Novel HCV NS5B polymerase inhibitors derived from 4-(1′,1′-dioxo-1′,4′-dihydro-1′λ6-benzo[1′,2′,4′]thiadiazin-3′-yl)-5-hydroxy-2H-pyridazin-3-ones. Part 3: Further optimization of the 2-, 6-, and 7′-substituents and initial pharmacokinetic assessments

Article abstract of DOI:10.1016/j.bmcl.2008.02.072

5-Hydroxy-3(2H)-pyridazinone derivatives were investigated as inhibitors of genotype 1 HCV NS5B polymerase. Lead optimization led to the discovery of compound 3a, which displayed potent inhibitory activities in biochemical and replicon assays [IC₅₀ (1b) < 10 nM; IC₅₀ (1a) = 22 nM; EC₅₀ (1b) = 5 nM], good stability toward human liver microsomes (HLM t_1/2 > 60 min), and high ratios of liver to plasma concentrations 12 h after a single oral administration to rats.

Full text of DOI:10.1016/j.bmcl.2008.02.072

Available online at www.sciencedirect.com
Bioorganic & Medicinal Chemistry Letters 18 (2008) 3446–3455
Novel HCV NS5B polymerase inhibitors derived
from 4-(10,10-dioxo-10,40-dihydro-10k6-benzo[10,20,40]
thiadiazin-30-yl)-5-hydroxy-2H-pyridazin-3-ones. Part 3:
Further optimization of the 2-, 6-, and 70-substituents
and initial pharmacokinetic assessments
Lian-Sheng Li, Yuefen Zhou,* Douglas E. Murphy, Nebojsa Stankovic, Jingjing Zhao,
Peter S. Dragovich, Thomas Bertolini, Zhongxiang Sun, Benjamin Ayida, Chinh V. Tran,
Frank Ruebsam, Stephen E. Webber, Amit M. Shah, Mei Tsan, Richard E. Showalter,
Rupal Patel, Laurie A. LeBrun, Darian M. Bartkowski, Thomas G. Nolan,
Daniel A. Norris, Ruhi Kamran, Jennifer Brooks, Maria V. Sergeeva,
Leo Kirkovsky, Qiang Zhao and Charles R. Kissinger
Anadys Pharmaceuticals, Inc., 3115 Merryfield Row, San Diego, CA 92121, USA
Received 5 February 2008; revised 27 February 2008; accepted 28 February 2008
Available online 5 March 2008
Abstract—5-Hydroxy-3(2H)-pyridazinone derivatives were investigated as inhibitors of genotype 1 HCV NS5B polymerase. Lead
optimization led to the discovery of compound 3a, which displayed potent inhibitory activities in biochemical and replicon assays
[IC50 (1b) < 10 nM; IC50 (1a) = 22 nM; EC50 (1b) = 5 nM], good stability toward human liver microsomes (HLM t1/2 > 60 min), and
high ratios of liver to plasma concentrations 12 h after a single oral administration to rats.
ꢀ 2008 Published by Elsevier Ltd.
Hepatitis C virus (HCV) is a major cause of chronic liver
disease, including cirrhosis and liver cancer. Globally,
an estimated 170 million individuals, 3% of the world’s
population, are chronically infected with HCV and an
estimated 4 million people are newly infected each year.1
Currently, there is no vaccine available to prevent hepa-
titis C nor a direct antiviral agent available for the treat-
ment of chronic hepatitis C.
current HCV therapy result in an urgent medical need
for improved treatments.3
As part of our efforts to develop small molecule, non-
nucleoside inhibitors of the HCV NS5B enzyme, a vi-
rally encoded RNA-dependent RNA polymerase
(RdRp) essential for HCV replication,4 we previously
described the discovery and structure-based optimiza-
tion of a series of 5-hydroxy-3(2H)-pyridazinone deriva-
tives which bind to the NS5B ‘palm’ site (1, Fig. 1).5,6 In
the course of this work, we identified compound 2
The current standard of care is a combination of pegy-
lated interferon (IFN) with ribavirin.2 Inadequate re-
sponse rates, in particular for patients infected with
genotype 1 HCV, along with significant side-effects of
O
O
O
S
O
O
O
OH
N
N
NH2
S
R3
OH
N
S
N
H
R1
N
H
N
Keywords: Pyridazinones; 5-Hydroxy-3(2H)-pyridazinone derivatives;
Hepatitis C virus (HCV); RNA-dependent RNA polymerase (RdRp);
Small molecule; Non-nucleoside NS5B inhibitors; DMPK profiling;
Ratio of liver to plasma concentration.
2
O
N
N
O
NS5B (1b) IC50 = 0.07 uM
NS5B (1a) IC50 = 0.34 uM
Replicon EC50 (1b) = 0.4 uM
R2
1
*
Corresponding author. Tel.: +1 858 273 2501x221; fax: +1 858 273
2697; e-mail: yuefen.zhou@gmail.com
Figure 1. HCV NS5B polymerase inhibitors.
