Design and optimisation of potent gp120-CD4 inhibitors

Article abstract of DOI:10.1016/j.bmcl.2009.06.102

The synthesis and structure-activity relationship of a series of novel gp120-CD4 inhibitors are described. Pharmacokinetic studies and antiviral spectrum assessment of lead compounds led to the identification of compound 36, a potent gp120-CD4 inhibitor w

Full text of DOI:10.1016/j.bmcl.2009.06.102

Bioorganic & Medicinal Chemistry Letters 19 (2009) 5250–5255
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design and optimisation of potent gp120-CD4 inhibitors
a
a
a
b
Thien-Duc Tran a, , Fiona M. Adam , Frederick Calo , David R. Fenwick , Juin Fok-Seang ,
Iain Gardner c, Duncan A. Hay a, Manos Perros b, Jaiessh Rawal c, Donald S. Middleton a,
Tanya Parkinson b, Christopher Pickford b, Michelle Platts a, Amy Randall a, Peter T. Stephenson a,
Hannah Vuong a, David H. Williams a
*
a Discovery Chemistry, Pfizer Global Research and Development, Sandwich Laboratories, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
b Discovery Biology, Pfizer Global Research and Development, Sandwich Laboratories, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
c Department of Pharmacokinetics, Dynamics and Metabolism, Pfizer Global Research and Development, Sandwich Laboratories, Ramsgate Road, Sandwich, Kent CT13 9NJ, UK
a r t i c l e i n f o
a b s t r a c t
Article history:
The synthesis and structure–activity relationship of a series of novel gp120-CD4 inhibitors are described.
Pharmacokinetic studies and antiviral spectrum assessment of lead compounds led to the identification
of compound 36, a potent gp120-CD4 inhibitor which exhibited antiviral potency across a spectrum of 25
clade B isolates.
Received 15 May 2009
Revised 23 June 2009
Accepted 23 June 2009
Available online 4 July 2009
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
HIV-1
gp120-CD4 inhibitor
Antiviral
Infection with human immunodeficiency virus (HIV) affects
millions of people worldwide.1 Despite the introduction of highly
active antiretroviral therapy (HAART) regimes using combination
therapies which target one of the essential viral proteins, there is
still a need for new therapeutic agents with improved dosing re-
Trimeris has also reported a series of small molecule gp120-CD4
inhibitors based on the modifications of the alpha keto amide with
an isosteric sulfonamide group.7
As part of our research efforts to identify new therapeutic
agents for the treatment of HIV, we recently reported the discovery
of compound 1,8 the prototype of a series of novel HIV-1 entry
inhibitors that bind directly to the viral gp120 envelope protein
and prevent the interaction between gp120 and host CD4 receptors
on T cells, the very first step of the viral entry process. Compound 1
has an exciting profile with a 200 mg bid dose predicted to give a
gimes that are better tolerated with
a reduced side effect
burden.2
Bristol-Myers Squibb has developed a series of small mole-
cule gp120-CD4 inhibitors based on an alpha keto amide struc-
ture (Fig. 1) which specifically inhibit HIV entry.3 These
compounds exhibit potent antiviral activity in cell culture and
BMS-488043 has demonstrated the ability to reduce viral loads
in man.4–6
9
Cmin of 500 nM, covering 70% of clade B viruses at IC90. The com-
pound is active against both -R5, -X4 and dual tropic viruses and
has an estimated half-life in human of 5 h. The present work
H
H
N
N
O
O
MeO
N
N
N
N
O
N
O
N
OMe
OMe
O
O
BMS-488043
BMS-378806
Figure 1. gp120-CD4 inhibitors reported by BMS.
* Corresponding author. Tel.: +44 1304 644548; fax: +44 1304 651819.
E-mail address: thien.tran@pfizer.com (T.-D. Tran).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.06.102

