-4-chlorophenyl}oxy)-N-[2-
methyl-4-(1-oxido-4-thiomorpholinyl)phenyl]acetamide (65a): A
mixture of 63a (0.443 g, 1.2 mmol), 1-hydroxybenztriazole (0.16
g, 1.2 mmol), 1-(3-dimethylaminopropyl)-3-ethylcarbodiimide hy-
drochloride (0.23 g, 1.2 mmol), triethylamine (0.34 mL, 2.4 mmol)
and aniline 64 (0.40 g, 1.8 mmol) in DMF (15 mL) was stirred at
room-temperature overnight. The reaction mixture was then par-
titioned between EtOAc and water. The organic layer was separated,
washed with water and brine, dried over MgSO4, filtered, and
concentrated in vacuo. Column chromatography on silica gel
(elution with 2% MeOH/CH2Cl2) gave 65a (0.154 g, 22%) as an
off-white foam: 1H NMR (400 MHz, DMSO-d6) δ 8.80 (s, 1H),
7.70-7.64 (m, 2H), 7.50-7.36 (m, 4H), 7.24 (d, J ) 8.8 Hz, 1 H),
7.15 (d, J ) 8.8 Hz, 1H), 6.85 (d, J ) 2.6 Hz, 1H), 6.79 (dd, J )
3, 8 Hz, 1H), 4.62 (s, 2H), 3.73-3.67 (m, 2H), 3.56-3.51 (m, 2H),
2.91-2.84 (m, 2H), 2.67-2.61 (m, 2H), 2.07 (s, 3H) ppm.
2-[4-Chloro-2-(2-cyanobenzoyl)phenoxy]-N-[2-methyl-4-(1-
oxido-4-thiomorpholinyl)phenyl]acetamide (66a): Copper cyanide
(0.037 g, 0.42 mmol) was added to a solution of 65a (0.120 g,
0.21 mmol) in DMF (5 mL), and the resulting mixture was heated
to 160 °C and stirred overnight. The mixture was then cooled, and
water was added. The resulting solid was filtered and washed with
EtOAc. The filtrate was dried over MgSO4 and concentrated in
vacuo. Column chromatography on silica gel (elution with 10-
100% EtOAc/hexanes) gave 66a (0.012 g, 11%) as an orange
foam: 1H NMR (400 MHz, DMSO-d6) δ 8.97 (s, 1H), 7.99-7.96
(m, 1H), 7.76-7.66 (m, 4H), 7.56 (d, J ) 2.7 Hz, 1H), 7.22 (d, J
) 8.9 Hz, 1H), 7.06 (d, J ) 8.6 Hz, 1H), 6.82 (d, J ) 3.0 Hz, 1H),
6.76 (dd, J ) 3, 9 Hz, 1H), 4.62 (s, 2H), 3.73-3.69 (m, 2H), 3.55-
3.53 (m, 2H), 2.90-2.84 (m, 2H), 2.65-2.60 (m, 2H), 1.99 (s,
3H) ppm; MS (ES-) m/z 521 (M - H); Anal. (C27H24N3O4ClS +
0.3 H2O) C, H, N, S.
[3,5-Bis(methyloxy)phenyl](5-chloro-2-hydroxyphenyl)metha-
none (69a): A solution of 2-bromo-4-chlorophenol (0.830 g, 4.0
mmol) in THF (20 mL) was cooled to -78 °C. n-Butyllithium (5.5
mL of a 16 M solution in hexanes, 8.8 mmol) was added dropwise
over 5 min, and the resulting mixture was stirred at -78 °C for 1
h. A solution of 68 (0.901 g, 4.0 mmol) in THF (5 mL) was added
dropwise over 4 min, and the resulting mixture was stirred at -78
°C for 1.25 h and then at room temperature for 14 h. The reaction
mixture was poured into water (50 mL) and extracted with two
50-mL portions of EtOAc. The combined organic layers were dried
over MgSO4, filtered, and concentrated in vacuo to give 1.193 g
of a brown oil. Column chromatography on silica gel (elution with
10% EtOAc/hexanes), followed by crystallization from hot ether
gave 69a (0.234 g, 20%) as yellow crystals: 1H NMR (300 MHz,
CDCl3) δ 11.83 (s, 1H), 7.62 (d, J ) 2.6 Hz, 1H), 7.45 (dd, J )
2.6, 8.9 Hz, H), 7.03 (d, J ) 8.9 Hz, 1H), 6.76 (d, J ) 2.3 Hz,
2H), 6.68 (t, J ) 2.3 Hz, 1H), 3.84 (s, 6H) ppm; 13C NMR (100
MHz, CDCl3) δ 200.6, 161.9, 160.9, 139.1, 136.5, 132.6, 123.6,
120.2, 119.9, 107.1, 104.5, 55.9 ppm; Anal. (C15H13O4Cl) C, H,
N.
