Discovery of a potent boronic acid derived inhibitor of the HCV RNA-dependent RNA polymerase

Article abstract of DOI:10.1021/jm400317w

A boronic acid moiety was found to be a critical pharmacophore for enhanced in vitro potency against wild-type hepatitis C replicons and known clinical polymorphic and resistant HCV mutant replicons. The synthesis, optimization, and structure-activity relationships associated with inhibition of HCV replication in a subgenomic replication system for a series of non-nucleoside boron-containing HCV RNA-dependent RNA polymerase (NS5B) inhibitors are described. A summary of the discovery of 3 (GSK5852), a molecule which entered clinical trials in subjects infected with HCV in 2011, is included.

Full text of DOI:10.1021/jm400317w

Subscriber access provided by FLORIDA STATE UNIV
Article
Discovery of a Potent Boronic Acid Derived Inhibitor
of the HCV RNA-Dependent RNA Polymerase
Andrew Maynard, Renae M. Crosby, Byron Ellis, Robert Hamatake, Zhi Hong, Brian A.
Johns, Kirsten M. Kahler, Cecilia Koble, Anna L. Leivers, Martin R. Leivers, Amanda Mathis,
Andrew J. Peat, Jeffrey J. Pouliot, Christopher D. Roberts, Vicente Samano, Rachel M.
Schmidt, Gary K. Smith, Andrew Spaltenstein, Eugene L. Stewart, Pia Thommes, Elizabeth
M. Turner, Christian Voitenleitner, Jill T. Walker, Greg Waitt, Jason Weatherhead, Kurt L
Weaver, Shawn Williams, Lois Wright, Zhiping Z. Xiong, David Haigh, and J. Brad Shotwell
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400317w • Publication Date (Web): 14 May 2013
Downloaded from http://pubs.acs.org on May 17, 2013
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 43
1
2
3
4
5
6
7
8
9
Discovery of a Potent Boronic Acid Derived
Inhibitor of the HCV RNA-Dependent RNA
Polymerase
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Andrew Maynard2, Renae M. Crosby1, Byron Ellis2, Robert Hamatake1, Zhi Hong1, Brian
Johns1, Kirsten M. Kahler2, Cecilia Koble1, Anna Leivers1, Martin R. Leivers1, Amanda Mathis4,
Andrew J. Peat1, Jeffrey J. Pouliot1, Christopher D. Roberts1, Vicente Samano1, Rachel M.
Schmidt2, Gary K. Smith2, Andrew Spaltenstein1, Eugene L. Stewart2, Pia Thommes3, Elizabeth
M. Turner1, Christian Voitenleitner5, Jill T. Walker1, Greg Waitt2, Jason Weatherhead1, Kurt
Weaver2, Shawn Williams2, Lois Wright2, Zhiping Z. Xiong1, David Haigh3, and J. Brad
Shotwell1,*
1GlaxoSmithKline, Infectious Diseases Medicines Discovery Unit, 5 Moore Drive, Research
Triangle Park, NC 27709ꢀ3398
2GlaxoSmithKline, Platform Technology and Science, 5 Moore Drive, Research Triangle Park,
NC 27709ꢀ3398
3 GlaxoSmithKline, Infectious Diseases Centre for Excellence for Drug Discovery, Gunnels
Wood Road, Stevenage, Hertfordshire, UK SG1 1NY
4 Current Address: Salix Pharmaceuticals Inc., 8510 Colonnade Center Drive, Raleigh, NC
27615
ACS Paragon Plus Environment

Page 3 of 43
Journal of Medicinal Chemistry
1
2
5Current Address: Gilead Sciences, 333 Lakeside Drive, Foster City, CA 94404
KEYWORDS: HCV RNAꢀDependent RNA Polymerase, boron, NS5B
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ABSTRACT: A boronic acid moiety was found to be a critical pharmacophore for enhanced in
vitro potency against wild type hepatitis C replicons and known clinical polymorphic and
resistant HCV mutant replicons. The synthesis, optimization, and structureꢀactivity relationships
associated with inhibition of HCV replication in a subꢀgenomic replication system for a series of
nonꢀnucleoside boronꢀcontaining HCV RNAꢀDependent RNA Polymerase (NS5B) inhibitors are
described. A summary of the discovery of 3 (GSK5852), a molecule which entered clinical trials
in subjects infected with HCV in 2011, is included.
Introduction. Of all new chemical entities approved by the US FDA only bortezomib
incorporates the element boron.1 Despite the opportunity for a tremendous increase in
compositional diversity and novelty2 via the incorporation of boronic acids or related boron
heterocycles into drug candidates, boron remains poorly represented in the patent literature and
in the clinic. This lack of representation has perhaps at its roots the perception within the
medicinal chemistry community that boron functions as an intrinsically (chemically or
metabolically) unstable Lewis acid. Recently a growing awareness has emerged that this view is
likely a false one.3,4,5 Reports describing the characterization of novel boronꢀcontaining
substructures6 in the context of both preꢀclinical7,8,9 and clinical10,11 efforts are becoming more
common.
Boronic acids have been wellꢀcharacterized as pharmacophores for the reversible covalent
inhibition of hydroxyl proteases,12 and recently we13 and others14 have disclosed several series of
potent inhibitors of the HCV protease which incorporate a boronic acid warhead at the P4 and P1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 43
1
2
3
4
5
6
7
8
9
sites respectively. We became interested in a unique opportunity to profile boronic acid isosteres
within the context of HCV RNAꢀdependent RNA polymerase (RdRp) inhibitors. The NS5B
protein consists of 591 amino acids and is responsible for synthesizing new viral RNA. We
envisioned 115 (Figure 1), a prototypical NS5B Palm II inhibitor, could serve as a starting
scaffold for leveraging boronꢀbased chemistry. We recognized that the orientation of the Palm II
pocket allowed for the positioning of an inhibitor with a boronic acid (or other polar moiety) near
the catalytic site and/or RNA/NTP channel. Here, a boronic acid could also interact directly with
the incoming NTP or the growing RNA chain itself. 16 Boronic acids are known to couple to the
ribose units of nucleotides, 17 and modeling suggested the benzofuran scaffold of 1 could position
such a group within proximity of a growing viral RNA chain. There was also an opportunity to
specifically interact with key nucleophilic residues (for example S365) flanking the Palm II
allosteric site by a direct reversible covalent bond, which could enhance the desirable slow
binding kinetics reported for 1.18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Consistent with the high sequence conservation of the Palm II site, 1 exhibits similar potency
across wildꢀtype genotypes 1ꢀ6 (as assessed by HCV replicons). This is in contrast to inhibitors
that bind to NS5B at other allosteric sites. Despite initially promising clinical results,
progression of 1 was ultimately discontinued due to safety findings.19 Additionally, a significant
shift in potency against genotype (GT) 1b 316N, a key polymorph present in 40% of GT1b
patients,20 potentially limited efficacy in a large number of patients. Additionally, the emergence
of highly resistant GT1a virus bearing a C316Y mutation during clinical trials in patients treated
with 1 was observed. Generally, the design of allosteric inhibitors of NS5B has been
complicated by (i) a lack of sequence homology within most known allosteric binding sites of
the diverse HCV genotypes 1ꢀ621, (ii) intraꢀgenotype polymorphism within the allosteric sites in
ACS Paragon Plus Environment

