Synthesis, antiproliferative and antiviral activity of imidazo[4,5- d]isothiazole nucleosides as 5:5 fused analogs of nebularine and 6- methylpurine ribonucleoside

Article abstract of DOI:10.1021/jm960605z

A series of imidazo[4,5-d]isothiazole nucleosides related to the antibiotic nebularine and the highly cytotoxic 6-methyl-9-β-D- ribofuranosylpurine have been synthesized from the corresponding heterocycles. The sodium salt glycosylation of the imidazo[4,5

Full text of DOI:10.1021/jm960605z

J . Med. Chem. 1997, 40, 771-784
771
Syn th esis, An tip r olifer a tive a n d An tivir a l Activity of Im id a zo[4,5-d ]isoth ia zole
Nu cleosid es a s 5:5 F u sed An a logs of Nebu la r in e a n d 6-Meth ylp u r in e
Ribon u cleosid e
Eric E. Swayze,† J ohn C. Drach, Linda L. Wotring, and Leroy B. Townsend*
Department of Chemistry, College of Literature, Sciences, and the Arts, Department of Biologic and Materials Sciences,
School of Dentistry, Departments of Pharmaceutical and Medicinal Chemistry, College of Pharmacy, University of Michigan,
Ann Arbor, Michigan 48109
Received August 22, 1996X
A series of imidazo[4,5-d]isothiazole nucleosides related to the antibiotic nebularine and the
highly cytotoxic 6-methyl-9-â-D-ribofuranosylpurine have been synthesized from the corre-
sponding heterocycles. The sodium salt glycosylation of the imidazo[4,5-d]isothiazoles proceeded
smoothly, giving mixtures of N-4 and N-6 regioisomers in generally good yields. The protected
derivatives were deblocked using standard conditions to afford the desired imidazo[4,5-d]-
isothiazole nucleosides, usually as crystalline solids. None of the new nucleosides or heterocycles
displayed selective activity against human cytomegalovirus (HCMV) or herpes simplex virus
type 1 (HSV-1). The N-6 glycosylated imidazo[4,5-d]isothiazoles were completely inactive up
to the highest concentration tested. The N-6 glycosylated imidazo[4,5-d]isothiazoles also were
inactive in antiproliferative and cytotoxicity assays, except for 3-methyl-6-â-D-ribofuranosylimi-
dazo[4,5-d]isothiazole (15a ) and 5-(benzylthio)-6-(2-deoxy-â-D-ribofuranosyl)imidazo[4,5-d]-
isothiazole (5e), which showed moderate inhibition of L1210 cell growth. However, the
heterocycles and several of the N-4 glycosylated derivatives were toxic to HFF, KB and L1210
cells; compounds with 5-benzylthio substituents were the most cytotoxic agents in this series.
In tr od u ction
Purine nucleosides and their analogs have been
extensively investigated as antitumor and antiviral
agents.1-14 While these studies have included com-
pounds with a variety of ring modifications and sub-
stituents, most have contained a fused 6:5 carbon-
nitrogen heterocyclic system analogous to the purine
ring, for example, 9-â-D-ribofuranosylpurine (nebu-
larine) (I, Figure 1)6 and 6-methyl-9-â-D-ribofurano-
F igu r e 1. Substituted 6-(â-D-ribofuranosyl)imidazo[4,5-d]-
sylpurine (II).18 As part of our program to prepare a
isothiazoles (III) as analogs of nebularine (I) and 6-methylpu-
rine ribonucleoside (II).
series of novel purine nucleoside analogs, we were
interested in the replacement of the C2-N3 grouping
of the purine ring with a single sulfur atom. The sulfur
atom has been proposed to be analogous to a -CHdCH-
grouping because of its steric and electronic properties.15
For example thiophene could be viewed as an analog of
benzene, and 5-amino-2-â-D-ribofuranosyl-1,2,4-thiadia-
zol-3-one might be considered an analog of cytidine.16
We have extended this analogy and envisioned imidazo-
[4,5-d]isothiazoles as analogs of purines having the
-(NdCH)- group in the pyrimidine ring of purines
replaced with a bivalent sulfur atom. We have recently
prepared19 a series of imidazo[4,5-d]isothiazoles and
shown that they retain similar spatial characteristics
relative to the parent purines17,18 via a modeling study,
which was later confirmed by an X-ray crystal struc-
ture.19 These considerations suggest that the nucleo-
sides of imidazo[4,5-d]isothiazoles should be good steric
mimics of the corresponding purine nucleosides, but
with very different electronic character, which may lead
to some interesting biological properties.
been prepared by coupling the previously described
imidazo[4,5-d]isothiazoles19,22,23 with the appropriate
sugar derivatives. In addition to the ribosyl nucleosides,
the 2′-deoxyribosyl and arabinosyl derivatives were
prepared. All compounds were tested for activity against
two herpes viruses and for effects on the proliferation
of one normal and two cancer cell lines.
Ch em istr y
The sodium salt glycosylation method20,21,24 was used
to prepare the N-4 and N-6 2′-deoxyribosyl derivatives
of the variously substituted imidazo[4,5-d]isothiazoles,
analogously to the imidazo[4,5-d]isothiazoles from
3-methyl-5-(methylthio)imidazo[4,5-d]isothiazole (1b)19
(Scheme 1). Reaction of the in situ generated sodium
salt of compounds 1a -e19 with 2-deoxy-3,5-di-O-p-
toluoyl-R-D-erythro-pentofuranosyl chloride (2)25 at room
temperature gave the corresponding N-6 and N-4 â-nu-
cleosides 3 and 4 in 63-88% overall yields, with no
evidence of the formation of the corresponding R-ano-
A series of imidazo[4,5-d]isothiazole nucleosides (III),
including the analog 15a of 6-methylpurine (II), has
1
mers seen in the H NMR spectra. The regiospecificity
of the reaction appeared to be governed by steric
principles, as heterocycles bearing both 3- and 5-sub-
stituents gave primarily N-6 nucleosides (for example,
3b:4b ) 7:2) while the less hindered derivatives afforded
* Author to whom correspondence should be addressed.
† Present address: ISIS Pharmacueticals, 2292 Faraday, Carlsbad,
CA 92008.
X Abstract published in Advance ACS Abstracts, J anuary 1, 1997.
S0022-2623(96)00605-X CCC: $14.00 © 1997 American Chemical Society

772 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
Sch em e 1. Synthesis of
(2′-Deoxyribofuranosyl)imidazo[4,5-d]isothiazoles
assignment of the site of glycosylation before deprotec-
tion by observing the TLC mobility (ethyl acetate/
hexane) of each isomer formed, since in all cases the
N-4 isomer had a larger Rf value than the corresponding
N-6 isomer.
The carbon-13 NMR spectra26 of the N-6 and N-4
derivatives also showed a distinctly different pattern in
the chemical shifts of the four imidazo[4,5-d]isothiazole
carbon atoms, which compared well with that of the
corresponding N-methyl derivatives. It can be seen
from Table 1 that the bridgehead carbons of N-4
derivatives are shifted dramatically with respect to the
corresponding N-6 isomers. The C3a signals of the N-4
compounds show a large upfield shift of 12-14 ppm,
while the C6a resonances are shifted downfield 15-18
ppm, relative to the corresponding N-6 derivatives. The
other imidazo[4,5-d]isothiazole 13C resonances exhibit
smaller shifts. The N-4 isomers also show a 3-5 ppm
upfield shift of the C3 signal and 1-3 ppm downfield
shift of the C5 resonance, relative to the N-6 isomers.
This pattern is consistent with that shown by the
4-methyl compounds, which show an upfield shift of 10
ppm for C3a and a downfield shift of 10 ppm for C6a,
with respect to the 6-methyl analog. Furthermore, this
pattern is also consistent with that shown by the related
purine ring system. For example, 7-methylpurine (analo-
gous to a 4-methylimidazo[4,5-d]isothiazole) shows an
upfield shift of 8 ppm and a downfield shift of 9 ppm
for the corresponding bridgehead carbon signals, rela-
tive to 9-methylpurine (analogous to a 6-methylimidazo-
[4,5-d]isothiazole).26 This distinctive pattern of carbon-
13 chemical shifts shown by N-substituted imidazo-
[4,5-d]isothiazoles provides a simple means for the as-
signment of the regiochemistry of substituted imidazo-
[4,5-d]isothiazole nucleosides.
The individual assignments of the imidazo[4,5-d]-
isothiazole carbon atoms for N-6 substituted derivatives
were made by observing the fully coupled and partially
decoupled 13C NMR spectra. The 3- and 5-unsubsti-
tuted derivatives have nonquaternary ring carbons,
which could readily be assigned due to the large one-
bond coupling (J CH). The bridgehead and other qua-
ternary 13C resonances were readily assigned by ob-
serving the partially decoupled spectra. For example,
irradiating at the frequency of the resonance due to the
3-methyl peak of 5b in the 1H NMR spectrum decouples
the C3 and C3a resonances, which appear as sharp
singlets at δ 152.1 and 147.3 ppm, respectively. The
C3 carbon is assigned to the most downfield resonance
on the basis of its chemical shift in relation to the
3-unsubstituted derivatives. The introduction of a
methyl group causes the expected large downfield shift
(∼9 ppm) of the C3 carbon, while the downfield shift of
the C3a is only 1-2 ppm. Furthermore, the absolute
intensity of the C3 signal is larger than that of the C3a
resonance, due to the NOE enhancement provided by
decoupling the adjacent methyl protons. The C6a shows
a three-bond coupling to the anomeric proton (3J CH ) 5
Hz) and appears as a doublet at δ 147.1 ppm. The C5
appears as a multiplet at δ 151.6 ppm, due to the
presence of two three-bond couplings. Assignment of
the other N-6 derivatives was accomplished in a similar
manner, or by analogy to these assignments.
the N-4 and N-6 products in an approximately 1:1 ratio
(for example 3a :4a ) 3c:4c ) 1:1). The N-4 and N-6
regioisomers were readily separated by flash column
chromatography to give the protected nucleosides, and
the blocking groups were then removed via methanoly-
sis to afford the desired nucleosides 5a -e and 6a -c,e
as crystalline solids.
The site of glycosylation was determined by a com-
parison of the UV spectra of compounds 5 and 6 with
the spectra of the corresponding alkylated compounds
prepared previously.19 The N-6 glycosylated 5-unsub-
stituted compound 5a had a UV spectrum which com-
pared well with that of the corresponding 6-methyl
derivative prepared by a direct synthesis, showing a
strong absorption with a λmax at 208 nm and medium
absorption having a λmax of 248 nm This was readily
discernible from the pattern shown by 6a , which in-
cluded a strong absorption at a λmax of 246 nm and a
distinctive shoulder in the 250 nm region. Assignment
of the 5-substituted derivatives 5b-e and 6b,c,e was
accomplished in a similar manner. The UV spectra of
the previously prepared 3,4-dimethyl-5-(methylthio)-
imidazo[4,5-d]isothiazole displays two distinct peaks at
215 and 276 nm and a shoulder at 258 nm. In contrast,
3,6-dimethyl-5-(methylthio)imidazo[4,5-d]isothiazole
shows an absorption with a λmax at 219 and 270 nm,
with no shoulder evident. These patterns were nearly
identical to those shown by the 5-substituted compounds
(series b-e). It was also possible to make a tentative
The carbon-13 resonances for the N-4 isomers were
assigned on the basis of their chemical shifts and also

