Novel benzthiodiazepinones as antiherpetic agents: SAR improvement of therapeutic index by alterations of the seven-membered ring

Article abstract of DOI:10.1016/S0960-894X(02)00578-4

A series of novel benzthiodiazepinones was studied as antiherpetic agents. Significant improvements in potency and therapeutic index in a viral replication assay were realized over the starting molecule. The role of stereospecific substitution on the diazepine ring and optimal nitrogen substitution were investigated.

Full text of DOI:10.1016/S0960-894X(02)00578-4

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2981–2983
Novel Benzthiodiazepinones as Antiherpetic Agents: SAR
Improvement of Therapeutic Index by Alterations of the
Seven-Membered Ring
Harriet W. Hamilton,a,* Gisele Nishiguchi,a Susan E. Hagen,a John D. Domagala,a
Peter C. Weber,a Stephen Gracheck,a Stefanie L. Boulware,a Eric C. Nordby,a
Hidetsura Cho,b Takeshi Nakamura,b Satoru Ikedab and Wataru Watanabeb
aDepartments of Chemistry and Infectious Diseases, Pfizer Global R&D, Ann Arbor, MI 48118, USA
bDepartments of Chemical and Biological Research, Japan Tobacco Inc., Central Pharmaceutical Research Institute,
Takatsuki, Osaka, Japan
Received 19 March 2002; accepted 28 June 2002
Abstract—A series of novel benzthiodiazepinones was studied as antiherpetic agents. Significant improvements in potency and
therapeutic index in a viral replication assay were realized over the starting molecule. The role of stereospecific substitution on the
diazepine ring and optimal nitrogen substitution were investigated.
# 2002 Elsevier Science Ltd. All rights reserved.
PD 146626, 1 (Table 1), was identified from a high-
throughput screening in a cellular assay as an anti-
herpetic agent. Extensive studies on the mechanism of
action suggest the compound targets a novel cellular
function that is critical for the expression of HSV-1
immediate early genes but not host genes.1 PD 146626
was not optimal for lead development as it possessed a
low therapeutic index, was found to be extensively
metabolized in vivo, and was not active orally in a
murine herpetic model. This communication describes
the structure–activity relationships (SARs) developed to
address the issue of separation of viral replication inhi-
introduction of nitrogen to the 3-position of the benzo-
thiophene was efficiently accomplished by oxidation of
the sulfur prior to nucleophilic attack with the diamine.
Reduction of the sulfoxide (trimethylsilyliodide gener-
ated in situ) allowed base catalyzed cyclization of the
seven-membered ring analogues 3.
Selective alkylation of the two nitrogens was accom-
plished by first derivatizing the amide nitrogen with a
silyl protecting group, followed by anion formation on
the second nitrogen and reaction with the appropriate
electrophile. Deprotection then afforded target com-
pounds 4.
bition from cellular toxicity, leading to
improvement in therapeutic index (TI).
a 3-log
Substitution on the carbons of the seven-membered ring
required synthesis of the mono-protected diamine from
the appropriate amino acid (Scheme 2). Either BOC or
CBZ could be used as the first protecting group, and
then the alternative protecting group selected if the
alternative stereochemistry was desired. These diamines
were then used in the synthetic Scheme with appropriate
deprotection prior to cyclization.
The general route to the synthesis of PD 146626 and related
analogues is given in Scheme 1. Synthesis of oxygen (2) and
sulfur substitutions in the seven-membered ring has been
published previously.2 Initial investigations on the nature
of the heteroatom in the seven-membered ring led us to
focus on substituted nitrogen in this position (Table 1).
New methodology was needed to improve yields and
synthetic flexibility for these analogues. Thus, direct
The viral replication inhibition assay used for these
studies has been described in detail elsewhere.1 Briefly,
HSV-1-lacZ replication inhibition assays were done in
triplicate and EC50’s determined from logarithmic plots
*Corresponding author. Fax: +1-734-622-7879; e-mail: harriet.
hamilton@pfizer.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00578-4

