BMS-955176 is a second-generation maturation inhibitor developed to improve upon the first generation maturation inhibitor bevirimat. BMS-955176 was designed through structure-activity relationships to maintain potency against viruses with gag polymorphisms associated with reduced susceptibility to bevirimat. BMS-955176 inhibits HIV protease cleavage of the gag polyprotein, binds tightly to HIV gag, and maintains antiviral activity against a diverse panel of HIV subtypes and drug resistant isolates. Phase I/II clinical trials demonstrated BMS-955176 monotherapy resulted in over a 1 log10 decline in HIV viral load. BMS-955176 has potential as an antiretroviral with a novel mechanism of action.

1. SUMMARY
Background: BMS-955176 is a second-generation HIV-1 maturation inhibitor (MI). A first-generation MI, bevirimat, showed
efficacy in early-phase studies, but ~50% of subjects had virus with reduced susceptibility, associated with naturally
occurring Gag polymorphisms. Assays designed to optimize target specificity, virologic potency, polymorphic coverage, and
human serum binding were used to identify an improved MI clinical candidate.
Methods: BMS-955176 inhibition of Gag cleavage in HIV-1-infected cells and specific binding to Gag in virus-like particles
(VLPs) was used to assess MI targeting. Potency was optimized using a panel of engineered reporter viruses containing
polymorphic changes in Gag that reduce susceptibility to bevirimat (including V362I, Q369H, V370A/M/Δ and T371A/Δ).
Candidates were then evaluated against a library of recombinant viruses containing gag/Pr genes from clinical isolates
to assess their spectra of activity. Further, BMS-955176 activity was also tested against a series of clinical isolates in
peripheral blood mononuclear cells (PBMCs) and a panel of antiretroviral-resistant viruses. Serum effect experiments were
performed in 40% human serum/10% fetal bovine serum (FBS) + 27 mg/mL human serum albumin (HuSA) vs. 10% FBS.
Results: BMS-955176 inhibits HIV-1 protease cleavage at the CA/SP1 junction within Gag in HIV-1-infected cells, and
binds tightly and reversibly to Gag in purified HIV-1 VLPs. The average antiviral EC50
was 3.9 ± 3.4 nM toward a library
of 87 gag/Pr recombinant subtype B viruses containing 96% of subtype B polymorphic Gag diversity near the CA/SP1
cleavage site. Seventy six clinical isolates of HIV-1 subtypes A, AE, B, C, D, F and G evaluated in PBMCs exhibited
EC50
s of 0.001–1.5 µM. Activity was maintained against a panel of reverse transcriptase-, protease- and integrase-inhibitor-
resistant viruses, with EC50
s similar to wild-type. Average human serum binding was 86±0.4%; a 5.4-fold reduction in
EC50
occurred in the presence of 40% human serum/10% FBS + 27 mg/mL HuSA.
Conclusions: BMS-955176 is a second-generation MI that inhibits HIV-1 protease cleavage at the CA/SP1 junction within
Gag. BMS-955176 has potent in vitro anti-HIV-1 activity in the presence of a range of Gag polymorphisms associated with
reduced susceptibility to a first-generation MI, broad coverage of HIV-1 subtypes, and low human serum binding. The antiviral
activity of BMS-955176 was also investigated in a Phase IIa trial (AI468002), where 10-day BMS-955176 monotherapy (doses
of 5–120 mg QD) in HIV-1 subtype B-infected subjects resulted in median declines in HIV-1 RNA of >1 log10
c/mL (40–120 mg
QD doses). This confirmed successful application of the preclinical development strategy to identify a second-generation MI with
improved polymorphic coverage and reduced serum effect. These data support further clinical development of BMS-955176.
BACKGROUND
■■ Life-long management of HIV-1 infection requires sequential combination antiretroviral (ARV) therapy,
preferably with simple and convenient regimens containing a minimum of three fully active agents.1
■■ ARV treatment options, particularly for treatment-experienced (TE) patients, may be limited due to
treatment-emergent or transmitted resistance, adverse events, drug-drug interactions (DDIs), and regimen
complexity.1,2
■■ Novel ARVs are needed that target different steps of the viral life cycle, have unique resistance profiles,
and display long-term tolerability with manageable DDIs.
■■ HIV-1 maturation is a crucial step of the viral life cycle that leads to the formation of morphologically
mature, infectious virus particles.3
■■ Maturation inhibitors (MIs) inhibit the last protease (Pr) cleavage event between capsid (CA) protein p24
and spacer peptide 1 (SP1) in HIV-1 Gag polyprotein, resulting in the release of immature, non-infectious
virions (Figure 1).3,4
Figure 1: Mode of action of MIs
Development of a second-generation MI: insights from a first-
generation MI, bevirimat (BVM)
■■ The development of a first-generation MI, BVM, was terminated owing to inadequate coverage of
polymorphic Gag variants represented in ~50% of the general HIV-1 population (Figure 2).5–7
■■ High protein binding of BVM (>99%)8
limited the available free fraction necessitating a high dose.
