Conformation-based restrictions and scaffold replacements in the design of hepatitis C virus polymerase inhibitors: Discovery of deleobuvir (BI 207127)

Article abstract of DOI:10.1021/jm4011862

Conformational restrictions of flexible torsion angles were used to guide the identification of new chemotypes of HCV NS5B inhibitors. Sites for rigidification were based on an acquired conformational understanding of compound binding requirements and the roles of substituents in the free and bound states. Chemical bioisosteres of amide bonds were explored to improve cell-based potency. Examples are shown, including the design concept that led to the discovery of the phase III clinical candidate deleobuvir (BI 207127). The structure-based strategies employed have general utility in drug design.

Full text of DOI:10.1021/jm4011862

Article
pubs.acs.org/jmc
Conformation-Based Restrictions and Scaffold Replacements in the
Design of Hepatitis C Virus Polymerase Inhibitors: Discovery of
Deleobuvir (BI 207127)
Steven R. LaPlante,*,† Michael Bos, Christian Brochu, Catherine Chabot, Rene Coulombe,
́
̈
James R. Gillard, Araz Jakalian, Martin Poirier, Jean Rancourt, Timothy Stammers,
Bounkham Thavonekham, Pierre L. Beaulieu, George Kukolj, and Youla S. Tsantrizos*,‡
Departments of Chemistry and Biological Sciences, Boehringer Ingelheim (Canada) Ltd., 2100 Cunard Street, Laval, Quebec, Canada
H7S 2G5
S
* Supporting Information
ABSTRACT: Conformational restrictions of flexible torsion
angles were used to guide the identification of new
chemotypes of HCV NS5B inhibitors. Sites for rigidification
were based on an acquired conformational understanding of
compound binding requirements and the roles of substituents
in the free and bound states. Chemical bioisosteres of amide
bonds were explored to improve cell-based potency. Examples
are shown, including the design concept that led to the
discovery of the phase III clinical candidate deleobuvir (BI
207127). The structure-based strategies employed have
general utility in drug design.
with the binding events, from which rigidification proposals and
analogues were then generated as a means of exploring new
chemotypes and chemically diverse series of compounds. The
examples discussed here involve targeting the thumb pocket 1
allosteric site of the NS5B polymerase of the hepatitis C virus
(HCV).6
INTRODUCTION
■
The optimization and exploitation of ligand binding to
macromolecules, which is one of the primary goals of medicinal
chemistry, may often be confounded by a variety of interrelated
factors. Structure−activity relationships (SARs) attempt to
elucidate the importance and impact of direct ligand−protein
interactions such as van der Waals surface contacts and
lipophilic and ionic attractions as well as the establishment of
hydrogen bonds and solvation influences.1 However, there are
some other less characterized but nonetheless important
factors. These include but are not limited to conformational
changes associated with the binding events (ligand and
protein). The deconvolution of the individual enthalpic and
entropic energy components detailed above often proves to be
very challenging or seemingly impossible. While significant
advances have led to receptor-induced fit and conformational
RESULTS AND DISCUSSION
■
Our initial HCV NS5B polymerase hit [compound 1 (Figure
1)] came from a high-throughput screen of our compound
collection.7 Ligand-based optimization led to the potent
tryptophan−benzimidazole analogue 2,7b,h but this series lacked
many desired properties for an effective anti-HCV drug (e.g.,
cell culture activity) presumably because of the effects of two
ionizable carboxylic acid functions. A knowledge-based strategy
successfully “linker hopped” to a new structural hinge that led
to a promising diamide chemical series [e.g., compound 3
(Figure 1)] that appropriately matched the ligand bioactive
conformation and pocket bioactive space.3,7f Subsequently, hit-
to-lead and lead optimization8a efforts led to several structural
modifications, including the replacement of the benzimidazole
scaffold with the more lipophilic indole isostere, and
identification of BILB 1941 [compound 4 (Figure 1)], which
was advanced to clinical development. This compound was the
selection models that describe bimolecular recognition,1,2
a
more comprehensive view of the impact of shape and
conformational changes upon binding is still evolving.3−5
Ligand−receptor association may be viewed as the collision
of two flexible objects where their transient shapes must have
sufficient complementarities to allow initial contact followed by
mutual adaptations to stabilize the interactions. Deciphering the
precise details of the collision and binding events remains
largely elusive and impractical in drug discovery, but if trends
are exposed and properly exploited, then breakthroughs may be
possible.3
Special Issue: HCV Therapies
Received: August 23, 2013
This work describes our approach in acquiring information
about the ligand shape and conformational changes associated
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 1. Key lead series discovered as thumb pocket 1 HCV polymerase inhibitors. Blue and red denote those atoms involved in the structural
hinge. For the IC50 assay, the Δ21 NS5B construct from GT-1b was used.7g For the EC50 assay, a GT-1b replicon reporter assay was employed.7e,12
first inhibitor of binding of HCV NS5B to the thumb pocket 1
allosteric site to demonstrate proof of concept in reducing the
viral load in patients infected with HCV genotype 1, albeit with
limited potency at the tolerated doses.8
Efforts were then focused on identifying new series of
compounds that bound to the same site and had improved cell-
based potency and pharmacological properties. The rationale
employed to transition to new and promising series was based
on an acquired understanding of ligand binding requirements
and substituent roles in the free and bound states of our initial
leads.