0960-894X/$ - see front matter ꢀ 2008 Published by Elsevier Ltd.
doi:10.1016/j.bmcl.2008.02.072

L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
3447
(Fig. 1) bearing an acetamide R3 moiety as one of the
most potent NS5B inhibitors and antiviral agents.
culture. As we also previously observed,6 appending
substituents to such heterocycles were detrimental to
the associated biological potencies (4d–e) as was the
incorporation of a six-membered aromatic ring at this
location (4f). Introduction of alkyl and alkenyl groups
(both cyclic and acyclic) at the R1 position of the inhib-
itor design (4g–4s) generally afforded compounds with
significantly poorer anti-NS5B and/or antiviral proper-
ties relative to 3a. The most notable exceptions were
compounds 4l and 4m, which exhibited replicon poten-
cies that were only about fourfold weaker than that dis-
played by 3a. While the exact reasons for these
structure–activity relationships are not all known, the
results described above confirmed that unsubstituted
five-membered heterocycles represent ideal R1 substitu-
ents that confer the most potent anti-NS5B and antiviral
properties to this series of NS5B inhibitors.
Here, we report the further optimization and prelimin-
ary pharmacokinetic profiling of this series of pyridazi-
none-containing NS5B inhibitors. Since during our
previous studies we observed that R3 substitution had
the most significant impact on inhibitor potency,5 we
chose to continue making modifications at this position.
As shown in Table 1, when the –OCH2CONH2 moiety
present in compound 2 was replaced by a –NHSO2Me
fragment, the resulting compound (3a) displayed dra-
matically improved NS5B inhibitory potency and antivi-
ral activity.10 This significant increase in potency could
be explained with the aid of the co-crystal structure of
3a bound to the NS5B protein.
As is depicted in Figure 2, the –NHSO2Me R3 moiety of
3a formed H-bonds with NS5B residues Asn291 and
Asp318. Similar H-bonding interactions were also ob-
served in the 2–NS5B co-crystal structure between the
acetamide moiety of compound 2 and NS5B.6 Impor-
tantly, the –NHSO2Me R3 moiety of 3a formed a third
H-bond network with the residue Ser288 via a water
molecule that was absent in the 2–NS5B co-crystal
structure. This additional H-bond network, which com-
plements a similar interaction from the –SO2– moiety in
the benzothiadiazine ring of 3a (Fig. 2), is likely respon-
sible for the potency differences observed between com-
pounds 2 and 3a.12
Table 3 summarizes the results obtained from varying
the R2 substituent of pyridazinones containing optimal
2-thiophene R1 and –NHSO2Me R3 moieties.
As previously observed,6 a R2 alkyl fragment in linear 4–
5 carbon chain length led to both good anti-NS5B and
antiviral activities (5a–f) comparable to that of 3a. How-
ever, more polar R2 groups with a methoxy moiety at
the terminus led to less potent compounds (5g and 5h)
(compared with 5c). Introducing branching at the C-
atom of the R2 fragment adjacent to the pyridazinone
N-2 atom was highly detrimental to both anti-NS5B
and antiviral properties (compounds 5i–k). Consistent
with what we observed previously,6 shorter R2 groups
such as iso-butyl and 2,2-dimethylpropyl (5l–m) resulted
in significant potency loss in 1b replicon assays as com-
pared with 3a. In general, the saturated cyclic alkyl
methyl/ethyl R2 groups (5n–t) led to less potent inhibi-
tors as compared with 3a with the exception of 5p. Inter-
estingly, analogs containing a benzyl moiety (with or
without substitution) or a 2-pyridyl methylene R2 group
(5u–x) all displayed poor antiviral potency in the 1b rep-
licon assay. Such inhibition properties could be im-
proved by the introduction of a 3-thiophene methylene
R2 group (5y). Collectively, the above results demon-
strate that numerous aliphatic moieties tethered to the
pyridazinone N-2 atom by a methylene group could
serve as effective R2 substituents for this series of
NS5B inhibitors.
Although the above co-crystal structure indicated that
the methyl substituent of the –NHSO2Me R3 moiety
present in 3a was oriented away from the NS5B surface,
increasing the size of this fragment was generally detri-
mental to anti-NS5B and antiviral properties (Table 1,
compounds 3b–3g). However, this methyl group could
be replaced with an –NH2 moiety without sacrificing
antiviral potencies (3h). Other, often drastic, losses in
NS5B inhibition activity and antiviral properties were
noted for compounds in which the NH of the R3 sulfon-
amide moiety was alkylated (3i–l) or removed (com-
pounds 3m–q). Such unfavorable property alterations
are likely due to the loss of the H-bond noted in Figure
2 between the R3 sulfonamide NH and the NS5B
Asp318 residue.13 The results described above confirmed
the large impact that R3 variation had on the anti-NS5B
and antiviral properties of this series of inhibitors and
identified the –NHSO2Me and –NHSO2NH2 fragments
as optimal R3 moieties.
