Inhibitors of human immunodeficiency virus type 1 (HIV-1) attachment. 5. An evolution from indole to azaindoles leading to the discovery of 1-(4-benzoylpiperazin-1-yl)-2-(4,7-dimethoxy-1H-pyrrolo[2,3-c]-pyridin-3-yl) ethane-1,2-dione (BMS-488043), a drug candidate that demonstrates antiviral activity in HIV-1-infected subjects

Article abstract of DOI:10.1021/jm900843g

Azaindole derivatives derived from the screening lead 1-(4- benzoylpiperazin-1-yl)-2-(1H-indol-3-yl)ethane-1,2-dione (1) were prepared and characterized to assess their potential as inhibitors of HIV-1 attachment. Systematic replacement of each of the unfused carbon atoms in the phenyl ring of the indole moiety by a nitrogen atom provided four different azaindole derivatives that displayed a clear SAR for antiviral activity and all of which displayed marked improvements in pharmaceutical properties. Optimization of these azaindole leads resulted in the identification of two compounds that were advanced to clinical studies: (R)-1-(4-benzoyl-2-methylpiperazin-1-yl)-2-(4- methoxy-1H-pyrrolo[2,3-b]pyridin-3-yl)ethane-1,2-dione (BMS-377806, 3) and 1-(4-benzoylpiperazin-1-yl)-2-(4,7-dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-yl) ethane-1,2-dione (BMS-488043, 4). In a preliminary clinical study, 4 administered as monotherapy for 8 days, reduced viremia in HIV-1-infected subjects, providing proof of concept for this mechanistic class.

Full text of DOI:10.1021/jm900843g

7778 J. Med. Chem. 2009, 52, 7778–7787
DOI: 10.1021/jm900843g
Inhibitors of Human Immunodeficiency Virus Type 1 (HIV-1) Attachment. 5. An Evolution from Indole to
Azaindoles Leading to the Discovery of 1-(4-Benzoylpiperazin-1-yl)-2-(4,7-dimethoxy-1H-pyrrolo[2,3-c]-
pyridin-3-yl)ethane-1,2-dione (BMS-488043), a Drug Candidate That Demonstrates Antiviral Activity in
HIV-1-Infected Subjects¥
Tao Wang,*,† Zhiwei Yin,† Zhongxing Zhang,† John A. Bender,† Zhong Yang,† Graham Johnson,† Zheng Yang,‡
Lisa M. Zadjura,‡ Celia J. D’Arienzo,‡ Dawn DiGiugno Parker,§ Christophe Gesenberg,§ Gregory A. Yamanaka,
,#
Yi-Fei Gong, Hsu-Tso Ho, Hua Fang, Nannan Zhou, Brian V. McAuliffe, Betsy J. Eggers, Li Fan, Beata Nowicka-Sans,
Ira B. Dicker, Qi Gao,^ Richard J. Colonno, Pin-Fang Lin, Nicholas A. Meanwell,† and John F. Kadow*,†
†Departments of Chemistry and ‡Metabolism and Pharmacokinetics, Preclinical Candidate Optimization and §Discovery Pharmaceutics and
Virology, Bristol-Myers Squibb Research and Development, 5 Research Parkway, Wallingford, Connecticut 06492, and ^Analytical Research and
Development, Bristol-Myers Squibb Research and Development, 1 Squibb Drive, New Brunswick, New Jersey 08901. # Deceased June 27th, 2005.
Received June 10, 2009
Azaindole derivatives derived from the screening lead 1-(4-benzoylpiperazin-1-yl)-2-(1H-indol-3-yl)ethane-
1,2-dione (1) were prepared and characterized to assess their potential as inhibitors of HIV-1 attachment.
Systematic replacement of each of the unfused carbon atoms in the phenyl ring of the indole moiety by a
nitrogen atom provided four different azaindole derivatives that displayed a clear SAR for antiviral activity
and all of which displayed marked improvements in pharmaceutical properties. Optimization of these
azaindole leads resulted in the identification of two compounds that were advanced to clinical studies: (R)-
1-(4-benzoyl-2-methylpiperazin-1-yl)-2-(4-methoxy-1H-pyrrolo[2,3-b]pyridin-3-yl)ethane-1,2-dione (BMS-
377806, 3) and 1-(4-benzoylpiperazin-1-yl)-2-(4,7-dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-yl)ethane-1,2-dione
(BMS-488043, 4). In a preliminary clinical study, 4 administered as monotherapy for 8 days, reduced viremia
in HIV-1-infected subjects, providing proof of concept for this mechanistic class.
Introduction
bitors, 11 protease inhibitors, 1 integrase inhibitor, 1 fusion
inhibitor, and 1 CCR5 coreceptor antagonist.11 In principle,
mechanistically novel inhibitors should not be cross-resistant
to the existing anti-HIV/AIDS drugs and would find utility in
drug combinations aimed at salvage or, eventually, first line
therapy. An early and important event in HIV infection
The human immunodeficiency virus (HIV) infection pan-
demic is now over 25 years old and continues to be a major
medical threat to human kind.1-4 By the end of 2005, 40
million people worldwide were estimated to be infected with
HIV, and deaths from its sequelae, acquired immunodefi-
ciency syndrome (AIDSa), were occurring at the rate of
approximately 3 million people per annum. Highly active
antiretroviral therapy (HAART) has helped to decelerate
the spread of the HIV/AIDS epidemic in many parts of the
world and has turned HIV into a chronic disease for many
infected individuals. However, with long-term HAART, the
development of resistance has emerged as a concern of con-
siderable importance.5-10 More than 70% of patients cur-
rently relying on antiretroviral therapy (ART) have developed
some detectable level of resistance,5 and 20% of new infec-
tions carry key mutations to existing drug classes.9,10Thus,
there remains a need for the discovery and development of
new anti-HIV agents with mechanisms that are distinct from
the 31 currently marketed drugs, a pharmacopeia comprising
17 nucleoside and non-nucleoside reverse transcriptase inhi-
a Abbreviations: HIV-1, human immunodeficiency virus type 1;
AIDS, acquired immunodeficiency syndrome; HAART, highly active
antiretroviral therapy; ART, antiretroviral therapy; gp120, glycoprotein
120; CCR5, chemokine (C-C motif) receptor 5; CXCR4, chemokine
(C-X-C motif) receptor 4; CD4, cluster of differentiation 4; IC50, 50%
inhibitory concentration; EC50, 50% effective concentration; CC50, 50%
cytotoxic concentration; DMEM, Dulbecco’s modified Eagle medium;
FBS, fetal bovine serum; Luc, luciferase; LTR, long terminal repeat;
XTT, 2,3-bis(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-car-
boxanilide inner salt; PBMC, peripheral blood mononuclear cells; HeLa
cells, Henrietta Lacks cell line; CYP 450, cytochrome P450; QT interval,
a measure of the time between the start of the Q wave and the end of the T
wave in the heart’s electrical cycle; hERG, the human ether-a-go-go
related gene; Caco cells, human colorectal adenocarcinoma cells; HLM,
human liver microsomes; SAR, structure-activity relationship; T1/2
,
half-life; AUC, area under the curve; Cmax, the maximum concentration
of a drug in the body after dosing; Cmin, the minimum concentration that
a drug achieves in a tested area after the drug has been administrated and
prior to the administration of a second dose; LC, liquid chromatogra-
phy; HPLC, high pressure liquid chromatography; MS, mass spectro-
metry; HRMS, high resolution mass spectrometry; DEBPT,
3-(diethoxyphosphoryloxy)-1,2,3-benzotriazin-4(3H)-one; PEG, poly-
(ethylene glycol); Bz, benzoyl; MeOH, methanol; EtOH, ethanol; Et3N,
triethylamine; THF, tetrahydrofuran; EtOAc, ethyl acetate; Mg-
SO4, magnesium sulfate; K2CO3, potassium carbonate; Na2CO3, sodium
carbonate; AlCl3, aluminum chloride; NH4Cl, ammonium chloride;
DMSO, dimethyl sulfoxide; NaOMe, sodium methoxide; KOMe, potas-
sium methoxide; CuI, copper(I) iodide; CH2Cl2, methylene chloride;
NaHCO3, sodium bicarbonate; HCl, hydrogen chloride; DMF, N,N-
dimethylformamide; iPr2NEt, diisopropylethylamine.
¥ We dedicate this article to the memory of Dr. Gregory Yamanaka, a
valued and highly respected colleague who contributed significantly to
antiviral drug discovery at Bristol-Myers Squibb. Greg was a biochemist
involved in fundamental aspects of the biochemical profiling of two
marketed antiviral drugs, the HIV protease inhibitor atazanavir, mar-
keted as Reyataz, and the nucleoside-based HBV inhibitor entecavir,
marketed as Baraclude.