Page 5 of 43
Journal of Medicinal Chemistry
1
2
significant portions of the HCV patient population (including 316N in Palm II)22, and (iii) the
propensity for rapid emergence of drugꢀresistant mutations as a result of the low genetic barrier
to resistance associated with previously reported allosteric inhibitors and the errorꢀprone
replication of HCV.23,24 It was therefore of interest to investigate whether introduction of a
boronic acid pharmacophore might provide unique advantages for targeting a subset of NS5B
polymorphs and resistant mutants. As multiple ligandꢀbound crystal structures were available
for this protein, structureꢀbased design of boronꢀbased Palm II inhibitors offered a direct
approach to explore robust HCV inhibition by addressing known mutational deficiencies (GT
1a316Y, GT 1b316N, GT 1b316Y) of 1. Herein we report the discovery, synthesis, and structure
activity relationships of potent, panꢀgenotype boronic acid HCV polymerase inhibitors where the
boronic acid functionality was found to be a key pharmacophore for replicon potency. We
include the first crystallographic data for clinically relevant GT1a and GT1b NS5Bs (316Y), an
efficient diastereoselective synthesis of novel chiral oxaborole 2, evidence of slow binding
kinetics to isolated NS5B protein, and the first disclosure of boronic acid 3 (GSK5852, Figure 1),
an aryl boronic acid which was progressed to proof of concept studies in human HCV subjects in
2011.25
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Palm II site structural biology, modeling, and initial design strategy. Of the four well
established NS5B allosteric inhibitor sites, the Palm II nonꢀnucleoside inhibitor (NNI) site is the
most proximal to the catalytic site of the polymerase.21 Crystallographic studies carried out with
1 demonstrated that the Palm II site is strategically positioned adjacent to the conserved catalytic
GDD motif at the junction formed between the incoming nucleotide triphosphate (NTP) substrate
and RNA template channels.18 This key RNA junction serves as the active site for RNA
priming, initiation, and elongation. The Palm II inhibitor pocket consists of a highly conserved
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 43
1
2
3
4
5
6
7
8
9
hydrophobic cleft formed between the serineꢀrich S365CSS loop and P197ꢀS210 helix of the palm.
Typically, this pocket is induced upon binding of the ligand, whereby the sidechains of
conserved R200 and S365 rotate to accommodate inhibitors.18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
A model of HCV RdRp with a growing RNA strand was constructed using the homologous polio
RdRp:RNA crystal structure. Boronic acid containing derivatives of 1 were then added and the
location of the boronic acid was optimized in silico. Optimization was guided by potential
interactions with the incoming NTP or growing RNA strand. Figure 2 depicts the binding mode
of 3 in the Palm II site. While the majority of inhibitor 3 contacts are located in the palm domain
(gray), there are also contacts with the βꢀhairpin flap (yellow), specifically Y448 (See Figure 5
and 7). Y448 is conserved across HCV genotypes. The βꢀflap has been implicated in the
regulation of polymerase initiation and the binding of template RNA. 26 More recently, Y448 has
been reported to orient GTP for polymerase de novo initiation.27 We superimposed NS5B with
experimental structures of related RdRps, some including the presence of coꢀcrystallized viral
RNA, to assess variation of the Palm II site in the presence or absence of RNA. Based on these
structural comparisons, we found either an aryl boronic acid (as depicted for 3 in Figure 2) or an
extension of the hydroxyl ethyl sidechain present in 1 with a methylene boronic acid moiety (i.e.
5, Scheme 1), would position the polar boronic acid moiety close to the putative incoming
NTP.27,28 Figure 2b depicts the location of the aryl boronic acid of 3 at the center of the
NTP/RNA channel. This view is magnified in Figure 2c in the context of incoming NTP (a GTP
dinucleotide primer) by superposition with a HCV GT1a construct.29 Similarly, Figure 2d
depicts the location of 3 relative to the viral RNA primerꢀtemplate duplex, based on NS5B
superposition with poliovirus RdRp in “active” complex with an extended RNA primerꢀtemplate
duplex.30 The orientation of the nucleotides in the GT1a NS5B:GTP structure31 is in register
ACS Paragon Plus Environment