Additionally, poor bio-pharmaceutical properties required a demanding, but ultimately unsuccessful,
oral formulation9
to achieve once-daily (QD) target exposures.
Gag polyprotein
Naturally occurring polymorphisms associated with
reduced susceptibility to BVM are highlighted in yellow
CA SP1
R V L A E A M S Q V T N S A T I M M Q R
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
NC
Figure 2: Naturally occurring polymorphisms associated with reduced susceptibility
to BVM7,10,11
BVM, bevirimat; CA, capsid; MA, matrix protein; NC, nucleocapsid; P6, protein 6; SP1, spacer peptide 1; SP2, spacer peptide 2.
Strategy for development of a second-generation MI with improved
polymorphic coverage
Use structure–activity relationship (SAR) to develop a clinical candidate with antiviral activity toward a
panel of polymorphic viruses with site-directed, BVM-resistance-associated,
Gag polymorphisms; optimize for potency, low human serum binding and QD dosing12,13
Confirm mode of action retains specificity for inhibition of viral maturation
both in infectious virus and at the biochemical level12
Confirm in vitro coverage by evaluation of candidate activity toward a library of 87
subtype B gag/Pr recombinant viruses (derived from clinical isolates) in MT-2 cells
Confirm in vitro coverage by evaluation of a diverse panel of clinical isolates evaluated in
peripheral blood mononuclear cells (PBMCs)
METHODS
Viruses and cells
■■ Recombinant lab-adapted HIV-1NL4-3
containing the Renilla luciferase (Rluc) reporter gene in the
nef locus (NL4-3
RepRluc) was used. Laboratory HIV-1 strains, ARV-resistant viruses (apart from the
non-nucleos(t)ide reverse transcriptase inhibitors [NNRTI], integrase strand transfer inhibitor (INI[STI]) and
ritonavir-resistant viruses), and clinical isolates (except those obtained from BMS-sponsored clinical trials)
were obtained from the NIH AIDS Research and Reference Reagent Program. Other ARV-resistant viruses
were created through site-directed mutagenesis in the NL4-3
RepRluc virus.
■■ Recombinant viruses with cloned gag/Pr sequences were generated by subcloning gag/Pr genes from
clinical isolates (subtype B) into NL4-3
RepRluc.
■■ Virus-like particles (VLPs) were prepared by transfection of HEK 293T cells with full length Gag DNA.
Assays
■■ All laboratory strains (wild-type [WT] or recombinant) were used to infect MT-2 cells, while purified PBMCs were
used as host cells for clinical isolates. Antiviral activity of viruses containing the Rluc reporter was quantified by
luciferase activity; others by either reverse transcriptase activity or yield of the p24 antigen in supernatants.
■■ MI dissociation half-lives were determined by pre-binding of radiolabeled MI to HIV Gag VLP, followed by
cold chase with excess unlabeled MI.12
■■ Serum effects were assessed in 10% fetal bovine serum (FBS) + 40% human serum supplemented
with 27 mg/mL human serum albumin vs. 10% FBS. Percentage serum binding was determined by an
ultracentrifugation technique.
■■ The cleavage of the CA precursor, p25, in cells was quantified by Western analysis of cells expressing
HIV-1 in the presence or absence of BMS-955176, upon prior transfection.
■■ p25 cleavage in VLPs was determined by incubation of HIV-1 Gag VLPs with/without BMS-955176 and
sequential cleavage with HIV-1 Pr and then trypsin. Peptides were quantified using an LC/MS method.12
RESULTS
Progression to clinical candidate, BMS-95517612
■■ SAR progression shows increased polymorphic coverage and decreased serum shift, leading to
identification of BMS-955176 (Table 1).
■■ BMS-955176 exhibits no substantial reduction in activity towards single key site-directed mutants (SDMs)
with BVM-resistance associated polymorphisms (Table 1).
■■ The lower serum binding of BMS-955176, compared to BVM, increases the free fraction of drug, which
facilitates QD dosing.
