Thiourea inhibitors of herpesviruses. Part 3: Inhibitors of varicella zoster virus

Article abstract of DOI:10.1016/j.bmcl.2004.06.025

The preparation of α-methylbenzyl thioureas and their biological activity against varicella zoster virus is described. Several analogs demonstrated IC₅₀s<0.1μM and their SAR are discussed. These compounds represent a novel class of potent and selective nonnucleoside inhibitors of varicella zoster virus.

Full text of DOI:10.1016/j.bmcl.2004.06.025

Bioorganic & Medicinal Chemistry Letters 14 (2004) 4157–4160
Thiourea inhibitors of herpesviruses. Part 3:
Inhibitors of varicella zoster virus
Martin J. Di Grandi,* Kevin J. Curran, Gregg Feigelson, Amar Prashad, Adma A. Ross,
Robert Visalli, Jeanette Fairhurst, Boris Feld and Jonathan D. Bloom
Wyeth Research, 401 N. Middletown Rd, Pearl River, NY 10965, USA
Received 27 April 2004; revised 9 June 2004; accepted 9 June 2004
Available online 26 June 2004
Abstract—The preparation of a-methylbenzyl thioureas and their biological activity against varicella zoster virus is described.
Several analogs demonstrated IC50s < 0.1 lM and their SAR are discussed. These compounds represent a novel class of potent and
selective nonnucleoside inhibitors of varicella zoster virus.
Ó 2004 Elsevier Ltd. All rights reserved.
1. Introduction
for the cleavage and packaging of viral DNA into
capsids, appears to be unique.4
Varicella zoster virus (VZV), an alpha herpesvirus, is the
causative agent of chickenpox in children and shingles in
adults. Though in certain instances, this viral infection
can lead to encephalitis, a potentially fatal complication
for both immunocompetent and immunocompromised
populations,1 more commonly, adulthood VZV disease
is accompanied by severe pain known as post-herpetic
neuralgia (PHN), which can persist for months after
initial skin lesions have resolved. To date, there are only
four available therapies for the treatment of VZV dis-
ease: acyclovir, valacyclovir, famciclovir, and foscarnet2
and while these agents can shorten lesion duration, they
are at best only marginally effective for the treatment of
PHN. Moreover, as these agents (except foscarnet) re-
quire intracellular activation by viral thymidine kinase
to inhibit viral DNA polymerase, they are likely to suffer
from cross-resistance. As part of a program to identify
inhibitors of herpesviruses,3a;4 we have discovered a
novel class of specific and potent thiourea inhibitors
of VZV replication. Additionally, their mechanism of
action, inhibition of the ORF54 protein which, based on
its homology to the HSV-1 UL6 protein, is responsible
We recently reported the antiviral activity of a series of
bis-aryl thioureas (1) against human cytomegalovirus
(HCMV).3b Refinement of this lead series by converting
the aryl A-ring to an a-methylbenzyl A-ring (2) im-
proved potency against this virus by 1–2 orders of
magnitude.3b Incorporation of this structural modifica-
tion into the previously reported HSV-1 inhibitor 35
ultimately led to ( )-8, a compound that demonstrated
activity against VZV (vide infra). With the goal of
enhancing potency against VZV, a separate program
was initiated (Fig. 1).
2. SAR
The three aromatic rings and the linker between the
A-ring and the thiourea of 8 offered ample opportunity
for lead optimization. The SAR of the benzylic carbon
of compounds 4–12 is shown in Table 1.6 The a-methyl
substituent of 5 is clearly superior to the unsubstituted
methylene (4), the gem-disubstituted derivative (6) or
a-ethyl linker (compare 7 vs 8). Moreover, unlike the
case with our CMV inhibitors, there was a strict stereo-
chemical requirement as only the S-enantiomers were
active (9–12).7
Keywords: Thiourea; Inhibitors; VZV; Varicella zoster virus; Novel
inhibitors.
* Corresponding author. Tel.: +1-845-602-2006; fax: +1-845-602-5561;
e-mail: digranm@wyeth.com
With regard to the aromatic A-ring, several substituents
were well tolerated, though there was a modest prefer-
ence for 4-halo groups in this portion of the molecule
Present address: Department of Biology, Indiana University-Perdue
University at Fort Wayne, IN 46805, USA.
0960-894X/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2004.06.025