*To whom correspondence should be addressed. For T.W.: phone, 203-
677-6584; fax, 203-677-7702; e-mail, tao.wang@bms.com. For J.F.K.:
phone, 203-677-6910; fax, 203-677-7702; e-mail, john.kadow@bms.com.
r
pubs.acs.org/jmc
Published on Web 09/21/2009
2009 American Chemical Society

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23 7779
Figure 1. HIV-1 attachment inhibitors.
involves the formation of a specific interaction between the
membrane-bound HIV-1 glycoprotein 120 (gp120) and CD4,
a host cell protein expressed on human T cells and macro-
phages that is the primary receptor for HIV-1.12 The interac-
tion of HIV-1 with CD4 is an essential prelude to subsequent
biochemical steps that are necessary for efficient HIV-1 entry
into the host cell cytosol where the process of replication can
be initiated. An interruption of the interaction of viral gp120
with CD4 would be expected to interfere with HIV-1 infectiv-
ity at a very early step of the viral life cycle, a potentially
attractive target for systemic therapy and topical (micro-
bicidal) prophylaxis.13-24 We have pioneered a series of small
molecules that interfere with the interaction between HIV-1
gp120 and the host cell receptor and have described progress
toward identifying inhibitors that meet target criteria for
clinical evaluation.25-32Preliminary mechanistic studies indi-
cate that this class of HIV-1 attachment inhibitor binds to
viral gp120 and stabilizes a conformation not recognized by
CD4.28-30 However, under different experimental conditions,
compounds of this class may form a ternary complex with
gp120 and CD4 and interfere with CD4-induced exposure of
the HIV gp41 heptad repeats, the assembly of which is a
critical event in virus-host cell membrane fusion.31,32 Optimi-
zation of a screening lead, the indoleglyoxamide derivative 1
(Figure 1) which was identified using a cell-based assay,
provided potent inhibitors of HIV-1 infection in vitro.26,27
However, additional preclinical profiling of 1 and its close
analogues surfaced limitations associated with the physico-
chemical properties of this class of HIV-1 inhibitor that
compromised efficient drug formulation and delivery. In an
effort to address this problem, which was attributed in part to
the properties inherent to the indole heterocycle, we prepared
the four possible unsubstituted azaindole analogues of indole
2 (Figure 1), used as the prototype for this study, and
compared their in vitro profiles. In this article, we describe
the results of this survey and the subsequent chemical optimi-
zation that led to the identification of two compounds that
were advanced to clinical studies: the 7-azaindole HIV-1
attachment inhibitor BMS-378806 (3)26,28and the 6-azaindole
derivative BMS-488043 (4).30 Clinical studies with 4 have
validated HIV-1 attachment inhibition as a potential ap-
proach to the control of HIV-1 infection in vivo.33
Results and Discussion
The indole 2 inhibited virus entry in a pseudotype assay
using envelopes from the CCR5-dependent JRFL strain and
CXCR4-dependent LAI strain of HIV-1 with EC50 values of
4.0 and 4.9 nM, respectively, and was not overtly cytotoxic to
the HeLa host cell line (Table 1). In order to provide bench-
mark data for the series, this compound was profiled exten-
sively ina series of invitroassaysdesignedtopredict aspectsof
the clinical performance of the molecule and identify potential
problems that would interfere with preclinical and/or clinical
development. When incubated in a human liver microsomal
preparation, the half-life for the disappearance of 2 was 16.9
min, an interval considered to predict an intermediate clear-
ance rate in humans based on comparison to standards.
However, compound 2 had the potential to be well absorbed
in vivo based on its high permeability across a layer Caco-2
cells grown to confluence (Pc = 169 nm/s at pH 6.5). The
compound inhibited three CYP 450 enzymes considered to be
of clinical importance, with IC50 values of 2.6, 5.2, and 4.8 μM
toward the 2C19, 2C9, and 2D6 isoforms, respectively.
Although this inhibition was of modest potency, the data
were sufficient to provoke some concern for the potential for
drug-drug interactions in vivo and assessment of this aspect
of the biochemical pharmacology was incorporated into the
screening paradigm. Indole 2 was predicted to have low
potential for cardiac toxicity due to the induction of QT
rhythm abnormalities related to inhibition of the rapid com-
ponent of the delayed rectifier cardiac Kþ channel encoded by
the hERG gene based on an IC50 value of >80 μM in a
channel flux assay. However, the solubility of a crystalline
form of 2 in an aqueous buffer solution at pH 6.5 was low (16
μg/mL) and the molecule offered limited potential for for-
mulation as a salt because of the absence of basic functionality
and the high pKa of the indole NH. Collectively, these data
identified several weaknesses in the profile of 2 that became
the focus of an optimization campaign, most notably the
modest stability in HLM and the low aqueous solubility.
To assess the potential for addressing liabilities associated
with the indole chemotype, modification of the phenyl ring
of the core heterocycle was examined in the context of the
four possible azaindole isomers 5-8 (Figure 1), which were

7780 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23
Wang et al.

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23 7781
evaluated in a pseudotype assay using the CCR5-dependent
JRFL HIV-1 strain envelope. As illustrated in Table 1, the
introduction of a nitrogen atom resulted in molecules with
substantially different in vitro profiles when compared to the
parent molecule 2. The HIV-1 inhibitory activity of 2 was fully
preserved with the 4-aza isomer 5 and the 7-aza isomer 8, EC50
values of 1.6 and 1.7 nM, respectively. However, when the
nitrogen atom was incorporated at a more exposed position of
the core, antiviral potency was reduced, with the 6-aza isomer
7 a modest 5-fold less potent than 2, EC50 = 22 nM, and the 5-
aza analogue 6 a more dramatic 100-fold weaker, EC50 = 576
nM. However, the azaindole series demonstrated uniformly
improved profiles in assays predictive of pharmacokinetic and
pharmaceutical performance. The half-life in human liver
microsomes (HLM) for the four isomers varied from 38.5
to >100 min, representing a marked enhancement in meta-
bolic stability compared to 2 that indicated reduced potential
for first-pass metabolism in humans. This result may be
related, in part, to the improved aqueous solubility deter-
mined for this family of azaindoles 5-8 at pH 6.5, which is at
least 25-fold higher than that measured for the indole 2. The
basic nitrogen atom of the azaindole ring (pKa = 2.0-6.2)
offered additional avenues and options for formulation based
on the potential for salt formation, although stronger acids
would be required for the less basic isomers.34 The increased
basicity associated with the azaindole core appeared to corre-
late with the permeability of these compounds across a Caco-2
monolayer at pH 6.5. The highly permeable 7-azaindole 8
(pKa = 2.0) would be expected to exist almost exclusively as
the free base at pH 6.5 whilea small fraction of the 4-azaindole
5 (pKa = 5.0) would be present as the pyridinium cation,
leading to reduced permeability. In contrast, a considerable
amount of both the 5-azaindole 6, pKa = 6.2, and 6-azaindole
7, pKa = 6.0, would exist in the protonated form at pH 6.5,
which would reduce the rate of permeation across the Caco-2
membrane. However, it was anticipated that the basicity of
the azaindoles could be modulated by the judicious deploy-
ment of substituents (vide infre). Altering the core heterocycle
did not introduce any additional safety liabilities based on the
in vitro profiling data which included assessments that predict
cytotoxicity, hepatotoxicity, and cardiotoxicity. Of some
concern with the incorporation of pyridine-like structural
elements was the potential for increased CYP 450 inhibi-
tion.35,36 However, although signals were observed within
the matrix of compounds and recombinant enzymes tested,
there were no clear and consistent trends and all four azain-
dole isomers appeared to offer some improvement over the
parent compound 2.
7, since this isomer uniquely combined good intrinsic potency
with an opportunity to simultaneously substitute at C-4 and
C-7, a pattern associated with improved antiviral activity in
the indole series.27 Of particular interest in this context, the
introduction of a methoxy substituent at C-7 was anticipated
to not only enhance antiviral activity but also reduce the
basicity of the azaindole nitrogen atom, thereby leading to
enhanced membrane permeability.37,38 The substantial elec-
tron withdrawing properties of a methoxy substituent adja-
cent to the N atom of pyridine are clearly reflected in the
increased acidity of the conjugate acid of 2-methoxypyridine,
pKa = 3.06, compared to protonated pyridine, which has
pKa = 5.23.37,38 This is attributed to an electronegative effect
mediated through the C-O σ-bond that depends on proxi-
mity. In contrast,
a 4-methoxy substituent increases
the basicity of pyridine (pKa = 6.47), the result of the
dominance of electron donation via resonance through the
π system. The reduced basicity of the 6-aza nitrogen atom
combined with steric shielding by the methoxy substituent
would also be expected to attenuate the potential for oxidative
metabolism to the N-oxide and help to mitigate the potential
for inhibition of CYP 450 enzymes.35,36 To further improve
the metabolic stability in HLM, the methyl subsituent at-
tached to the piperazine ring that was introduced with 3 was
omitted, since SAR from a related series suggested that this
group is susceptible to metabolic modification. The key
challenge was to identify a molecule with the correct balance
of potency and metabolic stability that did not sacrifice
solubility and membrane permeability, an analysis that ratio-
nalized the 4,7-dimethoxy-6-azaindole derivative 4 as a target
molecule.