In general, the pyridazinone-containing NS5B inhibitors
described in this work displayed somewhat weaker inhi-
bition activities when tested against genotype 1a HCV
polymerase than that against genotype 1b enzyme (Ta-
bles 1–3). The difference in these inhibition levels was
within sixfold for most compounds, although larger
and smaller variations were noted. All of the smaller
deviations were detected during the exploration of the
pyridazinone R1 moieties (Table 2: 4i, 4m, 4p, and 4s)
while the larger differences were mostly observed during
variation of the R2 substituents (Table 3: 5g, 5l, 5n–o,
5r–s, and 5u–x). Although the structural basis for these
potency differences is not known, these results suggest
that the described pyridazinone-containing molecules
have good potential to inhibit HCV polymerases derived
We next explored the systematic variation of the R1 and
R2 substituents attached to the pyridazinone ring of the
inhibitors (1). Due to its relative ease of synthesis, we
employed the –NHSO2Me moiety in lieu of the
–NHSO2NH2 fragment as the preferred R3 substituent.
As shown in Table 2, when the 2-thiophene R1 group
present in 3a was replaced by other unsubstituted five-
membered heterocycles identified during our previous
studies,6 the resulting compounds (4a–c) retained good
NS5B inhibition activity and antiviral properties in cell

3448
L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
Table 1. Optimization of R3 substituents
O
O
S
R3
OH
N
N
S
N
H
N
O
3a - 3q
Compounda
R3
IC50 (1b)b (lM)
IC50 (1a)c (lM)
EC50 (1b)b (lM)
CC50 (lM)
b
b
HLM t1/2 (min)
H
N
3a
<0.01
<0.025
0.005
>33
>33
>1
>60
>60d
>60
NDe
ND
>60
S
O
O
H
N
3b
3c
3d
3e
3f
0.045
0.027
0.03
0.054
0.092
0.16
0.098
0.043
0.25
S
O
O
H
N
S
O
O
H
N
>1
S
O
O
O
H
N
0.038
0.019
0.14
0.15
>1
S
O
H
N
0.026
0.069
>1
NH2
S
O
O
H
N
N
3g
3hf
3i
0.17
<0.01
0.023
0.25
3.2
0.49
0.028
0.46
ND
ND
ND
0.29
0.76
0.007
0.13
0.42
ND
4.8
>33
>1
>60
>60
28
S
S
O
O
H
N
NH2
O
O
N
O
>33
>1
S
O
O
O
3j
11
S
O
N
N
3k
3l
ND
>33
>33
ND
27
S
O
O
N
0.29
0.15
S
O
O
3mg
0.11
22
S
O
O

L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
3449
Table 1 (continued)
b
b
HLM t1/2 (min)
Compounda
R3
IC50 (1b)b (lM)
IC50 (1a)c (lM)
EC50 (1b)b (lM)
CC50 (lM)
S
3n
3o
3p
0.24
ND
1.8
>33
ND
>33
ND
28
O
O
S
S
0.3
ND
1.8
ND
1.5
5.6
13d
11
O
O
NH2
O
0.076
3.2
O
O
N
3q
ND
ND
S
O
a See Ref. 7.
b See Refs. 5 (for assay methods and experimental errors) and 8.
c See Refs. 6 (for assay method and experimental error) and 9.
d Compounds were screened at 5 lM and the other compounds were screened at 1 lM.
e ND, not determined.
f R2 = 3,3-dimethylbutyl (R2 where shown in compound 1).
g Racemic.
compounds tested exhibited moderate to long half-lives
(t1/2 > 30 min). Notable exceptions included molecules
in which the R3 sulfonamide NH was alkylated or re-
moved (Table 1: 3i–q) and the inhibitors bearing alkyl
or alkenyl R1 moieties (Table 2: 4g–s). Importantly,
many examples of pyridazinone-containing compounds
were identified that combined potent NS5B inhibition
properties with good stability toward human liver
microsomes.
Encouraged by the above results, we examined the phar-
macokinetic properties of two very potent and metabol-
ically stable NS5B inhibitors (3a and 5a) in monkeys
after both oral (po) and intravenous (iv) administration
of a 5 mg/kg single dose (Table 4).