4158
M. J. Di Grandi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4157–4160
N
H
S
CF3
N
N
H
S
H3C
S
N
S
F3C
O
B
N
H
N
H
A
O
F3C
N
H
N
H
CF3
1
2
Cl
H
N
H
C
H3CO
N
H3C
S
S
B
A
O
F
O
F
N
H
N
H
N
N
H
H
H3CO
F
3
8
Figure 1. Structures of herpesvirus inhibitors.
furan 16 was encouraging. However, numerous addi-
tional fused bicyclic A-rings, including 3-benzothio-
phenyl (17), 2-naphthyl (18), 3-indolyl (19) as well as a
variety of quinoline and isoquinoline derivatives, had no
appreciable activity against VZV (3.7 to >17 lM).
Table 1. Effect of A-ring linkers on VZV activity
H
N
S
A
O
F
N
H
N
H
X
In the B-ring, 3-substituted phenyl rings were generally
more potent than the corresponding unsubstituted
analogs. For instance, the 3-CF3 and 3-CN derivatives
of compound 8, (compounds 20 and 23, Table 3)
improved potency by 5-10-fold. That electronically dis-
similar groups (cf., 20–24) at this position afforded
improved activity in most instances argues for a con-
formational effect of this substituent. However, there is a
size limit as increasing from methoxy to ethoxy (25) or
N-phenyl benzamide (26) diminishes potency. The
2-substituted analogs (27–29) did not offer any
improvement over compound 8.
4-12
a
Compound
X
A
Chirality
––
IC50
4
H
H
H
F
–CH2–
>26.3
1.3
9.7
5
–C(CH3)H–
–C(CH3)2–
–C(C2H5)H–
–C(CH3)H–
–C(CH3)H–
–C(CH3)H–
–C(CH3)H–
–C(CH3)H–
6
––
7
>17.6
0.60
0.70
>19.1
0.13
49.1
8
F
9
H
H
Br
Br
S
R
S
R
10
11
12
a IC50 values are lM.
About the C-ring, there was little correlation between
this ring or its linker, and potency (Table 4). For in-
stance, aryl (30) or heteroaryl (31, 32) groups linked
through an amide bond were well tolerated as were
(Table 2; 13, 14). Interestingly, the 4-fluorophenyl
A-ring derivative 8 yielded the weakest inhibitor of the
halogen series but when combined with different B-rings
(vide infra), this moiety was most useful. Other aryl
groups were examined and initially the potency of benzo-
Table 3. Racemic B-ring modified VZV inhibitors
X
3
H
N
2
Table 2. A-ring modified VZV inhibitors
H3C
S
O
F
N
N
H
N
H
H
H3C
Ar
S
F
20-29
O
F
N
N
a
H
H
Compound
X
IC50
13-19
8
H
3-CF3
3,5-Br2
0.60
0.06
0.07
0.09
0.11
0.12
0.57
0.26
0.99
1.37
1.54
20
21
22
23
24
25
26
27
28
29
a
Compound
Ar
IC50
8
4-F–Ph
4-Cl–Ph
0.60
0.21
0.42
0.96
0.18
3.56
4.51
9.48
3-OCH3
3-CN
3-CH3
13
14
15
16
17
18
19
4-Br–Ph
4-CN–Ph
2-Benzofuran
3-OC2H5
3-CONHPh
2-CH3
2-Cl
3-Benzothiophene
2-Naphthalene
3-Indole
2-OCH3
a IC50 values are lM.
a IC50 values are lM.

M. J. Di Grandi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4157–4160
4159
Table 4. Racemic C-ring substituted VZV inhibitors
heteroaromatics attached directly to the B-ring (33, 34).
We therefore sought to install functional groups at this
position that would enhance the pharmacokinetic (PK)
profile of these compounds but this effort was unsuc-
cessful.
X
H3C
S
R
N
N
H
H
F
30-34
We also investigated the necessity of the thiourea func-
tionality (Table 5) and found the corresponding ureas to
be essentially equipotent with their thiourea counter-
parts, in contrast to the previously reported CMV
inhibitors.8 However, the ureas suffered from very poor
PK properties and were dropped from further consid-
eration.
a
Compound
X
R
IC50
0.19
30
31
32
33
34
NHCO
NHCO
2-(OCH3)-Ph
3-Isoquinoline 0.18
NHCO
3-Isoxazole
4-Isoxazole
2-Quinoline
4-Quinoline
PhCH2
0.09
0.19
0.19
a IC50 values are lM.
3. Chemistry
The final products were synthesized in a straightforward
manner using the two routes shown (Scheme 1). Treat-
ing a hot acetonitrile solution of isothiocyanate 38a or
39b with either aniline 39a or benzyl amine 38b,
respectively, afforded the desired thioureas. The prepa-
ration of both the 2- and 3-substituted B-rings started
with commercially available 2-substituted-4-nitroani-
lines (40). Acylation of this unreactive aniline under
forcing conditions (5 equiv acid chloride, refluxing THF
overnight), followed by reduction of the nitro group9
gave the desired 3-substituted B-rings (41). Alterna-
tively, if the amino moiety of 40 was first protected as its
trifluoroacetamide, followed by reduction, acylation and
deprotection, the result was the requisite 2-substituted
B-rings (42). The 3- and 4-substituted isoxazole con-
taining B-rings (i.e., 33, 34) were prepared via aryl nitrile
Table 5. Urea versus thiourea comparison
CN
H
N
H3C
Ar
X
O
F
N
N
H
H
23, 35-37
Compounda
Ar
X
IC50
b
35 (S)
36 ( )
37 (S)
23 ( )
2-Benzofuran
2-Benzofuran
4-F–Ph
O
S
0.11
0.11
0.03
0.11
O
S
4-F–Ph
a Compounds are either pure (S)-enantiomer or racemic (Æ) as indi-
cated.
b IC50 values are lM.
Y
H
H3C
Ar
N
Ar'
x
H
+
N
Ar'
a
R
H3C
Ar
S
O
(1)
X
O
N
N
38a R = NCS
38b R = NH2
39a X = NH2
39b X = NCS
H
H
X
H
N
R
O
H2N
b,c
41
X
NH2
(2)
O2N
40
d,e,f,g
H
X
N
R
O
H2N
42
R
HO
N
Ar
h,i
+
(3)
R
N
Cl
Ar
O
43a R = CH2Ph
43b R = Ph-4-NH2
44a Ar = Ph-4-NHBoc
44b Ar = 4-Quinoline
45a R = CH2Ph; Ar = Ph-4-NH2
45b R = Ph-4-NH2; Ar = 4-Quinoline
Scheme 1. Reagents: (a) hot acetonitrile; (b) acyl chloride, THF, reflux; (c) reduction; (d) TFAA; (e) reduction; (f) acyl chloride, TEA; (g) K2CO3,
MeOH; (h) TEA; (i) TFA.