The 6-azaindole 4 was initially prepared from commercially
available 5-bromo-2-chloro-3-nitropyridine (9), as described
in Scheme 1. Treating 9 with 3-4 equiv of vinylmagnesium
bromide at -78 to -20 ꢀC afforded 4-bromo-7-chloro-6-
azaindole (10) in yields of up to 35%.39 A copper-catalyzed
tandem displacement of the chlorine and bromine atoms by
NaOMe or KOMe provided 4,7-dimethoxy-6-azaindole (11)
in yields as a high as 75%.40,41 A Friedel-Crafts type reaction
of 11 with either methyl or ethyl chlorooxoacetate promoted
42
by an excess of AlCl3 generated an ester 12 which was
hydrolyzed to the acid salt 13 using K2CO3 in aqueous
MeOH. In the final step, the acid salt 13 was coupled with
N-benzoylpiperazine using 3-(diethoxyphosphoryloxy)-1,2,3-
benzotriazin-4(3H)-one (DEBPT)43 as the dehydrating agent
to afford 4 in an overall isolated yield of up to 37% from the
azaindole 11. The structure of 4 was confirmed by single
crystal X-ray structure analysis and compared with that of 3
(Figure 2).44 In the solid state the 2-keto moiety of both
molecules is oriented away from 4-methoxy substituent of
the azaindole ring, while the two carbonyl moieties are
orthogonal to each other. In addition, both piperazine nitro-
gen atoms in each molecule are sp2 hybridized, resulting in a
planar arrangement of the proximal atoms while the phenyl
group is distorted from coplanarity with the amide carbonyl
as a function of allylic 1,3-strain.45-47In the solid state, the
piperazine methyl group of 3 adopts an axial disposition to
avoid unfavorable allylic 1,3-strain with the amide carbonyl
oxygen atom.45-48
The in vitro profile of the 7-azaindole 8 was encouraging,
and optimization of this chemotype was facilitated by the
commercial availability of the parent heterocycle. This initia-
tive took advantage of early structure-activity relationships
(SARs) developed for the indole series and resulted in the
identification of the 4-methoxy-7-azaindole derivative 3,
the first HIV attachment inhibitor to enter clinical studies
designed to validate this mechanistic approach to the control
of HIV-1.26 Unfortunately, 3 failed toachieve targeted plasma
concentrations following oral administration to humans,
an outcome attributed to a combination of moderate stability
in HLM, T1/2 = 37 min, and Caco-2 cell permeability (Pc =
51 nm/s) at the lower end of the range that would predict good
intestinal absorption in humans.33
The in vitro profiling of 4 revealed a compound with
physical properties (pKa, logD, and melting point) that
exhibited considerable similarity to those of 3, as summarized
in Table 1. However, notable differences between the two
compounds are the enhanced Caco-2 cell permeability and
In an effort to identify compounds with improved potency,
attention was focused on the 6-azaindole series represented by

7782 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23
Wang et al.
Figure 2. X-ray crystallographic structures of 3 and 4.
Scheme 1. Discovery Synthesis of Compound 4
Table 2. Pharmacokinetic Profile of Compounds 3 and 4 in the Rat
increased stability in HLM associated with 4, positive attri-
butes counterbalanced by a 5-fold reduction in solubility at
pH 6.5, a difference that is, however, eroded at lower pH.
The antiviral properties of 4 compare favorably with those
of 3 toward the laboratory strain of HIV-1 used for screening
(Table 1). As is typical for this series of HIV-1 attachment
inhibitor, viruses expressing a B-clade envelope protein are
more susceptible than other clades and 4 displayed a median
EC50 of 36.5 nM toward 40 clinical isolates, which compares
to 61.5 nM for 3.49
compound
parameter
dose (mpk)
3
4
5
5
oral Cmax (μg/mL)
oral AUC ((μg/h)/mL)
iv terminal T1/2 (h)
bioavailability (%)
0.09
0.5
0.3
19
1.9
6.3
2.4
90
(3), with the potential for boosting by coadministration of the
CYP 3A4 inhibitor ritonavir, since 4 is primarily metabolized
by this enzyme in vitro.50 Azaindole 4 was, however, consi-
dered unlikely to be a perpetrator of drug-drug interactions
based on its lack of significant inhibitory activity toward a
panel of recombinant CYP 450 enzymes, IC50 > 30 μM.
The 6-azaindole 4 is a crystalline, nonhygroscopic solid that
melts at 229.5-232 ꢀC and displays good stability as the solid,
suffering no degradation for at least 4 weeks at temperatures up
to 50 ꢀC and 75% relative humidity at 40 ꢀC. The compound is
soluble at 0.9 mg/mL at pH 1.5 and 0.04 mg/mL at pH 4-8 and
shows excellent stability in solution at pH 2 and pH 7 with no
observed chemical changes observed after 48 h at 37 ꢀC.
Azaindole 4 possessed a clean preclinical safety profile,
showing no significant cytotoxicity toward 11human cell lines
that included T-cells, peripheral blood monocytes (PBMC),
hepatocarcinoma, neuronal tumors, cervical, lung, and larynx
carcinomas and embryonic kidney cells, CC50 g 120 μM.
Similarly, the compound was not toxic to hepatocytes in vitro
at concentrations up to 200 μM in the absence or presence of
expression of any of the four major CYP 450 enzymes. The
lack of detectable blockade of the hERG rapid delayed
rectifier Kþ ion channel at concentrations of 80 μM in vitro
(Table 1) was confirmed in vivo with no drug-related ECG
changes observed in dogs at doses as high as 200 mpk, which
afforded Cmax = 33 μM. In exploratory toxicology studies, 4
was welltoleratedwhengivenorallytorats for 2 weeks atdaily
The pharmacokinetic properties of 4 were examined in
detail in several preclinical species.50 The oral bioavailability
of 4 in rats, dogs, and monkeys following administration as a
solution in 90% PEG400/10% EtOH was 90%, 57%, and
60%, respectively. This compares to an oral bioavailability of
19%, 77%, and 24%, respectively, for 3 in the same species
and reflects the enhanced membrane permeability of 4 ob-
served in Caco-2 cells and improved metabolic stability in rat,
dog, and monkey liver microsomes.50 Detailed comparative
pharmacokinetic data for 3 and 4 in the rat are compiled in
Table 2. The overall exposure of 4 as assessed by AUC was
over 10-fold higher when compared to 3, with a terminal
intravenous half-life 8-fold longer and a Cmax 20-fold higher.
Total body clearance was low in rats, dogs, and monkeys,
predicted effectively by in vitro liver microsomal stability
data, with the result that plasma levels at 12 h after oral
dosing were 18- to 160-fold higher than the median antiviral
EC50 toward the panel of 25 clinical isolates. Clearance in
humans was predicted to be low (1.5-3.0 (mL/min)/kg) based
on human liver microsomal stability data or allometric scaling
of the animal pharmacokinetic profiles, with the latter pre-
dicting a half-life in humans of over 3 h.50 The oral bioavail-
ability of 4 after administering a crystalline suspension (mean
particle size of ∼7 μm) to rats at a dose of 5 mg/kg was 55%,
indicative of reasonable dissolution in gastrointestinal fluid.
Collectively, these profiling data predicted that 4 should have
improved oral exposure in humans compared to BMS-378806

Article
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23 7783
doses of up to 200 mg/kg and to dogs for 10 days at daily doses
of up to 100 mg/kg, with the Cmin being 58- and 160-fold the
median EC50 at the higher doses, respectively. Additionally, 4
was not mutagenic in an Ames reverse-mutation assay at
concentrations of up to 5 mg/plate and the compound exerted
no significant effects toward a panel of 53 in vitro biochemical
assays at 10 μM. These data revealed a safety profile consis-
tent with exploring 4 in phase 1 clinical studies in normal
healthy volunteers and proof-of-concept studies in HIV-1-
infected subjects.33
Pharmacokinetic studies in normal healthy volunteers under
fasting conditions revealed that Cmax and AUC increased in a
dose-related but not dose-proportional fashion following ad-
ministration of 200, 400, or 800 mg of 4 formulated in
capsules.51 Higher doses of 1200, 1800, or 2400 mg resulted in
no significant increase in exposure. A solution dose of 200 mg
resulted in a 3-fold increase in exposure of 4 compared to the
tablet formulation, while pretreatment with ritonavir increased
exposure by 43% at a dose of 400 mg and administration with
food resulted in a 3- to 5-fold increase in exposure, higher
following a high fat meal compared to a light meal, over a
400-1200 mg dose range.51 In a multiple-dose study conducted
in HIV-1-infected adults, the antiviral activity, safety, and
tolerability of 4 were evaluated at doses of 800 or 1800 mg of
drug administered with a high fat meal.33 Patients received doses
of either 4 or placebo every 12 h for 8 days, with no serious
adverse events during the study. While none of the placebo-
treated subjects had a maximal viral load decline of >0.5 log10
copies/mL, 58% (7/12) of the patients treated with 800 mg of 4
had a viral load decline of >1.0 log10 measured on day 8.33 The
results of this preliminary clinical study with 4 demonstrate that
an orally bioavailable, small molecule HIV-1 attachment in-
hibitor can have potent antiviral activity in infected subjects.