Table 1: Progression to clinical candidate, BMS-955176: evaluation against
SDMs with key BVM resistance-associated polymorphisms11
EC50
≤13 nM and FC EC50
≤7
Lower serum shift, increasing polymorphic coverage
BVM BMS-A/B/C
BMS-
955176
EC50
, nM FC-EC50
Virus
(HIV-1NL4-3
)
Subtype B, %
LANL DB*
BVM BMS-A BMS-B BMS-C
BMS-
955176
BVM
BMS-
955176
Serum shift†
130 7.0 9.3 21 5.4 - -
WT 51 10 15 15 2 1.9 - -
WT + serum‡
- 1300 105 139 43 10.3 - -
SDMsinNL4-3
background
V362I 12 74 210 77 18 4.5 7.4 2.4
Q369H 2.4 8 n.d. n.d. n.d. 2.0 0.8 1.0
V370A 15 552 233 19 8 2.7 55 1.4
V370M 5 111 4000 28 10 2.8 11 1.5
ΔV370§
0.6 4000 4000 415 31 13 400 6.8
V370A + ΔT371§
1.9 4000 4000 200 n.d. 6.6 400 3.5
T371A 5 10 28 11 3 3.0 1.0 1.5
ΔT371 3 77 292 38 3 4.7 7.7 2.5
Multiple cycle assay in MT-2 cells. * Percentage of subtype B isolates in the Los Alamos database (2010, N=1754).
†
Ratio of EC50
in serum condition versus no added serum ‡
Assay conducted in the presence of 40% human serum + 27 mg/mL human
serum albumin. §
Surrogates for clade C. BVM, bevirimat; FC, fold change; n.d., not determined; LANL, Los Alamos National Laboratory;
SDM, site-directed mutants; WT, wild-type.
Effect of human serum on BMS-955176 potency
■■ A 5.4-fold serum shift in BMS-955176 EC50
was observed, compared with 100-fold serum shift for
BVM (Table 1).
■■ BMS-955176 was 86±0.4% (n=4) bound to human serum proteins.
■■ Low serum shift/protein binding increases free fraction of drug, which facilitates QD dosing.
BMS-955176 is an MI that binds specifically and reversibly to HIV-1
Gag, with a slow dissociation rate
■■ BMS-955176 blocks cleavage of CA precursor p25 to CA p24 in HIV-1-infected cells (Figure 3a) and in
VLPs (Figure 3b), but not in HIV-1-infected cells or VLPs with the A364V substitution12
(confers high-level
resistance to BVM and BMS-955176).
■■ [3
H]BMS-955176 binds specifically and reversibly to HIV-1 VLPs12
(Figure 3c).
■■ BMS-955176 dissociates from BVM-resistant Gag polymorphs (V370A shown as an example in Figure 3d)
more slowly than BVM, correlating with its improved antiviral activity.12
p25
p24
WT A364V
BMS-955176 – –+ +
Countspermillion
Kd
= 5.5 nM
1000
800
600
400
200
0
0 100 200
[3
H]–BMS–955176 (nM)
100
50
0
0 1 2 3 4 5
WT
A364V
Percentageinhibition
ofp25cleavage
176
BVM
Time (hours)
4002000
0
50
100
V370A
Percentagedrugbound
Time (minutes)
a b
c d
Figure 3: Biochemical characterization of BMS-955176, a) cleavage of HIV-1 CA
precursor p25 to p24 in HIV-1-infected cells; b) inhibition of p25 cleavage in VLPs;
c) binding of radiolabelled BMS-955176 to HIV-1 Gag in WT VLPs; d) dissociation of
BMS-955176 and BVM from HIV-1 V370A Gag p25
BVM, bevirimat; Kd
, dissociation constant; VLP, virus-like particle.
Progression of SAR leading to BMS-955176: analysis of a panel of
recombinant subtype B viruses with gag/Pr genes derived from
clinical isolates
■■ Progression of SAR demonstrates sequential improvements in the coverage of polymorphic Gag variants,
leading to clinical candidate BMS-955176 (Table 2).
—— BMS-955176 shows broad activity spectrum toward a library of 87 gag/Pr recombinant subtype B
viruses (gag/Pr genes from clinical isolates).
Table 2: Progression of SAR leading to BMS-955176: analysis of activity toward a
panel of recombinant subtype B viruses with gag/Pr genes from clinical isolates
EC50
≤13 nM EC50
, nM EC50
, nM
Key single
polymorphisms
Na
BVMa
BMS-Ba
BMS-Ca
N BMS-955176
WT†
21 14 14 3 51 3
V362I* 3 36 29 6 5 5
Q369H 1 4 7 15 1 2
V370I/L/T 3 251 26 10 4 3
V370A 3 227 38 12 7 2
V370M 5 903 76 67 9 7
T371A/Q/N/S/TT 4 113 49 7 6 4
V370A + ΔT371 3 4000 629 236 4 12
* Secondary polymorphisms under study. †
No change at Gag amino acid positions V362, 369, V370 or T371. a
Sub-library N = 43.