Page 7 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with the baseꢀpairing of the polio RdRp:RNA structure. Thus, homology modeling suggested
that 3 and related analogs could disrupt polymerase initiation, as well as priming of the +0
nucleotide position. Therefore, locating a boronic acid at this position suggested potential
inhibition of NS5B through multiple mechanisms, including restriction of the βꢀflap to an
inactivated state and/or direct interference of RNA processing. We anticipated this strategy
could offer novel opportunities to resolve many of the deficiencies associated with 1 (and
generally other NS5B NNIs), particularly with respect to achieving potency across polymorphic
NS5B variants.32,33
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Chemistry and Structure-Activity Results and Discussion. Initially this hypothesis was tested
by directly modifying the sulfonamide of 1 by extension of the hydroxyethyl moiety or
introduction of an aryl linker group to access the NTP/RNA channel. Alkylation of 1 directly
with commercially available 2ꢀ(bromomethyl)ꢀ4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolane and
potassium carbonate in DMF followed by acidic deprotection of the pinacol ester in the presence
of polymerꢀsupported benzeneboronic acid to sequester residual pinacol afforded boronic acid 5
(Scheme 1).34,35 Similarly, direct alkylation of sulfonamide 6 with either 3ꢀ or 4ꢀ
(bromomethyl)phenyl)boronic acids afforded 7 and 8, respectively.34,35 At first, boronic acids 5,
7, and 8 were profiled alongside 1 in three cellꢀbased HCV replication systems (replicon GT1a,
GT1b, and GT1b 316N; after 2 day incubations) as well as in a GT1b biochemical polymerase
assay (Table 1). Boronic acids 5, 7, and 8 demonstrated improved potency in the HCV replicon
systems relative to 1 (Table 1). In addition to modest (2ꢀ5 fold) improvements against GT1a and
GT1b wildꢀtype replicons, 5 showed a roughly 40ꢀfold improvement in potency relative to 1
against a GT1b replicon containing key clinical polymorph 316N (Table 1). The aryl boronic
acid 7 also demonstrated improvements in in vitro potency against this polymorph. The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 43
1
2
3
4
5
6
7
8
9
differences in biochemical potency with isolated NS5B were not as apparent as IC50s in this
assay were within error for all analogs profiled and likely approached the limit of detection in
our enzyme assay (Table 1, see Supporting Information for protocol, NS5B concentration = 30
nM). To test whether the boronic acid was unique relative to other polar functionality with
respect to 1b316N replicon potency, we profiled a number of direct boronic acid replacements
for 7 including a phosphate (9), carboxylic acid (10), sulfonic acid (11), phosphonate ester (12)
and phenyl ring (13). The corresponding dimethyl phosphate, methyl amide, sulfone, nitrile,
methyl ester, phenol, sulfonamide, and hydroxyamide (see Supporting Information Table SIꢀ3)
were also profiled. Consistent with our hypothesis, 5 showed inhibition of GT1b C316N with a
shift relative to WT 1b replicon potency much less than other analogs. While polar anionic
functionality in general afforded potent biochemical inhibitors (i.e. 9ꢀ11), as anticipated these
inhibitors suffered from poor permeability. This was in contrast to the potency and good
permeability seen with boronic acids 5, 7, and 8. Notably, even those analogs which had good
permeability (1, 12, 13) demonstrated a characteristic shift between GT1b and GT1b C316N
replicon potency (≥ 5 fold) whereas linear boronic acid 5 was equipotent when profiled in both
replicons (See Supplementary Table SIꢀ7 for chromatographic log D, artificial membrane
permeability, and MDCK permeability data for 1, 3, 5, 7ꢀ13). We considered a shift in potency
against GT1b 316N for 1 a risk to be recapitulated in human clinical trials and as such we
considered it an important property requiring optimization. Incorporation of a boronic acid
functionality elsewhere on the benzofuran core, specifically replacement of the pendant methyl
amide with boron functionality anticipated to engage conserved S365 directly, did not result in
improved antiviral properties (see Supporting Information Table SIꢀ4). As such, additional
efforts focused on targeting the RNA/NTP tunnel with boronꢀcontaining groups.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Page 9 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Both boronic acids 5 and 7 exhibited poor in vivo DMPK in rat. Linear boronic acid 5 was
readily oxidized to 1 in vivo in rat (Cl = 25 ml/min/kg) while 7 demonstrated <3% bioavailability
from a solution dose with moderate clearance in rat (Cl = 36 mL/min/kg, see Supporting
Information Table SIꢀ6ꢀ9 for full in vivo rat PK characterization of 2, 3, 4, 5, 7, and 17 with
liver:plasma partitioning for 3). Addtionally the in vitro characterization of 3 including
microsomal stability, P450 inhibition, MDCK permeability, glutathioneꢀtrapping of metabolites,
replicon potency in the presence of 40% human serum, and aqueous solubility is included).
However, we were encouraged by the initial improvements in in vitro potency, especially the
dramatic improvement in potency for the 316N polymorph, and designed a number of additional
inhibitors to improve the pharmacokinetic properties of 5 via incorporation of the boronic acid
moiety into a ring. For example, direct alkylation of benzofuran 6 with known allylic bromide
1435 followed by rhodium catalyzed hydroboration afforded benzyl ether 15. Hydrogenation and
pinacol deprotection afforded oxaborole 16 in excellent yield (Scheme 2). A number of related
oxaboroles (Table 2) could be prepared in a similar manner utilizing appropriately substituted
olefins in place of 9 and subsequent hydrogenations.35 With these substrates, we did not observe
boronꢀcarbon bond reduction with hydrogenations performed at room temperature and up to 40
PSI H2. Overall oxaboroles 16ꢀ22 were stable, easily purified, and showed no cell toxicity in
Huhꢀ7 (replicon) cells (see Supporting Information Table SIꢀ5 for cellular toxicity data for all
analogs).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
This series of boronic acids maintained potent singleꢀdigit nanomolar inhibition of both GT1a
and GT1b replicons (Table 2). Comparison of 18 to linkerꢀshortened 16 revealed some tolerance
to the location of the presentation of the boronic acid within the NS5B NTP/RNA channel. In
general, 5ꢀmembered ring oxaboroles (16 and 18) were more potent than their 6ꢀmembered ring
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 43
1
2
3
4
5
6
7
8
9
congeners (21 and 22) against the GT1b 316N polymorph. Resolution of 16 using supercritical
fluid chromatography (SFC) was successful on milligram scale. The absolute configuration of 2
was initially assigned using ab initio vibrational circular dichroism spectroscopy (VCD) (See
Supporting Information Figure SIꢀ10). Assaying of the resolved enantiomers revealed no
preference for (R)ꢀ2 relative to (S)ꢀ17 or unsaturated 19 in terms of potency against 316N. A
much more dramatic difference in pharmacokinetic profile was observed however with clearance
in rat for 2 and 17 measured as 6.6 and 32 ml/min/kg respectively (see Supporting Information
Table SIꢀ6 for full PK characterization).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The initial SFCꢀbased resolution of 16 proved capricious, allowing access to only milligram (<
25 mg) quantities of 2. Accordingly, we turned toward the diastereoselective synthetic approach
illustrated in Scheme 3 to access quantities of 2 sufficient for in vivo PK in higher species and a
full characterization of its HCV resistance profile. Copperꢀmediated borylation of 23 afforded
the corresponding βꢀpinacol boronate.36 Subsequent exposure to TiCl4/DIPEA followed by
BOMCl afforded the desired aldol adduct 24 in 47% yield with >10:1 diastereoselectivity over
two steps. Direct reduction to alcohol 25 followed by Mitsunobu esterification with 6 and
subsequent hydrogenation afforded 2. The absolute configuration assigned earlier by VCD was
consistent with the anticipated diastereoselectivity from the aldol reaction and analysis using the
earlier separation method indicated 2 was obtained in >95% ee. Notably, the present approach is
complementary to recent reports by Molander wherein ephedrineꢀbased auxilliaries are
implemented to install boron moieties (via alkylation with bromoꢀmethyl dioxaborolanes) in the
diastereoselective aldol step.37 Due to the propensity for βꢀhydroxy amides to undergo
elimination under related alkylation conditions, the synthesis of 2 was best accomplished through
ACS Paragon Plus Environment

Page 11 of 43
Journal of Medicinal Chemistry
1
2
3
4
the strategy illustrated in Scheme 3. Indeed it was found that a βꢀboryl group was compatible
5
with the aldol chemistry and not prone to elimination.
6
7
8
9
Access to gram quantities of 2 allowed for its broader virologic characterization and higher
species DMPK profiling. Benzofuran 1 has shown significantly reduced potency against a
number of other mutations near the palm site and a marked shift in its ability to inhibit replicon
carrying the 316N polymorph. Specifically, other mutations at residues C316 and S365 have
been reported as particularly problematic, giving rise to EC50 shifts of >100 fold in most cases.
Oxaborole 2 shows an improvement in its ability to inhibit the replication of these mutant GT1b
replicons relative to 1 (Table 3), but showed a 100 fold loss in potency for a GT1a 316Y mutant
replicon (EC50 = 830 nM) relative to wild type replicon (EC50 = 7.6 nM, Table 2). The poor
activity against GT1a 316Y mutant virus was of particular concern, as the mutation was
observed in individuals treated with 1 monotherapy indicating that patient viruses containing
GT1a 316Y are viable resistant mutants. Additionally, further in vitro fitness increases of nearly
tenfold have been reported with secondary fitness mutations in combination with 316Y in
replicons.18, 20, 32
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
To minimize the potential GT1a 316Y liability for 2, we assessed the activity of other
compounds in the series to ascertain their profiles against GT1a 316Y. Initially we focused on
boronic acid 7 for further structureꢀactivity studies because it was more potent than 2 against
GT1a 316Y (Table 3). Because this potency was improved (7ꢀfold) relative to 2, in addition to
representing a smaller foldꢀshift relative to the GT1b 316N potency, we synthesized a set of
functionalized aryl boronic acids in an attempt to further enhance potency. To access bicyclic
boron containing motifs, 26 was subjected to sequential carbonylation, Sandmeyer, and radical
bromination conditions which afforded 27. Direct alkylation of 6 with 27 followed by borylation
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 43
1
2
3
4
5
6
7
8
9
under standard conditions afforded 28 (Scheme 4). The boronic acid moiety proved stable to a
variety of conditions, including direct reduction with superhydride to give 4 (29-33 were
prepared in an analogous manner).34,35 Clinical candidate 3 and related ortho substituted boronic
acids were prepared in a similar manner to 7 via direct alkylation with the corresponding benzyl
bromide containing either the boronic acid moiety (for 3,7, and 8) or an aryl bromide precursor
which was subsequently converted to the boronic acid (for 34 and 35).34,35
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The structureꢀactivity trends for the oxaborole analogs (Table 4) were consistent with those
observed for 7 and 8. Specifically, analogs which have a para substituted aryl boronic acid
relative to the connection to the core (4 and 29) proved superior to analogs with metaꢀdisposed
boronic acids (30 and 31) when profiled against the 1b316N polymorph. This observation is
consistent with the predicted positioning of the boronic acid within the NTP/RNA tunnel. Other
arrangements of the boronic acid about the aryl ring including chiral 32 and 6ꢀmembered ring 33
showed promising potencies across wild type replicons. Additionally, orthoꢀsubstitution
adjacent to the boronic acid was well tolerated, with fluoroꢀanalog 3 having higher potency to
electronꢀdonating methoxy 34 or the methyl analog 35 against the GT1b 316N replicon.
Installation of fluorine ortho to the boron moiety of 29 afforded 4, but did not result in a boost in
GT1b 316N potency relative to 29 as was observed for the 7, 3 pair.
As observed in other series, direct inhibition of isolated NS5B protein was relatively invariant to
compound, while the cellꢀbased replicon assays proved to be more discriminating. Although 3,
4, 32, and 33 showed similar activities against GT1b (≤ 6ꢀfold) and the GT1b 316N polymorph
(Table 4), substantial differences in potencies were observed against other HCV mutant replicons
(Table 3). For example a ≥ 50ꢀfold difference in in vitro GT1a 316Y potency was observed
between the compounds, with only 3 potently inhibiting this mutant (EC50 < 10 nM).
ACS Paragon Plus Environment