An overview of our understanding of the roles of each
substituent for 4 was assembled via multidisciplinary
approaches and is summarized in Figure 2. The ensemble of
information was consistent with the cyclopentyl ring playing an
important role, as it binds in a lipophilic pocket exposed upon
displacement of the Λ1 finger loop that interacts with the
thumb domain in the apo state.3 The pyridyl group lies over the
protein surface and appears to have minor binding attributes as
supported by the lack of NMR differential line broadening, and
SAR studies that show that a plethora of substituent
replacements are possible at this position.3,7a−d,h,8a This
substituent’s principal role is likely to help orient and sterically
rigidify the free state of the critical cyclopentyl ring to adopt the
bioactive conformation. The indole has an important
scaffolding role, orienting the appendages along the α, β, and
γ torsion angles (Figure 2). This bicylic scaffold also lies
perfectly flat against a receptor cliff. The methyl group helps
rigidify the α torsion angle and may be involved in weak
electrostatic interactions with the polymerase backbone
carbonyls of Leu 492 and Gly 493.3 The right-hand side of
Figure 2. Model structure of 4 bound to thumb pocket 1 of HCV
polymerase. The model was generated using a combination of X-ray
structures of related compounds, NMR data, and docking.3 Also
displayed are interpretations of the roles of each substituent derived
from multidisciplinary efforts. The red-colored Greek letters α−ν
designate the torsion angles for which conformational restrictions were
explored and discussed herein. Some amino acid residues and features
of the HCV polymerase binding site are colored yellow.
this compound employs an α,α-disubstituted amino acid as a
structural hinge that links with and allows the phenyl cinnamic
acid right-hand side to lie flat over the upper receptor plateau.
The carbonyl located between angles γ and δ is involved in a
hydrogen bond with Arg 503 (not highlighted in the surface
display in Figure 2).
B
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 3. Example of conformational restrictions along torsion angles α and β (colored red). For the IC50 assay, the Δ21 NS5B construct from GT-
1b was used.7g For the EC50 assay, a GT-1b replicon reporter assay was employed.7e,12
Figure 4. Structures and inhibition activities of methyl and aza analogues. NMR ROESY data are shown for the compounds in the diamide series.
ROESY data provide inter-hydrogen distance information for small molecules in the free state, allowing the determination of conformational
preferences in the free state.
Given the wealth of knowledge built around this diamide
series as well as the earlier tryptophan analogues,3,7,8 multiple
ideas for conformational restrictions and scaffold replacements
were proposed to identify new possible chemotypes. Studies
exemplified here involved the torsion angles designated by
Greek letters (red) in Figure 2.
C
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 5. (A) Structure of compound 14 with the hydrogen nomenclature colored on the basis of the observed DLB data given in panel B. Magenta
numbers denote hydrogen resonances that changed significantly upon the addition of HCV polymerase; green numbers denote hydrogens that
experienced insignificant changes, and black numbers signify ambiguous data. (B) Blue 1H NMR spectrum of free 14 (at 200 μM) and red spectrum
after addition of HCV polymerase (at 20 μM).
The first examples involving conformational rigidifications of
the left-hand side were focused on torsion angles α and β. We
reported earlier that an important role of the pyridyl group at
the indole C-2 position was to sterically orient the critical
cyclopentyl to its bioactive conformation.7h Thus, both
substituents appeared to be conformationally interdependent
through the rotation energy barriers for angles α and β.7h This
was corroborated with the creative efforts of other groups by
designing novel macrocycles of various ring sizes that attempt
to rigidify the C-2 substituent in its bioactive conformation (as
revealed by X-ray crystallography).9
Analogue 5 (Figure 3) provides a literature example of this
type of “top cyclization” (red), and additional examples are
shown in the Supporting Information.9 This compound had
impressive activity in cell culture (EC50 = 0.004 μM), and many
new compounds have been discovered that apply similar and
more elaborate top cyclization strategies. Other pharmaceutical
companies have found that some analogues with top
cyclizations had improved PK profiles, and compounds that
are based on this concept are in clinical development.9
amenable to biophysical analyses. Compound 10 exhibited
ROESY NMR cross-peaks between amide hydrogen 1 and both
aromatic hydrogens 16 and 18, implying conformational
heterogeneity along γ in the free state. Earlier energy
calculations reported low-energy minima when the directly
attached left-side amide and benzimidazole planes were mostly
coplanar, suggesting that the amide can adopt two orientations:
the bioactive “up orientation” where NH1 is close to hydrogen
18 and the nonbioactive “down orientation” where NH1 is
close to hydrogen 16.3,7h Therefore, these data were consistent
with only half of the molecules adopting the bioactive up
orientation along γ in the free state.
We therefore designed compounds 11 and 13 (Figure 4)
with chemical modifications that were meant to promote the
bioactive up orientation of the amide NH1 group. For
compound 11, it was envisioned that placing a methyl group
at position 18 would result in a steric clash with the proximate
amide carbonyl, thus forcing the amide NH1 group into or
close to the bioactive up orientation. As a control, compound
12 was also made with the methyl placed at position 16.
Unfortunately, compounds 11 and 12 were both much less
potent than unrestricted control compound 10. Calculations
suggested that this may be the result of distorting the amide
plane away from the benzimidazole plane,7h and ROESY NMR
data on the right of Figure 4 are consistent with the methyl
groups forcing a mix of conformations given that compounds
11 and 12 have peaks between the amide hydrogen 1 and both
hydrogens 16 and 18. It is noteworthy that the carbonyl group
of this amide bond is involved in a critical hydrogen bond with
the guanidinium side chain of Arg 503, and disturbing the
orientation of this group could have a negative impact on
potency. Thus, methylation did not force a single amide
orientation but rather allowed non-bioactive, free-state
conformations as was also noted for the unsubstituted
derivative 10. Other factors could also have contributed to
We also applied “bottom cyclization” strategies to constrict
rotations along both α and β torsion angles. Figure 3 provides
examples of bottom cyclizations (colored red) as in compounds
6−9. Compound 6 exhibited poor activity (80-fold decrease in
potency) compared to the corresponding unrestricted ana-
logues bearing a heterocycle such as 2-pyridyl at C-2 (IC50
=
0.23 μM).7c,h Nonetheless, this type of cyclization, when
combined with more optimal right-side groups that compensate
for the loss of intrinsic potency of the left-hand side, has the
potential to provide alternate chemotypes with good cell
culture activity as shown for compounds 7−9.
