Discovery of thiophene-2-carboxylic acids as potent inhibitors of HCV NS5B polymerase and HCV subgenomic RNA replication. Part 2: Tertiary amides

Article abstract of DOI:10.1016/j.bmcl.2003.10.068

Further SAR studies on the thiophene-2-carboxylic acids are reported. These studies led to the identification of a series of tertiary amides that show inhibition of both HCV NS5B polymerase in vitro and HCV subgenomic RNA replication in Huh-7 cells. Structural insights about the bioactive conformation of this class of molecules were deduced from a combination of modeling and transferred NOE (trNOE) studies.

Full text of DOI:10.1016/j.bmcl.2003.10.068

Bioorganic & Medicinal Chemistry Letters 14 (2004) 797–800
Discovery of thiophene-2-carboxylic acids as potent inhibitors
of HCV NS5B polymerase and HCV subgenomic RNA
replication. Part 2: Tertiary amides
Laval Chan,* Oswy Pereira, T. Jagadeeswar Reddy, Sanjoy K. Das, Carl Poisson,
Marc Courchesne, Melanie Proulx, Arshad Siddiqui, Constantin G. Yannopoulos,
Nghe Nguyen-Ba, Caroline Roy, Daniel Nasturica, Christophe Moinet, Richard Bethell,
Martine Hamel, Lucille L’Heureux, Maud David, Olivier Nicolas,
Philippe Courtemanche-Asselin, Stephanie Brunette, Darius Bilimoria and Jean Bedard
Shire BioChem Inc., 275 Armand-Frappier, Laval, Quebec, Canada H7V 4A7
Received 19 August 2003; accepted 31 October 2003
Abstract—Further SAR studies on the thiophene-2-carboxylic acids are reported. These studies led to the identification of a series
of tertiary amides that show inhibition of both HCV NS5B polymerase in vitro and HCV subgenomic RNA replication in Huh-7
cells. Structural insights about the bioactive conformation of this class of molecules were deduced from a combination of modeling
and transferred NOE (trNOE) studies.
# 2003 Elsevier Ltd. All rights reserved.
The severe morbidity associated with infection by the
hepatitis C virus (HCV) together with the lack of an
efficacious and well-tolerated agent has generated
intense research efforts in the discovery of novel anti-
HCV chemical entities.1 In the preceding paper,2 the
discovery of a novel class of low molecular weight HCV
NS5B RNA dependent RNA polymerase inhibitors was
reported and several analogues such as 1 were also
inhibitory to the replication of HCV RNA subgenomic
replicons in Huh-7 cells (replicon assay). Structure–
activity relationship studies on the 5-phenyl and sulfon-
amide moieties have provided insights on the require-
ment for activity in both in vitro HCV polymerase and
in surrogate HCV replication assays. Only compounds
with substituents in the precise 2,3,5 arrangement
depicted in 1 showed consistent activity and it was
demonstrated that sulfonamides were optimum whereas
secondary amides, ureas and carbamates were less
active. Electron withdrawing substituents at the para
position of 5-phenyl were also beneficial as they also
served the purpose of reducing cytotoxicity and thereby
increasing selectivity indices. It also became apparent
that in vitro inhibition of NS5B polymerase did not
necessarily correlate with inhibition of HCV RNA
replication in the replicon cell line. The exact cause of
this discrepancy is not known but the artificial condi-
tions utilized for the in vitro assay (e.g., a homo-
polymeric template) and the association of enzyme with
viral and cellular proteins in the replication assay are
probably key factors. Recent reports have provided
evidence that the activity of HCV NS5B polymerase is
modulated by NS3, NS4B and NS5A non structural
proteins.3 These proteins are not present in the enzy-
matic assay; it is thus possible that in the biochemical
assay, the conformation and/or accessibility of the
binding site is different from that of the replication
model. As a result, the affinity of certain analogues for
the NS5B binding site may be different in the more
complex cellular environment. The goal of this study
was therefore 2-fold; the first objective was to conduct
further SAR studies to optimize the profile of the com-
pounds such as potency enhancement in both biochem-
ical and replication assays. Since analogues that were
active in both the enzymatic and cell-based assays could
be identified in the previous study, the second objective
was therefore to identify structural features that show
consistent and predictable correlation between these
two assays.
* Corresponding author: Tel.: +1-450-978-7802; fax: +1-450-978-
7777; e-mail: lchan@ca.shire.com
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.10.068