Page 13 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Although 3 was a potent inhibitor, its mechanism of action was not well understood. Potency in
the replicon system had been systematically optimized, but a relationship between the cellꢀbased
and biochemical potency as measured for the series had not been established (especially for 1 vs.
3). Additionally, although in silico modeling has been utilized by others,18 no crystallographic
data for GT1a or GT1b NS5B possessing a 316Y mutation had been previously reported.
Unfortunately, the models offered little explanation for the exceptional potency of 3 relative to 2
or 4 against the GT1a 316Y mutant.18 To address these concerns, we prioritized obtaining both
biochemical kinetic data and crystallographic structural data so as to better understand the unique
profile of 3 and to guide future compound design.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Measurement of koff for 3 and 1. The enzyme IC50 data in Tables 1, 2, and 4 were determined
after a preꢀincubation of enzyme with inhibitor for 15 minutes. Suggestive of slowꢀbinding
kinetics, increased preincubation time to 45 or 90 minutes results in increased potency of 3 (data
not shown). Accordingly, the primary driver for compound potency throughout the optimization
phase of this series focused on the cellꢀbased replicon system (2ꢀday compound incubation).
Offꢀrates for 3 and 1 were measured to further explore the slowꢀbinding kinetics with isolated
GT1b 316N protein. The enzyme was incubated with either 3 or 1 to preꢀform an
enzyme:inhibitor complex. The solution was then rapidly diluted 20ꢀfold into an excess of the
other inhibitor (that is, the 3:NS5B complex was diluted into excess 1, and vice versa) to allow
the bound inhibitor to dissociate and to prevent rebinding. The rate of dissociation of the
inhibitor from the enzyme was determined by rapid size exclusion filtration of the
NS5B:inhibitor complexes. This procedure was followed by measurement of the concentration
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 43
1
2
3
4
5
6
7
8
9
of each bound inhibitor by mass spectrometry. The results of these experiments are shown in
Figure 3. Our method measures both 3 and 1 simultaneously, so that 3 (top panel) dissociation is
measured both as a decrease in bound 3 and an equal increase in bound 1. Similarly, dissociation
of 1 (bottom panel) is measured both as a decrease in bound 1 and as an increase in bound 3. As
shown in Figure 3, the dissociation of 3 is qualitatively much slower than dissociation of 1 from
isolated GT1b 316N NS5B protein. These data are consistent with similar experiments carried
out using 3Hꢀ3 (not shown). Dissociation rate constants were calculated from these data. The
dissociation of 1 was best fit to a single exponential that gives a calculated t1/2 of 1.7 hours, and
is in good agreement with previous reports that state the 1:NS5B complex requires two hours to
reach equilibrium.18 We attempted to fit the dissociation of 3 to a single exponential, however,
the data are better fit to a mechanism in which 21% of the compound dissociates with a “fast”
offꢀrate, t1/2 value of 4.3 hrs, and 79% dissociates with a much slower offꢀrate, t1/2 value of 69.3
hours. The determination of two dissociation rates can be indicative of a structural
reorganization of the enzyme inhibitor complex. Overall, the data show that the majority of 3
dissociates from NS5B approximately 40ꢀfold slower than does 1. The slower dissociation rate
is consistent with an increased GT1b 316N replicon potency for 3 (1.9 nM) relative to 1 (100
nM).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 3 inhibits NS5B Initiation. Due to the partial overlap in compound binding sites, it
has been suggested that allosteric NS5B Palm II binders should behave mechanistically similar
to NS5B Palm site I binders and inhibit the initiation step of the polymerase cycle. This concept
was tested directly for 3 by measuring inhibition of a biochemical assay that used purified NS5B
to catalyze de novo initiation. An RNA template of sequence 5’ꢀA8CGꢀ3’ was used to drive
production of a pCpG dimer product in a nonꢀprimer driven reaction. NS5B was capable of
ACS Paragon Plus Environment

Page 15 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
adding GTP to a radiolabeled CTP nucleoside in uninhibited reactions (Figure 4, Lane 1).
Addition of increasing amounts of 3 inhibited this reaction (Figure 4, Lanes 3ꢀ7) until no product
was visible at high concentrations of inhibitor. This de novo initiation assay demonstrates 3 is
capable of blocking initiation and 3 is the first Palm II NS5B inhibitor for which experimental
evidence has been published supporting this mechanism of action.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NS5B:Inhibitor Crystallographic Studies. HCV NS5B proteins from GT1b 316N, GT1a,
GT1b 316Y, and GT1a 316Y were cloned, expressed, purified, and found to be suitable for
crystallography. All proteins were crystallized and then coꢀincubated with 1, 3, 4 and 32. Figure
5 shows crystal structures of GT1a, GT1a 316Y, and GT1b 316N NS5B proteins with 3. In all
structures the orientation of the inhibitor core is essentially identical. The pendant fluoro phenyl
ring occupies the hydrophobic recess of the Palm II cleft and the benzofuran core forms a cationꢀ
π interaction with R200. R200 is conserved across all HCV genotypes and plays a critical role in
the binding mode of 3 in all NS5B constructs. Relative to the apo crystal structure, R200 pivots
to accommodate and interact with 3. Additionally, R200 anchors a network of key hydrogen
bond interactions to the sulfonamide and carboxamide groups of the inhibitor, the backbone
carbonyl of G192, and theY448 hydroxyl. S365 also consistently hydrogen bonds with the
carboxamide of 3 in the pocket. Towards the solvent/RNA front, the benzyl boronic acid group
of 3 forms favorable interactions with M414 and Y448. The boronic acid also has interactions
with bound water molecules. In the GT1a 316Y mutant (Figure 5b) F193 rotates to
accommodate the C316Y mutation, as 316Y is oriented towards the solvent front in proximity to
the boronic acid of 3. In Figure 6a, the boronic acid of 3 is observed to form a throughꢀwater
hydrogen bond with 316Y in the 1b316Y mutant, while compound 4 and 32 are able to make
direct hydrogen bonds with 316Y (Figure 6b and Figure 6c). In contrast, the hydroxyl ethyl
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 43
1
2
3
4
5
6
7
8
9
sidechain of 1 is oriented away from 316Y, though a water is observed to contact 316Y (Figure
6d). R200 is also flipped in the interaction with 1 compared its orientation with 3, 4, and 32.
Contrary to our original hypothesis, the tremendous potency enhancements associated with the
boronic acid functionality of 3 do not appear to be due to the formation of a covalent bond with
any nearby residue. Additionally, we have seen no crystallographic evidence of hydration or a
charged boronate.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
From the standpoint of inactivating the βꢀhairpin flap which regulates polymerization, the
binding of 3 completes the formation of a hydrophobic pocket (comprised of F193, P197, M414,
F551 and 3) in which Y448 is buried (Figure 7). Y448 is located at the tip of the βꢀflap and is
conserved across all HCV genotypes. The benzyl ring of 3 provides a hydrophobic buttress
against Y448, while the Y448 hydroxyl forms an intermolecular hydrogen bond with the
sulfonamide of 3 and/or an intraꢀNS5B hydrogen bond with R200 in the back of this pocket.
This is consistent with the SAR of 13, which has reasonable replicon potency (Table 1, GT1a
EC50 = 58 nM, GT1b EC50 = 18 nM) and features only a simple unsubstituted benzyl ring.
Removal of the benzyl ring substantially degrades replicon potency (EC50 > 500 nM).
Restricting Y448 in this manner likely traps the βꢀflap in an inactivated state, inhibiting
polymerase initiation, RNA binding, and progression to RNA elongation.
All attempts to coꢀcrystallize RNA in the presence of inhibitors were unsuccessful. Recently, a
HCV NS5B GT2a construct, with the βꢀflap deleted, enabled resolution of NS5B in complex
with viral primerꢀtemplate RNA.27 Applying NS5B homology, superposition of these RNAꢀ
bound structures with inhibitor coꢀcrystal structures reported here suggests 3 can also interact
with the +0 nucleotide of the primer strand to disrupt polymerization, in addition to inactivating
ACS Paragon Plus Environment