In parallel with the studies described above, chemical
modifications were applied to restrict conformational flexibility
along torsion angle γ (Figure 4). Studies were performed on a
related benzimidazole series with improved solubility and
D
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 6. (A) Superposition of the four orientations of the right-hand side of 14 in the bound state that are consistent with the ensemble of
transferred NOESY partially displayed in panels B−D. The methods employed for determining these structures are similar to those described in ref
3. Compounds 15 and 16 are purified enantiomers (exact stereochemistry undetermined).
the loss of potency. For example, methylation at position 16
may have also introduced unfavorable steric clashes with the
receptor pocket (e.g., compound 12). However, this should not
be the case for methylation at position 18 given that it should
be solvent-exposed in the bound state. More likely in this case,
methylation distorts the amide away from the bioactive
conformation. Given the information provided above, we
concluded that methylated analogues had undesirable proper-
ties, and alternative strategies for rigidifying γ should be
considered.
tools, in particular NMR, may be used to better understand
undesired and unexpected properties.
Conformational rigidification efforts were also directed to the
extended diamide right-hand side of inhibitors. It was
noteworthy that the potencies of analogues bearing a cinnamyl
moiety in either the meta or the para position with respect to
the aniline nitrogen were similar (unpublished data; refer also
to the example in Appendix 3 of the Supporting Information).
Consistent with this positional tolerability, the flexible ν torsion
angle of the cinnamic acid moiety could be replaced with a
variety of rigid and planar five-membered rings (data not
shown). One interesting example is shown in Figure 5A as
compound 14. This compound had attractive potency and
solubility properties that made it amenable to NMR solution
studies for improving our understanding of its binding to the
polymerase. NMR experiments were performed using differ-
ential line broadening (DLB) techniques13 to differentiate the
ligand substituents that contacted the receptor from those that
were solvent-exposed. This was done by superimposing the
Calculations7h suggested that incorporation of nitrogen at
position 18 would result in low-energy minima such that amide
NH1 would preferentially promote the bioactive up orientation
of the amide (via the nitrogen lone pair electrons and the amide
NH1 group). The corresponding aza-benzimidazole was made
(compound 13), and the expected free-state conformation was
corroborated by NMR data, where no ROESY peak was
observed between hydrogens 1 and 16 (bottom right of Figure
4), a result that is possible only if compound 13 adopts the
lowest-energy up orientation of the amide NH1 group.
However, a comparison of the activities of the control
compound 10 versus aza-benzimidazole 13 showed a 4-fold
loss of potency, despite conformational rigidification toward the
bioactive conformation. It is likely that the loss of potency upon
incorporation of a nitrogen arose mainly from another
unfavorable source, such as electrostatic incompatibility
between the negative electrostatic potential introduced by
incorporation of the nitrogen on the ligand and the negative
electrostatic potential of the corresponding site of the receptor
(see the Supporting Information). Despite the fact that the
nitrogen and methylation strategy described here failed to
spawn attractive chemotypes, it did demonstrate that analysis
1
one-dimensional H NMR spectra of free 14 (blue in Figure
5B) with that of 14 in the presence of HCV polymerase (red in
Figure 5B). Ligand resonances that differed the most (see
dotted ovals in Figure 5B) are those that report contacts with
the receptor, and those that do not change are assumed to be
solvent-exposed in the free and bound states.
The DLB data are summarized in Figure 5A where the
hydrogen numbers are colored magenta for resonances that
changed the most and hydrogen numbers are colored green for
those resonances that experienced little or no change. No
conclusions could be made for hydrogen numbers colored
black. The ensemble of this DLB data suggested that the
bottom part of 14 contacted the receptor (magenta numbers
are all at the bottom), whereas the top part was solvent-exposed
E
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
(green numbers are at the top). Overall, the DLB data are
consistent with a compound binding in a manner similar to that
of the related complex shown for 4 in Figure 2.
Attempts were then made to modulate the free-state
conformational properties by changing the position of the
nitrogen of the central structural hinge as was described in
detail for an earlier series.3 However, Figure 6 shows that the
related enantiomeric analogues 15 and 16 did not provide
improved potencies. Nonetheless, all three compounds were
then subjected to NMR transferred NOESY experiments13 in
the presence of the polymerase to determine the bound
conformations using cross-peaks that report intraligand hydro-
gen distance information (<5 Å) (Figure 6). Our intention was
to identify a bound conformation that would be representative
of the series. However, an analysis of the transferred NOESY
data for all three compounds in Figure 6 suggested that the
right side of these compounds did not adopt identical
conformations; instead, they assumed alternate torsion angles
along ζ and ε. For example, the relative cross-peak intensities
between hydrogens 4 and 5 and between hydrogens 4 and 9 are
similar in panels C and D of Figure 6 but different in Figure 6B.
Also, a cross-peak was observed between hydrogens 4 and 18 in
Figure 6D, whereas it was lacking in panels B and C. Likewise, a
cross-peak was noted between hydrogens 4 and 1 in Figure 6B,
but lacking in panels C and D.
Collectively, the data suggested that one may not deduce a
single accurate bound conformation that would be representa-
tive of this series. A more appropriate approach would be to
consider that the right-hand sides may adopt multiple possible
orientations that lie flat on the upper plateau of the receptor.
The transferred NOESY data indicated that these conforma-
tions were due mainly to differences in the ζ and ε torsion
angles with a significant interdependence on δ. These multiple
orientations were modeled using compound 14 for the sake of
convenience and taking into account the above differences
noted for compounds 14−16. Four complementary flat
conformations involving torsion angles ζ and ε were modeled
for compound 14 as shown in Figure 6A. These binding modes
are consistent with our extensive structure−activity studies
where it was found that a variety of substituents were tolerated
and complemented the flat surface feature of the upper plateau
of the receptor (data not shown). For example, Appendix 3 of
the Supporting Information shows a series of substituents meta,
para, or ortho to the aniline nitrogen for probing the spacial
range that was tolerated. Although all compounds maintained
flat conformations in the free state, the para and meta
substituents exhibited similar potencies whereas the ortho
substituent resulted in a loss of activity (also noted for other
unpublished examples). Thus, it became apparent that many
substitutions were tolerated as long as planarity was maintained
where the flat ligand conformations properly complemented
the available surface feature of the upper plateau.