(3-Bromo-5-chlorophenyl)(5-chloro-2-methoxyphenyl)metha-
none (53a): A solution of 51a (0.270 g, 1.0 mmol) in Et2O (5 mL)
was cooled to -78 °C, and n-BuLi (0.7 mL of 1.6 M solution in
hexanes, 1.1 mmol) was added dropwise over 2 min. The reaction
mixture was stirred at -78 °C for an additional 10 min, and then
52 (0.230 g, 1.0 mmol) was added in small portions over 4 min.
The reaction mixture was stirred at -78 °C for 1.25 h and then
allowed to warm to room temperature and stir for 14 h. The mixture
was then poured into water (25 mL) and extracted with EtOAc (25
mL) and CH2Cl2 (25 mL). The combined organic layers were dried
over MgSO4, filtered, and concentrated in vacuo to give 53a (0.351
1
g, 97%): MS (APCI-) m/z 357.9 (M - H); H NMR (300 MHz,
CDCl3) δ 7.76 (app t, J ) 1.5 Hz, 1H), 7.70 (app t, J ) 1.8 Hz,
1H), 7.65 (app t, J ) 1.5 Hz, 1H), 7.47 (dd, J ) 2.7, 8.9 Hz, 1H),
7.36 (d, J ) 2.7 Hz, 1H), 6.95 (d, J ) 8.9 Hz, 1H), 3.72 (s, 3H)
ppm; 13C NMR (100 MHz, CDCl3) δ 192.4, 156.1, 140.3, 135.7,
135.6, 132.7, 130.9, 129.7, 128.8, 128.5, 126.3, 123.0, 113.2, 56.2
ppm.
3-Chloro-5-(5-chloro-2-methoxybenzoyl)benzonitrile (54a): A
mixture of 53a (0.299 g, 0.83 mmol), NaCN (0.086 g, 1.76 mmol),
CuI (0.028 g, 0.15 mmol), and Pd(PPh3)4 (0.113 g, 0.10 mmol) in
acetonitrile (8 mL) was warmed to reflux for ca. 45 min. The
reaction mixture was cooled to room temperature, diluted with
EtOAc (50 mL), and filtered through Celite. The solution was then
washed with water (25 mL), dried over MgSO4, filtered, and
concentrated in vacuo. Column chromatography on silica gel
(elution with 5% EtOAc in hexanes) gave 54a (0.171 g, 56%): 1H
NMR (400 MHz, CDCl3) δ 7.93 (app t, J ) 1.8 Hz, 1H), 7.82
(app t, J ) 1.5 Hz, 1H), 7.76 (app t, J ) 1.7 Hz, 1H), 7.47 (dd, J
) 2.5, 8.9 Hz, 1H), 7.37 (d, J ) 2.5 Hz, 1H), 6.93 (d, J ) 8.9 Hz,
1H), 3.67 (s, 3H) ppm; 13C NMR (100 MHz, CDCl3) δ 191.8, 156.2,
140.1, 135.9, 135.6, 133.6, 133.3, 1314, 130.0, 128.0, 126.7, 116.9,
114.4, 113.3, 56.1 ppm.
3-Chloro-5-(5-chloro-2-hydroxybenzoyl)benzonitrile (55a):
Prepared from 54a according to the procedure described above for
7. Isolated 55a (0.174 g, quant) as a yellow solid: 1H NMR (400
MHz, CDCl3) δ 11.43 (s, 1H), 7.84-7.82 (m, 2H), 7.78 (t, J ) 1.5
Hz, 1H), 7.49 (dd, J ) 2.7, 9.0 Hz, 1H), 7.34 (d, J ) 2.7 Hz, 1H),
7.05 (d, J ) 9.0 Hz, 1H) ppm; 13C NMR (100 MHz, CDCl3) δ
196.9, 162.1, 139.9, 137.7,136.3, 135.1, 133.2, 131.6, 130.5, 124.4,
120.9, 118.9, 116.6, 114.8 ppm.