Page 17 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the βꢀflap/Y448. Intriguingly, superposition yields the ribose of the +0 nucleotide oriented
towards the boronic acid of 3, such that the 2’ hydroxyl is within 3Å of the boron (Figure SIꢀ12,
Supplementary Information), suggestive of a transient covalent interaction with the ribose or
potential steric interference. Figure 8 depicts the superposition of the HCV NS5B βꢀflap along
the axis of the double stranded RNA product resulting from polymerization, emphasizing the
inactivated conformation of the regulatory βꢀflap (yellow).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Taken together, comparison of crystallographic data suggests that 3 inhibits polymerization
primarily through two mechanisms: (i) stabilization of the βꢀflap in a closed, inactivated state to
inhibit template binding and primer initiation, including restricting Y448 to participate in de
novo initiation (consistent with the initiation studies in Figure 4), and/or (ii) disruption of the
RNA processing channel either through direct steric contacts or transient adduct formation
between the +0 ribose and the boronic acid of 3. Notably, the introduction of boronic acid
moieties uniquely enhance the robustness of replicon inhibition. Structural data reveals direct
and throughꢀwater hydrogen bond contacts with the boronic acid and key mutants, (particularly
316Y). Additional interactions, potentially including interaction with viral RNA, or altered
NS5B:inhibitor interactions in the presence of RNA, await further characterization.
Conclusions. Despite a growing number of reports highlighting applications of boronic acids
within the context of inhibitor design, few studies describe the optimization of boronꢀcontaining
molecules with a balance of properties ultimately suitable for clinical investigations in man. We
have discovered a series of potent inhibitors of the NS5B polymerase wherein a boronic acid
moiety has been demonstrated to be a critical pharmacophore. Potency against key resistance
mutants was optimized to provide clinical candidate 3. Clinical candidate 3 shows slow binding
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 43
1
2
3
4
5
6
7
8
9
kinetics with isolated GT1b 316N protein, has been characterized with a dissociation halfꢀlife of
> 40 hours (Figure 3), and inhibits the initiation step of the polymerase RNA replication cycle
(Figure 4). We have reported elsewhere a full virologic (including several resistant replicons
specific to 3)33 and pharmacokinetic25 characterization of 3 in mouse, rat, dog, monkey, and man.
Notably, 3 showed a statistically significant reduction in serum HCV ribonucleic acid (RNA)
following a single dose of 420 mg (ꢀ1.33 log10 copies/mL) compared with placebo (ꢀ0.09 log10
copies/mL) at 24 h postꢀdose in subjects infected with hepatitis C virus but is not being further
pursued in clinical trials.25
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Acknowledgements. We thank Dean Phelps for the VCD analysis for compound 2. All authors
are current or former employees of GlaxoSmithKline.
Experimental Section
HCV Stable and Transient Replicon assays and NS5B polymerase assays. Experimental
protocols and method for IC50 determination can be found in the Supporting Information.
NS5B Initiation Assay. The initiation assay utilized genotype 1b NS5B (strain BK) containing
the mutations L47Q, F101Y, and K114R, a Cꢀterminal ꢀ21 truncation, and a Cꢀterminal
hexahistidine tag to drive de novo initiation from an oligo template of the sequence 5’ꢀ
AAAAAAAAGCꢀ3’ (Integrated DNA Technologies, Coralville IA). Reactions contained 2.7 ꢁM
NS5B, 1 mM GTP, 100 ꢁM CTP, 10 ꢁCi [α33P]ꢀCTP, and 23 ꢁM RNA template in a buffer of
50 mM HEPES pH 7.3, 10 mM magnesium chloride, and 5 mM dithiothreitol, and were begun
either by addition of enzyme or by addition of RNA template. After a 30 minute incubation at
30oC, reactions were terminated by addition of 90% formamide and freezing on dry ice.
ACS Paragon Plus Environment