Figure 7. Redesign of the structural hinge leading from diamide 4 to
17.
shown in Figure 7 ultimately led to the discovery of deleobuvir
(BI 207127) (compound 17), which is currently in phase III
clinical trials.10
Figure 8B shows a model of 17 bound to the thumb pocket 1
subdomain of HCV polymerase. A close-up view given in
Figure 8C reveals details of the binding mode where the red
dotted box highlights the cyclized ring discussed above that lies
flat on the upper plateau of the receptor. The colored surfaces
in Figure 8C highlight the position of amino acid substitutions
associated with compound resistance that lie immediately
below the new right-side appendage that provided biological
corroboration of this binding site.7e
Compound 17 is a potent and specific inhibitor of GT-1
HCV polymerase activity (IC50 = 50 nM) and subgenomic
antiviral activity (EC50 values of 11 and 23 nM in cell-based
replicon GT1b and GT1a assays, respectively) and shows weak
or no inhibition in specificity assays that include poliovirus
RdRp, mammalian DdRp II, and DNA polymerase α, β, and
γ.10,12 Furthermore, its in vitro ADME and in vivo cross-species
PK profiles are consistent with further progression into drug
development.14 Compound 17 displayed good antiviral potency
and tolerability in early clinical trials of short-term treatment
either as a single agent or in combination with pegylated
interferon-α2a/ribavirin (RBV) in HCV GT1 patients.8d
Moreover, the IFN-free combination of our NS3 protease
inhibitor faldaprevir with 17 and RBV has demonstrated high
efficacy and good tolerability in GT1b treatment-naive patients
in phase II clinical trials.10
The ensemble of conformations shown in Figure 6A
suggested that conformationally rigidified bioisosteres of the
right-hand side amide moiety may be acceptable replacements.
Bioisosteres of amide bonds are of significant utility in drug
design as they often lead to compounds with improved cell-
based potency and physicochemical properties.11 The most
fruitful of our approaches is illustrated in Figure 7. The amide−
aryl segment of the monocyclic diamide series 4 was replaced
by a benzimidazole moiety to eliminate the amide bond and
restrict rotation along ζ. Although multiple types of bioisosteres
were tolerated (data not shown), the structural modification
CONCLUSIONS
■
This work presents examples of how NMR-guided conforma-
tional restrictions and scaffold replacements are valuable
strategies in drug design. Knowledge building via structure
and dynamics approaches was found to be critical for the
generation of hypotheses and follow-up studies. Here, we
reported examples in which promising new chemotypes were
discovered, and a new series led to a compound that advanced
to clinical development.
F
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Figure 8. (A) Structure of deleobuvir (17) shown with a red dotted box that highlights the region implicated in the novel structural hinge. (B)
Space-filling view of 17 docked to NS5B based on the X-ray structures of the apo form and a related complex as described in detail earlier (Protein
Data Bank entries 3MWV and 3MWW, respectively).3 (C) Stick view of 17 in the thumb 1 binding pocket.
For each compound, a multitude of transferred NOESY data were
collected using many samples on 400, 600, and 800 MHz NMR
spectrometers. NOESY data were processed using XWIN-NMR and
MATERIALS AND METHODS
■
HCV Assays. Inhibition of GT-1b HCV NS5BΔ21 enzymatic
activity was performed as previously described.7g The bicistronic
Win2D, and the NOESY cross-peaks were completely assigned by
luciferase reporter replicon, encoding the Con1 GT-1b NS2-NS5B
applying a combination of ROESY, COSY, NOESY, one-dimensional
coding region, and the experimental procedures for measuring EC50
values in the experiments reported above have been described
elsewhere.12 Compounds were incubated with cells for 72 h, and the
relative levels of luciferase present were determined using the Bright-
Glo luciferase substrate (Promega) on a Packard Topcount instru-
ment. Alternatively, GT-1a or GT-1b HCV subgenomic stable cell
lines (129-S.16 or S22.3 cells, respectively) established at the
Boehringer Ingelheim (Canada) Ltd. R&D Laboratories were also
used to assess inhibition of HCV RNA replication in cell culture
through quantification of HCV RNA levels by TaqMan quantitative
real-time reverse transcription polymerase chain reaction.7e EC50
values were determined by the nonlinear regression routine NLIN
procedure of SAS (EC50). All reported values are the average of at least
two measurements.
Bound Structures Based on X-ray, NMR, and Docking. The
data presented in this work were gathered using the same constructs,
enzymatic assay, and cell culture assay as reported previously.3 The
bound structures of compounds 4, 17, and 14−16 were determined
using a combination of X-ray, NMR, and docking as described
elsewhere.3 Overall, they were docked to NS5B based on the X-ray
structures of the apo form and a related complex as described in detail
previously (Protein Data Bank entries 3MWV and 3MWW,
respectively).3
NMR Transferred NOESY Data. NOESY data were collected to
extract intramolecular hydrogen distances of compounds 14−16 and
S3 bound to HCV polymerase. These distances were then applied as
constraints in simulated annealing calculations to identify low-energy
structures that also satisfy the NMR constraints. All of the compounds
were soluble in buffer at concentrations equal to or greater than 0.2
mM, showed no evidence of self-aggregation as determined by
NOESY and/or DLS data, had stability equal to or greater than that at
24 h when in complex with HCV polymerase, and had relatively low
levels of spin diffusion artifacts.
spectra, HMQC, and HMBC data. Volumes of NOESY cross-peaks
were measured and translated to inter-hydrogen distances. The three-
dimensional structures were calculated using MSI or CCG modeling
software. A protocol that is similar to that previously described by
LaPlante et al.15 was applied
ASSOCIATED CONTENT
■
S
* Supporting Information
Materials and methods and other relevant experiments and a
synthesis procedure for compound 17. This material is available
free of charge via the Internet at http://pubs.acs.org.
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: RESGeneral.LAV@boehringer-ingelheim.com.
Present Addresses
†S.R.L.: NMX Solutions, 500 boulevard Cartier Ouest, Suite
6000, Laval, QC, Canada H7V 5B7. E-mail: stevenrlaplante@
gmail.com.
‡Y.S.T.: Department of Chemistry, McGill University, 801
Sherbrooke St. W., Montreal, QC, Canada H3A 0B8. E-mail:
youla.tsantrizos@mcgill.ca.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We are grateful to the following colleagues: P. Anderson, N.