2982
H. W. Hamilton et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2981–2983
Table 2. Benzyl ring substitution
Table 1. Nature of heteroatom in seven-membered ring
Example
X
EC50 (mM)
TC50 (mM)
Tl
60
10
60
35
28
1750
1600
1 (PD146626)
5
2
3a
4a
4b
4c
S
S(O)
O
NH
N-CH3
N-Allyl
N-Benzyl
0.125
2.55
0.21
0.13
0.44
7.6
25.8
13.1
4.5
12.5
1.75
8.0
Compd
R
EC50 (mM)
TC50 (mM)
Tl
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
H
p-Cl
p-CH3
p-OCH3
p-NO2
m-Cl
m-CH3
m-NO2
m-OCH3
o-Cl
0.005
0.040
0.030
0.012
0.050
0.005
0.005
0.004
0.0004
0.008
0.021
8.0
12.0
11.0
10.0
25.0
20.0
12.0
25.0
10
1600
300
360
830
0.001
0.005
500
4000
2400
6250
25,000
1500
1200
of compound concentration versus mean percent inhi-
bition. Toxicity was determined separately in uninfected
cells treated with drug and assayed for their ability to
12.0
25.0
4m
o-CH3
generate XTT formazan. Table
1 illustrates the
improvement seen in TI by changing the heteroatom of
the seven-membered ring from sulfur to substituted
nitrogen. This 2-log improvement was seen with most
small alkyl substitutions (propyl, allyl, i-butyl, and
i-pentyl), and with benzyl.
Finally, we looked at substitution on carbons of the seven-
membered ring (Table 3). This was done to hinder sites of
metabolism seen with some of these analogues (to be dis-
cussed in a separate publication), as well as to investigate
whether there was an influence on TI. In the case of ste-
reoisomers adjacent to the amide nitrogen, the only sig-
nificant improvement was seen with the (S) analogue 3h.
When the stereocenter was adjacent to the other nitrogen
there was a preference for the S-configuration (3j and 3l).
We next investigated the effect of substitution on the
benzyl ring (Table 2). The highest TI values were
observed with meta-substitutions. Compound 4k was
the most potent compound of this series, with a sub-
nanamolar EC50 against viral replication. Since the cellular
toxicity remained above micromolar this compound had
the greatest therapeutic index as well.
Combination of carbon substitution and nitrogen sub-
stitution led to compounds less potent than the original
Scheme 1. Synthesis of target compounds. (a) 2-Aminoethanethiol/DBU; (b) ClCH2CN/KO-t-Bu/DMSO; (c) BH3.Me2S; (d) NaOMe; (e) mono-
protected 1,2-diamino-ethane where protecting group=BOC or CBZ/base (Scheme 2); (f) TMS-Cl/NaI; (g) deprotection/NaOMe; (h) base/Cl-Si-
(Ph)2Me; (i) base/R-X/deprotection.
Scheme 2. Synthesis of diamines. (a) CDl; (b) NH3 or NH4HCO3; (c) cyanuric chloride; (d) H2/Raney nickel; (e) alternative protecting group; (f)
deprotection PG. Intermediates and final diamines checked for chiral integrity by chiral HPLC.

H. W. Hamilton et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2981–2983
2983
Table 3. Carbon- and nitrogen-substitutions on seven-membered ring
nitrogen-only substituted analogue (Table 4). Thus the
compounds 4n through 4q showed no advantage in TI
over the analogous 4k. In each case the (S) isomer was
somewhat better than the corresponding (R) isomer in
terms of TI. The analogues with R1 or R2 substituents
could not be alkylated selectively on the desired nitrogen.
Compd
R1
R2
R3
R4
EC50 (mM) TC50 (mM) Tl
The results of these studies show the dramatic improve-
ment in TI achieved by alteration of the seven-mem-
bered ring of compound 1. The antiviral effect seen with
these compounds is not due to direct interaction with
the virus itself, rather to an interference with cellular
machinery utilized by the virus in its replication. Thus
the importance of a large therapeutic index is critical.
The exquisite selectivity of these compounds for halting
the virus without harming the cell as seen with com-
pound 4k offers an alternative therapeutic approach to
the treatment of herpetic infections. Unfortunately
metabolism issues with selected analogues precluded in
vivo activity, and the SAR to address these problems is
the subject of a forthcoming communication.
3a
H
H
H
H
H
H
H
H
H
CH3
H
Bz
H
H
H
H
H
H
H
H
H
H
H
H
CH3
H
Bz
H
iPr
H
iBu
H
H
H
H
H
CH3
H
Bz
H
iPr
H
iBu
H
H
H
H
H
0.13
0.17
0.02
0.28
0.34
0.06
0.09
0.02
0.39
0.05
0.13
0.01
0.09
4.5
6.0
1.0
1.9
10.0
4.0
3.3
2.0
7.0
2.0
7.0
1.1
1.4
35
35
50
7
29
67
37
100
18
40
54
110
16
3b (S)
3c (R)
3d
3e
3f
3g
3h
3i
3j (S)
3k (R)
3l
H
CH3
H
3m
Bz
Table 4.
Acknowledgements
The authors would like to thank the Analytical Chem-
istry Department of Pfizer Global Research and Devel-
opment Ann Arbor for their help with chiral analysis of
our diamines.
Compd
R3
R4
EC50 (mM)
TC50 (mM)
Tl
References and Notes
4k
4n
4o
4p
4q
H
H
CH3
H
H
CH3
H
Bz
H
0.0004
0.02
0.03
0.09
0.45
10.0
11.0
3.4
6.0
4.0
25,000
550
170
67
1. Boulware, S. L.; Bronstein, J. C.; Nordby, E. C.; Weber,
P. C. Antiviral Res. 2001, 51, 111.
2. Khatana, S. S.; Boschelli, D. H.; Kramer, J. B.; Connor,
D. T.; Barth, H.; Stoss, P. J. Org. Chem. 1996, 61, 6060.
Bz
9