Page 19 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
Samples were run on 25% polyacrylamide 7M urea TBE gels and quantitated with a Typhoon
system.
7
8
9
Compound Purities & Identity. All compounds underwent an immediate processing quality
control protocol to confirm identity and determine relative purity immediately before processing
in the replicon or NS5B polymerase assays. Analysis is by UPLCꢀMS with UV diode array
detection to determine purity and MS was used to confirm molecular weight. A Waters Acquity
UPLC system comprising Binary Solvent manager, Sample Manager, PDA Detector, Waters ZQ
or SQD mass spectrometer, Waters Acquity Evaporative Light Scattering Detector or Polymer
Laboratories Evaporative Light Scattering Detector were employed. Mobile phases: acetonitrile
+ 0.1% formic acid; water + 0.1% formic acid. Wash solutions: strong wash 100% acetonitrile +
0.1% formic acid; weak wash 50:50 acetonitrile:water + 0.1% formic acid. All individual lots of
compounds tested ≥ 95% pure (unless otherwise noted) according to this protocol (see
Supporting Information Table SIꢀ2 for purity ranges).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Absolute Stereochemical Assignments by VCD. Absolute stereochemical assignments for 2
and 17 were initially made by comparing an ab initio VCD analysis (see Supporting Information
Section VI) with spectra obtained from enantiomerically pure (> 95% ee) samples of 2 and 17
obtained by resolution of 16 using supercritical fluid chromatography (25% MeOH/CO2, 140
bar, 10 mL/min, ChiralPak IC 10X250 mm, 254 nM). The estimated confidence limit of this
analysis (see Supporting Information Figure SIꢀ11) was 94% and afforded absolute
configurational assignments consistent with the synthesis of 2 using the diastereoselective route
described in Scheme 3.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 43
1
2
3
4
5
6
7
8
9
Diastereoselective synthesis of 2. (4R,5S)-3-acryloyl-4-methyl-5-phenyl-1,3-oxazolidin-2-one
(23). A solution of 2ꢀpropenoic acid (12.59 mL, 183 mmol) and triethylamine (49.2 mL, 353
mmol) in tetrahydrofuran (900 mL) maintained at ꢀ20°C was treated dropwise with acryloyl
chloride (13.8 mL, 169 mmol). A thick white precipitate formed and the mixture was maintained
with stirring at ꢀ20°C for 1 hour. Lithium chloride (7.78 g, 183 mmol) was added (as a solid) in
one portion followed by (4R,5S)ꢀ4ꢀmethylꢀ5ꢀphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone (25g, 141 mmol) as a
solution in tetrahydrofuran (150 mL). The mixture was warmed to room temperature,
maintained with stirring for four hours, and concentrated under reduced pressure. The solids
were dissolved in dichloromethane and water and the organic layer was washed with brine, dried
over sodium sulfate, filtered, stripped onto celite, and purified by column chromatography to
afford (4R,5S)ꢀ3ꢀacryloylꢀ4ꢀmethylꢀ5ꢀphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone (23) (25.26 g, 109 mmol, 77
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
% yield) as a clear oil. H NMR (400 MHz, CDCl3) δ ppm 7.36 ꢀ 7.60 (m, 4 H) 7.29 ꢀ 7.36 (m, 2
H) 6.58 (dd, J=17.00, 1.76 Hz, 1 H) 5.93 (dd, J=10.36, 1.76 Hz, 1 H) 5.72 (d, J=7.23 Hz, 1 H)
4.73 ꢀ 4.93 (m, 1 H) 0.95 (d, J=6.64 Hz, 3 H).
(4R,5S)ꢀ3ꢀ((S)ꢀ3ꢀ(benzyloxy)ꢀ2ꢀ((4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)methyl)propanoyl)ꢀ4ꢀmethylꢀ5ꢀphenyloxazolidinꢀ2ꢀone (24). A suspension of sodium tertꢀ
butoxide (0.419 g, 4.36 mmol), DPEPhos (0.78 g, 1.45 mmol), and copper(I) chloride (0.14 g,
1.45 mmol) in THF (50 mL) was stirred at room temperature, cooled to 0°C, and treated
dropwise with bis(pinacolato)diboron (12.91 g, 50.9 mmol) in THF (50 mL). Stirring was
maintained for 45 minutes and then (4R,5S)ꢀ3ꢀacryloylꢀ4ꢀmethylꢀ5ꢀphenylꢀ1,3ꢀoxazolidinꢀ2ꢀone
(23) (11.20 g, 48.4 mmol, in THF, 50 mL) was added dropwise. The solution was warmed to
room temperature and maintained with stirring for 3 hours. The mixture was passed through
ACS Paragon Plus Environment

Page 21 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
glass filter paper, diluted with celite, concentrated, and purified by column chromatography to
afford intermediate (4R,5S)ꢀ4ꢀmethylꢀ5ꢀphenylꢀ3ꢀ[3ꢀ(4,4,5,5ꢀ tetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
7
8
9
1
yl)propanoyl]ꢀ1,3ꢀoxazolidinꢀ2ꢀone (8.13 g, 22.63 mmol, 47% yield) as a white solid. H NMR
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(400 MHz, METHANOLꢀd4) δ ppm 7.32 ꢀ 7.47 (m, 5 H) 5.78 (d, J=7.41 Hz, 1 H) 4.77 (t, J=6.93
Hz, 1 H) 2.86 ꢀ 3.16 (m, 2 H) 1.16 ꢀ 1.31 (m, 12 H) 0.95 (q, J=6.50 Hz, 2 H) 0.83 (d, J=6.63 Hz,
3 H). A solution of (4R,5S)ꢀ4ꢀmethylꢀ5ꢀphenylꢀ3ꢀ[3ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)propanoyl]ꢀ1,3ꢀoxazolidinꢀ2ꢀone (5.0 g, 13.92 mmol) in DCM (69.6 ml) was cooled to 0°C
and then treated by the addition of TiCl4 (1M in DCM, 16.70 ml, 16.70 mmol). The yellow
reaction was stirred for 10 minutes, then DIEA (3.04 ml, 17.40 mmol) was added to the reaction.
The reddish reaction was treated with chloromethyl phenylmethyl ether (2.51 ml, 18.09 mmol).
The mixture was slowly allowed to warm to room temperature with stirring for 2 hours. The
reaction was then diluted with additional dichloromethane, saturated aqueous NH4Cl, and water.
The layers were separated and the combined organic layers were washed with saturated NaHCO3
and brine. The combined organic layers were dried over Na2SO4, filtered, and concentrated to
afford
(4R,5S)ꢀ3ꢀ((S)ꢀ3ꢀ(benzyloxy)ꢀ2ꢀ((4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)methyl)propanoyl)ꢀ4ꢀmethylꢀ5ꢀphenyloxazolidinꢀ2ꢀone as a crude residue (quant. yield, used
without further purification).
(R)-3-(benzyloxy)-2-((4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)methyl)propan-1-ol (25).
A
solution of (4S,5R)ꢀ4ꢀmethylꢀ5ꢀphenylꢀ3ꢀ{(2R)ꢀ3ꢀ[(phenylmethyl)oxy]ꢀ2ꢀ[(4,4,5,5ꢀtetramethylꢀ
1,3,2ꢀdioxaborolanꢀ2ꢀyl)methyl]propanoyl}ꢀ1,3ꢀoxazolidinꢀ2ꢀone (24) (3.46g, 8.30 mmol) in
THF (41.5 ml) and MeOH (0.40 ml, 9.96 mmol) was placed under a nitrogen atmosphere, cooled
to 0°C, and treated with LiBH4 (2M solution in THF, 4.98 ml, 9.96 mmol). After stirring at room
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 43
1
2
3
4
5
6
7
8
9
temperature overnight, the reaction was quenched by the addition of MeOH and saturated NH4Cl
and extracted with ethyl acetate. The combined organics were washed with water, dried over
Na2SO4, filtered, and concentrated. The crude residue was purified by column chromatography
to afford (R)ꢀ3ꢀ(benzyloxy)ꢀ2ꢀ((4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)methyl)propanꢀ1ꢀol
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
(25) as a clear oil (1.03g, 41% yield). H NMR (400 MHz, METHANOLꢀd4) δ ppm 7.16 ꢀ 7.39
(m, 5 H) 4.48 (s, 2 H) 4.09 (q, J=7.23 Hz, 1 H) 3.51 ꢀ 3.62 (m, 1 H) 3.41 ꢀ 3.51 (m, 2 H) 3.33 ꢀ
3.41 (m, 1 H) 2.02 ꢀ 2.11 (m, 1 H) 1.20 (d, J=2.93 Hz, 12 H) 0.75 (dd, J=7.23, 2.93 Hz, 2 H).
5-cyclopropyl-2-(4-fluorophenyl)-6-[{[(4R)-2-hydroxy-1,2-oxaborolan-4
yl]methyl}(methylsulfonyl)amino]-N-methyl-1-benzofuran-3-carboxamide (2). A solution of 5ꢀ
cyclopropylꢀ2ꢀ(4ꢀfluorophenyl)ꢀNꢀmethylꢀ6ꢀ[(methylsulfonyl)amino]ꢀ1ꢀbenzofuranꢀ3ꢀ
carboxamide (175 mg, 0.435 mmol), (2R)ꢀ3ꢀ[(phenylmethyl)oxy]ꢀ2ꢀ[(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)methyl]ꢀ1ꢀpropanol (233 mg, 0.761 mmol), and triphenylphosphine (257 mg,
0.98 mmol) in tetrahydrofuran (5 mL) was maintained at 0°C with stirring and treated dropwise
with DIAD (0.169 mL, 0.870 mmol). The mixture was warmed to room temperature, maintained
with stirring for 16 hours, diluted with celite, concentrated, and the residue was absorbed on
celite. The product was purified by column chromatography (20 to 50% EtOAc/hexanes) to
afford
5ꢀcyclopropylꢀ2ꢀ(4ꢀfluorophenyl)ꢀNꢀmethylꢀ6ꢀ((methylsulfonyl){(2R)ꢀ3ꢀ
[(phenylmethyl)oxy]ꢀ2ꢀ[(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)methyl]propyl}amino)ꢀ1ꢀ
benzofuranꢀ3ꢀcarboxamide (235 mg, 0.340 mmol, 78 % yield) as a white foam with NMR
spectroscopic properties identical to 10. A suspension of 5ꢀcyclopropylꢀ2ꢀ(4ꢀfluorophenyl)ꢀNꢀ
methylꢀ6ꢀ((methylsulfonyl){(2R)ꢀ3ꢀ[(phenylmethyl)oxy]ꢀ2ꢀ[(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)methyl]propyl}amino)ꢀ1ꢀbenzofuranꢀ3ꢀcarboxamide (0.600 g, 0.869 mmol)
ACS Paragon Plus Environment

