Inhibitors of HCV NS5B polymerase: Synthesis and structure-activity relationships of N-1-heteroalkyl-4-hydroxyquinolon-3-yl-benzothiadiazines

Article abstract of DOI:10.1016/j.bmcl.2005.01.071

N-1-Alkylamino and N-1-alkyloxy-4-hydroxyquinolon-3-yl benzothiadiazines were synthesized and evaluated as inhibitors of genotype 1 HCV polymerase. The N-1-alkyloxy derivatives were not potent inhibitors, however N-1-alkylamino derivatives displayed compa

Full text of DOI:10.1016/j.bmcl.2005.01.071

Bioorganic & Medicinal Chemistry Letters 15 (2005) 1577–1582
Inhibitors of HCV NS5B polymerase: synthesis and
structure–activity relationships of N-1-heteroalkyl-4-
hydroxyquinolon-3-yl-benzothiadiazines
John K. Pratt,* Pamela Donner, Keith F. McDaniel, Clarence J. Maring, Warren M. Kati,
Hongmei Mo, Tim Middleton, Yaya Liu, Teresa Ng, Qinghua Xie, Rong Zhang,
Debra Montgomery, Akhteruzzaman Molla, Dale J. Kempf and William Kohlbrenner
Infectious Disease Research, Global Pharmaceutical Research and Development, Abbott Laboratories, Abbott Park, IL 60064, USA
Received 9 December 2004; revised 26 January 2005; accepted 28 January 2005
Abstract—N-1-Alkylamino and N-1-alkyloxy-4-hydroxyquinolon-3-yl benzothiadiazines were synthesized and evaluated as inhibi-
tors of genotype 1 HCV polymerase. The N-1-alkyloxy derivatives were not potent inhibitors, however N-1-alkylamino derivatives
displayed comparable potency to carbon analogs. Analogs with aliphatic substituents were significantly more potent than those with
benzylic substituents against genotype 1a polymerase. The most potent inhibitors contained small alkyl or carbocyclic substituents
and exhibited IC50Õs of 50–100 and 200–400 nM against genotype 1b and 1a HCV polymerase, respectively.
Ó 2005 Elsevier Ltd. All rights reserved.
Hepatitis C virus (HCV) infection is a leading cause of
chronic liver disease and the leading cause of death from
liver disease in the United States.1 The Centers for Dis-
ease Control and Prevention estimate that more than 2.7
million individuals in the United States and 170 million
people worldwide have ongoing HCV infection.2 HCV
has six major genotype classes,3 with genotypes 1 and
2 being most prevalent in the United States, Europe,
and Japan.
this combination therapy, demonstrating SVR rates of
<50% even after treatment therapies of 48 weeks in
duration.4,5 In addition, nearly three out of four people
treated with PEG-IFN-a/ribavirin experience systemic
side effects, which can include neutropenia, depression,
irritability, headaches, nausea, anemia, etc.6 These fac-
tors result in a therapeutic need for new HCV therapies,
particularly those directed at genotype 1 HCV.
Because of their demonstrated and vital roles in viral
replication, the NS3 protease/helicase and NS5B RNA
dependent RNA polymerase (RdRp) have been the most
studied viral protein targets for small molecule HCV
therapy. Numerous efforts directed at inhibiting HCV
NS5B RdRp include nucleoside7,8 and non-nucleoside
inhibitors.9–11 A series of N-1-alkyl-4-hydroxy-quino-
lon-3-yl benzothiadiazine non-nucleoside inhibitors 1
(Fig. 1) were identified from a high throughput screen-
ing assay12 and were found to inhibit polymerase at
the initiation step of RNA replication.13 Most import-
antly, in the cell-based HCV replicon assay the com-
pounds of class 1 potently inhibited the production of
viral RNA.
The goal in treating patients infected with HCV is eradi-
cation of the infection. The infection is considered erad-
icated when there is a sustained virologic response
(SVR), defined as the absence of HCV RNA in serum
at the end of treatment and six months post-treatment.
