This document describes efforts to discover novel small molecule HIV-1 entry inhibitors. Researchers used field-based virtual screening and bioisosteric replacement to identify potential inhibitors. They tested 50 compounds and found 4 with sub-100nM potency against HIV-1, representing new "core" chemotypes. The most potent compound had an IC50 of 0.6nM. Through de novo design and structure-property analysis, they further improved potency and drug-likeness, identifying additional leads with low nanomolar potency and improved predicted drug-like properties. This work demonstrates an effective structure-based approach to discover new HIV-1 entry inhibitors.

1. Discovery and optimization of novel
small molecule HIV-1 entry inhibitors
using field-based virtual screening and
bioisosteric replacement
Simon Cocklin, Ph.D.

2. HIV-1 Global Burden
Group M viruses are the most widespread, accounting for ~99% of global infections
and can be further subdivided into nine distinct genetic subtypes, or clades. Clade B
is prevalent North America, Western Europe, and Australia. Clades A, C, and D are
most prevalent in Latin America, Africa and Asia.

3. Human Immunodeficiency Virus-1 (HIV-1) Life cycle and Intervention points
Attachment
e.g., sCD4
Fusion
e.g., T-20
Reverse Transcription
e.g., NRTIs
Integration
Translation
Transcription
Budding and Maturation
Viral Assembly
e.g., integrase
inhibitors
e.g., protease
inhibitors
Attachment of virus to cells
Fusion
Reverse Transcription
Integration
Transcription
Translation
Virus assembly
Budding and Maturation
In development – disrupts viral attachment to cellular receptors
In clinical use – disrupts the processes involved in viral and cell
membrane fusion
In clinical use – disrupts the process of changing the RNA
genome to DNA
In clinical use– disrupts the process of changing the RNA genome
to DNA
In clinical use – disrupts the HIV-1 protease maturation of viral
gene products

4. A. The envelope protein complex is a heterotrimer of three gp41 molecules
and three gp120 molecules, embedded in the viral envelope.
HR1
HR2
Entry of HIV-1 is a multistep process mediated by the Envelope complex
Schematic
Viral membrane
PDB code: 4NCO
Julien et al. (2013) Science. Dec. 20 (342(6165): 1477-83
gp120
gp41

5. A. The envelope protein complex is a heterotrimer of three gp41 molecules
and three gp120 molecules, embedded in the viral envelope
B. The first specific high affinity interaction occurs between CD4 and the
gp120 subunit of the envelope complex.
HR1
HR2
Entry of HIV-1 is a multistep process mediated by the Envelope complex

6. A. The envelope protein complex is a heterotrimer of three gp41 molecules
and three gp120 molecules, embedded in the viral envelope
B. The first specific high affinity interaction occurs between CD4 and the
gp120 subunit of the envelope complex.
C. This interactions promotes a subsequent interaction with another cell
surface receptor, which is usually a G-protein coupled receptor, and
usually either CXCR4 or CCR5
HR1
HR2

7. A. The envelope protein complex is a heterotrimer of three gp41 molecules
and three gp120 molecules, embedded in the viral envelope
B. The first specific high affinity interaction occurs between CD4 and the
gp120 subunit of the envelope complex.
C. This interactions promotes a subsequent interaction with another cell
surface receptor, which is usually a G-protein coupled receptor, and most
predominantly either CXCR4 or CCR5
D. This interaction causes rearrangement in the gp41 subunit, such that the
fusion peptide is inserted into the membrane, and the helical regions (HR1
and HR2) fold back and interact with each other, facilitating the bringing
together of the two membranes and eventual fusion.
HR1
HR2

8. The piperazine derivatives developed by BMS are the most potent entry
inhibitors to date
BMS-378806
BMS-488043
BMS-626529
EC50 1.47 nM
CC50 >300 µM
EC50 0.88 nM
CC50 >300 µM
EC50 0.4 nM
CC50 >300 µM
Lin et al. 2003. PNAS.
100(19):11013
Wang et al. 2009. JMedChem.
52:7778-7787
Nowicka-Sans. 2012. AAC.
56(7):3498-3507

9. • The BMS compounds have a broad therapeutic spectrum indicating
their binding site is conserved and available
• They are specific to HIV-1
• Limited by their low solubility and low bioavailability
Therefore, we proposed to find other chemotypes that bind to the BMS
binding site on Env but that potentially have improved drug-like properties

10. As there is no information regarding BMS binding site, we templated BMS-377806,
BMS-488043 and BMS-626529 to develop a Field Point pharmacophore.
This was used to probe the Cresset in silico compound library using Blaze.