Page 23 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 10% Pd/C (Degussa Type (wet). (300mg, 0.869 mmol) in tetrahydrofuran (10 mL) was
maintained under 40 psi hydrogen gas with rapid stirring for 16 hours. The mixture was filtered
through celite, concentrated under reduced pressure, dissolved in tetrahydrofuran (10 mL) and
aqueous 5.0N HCl (1.216 mL, 6.08 mmol), and polymer supported benzene boronic acid (2.6
mmol/g) (1.67 g, 4.34 mmol) was added. The mixture was maintained with stirring at room
temperature for 2 hours, filtered through celite, concentrated, and partitioned between
dichloromethane and HCl (pH < 3). The organic layer was dried over sodium sulfate, filtered,
taken to a residue under reduced pressure, and purified by column chromatography to afford 5ꢀ
cyclopropylꢀ2ꢀ(4ꢀfluorophenyl)ꢀ6ꢀ[{[(4R)ꢀ2ꢀhydroxyꢀ1,2ꢀoxaborolanꢀ4ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
yl]methyl}(methylsulfonyl)amino]ꢀNꢀmethylꢀ1ꢀbenzofuranꢀ3ꢀcarboxamide (2) (297 mg, 0.594
mmol, 68.3 % yield) as a white foam with > 95% purity and > 95% ee% as determined by chiral
chromatography (see supporting information). The residue was dissolved in acetonitrile/water,
frozen, and stored on a lyophilizer for 16 hours to give a freeꢀflowing white solid. 1H NMR
(CDCl3) δ: 7.78ꢀ7.95 (m, 2H), 7.47 (d, J = 7.6 Hz, 1H), 7.29 (d, J = 2.1 Hz, 1H), 7.20 (m, 2H),
5.75 (br. s.,1H), 4.32 (br. s., 1H), 4.05 ꢀ 4.21 (m, 1H), 3.54 ꢀ 4.00 (m, 3H), 2.91 ꢀ 3.14 (m, 6H),
2.17 2.68 (m, 2H), 1.03 ꢀ 1.25 (m, 3H), 0.66 ꢀ 1.02 (m, 3H). LCMS (m/z, ES+) = 501 (M+1).
Author Information
Corresponding Author: For JBS: phone, (919) 483ꢀ3918; email, brad.j.shotwell@gsk.com
Abbreviations used
HCV, hepatitis C virus; NTP, nucleotide triphophate; NS5B, nonstructural protein 5B; GT,
genotype; NNI, nonꢀnucleoside inhibitor; HCV RdRp, Hepatitis C Virus RNAꢀDependent
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 43
1
2
3
4
5
6
7
RNA Polymerase; MDCK, MadinꢀDarby Canine Kidney epithelial cells; SFC, supercritical
fluid chromatography; VCD, vibrational circular dichroism spectroscopy
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Associated Content. Supplementary information including references to chemical syntheses,
purities of final compounds, potency and structureꢀactivity relationships for additional analogs,
protocols for the NS5B polymerase and replicon assays, in vivo and in vitro DMPK protocols
and data, NMR spectra and chiral HPLC traces, and a full description of the VCD analysis can
be accessed free of charge via the internet at http://pubs.acs.org.
References
1. Center for Drug Evaluation and Research, Food and Drug Administration, US Department of
Health and Human Services. 2011 Novel New Drugs. (Accessed January 2012).
http://www.fda.gov/Drugs/DevelopmentApprovalProcess/DrugInnovation/ucm285554.htm
2. Bosdet, M. J. D.; Piers, W. E. BꢀN as a CꢀC Substitute in Aromatic Systems. Can. J. Chem.
2009, 87, 8ꢀ29.
3. Nielsen, F. H. Boron in Human and Animal Nutrition. Plant Soil 1997, 193, 199ꢀ208.
4. Hunter, P. Not Boring at All. EMBO Rep. 2009, 10, 125ꢀ128.
5. Ciani, L.; Ristori, S. Boron as a Platform for New Drug Design. Expert Opin. Drug Discov.
2012, 7, 1017ꢀ1027.
ACS Paragon Plus Environment

Page 25 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
6. Zhdankin, V. V.; Persichini, P. J.; Zhang, L.; Fix, S.; Kiprof, P. Synthesis and Structure of
7
8
9
Benzoboroxoles: Novel Organoboron Heterocycles. Tetrahedron Lett. 1999, 40, 6705ꢀ6708.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7. Rock, F. L.; Mao, W.; Yaremchuk, A.; Tukalo, M.; Crépin, T.; Zhou, H.; Zhang, Y.ꢀK.;
Hernandez, V.; Akama, T.; Baker, S. J.; Plattner, J. J.; Shapiro, L.; Martinis, S. A.; Benkovic, S.
J.; Cusack, S.; Alley, M. R. K. An Antifungal Agent Inhibits an AminoacylꢀtRNA Synthetase by
Trapping tRNA in the Editing Site. Science 2007, 316, 1759ꢀ1761.
8. Tomsho, J. W.; Arnab, P.; Hall, D. G.; Benkovic, S. J. Ring Structure and Aromatic
Substituent Effects on the pKa of the Benzoxaborole Pharmacophore. ACS Med. Chem. Lett.
2012, 3, 48ꢀ52.
9. Baker, S. J.; Tomsho, J. W.; Benkovic, S. J. Boronꢀcontaining Inhibitors of Synthetases.
Chem. Soc. Rev. 2011, 40, 4279-4285.
10. Baker, S. J.; Zhang, Y.ꢀK.; Akama, T.; Lau, A.; Zhou, H.; Hernandez, V.; Mao, W.; Alley,
M. R. K.; Sanders, V.; Plattner, J. J. Discovery of a New Boronꢀcontaining Antifungal Agent, 5ꢀ
Fluoroꢀ1,3ꢀdihydroꢀ1ꢀhydroxyꢀ2,1ꢀbenzoxaborole (AN2690), for the Potential Treatment of
Onychomycosis. J. Med. Chem. 2006, 49, 4447ꢀ4450.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 43
1
2
3
4
5
6
7
11. Sutcliffe, J. A. Antibiotics in Development Targeting Protein Synthesis. Ann. N. Y. Acad.
Sci. 2011, 1241, 122ꢀ152.
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
12. Smoum, R.; Rubinstein, A.; Dembitsky, V. M.; Srebnik, M. Boron Containing Compounds
as Protease Inhibitors. Chem. Rev. 2012, 112, 4156ꢀ4220.
13. Li, X.; Zhang, S.; Zhang, Y.ꢀK.; Liu, Y.; Ding, C. Z.; Zhou, Y.; Plattner, J. J.; Baker, S. J.;
Bu, W.; Liu, L.; Kazmierski, W. M.; Duan, M.; Grimes, R. M.; Wright, L. L.; Smith, G. K.;
Jarvest, R. L.; Ji, J.ꢀJ.; Cooper, J. P.; Tallant, M. D.; Crosby, R. M.; Creech, K.; Ni, Z.ꢀJ.; Zou,
W.; Wright, J. Synthesis and SAR of Acyclic HCV NS3 Protease Inhibitors with Novel P4ꢀ
benzoxaborole Moieties. Bioorg. & Med. Chem. Lett. 2011, 21, 2048ꢀ2054.
14. Priestley, E. S.; De Lucca, I.; Ghavimi, B.; EricksonꢀViitanen, S.; Decicco, C. P. P1
Phenethyl Peptide Boronic Acid Inhibitors of HCV NS3 Protease. Bioorg. & Med. Chem. Lett.
2002, 12, 3199ꢀ3202.
15. Kneteman, N. M.; Howe, A. Y. M.; Gao, T.; Lewis, J.; Pevear, D.; Lund, G.; Douglas, D.;
Mercer, D. F.; Tyrrell, D. L. J.; Immermann, F.; Chaudhary, I.; Speth, J.; Villano, S. A.;
O’Connell, J.; Collett, M. HCV796: A Selective Nonstructural Protein 5B Polymerase Inhibitor
with Potent Antiꢀhepatitis C Virus Activity In Vitro, in Mice with Chimeric Human Livers, and
in Humans Infected with Hepatitis C Virus. Hepatology. 2009, 49, 745ꢀ752.
ACS Paragon Plus Environment

