Synthesis and SAR studies of novel heteroaryl fused tetracyclic indole-diamide compounds: Potent allosteric inhibitors of the hepatitis C virus NS5B polymerase

Article abstract of DOI:10.1016/j.bmcl.2012.02.063

Presented here are initial structure-activity relationship (SAR) studies on a series of novel heteroaryl fused tetracyclic indole-based inhibitors of the hepatitis C viral polymerase, NS5B. The introduction of alternative heterocyclic moieties into the indolo-fused inhibitor class significantly expands the reported SAR and resulted in the identification of pyridino analogs, typified by compounds 44 and 45 that displayed excellent potency against the NS5B polymerase of both HCV 1a and HCV 1b genotypes.

Full text of DOI:10.1016/j.bmcl.2012.02.063

Bioorganic & Medicinal Chemistry Letters 22 (2012) 2866–2871
Contents lists available at SciVerse ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and SAR studies of novel heteroaryl fused tetracyclic
indole-diamide compounds: Potent allosteric inhibitors
of the hepatitis C virus NS5B polymerase
Min Ding a, , Feng He a, Thomas W. Hudyma a, Xiaofan Zheng a, Michael A. Poss a, John F. Kadow c,
⇑
Brett R. Beno a, Karen L. Rigat b, Ying-Kai Wang b, Robert A. Fridell b, Julie A. Lemm b, Dike Qiu b,
Mengping Liu b, Stacey Voss b, Lenore A. Pelosi b, Susan B. Roberts b, Min Gao b, Jay Knipe a,
Robert G. Gentles a
a Molecular Sciences and Candidate Optimization, Bristol-Myers Squibb, Research and Development, 5 Research Parkway, Wallingford, CT 06492, United States
b Department of Virology, Bristol-Myers Squibb, Research and Development, 5 Research Parkway, Wallingford, CT 06492, United States
c Medicinal Chemistry, Bristol-Myers Squibb, Research and Development, 5 Research Parkway, Wallingford, CT 06492, United States
a r t i c l e i n f o
a b s t r a c t
Article history:
Presented here are initial structure–activity relationship (SAR) studies on a series of novel heteroaryl
fused tetracyclic indole-based inhibitors of the hepatitis C viral polymerase, NS5B. The introduction of
alternative heterocyclic moieties into the indolo-fused inhibitor class significantly expands the reported
SAR and resulted in the identification of pyridino analogs, typified by compounds 44 and 45 that dis-
played excellent potency against the NS5B polymerase of both HCV 1a and HCV 1b genotypes.
Ó 2012 Elsevier Ltd. All rights reserved.
Received 6 January 2012
Accepted 21 February 2012
Available online 1 March 2012
Keywords:
HCV NS5B polymerase inhibitors
Tetracyclic indoles
The hepatitis C virus (HCV) currently infects about 180 million
people globally,1 and frequently results in serious and often fatal
liver disease. The standard of care (SOC) for the treatment of
HCV infection is expected to evolve significantly in coming years,
from the current protocol of co-administration of pegylated inter-
feron-a and ribavirin, (a regime that provides limited efficacy
against HCV genotypes 1a and 1b, and is associated with significant
adverse side-effects2), to protocols incorporating more selective
agents, such the newly approved HCV specific protease inhibitors
VICTRELIS™ and INCIVEK™.3 In addition to the viral protease, the
HCV non-structural protein 5B (NS5B) is an RNA-dependent RNA
polymerase that is essential for viral replication. Several active-site
and allosteric NS5B inhibitors have been reported, and recent clin-
ical data4 on both inhibitor classes have provided support for NS5B
as a viable target for HCV therapy, with inhibitors of this enzyme
expected to have utility in a number of potential combination
therapies.
Figure 1. Aryl-fused tetracyclic indolo NS5B inhibitors 1.5 Generic structure of
fused heterocyclic analogs, 2. [IC50 values are the means of at least 2 experiments
and were obtained using a previously reported enzymatic assay.7]
Described herein is our initial investigation into a series of indole
based inhibitors related to the previously reported indolo analog 1.
