Arylalkylidene rhodanine with bulky and hydrophobic functional group as selective HCV NS3 protease inhibitor

Article abstract of DOI:10.1016/S0960-894X(00)00610-7

Arylalkylidene rhodanines 2(a-d) inhibit HCV NS3 protease at moderate concentrations. They are better inhibitors of other serine proteases such as chymotrypsin and plasmin. However, the selectivity of arylmethylidene rhodanines (8a, 9a) with bulkier and more hydrophobic functional groups increases by 13- and 25-fold towards HCV NS3 protease respectively.

Full text of DOI:10.1016/S0960-894X(00)00610-7

Bioorganic & Medicinal Chemistry Letters 11 (2001) 91±94
Arylalkylidene Rhodanine with Bulky and Hydrophobic Functional
Group as Selective HCV NS3 Protease Inhibitor
Wan Theng Sing,a Cheng Leng Lee,a Su Ling Yeo,b Siew Pheng Limb
and Mui Mui Sima,Ã
aMedicinal and Combinatorial Chemistry Laboratory, Institute of Molecular and Cell Biology, 30 Medical Drive,
Singapore 117609, Singapore
bCollaborative Antiviral Research Laboratory, Institute of Molecular and Cell Biology, 30 Medical Drive,
Singapore 117609, Singapore
Received 10 July 2000; accepted 12 October 2000
AbstractÐArylalkylidene rhodanines 2(a±d) inhibit HCV NS3 protease at moderate concentrations. They are better inhibitors of
other serine proteases such as chymotrypsin and plasmin. However, the selectivity of arylmethylidene rhodanines (8a, 9a) with
bulkier and more hydrophobic functional groups increases by 13- and 25-fold towards HCV NS3 protease respectively. # 2001
Elsevier Science Ltd. All rights reserved.
Hepatitis C virus (HCV), identi®ed in 1975, is classi®ed
as a new genus of the Flaviviridae family.1 HCV infection
leads to acute and chronic hepatitis, liver cirrhosis, and
in some cases, hepatocellular carcinoma. Treatment
with interferon is only e€ective in 20% of patients. The
high rate of relapse, signi®cant side e€ects, and exorbi-
tant cost associated with the current treatments have,
therefore, prompted extensive research in ®nding other
e€ective means of therapy.
an interesting and convenient core for further develop-
ment of a selective NS3 protease inhibitor. Here, we
report the synthesis of an arylalkylidene rhodanine
library and the modi®cation of the rhodanine side chain
in an attempt to increase the selectivity towards the NS3
protease.
Synthesis
Several potential therapeutic targets have emerged fol-
lowing the sequencing of the HCV genome. The non-
structural protein 3 (NS3), a serine protease with a
chymotrypsin-like fold,2 plays an important role in cleav-
ing the nonstructural HCV proteins that are necessary
for HCV replication. Several attempts to seek HCV
NS3 protease inhibitors have been reported. However,
most of the nonpeptidic small molecule inhibitors
reported are nonspeci®c and inhibit NS3 protease in the
micromolar range.3
Rhodanine 1 was synthesized by reacting methyl thio-
glycolate with isothiocyanate at room temperature (Fig. 1).
Knoevenagel condensation with aldehyde proceeded
very smoothly to completion just by re¯uxing in ethanol
for 6 h without using acid or base as condensing agent.
The condensation of ketone with rhodanine 1 was
achieved using ammonium acetate. Arylalkylidene rho-
danine 2 was obtained as crystals precipitating out of
the reaction mixture.5 In some cases, black and sticky
syrup obtained was washed with a small amount of
CH2Cl2 and dried.
Rhodanine derivatives are known to possess biological
activities such as anticonvulsant, antibacterial, antiviral
and antidiabetic.4 Sudo et al. have also reported rhodanine
derivatives as HCV protease inhibitor.3a It is therefore
Of approximately 2000 compounds that were synthe-
sized and characterized by 1H NMR and MS, eight were
randomly selected and their purities determined by HPLC.
The results are summarized in Table 1. In general,
aldehydes gave better condensation yields and purities
as compared to ketones, and the more hindered ketones
gave poorer yields than the less hindered ketones. The
ÃCorresponding author. Tel.: +65-874-7823; fax: +65-779-1117;
e-mail: mcbsimmm@imcb.nus.edu.sg
0960-894X/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(00)00610-7