Aubry, R. Bethell, D. Cameron, M. Cordingley, J.-M. Ferland,
M.-A. Poupart, and F. Stasi. We also acknowledge the valuable
G
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
analogues correlates with their affinity to E-selectin. Angew. Chem.,
Int. Ed. 2001, 125, 1941−1945.
teamwork and contributions of our colleagues at Boehringer
Ingelheim.
(5) (a) Kramer, B.; Rarey, M.; Lengauer, T. Evaluation of the FLEXX
incremental construction algorithm for protein-ligand docking. Proteins
1999, 37, 228−241. (b) Verdonk, M. L.; Cole, J. C.; Hartshorn, M. J.;
Murray, C. W.; Taylor, R. D. Improved protein-ligand docking using
GOLD. Proteins 2003, 52, 609−623. (c) Goodsell, D. S. Computa-
tional docking of biomolecular complexes with AutoDock CSH
Protocols 2009, 2009 (pdb prot5200). (d) Morris, G. M.; Huey, R.;
Olson, A. J. Using AutoDock for ligand-receptor docking. Current
Protocols in Bioinformatics; Wiley: New York, 2008. (e) Ewing, T. J.;
Makino, S.; Skillman, A. G.; Kuntz, I. D. DOCK 4.0: Search strategies
for automated molecular docking of flexible molecule databases. J.
Comput.-Aided Mol. Des. 2001, 15, 411−428. (f) McGann, M. R.;
Almond, H. R.; Nicholls, A.; Grant, J. A.; Brown, F. K. Gaussian
docking functions. Biopolymers 2003, 68, 76−90. (g) Goto, J.; Kataoka,
R.; Muta, H.; Hirayama, N. ASEDock-docking based on α spheres and
excluded volumes. J. Chem. Inf. Model. 2008, 48, 583−590. (h) Tondel,
K.; Anderssen, E.; Drablos, F. Protein Alpha Shape (PAS) Dock: A
new gaussian-based score function suitable for docking in homology
modeled protein structures. J. Comput.-Aided Mol. Des. 2006, 20, 131−
144. (i) Jackson, S.; DeGrado, W.; Dwivedi, A.; Parthasarathy, A.;
Higley, A.; Krywko, J.; Rockwell, A.; Markwalder, J.; Wells, G.
Template-constrained cyclic peptides: Design of high-affinity ligands
for GPIIb/IIIa. J. Am. Chem. Soc. 1994, 116, 3220−3230. (j) Miller, C.
P.; Collini, M. D.; Harris, H. A. Constrained phytoestrogens and
analogues as ERβ selective ligands. Bioorg. Med. Chem. Lett. 2003, 13,
2399−2403. (k) Udugamasooriya, D. G.; Spaller, M. R. Conforma-
tional constraint in protein ligand design and the inconsistency of
binding entropy. Biopolymers 2008, 89, 653−667. (l) Sharma, R.; Lee,
J.; Wang, S.; Milne, G. W. A.; Lewin, N. E.; Blumberg, P. M.; Marquez,
V. E. Conformationally constrained analogues of diacylglycerol. 10.
Ultrapotent protein kinase C ligands based on a racemic 5-
disubstituted tetrahydro-2-furanone template. J. Med. Chem. 1996,
39, 19−28. (m) Martin, S. F. Preorganization in biological systems:
Are conformational constraints worth the energy? Pure Appl. Chem.
ABBREVIATIONS USED
■
ADME, absorption, delivery, metabolism, excretion; clogP,
calculated partition coefficient; DdRP, DNA-dependent RNA
polymerase; DLB, differential line broadening; DMSO,
dimethyl sulfoxide; HCV, hepatitis C virus; HPLC, high-
pressure liquid chromatography; HTS, high-throughput screen;
IFN, interferon; NMR, nuclear magnetic resonance; NOESY,
nuclear Overhauser exchange spectroscopy; NS3, nonstructural
HCV protease; RBV, ribavirin; RdRp, RNA-dependent RNA
polymerase; ROESY, rotating-frame exchange spectroscopy;
SAR, structure−activity relationship
REFERENCES
■
(1) (a) Teague, S. Implications of protein flexibility for drug
discovery. Nat. Rev. Drug Discovery 2003, 2, 527−541. (b) Weikl, T.
R.; von Deuster, C. Selected-fit versus induced-fit protein binding:
Kinetic differences and mutational analysis. Proteins: Struct., Funct.,
Bioinf. 2009, 75, 104−110. (c) Hammes, G. G.; Chang, Y. C.; Oas, T.
G. Conformational selection or induced fit: A flux description of
reaction mechanism. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 13737−
13741. (d) Bakan, A.; Bahar, I. The intrinsic dynamics of enzymes
plays a dominant role in determining the structural changes induced
upon inhibitor binding. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 14349−
14354. (e) Lakomek, N. A.; Lange, O. F.; Oliver, F.; Walter, K. F.;
Korvin, F. A.; Fares, C.; Egger, D.; Lunkenheimer, P.; Meiler, J.;
Grubmueller, H.; Becker, S.; de Groot, B. L.; Griesinger, C. Residual
dipolar couplings as a tool to study molecular recognition of ubiquitin.
Biochem. Soc. Trans. 2008, 36, 1433−1437. (f) Lange, O. F.; Lakomek,
N. A.; Fares, C.; Schroeder, G. F.; Gunnar, G.; Walter, K. F. A.; Becker,
S.; Meiler, J.; Grubmueller, H.; Griesinger, C.; de Groot, B. L.
Recognition dynamics up to microseconds revealed from an RDC-
derived ubiquitin ensemble in solution. Science 2008, 320, 1471−1475.
(g) Kern, D.; Zuiderweg, E. R. The role of dynamics in allosteric
regulation. Curr. Opin. Struct. Biol. 2003, 13, 748−757. (h) Mittermaier,
A.; Kay, L. E. New tools provide new insights in NMR studies of
protein dynamics. Science 2006, 312, 224−228.