Compounds structurally related to 1 have been demonstrated to
bind in the ‘thumb’ domain of NS5B, as evidenced by a consistent
P495L resistance profile, and more definitively, by a reported co-
crystal structure.5 A focus of our recent research6 in this area has
been the evaluation of indolo ring systems in which the fused aryl
moiety in 1 has been replaced with a variety of heterocycles, as
shown by 2 in Figure 1.
It can be anticipated that such substitutions would extend the
range of physicochemical properties displayed by these ring
systems and significantly expand the SAR reported in this area.
⇑
Corresponding author.
E-mail address: min.ding@bms.com (M. Ding).
0960-894X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2012.02.063

M. Ding et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2866–2871
2867
Figure 2. Novel heteroaryl-fused, indoloazepine, indolodiazepinone and indolooxazepine NS5B inhibitor classes.
Several of the heterocyclic intermediates 3–7 required for these
studies are shown in Figure 2.8
Heck strategy that we have discussed previously.8 Correspond-
ingly, we developed the alternative methodology depicted below
that utilizes an intermolecular Mitsunobu reaction to construct
the oxazepine ring. In this reaction sequence, 3-bromo-2-hydroxy-
pyridine (31) is first derivatized as the O-benzyl ether 32. This
compound was coupled under Suzuki conditions with the pinacol
boronate ester 10 to afford the pyridinyl derivative 33 in good
yield. This intermediate was alkylated using benzyl 2-bromoethy-
lether to give the di-ether 34. Subsequent hydrogenolysis afforded
the hydroxyl-pyridone 35 that under Mitsunobu conditions cy-
clized to give the target indolo-pyridooxazepine indole 36. This
was then hydrolyzed to 37 in moderate overall yield.
The fused thiazolo derivative 9 was prepared in a similar manner
to that used to access compounds 3 and 4, and utilized the previ-
ously reported8 indoloazepinone intermediate 8 in a condensation
reaction with thiourea to give the target thiazolo derivative 9 in
essentially quantitative yield, as shown in Scheme 1.
Additional pyrazino- and imidazolo-fused indoloazepine inter-
mediates 16 and 23 were prepared as shown in Scheme 2. In the
case of the former, this involved an initial Suzuki coupling of 2,3-
dichloropyrazine 11 with the pinacol boronate derivative 10.8
A
subsequent Stille reaction with tributylvinyltin furnished the vinyl
intermediate 13, that was then alkylated using allylbromide to give
the di-olefinic indole 14 in good overall yield. 14 was converted to
the indolo-fused pyrazinoazepine 15 using the Grubb’s metathesis
reaction employing a second generation catalyst, and the desired
target 16 was obtained by a final catalytic hydrogenation and basic
hydrolysis. The related imidazo derivative 23 was prepared in an
essentially analogous fashion using the SEM-protected 4,5-diiodo-
imidazole intermediate (18), prepared as shown in Scheme 2.
A further furano-fused intermediate 30 was prepared as shown
in Scheme 3 using a reaction sequence similar to that outlined
above. 3-Bromofuran (24) was formylated using LDA and DMF to
give intermediate 25. This was then subjected to Suzuki coupling
with the pinacol boronate 10 to afford the di-substituted furan
26. This compound was used without additional protection in a
Wittig reaction using an excess of base and triphenylphosphonium
bromide to give 27. Alkylation of 27 with allylbromide followed by
olefin metathesis under conditions described above provided the
indolo-fused compound 29. This was converted to the desired sat-
urated target 30 by catalytic hydrogenation and basic hydrolysis.