All liquid chromatography (LC) data were recorded on a
Shimadzu LC-10AS liquid chromotograph using a SPD-10AV
UV-vis detector, and mass spectrometry (MS) data were
determined with a Micromass Platform for LC operating in
electrospray mode. The column used was selected from one of
three: column A is a YMC ODS-A S7 3.0 mm ꢀ 50 mm column;
column B is a PHX-LUNA C18 4.6 mm ꢀ 30 mm column;
column C is an Xterra 4.6 mm ꢀ 50 mm C18 5 μm column.
Gradient was performed from 100% solvent A/0% solvent B to
0% solvent A/100% solvent B (solvent A was 10% MeOH/90%
H2O/0.1% CF3CO2H; solvent B was 10% H2O/90% MeOH/
0.1% CF3CO2H). The flow rate was 5 mL/min, and the gradient
time was 2 min with a holding time of 1 min. The detector
wavelength was 220 nm.
Preparative HPLC methods were used to purify target com-
pounds. Compounds purified by preparative HPLC were di-
luted in MeOH (1.2 mL) and purified on a Shimadzu LC-10A
automated preparative HPLC system using the following meth-
ods: The column used was a YMC C18 S5 20 mm ꢀ 100 mm
column with a detector operating at a wavelength of 220 nm.
Purification ramped from the initial gradient (30% B, 70% A)
to the final gradient (100% B, 0% A) over 20 min with holding
for 3 min (100% B, 0% A). Solvent A was 10% MeOH/90%
H2O/0.1% CF3CO2H, and solvent B was 10% H2O/90%
MeOH/0.1% CF3CO2H.
Solubilities were measured 24-72 h after samples were added
into phosphate buffered solution (pH 6.5) at room temperature.
pKa values were obtained via spectrophotometric titration
method by using a pION GLpKa instrument. The log D values
were determined using a classical octanol/water shake flask
method using phosphate buffered solution at pH 6.5.
Melting points were determined using a TA SC 2920 modu-
lated differential scanning calorimeter or EZ-Melt automated
melting point apparatus.
Optical rotation values [RD] were determined on solutions of
the indole 2 and azaindoles 5-8 in MeOH-CHCl3 (1:1) at a
wavelength of 589 nm at 20 ꢀC or a solution in MeOH for
compound 3 at 23 ꢀC.
The experimental methodology used to profile the Caco-2
permeability and liver microsomal stability of compounds 2-8
followed protocols previously described.56
(R)-1-(4-Benzoyl-2-methylpiperazin-1-yl)-2-(1H-pyrrolo[2,3-
b]pyridin-3-yl)ethane-1,2-dione (8) and its precursors were pre-
pared as previously reported.26 The synthetic intermediates
methyl 2-oxo-2-(1H-pyrrolo[3,2-b]pyridin-3-yl)acetate, methyl
2-oxo-2-(1H-pyrrolo[3,2-c]pyridin-3-yl)acetate, and methyl 2-
oxo-2-(1H-pyrrolo[2,3-c]pyridin-3-yl)acetate were prepared as
previously described.42
(R)-1-(4-Benzoyl-2-methylpiperazin-1-yl)-2-(1H-indol-3-yl)-
ethane-1,2-dione (2). Et3N (2 mL) was added to a solution of
(R)-2-methyl-N-benzoylpiperazine (0.7 g, 3.43 mmol) and
2-(1H-indol-3-yl)-2-oxoacetyl chloride (0.5 g, 2.41 mmol) in
dry THF. The mixture was stirred at room temperature for
16 h before being quenched by adding 10% brine solution (100
mL). The aqueous layer was extracted with EtOAc (2 ꢀ 100
mL) and the combined organic layers were dried over MgSO4
and concentrated to give a residue, a portion of which (100 mg)
was purified by preparative HPLC to afford 2. HPLC reten-
Conclusions
In summary, an initial comparative survey demonstrated
that azaindoles had the potential to address several of the key
issues identified with indole-based HIV-1 attachment inhibi-
tors, particularly low aqueous solubility and poor metabolic
stability. Insights derived from the SAR associated with
indole-based inhibitors anticipated the 4,7-dimethoxy-6-
azaindole derivative 4. This HIV-1 attachment inhibitor
offered improved in vivo pharmacokinetic profiles in three
species and appeared to address the low permeability and
limited metabolic stability thatwere the mostserious liabilities
of 3, a prototypical compound that showed insufficient
exposure in human clinical studies.33 The 6-azaindole 4 has
established proof-of-concept for this mechanistically novel
class of HIV-1 inhibitor in infected human subjects. However,
despite the improvements described above, the pharmaceuti-
cal properties of 4 remain less than ideal. Oral dosing in
conjunction with a high fat meal is required to provide
targeted exposure in humans, an observation that prompted
an exploration of additional formulation approaches de-
signed to improve drug dissolution in the gastrointestinal
tract.44,52 An alternative solution to formulation and drug
dissolution issues in the form of a prodrug will be described in
future publications.53
tion time: 1.47 min (column C). Mp: 212-216 ꢀC. [R]20
D
-7.18ꢀ (c 4.15 mg/mL). 1H NMR (DMSO-d6): δ 1.18 (b,
3H), 2.70-4.90 (m, 7H), 7.26-7.55 (m, 8H), 8.12-8.22 (m,
2H). 13C NMR (DMSO-d6): δ 14.6, 15.6, 44.0, 49.6, 58.8,
112.6, 113.0, 120.9, 122.5, 123.5, 124.8, 126.9, 128.4, 129.5,
135.4, 136.8, 136.9, 166.8, 170.1, 186.6. MS m/z (MH)þ calcd
for C22H22N3O3: 376.17; found 376.10. HRMS m/z (MH)þ
calcd for C22H22N3O3: 376.1661; found 376.1673. Anal. Calcd
for C22H21N3O3: C, 70.38; H, 5.64; N, 11.19. Found: C, 69.99;
H, 5.54; N, 10.99.
Experimental Section
General Directions. 1H and 13C NMR spectra were obtained
on a Bruker 500 Ultrashield NMR spectrometer operating at
500 MHz for 1H and 125 MHz for 13C.
(R)-1-(4-Benzoyl-2-methylpiperazin-1-yl)-2-(1H-pyrrolo[3,2-b]-
pyridin-3-yl)ethane-1,2-dione (5). Methyl 2-oxo-2-(1H-pyrrolo[3,

7784 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23
Wang et al.
2-b]pyridin-3-yl)acetate (160 mg, 0.78 mmol) and K2CO3 (216 mg,
1.56 mmol) were dissolved in a 1:1 mixture of MeOH and H2O (10
mL). After 16 h, LCMS analysis revealed 2-oxo-2-(1H-pyrrolo[3,2-b]-
pyridin-3-yl)acetic acid (MS m/z calcd for C9H7N2O3 191.05;
found 190.87; HPLC retention time, 0.19 min (column A)) as the
major product. A sample of 1 N HCl was added to the reaction
mixture to adjust the pH to 5, the solvent removed under vacuum,
and the residue dissolved in DMF (4 mL). (R)-2-Methyl-N-ben-
being dried under high vacuum to yield 10 (52.3 g, 22%) as a
brown solid. LC purity: 95.7%. HPLC retention time: 1.62 min
(column B). 1H NMR (DMSO-d6): δ 6.58 (s, 1H), 7.79 (s, 1H),
8.05 (s, 1H), 12.5 (br s, 1H). 13C NMR (DMSO-d6): δ 102.2,
110.9, 129.3, 131.6, 132.9, 134.4, 137.2. MS m/z (MH)þ calcd for
C7H4BrClN2: 230.93; found 231.15.
4,7-Dimethoxy-1H-pyrrolo[2,3-c]pyridine (11). A three-neck,
3 L flask equipped with a Dean-Stark trap and an overhead
stirrer was charged with MeOH (500 mL) and stirred under an
atmosphereof N2. KOMe (75.6 g, 1.08 mol) was added in portions
followed by toluene (500 mL) and the solution stirred for 15 min
until it became clear. Celite 545 (40 g), toluene (250 mL),
4,4-dibromo-7-chloro-1H-pyrrolo[2,3-c]pyridine (25.0 g, 108
mmol), and additional toluene (250 mL) were added, and the
mixture was brought to reflux and held at 115 ꢀC for 7 h while
collecting a mixture of MeOH and toluene (600 mL total). The
solution was cooled to ambient temperature, and LCMS ana-
lysis revealed 4-bromo-7-methoxy-1H-pyrrolo[2,3-c]pyridine
(MS m/z (MH)þ calcd for C8H7BrN2O 226.97; found 227.00.
