Discovery of benzimidazole-diamide finger loop (Thumb Pocket I) allosteric inhibitors of HCV NS5B polymerase: Implementing parallel synthesis for rapid linker optimization

Article abstract of DOI:10.1016/j.bmcl.2009.10.136

Previously described SAR of benzimidazole-based non-nucleoside finger loop (Thumb Pocket I) inhibitors of HCV NS5B polymerase was expanded. Prospecting studies using parallel synthesis techniques allowed the rapid identification of novel cinnamic acid rig

Full text of DOI:10.1016/j.bmcl.2009.10.136

Bioorganic & Medicinal Chemistry Letters 20 (2010) 196–200
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of benzimidazole-diamide finger loop (Thumb Pocket I) allosteric
inhibitors of HCV NS5B polymerase: Implementing parallel synthesis for
rapid linker optimization
*
*
Sylvie Goulet, Marc-André Poupart , James Gillard, Martin Poirier, George Kukolj, Pierre L. Beaulieu
Boehringer Ingelheim (Canada) Ltd, Research and Development, 2100 Cunard Street, Laval, Quebec, Canada H7S 2G5
a r t i c l e i n f o
a b s t r a c t
Article history:
Previously described SAR of benzimidazole-based non-nucleoside finger loop (Thumb Pocket I) inhib-
itors of HCV NS5B polymerase was expanded. Prospecting studies using parallel synthesis techniques
allowed the rapid identification of novel cinnamic acid right-hand sides that provide renewed oppor-
tunities for further optimization of these inhibitors. Novel diamide derivatives such as 44 exhibited
comparable potency (enzymatic and cell-based HCV replicon) as previously described tryptophan-
based inhibitors but physicochemical properties (e.g., aqueous solubility and lipophilicity) have been
improved, resulting in molecules with reduced off-target liabilities (CYP inhibition) and increased met-
abolic stability.
Received 27 August 2009
Accepted 30 October 2009
Available online 4 November 2009
Keywords:
Hepatitis C virus
HCV
NS5B polymerase
Inhibition
Ó 2009 Elsevier Ltd. All rights reserved.
Allosteric inhibitors
Replicon
Replication
The hepatitis C virus (HCV) has infected an estimated 170–200
million people world-wide.1,2 The infection becomes chronic in a
majority of cases and may lead to severe or fatal liver damage (cir-
rhosis and hepatocellular carcinomas). Currently available therapies
are based on the combined use of pegylated interferons (e.g., PEG-
Traditionally, the use of parallel synthesis in lead optimization
has focused mostly on expanding the diversity of substituents
around a core scaffold. While certainly worthwhile and useful in
rapidly providing structure–activity relationships (SAR), comple-
mentary uses of parallel chemistry techniques and principles can
be envisaged that concretely advance lead optimization projects
by providing more profound structural alterations and novel
opportunities for further optimization. Herein, we report a novel
strategy for identifying an alternative lead series through the syn-
thesis of small libraries based on established SAR and acquired
knowledge of preferred pharmacophores. This approach was suc-
cessfully applied to our HCV NS5B polymerase inhibitor project
and provided new opportunities for expansion and diversification
of a previously established lead compound.
We previously reported the discovery of specific benzimid-
azole-based non-nucleoside inhibitors (NNI) that bind to an allo-
steric site in the thumb domain of HCV NS5B polymerase
(Thumb Pocket I).6,7 This class of compounds is thought to interfere
with a protein conformational change that is required to initiate
viral RNA synthesis by the enzyme. The inhibitors bind in a hydro-
phobic pocket at the top of the thumb domain, and compete with
intra-molecular binding of a loop that extends from the finger do-
main of the enzyme, to prevent the formation of a completely en-
closed active site required for productive template binding and
RNA synthesis.7
IFN-
a
) and ribavirin which have limited efficacy (ꢀ50% for genotype
1 infected patients) and tolerability.3 The last years have witnessed
extensive efforts from the pharmaceutical industry, devoted to the
development of specific antiviral agents for the treatment of HCV
infection that would provide improved efficacy, tolerability, and
compliance in addressing the unmet medical need.4,5
Since its inception, combinatorial chemistry has become an
integrated part of the drug discovery process. On the one hand,
the synthesis of large screening libraries has contributed to com-
pound collections in pharmaceutical companies, occupying a sig-
nificant proportion of screening pools. On the other hand, the
synthesis of small focused libraries through parallel synthesis tech-
niques has played an equally important, yet complementary role in
lead optimization. Recent developments in new synthetic method-
ologies as well as new instrumentation for synthesis and purifica-
tion of compounds have contributed to significantly increasing the
utilization of parallel synthesis in lead optimization and the
generalization of these techniques to increasingly complex
chemistries.