The synthesis of the final heterocyclic intermediate included in
this study, the indolo-pyridooxazepine 37, was prepared using the
chemistry depicted in Scheme 4. Although 37 is a regioisomer of
compound 6, it cannot be accessed by the same intramolecular
With all of the key intermediates in hand, each was subsequently
converted to the desired (E)-3-(4-(1-aminocyclopentanecarboxam-
ido)phenyl)acrylic acid carboxamide9 using the methodology
shown in Scheme 5. Each of the core heterocyclic acids was first
condensed with (E)-methyl 3-(4-(1-aminocyclopentanecarboxa-
mido)phenyl)acrylate using TBTU as coupling agent with the final
target acids obtained by basic hydrolysis with sodium hydroxide,
except in the case of 45. Compound 45 was obtained by the use of
LiI in pyridine as the lactam moiety in the molecule was sensitive
to base-catalyzed hydrolysis conditions. Compound 40 was ob-
tained by treating the SEM-protected imidazole precursor with
TBAF in THF. The structures of the final products are shown in
Scheme 6.
Table 1 lists the activities of these compounds against both
wild-type and P495L mutant genotype 1b NS5B enzymes, as well
as their activities in HCV genotype 1a and 1b replicon systems.10
In general, all of the compounds showed good concordance be-
tween their activities against the isolated wild-type enzyme and in
both replicon systems. The analogs also consistently displayed a
reduced activity toward the P495L mutant enzyme. Given this,
and their close structural relationship to compounds known to
bind in the well-studied5,6,11 NS5B inhibitor binding pocket proxi-
mal to P495, it would appear reasonable to assume that these com-
pounds bind in a similar fashion, vide infra.
In terms of discernable SAR’s, the two fused thiazoles 41 and 42
and the imidazole analog 40 show a significant reduction in po-
tency relative to the phenyl analog 2. In relation to 41 and 42, each
has a pendant functionality that can encounter a steric clash with
the enzyme with an obvious negative impact on potency, vide in-
fra. This is in contrast to the two other oxygen containing five-
membered rings 38 and 39 that display similar potency to 2. With
respect to imidazole 40, reported12 experimental pKa values of the
conjugate acids of dialkylsubstituted imidazoles are between 7.2
and 8.7. Correspondingly, this compound may be significantly
Scheme 1. Synthesis of compound 9.

2868
M. Ding et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2866–2871
Scheme 2. Reagents and conditions: (a) Pd(PPh3)4, LiCl, Na2CO3, EtOH/toluene, 80 °C, 41–43%; (b) tributylvinyltin, PdCl2(PPh3)2, LiCl, DMF, 100 °C, 29–88%; (c) KH, allyl
bromide, 75%; (d) Grubbs catalyst 2nd generation, CH2Cl2, reflux, 64–100%; (e) H2/Pd-C, 84–91%; (f) NaOH, MeOH/THF, 87–100%; (g) SEMCl, K2CO3, DMF, 92%.
Scheme 3. Reagents and conditions: (a) LDA, DMF, À78 °C to rt, 60%; (b) 10, Pd(PPh3)4, LiCl, Na2CO3, EtOH/toluene, 80 °C, 50%; (c) n-BuLi, methyl triphenylphosphonium
bromide, 65 °C, 54%; (d) KH, allyl bromide, 67%; (e) Grubbs catalyst 2nd generation, CH2Cl2, reflux, 85%; (f) H2/Pd-C, 80%; (g) NaOH, MeOH/THF, 90%.
ionized under our assay conditions (pH 7.2–7.5), and this may ac-
count for the observed reduction in potency.
The six-membered isosteric fused heterocyclic systems 43, 44,
46 and 47 are essentially equipotent with the phenyl analog 2, as

M. Ding et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2866–2871
2869
Scheme 4. Reagents and conditions: (a) NaH, benzyl bromide, DMF, 52%; (b) Pd(PPh3)4, LiCl, Na2CO3, EtOH/toluene, 80 °C, 70%; (c) NaH, benzyl 2-bromoethyl ether, DMF,
49%; (d) H2/Pd-C, 90%; (e) PPh3, DBAD, THF, 85 °C, 58%; (f) NaOH, MeOH/THF, 65%.
type 1a (and correspondingly, can only speculate on the origin of
the observed potency differences).
However, we have previously reported the co-crystal structure
of compound 48 with a genotype 1b NS5B construct,13 the model
of which is shown in Figure 3.