The current standard of care for treating HCV is a com-
bination therapy of pegylated interferon-a (PEG-IFN-a)/
ribavirin. Response rates for HCV patients having geno-
types 2 or 3 on a 24-week treatment regimen of PEG-
IFN-a/ribavirin show a SVR approaching 80%. Patients
infected with genotype 1 HCV do not respond as well to
Keywords: Hepatitis C; HCV; RNA dependent RNA polymerase;
NS5B.
Our preliminary SAR of the N-1-alkyl benzothiadi-
azines demonstrated equivalent polymerase activity for
the quinolone core 1 and 1,8-naphthyridone core 2.
*
Corresponding author. Tel.: +1 847 937 0432; fax: +1 847 938
2756; e-mail: john.k.pratt@abbott.com
0960-894X/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2005.01.071

1578
J. K. Pratt et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1577–1582
stricted or that the branched analogs could not adopt
a low energy conformation suitable for binding. This
paper describes the synthesis and structure–activity
relationships of quinolone and naphthyridione benzo-
thiadiazines in which the a-carbon of the N-1 side chain
is replaced with either an oxygen (compound 3) or a
nitrogen (compounds 4 and 5). We rationalized that
these replacements could allow the attached alkyl
groups access to hydrophobic binding via slightly differ-
ent bond angles/conformations. However, the effect of
changing a lipophilic carbon atom to a more polar nitro-
gen or oxygen atom on the interaction of the inhibitor
with the protein was unknown.
Figure 1.
The preparation of the 1,8-naphthyridone benzothiadi-
azines 2, shown in Scheme 1, is similar to the synthesis
of compound 1.14 Treatment of 8-azaisatoic anhydrides
6 with NaH and (1,1-dioxo-1,4-dihydro-1k6-benzo-
[1,2,4]- thiadiazin-3-yl)-acetic acid ethyl ester15 (com-
pound 7) followed by acetic acid provided 2 directly.
Alternatively, the 8-azaisatoic anhydrides were allowed
to react with diethyl malonate and NaH to give com-
pound 8, which were in turn treated with 2-aminoben-
zenesulfonamide in refluxing toluene to give the
carboxanilides 9. The thiadiazine ring was then formed
by an intramolecular dehydration using 15% aqueous
KOH at reflux to give 2.
Table 1. IC50 values against genotype 1b polymerase: effects of
branching at the a-carbon for analogs of 2
Compd
R
1b IC50, lMa
2c
2d
2e
2f
–CH2Ph
0.083
4.63
(S)–CHMePh
–CH2CH(Et)2
–CH(Et)2
0.058
13.2
2g
2h
0.093
50.0
a IC50 values in all tables are means of at least two independent
determinations, standard deviation 10%. Detailed protocols can be
found in the Supplementary data.
The synthesis of N-1-alkyloxy derivatives 3, shown in
Scheme 2, entailed treating ethyl 2-chloronicotinate 10
with an alkoxyamine in the presence of HunigÕs base
at high temperature to give 11 in poor yields. Attempts
to convert 11 or the corresponding nicotinic acid to an
isatoic anhydride with phosgene or phosgene equiva-
lents were unsuccessful. However, 11 could be acylated
with ethyl 3-chloro-3-oxopropionate followed by cycli-
zation with sodium ethoxide to give the naphthyridone
13. Treatment of 13 with 2-aminobenzenesulfonamide
We also observed that the nature of the hydrophobic
functionality at the N-1 position of the quinolone ring
substantially impacted biochemical potency. In particu-
lar, branching at the a-carbon of the alkyl group dra-
matically decreased inhibitory activity (Table 1)
suggesting either that this position was sterically re-
Scheme 1. Reagents and conditions: (a) compound 7, NaH, THF, reflux, 3 h then AcOH, reflux 1 h, 50–88%; (b) diethylmalonate, NaH, DMA,
120 °C, 2 h, 74–95%; (c) 2-aminobenzenesulfonamide, toluene, reflux, 5 h, 50–60%; (d) 15% aqueous KOH, reflux, 12 h, 95%.