2007, 79, 193−200. (n) Harris, P. W. R.; Hugel, H. M.; Nurlawis, F. A
̈
review of the molecular conformations of melatonin ligands at the
melatonin receptor. Mol. Simul. 2002, 28, 889−902. (o) Giebel, L. B.;
Cass, R.; Milligan, D. L.; Young, D.; Arze, R.; Johnson, C. Screening of
cyclic peptide phage libraries identifies ligands that bind streptavidin
with high affinities. Biochemistry 1995, 34, 15430−15435.
(2) Koshland, D. The active site and enzyme action. Adv. Enzymol.
Relat. Subj. Biochem. 1960, 22, 45−97.
(3) LaPlante, S. R.; Gillard, J. R.; Jakalian, A.; Aubry, N.; Coulombe,
R.; Brochu, C.; Tsantrizos, Y. S.; Poirier, M.; Kukolj, G.; Beaulieu, P. L.
Importance of ligand bioactive conformation in the discovery of potent
indole-diamide inhibitors of the hepatitis C virus NS5B. J. Am. Chem.
Soc. 2010, 132, 15204−15212.
(6) (a) Choo, Q.-L.; Kuo, G.; Weiner, A. J.; Overby, L. R.; Bradley,
D. W.; Houghton, M. Isolation of a cDNA clone derived from a blood-
borne non-A, non-B viral hepatitis genome. Science 1989, 244, 359−
362. (b) Lavanchy, D. The global burden of hepatitis C. Liver Int.
2009, 29, 74−81. (c) O’Brien, C.; Godofsky, E.; Rodriguez-Torres, M.;
Afdhal, N.; Pappas, S. C.; Pockros, P.; Lawitz, E.; Bzowej, N.; Rustgi,
V.; Sulkowski, M.; Sherman, K.; Jacobson, I.; Chao, G.; Knox, S.;
Pietropaolo, K.; Brown, N. A. Randomized trial of valopicitabine
(NM283), alone or with peg-interferon, vs. retreatment with peg-
interferon plus ribavirin (pegIFN/RBV) in hepatitis C patients with
previous non-response to PegIFN/RBV: First interim results.
Hepatology 2005, 42, 234A. (d) Brown, N. Progress towards improving
antiviral therapy for hepatitis C with hepatitis C virus polymerase
inhibitors. Part I: Nucleoside analogues. Expert Opin. Invest. Drugs
2009, 18, 709−725. (e) Huang, Z.; Murray, M. G.; Secrist, J. A., III.
Recent development of therapeutics for chronic HCV infection.
Antiviral Res. 2006, 71, 351−362.
(4) (a) Peng, J. W. New probes of ligand flexibility in drug design:
Transferred 13C CSA-dipolar cross-correlated relaxation at natural
abundance. J. Am. Chem. Soc. 2003, 125, 11116−11130. (b) LaPlante,
S. R.; Aubry, N.; Dez
́
iel, R.; Ni, F.; Xu, P. Transferred 13C T1
relaxation at natural isotopic abundance: A practical method for
determining site-specific changes in ligand flexibility upon binding to a
macromolecule. J. Am. Chem. Soc. 2000, 122, 12530−12535. (c) Peng,
W.; Wilson, B. D.; Namanja, A. T. Mapping the dynamics of ligand
reorganization via 13CH3 and 13CH2 relaxation dispersion at natural
abundance. J. Biomol. NMR 2009, 45 (1−2), 171−183. (d) Peng, J. W.
Communication breakdown: Protein dynamics and drug design.
Structure 2009, 17, 319−320. (e) Mauldin, R. V.; Carroll, M. J.; Lee,
A. L. Dynamic dysfunction in dihydrofolate reductase results from
antifolate drug binding: Modulation of dynamics within a structural
state. Structure 2009, 17, 386−394. (f) Taylor, R. E.; Chen, Y.; Beatty,
A.; Myles, D. C.; Zhou, Y. Conformation−activity relationships in
polyketide natural products: A new perspective on the rational design
of epothilone analogues. J. Am. Chem. Soc. 2003, 125, 26−27.
(g) Thoma, G.; Magnani, J. L.; Patton, J. T.; Ernst, B.; Jahnke, W.
Preorganization of the bioactive conformation of sialyl LewisX
(7) (a) Beaulieu, P. L.; Bos, M.; Bousquet, Y.; Fazal, G.; Gauthier, J.;
̈
Gillard, J.; Goulet, S.; LaPlante, S. L.; Poupart, M. A.; Lefebvre, S.;
McKercher, G.; Pellerin, C.; Austel, V.; Kukolj, G. Non-nucleoside
inhibitors of the hepatitis C virus NS5B polymerase: Discovery and
preliminary SAR of benzimidazole derivatives. Bioorg. Med. Chem. Lett.
2004, 14, 119−124. (b) Beaulieu, P. L.; Bos, M.; Bousquet, Y.; DeRoy,
̈
P.; Fazal, G.; Gauthier, J.; Gillard, J.; Goulet, S.; Poupart, M.-A.; Valois,
S.; McKercher, G.; Kukolj, G. Non-nucleoside inhibitors of the
hepatitis C virus NS5B polymerase: Discovery of benzimidazole 5-
carboxylic amide derivatives with low-nanomolar potency. Bioorg. Med.
H
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Chem. Lett. 2004, 14, 967−971. (c) Beaulieu, P. L.; Gillard, J.;
Amssoms, K.; Canard, M.; Last, S.; van den Steen, I.; Devogelaere, B.;
Rouan, M.-C.; Vijgen, L.; Berke, J. M.; Dehertogh, P.; Fransen, E.;
Cleiren, E.; van der Helm, L.; Fanning, G.; van Emelen, K.; Nyanguile,
O.; Simmen, K.; Raboisson, P. Finger loop inhibitors of the HCV
NS5b polymerase. Part II. Optimization of tetracyclic indole-based
macrocycle leading to the discovery of TMC647055. Bioorg. Med.
Chem. Lett. 2012, 22, 4437−4443. (f) Cummings, M. D.; Lin, T.-I.;
Hu, L.; Tahri, A.; McGowan, D.; Amssoms, K.; Last, S.; Devogelaere,
B.; Rouan, M.-C.; Vijgen, L.; Berke, J. M.; Dehertogh, P.; Fransen, E.;
Cleiren, E.; van der Helm, L.; Fanning, G.; van Emelen, K.; Nyanguile,
O.; Simmen, K.; Raboisson, P.; Vendeville, S. Structure-based
macrocyclization yields hepatitis C virus NS5B inhibitors with
improved binding affinities and pharmacokinetic properties. Angew.