T.-D. Tran et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5250–5255
5251
OMe
N
O
O
Compound 1
N
Fusion IC50 = 14nM
N
MeNH
AV IC50 = 2nM
clogP = 2.2 LogD = 0.7
O
O
Figure 2. Profile of compound 1.
O
O
Br
O
OH
(a)
N
R
N
+
R
N
N
O
O
2
compounds 3-15
Scheme 1. Reagents and conditions: (a) Cs2CO3, acetone, rt, 24 h, y = 60–90%.
describes our attempts to identify a back-up with improved profile
in terms of antiviral spectrum and/or pharmacokinetic properties
over compound 1 (Fig. 2).
Our goals were to identify a more potent compound with com-
parable pharmacokinetic profile to compound 1 to lower the dose
Table 1 (continued)
Compound
R
Fusion IC50
(nM)
Anti-viral IC50
(nM)
clog Pd
MW
403
10
103
99
492
271
ND
3.42
Table 1
N
gp160 fusion inhibitory activitya and antiviral activityb of piperazine analogues
Compound
R
Fusion IC50
(nM)
Anti-viral IC50
(nM)
clog Pd
MW
407
11
12
13
3.63
3.75
3.54
403
403
403
N
N
O
3
2040
834
16
NDc
ND
36
3.28
7430
9370
N
N
N
4
5
6
3.44
3.54
3.34
406
406
420
ND
N
N
N
N
N
14
68
1050
2.31
2.98
419
404
NH
O
N
N
15
79
15
493
ND
7
8
9
123
433
575
1070
449
4.54
3.63
3.42
402
403
403
a,b
See Refs. 5 and 6 for complete details of assay conditions.
ND = not determined.
Biobyte Corporation, 201 W. 4th St.,#204, Claremont CA 91711-4707, CLOGP
c
d
version 4.3,
N
prediction and to search for significantly structurally-differenti-
ated leads to give a different antiviral spectrum to compound 1.
Our work focused initially on finding a novel expression of the
methoxy pyridine amide moiety as this fragment could be easily
replaced from a synthetic point of view. A range of polar bicyclic
analogues were prepared to mimic the methoxy substitution pat-
tern of the pyridine template. Test compounds were conveniently
2040
N

5252
T.-D. Tran et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5250–5255
O
N
OMe
O
OH
N
O
O
a
b-c
O
HO
O
OMe
Het
N
+
Het
N
Het
N
17
16
O
d-g
h-i
N
N
O
O
O
18
N
Het
N
N
Cl
O
O
j
k-m
Het
N
O
O
NR1R2
O
N
N
compounds 19-21
N
O
N
O
O
O
Het
N
Het
N
R4R3N
compounds 22-24
O
compounds 25-27
NR5R6
Scheme 2. Reagents and conditions: (a) DIAD, PPh3, THF, rt, o/n, y ꢀ 70%; (b) NaOH, dioxane, 55 °C, 3 h, y ꢀ 80%; (c) HBTU, triethylamine, 2-(R)-4-benzoyl-2-methyl-
piperazine, DCM, DMF, rt, o/n, y = 40–90%; (d) N-bromosuccinimide, acetonitrile, rt, o/n, y ꢀ 40%; (e) Pd(dppf)Cl2, CO, triethylamine, methanol, 100 psi, 100 °C, y ꢀ 60% (f)
NaOH, dioxane, 55 °C, 3 h, y ꢀ 80%; (g) HBTU, triethylamine, R1R2NH, DCM, DMF, y = 40–90%; (h) m-CPBA, DCM, rt, 3 h; y = 30–70%; (i) POCl3, DCM; 50 °C, 3 h, y ꢀ 70% (j)
R3R4NH, triethylamine, caesium fluoride, DMSO, 130 °C, y ꢀ 70%; (k) Pd(dppf)Cl2, CO, triethylamine, methanol, 100 psi, 100 °C, y ꢀ 60% (l) NaOH, dioxane, 55 °C, 3 h, y ꢀ 80%;
(m) HBTU, triethylamine, R5R6NH, DCM, DMF, rt, o/n, y = 40–90%.
prepared from commercially available hydroxy heterocycles by the
general methodology described in Scheme 1. Standard bromide
displacement of intermediate 2 with the appropriate hydroxy het-
erocycles under basic conditions provided required targets 3–15
(Table 1).
ate 17. Intermediate 17 was then brominated with N-bromosuccin-
imide in acetonitrile. The corresponding bromo intermediate was
carbonylated under carbon monoxide pressure to give the corre-
sponding methyl ester intermediate. Ester hydrolysis under basic
conditions followed by standard amide coupling provided target
compounds 19–21. Reaction of intermediate 17 with m-CPBA pro-
vided the N-oxide intermediate which was then chlorinated to give
the chloro intermediate 18. Carbonylation of 18 followed by ester
hydrolysis and amide coupling provided target compounds 25–
27. The chloro intermediate 18 could also be displaced by amines
under basic conditions to provide target compounds 22–24.
Initial attempts to incorporate the potency enhancing amide
substituent of compound 1 (Table 2) into the isoquinoline, quino-
line and indazole templates resulted in a loss in potency (com-
pounds 19–21). Due to the difficulty in functionalising the
indazole template, we subsequently focused our efforts on the iso-
quinoline and quinoline templates. Despites our initial disappoint-
ments, we continued to explore a range of substituents to build up
SAR around the heterocyclic rings. Primary amines (compound 22)
improved potency by 20-fold in both fusion and anti-viral assays,
while introduction of secondary amines resulted in a loss of po-
tency (compound 23) suggesting that the H-bond donor was essen-
tial for potency. Amides such as 25 were also tolerated albeit
significantly weaker in the antiviral assay. Within the isoquinoline
template, both amines and amides were tolerated (compounds 24
and 26), with potency comparable to the quinoline 22. The most
A number of derivatives showed promising activity in our
gp160 fusion assay,10 however, this did not always translate to
equivalent antiviral activity.11 Interestingly, indazole analogue 5
had promising potency in both assays as well as its N-methylated
version 6 suggesting that the hydrogen on the indazole was not re-
quired for potency. Isoquinoline analogue 10 and quinoline ana-
logue 11 had also promising potency in both assays, prompting
us to explore the SAR around these novel heterocycles to enhance
the potency of our targets. Structurally, the quinoline, isoquinoline
and indazole templates looked very similar to the methoxy pyri-
dine template; they just lack the amide feature which was essen-
tial for the potency of compound 1.2 Our priority was to
introduce this key amide substituent onto the newly discovered
rings to assess its influence on potency and lower the lipophilicity
of the lead molecules. All single diastereoisomeric compounds
were synthesised by the general method shown in Scheme 2.
Reaction of (R)-methyl lactate and the appropriate phenol in
THF under standard Mitsunobu conditions afforded the aryl ether
16 generally in 70% yield. Ester hydrolysis under basic conditions
followed by standard amide coupling with 2-(R)-4-benzoyl-2-
methyl-piperazine provided the single diastereoisomer intermedi-