Figure 2. R3 region of the co-crystal structure of 3a bound to the
NS5B protein.11 Other protein–ligand interactions were similar to
those previously observed between compound 2 and NS5B.6
As expected based on its good in vitro stability toward
monkey liver microsomes (MLM), compound 3a exhib-
ited a low in vivo clearance value after iv administration
(Table 4). Somewhat surprisingly, 5a displayed similar
in vivo clearance properties in spite of its reduced
MLM stability relative to 3a (Table 4 and Fig. 3). These
results suggested that the in vivo clearance of one or
both compounds may not be primarily affected by
CYP-mediated biotransformation. Both inhibitors also
exhibited low distribution volumes in the iv experiments.
from both 1a and 1b genotypes. However, the data also
caution that the selection of truly optimal inhibitor sub-
stituents, particularly at the pyridazinone R2 position,
should be made using inhibition values obtained from
both genotype 1a and 1b enzymes.
In addition, the antiviral potencies (EC50) of the inhibi-
tors under study measured against genotype 1b HCV
replicon were usually weaker (2- to 20-fold) than the
corresponding genotype 1b NS5B IC50 values (Tables
1–3).14 However, several exceptions to this general trend
were noted (compounds 3m, 4l–m, 4p, 5b–d, 5g–h, and
5p). The differences in the EC50/IC50 ratios are likely
caused by variations in compound cell permeability
and/or protein binding properties. The stability of many
pyridazinone-containing inhibitors toward human liver
microsomes was also assessed (Tables 1–3), and most
Unfortunately, the bioavailabilities of both compounds
(3a and 5a) after oral administration to cynomolgus
monkeys were very low. Since the molecules displayed
low in vivo clearance values and reasonable solubilities
in H2O at pH 7.4, we believe that their undesirable oral
PK properties primarily result from poor gut permeabil-
ity. Accordingly, both inhibitors displayed very low Papp
values in Caco-2 permeability assessments (Table 4).
These poor permeability properties are consistent with
the highly polar nature of both 3a and 5a (polar surface

3450
L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
Table 2. Optimization of R1 substituents
O
O
H
N
S
OH
N
N
S
O
R1
O
N
H
N
O
4a - 4s
b
b
HLM t1/2 (min)
Compounda
R1
IC50 (1b)b (lM)
IC50 (1a)c (lM)
EC50 (1b)b (lM)
CC50 (lM)
3a
<0.01
<0.025
0.005
>33
>33
>1
>60
>60e
31
S
N
4a
4bd
4c
0.01
0.048
0.047
0.05
0.008
0.005
0.007
0.24
S
<0.01
0.01
S
HN
N
>1
30
4d
0.047
0.34
>1
38
4ed
4f
0.14
0.22
0.18
0.19
0.51
0.097
0.4
1.5
>33
>1
>60
38
7
Cl
S
0.031
0.061
0.057
0.071
0.13
0.36
0.53
0.41
0.12
0.21
4g
4h
4i
>1
>1
17
19
25
>33
>33
4j
4k
4l
0.22
NDf
0.21
0.56
>33
>1
9
0.049
0.029
0.13
0.023
0.020
0.81
56
33
10
4m
4n
<0.025
0.3
>1
>33
CF3

L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
3451
Table 2 (continued)
Compounda
R1
IC50 (1b)b (lM)
IC50 (1a)c (lM)
EC50 (1b)b (lM)
CC50 (lM)
b
b
HLM t1/2 (min)
4o
4p
4q
4r
0.029
0.093
0.16
>1
6
0.1
0.075
0.6
0.1
0.8
1.1
1.04
>1
7.4
6
0.18
0.28
0.17
>1
ND
0.26
>33
>33
10
31
F3C
4s
a See Ref. 7.
b See Refs. 5 (for assay methods and experimental errors) and 8.
c See Refs. 6 (for assay method and experimental error) and 9.
d R2 = cyclopropyl ethyl group (R2 where shown in compound 1).
e Compound was screened at 5 lM and the other compounds were screened at 1 lM.
f ND, not determined.
Table 3. Optimization of R2 substituents
O
O
H
N
S
OH
N
S
O
O
S
N
H
N
N
O
5a - 5y
R2
Compounda
R2
IC50 (1b)b (lM)
IC50 (1a)c (lM)
EC50 (1b)b (lM)
CC50 (lM)
b
b
HLM t1/2 (min)
3a
<0.01
<0.025
0.005
>33
>1
>60
>60
>60
5a
5b
0.01
0.059
0.054
0.016
0.008
0.024
>1
5c
5dd
5e
5f
0.019
0.011
<0.01
<0.01
0.076
0.072
0.026
<0.025
0.009
0.005
0.004
0.019
>1
>60
>60
>60
>60
>1
>33
>1
O
5g
5h
0.027
0.027
2.5
0.06
>1
>60e
0.11
0.030
>33
>60
O
(continued on next page)

3452
L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
Table 3 (continued)
b
b
HLM t1/2 (min)
Compounda
R2
IC50 (1b)b (lM)
IC50 (1a)c (lM)
EC50 (1b)b (lM)
CC50 (lM)
5id
5jd
0.59
NDf
15
>33
>33
>60
ND
1
ND
16
5kd
5l
0.45
ND
0.3
11
>33
>1
>1
>1
>1
>60
>60
>60
>60
>60
<0.01
0.014
0.01
0.041
0.016
0.015
0.034
5m
5n
0.077
1.2
5o
<0.01
0.38
CF3
5p
5q
0.013
0.023
0.07
0.1
0.007
0.032
>1
>1
>60
60
CF3
5r
5s
5t
5u
5v
<0.01
0.018
0.19
0.2
0.08
0.040
3.3
>1
>1
>17
>1
>1
>60
>60
>60
>60
>60
0.21
0.69
0.45
0.48
0.011
0.033
0.19
0.75
N
Cl
F
5w
5x
0.012
0.15
0.26
>1
>60
<0.01
0.03
0.12
0.14
>1
>60
F
5y
0.091
0.046
>33
>60e
S
a See Ref. 7.