(2-Bromophenyl)(5-chloro-2-hydroxyphenyl)methanone (61a):
A mixture of 2-bromobenzoyl chloride (10 g, 46 mmol), aluminum
chloride (6.2 g, 46 mmol), and 4-chloroanisole (5.6 mL, 46 mmol)
in CH2Cl2 (250 mL) was stirred at reflux overnight. The reaction
mixture was then cooled to room temperature, poured into ice-
water, and extracted with CH2Cl2. The organic layer was dried over
MgSO4, filtered, and concentrated in vacuo to give 61a (13.76 g,
96%) as a tan solid: 1H NMR (400 MHz, DMSO-d6) δ 10.92 (s,
1H), 7.70 (d, J ) 7.9 Hz, 1H), 7.49-7.42 (m, 3H), 7.54 (dd, J )
2.7, 9.0 Hz, 1H), 7.29 (d, J ) 2.8 Hz, 1H), 6.98 (d, J ) 9.0 Hz,
1H) ppm.
N-[4-(Aminosulfonyl)-2-methylphenyl]-2-[4-chloro-2-(3-chloro-
5-cyanobenzoyl)phenoxy]acetamide (70h): Prepared from 55a and
21b according to the procedure described above for 24a. Column
chromatography on silica gel (elution with 0.5-1% MeOH/CH2-
Cl2) gave 70h (0.033 g, 12%) as a light yellow solid: 1H NMR
(400 MHz, DMSO-d6) δ 9.42 (s, 1H), 8.26 (s, 1H), 8.11 (s, 1H),
8.03 (t, J ) 1.6 Hz, 1H), 7.63 (dd, J ) 2.7, 8.9 Hz, 1H), 7.60-
7.53 (m, 3H), 7.49 (d, J ) 2.7 Hz, 1H), 7.22 (s, 2H), 7.19 (d, J )
9.1 Hz, 1H), 4.77 (s, 2H), 2.14 (s, 3H), ppm; 13C NMR (100 MHz,
DMSO-d6) δ 192.0, 166.6, 155.2, 141.0, 139.9, 139.1, 136.7, 135.1,
133.8, 133.4, 132.7, 131.9, 130.1, 128.6, 128.3, 126.0, 124.5, 124.3,
117.5, 116.0, 114.3, 67.7, 18.4 ppm; MS (APCI-) m/z 517 (M -
H); Anal. (C23H17N3O5Cl2S) C, H, N.
({2-[(2-Bromophenyl)carbonyl]-4-chlorophenyl}oxy)acetic acid
(63a): Intermediate 61a was converted in two steps to 63a (17.24
g, 98%) according to the procedures described above for conversion
Activity against HIV-1 in an Acute Infection Assay Using
MT4 Cells. The human T-cell lymphotropic virus type-1 trans-

738 Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2
Romines et al.
formed cell line MT4 was infected with HIV-1 (strain IIIB) at 100
times the amount necessary to cause a 50% reduction in cell growth.
The infected cells were incubated for 5 days in the presence of
various concentrations of each of the compounds to be tested. HIV-
1-mediated cytopathic effects (CPE) were assessed by an MTS
Staining method. A compound that expresses activity against HIV-1
prevents this CPE. Each value represents the mean from a single
experiment performed in triplicate. Compound induced cytotoxicity
was measured in parallel in uninfected cells (data not shown). For
the benzophenone compound series, cytotoxicty was not observed
up to 0.5 µM, which was, in most cases, the highest concentration
tested.
determined by reference to a corresponding calibration curve
constructed by linear or quadratic regression analysis of 1/x- or
1/x2-weighted standard peak areas. Pharmacokinetic parameter
values were estimated by noncompartmental analysis of the plasma
concentration-time curves.
Acknowledgment. The authors thank Gwyer Lovell, Brian
Owens and Pat Wheelan for conducting the pharmacokinetics
studies and Thimysta Burnette for providing the pharmacoki-
netics data.
HeLa MAGI Assay. The assay used in this report is a modified
version of the HeLa MAGI system14 in which HIV-1 infection is
detected by the activation of an HIV-LTR driven â-galactosidase
reporter in HeLa CD4-LTR- â-gal cells (obtained from Dr. Michael
Emerman through the AIDS Research and Reference Reagent
Program, Division of AIDS, NIAID, NIH). Quantitation of â-ga-
lactosidase is achieved by measuring the activation of a chemilu-
minescent substrate (Tropix/Applied Biosystems, Foster City, CA)
3 days after infection. The concentration of each compound required
to inhibit 50% (IC50) of the HIV-1-induced â-galactosidase signal,
relative to untreated controls was determined for each isogenic
recombinant virus. Cytotoxicity was measured in parallel using the
CellTiter96 assay (Promega, Madison, WI) to ensure that the loss
of signal was in fact due to a reduction in HIV replication and not
cytotoxicity (data not shown). For the benzophenone compound
series, cytotoxicty was not observed up to 0.5 µM, which was, in
most cases, the highest concentration tested.