The initial benzimidazole screening hit (1, NS5B IC50 = 14
was optimized to provide 5-hydroxytryptophan derivatives such
lM)
* Corresponding authors.
E-mail address: pbeaulieu@lav.boehringer-ingelheim.com (P.L. Beaulieu).
0960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.10.136

S. Goulet et al. / Bioorg. Med. Chem. Lett. 20 (2010) 196–200
197
as 2 with nanomolar potency in the enzymatic assay (Fig. 1).8,9
Unfortunately the polarity and ionization of these structures was
incompatible will cell permeation and the compounds were found
inactive in a cell-based HCV replicon assay.
enting group that positions the benzimidazole scaffold and the
right-hand side indole scaffold in an optimal arrangement for
interacting with the protein binding site.10,15 Attachment of a car-
boxylic acid function at the far right-hand side of the inhibitor was
previously shown to provide solubility and improve potency (cf.
compound 2).9 Thus, this function was maintained in the design
of our prospecting libraries. A hypothetical pharmacophore model
for benzimidazole inhibitors and the library designs are depicted in
Figure 2.
The library based on generic structure 6 suggested initial cou-
pling of benzimidazole carboxylic acid 9 to various diamino spac-
ers that were selected based on their ability to ultimately project
the terminal carboxylic acid pharmacophore at variable distances
and regions in space relative to the left-hand side of the molecule.
As shown in Scheme 1, coupling of benzimidazole 98 with Boc
mono-protected diamines 10 a–f under standard conditions pro-
vided amides 11a–f (ca. 75% yield) that were subsequently depro-
tected under acidic conditions to provide amines 12a–f. Each of
these amine fragments were then reacted with a set of aldehydes
(13–18) under reductive amination conditions to provide benzyl-
amine library I or coupled with isocyanate esters (19–23) to pro-
vide urea library II as shown in Scheme 1. Following
deprotection with NaOH (library II) and purification by reversed-
phase HPLC, inhibitors of general formula 6 were obtained in yields
ranging from 29% to 60% (library I, 34/36 compounds) and 12% to
80% (library II, 20/30 compounds). The 54 compounds thus ob-
tained were tested in our standard enzymatic NS5B assay,16 but
Removal of the two ionizable carboxyl groups led to neutral
inhibitors (3) with low micromolar potency in the cell-based assay
and additional biological validation of the mechanism of polymer-
ase inhibition.10 While replacement of the benzimidazole scaffold
of these inhibitors by a more lipophilic indole isostere provided
significant improvements in cell culture activity with EC50 values
in the low nanomolar range (e.g., 4),11,12 these molecules exhibited
undesirable physicochemical characteristics which we believed
would eventually compromise their further development. In par-
ticular, aqueous solubility was extremely low and the high lipo-
philicity of the molecules (cLog P = 6.1) increased potential for
off-target activity.13 In addition, electron-rich indole derivatives
are potential substrates for bioactivation that may lead to toxic-
ity.14 As a result, we initiated a search for alternative right-hand
sides with potential to confer superior drug-like properties to the
molecules and that would be more readily amenable to rapid ana-
log synthesis. Our objective was the identification of novel right-
hand sides that would provide inhibitors with enzymatic potency
in the nanomolar range, at least low micromolar cellular potency
and structural diversity from preceding series. Initial SAR in the
benzimidazole series had identified the cyclohexyl ring and the
carbonyl group on the benzimidazole scaffold as key contributors
to potency, whereas the C-2 position was more tolerant of struc-
tural changes.7,8 Our search for replacements of the right-hand side
portion of inhibitors was carried out using the synthetically acces-
sible benzimidazole scaffold, retaining these two anchor points.