BVM, bevirimat; LANL, Los Alamos National Laboratory; SAR, structural–activity relationships; WT, wild-type.
BMS-955176 is active against a panel of ARV-resistant
HIV-1 isolates
■■ BMS-955176 is active towards a range of mutations that confer reduced sensitivity to other classes of ARV
agents, including NRTIs, NNRTIs, PIs and INI(STI)s (Table 3).
Table 3: BMS-955176 is active against a panel of ARV-resistant HIV-1 isolates
ARV-resistant mutants
NRTI
(M184V)
NRTI
(41L, 67N,
215F, 219E,
69N)
NNRTI
(K103N,
Y181C)
PI14
(46I, 82F,
84V, 90M)
PI*
(54V, 82A
84V)
INI(STI)
(140S, 148H)
ARV Class Fold-change in EC50
relative to parental strain
BMS-955176 MI 1.1 1.9 1.0 0.9 1.4 0.7
Lamivudine NRTI 60 6.8 1.1 1.3 n.d. 1.4
Zidovudine NRTI 0.78 75 1.0 1.6 n.d. 0.82
Nevirapine NNRTI 1.3 1.6 131 0.83 n.d. 0.89
Efavirenz NNRTI 1.6 n.d. 58 0.29 1.4 0.84
Rilpivirine NNRTI 1.8 n.d. 6.6 n.d. 0.9 n.d.
Darunavir PI 1.8 n.d. 0.85 42 2.3 1.0
Atazanavir14
PI n.d. n.d. 1.6 3814
39 n.d.
Raltegravir INI(STI) 1.9 1.6 1.4 0.21 n.d. 350
* gag/Pr gene from PI-resistant virus (provided by Dr. Monique Nijhuis) transferred to NL4-3
RepRluc backbone; other viruses prepared as
site-directed mutants in NL4-3
RepRluc backbone. ARV, antiretroviral; INI(STI), integrase strand transfer inhibitor; NNRTI, non-nucleoside
reverse transcriptase inhibitor; n.d., not determined; NRTI, nucleoside reverse transcriptase inhibitor; PI, protease inhibitor.
BMS-955176 is active against a broad range of HIV-1 subtypes
■■ BMS-955176 shows antiviral activity toward a panel of HIV-1 clinical isolates representing two major
HIV-1 groups (Figure 4a):
—— Group M including subtypes A, AE, B, C, D, F and G,
—— Group N.
■■ As expected mechanistically, BMS-955176 was active against both CCR5- and CXCR4-tropic laboratory
strains of HIV-1 (Figure 4b).
10
1
0.1
0.01
0.001
0.0001
A
AE
B
C
D
F
G
N
O
Group
Group
Group M
BMS–955176EC50
,µM
(interquartilerange)
a)
b) HIV-1 strain Co-receptor usage BMS-955176 EC50
, nM (± SD)
IIIB CXCR4 2.8±0.3
LAI CXCR4 2.6±0.3
Bal CCR5 0.8
JRFL CCR5 1.4
Bars are medians with interquartiles.
Figure 4: Activity of BMS-955176 toward, a) a panel of HIV-1 clinical isolates;
b) CXCR4- and CCR5-tropic laboratory strains of HIV-1
Antiviral activity of BMS-955176 in a Phase IIa, randomized,
multi-part trial (AI468002)13
■■ The preclinical data supported further development of BMS-955176, and clinical data from the AI468002
study confirm the successful application of the preclinical development strategy to identify a
second-generation MI with improved polymorphic coverage and reduced serum effect.
■■ In AI468002, during 10-day BMS-955176 monotherapy (doses of 5–120 mg QD) in HIV-1 subtype
B-infected subjects, median declines in HIV-1 RNA of 1 log10
c/mL were observed at the 40–120 mg
doses from ~Day 7 onwards (Figure 5).13
-1.8
-1.6
-1.4
-1.2
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14// 17// 19// 24 25
Placebo
5 mg
10 mg
20 mg
40 mg
80 mg
120 mg
Study days
MedianchangeinHIV-1RNAfrom
baseline,log10
copies/mL
Dosing period
Figure 5: BMS-955176: median change in HIV-1 RNA over time in the
Phase IIa study13
Baseline polymorphisms at Gag V362, A364, Q369, and V370 were evaluated, but no baseline polymorphisms at position 364 were
present in study.