11. Blaze analysis resulted in the ranked identification of 1000 potential hit
compounds
50 compounds were individually chosen for antiviral testing using the single-
round infection assay
Pseudo-virus
• Transfect producer cells (293T) with:
• Env (AMLV, YU-2) Plasmid
• Rev plasmid
• Backbone Plasmid
• (NL4-3; Luc Reporter gene)
Co-incubation
• Incubate CD4+ CCR5+ cells
with:
• Pseudo-virus
• Compound of interest
Measure
Luminescence
• Determine IC50
of each
compound

12. Results from initial single-round infection screen
Compound IC50 YU-2 (µM) IC50 AMLV (µM)
11
13.1 ± 1.7 N.A.
12
53.5 ± 3.0 N.A.
28
33.7 ± 4.5 N.A.
32
79.4 ± 11 N.A.
34
153 ± 44 N.A.
Table 1: Structure and
potency of compounds
identified within this
study with specific
activity against HIV-1
YU-2. N.A. = not active
over concentration
range tested.

13. Piperazine is core scaffold
of BMS compounds
SC03 BMS-378806

14. Improved novelty - search for piperazine bioisosteres
Mowbray, C.E., et al. Challenges of drug discovery in novel target space. The
discovery and evaluation of PF-3893787: a novel histamine H4 receptor antagonist.
Bioorganic & medicinal chemistry letters 21, 6596-6602 (2011)
2-methyloctahydropyrrolo[3,4-c]pyrrole
BIF% with piperazine = 57%

15. ~7-fold reduction in potency
SC03 SC04
IC50 = 15µM IC50 = 100µM

16. SC04 has a new central scaffold and is specific
for HIV-1
But acenanapthene can be genotoxic and
Very low potency

17. Chose “headgroups” from BMS patent literature and checked BIF% to
acenapthene ring
BIF% = 100 BIF% = 57 BIF% = 64

18. SC07 SC08
IC50 = 16µM
SC04
IC50 = 100µM IC50 = 0.19µM

19. BMS-488043
BMS-626529
EC50 0.88 nM
CC50 >300 µM
EC50 0.4 nM
CC50 >300 µM
BMS improved compound by substituting methoxy for a methyltriazole

20. SC08
IC50 = 0.19µM
SC11
IC50 = 1 nM

21. SC11 has poor drug-like score, just like BMS compounds. Spark experiment on phenyl group and
assessment using StarDrop. Chose compound for synthesis based upon BIF% and drug-like score.

22. SC11
IC50 = 1 nM
SC26
IC50 = 0.6nM
Drug Score = 0.3815 ± 0.2323Drug Score = 0.0179 ± 0.026

23. De novo and Spark-inspired chemotype diversification
BMS-626529
Drug Score = 0.02 ± 0.027
SC26
IC50 = 0.6nM
Drug Score = 0.3815 ± 0.2323
IC50 = 0.4nM
SC15
IC50 = 3nM
AP Tanimoto = 0.69
AP Tanimoto = 0.47
SC12
IC50 = 80nM
AP Tanimoto = 0.58
SC45
IC50 = 100nM
AP Tanimoto* = 0.55
piperazine dipyrrolidine azetidine
pyrrolo-pyrazole
tetrahydropyridine

24. Potency improvement on the pyrrolo-pyrazole scaffold
SC12
IC50 = 80nM
AP Tanimoto = 0.58
SC38
IC50 = 8nM
AP Tanimoto = 0.43
TI (CC50/IC50) = 81,000

25. Summary
• Successfully used Fieldscreen to identify new entry inhibitors
• Improved novelty of hit using a combination of literature searching
coupled with Spark analysis
• Improved potency using patent literature coupled with Spark analysis
•Used Spark coupled to StarDrop to improve predicted drug-like
properties of dipyrrolidine lead.
•Identified a total of 4 new “core” chemotypes with ≤100nM potency

26. Acknowledgements
Drexel University College of Medicine
Marina Tuyishime
Mathew Danish
DFCI/Harvard Medical School
Navid Madani
Amy Princiotto
Joseph Sodroski
Cresset Group
Rae Lawrence
Compounds were synthesized by outsourcing to WuXi Apptec and HD Biosciences (China) and AsisChem
(USA). Special thanks to Drs. Henry-Georges Lombart and Joel Berniac from AsisChem.
Funding – NIH and W. W. Smith Charitable Trust