HPLC retention time of 1.20 min (column A)) as the major
product. MeOH (500 mL) was added to the solution, and the
temperature was once again increased to 115 ꢀC. At 60 ꢀC, CuI
(22.6 g, 119 mmol) and MeOH (100 mL) were added and the
mixture was held at 115 ꢀC for 1.5 h while collecting a mixture of
MeOH and toluene (750 mL total). The mixture was cooled to
ambient temperature, and MeOH (500 mL) and NH4Cl (58 g)
were added. The solution was stirred for 1 h and concentrated to
dryness. The residue was dissolved in EtOAc and filtered
through SiO2 (500 g), eluting with EtOAc. The filtrate was
collected, concentrated, and purified by flash chromatography
on a Biotage apparatus (75 M SiO2, 17.5% v/v EtOAc in
hexanes) to yield 4,7-dimethoxy-1H-pyrrolo[2,3-c]pyridine
(9.80 g, 51%) as a white solid. HPLC retention time: 1.36 min
(column B). 1H NMR (DMSO-d6): δ 3.87 (s, 3H), 3.96 (s, 3H),
6.45 (d, J = 2.6 Hz, 1H), 7.21 (s, 1H), 7.35 (dd, J = 2.6, 2.5 Hz,
1H), 11.74 (br s, 1H). 13C NMR (DMSO-d6): δ 52.4, 55.6, 99.0,
114.2, 121.1, 125.5, 126.6, 145.6, 145.8. MS m/z (MH)þ calcd for
C9H10N2O2: 179.08; found 179.05. HRMS m/z (MH)þ calcd for
C9H11N2O2: 179.0821; found 179.0811. Anal. Calcd for
zoylpiperazine HCl (190 mg, 0.79 mmol), DEPBT (233 mg, 0.78
mmol), and Pr2NEt (0.4 mL) were added to the solution. The
3
i
mixture was stirred at room temperature for 24 h, the solvents
were removed in vacuo, and the residue was partitioned between
10% Na2CO3 and EtOAc. The organic layer was separated, dried
over MgSO4, and concentrated under vacuum and the residue
purified by HPLC to afford 5 (13.2 mg, 4.5%). HPLC retention
time: 0.96 min (column A). Mp: 298-301 ꢀC. [R]20D -19.88ꢀ g-1
mL-1 dm-1 (c 3.77 mg/mL). 1H NMR (CD3OD-CDCl3): δ 1.31
(b, 3H), 5.10-3.00 (m, 7H), 7.28 (m, 1H), 7.44 (s, 5H), 7.87
(m, 1H), 8.27 (d, 1H, J = 10.5 Hz), 8.56 (m, 1H). 13C NMR
(CD3OD-CDCl3): δ 14.7, 16.1, 45.5, 50.6, 113.2, 119.1, 121.5,
127.2, 129.0, 130.6, 131.0, 135.2, 138.2, 143.0, 145.2, 167.0,
172.4, 185.6. MS m/z (MHþ) calcd for C21H21N4O3:
377.16; found 377.14. HRMS m/z (MH)þ calcd for
C21H21N4O3: 377.1614; found 377.1625. Anal. Calcd for
C21H20N4O3 0.4H2O: C, 65.75; H, 5.47; N, 14.60. Found: C,
3
65.52; H, 5.07; N, 14.49.
(R)-1-(4-Benzoyl-2-methylpiperazin-1-yl)-2-(1H-pyrrolo[3,2-c]-
pyridin-3-yl)ethane-1,2-dione(6). Compound6 was preparedfrom
methyl 2-oxo-2-(1H-pyrrolo[3,2-c]pyridin-3-yl)acetate by follow-
ing the general procedure described for 5 to afford 12.5 mg of
product (4.3%). HPLC retention time: 0.89 min (column A). Mp:
227 ꢀC. [R]20D -3.39ꢀ g-1 mL-1 dm-1 (c 3.85 mg/mL). 1H NMR
(CD3OD): δ 1.35 (b, 3H), 4.90-3.10 (m, 7H), 7.51 (b, 6H), 8.16
(d, 1H, J = 5.8 Hz), 8.61 (d, 1H, J = 4.5 Hz), 8.72 (m, 1H), 9.62
(b, 1H). 13C NMR (CD3OD): δ 14.6, 15.8, 45.8, 50.6, 108.5, 114.2,
122.6, 127.1, 128.9, 130.5, 135.2, 138.7, 141.9, 142.4, 143.6, 166.6,
172.6, 185.1. MS m/z (MH)þ calcd for C21H21N4O3: 377.16;
found 377.15. HRMS m/z (MH)þ calcd for C21H21N4O3:
377.1614; found 377.1630. Anal. Calcd for C21H20N4O3: C,
67.00; H, 5.35; N, 14.88. Found: C, 66.69; H, 5.21; N,
14.60.
C9H10N2O2 0.2H2O: C, 59.46; H, 5.77; N, 15.41. Found: C,
59.48; H, 5.41; N, 15.63.
3
(R)-1-(4-Benzoyl-2-methylpiperazin-1-yl)-2-(1H-pyrrolo[2,3-c]-
pyridin-3-yl)ethane-1,2-dione(7). Compound7 was preparedfrom
methyl 2-oxo-2-(1H-pyrrolo[2,3-c]pyridin-3-yl)acetate following
the proceduredescribedfor5 to afford12.7mg ofproduct (4.3%).
HPLC retention time: 0.89 min (column A). Mp: 249-253 ꢀC.
(4,7-Dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-yl)oxoacetic Acid
Ethyl Ester (12). Ethyl chlorooxoacetate (22.8 g, 167 mmol) was
added to a stirred slurry of AlCl3 (14.8 g, 111 mmol) in CH2Cl2
(320 mL). The solution was stirred for 1 h and then added
dropwise over 1.5 h to a stirred slurry of 4,7-dimethoxy-1H-
pyrrolo[2,3-c]pyridine (9.90 g, 56 mmol) and AlCl3 (7.4 g,
56 mmol) in CH2Cl2 (265 mL). After 1 h, the red reaction
mixture was poured over ice (1.1 kg) to afford a yellow solution
which was extracted with CH2Cl2 (2 ꢀ 300 mL). The combined
organic layers were concentrated to ∼100 mL, diluted with
brine (200 mL), and concentrated to remove the remaining
organic solvent. The residue was dissolved into EtOAc (800
mL), washed with H2O (200 mL) and brine (200 mL), dried
over MgSO4, and concentrated. The combined aqueous layers
were neutralized with NaHCO3 (42 g) and extracted with EtO-
Ac (3 ꢀ 600 mL) and the combined organic layers were dried
over MgSO4, concentrated, and combined with the previously
obtained organic material to give a crude, yellow-brown
oil (∼25 g). This material was purified by flash chromatography
using a Biotage apparatus (75 M SiO2, 50% v/v EtOAc in
hexanes) to afford (4,7-dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-
yl)oxoacetic acid ethyl ester (13.5 g, 87%) as a yellow semisolid.
LC purity: 98.0%. HPLC retention time: 1.28 min (column B).
1H NMR (DMSO-d6): δ 1.28 (t, J = 7.2 Hz, 3H), 3.83 (s,
3H), 3.99 (s, 3H), 4.32 (q, J = 7.2 Hz, 2H), 7.47 (s, 1H), 8.23
(s, 1H), 13.07 (br s, 1H). 13C NMR (DMSO-d6): δ 13.8,
52.9, 56.4, 61.3, 113.4, 118.8, 121.9, 122.5, 135.9, 145.5,
146.2,164.5. MS m/z (MH)þ calcd for C13H15N2O5: 279.10;
found 279.16.
[R]20 -5.15ꢀ g-1 mL-1 dm-1 (c 4.00 mg/mL). 1H NMR
D
(CD3OD-CDCl3): δ 1.28 (b, 3H), 5.10-3.10 (m, 7H), 7.46 (m,
5H), 8.34-8.18 (m, 3H), 8.81 (s, 1H). 13C NMR (CD3OD-
CDCl3): δ 14.5, 15.8, 45.7, 50.7, 113.8, 116.4, 127.1, 128.9, 130.4,
131.7, 134.8, 134.9, 135.3, 140.3, 140.9, 166.8, 185.8. MS m/z
(MH)þ calcd for C21H21N4O3: 377.16; found 377.10. HRMS m/z
(MH)þ calcd for C21H21N4O3: 377.1614; found 377.1632. Anal.
Calcd for C21H20N4O3 0.3H2O: C, 66.06; H, 5.44; N, 14.67.
3
Found: C, 65.94; H, 5.14; N, 14.46.