T.-D. Tran et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5250–5255
5253
Table 2
Table 2 (continued)
gp160 fusion inhibitory activity and antiviral activity of functionalised piperidine
analogues
Compound
R
Fusion IC50 Anti-viral clog P MW
(nM)
IC50 (nM)
O
N
N
R
26
21
1.5
3.21
460
N
O
O
O
NHMe
Compound
R
Fusion IC50 Anti-viral clog P MW
(nM)
IC50 (nM)
OMe
27
31
20
3.37
478
N
1
14
2
2.2
440
N
F
O
NHMe
O
MeNH
potent compound from each novel template, 22 and 26, was pro-
gressed into rat pharmacokinetic studies (Table 3).
N
19
>10,000
ND
2.65
463
N
Compound 22 had a lower absorption, lower bioavailability
and a higher unbound clearance compared to compound 1, attrib-
uted to its high lipophilicity. This, coupled with a lower potency
in the antiviral assay, resulted in an inferior profile to compound
1. Compound 26 was equipotent to compound 1 in the antiviral
assay and based on rat pharmacokinetic profile, we estimated
that to cover 70% of clade B virus at IC90, a 300 mg bid dose
would be required, resulting in an overall inferior profile to com-
pound 1. Subsequently, compound 26 showed a positive signal in
the glutathione trapping assay. The source of the reactive metab-
olite in compound 26 is believed to be P450-mediated oxidation
para to the oxygen on the isoquinoline leading to a quinone-like
species. We then prepared several analogues to block metabolism
at this position. Unfortunately, the direct fluoro analogue 27
showed a significant drop in potency in the antiviral assay, putt-
ing an end to our efforts in exploring an alternative expression of
the pyridine template.
O
O
O
MeNH
20
21,200
ND
2.71
460
N
MeNH
MeNH
21
22
1260
ND
2.92
4.12
460
432
N
We then turned our attention to the benzamide group in com-
pound 1. Our strategy was to identify a conformationally restricted
expression of the benzamide, the hypothesis being that if we could
access a favoured locked conformation then significant potency
gain could be realised. All targets were synthesised by the general
method shown in Scheme 3 (Table 4).
5
13
NHMe
N
Replacement of benzamide with directly attached imidazole 30
or purine 33 gave disappointing potency in the fusion assay.
Replacement with pyridine 29, indazole 31 or azaindole 32 gave
some encouraging potency in the fusion assay; however this did
not translate into antiviral activity. The most encouraging replace-
ment was with polar heterocyclic systems such as azaindole 34
and azabenzimidazole 35, which had similar polarity and antiviral
potency to compound 1. Unfortunately, both compounds 34 and 35
23
24
3200
ND
ND
4.34
3.91
446
432
N
N
61
N
Table 3
Rat pharmacokinetics (IV 1 mg/kg, po 1 mg/kg)
NHMe
Compound 1
Compound 22
Compound 26
Cl (mL/min/kg)
Clu (mL/min/kg)
T1/2 (h)
Vdu (L/kg)
F (%)
18
26
1
2.2
65
100
27
41
90
0.5
3.5
45
55
129
0.9
10.5
42
25
78
>4000
3.42
460
NHMe
N
Fabs (%)
60
O
Cl = Clearance; Clu = unbound clearance; Vdu = unbound volume of distribution,
F = bioavailability; Fabs = fraction of compound absorbed.