J. K. Pratt et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1577–1582
1579
Scheme 2. Reagents and conditions: (a) alkoxyamine hydrochloride, N,N-diisopropylethylamine, dioxane, 140 °C, 24 h, 30–35%; (b) ethyl 3-chloro-
3-oxopropionate, triethylamine, dichloromethane, 2 h, 42–65%; (c) NaOEt, EtOH, 82–98%; (d) 2-aminobenzenesulfonamide, toluene, reflux, 12 h,
86–89%; (e) 10% aqueous KOH, reflux, 12 h, 90–93%.
followed by heating in aqueous KOH gave the N-1-
alkyloxy benzothiadiazine derivatives 3.
and the cycloalkyl analogs 2g–h, with the cyclohexyl
analog 2h displaying >500-fold lower potency than the
cyclohexane methyl analog 2g.
Preparation of the N-1-alkylamino benzothiadiazines16
with the general structures of 4 and 5 could be accom-
plished using a similar approach, but it required the
development of a protecting group strategy for the
terminal hydrazine nitrogen, as shown in Scheme 3.
Commercially available 2-hydrazinobenzoic acid hydro-
chloride 15 was treated with benzaldehyde to give the
hydrazone 1617 and the carboxylic acid was converted
to the methyl ester 17 with trimethylsilyldiazomethane.
Treatment of 17 with ethyl 3-chloro-3-oxopropionate
in refluxing toluene afforded 18, followed by cyclization
to the 4-hydroxyquinolone 19 with sodium ethoxide in
ethanol. Compound 19 was allowed to react with 2-
aminobenzenesulfonamide in refluxing toluene to give
the carboxamide 20 in moderate yield. Treatment of
20 with refluxing 20% aqueous potassium hydroxide re-
sulted in cyclization of the sulfonamide to the benzothi-
adiazine accompanied by concomitant hydrolysis of the
benzaldehyde hydrazone to give 21. This intermediate
was treated with excess aldehyde or ketone (typically
5–10 mol equiv) and heated in a microwave reactor to
produce the hydrazones of choice. Similar transforma-
tions using a sealed tube in an oil bath proved unsatis-
factory due to long reaction times and low conversion
of the hydrazine to the hydrazone, especially with hin-
dered ketones. The hydrazone derivatives 22 were then
reduced to the final products 4 or 5 with lithium borohy-
dride in THF.
Table 2 highlights the effect of substituting a heteroatom
for the methylene at the N-1 a-position. Compounds 2a
and 2b, with a methylene link at N-1, displayed IC50 val-
ues of 1.49 and 0.069 lM, respectively. The oxygen-
linked homologs 3a and 3b were much less potent, with
IC50Õs of 58 and 20.5 lM, respectively. In contrast, the
nitrogen-linked analogs 4a and 5e inhibited HCV 1b
polymerase at <0.2 lM, indicating that nitrogen
replacement for the methylene is tolerated and prompts
further investigation of this series. The large discrep-
ancy in potency between the oxygen- and nitrogen-
linked analogs is of interest. Potentially, the latter might
preferentially interact with the protein via hydrogen
bond donation. Alternately, an unfavorable interaction
between the lone pairs of the nitrogen at the 8-position
of the napthyridine ring and the oxygen linker could dis-
favor the conformation required for productive binding
of the alkyl group. Oxygen-linked analogs lacking the 8-
heteroatom and nitrogen-linked napthyridines were not
readily synthetically accessible to test this hypothesis.
