HCV NS5A replication complex inhibitors. Part 2: Investigation of stilbene prolinamides

Article abstract of DOI:10.1016/j.bmcl.2012.08.049

In a previous disclosure,¹ we reported the dimerization of an iminothiazolidinone to form 1, a contributor to the observed inhibition of HCV genotype 1b replicon activity. The dimer was isolated via bioassay-guided fractionation experiments and shown to be a potent inhibitor of genotype 1b HCV replication for which resistance mapped to the NS5A protein. The elements responsible for governing HCV inhibitory activity were successfully captured in the structurally simplified stilbene prolinamide 2. We describe herein the early SAR and profiling associated with stilbene prolinamides that culminated in the identification of analogs with PK properties sufficient to warrant continued commitment to this chemotype. These studies represent the key initial steps toward the discovery of daclatasvir (BMS-790052), a compound that has demonstrated clinical proof-of-concept for inhibiting the NS5A replication complex in the treatment of HCV infection.

Full text of DOI:10.1016/j.bmcl.2012.08.049

Bioorganic & Medicinal Chemistry Letters 22 (2012) 6063–6066
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
HCV NS5A replication complex inhibitors. Part 2: Investigation of stilbene
prolinamides q
Denis R. St. Laurent a, Makonen Belema a, Min Gao b, Jason Goodrich a, Ramesh Kakarla a, Jay O. Knipe c,
Julie A. Lemm b, Mengping Liu b, Omar D. Lopez a, Van N. Nguyen a, Peter T. Nower b, Donald O’Boyle II b,
Yuping Qiu a, Jeffrey L. Romine a, Michael H. Serrano-Wu a, Jin-Hua Sun b, Lourdes Valera b, Fukang Yang a,
Xuejie Yang a, Nicholas A. Meanwell a, Lawrence B. Snyder a,
⇑
a Department of Medicinal Chemistry, Bristol-Myers Squibb Research and Development, 5 Research Parkway, Wallingford, CT 06492, USA
b Department of Virology, Bristol-Myers Squibb Research and Development, 5 Research Parkway, Wallingford, CT 06492, USA
c Department of Metabolism and Pharmacokinetics, Bristol-Myers Squibb Research and Development, 5 Research Parkway, Wallingford, CT 06492, USA
a r t i c l e i n f o
a b s t r a c t
In a previous disclosure,1 we reported the dimerization of an iminothiazolidinone to form 1, a contributor
to the observed inhibition of HCV genotype 1b replicon activity. The dimer was isolated via bioassay-
guided fractionation experiments and shown to be a potent inhibitor of genotype 1b HCV replication
for which resistance mapped to the NS5A protein. The elements responsible for governing HCV inhibitory
activity were successfully captured in the structurally simplified stilbene prolinamide 2. We describe
herein the early SAR and profiling associated with stilbene prolinamides that culminated in the identifi-
cation of analogs with PK properties sufficient to warrant continued commitment to this chemotype.
These studies represent the key initial steps toward the discovery of daclatasvir (BMS-790052), a com-
pound that has demonstrated clinical proof-of-concept for inhibiting the NS5A replication complex in
the treatment of HCV infection.
Article history:
Received 6 June 2012
Revised 7 August 2012
Accepted 13 August 2012
Available online 21 August 2012
Keywords:
Hepatitis C virus (HCV)
NS5A
Daclatasvir
Antiviral
Ó 2012 Elsevier Ltd. All rights reserved.
Clinical treatment for hepatitis C virus (HCV), a disease infecting
almost 170 million individuals world-wide, is associated with low
response rates for genotype 1-infected patients in conjunction
with a high incidence of side effects for regimens that include
Commensurate with in vitro data observed for the iminothiazo-
lidinone series, we showed previously that antiviral activity ob-
served with stilbene prolinamides was closely associated with
the embedded amino acid with the S-configuration in 2 150-fold
more potent than the R-antipode.1 In order to further delineate
the minimal molecular footprint required to effect the observed
inhibitory potency and to understand the role of symmetry, a sys-
tematic truncation of the lead molecule was undertaken.