b See Refs. 5 (for assay methods and experimental errors) and 8.
c See Refs. 6 (for assay method and experimental error) and 9.
d Racemic.
e Compounds were screened at 1 lM and other compounds were screened at 5 lM.
f ND, not determined.

L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
Table 4. In vitro and in vivo (monkeys) DMPK parameters for compounds 3a and 5a
3453
Compounda Solb (lM) Papp (cm/s) MLMd t1/2 Papp BA/AB
n
CL (iv)
(mL/min/kg) (L/kg)
Vss (iv) AUCinf
c
e
C12h(po)/ Fpo (%)
·10ꢀ6 (AB) (min)
ratio
(ng/h/mL)
EC50
3a
5a
55
44
0.016
>60
2375
2 (po) 9.0
2 (iv)
0.096
0.039
124 (po)
0.96
0.66
1f
1
10,543 (iv)
241 (po)
17,640 (iv)
0.060
22
214
8 (po) 5.0
6 (iv)
a Dose: 5 mg/kg in cynomolgus monkeys.
b Solubility in H2O at pH 7.4.
c The Caco-2 assay conditions are described in Ref. 15. Controls: Papp Atenolol (low) = 0.40 · 10ꢀ6 cm/s, Papp Propranolol (high) = 10 · 10ꢀ6 cm/s.
d The monkey liver microsomal (MLM) assay conditions are described in Ref. 16. MLM data were collected with 5 lM of compounds.
e Formulation (both iv and po administrations): 1% DMSO, 9.9% Cremophor EL in 50 mM sodium phosphate buffer, pH 7.4 (compound 3a) or 5%
N,N-dimethylacetamide, 9.5% Cremophor EL in 50 mM sodium phosphate buffer, pH 7.4 (compound 5a).
f Fpo = 2% with 1 mg/kg dose of compound 3a in cynomolgus monkeys; n = 8 (po), 8 (iv).
10000
1000
100
10
100000
10000
1000
100
plasma PO
liver PO
plasma IV
Liver IV
L/P
L/P
PO
IV
431:1
243:1
L/P
L/P
78:1
64:1
L/P
4:1
1 x EC50
10
1xEC50
‡
1
1
‡
0.1
0.1
0
6
12
18
24
1
4
8
12
Time (h)
Time (h)
Figure 3. Plasma concentrations of compound 5a at different times (h)
after a single dose of 5 mg/kg in cynomolgus monkeys (administered
both iv and po).
Figure 4. The plasma and liver concentrations and the ratios of the
liver to plasma concentrations of 3a at different times (h) after a single
administration (iv and po) to Sprague–Dawley rats with 3 mg/kg dose.
The liver concentrations at 1, 4, and 8 h after iv administration were
not determined. The dashed line shows the antiviral 1b EC50 value in
ng/mL for compound 3a.
2
area (PSA) = 203 A for both compounds) which
˚
places them considerably outside the polarity range typ-
ically associated with gut-permeable molecules
2
17
˚
(PSA < 140 A ). In addition, Papp BA/AB ratios for
both compounds 3a and 5a were very high suggesting
that the low oral bioavailabilities may be partially due
to their ability to serve as efficient efflux substrates. In
spite of the low oral bioavailabilities exhibited by 3a
and 5a, we found that the plasma concentrations of both
compounds 12 h after oral administration to monkeys
were close to their corresponding antiviral EC50 values
(Table 4 and Fig. 3).
a major site of HCV replication and this property may
enhance its ability to function as an effective anti-HCV
agent in vivo.