With the available synthesis for the N-1-alkylamine ser-
ies 4 and 5, the scope of the SAR was determined. The
inhibitory potencies for these derivatives are shown in
Tables 3 and 4. The compounds were tested against
polymerases representing both HCV 1a and 1b geno-
types. Table 3 summarizes benzylic and heteroaromatic
derivatives 4, all of which have a methylene spacer be-
tween the a-nitrogen and the aromatic ring. Compound
4a, the unsubstituted benzyl analog, inhibited genotype
1b polymerase with an IC50 of 0.129 lM, but was signifi-
cantly less potent against 1a (5.1 lM). Substitution at
the ortho position of the phenyl ring with either a bromo
or methyl substituent produced a 4-fold decrease in
Table 1 illustrates the effect of branching at the N-1 a-
carbon on biochemical potency.16 The IC50 of the benzyl
derivative 2c against 1b HCV polymerase was <0.1 lM,
whereas the a-methyl benzyl amine 2d was much less ac-
tive (IC50 4.6 lM). A more pronounced potency differ-
ence was observed between the alkyl derivatives 2e–f

1580
J. K. Pratt et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1577–1582
Scheme 3. Reagents and conditions: (a) benzaldehyde, EtOH/H2O, 98%; (b) (trimethylsilyl)diazomethane, THF/MeOH, 16 h, 90%; (c) ethyl 3-
chloro-3-oxopropionate, toluene, reflux, 4 h, 95%; (d) NaOEt, EtOH, 95%; (e) 2-aminobenzenesulfonamide, toluene, reflux, 5 h, 60%; (f) 20%
aqueous KOH, reflux, 16 h, 95%; (g) aldehyde/ketone, 135 °C microwave, 30 min, 60–80%; (h) LiBH4, cat. MeOH, THF, 0 °C to rt, 1 h, 50–80%.
tion to the binding pocket. Compounds 4h–n represent
heteroaromatic replacements for the phenyl ring. In gen-
eral, none of these derivatives displayed significantly im-
proved activity over the phenyl analog.
Table 2. Comparison of N-1 linking atoms (C vs N vs O)
Compd
R
R1
1b IC50, lM
2a
2b
3a
3b
4a
5e
–CH2CH2Ph
–CH2CH2CH(Me)2
–CH2Ph
—
—
—
—
H
1.49
0.069
58.0
Table 4 shows the SAR of three classes of alkyl and
cycloalkyl derivatives: aliphatic groups linked by a
methylene to the nitrogen (compounds 5a–g), branched
aliphatics (compounds 5h–m) and carbocyclic deriva-
tives (compounds 5n–r). In general, these aliphatic ana-
logs were equipotent to the aromatic analogs against 1b
polymerase, but several compounds 5d, 5f, 5n, and 5p
were significantly more active against the 1a genotype.
This suggests that the 1a enzyme is more discriminating
than 1b polymerase toward binding quinolone benzothia-
diazines in this pocket. The branched aliphatic analogs
were considerably less active against both 1b and 1a,
compared to the methylene-linked alkyls. The carbo-
–CH2CH(Me)2
Ph
–CH(Me)2
20.5
0.129
0.101
H
potency against 1b polymerase but no change in activity
against 1a. The 3-bromo derivative 4c was 6-fold more
active than the 2-bromo analog 4b. In comparison, the
3-methyl analog 4f was less potent against both 1a and
1b polymerases. A dramatic decrease in activity against
1b polymerase was observed with derivatives substituted
at the para position, 4d and 4g, suggesting a size limita-

J. K. Pratt et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1577–1582
1581
Table 3. Biochemical potency of benzylic and heterocyclic derivatives Table 5. 1b replicon potency of selected N-1-alkylamino derivatives
4
Compd
R
R1
1b EC50, lM
15.7
3.48
Compd
R
R1 1b IC50, lM 1a IC50, lM
4a
5a
5e
5f
Ph
Pr
H
H
H
H
4a
4b
4c
4d
4e
4f
Ph
2-BrPh
H
H
H
H
H
H
H
H
H
H
H
H
H
H
0.129
0.469
0.075
4.69
5.09
6.72
1.04
10.36
6.12
8.21
>5
iPr
Cyclopropyl
1.35
1.05
0.103
1.27
3-BrPh
4-BrPh
5n
5p
Cyclobutyl
Cyclohexyl
2-MePh
3-MePh
4-MePh
0.481
0.212
4g
4h
4i
>5
3-Pyridine
2-Thiophene
2-Thiazole
2-Furan
0.221
0.125
0.674
0.305
0.133
1.19
>5
cyclics. The structure–activity relationship data imply
that preferred N-1 substituents extend in a linear fashion
away from the quinolone core (e.g., through a CH2 or
NH link) and bend into a sterically restrictive, highly
lipophilic environment that is less discriminating in
strain 1b than in strain 1a.