pegylated interferon-
a
(PEG-IFN)/ribavirin (RBV).2,3 However, the
development of direct acting antiviral agents (DAAs) that can be
used in combination holds considerable promise for the identifica-
tion of better tolerated and more effective anti-HCV therapies in
the future.4 Clinical trials with the potent HCV NS5A inhibitor
daclatasvir (BMS-790052) in HCV genotype 1-infected subjects
have established the relevance of HCV NS5A as a therapeutic target
and additional advanced studies are ongoing.5,6 In this Letter we
describe the preliminary structure–activity relationships (SAR)
and early PK characterization for a series of symmetrically-dis-
posed stilbene prolinamides that formed the basis for the ultimate
identification of daclatasvir. These studies were predicated by the
delineation of the key pharmacophoric elements in dimer 1 that
were successfully captured in stilbene 2.1
Synthetic access to the unsymmetrical N-phenacetyl E-stilbene
prolinamides is shown in Scheme 1. (E)-4,40-(Ethene-1,2-diyl)dian-
iline (3) and N-phenacetyl-L-proline were treated with EEDQ to
give intermediate 4. A second coupling reaction with the carbox-
ylic acids corresponding to the products compiled in Table 1 gave
5. Alternatively, symmetrical E-stilbene prolinamides were ob-
tained upon coupling with Boc-proline to afford 6. Deprotection
followed by a second coupling reaction with a series of carboxylic
acids provided products 7 presented in Table 2, as shown in the
lower path of Scheme 1.
q
See Ref. 1.
⇑
Corresponding author. Tel.: +1 203 677 6191; fax: +1 203 677 7702.
E-mail address: Lawrence.Snyder@bms.com (L.B. Snyder).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.bmcl.2012.08.049

6064
D. R. St. Laurent et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6063–6066
Table 1
Ph
Ph
Truncation of stilbene prolinamides
O
O
HN
Me
O
O
N
O
O
N
HN
HN
O
S
N
HN
N
NH
R
Ar
O
Ar
N
S
NH Me
a,b
NH
O
N
Compd
R
EC50 1b (lM)
O
O
O
NH
N
O
O
Ph
O
O
Ph
N
2
0.000086
2
1
O
Me
H
N
MeO2CHN
N
8
9
0.0015
1.39
O
N
N
N
O
O
N
N
H
O
NHCO2Me
H
N
Daclatasvir
(BMS-790052)
Me
10
11
1.64
O
H
The data presented in Table 1 captures the effect of truncation on
the inhibitory potency towards HCV genotype 1b replication. It
was immediately apparent that excision of the terminal phenyl ring,
as in 8, resulted in significantly reduced potency (>200ꢀ). This
trend of potency loss continued with further truncation of 8 result-
ing in micromolar inhibitory activity for 9 and 10 and, finally, essen-
tially complete loss of replicon inhibitory activity with 11. Alanine
12 exhibited nanomolar potency albeit with a >10ꢀ loss when com-
pared to 2, which might be attributable to a loss in conformational
restriction of the terminal moiety. Excision of the phenyl ring from
12, to provide 13, or removal of the alanine methyl group from 12
resulting in 14 caused a marked diminution in replicon inhibitory
activity. These early results demonstrated the need for pursuing
the full molecular structure where optimal design resided in a sym-
metrical pharmacophore. This structural characteristic comple-
mented the dimeric nature of the crystal structures of HCV NS5A
domain I that were published after our discovery and is a key con-
cept that would ultimately be crucial to the discovery of daclatasvir
(BMS-790052).7
>10.00
O
O
O
NH
12
13
14
0.00257
0.153
O
O
Me
NH
NH
0.768
O
a
Data represent mean values of at least two experiments.
1b CC50 and BVDV EC50 >10.00 lM.
b
In a parallel effort, commercially-available Cbz-protected amino
acids were used to probe the ring size of the proline moiety as well
as varying substitution patterns on the pyrrolidine ring (Table 2).
The findings once again revealed alanine to be an inferior substi-
tute for proline (compare 15 and 16 to 17) while ring expansion
(19), contraction (18), or ring fusion (20–22) resulted in lowered
potency when compared to 17. Inclusion of polarity in the form
of an oxygen atom embedded within the ring, as in 23, or appended
to the pyrrolidine ring, as exemplified by 24–30, appeared to be
well tolerated provided that the appendage was smaller than eth-
oxy (the large carbamate 31 resulted in poor inhibitory activity). It
is concluded from the comparison of 24 and 25 with 26 and 28 that
hydrogen bond donors are not as well tolerated as hydrogen bond
acceptors in this region of the pharmacophore. It is noteworthy
that in conjunction with the quest to identify analogs with potent
genotype 1b inhibitory activity, we were concurrently developing a
genotype 1a replicon assay and surveying inhibitory potency with
select compounds. The t-Bu ether 30 was one of the first com-
pounds to demonstrate inhibitory activity towards the genotype
1a replicon with a relatively modest EC50 of 500 nM.
