HCV NS5B polymerase-bound conformation of a soluble sulfonamide inhibitor by 2D transferred NOESY

Article abstract of DOI:10.1016/j.bmcl.2004.08.018

HCV NS5B RNA-dependent RNA polymerase (NS5B) is essential for viral replication and is therefore considered a target for antiviral drug development. From our ongoing screening effort in the search for new anti-HCV agents, a novel inhibitor 1 with low μM a

Full text of DOI:10.1016/j.bmcl.2004.08.018

Bioorganic & Medicinal Chemistry Letters 14 (2004) 5333–5337
HCV NS5B polymerase-bound conformation of a soluble
sulfonamide inhibitor by 2D transferred NOESY
Constantin G. Yannopoulos,a,* Ping Xu,b Feng Ni,b Laval Chan,a Oswy Z. Pereira,c
T. Jagadeeswar Reddy,c Sanjoy K. Das,a Carl Poisson,a Nghe Nguyen-Ba,a
a
Nathalie Turcotte, Melanie Proulx, Lilianne Halab, Wuyi Wang, Jean Bedard,
c
a
c
a
´
Nicolas Morin,c Martine Hamel,c Olivier Nicolas,a Darius Bilimoria,a Lucille LÕHeureux,a
Richard Bethellc and Gervais Dionnea
aViroChem Pharma Inc., 275 Armand-Frappier Blvd., Laval, QC, Canada H7V 4A7
bBiotechnology Research Institute, 6100 Royalmount Avenue, Montreal, QC, Canada H4P 2R2
cShire BioChem Inc., 275 Armand-Frappier Blvd., Laval, QC, Canada H7V 4A7
Received 14 June 2004; revised 9 August 2004; accepted 9 August 2004
Available online 8 September 2004
Abstract—HCV NS5B RNA-dependent RNA polymerase (NS5B) is essential for viral replication and is therefore considered a tar-
get for antiviral drug development. From our ongoing screening effort in the search for new anti-HCV agents, a novel inhibitor 1
with low lM activity against the HCV NS5B polymerase was identified. SAR analysis indicated the optimal substitution pattern
required for activity, for example, carboxylic acid group at 2-position of thiophene ring. We describe the steps taken to identify
and solve the bioactive conformation of derivative 6 through the use of the transferred NOE method (trNOE).
Ó 2004 Elsevier Ltd. All rights reserved.
Transferred NOE1 (trNOE) is increasingly used in the
drug discovery2 process, as it provides structural insights
about the conformation of an inhibitor bound to an
appropriate macromolecular target. In order to observe
trNOEs, the inhibitor has to undergo rapid on/off
exchange between the free and target-bound state mak-
ing this method particularly attractive during the early
steps of a drug discovery program. We have successfully
applied the trNOE method and provided structural
insight to our ongoing development of antiviral agents
against the Hepatitis C virus (HCV). The urgent need
of new anti-HCV agents is accentuated by increasing
global infection levels to more that 170 million and the
lack of effective therapies or vaccines. In addition, the
incidence increases by 3–4 million per year,3 and it is
estimated that 80% of those infected will develop
chronic infection, and about 20% and 5% will develop
cirrhosis and hepatocellular carcinoma, respectively.4,5
HCV is a member of the Flaviviridae family and is a sin-
gle-stranded positive RNA6 virus that encodes a large
polypeptide (9.6kb; about 3000 amino acids). Complete
polypeptide maturation yields structural C, E1, E2 and
non-structural NS2, NS3, NS4A, NS4B, NS5A, NS5B
proteins. It has been demonstrated that the virally
encoded NS5B RNA-dependent RNA polymerase is
essential for viral replication7–9 and is therefore consid-
ered a viable target for antiviral drug development.
Three groups have solved independently the NS5B crys-
tal structure10–12 and no bioactive solution conforma-
tion has been reported for any of the known
inhibitors.13 Recently, we have reported the first X-ray
structure of a N,N-disubstituted phenylalanine bound
to NS5B polymerase at an allosteric site14 followed by
structure–activity relationship (SAR) studies.15a,b Our
ongoing efforts in the search for new anti-HCV agents
resulted in the discovery of a novel inhibitor 1 with
low lM activity against HCV polymerase. SAR analysis
indicated that the carboxylic acid at position 2 of the
thiophene was beneficial for activity; carboxylic acid 2
was almost threefold more potent than the correspond-
ing carboxamide 1. The complete details of our optimi-
zation program are described elsewhere.15c,d
Keywords: HCV; NS5B Polymerase; NMR; Inhibitor.
*
Corresponding author. Tel.: +1 450 978 6435; fax: +1 450 978
7777; e-mail: cyannopoulos@virochempharma.com
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.08.018