For the synthesis of the pyridazinone derivatives bear-
ing sulfonamide R3 moieties (1a, 10, and 11), three
routes (A, B, and C) were used most often as shown
in Scheme 1. In Route A, ester 620 was condensed with
2-aminobenzensulfonamide 7a21 by heating in pyridine
to generate the corresponding amide intermediates,
which were then cyclized in the presence of DBU in pyr-
idine to yield the desired compound 1a. In Route B,
hydrazone 820 was coupled with acids 9a22 in the pres-
ence of DCC or EDC to form the corresponding amide
intermediates which were further cyclized in the pres-
ence of NaOEt in absolute ethanol to form the desired
compound 1a. In Route C, compound 1b bearing an
iodo group at the 7-position was transformed to the cor-
responding sulfonamide analogs 10 or 11 via CuI cata-
lyzed displacement of the iodo atom.23 Compound 1b
(R3 = I) was prepared using Route A (from compounds
6 and 7b22) or Route B (from compounds 8 and 9b22).
Alternatively, the N-alkylated sulfonamide analog 11
was also accessed via alkylation of the unsubstituted
sulfonamide 10.
As the liver is a major site of HCV replication, we also
conducted a PK study with 3a in Sprague–Dawley rats
in order to measure the liver concentrations of the com-
pound.18 As shown in Figure 4, this study revealed that
the concentration of 3a in rat liver after oral administra-
tion was considerably higher than that in plasma for all
time points examined. The oral liver/plasma (L/P) ratios
declined over time, but consistent with the liver being
the organ initially exposed to orally dosed agents, re-
mained higher than the corresponding L/P ratios ob-
served 12 h after iv administration of 3a. Importantly,
the concentration of 3a in rat liver 12 h after po admin-
istration substantially exceeded its antiviral EC50 value
(C12h/EC50 = 24).19 The high L/P ratio (64:1 12 h post
oral dose) implied the tendency of 3a to accumulate at

3454
Route A
L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
2. (a) Sidwell, R. W.; Huffman, J. H.; Khare, G. P.; Allen, L.
O
O
OH
O
O
S
R3
R1
B.; Witkowski, J. T.; Robins, R. K. Science 1972, 177, 705;
(b) Smith, R. A.; Kirkpatrick, W.. In Ribavirin, a Broad
Spectrum Antiviral Agent; Academic Press: New York,
1980; vol. xiii, pp. 237; (c) De Clercq, E. Adv. Virus Res.
1993, 42, 1.
H2N
H2N
a, b (7a)
(7b)
OEt
+
N
a
N
R2
7
R
R
3 = NHSO2CH3 (7a)
3 = I (7b)
6
Route B
3. Hoofnagle, J. H.; Seeff, L. B. N. Eng. J. Med. 2007, 355,
2444.
4. (a) Kolykhalov, A. A.; Agapov, E. V.; Blight, K. J.;
Mihalik, K.; Feinstone, S. M.; Rice, C. M. Science 1997,
277, 570; (b) Koch, U.; Narjes, F. Curr. Top. Med. Chem.
2007, 7, 1302.
5. Zhou, Y.; Webber, S. E.; Murphy, D. E.; Li, L.-S.;
Dragovich, P. S.; Tran, C. V.; Sun, Z.; Ruebsam, F.; Shah,
A. M.; Tsan, M.; Showalter, R. E.; Patel, R.; Li, B.; Zhao,
Q.; Han, Q.; Hermann, T.; Kissinger, C. R.; LeBrun, L.;
Sergeeva, M. V.; Kirkovsky, L. Bioorg. Med. Chem. Lett.
2008, 18, 1413.
O
O
S
R3
O
O
O
OH
N
S
R3
R1
R1
N
OEt
c, d
N
H
+
N
H
N
N
NH
R2
N
O
HO
O
1
R2
9
3 = NHSO2CH3 (1a)
R3 = NHSO2CH3 (9a)
R
8
R3 = I (1b)
R3 = I (9b)
When R3 = I
e
Route C
When R5 = H
R5
When R5 = H
O
S
O
H
6. Zhou, Y.; Li, L.-S.; Dragovich, P. S.; Murphy, D. E.;
Tran, C. V.; Ruebsam, F.; Webber, S. E.; Shah, A. M.;
Tsan, M.; Averill, A.; Showalter, R. E.; Patel, R.; Han, Q.;
Zhao, Q.; Hermann, T.; Kissinger, C. R.; LeBrun, L.;
Sergeeva, M. V. Bioorg. Med. Chem. Lett. 2008, 18, 1419.
7. All structures of compounds under study were consistent
with 1H NMR and LC–MS analysis (P95% HPLC
purity). They are arbitrarily drawn in one of the possible
tautomer forms.