798
L. Chan et al. / Bioorg. Med. Chem. Lett. 14(2004) 797–800
N-Methylation of peptidic amides is an established tool
that has traditionally been employed in medicinal
chemistry to both, improve pharmacological proper-
ties, such as in vivo stability and to alter the overall
conformation and biological profile of peptides.4
Unfortunately, when a similar transformation was
applied to sulfonamide 1, N-methyl sulfonamide 2
showed a 5- to 7-fold loss in HCV polymerase inhibition
as well as in the cell-based assay indicating that there is
no beneficial effect in alkylating the sulfonamide nitro-
gen (Table 1). However, it was gratifying to observe that
when the amide nitrogen of p-toluamide 3 was methyl-
ated, tertiary amide 4 showed a 6-fold enhancement in
polymerase inhibition (Table 2). Furthermore, IC50 of 4
in replicon cells was 22 mM and no toxicity was
observed at 200 mM. It is also worthwhile to note that
the secondary amide is inactive in the replicon cell line
(IC50 >100 mM). Having thus identified a new direction
for the discovery for potential anti-HCV agents, the
optimization effort was therefore concentrated towards
identifying optimum amide as well as N-alkyl moieties.
The compounds in this study were prepared according
to the methods described in Scheme 1.
Although substitution of the amide nitrogen with
methyl resulted in a dramatic increase in potency of
para-toluamide 4, no such effect was observed with the
ortho and meta toluamides 5 and 6 indicating that the
beneficial effect of methylating the amide nitrogen was
limited to the presence of a para susbtituent. This is in
contrast to the sulfonamide class where a single ortho
methyl substitution was generally sufficient to provide
active entities in both enzymatic and cellular assays. In
addition, among all the para substitutions evaluated
only chloro 7 and methyl 4 retained activity, substitu-
tion with ethyl 8 or even trifluoromethyl 9 significantly
reduced activity against HCV polymerase. Unsub-
stituted benzamide 10 was also inactive underscoring
the necessity of having a substituent at the para posi-
tion. The chloro analogue was also equipotent to the
methyl analogue in the replication assay albeit with
higher toxicity. A meta methyl substituent in addition to
the para methyl group (analogue 11) was detrimental
to activity but the corresponding 2,4 disubstituted deri-
vative 12 had a similar profile to a single para methyl
substitution. Increasing the bulk of the alkyl substituent
on the amide nitrogen was in general beneficial for
activity. N-Ethyl derivative 13 was approximately equi-
potent to the corresponding methyl substituent, whereas
introduction of an isopropyl unit (14) resulted in a 4-
fold increase in activity in the replicon assay. It is also
interesting to note that no corresponding increase in the
inhibition of polymerase activity in the enzymatic assay
was observed for these compounds. Substitution with a
2,4-dichlorobenzamide moiety (15) which was found to
be important for activity in the phenylalanine series,5
also yielded an active compound.
Table 1. Activity of N-methylsulfonamide 2
R
HCV NS5B
polymerase
IC50 (mM)
Replicon
IC50 (mM)
Huh-7
CC50 (mM)
1
2
H
Me
1
5
5
35
57
115
Table 2. Activity against HCV polymerase and replicon cells10
The precise requirement of the substitution pattern on
the benzamide moiety indicates that this group is likely
to be accommodated in a structurally well-defined
pocket on the enzyme. Since all the amides described so
far in this study are aromatic and thus planar, the effect
of replacing the benzamides with a non-planar and
bulkier cyclohexylcarboxamide moiety was examined as
this can provide structural insight about the binding
site. As substitution at the para position was shown to
R1
R2
IC50 (mM)
HCV NS5B
polymerase
IC50 (mM)
Replicon
CC50 (mM)
Huh-7
3
4
5
6
7
8
9
10
11
12
13
14
15
H
4-MePh
4-MePh
3-MePh
2-MePh
4-ClPh
4-EtPh
4-CF3Ph
Ph
3,4-Me2Ph
2,4-Me2Ph
4-MePh
4-MePh
2,4-Cl2Ph
12
2
>50
21
1.2
>50
34
>32
>50
1.4
2
>100
22
>100
ND
19
>100
>50
ND
ND
16
12
5
8.9ꢀ2.4
0.6
0.3a
118
>200
>200
ND
88
ND
ND
ND
ND
160
Me
Me
Me
Me
Me
Me
Me
Me
Me
Et
100
Scheme 1. Preparation of compounds 3–18. Method: Compounds 3–
6, 8, 9, 11, and 12: (i) NaH, iodomethane DMF, 0 ꢁC!rt, 62%; (ii)
LiOH, THF/MeOH/H2O (3:2:1), 80 ꢁC, 73%; (iii) benzoyl chloride,
NaOH, dioxane, H2O, 0 ꢁC!rt. Compound 13 (i) 4-methylbenzoyl
chloride, (CH2Cl)2, reflux, 88%; (ii) NaH, iodoethane, DMF,
0 ꢁC!rt, 61%; (iii) LiOH, dioxane, H2O, rt, 30%. Compounds 7, 10,
14, 15, and 18; (i) NaH, iodoalkane, DMF, 0 ꢁC!rt; (ii) benzoyl
chloride, (CH2Cl)2, reflux; (iii) LiOH, THF/MeOH/H2O (3:2:1), rt.
Compounds 16–17: (i) NaH, 2-iodopropane, DMF, 0 ꢁC!rt, 32%; (ii)
4-methyl-1-cyclohexanecarboxylic acid, NCS, PPh3, (CH2Cl)2, reflux;
(iii) preparative TLC; (iv) LiOH, THF/MeOH/H2O (3:2:1), rt.
iPr
iPr
3
1.2
>200
62ꢀ10
123
16
iPr
1.5
17
18
iPr
iPr
24
9
>50
ND
69
15
a Real time PCR.