2.33
35.8
3.26
1.55
15.0
4.78
4j
4k
4l
3-Furan
2-Thiophene-3-Me
2-Thiophene-5-Cl
4m
4n
1.45
Detailed biological protocols for biochemical IC50 deter-
minations and replicon assay EC50 determinations are
available in the Supplementary data.
Table 4. Biochemical potency of aliphatic derivatives 5
Compd
R
R1
1b IC50, lM
1a IC50, lM
5a
5b
5c
5d
5e
5f
Pr
Bu
H
0.161
0.100
0.134
0.135
0.101
0.145
0.150
0.640
0.120
0.386
1.80
0.951
0.928
0.629
0.411
0.941
0.285
2.528
6.35
Supplementary data
H
i-Bu
Neopentyl
H
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.bmcl.
2005.01.071.
H
iPr
Cyclopropyl
Cyclohexyl
H
H
5g
5h
5i
H
Me
Et
Me
Et
Pr
Pr
i-Pr
Ph
References and notes
2.46
1.89
5j
Et
Pr
1. Kim, W. R. Hepatology 2002, 36, s30.
5k
5l
2.24
6.40
2. Alter, M. J.; Kruszon-Moran, D.; Nainan, O. V.; McQui-
llan, G. M.; Gao, F.; Moyer, L. A.; Kaslow, R. A.;
Margolis, H. S. N. Engl. J. Med. 1999, 341, 556.
3. Simmonds, P. J. Hepatol. 1999, 31(Suppl. 1), 54.
4. Manns, M. P.; McHutchison, J. G.; Gordon, S. C.; Rustgi,
V. K.; Shiffman, M.; Reindollar, R.; Goodman, Z. D.;
Koury, K.; Ling, M.; Albrecht, J. K. Lancet 2001, 358,
958.
Pr
Me
1.13
1.49
5m
5n
5o
5p
5q
5r
16.5
Cyclobutyl
0.108
0.150
0.060
0.217
0.439
0.278
0.747
0.356
1.04
Cyclopentyl
Cyclohexyl
Cycloheptyl
4-Pyranyl
4.63
5. Fried, M. W.; Shiffman, M. L.; Reddy, K. R.; Smith, C.;
Marinos, G.; Goncales, F. L., Jr.; Haussinger, D.; Diago,
M.; Carosi, G.; Dhumeaux, D.; Craxi, A.; Lin, A.;
Hoffman, J.; Yu, J. N. Engl. J. Med. 2002, 347, 975.
6. Walker, M. P.; Yao, N.; Hong, Z. Expert Opin. Invest.
Drugs 2003, 12, 1269.
7. Eldrup, A. B.; Allerson, C. R.; Bennett, C. F.; Bera, S.;
Bhat, B.; Bhat, N.; Bosserman, M.; Brooks, J.; Burlein, C.;
Carroll, S. S.; Cook, P. D.; Getty, K. L.; MacCoss, M.;
McMasters, D. R.; Olsen, D. B.; Prakash, T. P.; Prhavc,
M.; Song, Q.; Tomassini, J. E.; Xia, J. J. Med. Chem.
2004, 47, 2283.
8. Carroll, S. S.; Tomassini, J. E.; Bosserman, M.; Getty, K.;
Stahlhut, M. W.; Eldrup, A. B.; Bhat, B.; Hall, D.;
Simcoe, A. L.; LaFemina, R.; Rutkowski, C. A.; Wolan-
ski, B.; Yang, Z.; Migliaccio, G.; De Francesco, R.; Kuo,
L. C.; MacCoss, M.; Olsen, D. B. J. Biol. Chem. 2003, 278,
11979.