Imidazo[4,5-d]isothiazole Nucleosides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 773
Sch em e 2. Synthesis of
Arabinofuranosylimidazo[4,5-d]isothiazoles
by observation of the fully coupled and partially de-
coupled 13C spectra. The C6a carbon of N-4 imidazo-
[4,5-d]isothiazoles is uncoupled and appears as a sharp
singlet as the most downfield resonance at δ 163 ppm.
Irradiation at the frequency of the 3-methyl signal in
the 1H NMR spectrum of 6b decouples the two- and
three-bond couplings to the C3 and C3a carbons,
respectively. Therefore, C3 appears as a singlet with
high relative intensity (due to NOE enhancement from
decoupling), while C3a appears as a doublet (3J CH ) 5
Hz) at the most upfield resonance of 135-136 ppm, due
to a three-bond coupling from the anomeric proton. The
signal corresponding to C5 is complex, due to multiple
couplings, and appears significantly downfield of the
C3a signal. The C3 resonance shows either a large one-
bond coupling for the 3-unsubstituted derivatives or a
small (3J CH ) 7 Hz for 6b) two-bond coupling for the
3-methyl derivatives in the fully coupled 13C spectrum.
The large differences in the chemical shifts of the
carbon-13 resonances for the N-4 isomers readily allows
the assignment of a new derivative via the application
of standard carbon-13 substituent effect chemical shift
rules onto an analogous N-4 imidazo[4,5-d]isothiazole.
The anomeric configuration was assigned as â by an
inspection of the 1H NMR spectra, which showed the
anomeric proton as a pseudotriplet for the 5-unsubsti-
tuted compounds 5a and 6a . When a substituent was
present at the 5-position (5b-e and 6b,c,e), the pseudo-
triplet was not observed, and the resonance assigned
to the anomeric proton appeared as a doublet of doublets
with a peak width of 15-17 Hz. These observations are
consistent with the spectra observed for similar 8-un-
substituted and 8-substituted 2′-deoxyribofuranosylpu-
rine nucleosides, and led to the assignment of these
compounds as â-nucleosides.27 Furthermore, the start-
ing halogenose is known to have the R-configuration in
the solid state28 and has previously led to the exclusive
formation of â-anomers utilizing the sodium salt glyco-
sylation procedure at room temperature,29,30 presumably
due to a direct Walden inversion at the C1 carbon.
trans), since this pattern of chemical shifts is indicative
of â-anomers.27 That the major product is the â-anomer
is expected, since the starting halogenose is predomi-
nantly the R-anomer, and the gycosylation is presumed
to occur via an SN2 mechanism.
The blocking groups were removed via treatment with
boron trichloride in dichloromethane (-78 °C to 0 °C)
to afford the nucleosides 10a -e, 11c, and 11e. The
workup of this reaction proved critical, as these nucleo-
sides were quite sensitive to a cleavage of the sugar
moiety by acidic conditions. If the quenched acidic
reaction mixture was not completely neutralized at 0
°C, the desired nucleoside was converted to the corre-
sponding heterocycle. The nucleoside derivative 10a
was obtained in crystalline form and in a nearly
quantitative yield. However the 5-substituted com-
pounds proved more difficult to crystallize and were
subject to decomposition. The yields of pure 10b-e,
11c, and 11e were therefore modest, despite the appar-
ent clean conversion of the protected nucleosides into a
single product (by TLC and 1H NMR of the crude
product). In several cases, the nucleosides could not be
induced to crystallize and were isolated as hard foams
by lyophilization. Despite the presence of traces of the
corresponding R-anomer before deprotection, only a
The arabinofuranosides 10 and 11 (Scheme 2) were
also prepared using the sodium salt glycosylation method,
employing 2,3,5-tri-O-benzyl-R-D-arabinofuranosyl chlo-
ride (7)31 as the requisite halogenose. As found for the
corresponding 2-deoxyribofuranosyl synthon, the re-
giospecificity of the reaction appeared to be governed
by steric parameters. However, the halogenose 7 is
more hindered than the corresponding 2-deoxyhalo-
genose 2, and the reaction therefore proceeded with
more selectivity for the N-6 isomers. The N-6 isomers
8a -e were isolated in 46-68% yield after chromatog-
raphy to remove traces of the minor N-4 isomer (4-
30%), as well as unreacted starting material (11-17%).
Only when no 3-substituent was present was a signifi-
cant quantity of the N-4 isomers 9c and 9e isolated (24%
and 30%, respectively). The increased steric require-
ments of the arabinofuranosyl synthon were also evident
by slower reaction rates. Although homogeneous by
TLC, a small amount of the R-anomer was evident for
products 8 (3-5%), 9c (20%), and 9e (7%) upon exami-
nation of the 1H NMR spectrum. In all cases, the
anomeric proton of the major component appeared
downfield of that of the minor component. Thus, the
configuration of the major product was assigned as â
(C1′-C2′ cis) and the minor product as R (C1′-C2′
1
single anomer could be detected by H NMR after the
products had been isolated by chromatography or
crystallization.
The site of glycosylation was determined by examina-
tion of UV and 13C NMR spectra in a similar manner
as for the 2′-deoxy derivatives 5 and 6. The assignments
for the imidazo[4,5-d]isothiazole carbon-13 resonances

774 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
Ta ble 1. Heteroaromatic Carbon-13 Resonances of Imidazo[4,5-d]isothiazole Nucleosides
substituent
chemical shift (ppm)a
compd no.
R3
R5
sugard
Deoxy
C3
C3a
C6a
C5
5a
6a
5b
6b
5c
6c
5d
5e
6e
CH3
CH3
CH3
CH3
H
H
CH3
H
H
H
SCH3
SCH3
SCH3
N-6
N-4
N-6
N-4
N-6
N-4
N-6
N-6
N-4
153.3
148.4
152.1
147.3b
143.7b
140.0b
152.3
143.5b
140.2b
148.8
134.9
147.3
135.2b
148.9
136.6
147.3
148.6
136.4
146.0
163.6
147.1b
163.1b
147.9
163.2
146.9
147.4
163.0
145.0b
144.5b
151.6b
154.2b
152.4
154.9
149.4
150.0
153.0
SCH3
SCH2C6H5
SCH2C6H5
SCH2C6H5
H
Arabino
10a
10b
10c
11c
10d
10e
11e
CH3
CH3
H
H
CH3
H
H
SCH3
SCH3
SCH3
SCH2C6H5
SCH2C6H5
SCH2C6H5
N-6
N-6
N-6
N-4
N-6
N-6
N-4
152.0b
151.1
143.0
141.6
151.3
143.2
141.7
148.2b
147.0
148.6
138.8
147.0
148.6
138.6
148.4b
150.1b
150.9
161.8
149.9c
150.7c
161.7
145.3b
151.4b
152.5
153.5
149.5c
150.6c
151.8
H
Ribo
15a
16a
15b
15c
16c
15d
CH3
CH3
CH3
H
H
CH3
H
H
N-6
N-4
N-6
N-6
N-4
N-6
153.2
148.2
152.0
143.7
140.2
152.2
148.9
134.7
147.5
149.2
136.7
147.5
146.0
163.5
147.2
148.0
163.3
147.1
145.3
144.5
152.8
153.7
155.9
150.8
SCH3
SCH3
SCH3
SCH2C6H5
a
b
Assignments by analogy and from chemical shifts unless otherwise specified. Unequivocally assigned from fully or partially coupled
spectra. c Assignment uncertain. Abbreviation used: deoxy ) 2′-deoxy-â-D-erythro-pentofuranosyl; arabino ) â-D-arabinofuranosyl; ribo
) â-D-ribofuranosyl.
d
(Table 1) were determined by analogy of the chemical
shifts to the corresponding 2′-deoxy derivatives and by
selective decoupling experiments. For example, decou-
pling the 3-methyl proton signal of 10a showed C3 as a
singlet, C3a as a doublet (3J CH ) 12 Hz), C6a as a
pseudotriplet (3J CH ) 5 Hz), and C5 as a doublet of
3-methylimidazo[4,5-d]isothiazole (1a ) with 12 under
modified32 Vorbruggen33 glycosylation conditions gave
the nucleosides 13a and 14a as single anomers in 37%
and 23% yields, respectively, after a separation of
regioisomers by column chromatography. It was prefer-
able to use (trimethylsilyl)trifluoromethanesulfonate
(TMSOTf) as the catalyst, since use of the commonly
employed tin tetrachloride (SnCl4) gave low yields and
required longer reaction times. This may be due to the
ability of SnCl4 to complex with the isothiazole sulfur
atom, leading to a reduction in reactivity. Despite the
high temperature (80 °C) necessary for completion of
the reaction, some regioselectivity is evident, with
formation of the N-6 isomer being preferred. Depro-
tection with methanolic ammonia provided the 6-meth-
yl-9-â-D-ribofuranosylpurine analogs 15a and 16a as
crystalline solids in nearly quantitative yields.
3
doublets (J CH ) 166 Hz, J CH ) 3 Hz). For a few of the
compounds, the chemical shifts of a pair of quaternary
carbons were extremely close together, and the assign-
ments could not be made definitively based on observed
trends. Selective decoupling experiments were not
performed, due to a lack of sufficient material.
The patterns of chemical shifts and couplings seen
for the imidazo[4,5-d]isothiazole carbon atoms in the
â-D-arabinofuranosyl series generally follow the same
trends seen for the corresponding 2′-deoxy nucleosides.
This similarity readily allows the assignment of regio-
isomers and is expected since the carbohydrate portion
of the molecule is relatively distant from the imidazo-
[4,5-d]isothiazole carbon atoms. A characteristic of
these â-D-arabinofuranosyl nucleosides is a small (∼3
ppm) downfield shift of the C6a signal. This is presum-
ably due to the anisotropic effect of the 2′-hydroxyl
moiety, which could be relatively close to the C6a
carbon, since it is in the cis orientation with respect to
the heterocycle.
Unfortunately, this reaction was not as successful
with the 5-substituted imidazo[4,5-d]isothiazoles 1b-
d , giving variable yields of nucleosides, along with deep
red decomposition products. Use of SnCl4 as the
catalyst, and/or change of solvent to 1,2-dichloroethane,
did not lead to the formation of the deep red decomposi-
tion products but resulted in the formation of little or
no nucleoside material, even upon extended reaction
times and elevated temperatures. The reaction of
compounds 1b and 1d with 12 at 50 °C in the presence
of TMSOTf under similar conditions gave the corre-
sponding N-6 nucleosides 13b and 13d as the only
isolable products in 52% and 21% yields, respectively.
The formation of a single regioisomer is presumably due
We next prepared several â-D-ribofuranosylimidazo-
[4,5-d]isothiazole derivatives via condensation of the
silylated heterocycles with 1-O-acetyl-2,3,5-tri-O-ben-
zoyl-â-D-ribofuranose (12, Scheme 3) in the presence of
a Lewis acid catalyst. In this manner, reaction of