L. Chan et al. / Bioorg. Med. Chem. Lett. 14(2004) 797–800
799
be crucial for activity in both the cellular and enzymatic
assays, cyclohexyl analogues with a methyl substitution
at the 4-position were therefore selected. This modifi-
cation resulted in the identification of the most active
analogue of this study; trans isomer 16 had an IC50 of
0.6 mM in the inhibition of HCV subgenomic RNA
replication. This molecule had a selectivity index (CC50/
IC50) greater than 200 and did not show any inhibition
to human polymerases a, b and g (IC50 >100 mM). In
addition, stable cell lines expressing only the luciferase
gene or luciferase driven by either EMCV (Internal
Ribosome Entry Site, IRES) or HCV IRES were used
to ascertain that 16 is indeed specific to a HCV replic-
ation process in the replicon assay.6 No activity (IC50
>100 mM) was detected in any of the three stable cell
lines indicating that this compound is not targeting the
IRES-mediated translation or a luciferase gene product.
Real-time PCR analysis was also used to quantify the
amount of HCV subgenomic RNA produced and by
this method, IC50 was determined to be 0.3 mM. As for
the best benzamide analogue (14 and 15), the enzymatic
inhibition of the cyclohexylcarboxamide (IC50 1.5 mM)
did not correlate with the large gain of potency in the
cell-based assay. In contrast, both cis methyl analogue
17 and unsubstituted cyclohexyl analogue 18 were sig-
nificantly less potent than the trans isomer. This again
demonstrates that substitution at the 4-position is not
only necessary for activity but that group must maintain
an orientation that approximates the planar orientation
of a para substituted aromatic system.
erate molecular conformations7 of the secondary para-
toluamide analogue 3, N-Me amide 4 and the bulkier
N-isopropyl analogue 14. The first 200 conformers in
each search were then visually inspected for any shape
preferences among the analogues. Two overall shapes
were identified and Figure 1 depicts representative
examples, which are referred to as ‘closed’ or ‘open’
amide forms. The population of each of these two
shapes was also determined among the conformations
generated and the results are depicted in Table 3. In the
case of secondary amide 3, all the local minima gener-
ated were from the open or the fully extended form.
Introduction of an alkyl group on the amide nitrogen
resulted in the appearance of a number of closed forms.
Furthermore, as the bulk of the alkyl group became
more significant, the population became predominantly
of the closed mode; 77% of the local minima for the N-
isopropyl analogue 14 were of the closed form com-
pared to 37% for N-methyl analogue 4.
An enhancement in the inhibition of HCV subgenomic
RNA replication was observed although no corres-
ponding increase in polymerase inhibition was observed
in the biochemical assay. The N-isopropyl analogue was
thus 4-fold more potent than N-methyl analogue in the
replicon assay. It is thus possible that the bioactive
character of this class of compounds resembles the
closed form. In order to provide some experimental
basis for this hypothesis, a transferred NOE (trNOE)8
experiment was performed with 4 in the presence of
HCV NS5B polymerase/poly rA/oligo dT complex.
TrNOE’s were observed between the 5-phenyl aromatic
protons and the 3-benzamide protons as well as
between the 4-thiophene proton and the 3-benzamide
protons.9
In an attempt to provide a structural basis for the
remarkable effect of the N-alkyl substituent, a system-
atic conformational searching exercise was used to gen-
Since no trNOEs were detected in the absence of
enzyme, the bioactive character of this ligand most
likely adopts a ‘closed’ or concave shape in order to
allow positioning of the benzamide moiety within 5 A
to both thiophene and 5-phenyl moieties. The higher
potency of the bulkier N-isopropyl tertiary amide likely
arises from the larger population of the closed form in
solution. In addition, it appears that HCV NS5B poly-
merase in the replication complex opts for the closed
shaped ligand to a greater extent than the enzyme in the
biochemical assay. Further studies are under way to
exploit these findings.
Figure 1. Open and closed conformations of 4.
In summary, optimization studies led to the discovery
that N-alkylation of the amide nitrogen with a bulky
group and the use a cyclohexylcarboxamide moiety was
key in obtaining potent inhibitors of HCV sub-genomic
RNA replication. These findings provide useful leads
for further optimization in the search for a low-mole-
cular weight HCV therapy. Correlation between inhi-
bition of in vitro HCV NS5B polymerase activity and
the cell-based assay is not directly proportional but
nevertheless, a series that shows activity in both assays
was identified. The bioactive conformation is believed
to be of the ‘closed’ or concave form and is induced by
introduction of an alkyl substituent on the amide
nitrogen.
Table 3. Relationship between shape population and activity in the
replicon assay
R
% Closed Form
HCV Replication IC50 (mM)
3
4
14
H
Me
iPr
0
37
77
>100
22
5