4-Bromo-7-chloro-1H-pyrrolo[2,3-c]pyridine (10). A solution
of freshly prepared vinylmagnesium bromide in THF (3.92 L,
3.92 mol) was added via a cannula to a stirred solution of 5-
bromo-2-chloro-3-nitropyridine (250 g, 1.05 mol) in THF (5 L)
at -78 ꢀC at such a rate that the temperature was maintained
below -40 ꢀC. The mixture was stirred for 1.5 h at -78 ꢀC before
being quenched with saturated aqueous NH4Cl solution (1.5 L)
and stirred for 16 h. The layers were separated. The aqueous
layer was extracted with EtOAc (3 ꢀ 1 L) and the combined
organic layer washed with brine and concentrated. The residue
was dried under high vacuum and suspended in EtOAc (1 L)
before being concentrated on a rotary evaporator (removed
∼800 mL of EtOAc) and cooled in a refrigerator for 1 h. The
mixture was filtered and the collected solid washed with EtOAc
(2 ꢀ 50 mL) and EtOAc/hexanes (1:1, 50, and 100 mL) before

Article
Potassium (4,7-Dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-yl)-
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23 7785
of zeocin, 100 μg/mL of hygromycin, and 200 μg/mL of Genet-
icin (G418). The last three antibiotics, used as selection markers,
were omitted in the compound screen assay.
oxoacetate (13). K2CO3 (28.58 g, 207 mmol) was added to a
solution of (4,7-dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-yl)oxo-
acetic acid ethyl ester (28.74 g, 103 mmol) in a 1:1 mixture
of MeOH/H2O (500 mL), and the mixture was stirred for 6 h.
The white precipitate was collected by filtration and washed
with H2O (10 mL) to afford 13 (18.0 g, 60%). LC purity:
95.3%. HPLC retention time: 0.69 min (column B). 1H NMR
(DMSO-d6): δ 3.83 (s, 3H), 3.99 (s, 3H), 8.17 (s, 1H) 7.46 (s,
1H). 13C NMR (DMSO-d6): δ 52.8, 56.3, 113.6, 119.2, 121.9,
122.8, 135.8, 145.8, 146.2, 166.5, 183.0. MS m/z (MH)þ of
the corresponding acid of potassium 2-(4,7-dimethoxy-1
Pseudotype Reporter Virus. Pseudotype viruses HIV-1JRFL
LAI-Δenv-luc (JRFL-Luc) and HIV-1LAI LAI-Δenv-luc (LAI-
Luc) were prepared at Bristol-Myers Squibb by cotransfect-
ing 293T cells with a plasmid containing proviral DNA of HIV-1
LAI-Δenv-luc (firefly luciferase in place of the envelope
gene) and a plasmid expressing env (JRFL or LAI), driven by
the HIV-LTR. The Lipofectamine Plus kit (Invitrogen) was
used for transfection according to the package instructions. The
pseudotype virus was harvested as cell supernatants on day 3
post-transfection and stored in 10 mL aliquots at -80 ꢀC. The
virus titer was determined in HeLa 67 cells using the assay
protocol described below with a luciferase reporter gene assay
(high sensitivity, Roche) as the end point. Virus input in the
screen assays was normalized by the maximal luminescence
signal.
H-pyrrolo[2,3-c]pyridin-3-yl)-2-oxoacetate calcd for C11H11
-
N2O5: 251.07; found 251.09. HRMS m/z (MH)þof the
corresponding acid of potassium 2-(4,7-dimethoxy-1H-
pyrrolo[2,3-c]pyridin-3-yl)-2-oxoacetate calcd for C11H11N2-
O5: 251.0668; found 251.0659.
1-(4-Benzoylpiperazin-1-yl)-2-(4,7-dimethoxy-1H-pyrrolo-
[2,3-c]pyridin-3-yl)ethane-1,2-dione (BMS-488043, 4). iPr2Et-
N (44.2 mL, 253 mmol) was added to a slurry of potassium
2-(4,7-dimethoxy-1H-pyrrolo[2,3-c]pyridin-3-yl)oxoacetate (26.1 g,
90.6 mmol) and benzoylpiperazine hydrochloride (30.8 g, 135.9
mmol) in DMF (800 mL). DEPBT (51.5 g, 172.1 mmol) in DMF
(65 mL) was added dropwise over 15 min, and the mixture was
stirred under nitrogen for 24 h before being concentrated. The
residue was dissolved in EtOAc (1.6 L) and was washed with
saturated aqueous NaHCO3 (560 mL) and brine (700 mL). During
each washing, the solid that formed between the layers was collected
by filtration. The EtOAc layer was concentrated to ∼200 mL and
the precipitated solid collected and washed with EtOAc (2 ꢀ 100
mL). The solid material was combined to afford 4 (18. 0 g, 47%).
HPLC retention time: 1.33 min (column B). Mp: 229.5-232 ꢀC. 1H
NMR (DMSO-d6): δ 3.63-3.34 (m, 8H), 3.83 (s, 3H), 4.00 (s, 3H),
7.40 (m, 6H), 8.15 (s, 1H), 13.0 (s, 1H). 13C NMR (DMSO-d6): δ
39.9, 45.5, 52.9, 56.8, 114.4, 119.2, 122.1, 122.2, 126.9, 128.4, 129.6,
135.3, 136.6, 145.7, 146.2, 166.5, 169.3, 185.5. MS m/z (MH) þ calcd
for C22H23N4O5: 423.17; found 423.19. Anal. Calcd for
Test Sample Preparation. Compounds were initially dissolved
in 100% DMSO (30 mM stocks) and then 3-fold serially diluted
in DMSO in 10 steps.
Drug Susceptibility Assay Using Pseudotype Reporter Viruses.
Drug susceptibility was assessed in a single-cycle viral infection
system.54 The HIV LAI-Δenv-luc virus, pseudotyped with either
the JRFL or LAI envelope, was used to infect HeLa 67 cells in
the presence of serially diluted compounds. Cells were plated 1
day prior to infection at 105 cells per well in 100 μL of colorless
medium. On the next day, duplicates of 2 μL of serially diluted
compounds were per well into two replica plates. DMSO alone
was used for the no-drug and no-virus controls. An amount of
100 μL per well of diluted virus stocks or 100 μL per well of
medium (no-virus control) was added immediately after com-
pound addition. Assay plates were incubated at 37 ꢀC for 3 days.
Virus yield was monitored by firefly luciferase activity 3 days
after infection according to manufacturer’s protocol (luciferase
reporter gene assay, Roche).
Cytotoxicity Assay. Compound cytotoxicity was assayed in
parallel for all experiments by exposing uninfected MT-2 cells to
serially diluted compounds and measuring cell health after 3
days in an XTT assay,55 according to the manufacturer’s
recommendations.
Data Analysis. The 50% effective concentration (EC50) was
calculated using the one ligand binding site model in XLFit for
Microsoft Excel where percent inhibition = 1/[1þ (EC50/drug
concn)m] and m is a parameter that reflects the slope of the
concentration-response curve. The CC50 values were calculated
using the same equation.
C22H21N4O5 0.2H2O: C, 62.17; H, 5.07; N, 13.18. Found: C,
61.92; H, 5.41; 13.01.
3
X-ray Crystal Structures of BMS-377806 (3) and BMS-
488043 (4). Diffraction data were collected at room temperature
using a Bruker SMART 2K CCD diffractometer equipped with
˚
graphite-monochromated Cu KR radiation (λ = 1.540 56 A).
An empirical absorption correction utilized the SADABS rou-
tine (SMART and SAINT-Plus Area Detector Control and
Integration Software; Bruker AXS: Madison, WI, 1998). The
unit cell parameters were determined using the entire data set.
The structures were solved by direct methods and refined by the
full-matrix least-squares techniques, using the SHELXTL soft-
ware package (Sheldrick, G. M. SHELXTL. Structure Determi-
nation Programs, version 5.10; Bruker AXS: Madison, WI,
1997). Non-hydrogen atoms were refined with anisotropic
thermal displacement parameters. Hydrogen atoms involved
in hydrogen bonding were located in the final difference Fourier
maps, while the positions of the other hydrogen atoms were
calculated from an idealized geometry with standard bond
lengths and angles. They were assigned isotropic temperature
factors and included in structure factor calculations with fixed
parameters. Full crystallographic data have been deposited with
the Cambridge Crystallographic Data Center (reference num-
bers CCDC 740492 (3), CCDC 717399 (4)). Copies of the data
can be obtained free of charge via the Internet at http://www.
ccdc.cam.ac.uk.
Acknowledgment. The authors are grateful to Marie
D’Andrea and Elizabeth Bitel (HRMS and CHN data),
Dr. Xiaohua Stella Huang (NMR spectra), Dr. Richard A
Dalterio, Gottfried Wenke, and Lirisa Zueva (pKa and logD
measurements) for their assistance in obtaining analytical
and physicochemical data. In addition, the preclinical candi-
date optimization organization at Bristol-Myers Squibb is
acknowledged for in vitro and in vivo profiling of compounds
3 and 4.
References
(1) Gallo, R. C.; Montagnier, L. The discovery of HIV as the cause of
AIDS. N. Engl. J. Med. 2003, 349, 2283–2285.
(2) Pomerantz, R. J.; Horn, D. L. Twenty years of therapy for HIV-1
Antiviral Evaluation. Cells. 293T cells were obtained from the
NIH AIDS Research and Reference Reagent Program, grown in
DMEM with 10% heat inactivated FBS, 100 U/mL penicillin,
and 100 μg/mL streptomycin and subcultured twice a week. The
HeLa cell line expressing CD4, CCR5, and CXCR4 (HeLa 67
cells) was constructed at Bristol-Myers Squibb and maintained
in the medium described above, supplemented with 400 μg/mL
infection. Nat. Med. 2003, 9, 867–873.