Imidazo[4,5-d]isothiazole Nucleosides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 775
Sch em e 3. Synthesis of
Sch em e 4. Synthesis of an N-4
Ribofuranosylimidazo[4,5-d]isothiazoles
Ribofuranosylimidazo[4,5-d]isothiazole
binosyl derivatives. The carbon-13 assignments for the
imidazo[4,5-d]isothiazole carbon resonances are shown
in Table 1. As expected, the general patterns of chemi-
cal shifts and couplings seen for the imidazo[4,5-d]-
isothiazole carbon atoms in the â-D-ribofuranosyl series
are similar to those of the corresponding 2′-deoxy
nucleosides, which readily allows the unequivocal as-
signment of the N-4 and N-6 regioisomers.
Compounds 15a -d , 16a , and 16c were determined
to have the â-configuration by a conversion of a small
sample of each compound to the corresponding 2′,3′-O-
isopropylidine. This was readily accomplished by treat-
ment of each nucleoside derivative (∼5 mg) with per-
chloric acid (1 drop) in acetone (0.5 mL) for 1 h at room
temperature. After the reaction was quenched with
dilute aqueous sodium bicarbonate, followed by extrac-
tion with chloroform, a nearly quantitative yield of the
corresponding 2′,3′-O-isopropylidine was obtained. Ex-
amination of the 1H NMR spectra (CDCl3) of these crude
products showed a separation of 0.22, 0.26, 0.26, 0.25,
0.24, and 0.27 ppm between the isopropylidine methyl
peaks of the 2′,3′-O-isopropylidine nucleosides derived
from 15a , 15b, 15c, 15d , 16a , and 16c, respectively.
This separation (>0.15 ppm) is characteristic of â-D-
ribofuranosyl nucleosides.35 The exclusive formation of
â-anomers is expected due to the role of the neighboring
acyloxy group in the reaction, which promotes the
formation of trans nucleosides.36
[
]
to the steric demands of the 5-alkylthio moiety, which
prevents a reaction at the sterically more crowded N-4
position. Deprotection with methanolic ammonia gave
good yields of the 6-â-D-ribofuranosylimidazo[4,5-d]-
isothiazoles 15b and 15d . Reaction of the imidazo[4,5-
d]isothiazole 1c under similar conditions gave a complex
mixture, from which only one nucleoside product was
formed. However, this nucleoside could not be sepa-
rated from several highly colored decomposition prod-
ucts, even after careful column chromatography. The
impure, darkly colored crude product was therefore
deblocked with methanolic ammonia, and the crude
nucleoside was chromatographed, which provided 16c
as a syrup. This material was then crystallized to
provide the N-4 isomer 16c in 11% overall yield from
the corresponding heterocycle.
Since the N-6 isomer (13c) was not obtained using
silyl glycosylation conditions, we utilized the sodium salt
method, employing 2,3,5-tri-O-benzoyl-â-D-ribofuranosyl
bromide (17, Scheme 4)34 as the glycosylation reactant.
This reaction proved to be extremely sluggish, even at
elevated temperatures, and required a large excess of
halogenose to go to completion. The lower reactivity of
17 relative to the halogenoses 2 and 7 is presumably
due to the participation of the 2-acyloxy group in the
reaction. After extensive column chromatography to
separate the nucleoside products from unreacted sugar
products, the desired nucleoside 13c was obtained in
low yield. The crude material was deblocked with
sodium methoxide in methanol to provide the desired
nucleoside 15c in 11% yield from 1c after chromatog-
raphy and crystallization.
Biology
The five variously substituted imidazo[4,5-d]isothia-
zoles 1a -e and certain of their ribosyl, deoxyribosyl, and
arabinosyl derivatives (5a -e, 6a -c,e, 10a -e, 11a ,c,e,
15a -d , and 16a ,c) were evaluated in vitro for effects
on cell proliferation and for activity against herpes
simplex virus type 1 (HSV-1) and human cytomegalovi-
rus (HCMV). The heterocyles 1a -e inhibited prolifera-
tion of all the cell lines studied (Table 2). The most
potent was the 5-benzylthio derivative 1e with IC50
)
0.3 and 0.2 µM for L1210 and KB cells, respectively.
Compound 1e also strongly inhibited proliferation of
several other human tumor cell lines. In the presence
of 1 µM 1e, growth of G-361 cells was 7%, HT-29 cells
6%, and ZR-75-1 cells 15% of control. In the presence
of 10 µM, growth of G-361 cells was 2%, HT-29 cells 0%,
and ZR-75-1 cells 3% of control. All of the N-4 glyco-
sylated imidazo[4,5-d]isothiazoles except 6a were mod-
The site of glycosylation was determined by examina-
tion of UV and 13C NMR spectra in a manner similar
to that previously described for the 2′-deoxy and ara-

776 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
Ta ble 2. Antiproliferative and Antiviral Activity of Certain Imidazo[4,5-d]isothiazole Heterocycles and N-4 Nucleosides
substituent
R5
50% inhibitory concentration (µM)
L1210b
KBc
HFFd
HCMVe
HSV-1
a
f
compd no.
R3
R4
1a g
1b
1c
1d
1e
6a
6b
6c
6e
11c
11e
16a
16c
CH3
CH3
H
CH3
H
CH3
CH3
H
H
H
H
CH3
H
H
SCH3
SCH3
SCH2C6H5
SCH2C6H5
H
SCH3
SCH3
SCH2C6H5
SCH3
SCH2C6H5
H
H
H
H
H
H
17
13
7
20
60
15
9
32
26
32
10
14
32
18
32
21
21
45
60
45
45
15
7
0.3
78
33
20
2.2
11
1.5
100
8.6
0.2
deoxy
deoxy
deoxy
deoxy
arabino
arabino
ribo
>100h
>100
>100
>100
30
6
3
15
3
32
15
32
22
3.2
32
3.2
32
60
10
15
25
2.1
21
2.1
32
21
>100
>100
70
20
SCH3
ribo
15
32
a
b
Abbreviation used: deoxy, 2′-deoxy-â-D-erythro-pentofuranosyl; arabino, â-D-arabinofuranosyl; ribo, â-D-ribofuranosyl. Inhibition of
the growth of murine L1210 cells. The 50% inhibitory concentration is the concentration required to decrease growth rate to half of the
control rate; average of two or three experiments. c Inhibition of KB cell growth was determined as described in the text in quadruplicate
assays. Visual cytotoxicity scored on HFF cells at the time of plaque enumeration. e Plaque reduction assay in HFF cells; average of
d
duplicate experiments. f Compounds were assayed by ELISA in quadruplicate wells. Compounds 1a ,b reported in refs 19b and 23. At
g
h
100 µM, the highest concentration evaluated, 0-50% inhibition was observed.
Ta ble 3. Antiproliferative and Antiviral Activity of N-6 Imidazo[4,5-d]isothiazole Nucleosides
substituent
R5
50% inhibitory concentration (µM)
L1210b
KBc
HFFd
HCMVe
HSV-1f
a
compd no.
R3
R6
5a
5b
5c
5d
5e
10a
10b
10c
10d
10e
15a
15b
15c
15d
CH3
CH3
H
CH3
H
CH3
CH3
H
CH3
H
H
SCH3
SCH3
SCH2C6H5
SCH2C6H5
H
SCH3
SCH3
SCH2C6H5
SCH2C6H5
H
SCH3
SCH3
deoxy
deoxy
deoxy
deoxy
deoxy
arabino
arabino
arabino
arabino
arabino
ribo
>100h
>100
>100
>100
8
>100
>100
>100
>100
>100
10
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
90
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
150
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
>100
100
>100
>100
>100
>100
>100
>100
CH3
CH3
H
ribo
ribo
ribo
>100
>100
>100
>100
>100
>100
>100
>100
>100
CH3
SCH2C6H5
a
-f,hSee legend to Table 2 for footnotes.
erately toxic to human foreskin fibroblasts (HFF cells)
in a visual assay and also inhibited the growth of KB
and L1210 cells (Table 2). This toxicity was greatest
for the 5-benzylthio derivatives 6e and 11e and was
seen in both normal (HFF cells) and tumor (L1210 and
KB) cells. The lack of or low activity for the 5-unsub-
stituted compounds 6a and 16a suggested that the
presence of a 5-alkylthio or a 5-arylthio moiety may be
important for the cytotoxic properties of this series. The
5-benzylthio substituent conferred the highest potency,
as exemplified by 1e, 6e, and 11e. Activity, or lack
thereof, against herpes simplex virus type 1 (HSV-1)
and human cytomegalovirus (HCMV) paralleled cyto-
toxicity to L1210, KB, and HFF cells. Similarly, the
heterocycles 1a -e inhibited HCMV and HSV-1 to
approximately the same extent that they produced
cytotoxicity (Table 2), indicating that neither they nor
the N-4 nucleosides have specific antiviral activity.
In contrast, most of the N-6 glycosylated imidazo[4,5-
d]isothiazoles (Table 3) were inactive in the L1210 assay
as well as both antiviral assays and their respective
cytotoxicity controls. The two exceptions to this gen-
eralization were that the 3-unsubstituted 5-(benzylthio)-
6-(2-deoxy-â-D-ribofuranosyl) derivative 5e, and the
5-unsubstituted 3-methyl 6-â-D-ribofuranosyl derivative
15a , were moderately cytotoxic to L1210 cells. The
latter compound is particularly interesting in this
regard since it is the analog of 6-methyl-9-â-D-ribofura-
nosylpurine (II), which is extremely cytotoxic (IC50 ) 4
nM).18 Thus, replacing the purine ring system with the
imidazo[4,5-d]isothiazole ring system led to a major
decrease (g1000-fold) in cytotoxic potency. The struc-
tures of 6-methyl-9-â-D-ribofuranosylpurine (II) and 15a
overlap closely, as shown by modeling studies and X-ray
crystal data, implying that steric alterations are prob-
ably not responsible for the differing activity. The