5334
C. G. Yannopoulos et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5333–5337
NS5B activity (IC50 4lM). The solubility of 5 was sim-
CONH2
COOH
2
ilar to that of 3 but the complex stability was still less
than 2h. Weaker binders such as 4 (IC50 15lM) and 6
(IC50 25lM) were then evaluated. DLB was observed
for both analogues and surprisingly the weakest pyridyl
analogue 6 exhibited the best NS5B induced DLB. In
addition, the ternary complex was stable for almost 4h
and therefore 6 was chosen for NMR data collection.
S
S
H
H
N
N
3
5
S
Cl
S
Cl
O
O
O
O
1
2
HCV NS5B IC50 14µM
HCV NS5B IC50 5µM
Key trNOE correlations between the pyridyl ring proton
[a] (see labeled structure in Fig. 1), phenyl protons
[f,g,h], and the thiophene proton [e] can be clearly seen
In this report, we describe the steps taken to determine
the bioactive conformation of sulfonamide derivative 6
through the use of trNOE. Weakly binding sulfonamide
2 was chosen as a starting point for screening by 1D pro-
ton NMR differential line broadening (DLB) as a meas-
ure of specific complexation. Although 2 exhibited DLB
upon addition of the HCV NS5B polymerase pre-com-
plexed with poly[rA]/oligo[dT] template-primer,16 no
trNOEs were observed. Low concentration of the bound
state probably resulted in undetectable trNOEs due to
the observed low solubility of the inhibitor-polymerase
complex. It is interesting to note that in the absence of
the template-primer, weak DLB was observed and no
DLB resulted from template-primer in absence of
NS5B. Therefore, the presence of template-primer
appears to play an important role in the integrity of
the binding site.
1
in the 2D water flip-back H NOESY spectrum of 6
Synthetic efforts (Schemes 1 and 2) were therefore
undertaken to increase the solubility of this class of
inhibitors by introducing a pyridyl ring on either the
5- or 3-positions of the thiophene ring. Introducing a
2-pyridyl function at the 3-position of the thiophene ring
(Compound 3) retained weak anti-NS5B activity (IC50
4lM) and increased solubility by 10-fold. Unfortunately
no trNOEs were observed due to extensive protein pre-
cipitation within 2h after complex formation.
We therefore explored the possibility of introducing a 4-
pyridyl function at the 5-position of the thiophene ring
and pyridine analog 5 was found to retain weak anti-
Figure 1. The water flip-back 2D-NOESY for bound 6 at 800MHz
and 298K, with a mixing time of 50ms. The boxed NOE cross-peaks
appear upon NS5B binding and are of same sign as the diagonal peaks.
O
O
OH
O
Method A
S
2: R1=H; R2=R4=Me; R3=Cl; X=CH
H
S
N
3: R1=F; R2=R3=R4=H; X=N
4: R1=H; R2=R4=H; R3=CF3; X=N
6: R1=H; R2=R3=R4=H; X=N
O
Method B
R1
NH2
S
O
R2
X
R
R4
R= H,F
R3
Scheme 1.
O
O
O
O
OH
OH
OtBu
O
a
c
e
S
S
S
S
H
H
N
H
N
N
d
f
b
NH2
O
O
O
O
Br
O
S
S
S
N
O
5
Scheme 2. Reagents and conditions: Method A (i) LiOH, dioxane, H2O, reflux; (ii) ArSO2Cl, Na2CO3, dioxane, H2O. Method B (i) ArSO2Cl,
DMAP, CH2Cl2, H2O; (ii) LiOH, dioxane, H2O, reflux. (a) o-Toluenesulfonyl chloride, Et2O, DMAP, CH2Cl2. (b) LiOH, dioxane. (c) 2-Propene,
H2SO4, dioxane, 60°C. (d) LDA, (BrCF2)2, À78°C. (e) 4-Pyridyl-B(OH)2, Pd(PPh3)4. (f) TFA, CH2Cl2, rt.

C. G. Yannopoulos et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5333–5337
5335
(boxed cross-peaks) shown in Figure 1. Pyridyl ring pro-
ton [d] also correlated with phenyl [f,g,h] and thiophene
[e] protons, but to a weaker extent as compared to pro-
ton [a] correlations. This observation was key in deduc-
ing the orientation of the pyridyl nitrogen with respect
to proton [a] and thereof putting the [a] proton closest
to the thiophene proton [e]. Pyridyl proton [c] correlated
equally as [a] with all phenyl and thiophene protons.
The observed trNOE correlations maximized when
recorded at mixing times between 50 and 70ms and no
NOEÕs were observed for the free sulfonamide 6. Con-
formational searches17 generated low energy structures
(Fig. 2A) that could not explain the observed trNOE
correlations. Applying the strongest NOE contacts [d–
˚
˚
f] and [a–e] at ꢀ4A range followed by [a–f] at 4–5A
range quickly shaped 6 into a ÔbentÕ structure that was
Figure 2. (A) Lowest energy structure obtained from stochastic conformational search. (B) Low energy structure derived from NMR data of 6 when
bound to NS5B.
Figure 3. Back-calculated simulated NOESY of the NMR-derived bent-shape sulfonamide 6.