The 3-furyl group was chosen for the C-2 substituent as it had been
shown to confer good potency. SAR and NMR studies on the pro-
tein-bound conformation of inhibitors suggested that the central
unfortunately, all were found to be less active (IC50 ꢀ 3–50
lM)
than the free carboxylic acid 9 (IC50 = 1.6
l
M).8
Consequently, our strategy focused on the general motif repre-
sented by diamide structure 7 (Fig. 2) in which an aniline carbox-
ylic acid derivative is attached to benzimidazole scaffold 9 either
directly or through an amino acid linker (d,l-alanine or d,l-b-ala-
nine), thus providing various spacer lengths between the left-hand
side benzimidazole scaffold and the right-hand side carboxylic acid
pharmacophore. The libraries were synthesized on a solid support
as described in Scheme 2. Seventeen nitroarenes carrying a carbox-
ylic acid function (24–40) were attached to bromo-Wang resin un-
der standard conditions and the nitro functionality reduced to
provide anilines 41. Coupling of the aniline to fragment 9 provided
library A. Coupling with Fmoc-d,l-Ala or Fmoc-d,l-b-Ala followed
by deprotection provided intermediates 42 that were subsequently
substituent extending from the asymmetric a-carbon of the trypt-
amine portion of inhibitors such as 2–4, served mainly as an ori-
OMe
O
O
N
N
OMe
N
O
N
H
N
Initial hit from screening:
NS5B IC50 = 14 µM
1
O
O
O
COOH
H-bond to protein
N
N
N
H
scaffold
spacer
IC50 = 0.019 µM
COOH
O
O
O
O
O
OH
improves
N
N
2
3
4
N
N
H
S
H
O
O
R2
Ar
potency
N
IC50 = 0.3 µM
5
spacer
EC50 = 1.7 µM (1b replicon)
N
N
N
H
hydrophobic
pocket
O
O
R1
N
OH
N
N
OH
N
N
H
S
H
O
spacer
O
6
N
IC50 = 0.1 µM
EC50 = 0.05 µM (1b replicon)
O
N
N
H
N
N
( )
n
OH
N
H
N
O
OH
O
N
H
7
Figure 1. Allosteric ‘finger loop’ inhibitors of HCV polymerase from early optimi-
zation studies.
Figure 2. Pharmacophore model and library design for right-hand side exploration.

198
S. Goulet et al. / Bioorg. Med. Chem. Lett. 20 (2010) 196–200
R1
N
O
O
H2N
Boc
R1
N
10 a-f
N
N
N
N
OH
N
X
H
(a)
O
O
11 a-f X = Boc
12 a-f X = H
9
(b)
O
R1
R2
N
(c) Library I:
(d) Library II:
OHC-R-CO2H
N
N
H
O
O=C=N-R-CO2Et
N
6
R2
R1
Library I: Benzyl amines
Library II: Ureas
CO2H
O
HO2C
N
H
N
N
a
H
19
O
CO2H
13
N
H
H
N
N
HO2C
b
c
20
21
O
O
14
H
N
CO2H
N
H
N
H
CO2H
H
N
15
H
N
CO2H
N
H
d
e
CO2H
22
23
O
O
16
17
N
O
H
N
N
H
CO2H
CO2H
N
18
N
H f
CO2H
Scheme 1. Libraries of benzylamine and urea derivatives synthesized from a diamine spacer. Reagents and conditions: (a) amine 1.1 equiv, HATU (1.1 equiv, DIPEA
(1.5 equiv)/DMF; (b) 25% TFA/CH2Cl2, 30 min; (c) library I: OHC-R (10 equiv), NaCNBH3 (5 equiv), AcOH/trimethylorthoformate/DMF 1:1, 3 h then 0.1 M aqueous HCl; (c)
library II: (Note: the carboxylic acid was introduced as ester derivative) (i) O@N@C–R (1.5 equiv)/DMF, 1.5 h (ii) 1 M aqueous NaOH (3.3 equiv), 14 h; (iii) AcOH (6.6 equiv).