Chem., Int. Ed. 2012, 51, 4637−4640.
(10) (a) Larry, D.; Benhamou, Y.; Lohse, A. W.; Trepo, C.;
Moelleken, C.; Bronowicki, J.-P.; Arasteh, K.; Bourliere, M.; Heim, M.;
Enriquez, J.; Erhardt, A.; Zarski, J. P.; Wiest, R.; Gerlach, T.;
Wedemeyer, H.; Berg, T.; Stern, J.; Wu, K.; Abdallah, N.; Nehmiz, G.;
Boecher, W.; Steffgen, J. Safety, Pharmacokinetics and antiviral effect
of BI 207127, a novel HCV RNA polymerase inhibitor, after 5 days
oral treatment in patients with chronic hepatitis C. J. Hepatol. 2009, 50
(Suppl. 1), S383−S384 , Abstract 1054. (b) Zeuzem, S.; Asselah, S.;
Angus, P.; Zarski, J.-P.; Larrey, D.; Mullhaupt, H.; Gane, E.;
Schuchmann, M.; Lohse, A.; Pol, S.; Bronowicki, J.-P.; Roberts, S.;
Arasteth, A.; Zoulim, A.; Heim, M.; Stern, J. O.; Kukolj, G.; Nehmiz,
H.; Haefner, C.; Boecher, W. O. Efficacy of the protease inhibitor BI
201335, polymerase inhibitor BI 207127, and ribavirin in patients with
chronic HCV infection. Gastroenterology 2011, 141, 2047−2055.
(c) Zeuzem, S.; Soriano, V.; Asselah, T.; Bronowicki, J.-P.; Lohse, A.;
́
Bykowski, D.; Brochu, C.; Dansereau, N.; Duceppe, J. S.; Hache, B.;
́
Jakalian, A.; Lagace, L.; LaPlante, S. R.; McKercher, G.; Moreau, E.;
Perreault, S.; Stammers, T.; Thauvette, L.; Warrington, J.; Kukolj, G.
Improved replicon cellular activity of non-nucleoside allosteric
inhibitors of HCV NS5B polymerase: From benzimidazole to indole
scaffolds. Bioorg. Med. Chem. Lett. 2006, 16, 4987−4993. (d) Beaulieu,
P. L.; Bousquet, Y.; Gauthier, J.; Gillard, J.; Marquis, M.; McKercher,
G.; Pellerin, C.; Valois, S.; Kukolj, G. Non-nucleoside benzimidazole-
based allosteric inhibitors of the hepatitis C virus NS5B polymerase:
Inhibition of subgenomic hepatitis C virus RNA replicons in Huh-7
cells. J. Med. Chem. 2004, 47, 6884−6892. (e) Kukolj, G.; McGibbon,
G.; McKercher, G.; Marquis, M.; Lefebvre, S.; Thauvette, L.; Gauthier,
J.; Goulet, S.; Poupart, M. A.; Beaulieu, P. L. Binding site
characterization and resistance to a class of non-nucleoside inhibitors
of the hepatitis C virus NS5B polymerase. J. Biol. Chem. 2005, 280,
39260−39267. (f) Goulet, S.; Poupart, M. A.; Gillard, J.; Poirier, M.;
Kukolj, G.; Beaulieu, P. L. Discovery of benzimidazole-diamide finger
loop (thumb pocket I) allosteric inhibitors of HCV NS5B polymerase:
Implementing parallel synthesis for rapid linker optimization. Bioorg.
Med. Chem. Lett. 2010, 20, 196−200. (g) McKercher, G.; Beaulieu, P.
L.; Lamarre, D.; LaPlante, S. R.; LeFebvre, S.; Pellerin, C.; Thauvette,
L.; Kukolj, G. Specific inhibitors of HCV polymerase identified using
an NS5B with lower affinity for template/primer substrate. Nucleic
Acids Res. 2004, 32, 422−431. (h) LaPlante, S. R.; Jakalian, A.; Aubry,
N.; Bousquet, Y.; Ferland, J. M.; Gillard, J.; Lefebvre, S.; Poirier, M.;
Tsantrizos, Y. S.; Kukolj, G.; Beaulieu, P. L. Binding mode
determination of benzimidazole inhibitors of the hepatitis C virus
RNA polymerase by a structure and dynamics strategy. Angew. Chem.,
Int. Ed. 2004, 43, 4306−4311.
Mullhaupt, B.; Schuchmann, M.; Bourliere, M.; Buti, M.; Roberts, S.;
̈
(8) (a) Beaulieu, P. L.; Bos, M.; Cordingley, M. G.; Chabot, C.; Fazal,
̈
Gane, E.; Stern, J. O.; Kukolj, G.; Dai, L.; Bocher, W. O.; Mensa, F. J.
̈
G.; Garneau, M.; Gillard, J. R.; Jolicoeur, E.; LaPlante, S.; McKercher,
G.; Poirier, M.; Poupart, M.-A.; Tsantrizos, Y. S.; Duan, J.; Kukolj, G.
Discovery of the first thumb pocket 1 NS5B polymerase inhibitor
(BILB 1941) with demonstrated antiviral activity in patients
chronically infected with genotype 1 hepatitis C virus (HCV). J.
Med. Chem. 2012, 55, 7650−7666. (b) Erhardt, A.; Deterding, K.;
Benhamou, Y.; Reiser, M.; Forns, X.; Pol, S.; Calleja, J. L.; Ross, S.;
Spangenberg, H. C.; Garcia-Samaniego, J.; Fuchs, M.; Enriquez, J.;
Wiegand, J.; Stern, J.; Wu, K.; Kukolj, G.; Marquis, M.; Beaulieu, P.;
Nehmiz, G.; Steffgen, J. Safety, pharmacokinetics and antiviral effect of
BILB 1941, a novel hepatitis C virus RNA polymerase inhibitor, after 5
days oral treatment. Antiviral Ther. 2009, 14, 23−32. (c) Beaulieu, P.