92
W. T. Sing et al. / Bioorg. Med. Chem. Lett. 11 (2001) 91±94
Figure 1.
impurities are mainly the unreacted aldehydes or
rhodanines that co-precipitated out with the desired
products. The thermodynamically stable Z-isomer was
obtained for 5-benzylidene rhodanine.4 Either exclu-
sively Z-isomer or a mixture of the two geometrical
stereoisomers with the Z-isomer being the predominant
one (>75%) was obtained for the arylmethylidene
rhodanines.
and Leu135.7 Substrate speci®city studies also demon-
strate the importance of the acidic P6 residue to sub-
strate recognition.8 Based on the ®ndings, it might be
expected that a large inhibitor would be required to
fully occupy the substrate-binding site. Charged and
hydrophobic functional groups were therefore incorpo-
rated into arylalkylidene rhodanine 2 (Fig. 2).
The tert-butyl N-(2-aminoethyl)-carbamate 3a was di-
alkylated with methyl 4-(bromomethyl)-benzoate to give
4a. The Boc protecting group was removed by TFA,
and the free amine was mono-alkylated with 4-chloro-
benzyl chloride to give 5a. The amines 4b and 4c were
synthesized by mono-alkylating 3a and 3b with 4-chloro-
benzyl chloride and methyl 4-(bromomethyl)-benzoate,
respectively. Amines 5a, 4b and 4c were separately
coupled to Fmoc-l-leucine using the standard peptide
coupling condition. The Fmoc protecting group was
then removed to yield 6(a±c). Each step generally gave
more than 60% yield. Finally, the amine 6 was activated
by triphosgene and then coupled to 7 to give 8.9 The low
isolated yields of 8a and 8b were primarily due to the
formation of urea dimer (58% in the case of 8a).
Crystal structure of NS3 protease complexed with a
truncated NS4A cofactor peptide reveals that the S1
selectivity pocket is shallow and hydrophobic, and is
primarily formed by the side chains of Phe154, Ala157
Table 1. The yield6a and HPLC purity6b of compound 2
Compounds
Yield
Purity
88
2a
72
Arylalkylidene rhodanines 9(a±b) with a sulfonamide
functional group were also synthesized (Fig. 3). The less
reactive 3-aminoacetophenone was ®rst added dropwise
to an ice-cooled dilute solution of biphenyl-4,40-disulfonyl
chloride in CH2Cl2, followed by another amine. The
puri®ed ketone was condensed as described above with
rhodanine 1 to give 9.
2b
2c
2d
2e
2f
92
67
89
89
61
62
75
88
91
Results and Discussion
84
Arylalkylidene rhodanines 2 were tested in the NS3
protease assay in groups of 10 at 71 mM each (except 8
and 9 which were tested individually).10 12 Pooled sam-
ples showing an inhibition of greater than 60% were
screened individually at 71 mM again. The 50% inhibi-
tory concentration (IC50) of the active compounds was
determined. To check for their inhibitory selectivity, the
putative inhibitors were examined against two other
serine proteases such as chymotrypsin and plasmin.13
The results are summarized in Table 2.
80
100
73
Most of the arylalkylidene rhodanines 2 tested are not
active except 2(a±d), which are moderate and non-
selective protease inhibitors. Like most reported non-
peptidic small molecule inhibitors of NS3, all but 2c
inhibit chymotrypsin and plasmin better than NS3 pro-
tease.3 Compounds 8a and 9a inhibit both plasmin and
NS3 protease equally well, but they are not active
against chymotrypsin. Compounds 2(a±d), 8a and 9a
2g
2h
82

W. T. Sing et al. / Bioorg. Med. Chem. Lett. 11 (2001) 91±94
93
are also not active against elastase (data not shown).
The IC50's of 8a and 9a against chymotrypsin are 13-
and 25-fold (respectively) higher than that against NS3
protease. The fact that the IC50 of 8a and 9a against
chymotrypsin and plasmin is about 10-fold higher than
that of 2 suggests that selectivity for NS3 protease
improves with a long hydrophobic side chain. This may
also account for the activity but not the selectivity of 2b
against these serine proteases. Furthermore, the long
hydrophobic side chain is likely to be important for the
activity of 8a, 9a and 2c towards NS3 protease as com-
pared to 8(b±e), 9b and 2(a, b, d).
In this paper, we have demonstrated that a long hydro-
phobic side chain is important for the selectivity as well as
the activity of arylalkylidene rhodanine towards NS3
Figure 2.
Figure 3.