Virus Strains: HIV-1 strain IIIB was derived from cell-free
supernatants of cultures of the chronically infected cell line H93B
(H9/HTLV-IIIB). The nevirapine-resistant strain designated Nev-R
contains the Y181C mutation in RT. HIV-1 strain HxB2 was
derived from the molecular clone pHxB2-D.15 Construction of
recombinant DNA-derived isogenic strains containing wild type
(WT) RT or NNRTI-resistant mutations K103N, V106A, or Y181C
is described in ref 10.
Pharmacokinetic Studies. Compounds 70g and 70h (GW678248)
were each dissolved in PEG400:Solutol:H2O (30:20:50) at 0.5 mg/
mL and administered intravenously via bolus infusion at 1 mg/kg
(2 mL/kg) to male Han Wistar rats, beagle dogs, and cynomolgus
monkeys (two animals per species per compound). Blood samples
were collected in EDTA-containing tubes prior to dosing and at
selected times up to 24 h post-dose from each animal. The samples
were kept on ice until centrifuged to obtain plasma (typically within
30 min of collection), and the resulting plasma samples were stored
at or below -20 °C until analyzed. Rat studies were conducted at
GlaxoSmithKline (RTP, NC) in accordance with GSK Institutional
Animal Care and Use Committee (IACUC) protocols and guide-
lines. In-life portions of the dog and monkey studies were conducted
at Primedica Corporation/Charles River Laboratories (Worcester,
MA) under their IACUC-approved protocols. All plasma analyses
were performed at GlaxoSmithKline.
Plasma samples and calibration standards (0.1-mL aliquot) were
deproteinated by addition of acetonitrile (0.3-0.4 mL) or methyl-
tert-butyl ether (0.5 mL) followed by centrifugation (15800g, 4 °C,
5 min). The resulting supernatants were evaporated to dryness at
37 °C under nitrogen, and each dried extract was reconstituted with
20% acetonitrile in 0.1% acetic acid, pH 4.5 (0.1 mL). The plasma
extracts were analyzed by HPLC/MS using a Phenomenex Aqua
C18 Mercury column (3 micron; 20 × 2 mm) and a Sciex API
365 triple-quadrupole mass spectrometer equipped with a Turbo
Ion Spray source. Mobile phases were: (A) 5% acetonitrile in 0.1%
acetic acid, pH 4.6, and (B) 0.1% acetic acid in neat acetonitrile.
Samples (10- to 30-µL aliquots) were eluted from the analytical
column with a 1-min linear gradient from 80%A:20%B to 5%A:
95%B followed by a 1-min isocratic elution with 5%A:95%B at a
constant flow rate of 0.4 mL/min. The analytes were detected by
negative-ion MRM, monitoring the transitions m/z 496.1 to m/z
270.0 for 70g or m/z 516.1 to m/z 289.7 for 70h (GW678248) at a
collision energy of 35 eV. Drug concentrations in plasma were
Supporting Information Available: Procedures and charac-
terization data for 18a, 19, 24a,b, 41a-d, 42b-c, 42e-f, 56a,b,
57a,b, 58a,b, 59, 66b-c, 67, 70a-g, 70i and all intermediates used
in the preparation of these compounds; results from combustion
analyses. This material is available free of charge via the Internet
at http://pubs.acs.org.
References
(1) Staszewski, S.; Morales-Ramirez, J.; Tashima, K. T.; Rachlis, A.;
Skiest, D.; Stanford, J.; Stryker, R.; Johnson, P.; Labriola, D. F.;
Farina, D.; Manion, D. J.; Ruiz, N. M. Efavirenz plus zidovudine
and lamivudine, efavirenz plus indinavir, and indinavir plus zidovu-
dine and lamivudine in the treatment of HIV-1 infection in adults.
N. Engl. J. Med. 1999, 341, 1865-1873.
(2) Zhang, Z.; Hamatake, R.; Hong, Z. Clinical utility of current NNRTIs
and perspectives of new agents in this class under development.
AntiViral Chem. Chemother. 2004, 15, 121-134.
(3) Guidelines for the use of antiretroViral agents in HIV-1-infected adults
and adolescents. Published by the US Department of Health and
Human Services. Oct 29, 2004. Available at http://AIDSinfo.nih.gov.
(4) Grant, R. M.; Hecht, F. M.; Warmerdam, M.; Liu, L.; Liegler, T.;
Petropoulos, C. J.; Hellmann, N. S.; Chesney, M.; Busch, M. P.;
Kahn, J. O. Time trends in primary HIV-1 drug resistance among
recently infected persons. JAMA 2002, 288, 181-188.