L. Recent advances in the development of NS5B polymerase inhibitors
for the treatment of hepatitis C virus infection. Expert Opin. Ther. Pat.
2009, 19, 145−164. (d) Beaulieu, P. L. Successful strategies for the
discovery of antiviral drugs; Desai, M., Meanwell, N., Eds.; RSC Drug
Discovery Series 32; The Royal Society of Chemistry: London, 2013;
Chapter 9, p 279.
SVR4 and SVR12 with an interferon-free regimen of BI201335 AND
BI207127, +/− ribavirin, in treatment-naive patients with chronic
genotype-1 HCV infection: Interim results of SOUND-C2. J. Hepatol.
2012, 56 (Suppl. 2), S45. (d) Zeuzem, S.; Soriano, V.; Asselah, T.;
Bronowicki, J.-P.; Lohse, A. W.; Mullhaupt, B.; Schuchmann, M.;
̈
Bourlier
̀
e, M.; Buti, M.; Roberts, S. K.; Gane, E. J.; Stern, J. O.;
Vinisko, R.; Kukolj, G.; Gallivan, J.-P.; Bocher, W.-O.; Mensa, F. J.
Faldaprevir and deleobuvir for HCV genotype 1 infection. N. Engl. J.
Med. 2013, 369, 630−639.
̈
(11) Meanwell, N. A. Synopsis of some recent tactical application of
bioisosteres in drug design. J. Med. Chem. 2011, 54, 2529−2591.
(12) (a) Vaillancourt, F. H.; Pilote, L.; Cartier, M.; Lippens, J.; Liuzzi,
M.; Bethell, R. C.; Cordingley, M. G.; Kukolj, G. Identification of a
lipid kinase as a host factor involved in hepatitis C virus RNA
replication. Virology 2009, 387, 5−10. (b) Beaulieu, P. L.; Fazal, G.;
Goulet, S.; Kukolj, G.; Piorier, M.; Tsantrizos, Y.; Jolicoeur, E.; Gillard,
J.; Poupart, M.-A.; Rancourt, J. Viral Polymerase Inhibitors. Patent
WO 03/010141, 2003.
(9) (a) 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. Discovery of
conformationally constrained tetracyclic compounds as potent
hepatitis C virus NS5B RNA polymerase inhibitors. J. Med. Chem.
2006, 49, 6950−6953. (b) Stanfield, I.; Ercolani, C.; Mackay, A.;
Conte, I.; Pompei, M.; Koch, U.; Gennari, N.; Giuliano, C.; Rowley,
M.; Narjes, F. Tetracyclic indole inhibitors of hepatitis C virus NS5B-
polymerase. Bioorg. Med. Chem. Lett. 2009, 19, 627−632. (c) Haber-
mann, J.; Capito, E.; del Rosario Rico, M.; Ferreira, U.; Koch, U.;
Narjes, F. Discovery of pentacyclic compounds as potent inhibitors of
hepatitis C virus NS5B RNA polymerase. Bioorg. Med. Chem. Lett.
2009, 19, 633−638. (d) Ding, M.; He, F.; Hudyma, T. W.; Zheng, X.;
Poss, M. A.; Kadow, J. F.; Beno, B. R.; Rigat, K. L.; Wang, Y. K.;
Fridell, R. A.; Lemm, J. A.; Qiu, D.; Liu, M.; Voss, S.; Pelosi, L. A.;
Roberts, S. B.; Gao, M.; Knipe, J.; Gentles, R. Synthesis and SAR
studies of novel heteroaryl fused tetracyclic indole-diamide com-
pounds: Potent allosteric inhibitors of the hepatitis C virus NS5B
polymerase. Bioorg. Med. Chem. Lett. 2012, 22, 2866−2871. (e) Vende-
ville, S.; Lin, T.-I.; Hu, L.; Tahri, A.; McGowan, D.; Cummings, D.;
(13) LaPlante, S. R.; Aubry, N.; Bonneau, P.; Kukolj, G.; Lamarre, D.;
Lefebvre, S.; Li, H.; Llinas-Brunet, M.; Plouffe, C.; Cameron, D. R.
̀
NMR line-broadening and transferred NOESY as a medicinal
chemistry tool for studying inhibitors of the hepatitis C virus NS3
protease domain. Bioorg. Med. Chem. Lett. 2000, 10, 2271−2274.
(14) (a) Beaulieu, P. L.; Anderson, P. C.; Bos, M.; Brochu, C.;
̈
́
Cordingley, M. G.; Duan, J.; Garneau, M.; Lagace, L.; LaPlante, S. R.;
Marquis, M.; McKercher, G.; Poupart, M.-A.; Rancourt, J.; Stammers,
T.; Tsantrizos, Y. S.; Kukolj, G. Preclinical characterization of the
hepatitis C virus NS5B polymerase non-nucleoside inhibitor BI
207127. 19th International Symposium on Hepatitis C Virus and
Related Viruses, October 5−9, 2012, Venice, Italy. (b) Duan, J.;
Bolger, G.; Garneau, M.; Amad, M.; Batonga, J.; Montpetit, H.; Otis,
́
F.; Jutras, M.; Lapeyre, N.; Rheaume, M.; Kukolj, G.; White, P. W.;
Bethell, R. C.; Cordingley, M. G. The liver partition coefficient-
corrected inhibitory quotient and the pharmacokinetic-pharmacody-
namic relationship of directly acting anti-hepatitis C virus agents in
humans. Antimicrob. Agents Chemother. 2012, 56, 5381−5386.
(15) LaPlante, S. R.; Cameron, D. R.; Aubry, N.; Lefebvre, S.; Kukolj,
̀
G.; Maurice, R.; Thibeault, D.; Lamarre, D.; Llinas-Brunet, M. Solution
I
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry
Article
Structure of Substrate-based Ligands When Bound to Hepatitis C
Virus NS3 Protease Domain . J. Biol. Chem. 1998, 274, 18618−18624.
J
dx.doi.org/10.1021/jm4011862 | J. Med. Chem. XXXX, XXX, XXX−XXX