In an effort to elucidate the origin of the reduced potency of 41
and 42 relative to the most potent analog 47, the A-chain from the
NS5B 1b/48 co-crystal structure (PDB: 3Q0Z) was used to model
the binding of these three compounds to NS5B. The initial NS5B
1b/48 model was prepared by assigning bond orders and adding
hydrogen atoms to the atoms associated with the A-chain in the
co-crystal structure followed by optimization of the hydrogen
bonding network. These operations were performed with the Pro-
tein Preparation Wizard16 in Maestro™ v. 9.2.17 The resultant mod-
el was then subjected to 500 steps of PRCG18 minimization using
the OPLS2005 force-field19 and implicit water solvation20 with
the positions of all non-hydrogen atoms fixed. Two additional
rounds of 100 steps of steepest descent energy minimization were
then performed. In the first, the positions of the protein backbone
atoms were fixed, and in the second round, no positional restraints
were imposed. All energy minimization calculations utilized Mac-
roModelÒ v. 9.9.21 Models of compounds 41, 42, and 47 in their
putative bioactive conformations were then created by manual
editing of the structure of 48 from the refined complex model in
Scheme 5. Reagents and conditions: (a) TBTU, diisopropylethyl amine, DMF, 30–
Maestro™.17 Models of the complexes of 1b NS5B with compounds
87%; (b) NaOH, MeOH, THF, H2O, 90 °C, microwave, 26–86%; (c) LiI, pyridine, 150 °C,
41, 42, and 47 were then each energy minimized (OPLS2005, impli-
69%.
cit water solvation, 500 steps PRCG) with the positions of all atoms
except those of the ligands held fixed. Figure 4 shows the three
is analog 45 which possesses a more conformationally constrained
bridging element. Collectively these molecules demonstrate that
polarity can be tolerated both in the bridge as well as the fused
heterocyclic moieties. Interestingly, analogs 46 and 47 are isomeric,
differing only in the placement of a nitrogen atom. 47 displays
similar activity to the other six-membered heterocycles, but its
positional isomer 46 shows a significant reduction in potency
against genotype 1a. To date, we have not been able to obtain a
co-crystal of any of the analogs discussed above with NS5B geno-
compounds in the NS5B 1b P495 binding site after energy minimi-
zation. Both of the less potent compounds 41 and 42 are displaced
slightly upward and sit less deeply in the binding pocket compared
to potent analog 47. Furthermore, the methyl and amino substitu-
ents of 41 and 42, respectively, make close contacts with Ala393
and/or Leu492. This suggests that the relatively poor potency of
the two thiazole analogs may be a result of steric interactions
between the substituents and the protein which prevent the poly-
cyclic cores from achieving optimal contact with the enzyme. In

2870
M. Ding et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2866–2871
Table 1
Inhibitory activities of indolo-fused heterocycles as determined in biochemical assays
with NS5B wild-type and P495L mutant enzymes. Inhibitory activities in HCV
replicon 1b and 1a systems are also provided
Final product
acid
Enzyme WT
Enzyme P495L
Replicon 1b/1a
EC50 (lM)
a
a
a
IC50
(
lM)
IC50
(lM)
2
0.034
0.044
0.034
0.228
0.108
0.250
0.024
0.030
0.024
0.015
0.014
3.39
5.85
3.72
ND
5.1
4.84
2.28
1.731
3.1
>0.6
2.80
0.014/0.071
0.078/0.380
<0.069/ND
0.738/2.51
0.683*/4.19*
1.81/3.78
0.024/0.089
0.009/0.035
0.012/0.039
0.022/0.157
0.021/0.028
38
39
40
41
42
43
44
45
46
47
a
Values are means of at least three experiments, unless otherwise noted. Com-
pounds 2, 44 and 45 had CC50’s of 30, 63 and 45 lM, respectively in the replicon
system.
*
Data from a single test occasion.