9. Summa, V.; Petrocchi, A.; Pace, P.; Matassa, V. G.; De
Francesco, R.; Altamura, S.; Tomei, L.; Koch, U.;
Neuner, P. J. Med. Chem. 2004, 47, 14.
cyclic derivatives 5n–r displayed good potency against
both 1a and 1b polymerases. However, the introduction
of an oxygen heteroatom into the carbocyclic ring 5r
produced a marked decrease in activity, further illustrat-
ing the hydrophobic nature of this portion of the bind-
ing pocket.
Several compounds were evaluated in a cell-based HCV
replicon system18,19 against genotype 1b (Table 5). Alkyl
derivatives 5a, 5e, 5f, and 5p exhibited EC50 values in the
1–4 lM range, a 10–20-fold decrease from their respec-
tive biochemical potencies. The benzyl derivative 4a lost
over a hundred fold in potency having an EC50 of
15.7 lM. The cyclobutyl analog 5n was equipotent in
the biochemical and replicon assay suggesting this com-
pound has superior membrane permeability than the
other analogs that were tested.
In conclusion, we have determined that a nitrogen-for-
carbon replacement at the N-1 position of 4-hydroxy-
quinolin-3-yl benzothiadiazines yields potent inhibitors
of genotype 1 HCV polymerase. The optimal functional
groups at N-1 are small, branched aliphatics and carbo-
10. Chan, L.; Reddy, T. J.; Proulx, M.; Das, S. K.; Pereira, O.;
Wang, W.; Siddiqui, A.; Yannopoulos, C. G.; Poisson, C.;
Turcotte, N.; Drouin, A.; Alaoui-Ismaili, M. H.; Bethell,
R.; Hamel, M.; LÕHeureux, L.; Bilimoria, D.; Nguyen-Ba,
N. J. Med. Chem. 2003, 46, 1283.

1582
J. K. Pratt et al. / Bioorg. Med. Chem. Lett. 15 (2005) 1577–1582
11. Beaulieu, P. L.; Bos, M.; Bousquet, Y.; DeRoy, P.; Fazal,
G.; Gauthier, J.; Gillard, J.; Goulet, S.; McKercher, G.;
Poupart, M.; Valois, S.; Kukolj, G. Bioorg. Med. Chem.
Lett. 2004, 14, 967.
12. Dhanak, D.; Duffy, K. J.; Johnston, V. K.; Lin-Goerke, J.;
Darcy, M.; Shaw, A. N.; Gu, B.; Silverman, C.; Gates, A.
T.; Nonnemacher, M. R.; Earnshaw, D. L.; Casper, D. J.;
Kaura, A.; Baker, A.; Greenwood, C.; Gutshall, L. L.;
Maley, D.; DelVecchio, A.; Macarron, R.; Hofmann, G.
A.; Alnoah, Z.; Cheng, H.; Chan, G.; Khandekar, S.;
Keenan, R. M.; Sarisky, R. T. J. Biol. Chem. 2002, 277,
38322.
Duffy, K. J.; Kao, C. C.; Sarisky, R. T. J. Biol. Chem.
2003, 278, 16602.
14. Ukrainets, I. V.; Taran, E. A.; Gorokhova, O. V.; Jaradat,
N. A.; Voronina, L. N.; Porokhnyak, I. V. Chem.
Heterocycl. Compd. 2000, 36, 346.
15. Kovalenko, S. N.; Chernykh, V. P.; Shkarlat, A. E.;
Ukrainets, I. V.; Gridasov, V. I.; Rudnev, S. A. Chem.
Heterocycl. Compd. 1998, 34, 791.
16. Pratt, J. K. Patent Application WO 2004/041818 A1, 2004.
17. Fischer, E. B. Chem. Ber. 1902, 35, 2318.
18. Ikeda, M.; Yi, M.; Li, K.; Lemon, S. M. J. Virol. 2002, 76,
2997.
13. Gu, B.; Johnston, V. K.; Gutshall, L. L.; Nguyen, T. T.;
Gontarek, R. R.; Darcy, M. G.; Tedesco, R.; Dhanak, D.;
19. Blight, K. J.; Kolykhalov, A.; Rice, C. M. Science 2000,
290, 1972.