coupled to benzimidazole 9 to provide libraries B and C (Ala and b-
Ala spacers, respectively). Libraries A (no spacer) and C (d,l-b-Ala
spacer) did not provide compounds with significantly improved
potency compared to benzimidazole carboxylic acid derivative 9,
suggesting that no additional interactions with the protein could
be engaged with these structures. IC50 values in library A ranged
ably as a consequence of the high degree of rotational freedom. In
support of this argument, conformationally-restricted cinnamic
acid derivative B-39 was identified as the most potent inhibitor
in this series (IC50 = 0.34 lM). Furthermore, the decrease in ioniza-
tion associated with the cinnamic acid moiety compared to ben-
zoic or aliphatic carboxylic acid derivatives resulted in improved
cellular permeation and a beneficial effect on inhibition of the rep-
from 9 to 34
lM. The most potent analog in library C was the 4-
aminocinnamic acid derivative C-39 (IC50 = 0.6
lM) which pro-
licon (EC50 = 13 lM). Moving the cinnamic acid moiety from the
vided a modest 2.5-fold improvement in potency over 9.
para- to the meta-position of the aromatic ring (B-33) resulted in
a modest threefold reduction in intrinsic potency, suggesting that
the acid function is unlikely to be involved in a strong ionic inter-
action with protein residues. Conformationally restricted furoic
acid derivative B-35 had comparable intrinsic potency as cinnamic
derivative B-39 but the increased ionization of the carboxylic acid
function was detrimental to cell culture activity.
During this initial prospecting study, we identified novel dia-
mide right-hand side functionalities (e.g., B-39) for our benzimid-
azole-based NS5B inhibitors that provided comparable intrinsic
potency to inhibitors such as 3 and offered new opportunities for
further optimization. All compounds in library B had been synthe-
sized in racemic form, and one of our objectives was to probe con-
formational preferences at the central amino acid position by
Library B (d,l-Ala spacer) in contrast, provided more encourag-
ing results. The most interesting analogs displayed IC50 values
<500 nM in our enzymatic assay (Table 1) along with HCV geno-
type 1b replicon cell-based potency in the 10–30
For example, 4-benzoic amide analog B-37 provided a fourfold
improvement in intrinsic potency (IC50 = 0.39 M) compared to
carboxylic acid derivative 9 and unlike the latter, inhibited the
1b cell-based replicon with EC50 = 30
M.17 Extending the carbox-
lM range.
l
l
ylic acid function by a methylene atom was detrimental to potency
(B-38) as was modifying the substitution pattern from a para- to a
meta-arrangement (B-26). Phenoxyacetic acid derivative B-31
which extends the carboxylic acid function one atom further from
the aromatic ring was tolerated but offered no advantage, presum-

S. Goulet et al. / Bioorg. Med. Chem. Lett. 20 (2010) 196–200
199
O
O
O
a
b
c, d
H2N
H
+
Br
O2N
O
Ar
O2N
O
Ar
O
Ar
24-40
41
Br-Wang resin
CH3
(
O
O
O
H
e, f
N
N
Ar
N
OH
)
N
n
H
H
N
Ar
O
O
O
AA
Sub-library A: fragment missing
Sub-library B: n = 0
Sub-library C: n = 1
42
7
CO2H
N
N
CO2H
CO2H
O
N
N
N
N
N
CO2H
CO2H
27
9
H
H
H
H
26
25
24
H
N
H
N
CO2H
HO
O
O
CO2H
CO2H
CO2H
CO2H
N
N
N
H
N
H
31
30
28
H
32
H
29
O
H
N
N
CO2H
O
O
H
CO2H
CO2H
CO2H
N
H
N
N
36
33
34
35
H
H
N
CO2H
N
N
N
H
H
H
H
CO2H
CO2H
CO2H
37
38
39
40
Scheme 2. Libraries A–C prepared using no spacer or d,l-alanine and d,l-b-alanine spacers. Reagents and conditions: (a) nitro acid (4 equiv), DIPEA (6.8 equiv)/DMF, 15 h; (b)
SnCl2 dihydrate (13 equiv)/DMF, 24 h; (c) Fmoc-d,l-Ala or Fmoc-d,l-b-aminobutyric acid (1.5 equiv), TBTU (1.5 equiv), DIPEA (3 equiv)/DMF, 3 h at 60 °C and 5 h at rt; second
coupling: Fmoc-d,l-Ala or Fmoc-d,l-b-amino-butyric acid (4.5 equiv), HATU (4.5 equiv), DIPEA (6.5 equiv)/DMF, 15 h rt; (d) 20% v/v/DMF 20 min; (e) 9 (1.5 equiv), TBTU
(1.5 equiv), DIPEA (3 equiv)/DMF, 18 h; (f) 50% TFA/DCE, 1h.