800
L. Chan et al. / Bioorg. Med. Chem. Lett. 14(2004) 797–800
vidual structure at their ionization state that could be
Acknowledgements
reflected in aqueous conditions at ꢂpH 7. MOE (Chemi-
cal Computing Group Inc., Montreal, Canada) was used
to build and energy minimize each structure with the
MMFF94s force field using a gradient of 0.01 kcal/mol A.
The resulting minimized structures were used as starting
points for conformational searching conducted by using
the MOE implemented systematic conformational search
tool. To visualize the conformational preference of each
analogue, the first 200 lowest energy structures were
examined for each of the representative ‘closed’ or ‘open’
shapes. The average energy difference between the 1st and
200th structure was approximately 3 kcal/mol A.
We thank Jean Duchaine for providing the human
polymerase a, b and g inhibition data and Therese
Godbout for her assistance in the preparation of this
manuscript.
References and notes
1. Memon, M. I.; Memon, M. A. J. Viral Hepatitis 2002, 9,
84.
2. Chan, L.; Das, S. K.; Reddy, T. J.; Poisson, C.; Proulx,
M.; Pereira, O.; Courchesne, M.; Roy, C.; Wang, W.;
Siddiqui, A.; Yannopoulos, C. G.; Nguyen-Ba, N.; Lab-
recque, D.; Bethell, R.; Hamel, M.; Courtemanche-Asse-
lin, P.; L’Heureux, L.; David, M.; Nicolas, O.; Brunette,
S.; Bilimoria, D.; Bedard, J. Bioorg. Med. Chem. Lett.
2004, 14, preceding paper. doi:10.1016/j.bmcl.2003.10.067.
3. (a) Piccininni, S.; Varaklioti, A.; Nardelli, M.; Dave, B.;
Raney, K. D.; McCarthy, J. E. G. J. Biol. Chem. 2002,
277, 45670. (b) Shirota, Y.; Luo, H.; Qin, W.; Kaneko, S.;
Yamashita, T.; Kobayashi, K.; Murakami, S. J. Biol.
Chem. 2002, 277, 11149.
4. Dechantsreiter, M. A.; Planker, E.; Mathae, B.; Lohof,
E.; Hoelzemann, G.; Jonczyk, A.; Goodman, S. L.; Kess-
ler, H. J. Med. Chem. 1999, 42, 3033.
5. Chan, L.; Reddy, T. J.; Proulx, M.; Das, S. K.; Pereira,
O.; Wang, W.; Siddiqui, A.; Yannopoulos, C.; Poisson,
C.; Turcotte, N.; Drouin, A.; Alaoui, H. M. A.; Bethell,
R.; Hamel, M.; Bilimoria, D.; L’Heureux, L.; Nguyen-Ba,
N. J. Med. Chem. 2003, 46, 1283.
8. (a) Clore, G. M.; Gronenborn, A. M. J. Magn. Reson.
1982, 48, 402. (b) Ni, F.; Scheraga, H. A. Acc. Chem. Res.
1994, 27, 257.
9. Details of this experiment will be reported in a forth-
coming publication.
10. Polymerase IC50’s were determined from dose–response
curves using 11 concentrations for each compound in
uniplicate. Curves were fitted to data points using non-
linear regression analysis and IC50’s were interpolated
from the resulting curves using GraphPad Prism software,
version 2.0 (GraphPad Software, Inc. San Diego, CA,
USA). A positive control was included as an internal
standard in each set of experiments and results were con-
sidered accurate only when the IC50 value of the positive
control was within 0.45ꢀ0.16 mM. Similarly, IC50’s in
replicons were determined using 12 concentrations in dupli-
cate and 15 was included as positive control in each
experiment and data were considered accurate when IC50
of 15 was within 8.9ꢀ2.4 mM. CC50’s in Huh-7 Cells
were determined from five concentrations in quad-
ruplicate and data were taken when CC50 of 15 was
within 62ꢀ10 mM.
6. Bartenschlager, R. Nat. Rev. Drug Discov. 2002, 1, 911.
7. All three analogues were studied by modeling each indi-