Page 27 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
16. O’Farrell, D.; Trowbridge, R.; Rowlands, D.; Jager, J. Substrate Complexes of Hepatitis C
Virus RNA Polymerase (HCꢀJ4): Structural Evidence for Nucleotide Import and Deꢀnovo
Initiation. J. Mol. Biol. 2003, 326, 1025ꢀ1035.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17 Luvino, D.; Baraguey, C.; Smietana, M.; Vasseur, J.. Borononucleotides: Synthesis, and
.
Formation of a New Reversible Boronate Internucleosidic Linkage. Chem. Comm. 2008, 2352ꢀ
2354.
18. Hang, J. Q.; Yang, Y., Harris, S. F.; Leveque, V.; Whittington, H. J.; Rajyaguru, S.; Aoꢀ
leong, G.; McCown, M. F.; Wong, A.; Giannetti, A. M.; Le Pogam, S.; Talamas, F.; Cammack,
N.; Najera, I.; Klumpp, K.. Slow Binding Inhibition and Mechanism of Resistance of Nonꢀ
nucleoside Polymerase Inhibitors of Hepatitis C Virus. J. Biol. Chem. 2009, 284(23), 15517ꢀ
15529.
19. Feldstein, A.; Kleiner, D.; Kravetz, D.; Buck, M.. Severe Hepatocellular Injury With
Apoptosis Induced by a Hepatitis C Polymerase Inhibitor. J. Clin. Gastroenterol. 2009, 43(4),
374ꢀ381.
20. McCown, M. F.; Rajyaguru, S.; Kular, S.; Cammack, N.; Najera, I. GTꢀ1A or GTꢀ1b
SubtypeꢀSpecific Resistance Profiles for Hepatitis C Virus Inhibitors Telaprevir and HCVꢀ796.
Antimicrob. Agents & Chem. 2009, 53(5), 2129ꢀ2132.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 43
1
2
3
4
5
6
7
8
9
21. Beaulieu, P. Nonꢀnucleoside Inhibitors of the HCV NS5B Polymerase: Progress in the
Discovery and Development of Novel Agents for the Treatment of HCV Infections. Curr. Opin.
Invest. Drugs. 2007, 8(8), 614ꢀ634.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
22. Herlihy, K. J.; Graham, J. P.; Kumpf, R.; Patick, A. K.; Duggal, R.; Shi, S. T. Development
of Intergenotypic Chimeric Replicons To Determine the BroadꢀSpectrum Antiviral Activities of
Hepatitis C Virus Polymerase Inhibitors. Antimicrob. Agents & Chem. 2008, 52(10), 3523ꢀ
3531.
23. Kim, A. Y.; Timm, J. Resistance Mechanisms in HCV: From Evolution to Intervention.
Expert Rev. Anti Infect. Ther. 2008, 6(4), 463ꢀ478.
24. Sarrazin, C.; Zeuzem, S. Resistance to Direct Antiviral Agents in Patients with Hepatitis C
Virus Infection. Gastroenterology 2010, 138, 447ꢀ462.
25. Wilfret, D.; Walker, J.; Voitenleitner, C.; BaptisteꢀBrown, S.; Lovern, M.; Kim, J.; Adkison,
K.; Mathis, A.; Moss, L.; Yu, L.; Gan, J.; Spaltenstein, A. A Randomized, Double Blind, Dose
Escalation, Fusion, First Time in Human Study to Assess the Safety, Tolerability,
Pharmacokinetics, and Antiviral Activity of Single Doses of GSK5852 in Chronically Infected
Hepatitis C Subjects. <unpublished results>
ACS Paragon Plus Environment

Page 29 of 43
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26. Shim, J.ꢀH.; Larson, G.; Wu, J.ꢀ Z.; and Hong, Z. Selection of 3’ꢀTemplate Bases and
Initiating Nucleotides by Hepatitis C Virus NS5B RNAꢀDependent RNA Polymerase. J.
Virology 2002, 76(14), 7030ꢀ7039.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
27. Mosley, R. T.; Edwards, T. E.; Murakami, E.; Lam, A. M.; Grice, R. L.; Du, J.; Sofia, M. J.;
Furman, P. A.; Otto, M. J. Structure of Hepatitis C Virus Polymerase in Complex with Primerꢀ
Template RNA. J. Virology 2012, 86(12), 6503ꢀ6511.
28. Reich, S.; Golbik, R. P.; Geissler, R.; Lilie, H.; Behrens, S.ꢀE. Mechanism of Activity and
Inhibition of the Hepatitis C Virus RNAꢀdependent RNA Polymerase. J. Biol. Chem. 2010,
285(18), 13685ꢀ13693.
29. Butcher, S. J.; Grimes, J. M.; Makeyev, E. V.; Bamford, D. H.; and Stuart, D. I. A
Mechanism for Initiating RNAꢀdependent RNA Polymerization. Nature 2001, 410, 235ꢀ240.
30. Gong, P.; and Peersen O. B. Structural Basis for the Active Site Closure by the Poliovirus
RNAꢀdependent RNA Polymerase, Proc. Natl. Acad. Sci. USA 2010, 107, 22505ꢀ22510.
31. Harrus, D.; AhmedꢀElꢀSayed, N.; Simister, S.; Miller, M; Triconnet, M.; Hagedorn,C.;
Mahias, K.; Rey, F.; AstierꢀGin, T.; and Bressanelli, S. Further Insights into the Roles of GTP
and the C Terminus of the HCV Polymerase in Initiation of RNA Synthesis, J. Biol. Chem.,2010,
285, 32906ꢀ32918
ACS Paragon Plus Environment