(3) Barre-Sinoussi, F. The early years of HIV research: integrating
ꢀ
clinical and basic research. Nat. Med. 2003, 9, 844–846.
(4) Merson, M. H. The HIV-AIDS pandemic at 25;the global
response. N. Engl. J. Med. 2006, 354, 2414–2417.
(5) Richman, D. D.; Morton, S. C.; Wrin, T.; Hellmann, N.; Berry, S.;
Shapiro, M. F.; Bozzette, S. A. The prevalence of antiretroviral
drug resistance in the United States. AIDS 2004, 18, 1393–1401.

7786 Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23
Wang et al.
(6) Wainberg, M. A.; Brenner, B. G.; Turner, D. Changing patterns in
the selection of viral mutations among patients receiving nucleo-
side and nucleotide drug combinations directed against human
immunodeficiency virus type 1 reverse transcriptase. Antimicrob.
Agents Chemother. 2005, 49, 1671–1678.
(7) Vella, S.; Palmisano, L. The global status of resistance to antire-
troviral drugs. Clin. Infect. Dis. 2005, 41 (Suppl. 4), S239–S246.
(8) Shafer, R. W.; Schapiro, J. M. Drug resistance and antiretroviral
drug development. J. Antimicrob. Chemother. 2005, 55, 817–820.
(9) Little, S. J.; Holte, S.; Routy, J.-P.; Daar, E. S.; Markowitz, M.;
Collier, A. C.; Koup, R. A.; Mellors, J. W.; Connick, E.; Conway,
B.; Kilby, M.; Wang, L.; Whitcomb, J. M.; Hellman, N. S.;
Richman, D. D. Antiretroviral drug resistance among patients
recently infected with HIV. N. Engl. J. Med. 2002, 347, 385–394.
(10) Turner, D.; Wainberg, M. A. HIV transmission and primary drug
resistance. AIDS Rev. 2006, 8, 17–23.
(11) De Clercq, E. Anti-HIV drugs: 25 compounds approved within 25
years after the discovery of HIV. Int. J. Antimicrob. Agents 2009,
33, 307–320.
(12) Doms, R. W.; Trono, D. The plasma membrane as a combat zone
in the HIV battlefield. Genes Dev. 2000, 14, 2677–2688.
(13) Pierson, T. C.; Doms, R. W. HIV-1 entry inhibitors: new targets,
novel therapies. Immunol. Lett. 2003, 85, 113–118.
that blocks gp120-CD4 interactions. J. Virol. 2003, 77, 10528–
10536.
(30) Ho, H.-T.; Fan, L.; Nowicka-Sans, B.; McAuliffe, B.; Li, C.-B.;
Yamanaka, G.; Zhou, N.; Fang, H.; Dicker, I.; Dalterio, R.; Gong,
Y.-F.; Wang, T.; Yin, Z.; Ueda, Y.; Matiskella, J.; Kadow, J. F.;
Clapham, P.; Robinson, J.; Colonno, R. J.; Lin, P.-F. Envelope
conformational changes induced by human immunodeficiency
virus type 1 attachment inhibitors prevent CD4 binding and down-
stream entry events. J. Virol. 2006, 80, 4017–4025.
(31) Si, Z.; Madani, N.; Cox, J. M.; Chruma, J. J.; Klein, J. C.; Schon,
A.; Phan, N.; Wang, W.; Biorn, A. C.; Cocklin, S.; Chaiken, I.;
Freire, E.; Smith, A. B.; Sodroski, J. G. Small-molecule inhibitors
of HIV-1 entry block receptor-induced conformational changes in
the viral envelope glycoproteins. Proc. Natl. Acad. Sci U.S.A. 2004,
101, 5036–5041.
(32) Madani, N.; Perdigoto, A. L.; Srinivasan, K.; Cox, J. M.; Chruma,
J. J.; LaLonde, J.; Head, M.; Smith, A. B., III; Sodroski, J. G.
Localized changes in the gp120 envelope glycoprotein confer
resistance to human immunodeficiency virus entry inhibitors
BMS-806 and #155. J. Virol. 2004, 78, 3742–3752.
(33) Hanna, G.; Lalezari, J.; Hellinger, J.; Wohl, D.; Masterson, T.;
Fiske, W.; Kadow, J. F.; Lin, P.-F.; Giordano, M.; Colonno, R. J.;
Grasela, D. Antiviral Activity, Safety, and Tolerability of a Novel,
Oral Small-Molecule HIV-1 Attachment Inhibitor, BMS-488043,
in HIV-1-Infected Subjects. Presented at the 11th Conference
on Retroviruses Opportunistic Infections, San Francisco, CA,
February 8-11, 2004; Abstract 141.
(14) Kilby, J. M.; Eron, J. J. Novel therapies based on mechanisms of
HIV-1 cell entry. N. Engl. J. Med 2003, 348, 2228–2238.
(15) Meanwell, N. A.; Kadow, J. F. Inhibitors of the entry of HIV into
host cells. Curr. Opin. Drug Discovery Dev. 2003, 6, 451–461.
(16) Pierson, T. C.; Doms, R. W.; Pohlmann, S. Prospects of HIV-1
entry inhibitors as novel therapeutics. Rev. Med. Virol. 2004, 14,
255–270.
ꢀ
~
(34) Catalan, J.; Mo, O.; Perez, P.; Yanez, M. Influence of the tauto-
meric forms of azaindoles on their basicity in solution. J. Mol.
Struct. 1984, 107, 263–368.
(35) Hlavica, P. Functional interaction of nitrogenous organic bases
with cytochrome P450: a critical assessment and update of sub-
strate features and predicted key active-site elements steering the
access, binding, and orientation of amines. Biochim. Biophys. Acta
2006, 1764, 645–670.
(36) Rendic, S. Summary of information on human CYP enzymes:
human P450 metabolism data. Drug Metab. Rev. 2002, 34, 83–448.
(37) Tehan, B. G.; Lloyd, E. J.; Wong, M. G.; Pitt, W. R.; Gancia, E.;
Manallack, D. T. Estimation of pKa using semiempirical molecular
orbital methods. Part 2: application to amines, anilines and various
nitrogen containing heterocyclic compounds. Quant. Struct.-Act.
Relat. 2002, 21, 473–485.
(38) Perrin, D. D. Dissociation Constants of Organic Bases in Aqueous
Solution; Butterworths: London, 1965; Supplement to Pure Appl.
Chem.
(39) Zhang, Z.; Yang, Z.; Meanwell, N. A.; Kadow, J. F.; Wang, T. A
general method for the preparation of 4- and 6-azaindoles. J. Org.
Chem. 2002, 67, 2345–2347.
(40) Van de Poel, H.; Guillaumet, G.; Viaud-Massuard, M.-C. Synth-
esis of melatonin analogues derived from furo-[2,3-b]- and [2,3-c]-
pyridines by use of a palladium-copper catalyst system. Hetero-
cycles 2002, 57, 55–71.
(41) Yamato, T.; Hideshima, C.; Tashiro, M. A convenient preparation
and inclusion behavior of 1,3,5-tris-(hydroxyphenyl)benzenes.
Chem. Express 1990, 5, 845–848.
(42) Zhang, Z.; Yang, Z.; Wong, H.; Zhu, J.; Meanwell, N. A.; Kadow,
J. F.; Wang, T. An effective procedure for the acylation of
azaindoles at C-3. J. Org. Chem. 2002, 67, 6226–6227.
(43) Li, H.; Jiang, X.; Ye, Y.-h.; Fan, C.; Romoff, T.; Goodman, M.
3-(Diethoxyphosphoryloxy)-1,2,3-benzotriazin-4(3H)-one
(DEPBT): a new coupling reagent with remarkable resistance to
racemization. Org. Lett. 1999, 1, 91–93.
(17) Wang, H.-G. H.; Kadow, J. F.; Lin, P.-F. HIV gp120 envelope as a
therapeutic target. Drugs Future 2005, 30, 359–367.
(18) Castagna, A.; Biswas, P.; Beretta, A.; Lazzarini, A. The appealing
story of HIV entry inhibitors. Drugs 2005, 65, 879–904.
(19) Vermiere, K.; Schols, D. Anti-HIV agents targeting the interaction
of gp120 with the cellular CD4 receptor. Expert Opin. Invest. Drugs
2005, 14, 1199–1212.
(20) Kadow, J. F.; Wang, H.-G. H.; Lin, P.-F. Small-molecule HIV-1
gp120 inhibitors to prevent HIV-1 entry: an emerging opportunity
for drug development. Curr. Opin. Invest. Drugs 2006, 7, 721–726.
ꢀ
(21) Este, J. A.; Telenti, A. HIV entry inhibitors. Lancet 2007, 370, 81–
88.