5336
C. G. Yannopoulos et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5333–5337
13. For a recent review on HCV NS5B polymerase patent
literature, see: Zhang, X. Idrugs 2002, 5, 154.
14. Wang, M.; Ng, K. N. S.; Cherney, M. M.; Chan, L.;
not very different to the final NMR derived structure as
depicted in Figure 2B. The trNOE back-calculations of
this bent shape of 6 reproduced all intense experimental
trNOEÕs (more than 98% of boxed cross-peaks seen in
Fig. 1) as shown in Figure 3.18
´
Yannopoulos, C. G.; Bedard, J.; Morin, N.; Nguyen-Ba,
N.; Bethell, R. C.; Alaoui-Ismaili, M. H.; James, M. N. G.
J. Biol. Chem. 2003, 278, 9489.
15. (a) Chan, L.; Reddy, T. J.; Proulx, M.; Das, S. K.; Pereira,
O.; Wang, W.; Siddiqui, A.; Yannopoulos, C. G.; Poisson,
C.; Turcotte, N.; Drouin, A.; Alaoui, H. M. A.; Bethell,
R.; Hamel, M.; Bilimoria, D.; LÕHeureux, L. C. J. Med.
Chem. 2003, 46, 1283; (b) Reddy, T. J.; Chan, L.; Turcotte,
N.; Proulx, M.; Pereira, O. Z.; Das, S. K.; Siddiqui, A.;
Wang, W.; Poisson, C.; Yannopoulos, C. G.; Bilimoria,
D.; LÕHeureux, L.; Alaoui, H. M. A.; Nguyen-Ba, N.
Bioorg. Med. Chem. Lett. 2003, 13, 3341; (c) Chan, L.;
Reddy, S. K.; Das, S. K.; Reddy, T. J.; Poisson, C.;
Proulx, M.; Pereira, O.; Courchensne, M.; Roy, C.; Wang,
W.; Siddiqui, A.; Yannopoulos, C. G.; Nguyen-Ba, N.;
Labrecque, D.; Bethell, R.; Hamel, M.; Courtemanche-
Asselin, P.; LÕHeureux, L.; David, M.; Nicolas, O.;
Brunette, S.; Bilimoria, D.; Bedard, J. Bioorg. Med. Chem.
Lett. 2004, 14, 793; (d) Chan, L.; Pereira, O.; Reddy, T. J.;
Das, S. K.; Poisson, C.; Courchensne, M.; Proulx, M.;
Siddiqui, A.; Yannopoulos, C. G.; Nguyen-Ba; Roy, C.;
Nasturica, D.; Moinet, C.; Bethell, R.; Hamel, M.;
LÕHeureux, L.; David, M.; Nicolas, O.; Courtemanche-
Asselin, P.; Brunette, S.; Bilimoria, D.; Bedard, J. Bioorg.
Med. Chem. Lett. 2004, 14, 797.
The very weak trNOEÕs were not considered in the struc-
ture calculations or in the back-calculations as they may
arise from complex-mediated spin diffusion which is yet
of unknown nature.
In summary, a trNOE structure of a soluble sulfon-
amide is reported; the ÔbentÕ shape character provided
the first understanding towards better inhibitor design.
The solution conformation of 6 bound to HCV NS5B
polymerase has also been confirmed by X-ray crystallo-
graphy of a related sulfonamide analogue/NS5B com-
plex and these findings will be described elsewhere.
Finally, we are attempting to understand the role of
the template-primer since this study indicates that ade-
quate inhibitor binding to the NS5B is dependent on
the presence of template-primer.
Acknowledgements
16. The expression and purification of HCV NS5B genotype
1b strain BK is described in our publication (see Ref. 15a).
17. Sulfonamide 6 was studied by modeling the structure at
the ionization state that could be reflected in aqueous
conditions at ꢀpH7. MOE (Chemical Computing Group
Inc., Montreal, Canada) was used to build and energy
minimize 6 with the MMFF94s force field using a gradient
´ `
We thank Therese Godbout for her assistance in the
preparation of this manuscript.
References and notes
˚
of 0.01kcal/molA. The resulting minimized structure was
1. (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.
2. (a) Gonnella, N. C.; Bohacek, R.; Zhang, X.; Kolossvary,
I.; Paris, C. G.; Melton, R.; Winter, C.; Hu, S.-I.; Ganu,
V. Proc. Natl. Acad. Sci. U.S.A. 1995, 92(2), 462; (b)
LaPlante, S. R.; Aubry, N.; Bonneau, P. R.; Kukolj, G.;
Lamarre, D.; Lefebvre, S.; Li, H.; Llinas-Brunet, M.;