Imidazo[4,5-d]isothiazole Nucleosides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 777
70.05, 61.18, 40.10, 18.11; UV λmax (methanol) 207 (4170), 246
(9250), 250 (shoulder) nm. Anal. Calcd for C10H13N3O3-
S‚0.4H2O: C, H, N.
fusion of two five-membered rings is expected to lead
to an increased electron density of the overall ring
system, and their differing biological activity is probably
related to the profoundly different electronic character-
istics of these imidazo[4,5-d]isothiazoles.
4-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n o-
syl)-3-m eth yl-5-(m eth ylth io)im idazo[4,5-d]isoth iazole (4b)
a n d 6-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a -
n osyl)-3-m e t h yl-5-(m e t h ylt h io)im id a zo[4,5-d ]isot h ia -
zole (3b). 3-Methyl-5-(methylthio)imidazo[4,5-d]isothiazole
(1b, 926 mg, 5 mmol) was dissolved in warm, dry acetonitrile
(50 mL). Sodium hydride (80% w/w dispersion in mineral oil,
180 mg, 6 mmol) was added in one portion, and the resulting
suspension was stirred at room temperature for 1 h. The
suspension was cooled with an ice water bath, and 2-deoxy-
3,5-di-O-p-toluoyl-R-D-erythro-pentofuranosyl chloride (2, 2.33
g, 6 mmol) was added in one portion. The cold bath was
removed, and stirring was continued for 3 h at room temper-
ature. Ethyl acetate (50 mL) was added, and the suspension
was filtered through a thin pad (2 cm) of silica, which was
washed with ethyl acetate until no more product eluted (by
TLC). The combined filtrate washings were concentrated to
Exp er im en ta l Section
Gen er a l. Unless otherwise noted, general methods and
experimental procedures were the same as described previ-
ously.19 In all cases, the N-4 derivatives tested exhibited no
separation between antiviral activity and toxicity toward
uninfected cells.
4-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n o-
syl)-3-m et h ylim id a zo[4,5-d ]isot h ia zole (4a ) a n d 6-(2-
Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n osyl)-3-
m eth ylim idazo[4,5-d]isoth iazole (3a). 3-Methylimidazo[4,5-
d]isothiazole (1a , 0.28 g, 2 mmol) was dissolved in dry
acetonitrile (20 mL). Sodium hydride (80% w/w dispersion in
mineral oil, 90 mg, 3 mmol) was added in one portion, and
the resulting suspension stirred at room temperature for 1 h.
2-Deoxy-3,5-di-O-p-toluoyl-R-D-erythro-pentofuranosyl chloride
(2, 0.97 g, 2.5 mmol) was added in one portion, and stirring
was continued for 2.5 h at room temperature. Saturated
ammonium chloride (2 mL) was added, and the mixture was
partitioned between ethyl acetate (60 mL) and water (20 mL).
The aqueous layer was washed with ethyl acetate, and the
combined organics were washed with brine, dried over mag-
nesium sulfate, filtered, and concentrated to a syrup. Chro-
matography (70% ethyl acetate/hexane, 4 × 20 cm) afforded
4a (0.28 g, 29%) as a syrup which was homogeneous by TLC
and used without further purification: Rf 0.61 (70% ethyl
acetate/hexane); 1H NMR (300 MHz, CDCl3) δ 7.96 (s, 1H, 5-H),
7.94 (d, 2H), 7.82 (d, 2H), 7.29 (d, 2H), 7.22 (d, 2H), 6.38 (t, J
) 6.6 Hz, 1H, H-1′), 5.70 (m, 1H), 4.71-4.59 (m, 3H), 2.95-
2.75 (m, 2H), 2.67 (s, 3H), 2.45 (s, 3H), 2.40 (s, 3H). Continued
elution of the column provided 0.32 g (33%) of 3a , which was
homogeneous by TLC and used without further purification:
Rf 0.29 (70% ethyl acetate/hexane); 1H NMR (300 MHz, CDCl3)
δ 7.95 (d, 2H), 7.85 (s, 1H, 5-H), 7.84 (d, 2H), 7.28 (d, 2H),
7.20 (d, 2H), 6.26 (t, J ) 6.7 Hz, 1H, H-1′), 5.71 (m, 1H), 4.70-
4.54 (m, 3H), 2.82-2.75 (m, 2H), 2.59 (s, 3H), 2.44 (s, 3H), 2.40
(s, 3H).
6-(2-Deoxy-â-D-er yth r o-p en tofu r a n osyl)-3-m eth ylim id -
a zo[4,5-d ]isoth ia zole (5a ). To a solution of sodium meth-
oxide (20 mg) in methanol (20 mL) was added 3a (0.32 g, 0.66
mmol), and the suspension was stirred 18 h at room temper-
ature. The resulting solution was diluted with water (10 mL)
and then adjusted to pH 7 with 1 N HCl. The solution was
concentrated and then chromatographed (10% methanol/
chloroform, 2 × 15 cm) to provide 0.14 g (82%) of 5a : mp 148-
149 °C; [R]D -12.1° (c 1.0, methanol); Rf 0.26 (10% methanol/
chloroform); 1H NMR (300 MHz, DMSO-d6) δ 8.22 (s, 1H), 6.26
(t, J ) 6.7 Hz, 1H, H-1′), 5.40 (d, 1H, D2O exchangeable), 4.94
(t, 1H, D2O exchangeable), 4.30 (m, 1H), 3.87 (m, 1H), 3.47
(m, 2H), 2.48 (s, 3H), 2.31 (m, 2H); 13C NMR (90 MHz, DMSO-
d6) δ 153.32 (C3), 148.79 (C3a), 145.99 (C6a), 145.04 (C5),
87.74, 85.75, 70.82, 61.88, 40.21-38.82 (C-2′ hidden by solvent
peak), 16.28; UV λmax (methanol) 208 (16 300), 248 (5560) nm.
Anal. Calcd for C10H13N3O3S: C, H, N.
a
brown syrup, which was chromatographed (30% ethyl
acetate/hexane, 5 × 20 cm). Fractions containing the faster
moving minor product were pooled and concentrated to afford
0.52 g (19%) of crude 4b as a foam after drying at 0.1 mmHg.
This material contained a slightly faster moving impurity by
TLC, but was of significant purity for use in further reac-
tions: Rf 0.67 (50% ethyl acetate/hexane); 1H NMR (300 MHz,
CDCl3) δ 8.00-7.87 (m, 4H), 7.30-7.19 (m, 4H), 6.30 (dd, 1H,
H-1′), 5.71-5.64 (m, 1H), 4.89-4.65 (m, 2H), 4.55-4.49 (m,
1H), 2.96-2.50 (m, 2H), 2.69 (s, 3H), 2.62 (s, 3H), 2.44 (s, 3H),
2.40 (s, 3H). Fractions containing the slower moving major
product were pooled and concentrated to afford 1.74 g (65%)
3b as colorless crystals: mp 126-127 °C; Rf 0.45 (50% ethyl
1
acetate/hexane); H NMR (300 MHz, CDCl3) δ 8.00-7.82 (m,
4H), 7.33-7.18 (m, 4H), 6.31 (dd, 1H, H-1′), 5.70-5.65 (m, 1H),
4.71-4.58 (m, 3H), 2.82-2.50 (m, 2H), 2.74 (s, 3H), 2.58 (s,
3H), 2.44 (s, 3H), 2.39 (s, 3H). Anal. Calcd for C27H27N3O5S2:
C, H, N.
6-(2-De oxy-â-D-er yt h r o-p e n t ofu r a n osyl)-3-m e t h yl-5-
(m eth ylth io)im id a zo[4,5-d ]isoth ia zole (5b). To a solution
of 3b (538 mg, 1 mmol) in methanol (10 mL) was added 0.3
mL of Dowex-1 (-OH) washed with methanol. The resulting
suspension was stirred at room temperature for 24 h and then
filtered. The resin was washed well with methanol, and the
filtrates were concentrated. The residue was dissolved in ethyl
acetate and then partitioned between hexane (20 mL) and
water (20 mL). The precipitated solid was dissolved by heating
on a steam bath, and the aqueous phase was removed and
allowed to cool slowly to 0 °C. The resulting solid was collected
and dried to yield 198 mg (66%, mp 151-153 °C) of 5b.
A
second crop provided an additional 56 mg (19%, mp 150-151
°C), giving a combined yield of 85%: [R]D -0.8° (c 1.0,
1
methanol); Rf 0.44 (10% methanol/chloroform); H NMR (300
1
MHz, DMSO-d6) δ 6.12 (dd, 1H,
/ height peak width ) 15
2
Hz), 5.44 (d, 1H, D2O exchangeable), 4.92 (t, 1H, D2O ex-
changeable), 4.28 (m, 1H), 3.88 (m, 1H), 3.48 (m, 2H), 2.67 (s,
3H), 2.47 (s, 3H), 2.40-2.15 (m, 2H); 13C NMR (90 MHz,
DMSO-d6) δ 152.06 (C3), 151.60 (C5), 147.29 (C3a), 147.09
(C6a), 87.53, 84.74, 70.59, 61.64, 39.30, 15.88, 15.32; UV λmax
(methanol) 221 (15 470), 269 (7720) nm. Anal. Calcd for C11
H15N3O3S: C, H, N.
-
4-(2-Deoxy-â-D-er yth r o-p en tofu r a n osyl)-3-m eth ylim id -
a zo[4,5-d ]isoth ia zole (6a ). To a solution of sodium meth-
oxide (20 mg) in methanol (10 mL) was added 4a (0.28 g, 0.57
mmol), and the suspension was stirred 18 h at room temper-
ature. The resulting solution was diluted with an equal
volume of water and then adjusted to pH 7 with 1 N HCl. The
solution was concentrated and then chromatographed (10%
methanol/chloroform to 20% methanol chloroform, 2 × 15 cm)
to provide 0.11 g (76%) of 6a . A sample was crystallized from
water for analysis: mp 155-156 °C; [R]D -35.2° (c 0.9,
4-(2-De oxy-â-D-er yt h r o-p e n t ofu r a n osyl)-3-m e t h yl-5-
(m eth ylth io)im id a zo[4,5-d ]isoth ia zole (6b). To a solution
of 4b (500 mg, 0.93 mmol) in methanol (10 mL) was added 0.3
mL of Dowex-1 (-OH) washed with methanol. The resulting
suspension was stirred at room temperature for 24 h and then
filtered. The resin was washed well with methanol, and the
filtrates were concentrated. The residue was dissolved in ethyl
acetate and then partitioned between hexane (20 mL) and
water (20 mL). The precipitated solid was dissolved by heating
on a steam bath, and the aqueous phase was removed and
allowed to cool slowly to 0 °C. The resulting solid was collected
and dried to yield 213 mg (76%) of 6b: mp 68-70 °C; [R]D
+21.0° (c 1.1, methanol); Rf 0.43 (10% methanol/chloroform);
1
methanol); Rf 0.26 (10% methanol/chloroform); H NMR (300
MHz, DMSO-d6) δ 8.41 (s, 1H), 6.33 (t, J ) 6.4 Hz, 1H, H-1′),
5.38 (d, 1H, D2O exchangeable), 4.94 (t, 1H, D2O exchange-
able), 4.35 (m, 1H), 3.87 (dd, 1H), 3.49 (m, 2H), 2.60 (s, 3H),
2.66-2.31 (m, 2H); 13C NMR (90 MHz, DMSO-d6) δ 163.51
(C6a), 148.37 (C3), 144.48 (C5), 134.86 (C5), 87.89, 85.05,
1
1H NMR (300 MHz, DMSO-d6) δ 6.11 (dd, 1H, /2 height peak
width ) 17 Hz), 5.40 (d, 1H, D2O exchangeable), 4.95 (t, 1H,