(22) Schon, A.; Freire, E. Strategies for targeting the HIV-1 envelope
glycoprotein gp120 in the development of new antivirals. Future
HIV Ther. 2007, 1, 223–229.
(23) Strizki, J. Targeting HIV attachment and entry for therapy. Adv.
Pharmacol. 2008, 56, 93–120.
(24) Qian, K.; Morris-Natschke, S. L.; Lee, K.-H. HIV entry inhibitors
and their potential in HIV therapy. Med. Res. Rev. 2009, 29, 369–
393.
(25) Wang, H.-G. H.; Williams, R. E.; Lin, P.-F. A novel class of HIV-1
inhibitors that targets the viral envelope and inhibits CD4 receptor
binding. Curr. Pharm. Des. 2004, 10, 1785–1793.
(26) Wang, T.; Zhang, Z.; Wallace, O. B.; Deshpande, M.; Fang, H.;
Yang, Z.; Zadjura, L .M.; Tweedie, D. L.; Huang, S.; Zhao, F.;
Ranadive, S.; Robinson, B. S.; Gong, Y.-F.; Ricarrdi, K.; Spicer,
T. P.; Deminie, C.; Rose, R.; Wang, H.-G. H.; Blair, W. S.; Shi,
P.-Y.; Lin, P.-F.; Colonno, R. J.; Meanwell, N. A. Discovery of
4-benzoyl-1-[(4-methoxy-1H- pyrrolo[2,3-b]pyridin-3-yl)oxoacetyl]-
2-(R)-methylpiperazine (BMS-378806): a novel HIV-1 attachment
inhibitor that interferes with CD4-gp120 interactions. J. Med. Chem.
2003, 46, 4236–4239.
(44) Tobyn, M.; Brown, J.; Dennis, A. B.; Fakes, M.; Gao, Q.; Gamble,
J.; Khimyak, Y. Z.; McGeorge, G.; Patel, C.; Sinclair, W.;
Timmins, P. Amorphous drug-PVP dispersions: application of
theoretical, thermal and spectroscopic analytical techniques to the
study of a molecule with intermolecular bonds in both the crystal-
line and pure amorphous state. J. Pharm. Sci. 2009, 98, 3456–3468.
(45) Brameld, K. A.; Kuhn, B.; Reuter, D. C.; Stahl, M. Small molecule
conformational preferences derived from crystal structure data. A
medicinal chemistry focused analysis. J. Chem. Inf. Model. 2008,
48, 1–24.
(46) Johnson, F. Allylic strain in six-membered rings. Chem. Rev. 1968,
68, 375–413.
(47) Chevallier, F.; Beaudet, I.; Grognec, E. L.; Toupet, L.; Quintard,
J.-P. Allylstannation of N-acyliminium intermediates: a possible
method for the stereocontrolled synthesis of polyhydroxypiperi-
dines. Tetrahedron Lett. 2004, 45, 761–764.
(27) Meanwell, N. A.; Wallace, O. B.; Fang, H.; Wang, H.; Deshpande,
M.; Wang, T.; Yin, Z.; Zhang, Z.; Pearce, B. C.; James, J.; Yeung,
K.-S.; Qiu, Z.; Wright, J. J. K.; Yang, Z.; Zadjura, L.; Tweedie,
D. L.; Yeola, S.; Zhao, F.; Ranadive, S.; Robinson, B. A.; Gong,
Y.-F.; Wang, H.-G. H.; Blair, W. S.; Shi, P.-Y.; Colonno, R. J.; Lin,
P.-F. Inhibitors of HIV-1 attachment. Part 2: An initial survey of
indole substitution patterns. Bioorg. Med. Chem. Lett. 2009, 19,
1977–1981.
(28) Lin, P.-F.; Blair, W. S.; Wang, T.; Spicer, T. P.; Guo, Q.; Zhou, N.;
Gong, Y.-F.; Wang, H.-G. H.; Rose, R.; Yamanaka, G.;
Robinson, B.; Li, C.-B.; Fridell, R.; Deminie, C.; Demers, G.;
Yang, Z.; Zadjura, L.; Meanwell, N. A.; Colonno, R. J. A small
molecule HIV-1 inhibitor that targets the HIV-1 envelope and
inhibits CD4 receptor binding. Proc. Natl. Acad. Sci. U.S.A. 2003,
100, 11013–11018.
(29) Guo, Q.; Ho, H.-T.; Dicker, I.; Fan, L.; Zhou, N.; Friborg, J.;
Wang, T.; McAuliffe, B. V.; Wang, H.-G. H.; Rose, R. E.; Fang,
H.; Scarnati, H. T.; Langley, D. R.; Meanwell, N. A.; Abraham, R.;
Colonno, R. J.; Lin, P.-F. Biochemical and genetic characteriza-
tions of a novel human immunodeficiency virus type 1 inhibitor
(48) Wang, T.; Kadow, J. F.; Zhang, Z.; Yin, Z.; Gao, Q.; Wu, D.;
DiGiugno Parker, D.; Yang, Z.; Zadjura, L.; Robinson, B. A.;
Gong, Y.-F.; Blair, W. S.; Shi, P.-Y.; Yamanaka, G.; Lin, P.-F.;
Meanwell, N. A. Inhibitors of HIV-1 attachment. Part 4: A study of

Article
the effect of piperazine substitution patterns on antiviral potency in
the context of indole-based derivatives. Bioorg. Med. Chem. Lett.
Journal of Medicinal Chemistry, 2009, Vol. 52, No. 23 7787
(52) Fakes, M. G.; Vakkalagadda, B. J.; Qian, F.; Desikan, S.; Gandhi,
R. B.; Lai, C.; Hsieh, A.; Franchini, M. K.; Toale, H.; Brown, J.
Enhancement of oral bioavailability of an HIV-1 attachment
inhibitor by nanosizing and amorphous dispersion approaches.
Int. J. Pharm. 2009, 370, 167–174.
(53) Ueda, Y.; Connolly, T. P.; Kadow, J. F.; Meanwell, N. A.;
Wang, T.; Chen, C.-P. H.; Yeung, K.-S.; Zhang, Z.; Leahy,
D. K.; Pack, S. K.; Soundararajan, N.; Sirard, P.; Levesque, K.;
Thoraval, D. Prodrugs Based on Bicyclic Nitrogen-Contain-
ing Heterocyclic Antiviral Agents Having Substituted Piperazine
or Piperidine Rings. U.S. Pat. Appl. Publ. US 2005209246 A1,
2005.
2009, 19, 5140–5145.
(49) Lin, P.-F; Ho, H.-T.; Gong, Y.-F.; Dicker, I.; Zhou, N.; Fan, L.;
McAuliffe, B.; Kimmel, B.; Nowicka-Sans, B.; Wang, T.; Kadow,
J. F.; Yamanaka, G.; Yin, Z.; Yang, Z.; Zadjura, L.; Meanwell,
N. A.; Colonno, R. J. Characterization of a Small Molecule HIV-1
Attachment Inhibitor BMS-488043: Virology, Resistance and
Mechanism of Action. Presented at the 11th Conference on Retro-
viruses Opportunistic Infections, San Francisco, CA, February
8-11, 2004; Abstract 534.
(50) Yang, Z.; Zadjura, L.; Marino, A.; D’Arienzo, C.; Malinowski, J.;
Gesenberg, C.; Lin, P.-F.; Colonno, R. J.; Wang, T.; Kadow, J. F.;
Meanwell, N. A.; Hansel, S. Utilization of in vitro caco-2 perme-
ability and microsomal half-life screens in discovering BMS-
488043, a novel HIV-1 attachment inhibitor with improved phar-
macokinetic properties. Unpublished results.
(54) Johnson, V. A.; Byington, R. T. Infectivity Assay (Virus Yield
Assay). Techniques in HIV Research; New York, 1990.
(55) Cory, A. H.; Owen, T. C.; Barltrop, J. A.; Cory, J. G. Use of an
aqueous soluble tetrazolium/formazan assay for cell growth assays
in culture. Cancer Commun. 1991, 3, 207–212.
(51) Hanna, G.; Yan, J.-H.; Fiske, W.; Masterson, T.; Zhang, D.;
Grasela, D. Safety, Tolerability, and Pharmacokinetics of a Novel
Small-Molecule of HIV-1 Attachment Inhibitor, BMS-488043,
after Single and Multiple Oral Doses in Healthy Subjects.
Presented at the 11th Conference on Retroviruses Opportunistic
Infections, San Francisco, CA, February 8-11, 2004; Abstract 535.
(56) Yang, Z.; Zadjura, L. M.; D’Arienzo, C.; Marino, A.; Santone, K.;
Klunk, L.; Greene, D.; Lin, P.-F.; Colonno, R. J.; Wang, T.;
Meanwell, N. A.; Hansel, S. Preclinical pharmacokinetics of a
novel HIV-1 attachment inhibitor BMS-378806 and prediction of
its human pharmacokinetics. Biopharm. Drug Dispos. 2005, 26,
387–402.