R
NH2
b
HN
O
a
O
O
4
5
NH
O
NH
O
NH2
Ph
Ph
N
N
R
Boc
N
N
3
O
HN
HN
H2
N
O
O
c
d, e
Table 3 depicts the exploration of the capping element on the
proline, a region of the pharmacophore sensitive to structural
changes.1 Although a methyl group at the benzylic position seems
to be well tolerated (34 and 35), removal of a single carbon atom
from the phenacetyl cap of 2 afforded 36, a compound with
1000-fold attenuated potency. Further truncation to the acetyl-
substituted stilbene 37 proved to be even more detrimental to
genotype 1b inhibitory potency. Retaining the chain length found
in 2 with the introduction of heteroatoms proved to be a more
fruitful direction but showed a marked sensitivity to the position
O
O
N
6
7
NH
NH
O
N
Boc
R
Scheme 1. Preparation of stilbene prolinamides. Reagents and conditions: (a) N-
PhAc- -proline, EEDQ, CH2Cl2, rt; (b) RCO2H, EEDQ, CH2Cl2, rt; (c) N-Boc- -proline,
EEDQ, CH2Cl2, rt; (d) 4 N HCl in dioxane, MeOH, 0 °C to rt or TFA, CH2Cl2, rt; (e)
L
L
RCO2H, EDCI, DIEA, THF, rt or HATU, DIEA, DMF, rt.

D. R. St. Laurent et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6063–6066
6065
Table 2
Table 3
Evaluation of symmetrically-substituted proline and proline variants
Genotype 1b activity of symmetrically-disposed E-stilbene prolinamides—cap varia-
tions
R
HN
NH
R
H
N
Cap
N
O
N
Cap
N
H
O
a b
Compd
R
EC50
3.21
, 1b (lM)
a b
Compd
Cap
EC50 , 1b (lM)
O
CBz
HN
15
2
0.0000860
<0.005
O
O
CBz
Me
N
16
17
18
0.0731
Me
34
35
36
O
O
O
CBz
N
Me
0.00897
0.0580
0.005
0.102
CBz
O
N
CBz
N
O
O
19
20
0.0627
0.0669
Me
N
37
38
39
40
41
2.38
O
CBz
O
N
0.000784
0.00222
0.172
O
O
O
N
O
CBz
N
21
0.0890
N
O
CBz
N
0.00128
S
N
O
O
22
23
24
0.120
42
0.0102
O
O
CBz
N
Me
N
0.00614
0.0289
Me
Me
HN
43
44
45
0.0762
0.0867
1.00
O
N
O
O
O
CBz
N
N
HO
N
O
O
CBz
N
Me
N
46
47
0.0579
0.111
S
25
0.0115
N
O
HO
HN
O
O
CBz
N
26: R = Me; <0.005
27: R = Et; 0.038
26–27
28–30
O
O
Me
N
RO
CBz
N
48
49
50
0.140
0.011
0.056
28: R = Me; <0.005
29: R = Et; 0.015
30: R = t-Bu; 0.350
RO
O
O
CBz
N
O
31
32
33
>10.00
1.29
O
O
O
t-BuHN
O
CBz
N
O
Me
N
51
52
0.0868
1.63
H2N
CBz
N
O
O
N
0.0100
O
N
Me
Me
(continued on next page)
a
Data represent mean values of at least two experiments.
1b CC50 and BVDV EC50 >10.00 lM.
b

6066
D. R. St. Laurent et al. / Bioorg. Med. Chem. Lett. 22 (2012) 6063–6066
inhibitory potency ranging from 9 to 62 nM. The prototype stilbene
Table 3 (continued)
17 showed poor oral absorption, attributed to its high lipophilicity
and low solubility. Incorporation of a hydroxyl group on the pyrrol-
idine ring or removal of one of the lipophilic terminal benzene
rings provided analogs 24 and 8, respectively, with much improved
intravenous AUCs.
a b
Compd
Cap
EC50
,
1b (lM)
53
54
10.0
Replacement of the terminal benzene ring of 2 with a more po-
lar pyridine provided 38, which exhibited a marked improvement
in all PK parameters and 10% oral bioavailability. Finally, the cyclo-
propylmethyl derivative 55 exhibited an excellent PK profile with
an oral bioavailability of 64%. The data in Table 4 established that
adequate oral exposure is attainable with this chemotype.