4160
M. J. Di Grandi et al. / Bioorg. Med. Chem. Lett. 14 (2004) 4157–4160
3. (a) Bloom, J. D.; Di Grandi, M. J.; Dushin, R. G.; Curran,
K. J.; Ross, A. A.; Norton, E. B.; Terefenko, E.; Jones,
T. R.; Feld, B.; Lang, S. A. Bioorg. Med. Chem. Lett.
2003, 13, 2929–2932; (b) Bloom, J. D.; Dushin, R. G.;
Curran, K. J.; Donahue, F.; Norton, E. B.; Terefenko, E.;
Jones, T. R.; Ross, A. R.; Feld, B.; Lang, S. A.;
Di Grandi, M. J. Bioorg. Med. Chem. Lett. 2004, 14,
3401–3406.
4. Visalli, R. J.; Fairhurst, J.; Srinivas, S.; Hu, W.; Feld, B.;
Di Grandi, M. J.; Curran, K.; Ross, A.; Bloom, J. D.; Van
Zeijl, M.; Jones, T. R.; O’Connell, J.; Cohen, J. I.
J. Virology 2003, 77, 2349–2358.
5. Van Zeijl, M.; Fairhurst, J.; Jones, T. R.; Vernon, S. K.;
Morin, J.; LaRocque, J.; Feld, B.; O’Hara, B.; Bloom, J.
D.; Johann, S. V. J. Virology 2000, 74, 9054–9061.
6. All compounds were assayed as previously described
against VZV, CMV, and RSV (respiratory syncytial
virus)––see Refs. 3 and 4; to calculate IC50s, the com-
pounds were serially diluted and tested in duplicate. The
Sigmoidal Hill Slope (0–100) model was used for curve
fitting and IC50 determination. Each compound was tested
at least in two independent runs on different days; general
cellular toxicity was measured in the MTS assay (Barltrop,
J. A.; Owen, T. C.; Cory, A. H.; Cory, J. G. Bioorg. Med.
Chem. Lett. 1991, 1, 611–614). All compounds reported
had RSV and MTS IC50 values of >20 lM (data
not shown) and were substantially less active against
CMV.
oxide addition across an appropriately substituted
acetylene (43, Scheme 1, reaction 3)10 followed by
deprotection with TFA as necessary. These anilines were
then acylated with the appropriate isothiocyanate as
described above.
4. Conclusions
In this paper, we have described the synthesis of a series
of thioureas with potent and selective cellular activity
against VZV. The SAR pattern indicated a strong
preference for the S-enantiomers about the A-ring lin-
ker, a bias toward 3-substituted phenyl B-rings, the
presence of some structural tolerance about the C-ring
position and finally that ureas are equipotent with their
thiourea counterparts. Furthermore, these compounds
generated resistance consistent with the inhibition of the
cleavage and packaging of viral DNA. Despite the
unique mechanism of action of these selective VZV
inhibitors, the PK properties of these compounds were
limiting.
Acknowledgements
7. The trend shown in Table 1 held true for most analogs; in
the few instances where little to no preference for either
enantiomer was observed, both isomers were essentially
inactive.
8. In the case of our CMV inhibitors, the corresponding
ureas were about 10–15 times less active; see Ref. 3a.
9. Reduction was accomplished by hydrogenation (H2, 10%
Pd–C), transfer hydrogenation (cyclohexene, 10% Pd–C,
refluxing EtOH) or by dissolving metal reduction (SnCl2,
MeOH or Fe/NH4Cl, EtOH).
We thank Stan Lang, John O’Connell and Dave Shlaes
for their support, Bill Weiss and his group and the
Discovery Analytical Chemistry group in Pearl River for
their efforts during the course of this program.
References and notes
1. Kennedy, P. G. E. Rev. Med. Virol. 2002, 12, 327–334.
2. Snoeck, R.; Andrei, G.; De Clerq, E. Expert. Opin. Invest.
Drugs 2000, 9, 1743–1751.
10. Baskaran, S.; Baskaran, C.; Trivedi, G. K. J. Chem. Res.
Synopsis 1997, 394–395, and references cited therein.