778 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
D2O exchangeable), 4.27 (m, 1H), 3.77 (m, 1H), 3.66 (m, 2H),
2.70 (s, 3H), 2.58 (s, 3H), 2.41 (m, 1H), 2.21 (m, 1H); 13C NMR
(75 MHz, DMSO-d6) δ 163.12, (C6a) 154.18 (C5), 147.26 (C3),
135.24 (C3a), 86.95, 84.11, 69.39, 60.85, 40.43, 19.17, 15.08;
UV λmax (methanol) 258 (12 660), 277 (14 360) nm. Anal.
Calcd for C11H15N3O3S: C, H, N.
NMR (90 MHz, DMSO-d6) δ 163.17 (C6a), 154.88 (C5), 140.00
(C3), 136.57 (C3a), 87.87, 85.09, 70.59, 61.40, 39.70 (partially
obscured by solvent peak), 14.99; UV λmax (methanol) 220
(4740), 255 (9860), 279 (12 080) nm. Anal. Calcd for C10H13
N3O3S2: C, H, N.
-
4-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n o-
syl)-3-m et h yl-5-[(p h en ylm et h yl)t h io]im id a zo[4,5-d ]iso-
th iazole (4d) an d 6-(2-Deoxy-3,5-di-O-p-tolu oyl-â-D-er yth r o-
p en t ofu r a n osyl)-3-m et h yl-5-[(p h en ylm et h yl)t h io]im id -
a zo[4,5-d ]isoth ia zole (3d ). 3-Methyl-5-[(phenylmethyl)thio]-
imidazo[4,5-d]isothiazole (1d , 0.30 g, 1.15 mmol) was dissolved
in warm, dry acetonitrile (12 mL). Sodium hydride (80% w/w
dispersion in mineral oil, 42 mg, 1.38 mmol) was added in one
portion, and the resulting suspension stirred at room temper-
ature for 2 h. 2-Deoxy-3,5-di-O-p-toluoyl-R-D-erythro-pento-
furanosyl chloride (2, 0.54 g, 1.38 mmol) was added in one
portion, and stirring was continued for 3 h at room tempera-
ture. Ethyl acetate (20 mL) was added, and the suspension
was filtered through a thin pad (2 cm) of silica, which was
washed with ethyl acetate until no more product eluted (by
TLC). The combined filtrate washings were concentrated to
4-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n o-
syl)-5-(m eth ylth io)im idazo[4,5-d]isoth iazole (4c) an d 6-(2-
Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n osyl)-5-
(m eth ylth io)im idazo[4,5-d]isoth iazole (3c). 5-(Methylthio)-
imidazo[4,5-d]isothiazole (1c, 0.21 g, 1.24 mmol) was dissolved
in warm, dry acetonitrile (20 mL). Sodium hydride (80% w/w
dispersion in mineral oil, 45 mg, 1.49 mmol) was added in one
portion, and the resulting suspension was stirred at room
temperature for 1 h. 2-Deoxy-3,5-di-O-p-toluoyl-R-D-erythro-
pentofuranosyl chloride (2, 0.54 g, 1.38 mmol) was added in
one portion, and stirring was continued for 2 h at room
temperature. Saturated aqueous ammonium chloride (1 mL)
was added, and the mixture was partitioned between ethyl
acetate (30 mL) and water (15 mL). The aqueous layer was
washed with ethyl acetate (2 × 10 mL), and the combined
organic phases were washed with brine, dried over magnesium
sulfate, filtered, and concentrated. The resulting syrup was
chromatographed (35% ethyl acetate/hexane to 50% ethyl
acetate/hexane, 4 × 15 cm). Fractions containing the faster
moving minor product were pooled and concentrated to afford
0.32 g (49%) of crude 4c as an oil after drying at 0.1 mmHg:
Rf 0.80 (50% ethyl acetate/hexane); 1H NMR (300 MHz, CDCl3)
δ 8.38 (s, 1H), 7.96 (d, 2H), 7.82 (d, 2H), 7.29 (d, 2H), 7.20 (d,
2H), 6.24 (dd, 1H, H-1′), 5.70 (m, 1H), 4.70-4.55 (m, 3H), 2.90-
2.65 (m, 1H), 2.77 (s, 3H), 2.50-2.30 (m, 1H), 2.47 (s, 3H), 2.42
(s, 3H). Fractions containing the slower moving major product
were pooled and concentrated to afford 0.33 g (50%) of 3c,
which was used without further purification: Rf 0.48 (50%
a
brown syrup, which was chromatographed (30% ethyl
acetate/hexane, 4 × 15 cm). Fractions containing the faster
moving minor product were pooled and concentrated to afford
0.16 g (23%) of crude 4d as an oil after drying at 0.1 mmHg:
Rf 0.46 (30% ethyl acetate/hexane); 1H NMR (300 MHz, CDCl3)
δ 8.05-7.80 (m, 4H), 7.40-7.05 (m, 9H), 6.24 (dd, 1H, H-1′),
5.59 (m, 1H), 4.85-4.40 (m, 5H), 2.75-2.20 (m, 2H), 2.60 (s,
3H), 2.46 (s, 3H), 2.39 (s, 3H). Fractions containing the slower
moving major product were pooled and concentrated to afford
0.59 g (70%) of solid 3d , which was used without further
purification: Rf 0.31 (30% ethyl acetate/hexane); 1H NMR (300
MHz, CDCl3) δ 7.94 (d, 2H), 7.85 (d, 2H), 7.35-7.15 (m, 9H),
6.12 (dd, 1H, H-1′), 5.56 (m, 1H), 4.65-4.30 (m, 5H), 2.69-
2.25 (m, 2H), 2.62 (s, 3H), 2.45 (s, 3H), 2.39 (s, 3H).
1
ethyl acetate/hexane); H NMR (300 MHz, CDCl3) δ 8.39 (s,
1H), 7.96 (d, 2H), 7.84 (d, 2H), 7.29 (d, 2H), 7.84 (d, 2H), 6.28
(dd, 1H, H-1′), 5.67 (m, 1H), 4.68-4.62 (m, 3H), 2.85-2.60 (m,
2H), 2.75 (s, 3H), 2.45 (s, 3H), 2.39 (s, 3H).
6-(2-De oxy-â-D-er yt h r o-p e n t ofu r a n osyl)-3-m e t h yl-5-
[(p h en ylm eth yl)th io]im id a zo[4,5-d ]isoth ia zole (5d ).
A
suspension of 3d (0.45 g, 0.73 mmol) in methanolic ammonia
(15 mL) was stirred 48 h at room temperature. The resulting
solution was concentrated, and the residue was chromato-
graphed (80% ethyl acetate/hexane, 2 × 15 cm) to provide 0.26
g of a clear glass. Crystallization from methanol/water af-
forded 0.24 g (85%) of 5d hydrate: mp 90-91 °C; [R]D +84.8°
6-(2-De oxy-â-D -er yt h r o-p e n t ofu r a n osyl)-5-(m e t h yl-
th io)im id a zo[4,5-d ]isoth ia zole (5c). To a solution of 3c
(0.29 g, 0.55 mmol) in methanol (10 mL) was added 0.3 mL of
Dowex-1 (-OH) washed with methanol. The resulting suspen-
sion was stirred at room temperature for 48 h and then
filtered. The resin was washed well with methanol, and the
filtrates were concentrated. The residue was dissolved in ethyl
acetate and then partitioned between hexane (20 mL) and
water (20 mL). The precipitated solid was dissolved by heating
on a steam bath, and the aqueous phase was removed and
allowed to cool slowly to 0 °C. The resulting solid was collected
and dried to yield 72 mg (46%) of 5c after two crops: mp
131.5-132 °C; [R]D +1.4° (c 0.6, methanol); Rf 0.54 (10%
(c 1.1, methanol); Rf 0.58 (10% methanol/chloroform); 1H NMR
1
(360 MHz, DMSO-d6) δ 7.38-7.20 (m, 5H), 6.09 (dd, 1H,
/
2
height peak width ) 15.5 Hz), 5.41 (d, 1H, D2O exchangeable),
4.92 (t, 1H, D2O exchangeable), 4.62 (s, 2H), 4.25 (m, 1H), 3.85
(m, 1H), 3.55-3.40 (m, 2H), 2.49 (s, 3H), 2.21-2.05 (m, 2H);
13C NMR (90 MHz, DMSO-d6) δ 152.30 (C3), 149.44 (C5),
147.34 (C3a), 146.87 (C6a), 136.98, 128.94, 128.54, 127.53,
87.76, 85.11, 70.78, 61.77, 39.35, 37.65, 16.29; UV λmax
1
methanol/chloroform); H NMR (360 MHz, DMSO-d6) δ 8.55
(methanol) 220 (20 590), 272 (8800) nm. Anal. Calcd for C17-
H19N3O3S2‚H2O: C, H, N.
(s, 1H), 6.13 (dd, 1H, 1/2 height peak width ) 15.0 Hz), 5.47
(d, 1H, D2O exchangeable), 4.95 (t, 1H, D2O exchangeable),
4.31 (m, 1H), 3.90 (m, 1H), 3.51 (m, 2H), 2.68 (s, 3H), 2.40-
2.20 (m, 2H); 13C NMR (90 MHz, DMSO-d6) δ 152.41 (C5),
148.88 (C3a), 147.90 (C6a), 143.70 (C3), 87.87, 85.02, 70.88,
61.76, 39.41, 15.22; UV λmax (methanol) 221 (11 140), 272
(5490) nm. Anal. Calcd for C10H13N3O3S2: C, H, N.
4-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a n o-
syl)-5-[(p h en ylm eth yl)th io]im id a zo[4,5-d ]isoth ia zole (4e)
a n d 6-(2-Deoxy-3,5-d i-O-p-tolu oyl-â-D-er yth r o-p en tofu r a -
n osyl)-5-[(p h en ylm eth yl)th io]im id a zo[4,5-d ]isoth ia zole
(3e). 5-[(Phenylmethyl)thio]imidazo[4,5-d]isothiazole (1e, 90
mg, 0.33 mmol) was dissolved in warm, dry acetonitrile (5 mL).
Sodium hydride (80% w/w dispersion in mineral oil, 12 mg,
0.39 mmol) was added in one portion, and the resulting
suspension was stirred at room temperature for 1 h. 2-Deoxy-
3,5-di-O-p-toluoyl-R-D-erythro-pentofuranosyl chloride (152 mg,
0.39 mmol) was added in one portion, and stirring was
continued for 3 h at room temperature. The mixture was
filtered through a thin pad of silica, which was then washed
with ethyl acetate until no more product eluted (by TLC). The
filtrate and washings were combined and concentrated, and
the resulting syrup was chromatographed (30% ethyl acetate/
hexane, 2 × 15 cm). Fractions containing the faster moving
minor product were pooled and concentrated to afford 90 mg
(45%) of crystalline 4e: mp 115-116 °C; Rf 0.51 (30% ethyl
acetate/hexane); 1H NMR (270 MHz, CDCl3) δ 8.37 (s, 1H),
7.94 (d, 2H), 7.82 (d, 2H), 7.35-7.18 (m, 5H), 6.17 (dd, 1H,
H-1′), 5.62 (m, 1H), 4.70-4.45 (m, 5H), 2.75-2.35 (m, 2H), 2.44
(s, 3H), 2.40 (s, 3H). Fractions containing the slower moving
4-(2-De oxy-â-D -er yt h r o-p e n t ofu r a n osyl)-5-(m e t h yl-
th io)im id a zo[4,5-d ]isoth ia zole (6c). To a solution of 4c
(0.29 g, 0.55 mmol) in methanol (10 mL) was added 0.3 mL of
Dowex-1 (-OH) washed with methanol. The resulting suspen-
sion was stirred at room temperature for 48 h and then
filtered. The resin was washed well with methanol, and the
filtrates were concentrated. The residue was dissolved in ethyl
acetate and then partitioned between hexane (20 mL) and
water (20 mL). The precipitated solid was dissolved by heating
on a steam bath, and the aqueous phase was removed and
allowed to cool slowly to 0 °C. The resulting solid was collected
and dried to yield 43 mg (27%) of 6c after two crops: mp 156-
157 °C; [R]D -5.3° (c 0.7, methanol); Rf 0.51 (10% methanol/
chloroform); 1H NMR (360 MHz, DMSO-d6) δ 8.68 (s, 1H), 6.09
1
(dd, 1H,
/ height peak width ) 16.0 Hz), 5.40 (d, 1H, D2O
2
exchangeable), 5.03 (t, 1H, D2O exchangeable), 4.36 (m, 1H),
3.89 (m, 1H), 3.58 (t, 2H), 2.72 (s, 3H), 2.55-2.25 (m, 2H); 13
C