Table 1
racemic mixture. This result is consistent with previous findings
in the tryptophan series where a similar orientation of a car-
Most potent compounds from library B (d,l-Ala spacer)
Compds
NS5B IC50
,
l
M16
1b Replicon EC50, l
M17
boxyl, alkyl or heterocyclic substituent conferred preferential
binding to the protein (cf. compounds 2 and 3).15 In support of
this observation, replacement of the (R)-alanine linker by the
more flexible glycine residue (compound 45) resulted in a two-
fold loss in intrinsic potency. Diversifying SAR was performed
at the linker position with the aim of identifying superior linkers.
A variety of amino acids were introduced in place of the alanine
residue of 44 such as dl-Ile, dl-Phe, dl-cyclohexylalanine, dl-
norleucine, dl-phenylglycine, dl-cyclohexylglycine, dl-2-thienyl-
glycine, (R)-homoPhe, (R)-homoSer, (R)-4-phenylPhe, and (R)-eth-
B-37
B-38
B-26
B-31
B-39
B-33
B-35
0.39 ( 0.02)
1.52 (n = 1)
1.61 ( 0.04)
0.9 ( 0.2)
0.34 ( 0.07)
0.9 ( 0.2)
30 (n = 1)
13 (n = 1)
28 (n = 1)
0.45 (n = 1)
synthesizing configurationally-defined analogs. Previous studies in
the tryptophan series had identified a conformational bias for this
position, resulting from rigidification of the inhibitor backbone to-
ward the bioactive conformation.10,15 Compound B-39 was thus re-
synthesized using configurationally-defined (R)- and (S)-alanine
spacers and the results are shown in Table 2.
ylglycine. None proved superior to compound 44 (IC50 = 0.4
lM
for the homologated ethylglycine analog 46 to 8 M for the large
l
lipophilic 4-phenylPhe analog; results not shown). These obser-
vations are again consistent with the amino acid side chain
pointing to solvent and providing conformational rigidification
toward
a bioactive conformation. Modest improvements in
The (R)-configuration at the asymmetric center of the alanine
spacer provided a twofold improvement in potency over the
intrinsic potency could be achieved through conformational
rigidification of the right-hand side cinnamic acid moiety in com-
pound 44 (Fig. 3).
Both benzofuran (47) and indole (48) carboxylic acid deriva-
tives provided up to an additional twofold improvement in enzy-
matic potency with IC50 values in the 60–90 nM range (Table 2),
but the increased ionization of the carboxylic acid of these analogs
was detrimental to cell permeation and replicon inhibition (EC50
Table 2
Inhibitors with configurationally-defined Ala spacers
Compds
Spacer
NS5B
IC50
1b Replicon
EC50
M17
,
l
M16
, l
B-39
43
44
45
46
d,l-Ala
(S)-Ala
0.34 ( 0.07)
0.36 ( 0.08)
0.13 ( 0.02)
0.28 ( 0.08)
0.40 ( 0.07)
0.09 ( 0.02)
0.06 ( 0.01)
13 (n = 1)
25 (n = 1)
6.9 ( 0.8)
41 (n = 1)
6.5 (n = 1)
>42
>40 lM). Compound 44 was also tested against the RNA-depen-
dent RNA polymerase of the poliovirus and a mammalian DNA-
dependent RNA polymerase isolated from calf thymus.6,17 No inhi-
bition was detected against these enzymes at concentrations up to
(R)-Ala
Gly
(R)-Ethylglycine
(R)-Ala
(R)-Ala
47
48
200
lM, indicating that specificity against HCV was maintained in
>40
this series (TI >1500).