Figure 3. The co-crystal structure of compound 48 bound to the thumb domain of
HCV genotype 1b NS5B (Bartenschlager).14 The proximity of the binding site to
P495 is clearly shown. The binding site is represented by a surface (gray) and
cartoon (green). The side chains of Pro495, Arg498, and Arg503 and compound 48
are represented by sticks (cyan and yellow carbon atoms, respectively). Image
created with PyMOL.15
Scheme 6. Final product acids that were examined in enzymatic and replicon assay
systems.
the case of 42, the presence of the polar amino moiety in a rela-
tively hydrophobic protein environment may also contribute to
the observed sub-optimal potency.
A number of compounds from this series were investigated in
pharmacokinetic (PK) studies in the rat, with the data on two
analogs, 44 and 45 compiled in Table 2. Compound 44 displayed
Figure 4. Compounds 41, 42, and 47 (shown with purple, orange, and green carbon
atoms, respectively) are shown modeled into the P495 binding site in NS5B 1b.
Potent analog 47 sits more deeply in the binding pocket than either of the less
potent analogs which also have unfavorable steric interactions with Ala393 and/or
Leu492. Figure created with Maestro™.17
low oral bioavailability (4%) with an AUC of 1.7 lM h, but had
relatively low clearance characteristics (CL = 7.2 mL/min/kg; T1/
2 = 1 h), suggesting the molecule suffered from poor absorption.
Compound 45 had essentially no oral bioavailability and dis-
played similar clearance characteristics to 44 (CL = 8.7 mL/min/
kg). We attributed this lack of oral exposure to a combination
of poor solubility and membrane permeability. [In a Caco-2
assay designed to determine if these analogs were substrates
for P-gp, their poor solubility prevented the generation of inter-

M. Ding et al. / Bioorg. Med. Chem. Lett. 22 (2012) 2866–2871
2871
Table 2
Lemm, J.; Qiu, D.; Liu, M.; Voss, S.; Pelosi, L.; Roberts, S. B.; Gao, M.; Knipe, J.;
Gentles, R. G. Bioorg. Med. Chem. Lett. 2011, 21, 2925.
6. Hudyma, T. W.; Zheng, X.; He, F.; Ding, M.; Bergstrom, C. P.; Hewawasam, P.;
Martin, S. W.; Gentles, R. G. U.S. Patent 7,153,848, 2006.
7. (a) Wang, Y.-K.; Rigat, K. L.; Roberts, S. R.; Gao, M. Anal. Biochem. 2006, 106; (b)
Wang, Y.-K.; Rigat, K. L.; Sun, J.-H.; Roberts, S. R.; Gao, M. Arch. Biochem. Biophys.
2008, 470, 146.
8. Ding, M.; He, F.; Poss, M. A.; Rigat, K. L.; Wang, Y.-K.; Roberts, S. B.; Qiu, D.;
Fridell, R. A.; Gao, M.; Gentles, R. G. Org. Biomol. Chem. 2011, 9, 6654.
9. Beaulieu, P. L.; Dansereau, N.; Duan, J.; Garneau, M.; Gillard, J.; McKercher, G.;
LaPlante, S.; Lagacee, L.; Thauvette, L.; Kukolj, G. Bioorg. Med. Chem. Lett. 1825,
2010, 20.
Rat PK results observed with select examples of the indolo-fused tetracyclic NS5B
inhibitors shown in Scheme 6
Compound
Fa (%)
AUC (
l
M h)
Clearanceb (mL/min/kg)
44
45
4
<1
1.7
7.2
8.7
Not detected
The above results are the average values calculated from n = 3. Vehicle iv and po
PEG400.
a
Oral bioavailability.
Iv (dose 2 mg/kg).
b
10. (a) Fridell, R. A.; Qiu, D.; Wang, C.; Valera, L.; Gao, M. Antimicrob. Agents
Chemother. 2010, 54, 3641; (b) O’Boyle, D. R. I. I.; Nower, P. T.; Lemm, J. A.;
Valera, L.; Sun, J.-H.; Rigat, K.; Colonno, R.; Gao, M. Antimicrob. Agents
Chemother. 2005, 49, 1346–1353.
pretable data.] Addressing this issue will be the subject of a fu-
ture manuscript.