5254
T.-D. Tran et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5250–5255
O
Het
N
OH
N
O
O
N
O
N
MeO
O
O
a
O
MeO
b-c
O
MeO
N
Het-Cl
N
N
O
NHMe
O
NHMe
O
NHMe
28
27
compounds
Scheme 3. Reagents and conditions: (a) WSCDI, HOBt, DIPEA, 2-(R)-4-N-BOC-2-methyl-piperazine, THF, rt, o/n, y = 80%; (b) HCl 4 M dioxane, rt, 3 h, y = Q; (c) triethylamine,
butanol, 120 °C, 3 h, y ꢀ 50%.
Table 4
Table 5
Rat pharmacokinetic (IV 1 mg/kg, po 1 mg/kg)
Compound
Het
Fusion IC50
(nM)
Anti-viral IC50
(nM)
clog P
MW
414
Compound 36
Cl (mL/min/kg)
Clu (mL/min/kg)
T1/2 (h)
Vdu (L/kg)
F (%)
48
198
0.5
9.4
21
29
30
71
1620
ND
1.32
N
N
N
Fabs (%)
100
1900
1.7
458
452
N
N
Table 6
Antiviral activity of compound 1 and compound 36 against a panel of 25 clade B
isolates
1500
1400
1300
1200
1100
1000
900
31
32
33
34
42
979
>1000
ND
2.87
N
H
N
N
N
N
29
2.61
1.46
2.3
452
454
453
Compound 1
compound 36
800
700
600
500
400
300
200
100
0
N
H
N
296
24
N
H
N
1
3
5
7
9
11 13 15 17 19 21 23 25 27
Clade B isolates
6
N
N
H
had liver blood flow clearance and zero oral exposure in the rat
making both inferior to compound 1. Both compounds were be-
lieved to be substrates for aldehyde oxidase.12 To our surprise,
the isoquinoline modification (compound 36) produced a dramatic
increase in potency in both fusion and antiviral assays. However
the compound is not very stable metabolically with a half-life of
only 4 min in human liver microsomes due to its high lipophilicity.
Further heterocyclic analogues were prepared in the hope of main-
taining the excellent antiviral potency whilst reducing metabolic
vulnerability but all had disappointing activities as exemplified
by compound 37. With such an exquisite antiviral potency, com-
pound 36 emerged as our most promising agent and was subjected
to wider pharmacokinetic and virologic profiling. Compound 36
had a higher clearance and a lower bioavailability than compound
1 resulting in a half life of 0.5 h in the rat (Table 5). Scaling from
the rat, compound 36 was predicted to exhibit a half life in humans
of 3.5 h.
N
35
36
37
24
7
2.24
3.26
2.1
453
463
464
N
H
0.29
29
0.12
126
N
N
N