Imidazo[4,5-d]isothiazole Nucleosides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 779
major product were pooled and concentrated to afford 91 mg
(46%) of crystalline 3e, which was used without further
purification: mp 118-121 °C; Rf 0.34 (30% ethyl acetate/
phase was washed with hexane and then lyophilized. The
resulting crude solid was chromatographed (20% methanol/
chloroform, 2 × 15 cm) to provide 0.23 g (100%) of 10a : mp
175-176 °C; [R]D -13.0° (c 1.0, methanol); Rf 0.20 (10%
1
hexane); H NMR (270 MHz, CDCl3) δ 8.46 (s, 1H), 7.96 (d,
1
2H), 7.84 (d, 2H), 7.35-7.17 (m, 5H), 6.17 (t, 1H, H-1′), 5.61
(m, 1H), 4.65-4.40 (m, 5H), 2.50-2.34 (m, 2H), 2.46 (s, 3H),
2.41 (s, 3H).
methanol/chloroform); H NMR (300 MHz, DMSO-d6) δ 8.18
(s, 1H), 6.09 (d, J ) 4.1 Hz, 1H, H-1′), 5.98 (d, D2O exchange-
able), 5.57 (d, 1H, D2O exchangeable), 4.98 (t, D2O exchange-
able), 4.16 (m, 1H), 3.98 (dd, 1H), 3.70-3.50 (m, 2H), 2.45 (s,
3H); 13C NMR (75 MHz, DMSO-d6) δ 151.99 (C3), 148.43 (C6a),
148.15 (C3a), 145.31 (C5), 86.32, 84.81, 76.84, 75.75, 61.42,
16.04; UV λmax (methanol) 210 (15 820), 249 (5670) nm. Anal.
Calcd for C10H13N3O4S: C, H, N.
6-(2-De oxy-â-D-er yt h r o-p e n t ofu r a n osyl)-5-[(p h e n yl-
m eth yl)th io]im id a zo[4,5-d ]isoth ia zole (5e). To a solution
of sodium methoxide (20 mg) in methanol (20 mL) was added
3e (0.16 g, 0.27 mmol), and the suspension was stirred 24 h
at room temperature. The resulting solution was diluted with
an equal volume of water and then adjusted to pH 6 with 1 N
HCl. The solution was concentrated, chromatographed (5%
methanol/chloroform, 2 × 15 cm), and then crystallized from
toluene/ethanol to provide 81 mg (83%) of 5e: mp 99-102 °C;
[R]D +69.0° (c 0.7, methanol); Rf 0.24 (5% methanol/chloro-
form); 1H NMR (300 MHz, DMSO-d6) δ 8.60 (s, 1H), 7.40-
7.20 (m, 5H), 6.10 (t, J ) 6.8 Hz, 1H, H-1′), 5.42 (d, 1H, D2O
exchangeable), 4.94 (t, 1H, D2O exchangeable), 4.27 (s, 2H),
4.27 (m, 1H), 3.87 (m, 1H), 3.49 (m, 2H), 2.25-2.05 (m, 2H);
13C NMR (75 MHz, DMSO-d6) δ 149.96 (C5), 148.56 (C3a),
147.43 (C6a), 143.49 (C5), 136.67, 128.44, 128.09, 127.07,
87.60, 84.90, 70.58, 61.56, 39.33, 37.27; UV λmax (methanol)
218 (18 290), 275 (7690) nm. Anal. Calcd for C16H17N3O3S2‚
1/8H2O: C, H, N.
3-Meth yl-5-m eth ylth io-6-(2,3,5-tr i-O-ben zyl-â-D-ar abin o-
fu r a n osyl)im id a zo[4,5-d ]isot h ia zole (8b ). 3-Methyl-5-
(methylthio)imidazo[4,5-d]isothiazole (1b, 926 mg, 5 mmol)
was dissolved in warm, dry acetonitrile (50 mL). Sodium
hydride (80% w/w dispersion in mineral oil, 180 mg, 6 mmol)
was added in one portion, and the resulting suspension was
stirred at room temperature for 1 h. 2,3,5-Tri-O-benzyl-R-D-
arabinofuranosyl chloride (7, 2.63 g, 6 mmol) in 10 mL of
dichloromethane was added via cannula, and stirring was
continued for 24 h at room temperature. Saturated aqueous
ammonium chloride (5 mL) was added, and the suspension
was concentrated to a small volume. Water (50 mL) was
added, and the mixture was extracted with ethyl acetate (50
mL + 2 × 10 mL). The combined organics were washed with
brine (20 mL), dried over magnesium sulfate, filtered, and
concentrated to a brown syrup. This residue was chromato-
graphed (chloroform to 2% methanol/chloroform, 5 × 20 cm)
to separate 8b (Rf 0.61, 2% methanol/chloroform), which was
contaminated with a small amount of the N-4 isomer (Rf 0.74),
from unreacted starting material (0.16 g, 17%). Fractions
containing the faster moving products were pooled, concen-
trated, and chromatographed again (30% ethyl acetate/hexane,
5 × 20 cm) to afford the faster moving N-4 isomer (0.12 g, 4%)
as an oily solid containing an inseparable mixture of anomers
(R/â ) 1.2:1). Continued elution of the column gave 8b (2.00
g, 68%) as a syrup containing an inseparable mixture of
anomers (R/â ) 1:25) after drying at 0.1 mmHg: Rf 0.61 (2%
methanol/chloroform); 1H NMR (300 MHz, CDCl3) δ 7.36-6.90
(m, 13H), 6.92 (m, 2H), 6.10 (d, J ) 4.0 Hz, 1H, H-1′ of
â-anomer), 6.04 (d, J ) 3.4 Hz, 0.04H, H-1′ of R-anomer), 4.61-
4.45 (m, 4H), 4.47-4.20 (m, 3H), 4.14 (m, 1H), 4.04 (m, 1H),
3.75-3.58 (m, 2H), 2.66 (s, 3H), 2.60 (s, 3H).
4-(2-De oxy-â-D-er yt h r o-p e n t ofu r a n osyl)-5-[(p h e n yl-
m eth yl)th io]im id a zo[4,5-d ]isoth ia zole (6e). To a solution
of sodium methoxide (20 mg) in methanol (20 mL) was added
1a (0.16 g, 0.27 mmol), and the suspension was stirred 24 h
at room temperature. The resulting solution was diluted with
an equal volume of water and then adjusted to pH 6 with 1 N
HCl. The solution was concentrated, chromatographed (5-
10% methanol/chloroform, 2 × 15 cm), and then crystallized
from petroleum ether/methanol to provide 94 mg (94%) of 6e:
mp 144-145 °C; [R]D +44.6° (c 1.0, methanol); Rf 0.18 (5%
1
methanol/chloroform); H NMR (360 MHz, DMSO-d6) δ 8.70
(s, 1H), 7.43-7.24 (m, 5H), 6.08 (dd, 1H, 1/2 height peak width
) 16.0 Hz), 5.37 (d, D2O exchangeable), 5.03 (t, D2O exchange-
able), 4.56 (s, 2H), 4.33 (m, 1H), 3.86 (m, 1H), 3.57 (t, 2H),
2.38 (m, 1H), 2.14 (m, 1H); 13C NMR (90 MHz, DMSO-d6) δ
163.02 (C6a), 153.01 (C5), 140.17 (C3), 136.93, 136.41 (C5),
128.90, 128.51, 127.52, 87.86, 85.18, 70.53, 61.35, 39.73, 36.75;
UV λmax (methanol) 256 (12 000), 282 (14 620) nm. Anal.
Calcd for C16H17N3O3S2: C, H, N.
6-â-D-Ar a bin ofu r a n osyl-3-m eth yl-5-(m eth ylth io)im id -
a zo[4,5-d ]isoth ia zole (10b). To a solution of 8b (2.0 g, 3.4
mmol) in dry dichloromethane (15 mL) at -78 °C was added
boron trichloride (1.0 M in dichloromethane, 34 mL) dropwise
over 0.5 h. The solution was warmed to 0 °C via an ice bath
and then stirred for 2 h. The reaction mixture was cooled to
-78 °C and quenched via the dropwise addition of methanol
(17 mL), and the vigorously stirred solution was then neutral-
ized to pH 7 at 0 °C with an ice cold solution of saturated
sodium bicarbonate (ca. 80 mL). Chloroform (50 mL) was
added, and the organic layer was extracted with water (2 ×
50 mL). The combined aqueous layers were allowed to stand
overnight at 5 °C, and the resulting solid was removed by
filtration and then washed with ethanol. This process was
repeated, and the filtrate was concentrated onto 8 g of silica
gel and then chromatographed (10% methanol/chloroform, 4
× 20 cm) to yield a clear glass, which was homogeneous by
TLC. Attempts to crystallize this material from a variety of
solvents failed and led to decomposition of some product.
Chromatography (90:9:1 ethyl acetate/methanol/water, 4 × 20
cm) provided a glassy residue, which was dissolved in water
at room temperature and then lyophilized to afford 0.49 g
(44%) of 10b as the hemihydrate: mp 154-156 °C; [R]D -9.6°
(c 1.0, methanol); Rf 0.52 (90:9:1 ethyl acetate/methanol/water);
1H NMR (300 MHz, DMSO-d6) δ 5.97 (d, J ) 3.6 Hz, 1H, H-1′),
5.89 (br s, 1H, D2O exchangeable), 5.57 (br s, 1H, D2O
exchangeable), 4.95 (br s, 1H, D2O exchangeable), 4.14 (m, 1H),
3.99 (m, 1H), 3.82 (m, 1H), 3.62 (m, 2H), 2.67 (s, 3H), 2.44 (s,
3H); 13C NMR (90 MHz, DMSO-d6) δ 151.44 (C5), 151.13 (C3),
150.06 (C6a), 147.01 (C3a), 86.15, 85.57, 76.56, 76.24, 61.62,
16.38, 15.58; UV λmax (methanol) 221 (16 680), 269 (8450) nm.
Anal. Calcd for C11H15N3O4S2‚1/2H2O: C, H, N.
3-Met h yl-6-(2,3,5-t r i-O-b en zyl-â-D-a r a b in ofu r a n osyl)-
im id a zo[4,5-d ]isoth ia zole (8a ). Sodium hydride (80% w/w,
90 mg, 3 mmol) was added in one portion to a solution of 1a
(278 mg, 2 mmol) in dry acetonitrile (20 mL), and the resulting
suspension was stirred for 1 h at room temperature. A solution
of 2,3,5-tri-O-benzyl-R-D-arabinofuranosyl chloride (7, 3 mmol)
in dichloromethane (6 mL) was added via cannula, and the
mixture was stirred for 48 h at room temperature. Saturated
ammonium chloride (2 mL) was added, and the suspension
was partitioned between water (20 mL) and ethyl acetate (60
mL). The organic layer was washed with water and brine and
then dried over magnesium sulfate. The solvent was removed,
and the residue was chromatographed (50% ethyl acetate/
hexane to 70% ethyl acetate/hexane). Fractions containing the
faster moving product (Rf 0.54, 50% ethyl acetate/hexane) were
pooled and concentrated to afford 90 mg (8%) of the N-4 isomer.
Fractions containing the major product were combined to
provide 539 mg (50%) of 8a : Rf 0.25 (50% ethyl acetate/
hexane); 1H NMR (270 MHz, CDCl3) δ 7.74 (s, 1H), 7.40-6.95
(m, 15H), 5.92 (d, J ) 4.0 Hz, 1H, H-1′), 4.54-4.05 (m, 9H),
3.62 (m, 2H), 2.62 (s, 3H).
6-â-D-Ar a bin ofu r a n osyl-3-m eth ylim ida zo[4,5-d]isoth ia -
zole (10a ). To a solution of 8a (0.45 g, 0.84 mmol) in dry
dichloromethane (5 mL) at -78 °C was added boron trichloride
(1.0 M in dichloromethane, 8.4 mL) dropwise over 0.25 h. The
solution was warmed to 0 °C via an ice bath and then stirred
for 2 h. The reaction mixture was cooled to -78 °C and
quenched via the dropwise addition of methanol (5 mL), and
the vigorously stirred solution was then neutralized to pH 7
at 0 °C with an ice cold solution of 1 M sodium hydroxide (23-
25 mL). The organic layer was removed, and the aqueous