11. (a) Beaulieu, P. L.; Fazal, G.; Kukolj, G.; Jolicoeur, E.; Gillard, J.; Poupart, M.-A.;
Rancourt, J. Int. Patent Appl. 2003, WO 03010140.; (b) Ikegashira, K.; Oka, T.;
Hirashima, S.; Noji, S.; Yamanaka, H.; Hara, Y.; Adachi, T.; Tsuruha, J.-I.; Doi, S.;
Hase, Y.; Noguchi, T.; Ando, I.; Ogura, N.; Ikeda, S.; Hashimoto, H. J. Med. Chem.
2006, 49, 6950; (c) Beaulieu, P. L.; Gillard, J.; Bykowski, D.; Brochu, C.;
Dansereau, N.; Duceppe, J.-S.; Hache, B.; Jakalian, A.; Lagace, L.; LaPlante, S.;
McKercher, G.; Moreau, E.; Perreault, S.; Stammers, T.; Thauvette, L.;
Warrington, J.; Kukolj, G. Bioorg. Med. Chem. Lett. 2006, 16, 4987.
12. Lenarcik, B.; Ojczenasz, P. J. Heterocycl. Chem. 2002, 39, 287–290.
13. Lohmann, V.; Körner, F.; Koch, J. O.; Herian, U.; Theilmann, L.; Bartenschlager,
R. Science 1999, 285, 110.
In summary, we have described the syntheses of a series of
novel heteroaryl fused tetracyclic indole-diamide compounds and
have provided data on their inhibition of NS5B. Pyridino-fused ana-
logs with different bridging elements displayed excellent potency
against both HCV 1a and HCV 1b genotypes, with compounds 44
and 45 demonstrating EC50’s of 9 and 12 nM, respectively against
1b.
14. PDB: 3Q0Z.
References and notes
15. The PyMOL Molecular Graphics System, version 1.3, Schrödinger, LLC: New
York, NY, 2011.
16. Schrödinger Suite 2011 Protein Preparation Wizard; Epik version 2.2,
Schrödinger, LLC: New York, NY, 2011; Impact version 5.7, Schrödinger, LLC:
New York, NY, 2011; Prime version 3.0, Schrödinger, LLC: New York, NY, 2011.
17. Maestro, version 9.2, Schrödinger, LLC: New York, NY, 2011.
18. Polak, E.; Ribiere, G. Rev. Fr. Inf. Rech. Oper., Serie Rouge 1969, 16, 35.
19. Jorgensen, W. L.; Maxwell, D. S.; Tirado-Rives, J. J. Am. Chem. Soc. 1996, 118,
11225.
20. Still, W. C.; Tempczyk, A.; Hawley, R. C.; Hendrickson, T. A. J. Am. Chem. Soc.
1990, 112, 6127.
21. MacroModel, version 9.9, Schrödinger, LLC: New York, NY, 2011.
1. (a) Choo, Q.-L.; Kuo, G.; Weiner, A. J.; Overby, L. R.; Bradley, D. W.; Houghton, M.
Science 1989, 244, 359; (b) Lavanchy, D. Liver Int. 2009, 29(Suppl. 1), 74.
2. Farnik, H.; Mihm, U. A.; Zeuzem, S. Liver Int. 2009, 29, 23.
3. Ghany, M. G.; Nelson, D. R.; Strader, D. B.; Thomas, D. L.; Seeff, L. B. Hepatology
2011, 54, 1433.
4. (a) Brown, N. A. Expert Opin. Investig. Drugs 2009, 18, 709; (b) Beaulieu, P. L.
Expert Opin. Ther. Patents 2009, 19, 145; (c) Lin, K.; Hazuda, D. J.; Otto, M. J.
Expert Opin. Investig. Drugs 2010, 19, 963.
5. Zheng, X.; Hudyma, T. W.; Martin, S. W.; Bergstrom, C. P.; Ding, M.; He, F.;
Romine, J.; Poss, M. A.; Kadow, J. F.; Chang, C.-H.; Wan, J.; Witmer, M. R.; Morin,
P.; Camac, D. M.; Sheriff, S.; Beno, B. R.; Rigat, K. L.; Wang, Y.-K.; Fridell, R.;