T.-D. Tran et al. / Bioorg. Med. Chem. Lett. 19 (2009) 5250–5255
5255
W. S.; Shi, P.-Y.; Lin, P.-F.; Colono, R. J.; Meanwell, N. A. J. Med. Chem. 2003, 46,
4236.
4. Yang, Z.; Zadjura, L.; D’Arienzo, C.; Marino, A.; Santone, K.; Klunk, L.; Greene, D.;
Lin, P.-F.; Colonno, R.; Wang, T.; Meanwell, N.; Hansel, S. Bioparm. Drug Dispos.
2005, 26, 387.
Compound 36 was then tested against a panel of 25 clade B iso-
lates (Table 6). The compound was active against R5-, X4- and
dual-tropic viruses.13 Most importantly, compound 36 was active
against all 25 isolates tested and had an improved spectrum profile
compared to compound 1. Based on human predictions and antivi-
ral spectrum data, a 100 mg bid dose was predicted to give a Cmin of
50 nM, covering 70% of clade B at IC90. Assuming a linear pharma-
cokinetic profile across the dose range, a greater coverage could be
achieved by higher but still reasonable doses, for example, 600 mg
bid had a predicted Cmin of 300 nM would cover 96% of clade B at
IC90. Although compound 36 had an inferior pharmacokinetic pro-
file compared to compound 1, its exquisite potency provided in-
creased antiviral spectrum coverage. Further studies done on
compound 1 showed that cross clade inhibition14 would be very
difficult with this chemical class, so more resistance studies would
be required to progress compound 36 further.
5. (a) Hanna, G.; Lalezari, L.; Hellinger, J.; Wohl, D.; Masterson, T.; Fiskel, W.;
Kadow, J.; Lin, P.; Giordano, M.; Colonnol, R.; Grasela, D. 11th Conference on
Retroviruses, Opportunistic Infection, Feb 8–11, 2004, San Francisco, CA, 2004;
Abstract 141.; (b) Guo, Q.; Ho, H.-T.; Dickeer, I.; Fan, L.; Zhou, N.; Friborg, J.;
Wang, T.; McAuliffe, B. V.; Wang, H.-G. H.; Rose, R. E.; Fang, H.; Sarnati, H. T.;
Langley, D. R.; Meanwell, N. A.; Abraham, R.; Collono, . J.; Lin, P.-F. J. Virol. 2003,
77, 10528.
6. Wang, H.-G.; Williams, R. E.; Lin, P.-F. Curr. Pharm. Des. 2004, 10, 1785.
7. Lu, R.; Tucker, J. A.; Zinevitch, T.; Kirichenko, O.; Konoplev, V.; Kuznetsova, S.;
Sviridov, S.; Pickens, J.; Tandel, S.; Brahmachary, E.; Yang, Y.; Wang, J.; Freel, S.;
Fisher, S.; Sullivan, A.; Zhou, J.; Stanfield-Oakley, S.; Greenberg, M.; Bolognesi,
D.; Bray, B.; Koszalka, B.; Jeffs, P.; Khasanov, A.; Ma, Y.; Jeffries, C.; Liu, C.;
Proskurina, T.; Zhu, T.; Chucholowski, A.; Li, R.; Sexton, C. J. Med. Chem. 2007,
50, 6535–6544.
8. Tran, T. D.; Adam, F.; Fenwick, D. R.; Fok-Seang, J.; Gardner, I.; Hay, D.; Rawal, J.;
Middleton, D. S; Parkinson, T.; Pickford, C.; Platts, M.; Randall, A.; Stephenson,
P. T; Vuong, H.; Williams, D. H. Bioorg. Med. Chem. Lett. 2009, 19, 5246.
9. IC90 = the concentration of compound needed to reduce HIV replication in cell
culture by 90%.
In summary, a number of gp120-CD4 inhibitors for the treat-
ment of HIV-1 were synthesized. Evaluation of the potency and
pharmacokinetic properties of these compounds led to the identi-
fication of a broad antiviral spectrum agent with subnanomolar
levels of potency against several clade B isolates.
10. Fusion assay Bradley, J. et al J. Biomol. Screening 2004, 9, 516–524.
11. (a) Antiviral assays. Activity against NL4-3 virus in HeLa-P4 cells was
determined as follow. Diluted compound was added to wells of a sterile 96-
well flat-bottomed black tissue culture plate with clear base and lid (Costar).
HeLa-P4 cells were mixed with NL4-3 virus and added to the wells at a
concentration of 105 cells/mL in a volume of 90
lL. The plates were incubated
Supplementary data
for 5 days at 37 °C in a 5% CO2 humidified incubator, then virus infection was
quantified using the FluorAce b-galactosidase reporter assay kit (Bio-Rad)
according to the manufacturer’s instructions; (b) Antiviral assay against all
other strains was carried out by Monogram Biosciences using their phenosense
assay.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.bmcl.2009.06.102.
12. Oxidized metabolite formed in rat cytosol but not in rat liver microsome.
Inhibition of metabolism observed in the presence of an aldehyde oxidase
inhibitor, Raloxifene (data not shown).
References and notes
13. Antiviral activity of compounds 1 and 36 against 25 Clade B isolates reported in
Supplementary data.
1. UNAIDS/WHO 2008 AIDS epidemic updates.
14. Antiviral activity of compound 1 against Clade A, B, C and D isolates presented
at the twenty first International Conference on Antiviral Research, Montreal,
Quebec, Canada, April 13–17, 2008.
2. Pomerantz, R. J.; Horn, D. L. Nat. Med. 2003, 9, 867–873.
3. Wang, T.; Zhang, Z.; Wallace, O. B.; Deshpande, M.; Fang, H.; Yang, Z.; Zadujura,
L. M.; Tweedie, D. L.; Hunag, S.; Zhao, F.; Ranadive, S.; Robinson, B. S.;
Gong, Y.-F.; Ricarrdi, K.; Spicer, T. P.; Deminie, C.; Rose, R.; Wang, H.-D. H.; Blair,