CONCLUSIONS
■■ BMS-955176 is a second-generation MI that binds tightly and reversibly to HIV-1 Gag, blocking
cleavage of CA precursor p25 to capsid p24.12
■■ BMS-955176 displays greater potency and improved coverage of Gag polymorphisms compared with
a first-generation MI, and:12,13
—— dissociates slowly from Gag, translating to increased antiviral potency,12
—— shows low human serum binding (86%) resulting in a low EC50
shift in the presence of serum (5.4x),
—— displays broad spectrum antiviral activity against a library of polymorphic subtype B viruses,
representative of Gag diversity near the CA/SP1 cleavage site,
—— is active against a diverse panel of HIV-1 clinical isolates in PBMCs and a panel of HIV-1
isolates resistant to other classes of ARVs.
■■ Successful application of the preclinical development strategy was demonstrated following clinical
analysis of BMS-955176 activity in a 10-day BMS-955176 monotherapy study where median declines
in HIV-1 RNA of 1 log10
c/mL were observed at ~Day 7 (40–120 mg doses), and BMS-955176 was
shown to be generally safe and well tolerated.13
■■ These results support the further clinical development of BMS-955176; Phase IIb trials investigating
BMS-955176 in a dose-finding study in treatment-naïve patients and in a booster- and
nucleot(s)ide-sparing regimen in TE patients are underway.
ACKNOWLEDGMENTS
■■ We would like to thank all of the AI468002 clinical trial participants and their families.
■■ Special thanks to: Umesh Hanumegowda, Dawn Parker, Dirk Schürmann, Christian Sobotha, Susan Jenkins, Max Lataillade,
Carey Hwang, Heather Sevinsky, Neela Ray.
■■ Bristol-Myers Squibb: Palanikumar Ravindran, Hong Xiao, Dennis Grasela, Richard Bertz, Phyllis Chan, Albert DelMonte, Michael Child,
Yash Gandhi, Samit Joshi, Zheng Liu, Matthew Healy, Philip Ross, Anupama Sheoran, Varsha Chhatre, Yuan Tian, Yang Hong,
Joseph Cantone, Dieter Drexler, Jacob Swidorski, Bo Ding, Brian Mcauliffe, Xiuyang Guo, Todd Correll, and Eric Y Wong.
■■ Formerly at Bristol-Myers Squibb: Brian Terry, Himadri Samanta.
■■ Charité Research Organisation GmbH: Andreas Hüser and Anke Schulze.
■■ Professional medical writing and editorial assistance was provided by Sharmin Naaz at MediTech Media and 3D graphics were developed
by Olly Venning at Nucleus Global – both funded by Bristol-Myers Squibb.
REFERENCES
1. DHHS Panel on Antiretroviral Guidelines for Adults and
Adolescents. 2014. Available at: aidsinfo.nih.gov/contentfiles/
adultandadolescentgl.pdf (accessed Feb 2015).
2. Wittcop L et al. Lancet Infect Dis 2011; 11:363–371.
3. Sundquist W et al. Cold Spring Harb Perspect Med 2012;
2:a006924.
4. Adamson C et al. Expert Opin Ther Targets 2009; 13:895–908.
5. McCallister S et al. Antivir Ther 2008; 13:A10 (Abstract 8).
6. Castillo A et al. Antivir Ther 2006; 11:S37.
7. Adamson C et al. Retrovirology 2010; 7:36–43.
8. Beatty G et al. ICAAC 2005; Abstract LB-27.
9. Myriad Pharmaceuticals, Inc. (2010) Liquid Bevirimat Dosage
Forms for Oral Administration. US 20100216751 A1.
10. Choe S et al. CROI 2008; Abstract 880.
11. Baelen K et al. Antimicrob Agents Chemother 2009; 53:2185–2188.
12. Lin Z et al. CROI 2015; Abstract 42.
13. Lataillade M et al. CROI 2015; Oral Abstract 114LB.
14. Gong Y et al. Antimicrob Agents Chemother 2000; 44:2319–2326.
8th
IAS Conference on HIV Pathogenesis, Treatment and Prevention
BMS-955176: Characterization of a
Second-Generation HIV‑1 Maturation Inhibitor
B Nowicka-Sans, T Protack, Z Lin, Z Li,
S Zhang, N Meanwell, M Cockett,
A Regueiro-Ren, M Krystal, I Dicker
Bristol-Myers Squibb, Research and Development, Wallingford, CT, USA
Coresponding author
Ira Dicker, PhD
Research Fellow
Department of Virology
Bristol-Myers Squibb,
5 Research Parkway, 350M
Wallingford, CT 06492,USA
Email: ira.dicker@bms.com Phone: +1-203-677-7736
Poster:
TUPEA078
Modified from Lataillade et al.
CROI 2015, Abstract 114LB.
→ →