200
S. Goulet et al. / Bioorg. Med. Chem. Lett. 20 (2010) 196–200
O
Acknowledgments
H
N
N
N
N
H
We are grateful to Ginette McKercher and Louise Thauvette for
O
O
O
IC50 determinations and Martin Marquis for replicon EC50 values.
We thank Serge Valois, Colette Boucher and Michael Little for sol-
ubility and Log D determinations as well as Jianmin Duan and Josie
Demarte for CYP450 inhibition and microsomal stability data.
44
OH
References and notes
H
H
N
N
N
N
1. Choo, Q.-L.; Kuo, G.; Weiner, A. J.; Overby, L. R.; Bradley, D. W.; Houghton, M.
Science 1989, 244, 359.
2. Lavanchy, D. Liver Int. 2009, 29, 74.
COOH
COOH
H
H
O
O
O
N
H
47
48
3. Farnik, H.; Mihm, U.; Zeuzem, S. Liver Int. 2009, 29, 23.
4. For reviews on NS5B polymerase inhibitors, see: Beaulieu, P. L. Expert Opin.
Ther. Pat. 2009, 19, 145. and references cited therein.
Figure 3. Conformationally-restricted cinnamic acid isosteres.
5. For a review on inhibition of other HCV targets see, for example: (a) De
Francesco, R.; Migliaccio, G. Nature 2005, 436, 953; (b) Manns, M. P.; Foster, G.
R.; Rockstroh, J. K.; Zeuzem, S.; Zoulim, F.; Houghton, M. Nat. Rev. Drug Disc.
2007, 6, 991.
6. McKercher, G.; Beaulieu, P. L.; Lamarre, D.; LaPlante, S.; Lefebvre, S.; Pellerin, C.;
Thauvette, L.; Kukolj, G. Nucleic Acids Res. 2004, 32, 422.
7. Beaulieu, P. L. Curr. Opin. Drug Discovery Dev. 2006, 9, 618.
8. Beaulieu, P. L.; Bös, M.; Bousquet, Y.; Fazal, G.; Gauthier, J.; Gillard, J.; Goulet, S.;
LaPlante, S.; Poupart, M.-A.; Lefebvre, S.; McKercher, G.; Pellerin, C.; Austel, V.;
Kukolj, G. Bioorg. Med. Chem. Lett. 2004, 14, 119.
9. Beaulieu, P. L.; Bös, M.; Bousquet, Y.; DeRoy, P.; Fazal, G.; Gauthier, J.; Gillard, J.;
Goulet, S.; McKercher, G.; Poupart, M.-A.; Valois, S.; Kukolj, G. Bioorg. Med.
Chem. Lett. 2004, 14, 967.
Table 3
Comparison of ADMET parameters for tryptophan and diamide-based benzimidazole
NS5B inhibitors
Compds Log D
(pH 7.4)
Sola
HLMb
CYP450 IC50
1A2 2C9 2C19 2D6 3A4
>30 2.3 2.4 4.0 0.4
>30 >30 >30 >30 12.3
(lM)
(lg/mL) (min)
3
44
3.8
2.2
<0.07
171
26
44
10. Beaulieu, P. L.; Bousquet, Y.; Gauthier, J.; Gillard, J.; Marquis, M.; McKercher, G.;
Pellerin, C.; Valois, S.; Kukolj, G. J. Med. Chem. 2004, 47, 6884.
11. (a) Harper, S.; Pacini, B.; Avolio, S.; Di Filippo, M.; Migliaccio, G.; Laufer, R.; De
Francesco, R.; Rowley, M.; Narjes, F. J. Med. Chem. 2005, 48, 1314; (b) Ontoria, J.