O
O
R4
O
S
N
R4
O
N
S
OH
N
S
OH
N
O
O
R1
R1
f
N
N
H
H
N
N
N
R2
O
N
R2
O
10
11
Scheme 1. Reagents and conditions: (a) pyridine, 120 ꢁC, 3 h; (b)
pyridine, DBU (2–4 equiv), 120 ꢁC, 16 h, 12–43% over two steps; (c)
DCC, CH2Cl2/DMF, or EDC, NMM, DMF, rt, 12 h; (d) NaOEt,
EtOH, rt or 50–60 ꢁC, 8–12 h, 13–50% over two steps; (e)
NHR5SO2R4, CuI, sarcosine, K3PO4, DMF, 100 ꢁC, 4–16 h, 12–
52%; (f) R5-X (X = Cl, I), NaH, DMF, rt, 35–92%.
8. Potencies of compounds with IC50 values below 10 nM
cannot be distinguished from each other using the 1b
biochemical assay because this is the lower limit imposed
by the enzyme concentration. The EC50 values for these
compounds, however, do provide a comparative measure
of potency.
9. Potencies of compounds with IC50 values below 25 nM
cannot be distinguished from each other using the 1a
biochemical assay because this is the lower limit imposed
by the enzyme concentration.
10. Other thiadiazine or thiazine-containing NS5B inhibitors
bearing sulfonamide R3 fragments have been reported:
(a) Krueger, A. C.; Madigan, D. L.; Green, B. E.;
Hutchinson, D. K.; Jiang, W. W.; Kati, W. M.; Liu, Y.;
Maring, C. J.; Masse, S. V.; McDaniel, K. F.; Middle-
ton, T. R.; Mo, H.; Molla, A.; Montgomery, D. A.; Ng,
T. I.; Kempf, D. J. Bioorg. Med. Chem. Lett. 2007, 17,
2289; (b) Blake, F. J. et al. U.S. Patent US20060040972,
2006.
In summary, we synthesized a number of highly potent
inhibitors of genotype 1 HCV NS5B polymerase, with
single digit nM antiviral potency, derived from pyridaz-
inones with a methylsulfonamide R3 substituent. Our
SAR studies revealed that R1 and R3 substituents signif-
icantly affected the HLM stability of the resulting mole-
cules. A PK study in monkeys with the highly potent
inhibitor 5a (1b EC50 = 16 nM) using a single 5 mg/kg
oral dose showed poor oral bioavailability, but C12h in
plasma was close to its EC50 value. In addition, the ratio
of liver to plasma concentrations in rat of another very
potent inhibitor 3a (1b EC50 = 5 nM) 12 h after a single
oral dose of 3 mg/kg was very high (64-fold), and this
property may be beneficial since liver is a major site of
HCV replication. The further optimization of the PK
properties of this series of inhibitors will be discussed
in a future communication.
11. Crystals of HCV NS5B polymerase (genotype 1b, strain
BK, D21) were grown by the hanging drop method at
room temperature using a well buffer of 20% PEG 4K,
50 mM ammonium sulfate, 100 mM sodium acetate, pH
4.7 with 5 mM DTT. The crystals formed in space group
˚
P212121 with approximate cell dimensions, a = 85 A,
˚
˚
b = 106 A, c = 127 A containing two protein molecules in
the asymmetric unit. Protein–inhibitor complexes were
prepared by soaking these NS5B crystals for 3–24 h in
solutions containing 15–20% DMSO, 20% glycerol, 20%
PEG 4K, 0.1 M Hepes, 10 mM MgCl2 at pH 7.6 and
inhibitors at concentrations of 2–10 mM. Diffraction data
Acknowledgments
The authors thank Drs. Devron Averett and Steve Wor-
land for their support and helpful discussions during the
course of this work.
˚
were collected to a resolution of 2.1 A for compound 3a.
This crystal structure has been deposited in the Protein
Databank (www.rcsb.org) with entry code: 3CDE. Full
details of structure determination are given in the respec-
tive PDB entries.
References and notes
12. The reduced flexibility of the R3 –NHSO2CH3 fragment
relative to the –OCH2C(O)NH2 moiety may also contrib-
ute to the improved potency of 3a.
1. (a) Kim, W. R. Hepatology 2002, 36, 30; (b) Alter, M. J.;
Kruszon-Moran, D.; Nainan, O. V.; McQuillan, G. M.;
Gao, F.; Moyer, L. A.; Kaslow, R. A.; Margolis, H. S. N.
Eng. J. Med. 1999, 341, 556; (c) Alberti, A.; Benvegnu, L.
J. Hepatol. 2003, 38, S104.
13. The inability of the –SO2– moieties present in compounds
3i–q to duplicate the favorable H-bonding interactions
observed in the 3a–NS5B co-crystal structure may also

L.-S. Li et al. / Bioorg. Med. Chem. Lett. 18 (2008) 3446–3455
3455
contribute to the weaker biological potencies observed for
these molecules.
14. Genotype 1b NS5B IC50 values below 10 nM cannot be
accurately compared to the corresponding EC50 values
due to the limitation of the biochemical assay.
17. (a) Palm, K.; Stenberg, P.; Luthman, K.; Artursson, P.