(5) Ross, L. L.; Lim, M. L.; Liao, Q.; Wine, B.; Rodriguez, A. E.;
Weinberg, W.; Shaefer, M. Prevalence of antiretroviral drug resistance
and resistance mutations in antiretroviral therapy (ART) na¨ıve HIV
infected individuals from 44 cities during 2003. Presented at the 44th
Interscience Conference on Antimicrobial Agents and Chemotherapy
Meeting, Oct. 30 - Nov. 2, 2004, Washington, DC, Abstract H-173.
(6) Wyatt, P. G.; Bethell, R. C.; Cammack, N.; Charon, D.; Dodic, N.;
Dumaitre, B.; Evans, D. N.; Green, D. V. S.; Hopewell, P. L.;
Humber, D. C.; Lamont, R. B.; Orr, D. C.; Plested, S. J.; Ryan, D.
M.; Sollis, S. L.; Storer, R.; Weingarten, G. G. Benzophenone
derivatives: A novel series of potent and selective inhibitors of human
immunodeficiency virus type 1 reverse transcriptase. J. Med. Chem.
1995, 38, 1657-1665.
(7) Chan, J. H.; Freeman, G. A.; Tidwell, J. H.; Romines, K. R.; Schaller,
LT.; Cowan, J. R.; Gonzales, S. S.; Lowell, G. S.; Andrews, C. W.;
Reynolds, D. J.; St. Clair, M.; Hazen, R. J.; Ferris, R. G.; Creech,
K. L.; Roberts, G. B.; Short, S. A.; Weaver, K.; Koszalka, G. W.;
Boone, L. R. Novel benzophenones as nonnucleoside reverse
transcriptase inhibitors of HIV-1. J. Med. Chem. 2004, 47, 1175-
1182.
(8) (a) Stammers, D. K.; Stuart, D. I.; Ren, J.; Chamberlain, P. P.;
Weaver, K.; Short, S.; Andrews, C.; Romines, K.; Boone, L.; Schaller,
L.; Freeman, A.; Chan, J. Structural studies of benzophenone/HIV-1
RT complexes: Insights into the potency of the next generation
NNRTIs against wt and mutant HIV-1. Presented at the XIII
International HIV Drug Resistance Workshop, June 8-12, 2004,
Canary Islands, Spain, Poster 30. (b) Stammers, D. K., et al.,
unpublished results.
(9) Structure-activity relationships in this manuscript were determined
using HIV antiviral assays. These antiviral assays have been shown
to correlate well with enzyme inhibition in biochemical assays (ref
7, footnote 10).
(10) Ferris, R.; Hazen, R.; Roberts, G.; St. Clair, M.; Chan, J.; Romines,
K.; Freeman, G.; Tidwell, J.; Schaller, L.; Cowan, J.; Short, S.;
Weaver, K.; Selleseth, D.; Moniri, K.; Boone, L. Antiviral activity
of GW678248, a novel benzophenone nonnucleoside reverse tran-
scriptase inhibitor. Antimicrob. Agents Chemother. 2005, 49, 4046-
4051.
(11) 2-Bromo-1-methoxy-4-(trifluoromethyl)benzene was obtained by
O-methylation of triflouro-p-cresol using methyl idodide and potas-
sium carbonate in acetone, followed by bromination using bromine
and sodium acetate in acetic acid.

HIV Nonnucleoside ReVerse Transcriptase Inhibitor
Journal of Medicinal Chemistry, 2006, Vol. 49, No. 2 739
(12) The chloro starting material 43 was not commercially available, so
we prepared it in two steps from 6-methyl-2-pyridineamine. The
3-nitro group was installed using sulfuric and fuming nitric acids,
and the amine was converted to a chloro group using trimethylsilyl
chloride and tert-butyl nitrite.
(13) Kamm, O. â-Phenylhydroxylamine. Organic Syntheses CollectiVe
Volume I; Blatt, A. H., Ed.; John Wiley & Sons: New York, 1932;
pp 445-447.
(14) Kimpton, J.; Emerman, M. Detection of Replication-competent and
Pseudotyped Human Immunodeficiency Virus with a Sensitive Cell
Line on the Basis of Activation of an Integrated â-galactosidase Gene.
J. Virol. 1992, 66, 2232-2239.
(15) Fisher, A. G.; Collati, E.; Ratner, L.; Gallo, R. C.; Wong-Stall, F. A
molecular clone of HTLV-III with biological activity. Nature 1985,
316, 262-265.
JM050670L