94
W. T. Sing et al. / Bioorg. Med. Chem. Lett. 11 (2001) 91±94
Table 2. IC50 (mM)a of arylalkylidene rhodanines against serine pro-
teases
HPLC analysis was performed at 254 nm with a Hypersil ODS
C18 reverse-phase column (2.1Â200 mm). A gradient (30%
MeOH in H2O to 100% MeOH in 10 min, followed by 100%
MeOH for another 10 min) with ¯ow rate 0.3 mL/min was
employed.
7. (a) Love, R. A.; Parge, H. E.; Wickersham, J. A.;
Hostomsky, Z.; Habuka, N.; Moomaw, E. W.; Adachi, T.;
Hostomska, Z. Cell 1996, 87, 331. (b) Kim, J. L.; Morgen-
stern, K. A.; Lin, C.; Fox, T.; Dwyer, M. D.; Landro, J. A.;
Chambers, S. P.; Markland, W.; Lepre, C. A.; O'Malley, E. T.;
Harbeson, S. L.; Rice, C. M.; Murcko, M. A.; Caron, P. R.;
Thomson, J. A. Cell 1996, 87, 343.
Compounds
HCV NS3
Chymotrypsin
Plasmin
8a
9a
2a
2b
2c
2d
16
15
40
71
20
64
210
375
27
38
23
20
20
<1
<1
7
>200
13
aValues are means of two experiments.
8. (a) Zhang, R.; Durkin, J.; Windsor, W. T.; McNemar, C.;
Ramanathan, L.; Le, H. V. J. Virol. 1997, 71, 6208. (b) Ingal-
linella, P.; Altamura, S.; Bianchi, E.; Taliani, M.; Ingenito, R.;
Cortese, R.; Francesco, R. D.; Steinkuhler, C.; Pessi, A. Bio-
chemistry 1998, 37, 8906.
protease. Our results provide more information required
for further development of selective anti-HCV agents.
9. Preparation of arylalkylidene rhodanine 8: Triphosgene
(1/3 mol equiv) was added to a solution of 6(a±e) and diiso-
propylethylamine (2 mol equiv) in anhydrous CH2Cl2 and
stirred for 1 h followed by addition of 7 (1 mol equiv, prepared
by reacting ethyl isothiocyanatoacetate with methyl 3-
hydroxy-a-mercapto-b-methylcinnamate). The reaction mix-
ture was stirred for another 1 h, concentrated and chromato-
graphed on silica gel [hexane:ethyl acetate (75:25)] to yield
8(a±e). The two geometrical stereoisomers for 8a (E/Z=1/4)
and 9a (E/Z=1/5) were inseparable by silica gel chromato-
graphy. HPLC purity for both 8a and 9a are 100%.
10. Cloning and preparation of HCV NS3: The full length
HCV NS3 clone was obtained by PCR from the serum of a
patient infected with genotype 1b of HCV. The NS3 protease
fragment encoding only the protease domain (1-181aa) was
then obtained by PCR from the above clone. The fragment
was then cloned into the pET15B vector. The HIS tagged
fusion protein was transformed and expressed in BL21 (DE3)
bacterial cells. Expression and puri®cation was based on a
modi®ed method.11 The enzyme was quantitated by densit-
ometry with bovine serum albumin as standards. Fusion pro-
tein was frozen in 10 mL aliquots at 70 ꢀC.
11. Gallarini, P.; Brennan, D.; Nardi, C.; Brunetti, M.; Tomel,
L.; Steinkuhler, C.; Francesco, R. D. J. Virol. 1998, 72, 6758.
12. HCV NS3 protease assay: The peptide 4A (NH2-
LTTGSVVIVGRIILSGRPAVVPD-COOH) from amino
acids 18±40 of NS4A at 20 mM ®nal concentration was used to
enhance NS3 protease activity. pETNS3 (0.01 mg) was pre-
incubated with pep4A and chemical compounds diluted in
assay bu€er (50 mM Tris pH 7.5, 150 mM NaCl, 5 mM CaCl2,
10 mM DTT and 15% glycerol) at room temperature for 15±
30 min. The substrate 5A5B was generated by cloning 5A5B
cleavage site (EEASEDVVPCSMSYTWTGACCFGTM) into
the pGEX4T-1 vector. The substrate was expressed and pur-
i®ed in bacterial cells. GST beads containing 3 mg of substrate
5A5B were then added to start the reaction. Assay was per-
formed at 37 ꢀC for 15 min or at times giving less than 40%