In conclusion, the ability to translate the previously described
iminothiazolidinone series of HCV NS5A inhibitor1 into a progres-
sible, stilbene-based chemotype has been demonstrated. The stil-
bene prolinamides possess excellent HCV genotype 1b replicon
inhibitory activity with coherent and tractable SAR. The parent
phenacetyl analog 2 retained its inherent potency even after mod-
ification of the pyrrolidine ring and cap truncation or variation.
Notably, in the quest for pan-genotype coverage, analogs 30 and
54 showed modest levels of HCV genotype 1a inhibitory activity,
a prelude to the identification of a more balanced genotype 1a/
1b inhibitor that will be described in due course. Finally, rat PK
studies demonstrated that cap modification provided a direction
toward achieving oral exposure, ultimately optimized in the dis-
covery of daclatasvir as the first HCV NS5A inhibitor to demon-
strate clinical efficacy in HCV-infected subjects.
O
1.13
Et
O
O
55
0.062
a
Data represent mean values of at least two experiments.
1b CC50 and BVDV EC50 >10.00 lM.
b
Table 4
PK profile of symmetrically capped E-stilbene prolinamides
O
N
Cap
HN
Cap
NH
N
O
Compd
1b EC50
8
17
24
38
55
(l
M)a,b
0.02
82
0.009
77
<1
0.03
99
—
0.0008
88
0.062
84
4.4
RLMc (%)
Solubilityd
—
53
Caco-2 (nm/s)
Dose (mg/kg)
F (PO)%
<15
2.5
—
—
1.8
2
35
1.6
—
16
1.8
10
4.0
2.8
64
Acknowledgment
Cl (mL/min/kg)
4.0
5.2
1.2
—
—
10,535
19
11
6.4
1.0
—
—
2453
8.8
9.3
2.9
120
322
3077
2.6
4.9
800
1500
11,854
18,457
We thank Discovery Analytical Services (DAS) staff for provid-
ing HRMS and 1H NMR spectral data.
T
½
(h)
4.6
4.3
4.0
37
Vss (L/kg)
Cmax (ng/mL)
PO AUCe
References and notes
IV AUCe
1621
1. Paper 1 in the series: Romine, J. L.; Laurent, D. R., St.; Leet, J. E.; Martin, S. W.;
Serrano-Wu, M. H.; Yang, F.; Gao, M.; O’Boyle, D. R., II; Lemm, J. A.; Sun, J.-H.;
Nower, P. T.; Huang, X.; Deshpande, M. S.; Meanwell, N. A.; Snyder, L. B. ACS Med.
Chem. Lett. 2011, 2, 224.
2. Hoofnagle, J. H.; Seeff, L. B. N. Engl. J. Med. 2006, 355, 2444.
3. Lavanchy, D. Clin. Microbiol. Infect. 2011, 17, 107.
a
b
c
Data represent mean values of at least two experiments.
1b CC50 and BVDV EC50 >10.00 M.
% Remaining after 10 min incubation.
l
d
e
Solubility measured in
lg/mL.
Exposure measured in ng h/mL.
4. (a) Cheng, K.-C.; Gupta, S.; Wang, H.; Uss, A. S.; Njoroge, G. F.; Hughes, E. J.
Pharm. Pharmacol. 2011, 63, 883; (b) Yoshida, T.; Kondoh, M.; Yagi, K. Curr. Med.
Chem. 2011, 18, 1239; (c) Vermehren, J.; Sarrazin, C. Clin. Microbiol. Infect. 2011,
17, 122.
5. (a) Cordek, D. G.; Bechtel, J. T.; Maynard, A. T.; Kazmierski, W. M.; Cameron, C. E.
Drugs Future 2011, 36, 691; (b) Uli Schmitz, U.; Tan, S.-L. Recent Patents Anti-
Infect. Drug Disc. 2008, 3, 77; (c) Meanwell, N. A.; Belema, M. Annu. Rep. Med.
Chem. 2011, 46, 263.
of the heteroatom. For example, pyridine 38 (EC50 = 0.8 nM) is
essentially a log10 unit weaker than parent 2, while transposition
of the nitrogen to either the ortho- (39, EC50 = 2 nM) or para- (40,
EC50 = 170 nM) positions was associated with a further reduction
in potency.