780 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
5-(Meth ylth io)-4-(2,3,5-tr i-O-ben zyl-â-D-a r a bin ofu r a n o-
syl)im id a zo[4,5-d ]isoth ia zole (9c) a n d 5-(Meth ylth io)-6-
(2,3,5-tr i-O-ben zyl-â-D-a r a bin ofu r a n osyl)im id a zo[4,5-d ]-
isoth ia zole (8c). 5-(Methylthio)imidazo[4,5-d]isothiazole (1c,
650 mg, 3.8 mmol) was dissolved in warm, dry acetonitrile (75
mL). Sodium hydride (80% w/w dispersion in mineral oil, 137
mg, 4.56 mmol) was added in one portion, and the resulting
suspension was stirred at room temperature for 1 h. 2,3,5-
Tri-O-benzyl-R-D-arabinofuranosyl chloride (7, 5 mmol) in 11
mL of dichloromethane was added via cannula, and stirring
was continued for 48 h at room temperature. Saturated
aqueous ammonium chloride (5 mL) was added, and the
suspension was concentrated to a small volume and then
partitioned between water (40 mL) and ethyl acetate (40 mL).
The aqueous phase was extracted with ethyl acetate (2 × 20
mL) and the combined organics were washed with brine (20
mL), dried over magnesium sulfate, filtered, and concentrated
to a brown syrup. This residue was chromatographed (chlo-
roform to 2% methanol/chloroform, 5 × 20 cm) to afford
nucleoside material and unreacted starting material (79 mg,
12%). The combined nucleoside products were chromato-
graphed (30% ethyl acetate, 5 × 20 cm) to afford 0.53 g (24%)
of 9c as an inseparable mixture of anomers (R/â ) 1:5): Rf
filtrates were concentrated and then chromatographed (15%
methanol/chloroform, 2 × 20 cm). Fractions containing prod-
uct were concentrated and recrystallized from warm water to
yield 0.11 g (39%) of 11c: mp 187-189 °C dec; [R]D +12.9° (c
1
1.0, methanol); Rf 0.38 (10% methanol/chloroform); H NMR
(300 MHz, DMSO-d6) δ 8.48 (s, 1H), 5.97 (d, J ) 4.1 Hz, 1H,
H-1′), 5.60 (d, 1H, D2O exchangeable), 5.59 (d, 1H, D2O
exchangeable), 5.15 (t, 1H, D2O exchangeable), 4.10-4.01 (m,
2H), 3.84 (m, 1H), 3.70 (m, 2H), 2.72 (s, 3H); 13C NMR (360
MHz, DMSO-d6) δ 161.82 (C6a), 153.46 (C5), 141.57 (C3),
138.81 (C3a), 87.02, 84.97, 76.20, 75.82, 60.82, 15.11; UV λmax
(methanol) 221 (6130), 255 (10 950), 278 (12 940) nm. Anal.
Calcd for C10H13N3O4S2: C, H, N.
3-Meth yl-5-[(p h en ylm eth yl)th io]-6-(2,3,5-tr i-O-ben zyl-
â-D-a r a bin ofu r a n osyl)im id a zo[4,5-d ]isoth ia zole (8d ). So-
dium hydride (80% w/w, 90 mg, 3 mmol) was added in one
portion to a solution of 1d (653 mg, 2.5 mmol) in dry
acetonitrile (25 mL), and the resulting suspension was stirred
for 1 h at room temperature. A solution of 2,3,5-tri-O-benzyl-
â-D-arabinofuranosyl chloride (7, 3 mmol) in dichloromethane
(11 mL) was added via cannula, and the mixture was stirred
for 72 h at room temperature. Saturated ammonium chloride
(2.5 mL) was added, and the mixture was concentrated to a
small volume and then partitioned between water (25 mL) and
ethyl acetate (25 mL). The organic layer was washed with
brine, dried over magnesium sulfate, and filtered. The solvent
was removed, and the residue was chromatographed (chloro-
form to 2% methanol/chloroform, 4 × 20 cm) to afford nucleo-
side material, followed by unreacted starting material (74 mg,
11%). The nucleoside material was chromatographed again
(30% ethyl acetate/hexane, 4 × 20 cm), and fractions contain-
ing the major product were combined to provide 1.09 g (66%)
of 8d : Rf 0.51 (2% methanol/chloroform); 1H NMR (300 MHz,
CDCl3) δ 7.35-7.15 (m, 13H), 6.89 (m, 2H), 6.00 (d, J ) 3.9
Hz, 1H, H-1′), 4.59-4.35 (m, 6H), 4.20-4.11 (m, 3H), 3.98 (m,
1H), 3.94 (m, 1H), 3.63 (m, 2H), 2.62 (s, 3H).
1
0.84 (50% ethyl acetate/hexane); H NMR (300 MHz, CDCl3)
δ 8.28 (s, 1H), 7.38-7.18 (m, 13H), 6.79 (m, 2H), 6.05 (d, J )
3.2 Hz, 1H, H-1′ of â-anomer), 5.98 (d, J ) 4.1 Hz, 0.19H, H-1′
of R-anomer), 4.60-3.60 (m, 11H), 2.73 (s, 3H). Continued
elution of the column with 50% ethyl acetate/hexane gave 8c
(1.09 g, 50%) as a syrup containing an inseparable mixture of
anomers (R/â ) 1:30): Rf 0.58 (50% ethyl acetate/hexane); 1H
NMR (300 MHz, CDCl3) δ 8.39 (s, 1H), 7.38-7.15 (m, 13H),
6.89 (m, 2H), 6.06 (d, J ) 3.9 Hz, 1H, H-1′ of â-anomer), 6.00
(d, J ) 3.1 Hz, 1H, H-1′ of R-anomer), 4.53 (m, 4H), 4.30-4.03
(m, 5H), 3.68 (m, 2H), 2.69 (s, 3H).
6-â-D-Ar a bin ofu r a n osyl-5-(m eth ylth io)im id a zo[4,5-d ]-
isoth ia zole (10c). To a solution of 8c (1.09 g, 1.90 mmol) in
dry dichloromethane (9 mL) at -78 °C was added boron
trichloride (1.0 M in dichloromethane, 19 mL) dropwise over
0.5 h. The solution was warmed to 0 °C via an ice bath and
then stirred for 2 h. The reaction mixture was cooled to -78
°C, the reaction was quenched via the dropwise addition of
methanol (10 mL), and the vigorously stirred solution was then
neutralized to pH 7 at 0 °C with an ice cold solution of
saturated sodium bicarbonate. Chloroform (30 mL) was added,
and the organic layer was extracted with water (2 × 30 mL).
The aqueous layer was washed with ether, concentrated,
slurried with 2:1 ethanol/ethyl acetate, and filtered. The
filtrates were concentrated and chromatographed (15% metha-
nol/chloroform, 4 × 20 cm) to yield a clear glass, which was
taken up in methanol, diluted with water, and lyophilized to
yield 0.33 g (58%) of 10c hemihydrate. This material was very
hygroscopic, and a portion was recrystallized from methanol
to provide anhydrous 10c: mp 138-140 °C; [R]D -3.7° (c 1.2,
6-â-D-Ar a b in ofu r a n osyl-3-m et h yl-5-[(p h en ylm et h yl)-
th io]im id a zo[4,5-d ]isoth ia zole (10d ). To a solution of 8d
(1.09 g, 1.64 mmol) in dry dichloromethane (8 mL) at -78 °C
was added boron trichloride (1.0 M in dichloromethane, 16 mL)
dropwise over 0.25 h. The solution was warmed to 0 °C via
an ice bath and then stirred for 2 h. The reaction mixture
was cooled to -78 °C, the reaction was quenched via the
dropwise addition of methanol (8 mL), and the vigorously
stirred solution was then neutralized to pH 7 at 0 °C with an
ice cold solution of saturated sodium bicarbonate. The mixture
was extracted with chloroform (3 × 25 mL) and then ethyl
acetate (3 × 25 mL). The combined organic layers were
washed with brine, concentrated, and then chromatographed
(5% methanol/chloroform to 10% methanol/chloroform, 4 × 20
cm). Fractions containing product were concentrated and
chromatographed again (325:9:1 ethyl acetate/methanol/water)
to yield 0.42 g (65%) of a clear foam, which was homogeneous
by TLC. Attempts to crystallize this material from a variety
of solvents failed and led to decomposition of some product.
Chromatography (5% methanol/chloroform to 10% methanol/
chloroform, 4 × 20 cm) provided a glassy residue, which was
dissolved in ethanol, diluted with toluene, and concentrated
afford 0.37 g (51%) of 10d as a hydroscopic foam, which
retained 0.5 mol of toluene: [R]D -3.1° (c 0.9, methanol); Rf
1
methanol); Rf 0.27 (10% methanol/chloroform); H NMR (360
MHz, DMSO-d6) δ 8.46 (s, 1H), 6.00 (d, J ) 5.2 Hz, 1H, H-1′),
5.97 (d, 1H, D2O exchangeable), 5.64 (d, 1H, D2O exchange-
able), 5.03 (t, 1H, D2O exchangeable), 4.16 (m, 1H), 4.01 (m,
1H), 3.83 (m, 1H), 3.62 (m, 2H), 2.68 (s, 3H); 13C NMR (90
MHz, DMSO-d6) δ 152.46 (C5), 150.87 (C6a), 148.63 (C3a),
142.99 (C3), 86.05, 85.67, 76.66, 76.26, 61.60, 15.38; UV λmax
1
(methanol) 222 (13 450), 272 (7040) nm. Anal. Calcd for C10
13N3O4S2: C, H, N.
4-â-D-Ar a bin ofu r a n osyl-5-(m eth ylth io)im id a zo[4,5-d ]-
-
0.42 (10% methanol/chloroform); H NMR (360 MHz, DMSO-
H
d6) δ 8.32 (s, 1H), 7.42-7.24 (m, 5H), 5.96 (d, J ) 3.7 Hz, 1H,
H-1′), 5.92 (d, 1H, D2O exchangeable), 5.58 (d, 1H, D2O
exchangeable), 5.00 (t, 1H, D2O exchangeable), 4.47 (m, 2H),
4.07 (m, 1H), 3.97 (m, 1H), 3.79 (m, 1H), 3.60 (m, 2H), 2.47 (s,
3H); 13C NMR (90 MHz, DMSO-d6) δ 151.33 (C3), 149.87 (C6a
or C5), 149.41 (C5 or C6a), 147.03 (C3a), 136.85, 128.94,
128.48, 127.47, 86.18, 85.52, 76.59, 76.22, 61.59, 37.54, 11.31;
UV λmax (methanol) 221 (18 710), 271 (8740) nm. Anal. Calcd
for C17H19N3O4S2‚1/2C6H5CH3: C, H, N.
5-[(P h en ylm et h yl)t h io]-4-(2,3,5-t r i-O-b en zyl-â-D-a r a -
b in ofu r a n osyl)im id a zo[4,5-d ]isot h ia zole (9e) a n d 5-
[(P h en ylm et h yl)t h io]-6-(2,3,5-t r i-O-b en zyl-â-D-a r a b in o-
fu r a n osyl)im id a zo[4,5-d ]isoth ia zole (8e). 5-(Methylthio)-
imidazo[4,5-d]isothiazole (1e, 247 mg, 1 mmol) was dissolved
isoth ia zole (11c). To a solution of 9c (0.53 g, 0.93 mmol) in
dry dichloromethane (5 mL) at -78 °C was added boron
trichloride (1.0 M in dichloromethane, 9.4 mL) dropwise over
0.5 h. The solution was warmed to 0 °C via an ice bath and
then stirred for 2 h. The reaction mixture was cooled to -78
°C, the reaction was quenched via the dropwise addition of
methanol (5 mL), and the vigorously stirred solution was then
neutralized to pH 7 at 0 °C with an ice cold solution of
saturated sodium bicarbonate. Chloroform (15 mL) was added,
and the organic layer was extracted with water (2 × 15 mL).
The aqueous layer was washed with ether, concentrated,
slurried with 2:1 ethanol/ethyl acetate, and filtered. The