M.; Martin Hernando, J. I.; Malancona, S.; Attenni, B.; Stansfield, I.; Conte, I.;
Ercolani, C.; Habermann, J.; Ponzi, S.; Di Filippo, M.; Koch, U.; Rowley, M.;
Narjes, F. Bioorg. Med. Chem. Lett. 2006, 16, 4026.
12. Beaulieu, P. L.; Gillard, J.; Bykowski, D.; Brochu, C.; Dansereau, N.; Duceppe, J.-
S.; Haché, B.; Jakalian, A.; Lagacé, L.; LaPlante, S.; McKercher, G.; Moreau, E.;
Perreault, S.; Stammers, T.; Thauvette, L.; Warrington, J.; Kukolj, G. Bioorg. Med.
Chem. Lett. 2006, 16, 4987.
a
Twenty-four hours solubility of amorphous material at pH 7.2.
T1/2 at 10 lM in human liver microsomes (min).
b
In addition to identifying novel structural features that would
provide opportunities for diversification of the SAR in the benz-
imidazole series, we also aimed to improve the physicochemical
properties of our inhibitors. 5-Hydroxytryptophan derivatives such
as 3 suffered from high lipophilicity (Log D = 3.8 at pH 7.4) and
13. (a) Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881; (b) Gleeson,
M. P. J. Med. Chem. 2008, 51, 817.
14. Kalgutkar, A. S.; Gardner, I.; Obach, R. S.; Shaffer, C. L.; Callegari, E.; Henne, K.
R.; Mutlib, A. E.; Dalvie, D. K.; Lee, J. S.; Nakai, Y.; O’Donnell, J. P.; Boer, J.;
Harriman, S. P. Curr. Drug Metab. 2005, 6, 161.
15. LaPlante, S.; Jakalian, A.; Aubry, N.; Bousquet, Y.; Ferland, J.-M.; Gillard, J.;
Lefebvre, S.; Poirier, M.; Tsantrizos, Y.; Kukolj, G.; Beaulieu, P. L. Angew. Chem.,
Int. Ed. 2004, 43, 4406.
poor solubility (typically <0.1 lg/mL in pH 7.2 buffer). These unde-
sirable properties often lead to liabilities in ADMET (e.g., CYP450
inhibition, poor metabolic stability) and toxicity in vivo.13 In con-
trast, compound 44, with comparable potency to 3 against NS5B
and the cell-based replicon, provided for improved physicochemi-
cal properties (Log D = 2.2 at pH 7.4 and solubility = 171 lg/mL at
pH 7.2), reduced potential for metabolism by human liver micro-
16. All inhibitors in this study were purified to >95% homogeneity by reversed-
phase HPLC and isolated as TFA salts. All compounds were characterized by
mass spectrometry and gave 1H NMR spectra consistent with expected
structures. IC50 values were determined as described in Ref. 6 and are the
average of at least two independent determinations unless indicated
otherwise.
17. EC50 determinations in the cell-based replicon assay were performed in
duplicates using RT-PCR for RNA quantification as described in: Beaulieu, P. L.;
Fazal, G.; Goulet, S.; Kukolj, G.; Poirier, M.; Tsantrizos, Y.; Jolicoeur, E.; Gillard,
J.; Poupart, M.-A.; Rancourt, J. WO Patent WO 03/010141, 2003. Total replicon
RNA levels were normalized to control cellular GAPDH mRNA. Compounds
were devoid of cytotoxicity at inhibitory concentrations. For an overview and
basic protocols on the use of HCV replicons, see Lohmann, V. Meth. Mol. Biol.
2009, 510, 145: Hepatitis C: Methods and Protocols, 2nd ed.; Tang, H., Ed.;
Humana Press.
somes and CYP450 inhibition as depicted in Table 3.
In conclusion, a prospecting study using parallel synthesis
techniques allowed the rapid identification of novel right-hand
sides that offer renewed opportunities for further optimization
of our benzimidazole-based HCV NS5B polymerase inhibitors.
The novel diamide derivatives exhibit comparable potency
(enzymatic and cell-based replicon) as previously described tryp-
tophan-based inhibitors and improved physicochemical proper-
ties. Further optimization of these inhibitors will be reported in
the near future.