6. (a) Gao, M.; Nettles, R. E.; Belema, M.; Snyder, L. B.; Nguyen, V. N.; Fridell, R. A.;
Serrano-Wu, M. H.; Langley, D. R.; Sun, J.-H.; O’Boyle, D. R., II; Lemm, J. A.; Wang,
C.; Knipe, J. O.; Chien, C.; Colonno, R. J.; Grasela, D. M.; Meanwell, N. A.; Hamann,
L. G. Nat. Rev. Drug Discov. 2010, 465, 96; (b) Lok, A. S.; Gardiner, D. F.; Lawitz, E.;
Martorell, C.; Everson, G. T.; Ghalib, R.; Reindollar, R.; Rustgi, V.; McPhee, F.;
Wind-Rotolo, M.; Persson, A.; Zhu, K.; Dimitrova, D. I.; Eley, T.; Guo, T.; Grasela,
D. M.; Pasquinelli, C. N. Engl. J. Med. 2012, 366, 216; (c) Chayama, K.; Takahashi,
S.; Toyota, J.; Karino, Y.; Ikeda, K.; Ishikawa, H.; Watanabe, H.; McPhee, F.;
Hughes, E.; Kumada, H. Hepatology 2012, 55, 742; (d) Nettles, R. E.; Gao, M.;
Bifano, M.; Chung, E.; Persson, A.; Marbury, T. C.; Goldwater, R.; DeMicco, M. P.;
Rodriguez-Torres, M.; Vutikullird, A.; Fuentes, E.; Lawitz, E.; Lopez-Talavera, J.
C.; Grasela, D. M. Hepatology 1956, 2011, 54; (e) Fridell, R. A.; Wang, C.; Sun, J.-
H., ; O’Boyle, D. R., II; Nower, P.; Valera, L.; Qiu, D.; Roberts, S.; Huang, X.;
Kienzle, B.; Bifano, M.; Nettles, R. E.; Gao, M. Hepatology 1924, 2011, 54.
7. (a) Tellinghuisen, T.; Marcotrigiano, J.; Rice, C. M. Nat. Rev. Drug Discov. 2005,
435; (b) Love, R. A.; Brodsky, O.; Hickey, M. J.; Wells, P. A.; Cronin, C. N. J. Virol.
2009, 83, 4395; (c) Marcotrigiano, J.; Tellinghuisen, T. Methods Mol. Biol. 2009,
510, 85.
8. (a) Meanwell, N. A. J. Med. Chem. 2011, 54, 2529; (b) Patani, G. A.; LaVoie, E. J.
Chem. Rev. 1996, 96, 3147; (c) Burger, A. Prog. Drug Res. 1991, 37, 288.
9. (a) Meanwell, N. A. Chem. Res. Toxicol. 2011, 24, 1420; (b) Lipinski, C. A. J.
Pharmcol. Toxicol. Methods 2000, 44, 235; (c) Leeson, P. D.; Empfield, J. R. Annu.
Rep. Med. Chem. 2010, 45, 393; (d) Gleeson, M. P.; Hersey, A.; Montanari, D.;
Overington, J. Nat. Rev. Drug Discov. 2011, 10, 197; (e) Veber, D. F.; Johnson, S. R.;
Cheng, H.-Y.; Smith, B. R.; Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45,
2615.
Conversion of the terminal phenyl ring to a thiophene (see 41)
restored potent inhibitory activity, which is congruent with the
practice of using thiophene as a benzene bioisostere.8
A variety of five-membered heterocycles, 42–45, installed in a
fashion that maintained the optimal chain length were observed
to be less active than the thiophene isostere and showed a rela-
tively flat SAR profile with the exception of imidazole 45, which
exhibited poor inhibitory potency. A series of fused acyl aromatics
captured by 46–54 displayed diminished inhibitory activity. Sub-
stitution with the non-aromatic cyclopropylmethyl moiety gave
55, a modification that provided useful levels of antiviral activity.
Finally, it is noteworthy that 54 showed signs of genotype 1a inhib-
itory activity with an EC50 of ꢁ1
lM.
Since exploration of the SAR about the stilbene prolinamides
demonstrated that truncation was not a viable strategy and
symmetry was optimal for picomolar in vitro HCV replication inhi-
bition, there was concern that these high molecular weight mole-
cules might not have desirable PK properties.9 In order to
address these concerns, selective surveying was conducted. Table
4 shows PK properties for 5 analogs with genotype 1b replication