Plouffe, C.; Cameron, D. R. Bioorg. Med. Chem. Lett.
2000, 10(20), 2271.
used as a starting point for generating a set of conformers
obtained from the use of MOEÕs implemented systematic
conformational searching tool. To visualize the conform-
ational preference of 6, the first 50 lowest energy structures
were examined for each of the representative ÔclosedÕ or
ÔopenÕ shapes. Only the ÔopenÕ shape of 6 was observed.
The average energy difference between the 1st and 50th
˚
structure was approximately 2–3kcal/molA.
18. The MOE minimized structure of 6 was used as starting
points for introducing distance restrains observed from the
trNOE spectrum. The strongest observed H–H correla-
3. Hepatitis C: Global Prevalence. Weekly Epidemiological
Record, 2000; Vol. 75, p 3.
4. WHO 1999, J. Viral Hepatitis 6, 35.
˚
tions were initially assigned to distances of 4–5A and
imposed geometric constraints were set with strengths of
100. Conformations of 6 consistent with observed NOEs
were generated by minimization using the MMFF94s
force field and were saved as .pdb files needed for trNOE
back-calculations using the trNOEPAK program. trNOE-
PAK is part of the ECEPP/NMR (ECEPP algorithm was
originally developed in Prof. ScheragaÕs group at Cornell
University) suite that was developed by Dr. Feng Ni group
at the Biotechnology Research Institute and is currently
used for conformational structure calculations/validations
through comparisons of experimental and calculated
transferred NOE18 spectra. The conformer .pdb file was
used by the PDB2NOE program (part of ECEPP/NMR)
in order to calculate the predicted transferred NOE
intensities that were saved into a .noe file. The .noe was
used by GFIDSJ (part of ECEPP/NMR) to generate a
free-induction decay (FID) matrix (.dat file) using the
calculated NOEs and the experimental chemical shifts, line
widths, coupling constants and attenuation factors for all
observable protons. Finally, the NMRDSP program (part
of ECEPP/NMR) was used as a digital signal processing
5. Saito, I.; Miyamura, T.; Ohbayashi, A.; Harada, H.;
Katayama, T.; Kikuchi, S.; Watanabe, Y.; Koi, S.; Onji,
M., et al. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 6547.
6. Purcell, R. Hepatology 1997, 26, 11S.
7. Takamizawa, A.; Mori, C.; Fuke, I.; Manabe, S.; Mura-
kami, S.; Fujita, J.; Onishi, E.; Andoh, T.; Yoshida, I.;
Okayama, H. J. Virol. 1991, 65, 1105.
8. Kato, N.; Hijikata, M.; Ootsuyama, Y.; Nakagawa, M.;
Ohkoshi, S.; Sugimura, T.; Shimotohno, K. Proc. Natl.
Acad. Sci. U.S.A. 1990, 87, 9524.
9. Grakoui, A.; Wychowski, C.; Lin, C.; Feinstone, S. M.;
Rice, C. M. J. Virol. 1993, 67, 1385.
10. Ago, H.; Adachi, T.; Yoshida, A.; Yamamoto, M.;
Habuka, N.; Yatsunami, K.; Miyano, M. Structure
(London) 1999, 7, 1417.
11. Lesburg, C. A.; Cable, M. B.; Ferrari, E.; Hong, Z.;
Mannarino, A. F.; Weber, P. C. Nat. Struct. Biol. 1999, 6,
937.
12. Bressanelli, S.; Tomei, L.; Roussel, A.; Incitti, I.; Vitale, R.
L.; Mathieu, M.; De Francesco, R.; Rey, F. A. Proc. Natl.
Acad. Sci. U.S.A. 1999, 96, 13034.

C. G. Yannopoulos et al. / Bioorg. Med. Chem. Lett. 14 (2004) 5333–5337
5337
system for converting the simulated and corresponding
experimental FID matrices into the NOESY spectrum
.mat file. NMRVIEW 5.0.2 was used to visualize the
simulated .mat spectrum seen in Figure 3. Additional
references: Nemethy, G.; Gibson, K. D.; Palmer, K. A.;
Yoon, C. N.; Paterlini, G.; Zagari, A.; Rumsey, S.;
Scheraga, H. A. J. Phys. Chem. 1992, 96, 6472; (b) Ni, F.
Prog. NMR Spectrosc. 1994, 26, 517.