Pharm. Res. 1997, 14, 568; Our PSA values were calcu-
lated by a different method according to the method as
described in (b) Ertl, P.; Rohde, B.; Selzer, P. J. Med.
Chem. 2000, 43, 3714.
15. Assay condition for Caco-2: Express three day Caco-2 cell
cultures (Celsis/IVT, Baltimore, MD) were used on 24-well
polycarbonate membrane inserts. The integrity of the
monolayers was confirmed by measurement of the trans-
epithelial electrical resistance (TEER) and the apparent
permeability of the fluorescent marker compound lucifer
yellow. Apparent permeabilities of a series of test articles
and of selected control compounds were determined in
duplicate at a single concentration of 10 lM in the apical
to basolateral direction at 37 ꢁC. The basolateral com-
partments were sampled for up to 3 h after compound
addition to the apical compartment (collecting a total of
four times points) and were analyzed for test article
content by LC/MS/MS. Samples from the final apical
compartments were also collected and analyzed. The
apparent permeability of each test article was calculated
from these data.
16. Monkey liver microsomal half-life (MLM t1/2) assay
conditions: The final concentrations in the assay were
0.5 mg/mL monkey liver microsomes, 5 lM test article,
3.3 mM glucose-6-phosphate, 1.3 mM NADP+, and 0.4 U/
mL glucose-6-phosphate dehydrogenase in 50 mM potas-
sium phosphate (pH 7.4) and 3.3 mM MgCl2 buffer.
Samples were incubated at 37 ꢁC. At 0, 5, 10, 20, 30, 45,
and 60 min, the reactions were terminated by removing
125 lL-aliquots and quenching in 375 lL of ACN. The
samples were vortexed and centrifuged. The supernatant
was removed and dried under nitrogen. The dried samples
were reconstituted in 2% DMSO in water and analyzed by
LC/MS/MS for the remaining parent molecules. The t1/2
values were determined by fitting the data to a single
exponential decay curve. Comparison of t1/2 values from
multiple replicate experiments on a representative com-
pound gave a 16% standard deviation from the mean
value.
18. The PK parameters of compound 3a in rats: n = 12 (po)
and 3 (iv); CL (iv) = 13.8 (mL/min/kg); Vss (iv) = 0.453 (L/
kg); AUCinf = 27 (po); 3631 (iv) (ng/h/mL); C12h(po)/
EC50 = 0.35; Fpo = 1%.
19. The ratio of the C12h(liver)/EC50 was calculated assuming
the density of the liver was 1.
20. The procedures for the synthesis of compounds 6 and 8 are
described in (a) Li, L.-S.; Zhou, Y.; Zhao, J.; Dragovich,
P. S.; Stankovic, N.; Bertolini, T. M.; Murphy, D. E.; Sun,
Z.; Tran, C. V.; Ayida, B. K.; Ruebsam, F.; Webber, S. E.
Synthesis 2007, 3301; (b) Murphy, D. E.; Dragovich, P. S.;
Ayida, B. K.; Bertolini, T. M.; Li, L.-S.; Ruebsam, F.;
Stankovic, N. S.; Sun, Z.; Zhao, J.; Zhou, Y. Tetrahedron
Lett. 2008, 49, 811.
21. Intermediate 7a was prepared using modified literature
procedures as described in (a) Krueger, A. C.; Madigan,
D. L.; Green, B. E.; Hutchinson, D. K.; Jiang, W. W.;
Kati, W. M.; Liu, Y.; Maring, C. J.; Masse, S. V.;
McDaniel, K. F.; Middleton, T. R.; Mo, H.; Molla, A.;
Montgomery, D. A.; Ng, T. I.; Kempf, D. J. Bioorg. Med.
Chem. Lett. 2007, 17, 2289; (b) Hutchinson, D. K. et al.
U.S. Patent 2005/0107364; (c) Goldfarb, A. R.; Berk, B.
J. Am. Chem. Soc. 1943, 65, 738; (d) Dragovich, P. S.;
Murphy, D. E.; Tran, C. V.; Ruebsam, F. Synth.
Commun., 2008, in press.
22. The procedures for the synthesis of 7b and 9a–b are
described in Ruebsam, F.; Tran, M. T.; Webber, S. E.;
Dragovich, P. S.; Li, L.-S.; Murphy, D. E.; Kucera, D. J.;
Sun, Z.; Tran, C. V. PCT International Patent Application
WO 2007150001 A1, 2007. The synthesis of 3m containing
a cyclic sulfone R3 moiety was similar to that described in
the above patent. The syntheses of compounds 3l–q will be
discussed elsewhere.
23. Deng, W.; Liu, L.; Zhang, C.; Liu, M.; Guo, Q.-X.
Tetrahedron Lett. 2005, 46, 7295.