conversion. N-Tosyl-l-phenylalanine chloromethyl ketone at
400 mM was used as the positive control. SDS loading bu€er
was added to stop the reaction. Ten microliters of sample were
loaded and run in 15% polyacrylamide gels. Bands were
visualized with Coomassie Blue and the intensity of the bands
was quantitated using the densitometer.
13. Chymotrypsin and plasmin assay: The ¯uorogenic assays
were performed according to the manufacturer's directions from
Molecular Probes (Enchek Protease Molecular probes) that use
casein as the substrate. 0.005U of a-chymotrypsin and 0.002U
plasmin were used per assay. All reactions were performed at
30 ꢀC in black 96-well plates (Nunclon) and stopped at the
linear portion of the enzyme reactions. The enzymes a-chy-
motrypsin was purchased from Sigma (Cat No.: 7762) while
plasmin was purchased from Calbiochem (Cat No.: 527621).
Acknowledgements
The authors would like to thank Drs. Anthony E. Ting
and Agnes Tan for their extremely important sugges-
tions and Miss Yook Wah Choi for excellent technical
support. This work was supported by the National
Science and Technology Board of Singapore.
References and Notes
1. (a) Hagedorn, C. H.; Rice, C. M. Curr. Top. Microbiol.
Immunol.; The Hepatitis C Viruses; Springer: Berlin, New
York, 2000; Vol. 242. (b) Walker, M. A. DDT 1999, 4, 518. (c)
Robert, B. P. Drug News Perspect. 2000, 13, 69.
2. Hahm, B.; Han, D. S.; Back, S. H.; Song, O.-K.; Cho, M.-J.;
Kim, C.-J.; Shimotohno, K.; Jang, S. K. J. Virol. 1995, 69, 2534.
3. (a) Sudo, K.; Matsumoto, Y.; Matsushima, M.; Fujiwara,
M.; Konno, K.; Shimotohno, K.; Shigeta, S.; Yokota, T. Bio-
chem. Biophys. Res. Commun. 1997, 238, 643. (b) Kakiuchi,
N.; Komoda, Y.; Komoda, K.; Takeshita, N.; Okada, S.;
Tani, T.; Shimotohno, K. FEBS Lett. 1998, 421, 217. (c) Sudo,
K.; Matsumoto, Y.; Matsushima, M.; Konno, K.; Shimo-
tohno, K.; Shigeta, S.; Yokota, T. Antiviral Chem. Chemother.
1997, 8, 541. (d) Chu, M.; Mierzwa, R.; Truumees, I.; King, A.;
Patel, M.; Berrie, R.; Hart, A.; Butkiewicz, N.; DasMahapatra,
B.; Chan, T.; Puar, M. S. Tetrahedron Lett. 1996, 37, 7229.
4. (a) Ohishi, Y.; Mukai, T.; Nagahara, M.; Yajima, M.;
Kajikawa, N.; Miyahara, K.; Takano, T. Chem. Pharm. Bull.
1990, 38, 1911. (b) Momose, Y.; Meguro, K.; Ikeda, H.;
Hatanaka, C.; Oi, S.; Sohda, T. Chem. Pharm. Bull. 1991, 39,
1440. It was reported that only the thermodynamically stable
Z-isomer was observed for all arylidene rhodanines. The
methylene proton of the Z-isomer was more down®eld (7.9
ppm) than that of the E-isomer (7.4 ppm) due to the interac-
tion with the carbonyl group at the 4-position.
5. Preparation of arylalkylidene rhodanine 2: A mixture of
isothiocyanate (0.11 mmol), methyl thioglycolate (0.1 mmol)
and Et3N (0.03 mmol) in CH2Cl2 was stirred for 1 h. Excess
isothiocyanate was removed by amino-methylated polystyrene
resin (0.015 mmol, Novabiochem). The solution was ®ltered and
concentrated to give 1. Rhodanine 1 was heated with an alde-
hyde (0.1 mmol) in anhydrous EtOH (200 mL) for 6 h at 80 ꢀC.
Alternatively, rhodanine 1 (0.1 mmol), ketone (0.1 mmol) and
NH4OAc (0.2 mmol) were re¯uxed in toluene (500 mL) for 3
days. The crystalline compound 2 was ®ltered, washed (EtOH
and hexane, 100 mL each) and then dried. The average overall
yield and HPLC purity are more than 70%.
6. (a) Yield was calculated based on the ratio of the mass of
product to methyl thioglycolate expressed as a percentage. (b)