Imidazo[4,5-d]isothiazole Nucleosides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 781
in warm, dry acetonitrile (20 mL). Sodium hydride (80% w/w
dispersion in mineral oil, 36 mg, 1.2 mmol) was added in one
portion, and the resulting suspension stirred at room temper-
ature for 1 h. 2,3,5-Tri-O-benzyl-R-D-arabinofuranosyl chloride
(7, 1.3 mmol) in 5 mL dichloromethane was added via cannula,
and stirring was continued for 48 h at room temperature.
Saturated aqueous ammonium chloride (1 mL) was added, and
the suspension was partitioned between water and ethyl
acetate. The organic phase was washed with brine, dried over
magnesium sulfate, filtered, and concentrated to a syrup. This
residue was chromatographed (20% ethyl acetate/hexane, 4 ×
15 cm) to afford 198 mg (30%) of 9e as an inseparable mixture
of anomers (R/â ) 1:5): Rf 0.75 (30% ethyl acetate/hexane);
1H NMR (270 MHz, CDCl3) δ 8.27 (s, 1H), 7.45-7.02 (m, 18H),
6.73 (m, 2H), 6.03 (d, J ) 3.1 Hz, 1H, H-1′ of â-anomer), 5.98
(d, J ) 3.3 Hz, 0.25H, H-1′ of R-anomer), 4.65-3.45 (m, 13H).
Continued elution of the column gave 8e (297 mg, 46%) as a
syrup containing an inseparable mixture of anomers (R/â )
1:13): Rf 0.60 (30% ethyl acetate/hexane); 1H NMR (270 MHz,
CDCl3) δ 8.42 (s, 1H), 7.45-7.07 (m, 18H), 6.86 (m, 2H), 6.01
(s, 1H, H-1′ of â-anomer), 5.95 (s, 0.08H, H-1′ of R-anomer),
4.65-3.55 (m, 13H).
3-Meth yl-4-(2,3,5-tr i-O-ben zoyl-â-D-r ibofu r a n osyl)im id -
a zo[4,5-d ]isoth ia zole (14a ) a n d 3-Meth yl-6-(2,3,5-tr i-O-
ben zoyl-â-D-r ibofu r an osyl)im idazo[4,5-d]isoth iazole (13a).
3-Methylimidazo[4,5-d]isothiazole (1a , 0.20 g, 1.4 mmol) was
suspended in dry acetonitrile (15 mL), to which was added
N,O-bis(trimethylsilyl)acetamide (0.37 mL, 1.5 mmol) in one
portion. The suspension was heated to 80 °C for 0.5 h, 1-O-
acetyl-2,3,5-tri-O-benzoyl-â-D-ribofuranose (12, 0.72 g, 1.4
mmol) was added in one portion, and then (trimethylsilyl)-
trifluoromethanesulfonate (0.30 mL, 1.6 mmol) was added
dropwise over 5 min. The solution was stirred for 1.25 h at
80 °C, allowed to cool to room temperature, and then poured
into ice cold dichloromethane (15 mL) and saturated aqueous
sodium bicarbonate (15 mL). The aqueous layer was washed
with dichloromethane (15 mL), and the combined organic
phases were dried over magnesium sulfate, filtered, and
concentrated. The residue was chromatographed (50% ethyl
acetate/hexane, 4 × 20 cm), and fractions containing the faster
moving product were concentrated to yield 0.22 g (26%) of 14a
1
as a foam: Rf 0.49 (50% ethyl acetate/hexane); H NMR (300
MHz, CDCl3) δ 8.08-7.90 (m, 6H), 7.62-7.36 (m, 10H), 6.36
(d, J ) 5.5 Hz, 1H, H-1′), 6.02 (t, 1H), 5.94 (m, 1H), 4.91-4.65
(m, 3H), 2.68 (s, 3H). Fractions containing the slower moving
product were combined, and concentrated to yield 0.31 g (37%)
of 13a as a foam: Rf 0.26 (50% ethyl acetate/hexane); 1H NMR
(300 MHz, CDCl3) δ 8.08-7.93 (m, 6H), 7.60-7.39 (m, 10H),
6.23 (d, J ) 4.7 Hz, 1H, H-1′), 5.88-5.82 (m, 2H), 4.91-4.66
(m, 3H), 2.59 (s, 3H).
3-Me t h yl-6-â-D-r ib ofu r a n osylim id a zo[4,5-d ]isot h ia -
zole (15a ). A suspension of 13a (309 mg, 0.53 mmol) in 15
mL of methanolic ammonia was stirred at room temperature
in a pressure tube for 48 h. The solvent was removed, and
the residue was taken up in water (15 mL) and extracted with
chloroform (2 × 5 mL) and then hexane (10 mL). The aqueous
phase was lyophilized, and the resulting solid was chromato-
graphed (15% 9:1 methanol-water/ethyl acetate, 2 × 15 cm)
to afford 135 mg (94%) of 15a (mp 159-161 °C) which was
homogeneous on TLC and contained no impurities in the 1H
NMR spectrum. A sample was recrystallized from methanol
for analysis: mp 171-172.5 °C; [R]D -59.2° (c 1.0, methanol);
Rf 0.27 (10% methanol/chloroform); 1H NMR (300 MHz,
DMSO-d6) δ 8.22 (s, 1H), 5.77 (d, J ) 6.6 Hz, 1H, H-1′), 5.55
(d, 1H, D2O exchangeable), 5.31 (d, 1H, D2O exchangeable),
5.04 (t, 1H, D2O exchangeable), 4.16 (dd, 1H), 4.03 (m, 1H),
3.98 (m, 1H), 3.55 (m, 2H), 2.50 (s, 3H); 13C NMR (90 MHz,
DMSO-d6) δ 153.19 (C3), 148.90 (C3a), 146.00 (C6a), 145.31
(C5), 89.00, 86.01, 73.93, 70.81, 61.64, 16.26; UV λmax (metha-
6-â-D-Ar abin ofu r an osyl-5-[(ph en ylm eth yl)th io]im idazo-
[4,5-d]isoth ia zole (10e). To a solution of 8e (297 mg, 0.46
mmol) in dry dichloromethane (5 mL) at -78 °C was added
boron trichloride (1.0 M in dichloromethane, 4.6 mL) dropwise
over 0.25 h. The solution was warmed to 0 °C via an ice bath
and then stirred for 2 h. The reaction mixture was cooled to
-78 °C and quenched via the dropwise addition of methanol
(5 mL), and the vigorously stirred solution was then neutral-
ized to pH 7 at 0 °C with an ice cold solution of 1 M sodium
hydroxide (ca. 13 mL). The organic layer was removed, and
the aqueous phase was extracted with chloroform (3 × 10 mL)
and then with ethyl acetate (2 × 10 mL). The combined
organics were dried over magnesium sulfate, filtered, concen-
trated, and chromatographed (5% methanol/chloroform, 2 ×
15 cm) to provide a residue which was taken up in methanol,
diluted with water, and lyophilized to afford 117 mg (67%) of
10e as a hydroscopic foam: [R]D +4.8° (c 1.0, methanol); Rf
1
0.44 (10% methanol/chloroform); H NMR (360 MHz, DMSO-
d6) δ 8.51 (s, 1H), 7.44-7.24 (m, 5H), 6.00 (d, 1H, D2O
exchangeable), 5.96 (d, J ) 3.7 Hz, 1H, H-1′), 5.61 (d, 1H, D2O
exchangeable), 5.02 (t, 1H, D2O exchangeable), 4.50 (m, 2H),
4.10 (m, 1H), 3.98 (m, 1H), 3.81 (m, 1H), 3.62 (m, 2H); 13C NMR
(90 MHz, DMSO-d6) δ 150.73 (C6a or C5), 150.63 (C5 or C6a),
148.61 (C3a), 143.22 (C3), 136.99, 128.97, 128.51, 127.49,
86.11, 85.66, 76.72, 76.26, 61.59, 37.18; UV λmax (methanol)
219 (19 551), 274 (8490) nm. Anal. Calcd for C16H17N3O4S2:
C, H, N.
nol) 209 (16 420), 249 (6730) nm. Anal. Calcd for C10H13
N3O4S: C, H, N.
-
3-Me t h yl-4-â-D-r ib ofu r a n osylim id a zo[4,5-d ]isot h ia -
zole (16a ). A suspension of 14a (192 mg, 0.33 mmol) in 10
mL of methanolic ammonia was stirred at room temperature
in a pressure tube for 48 h. The solvent was removed, and
the residue was taken up in water (15 mL) and extracted with
chloroform (2 × 5 mL) and then hexane (10 mL). The aqueous
phase was lyophilized, and the resulting solid was chromato-
graphed (15% 9:1 methanol-water/ethyl acetate, 2 × 15 cm)
to afford 73 mg (82%) of 16a (mp 150-151 °C) which was
homogeneous on TLC and contained no impurities in the 1H
NMR spectrum. A sample was recrystallized from methanol
for analysis: mp 157-158 °C; [R]D -11.7° (c 0.8, methanol);
Rf 0.21 (10% methanol/chloroform); 1H NMR (300 MHz,
DMSO-d6) δ 8.45 (s, 1H), 5.90 (d, J ) 5.2 Hz, 1H, H-1′), 5.61
(d, 1H, D2O exchangeable), 5.26 (d, 1H, D2O exchangeable),
5.09 (t, 1H, D2O exchangeable), 4.24 (dd, 1H), 4.09 (dd, 1H),
3.97 (dd, 1H), 3.63 (m, 2H), 2.61 (s, 3H); 13C NMR (90 MHz,
DMSO-d6) δ 163.50 (C6a), 148.23 (C3), 144.51 (C5), 134.73
(C3a), 89.29, 85.54, 75.27, 69.64, 60.75, 18.14; UV λmax
(methanol) 207 (4860), 246 (10 000), 252 (shoulder) nm. Anal.
Calcd for C10H13N3O4S: C, H, N.
4-â-D-Ar abin ofu r an osyl-5-[(ph en ylm eth yl)th io]im idazo-
[4,5-d ]isoth ia zole (11e). To a solution of 9e (198 mg, 0.30
mmol) in dry dichloromethane (5 mL) at -78 °C was added
boron trichloride (1.0 M in dichloromethane, 3.0 mL) dropwise
over 0.25 h. The solution was warmed to 0 °C via an ice bath
and then stirred for 2 h. The reaction mixture was cooled to
-78 °C, quenched via the dropwise addition of methanol (5
mL), and the vigorously stirred solution was then neutralized
to pH 7 at 0 °C with an ice cold solution of 1 M sodium
hydroxide (ca. 9 mL). The organic layer was removed, and
the aqueous phase was extracted with chloroform (3 × 10 mL)
and then with ethyl acetate (2 × 10 mL). The combined
organics were dried over magnesium sulfate, filtered, concen-
trated, and chromatographed (5% methanol/chloroform, 2 ×
15 cm) to provide a residue which was taken up in methanol,
diluted with water, and lyophilized to afford 32 mg (28%) of
11e hemihydrate as a hygroscopic foam: [R]D -28.4° (c 0.4,
1
methanol); Rf 0.40 (10% methanol/chloroform); H NMR (360
MHz, DMSO-d6) δ 8.50 (s, 1H), 7.45-7.24 (m, 5H), 5.97 (d, J
) 4.1 Hz, 1H, H-1′), 5.61 (d, 1H, D2O exchangeable), 5.57 (d,
1H, D2O exchangeable), 5.15 (t, 1H, D2O exchangeable), 4.57
(q, 2H), 4.02 (m, 2H), 3.82 (m, 1H), 3.70 (m, 2H); 13C NMR (90
MHz, DMSO-d6) δ 161.68 (C6a), 151.76 (C5), 141.72 (C3),
138.63 (C3a), 136.87, 128.93, 128.51, 127.52, 87.05, 84.90,
76.30, 75.75, 60.75, 36.75; UV λmax (methanol) 256 (10 480),
281 (12 560) nm. Anal. Calcd for C16H17N3O4S2‚1/2H2O: C,
H, N.
3-Meth yl-5-(m eth ylth io)-6-(2,3,5-tr i-O-ben zoyl-â-D-r ibo-
fu r a n osyl)im id a zo[4,5-d ]isoth ia zole (13b). 3-Methyl-5-
(methylthio)imidazo[4,5-d]isothiazole (1b, 0.56 g, 3 mmol) was
suspended in dry acetonitrile (15 mL) to which was added N,O-
bis(trimethylsilyl)acetamide (0.78 mL, 3.15 mmol) in one
portion. The suspension was heated to 80 °C for 0.5 h, 1-O-

782 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
acetyl-2,3,5-tri-O-benzoyl-â-D-ribofuranose (12, 1.51 g, 3 mmol)
was added in one portion, and then the mixture was allowed
to cool to 50 °C over 0.5 h. (Trimethylsilyl)trifluoromethane-
sulfonate (0.61 mL, 3.15 mmol) was added dropwise over 5
min, and the solution was stirred for 0.75 h at 50 °C. The
dark reaction mixture was allowed to cool to room temperature
and then poured into ice cold dichloromethane (35 mL) and
saturated aqueous sodium bicarbonate (35 mL). The aqueous
layer was washed with dichloromethane (35 mL), and the
combined organic phases were dried over magnesium sulfate,
filtered and concentrated. The residue was chromatographed
(5 × 15 cm, chloroform to 1% methanol/chloroform), and
fractions containing product were concentrated to yield 1.15
g (61%) of crude 13b as a syrup, which was used without
further purification: Rf 0.60 (2% methanol/chloroform); 1H
NMR (360 MHz, CDCl3) δ 8.12 (d, 2H), 7.99 (d, 2H), 7.96 (d,
2H), 7.48-7.35 (m, 9H), 6.42 (d, J ) 6.5 Hz, 1H, H-1′), 5.93
(dd, 1H), 5.80 (t, 1H), 4.88-4.68 (m, 3H), 2.71 (s, 3H), 2.56 (s,
3H).
2 mmol) was suspended in dry acetonitrile (20 mL), to which
was added N,O-bis(trimethylsilyl)acetamide (0.52 mL, 2.1
mmol) in one portion. The suspension was heated to 80 °C
for 0.5 h, and the resulting solution was cooled to 50 °C. 1-O-
Acetyl-2,3,5-tri-O-benzoyl-â-D-ribofuranose (12, 1.02 g, 2 mmol)
was added in one portion, and then (trimethylsilyl)trifluo-
romethanesulfonate (0.43 mL, 2.2 mmol) was added dropwise
over 5 min. The solution was stirred for 1.5 h at 50 °C, and
the resulting dark red solution was poured into dichlo-
romethane (50 mL) and saturated aqueous sodium bicarbonate
(50 mL). The aqueous layer was washed with dichloromethane
(25 mL), and the combined organic phases were dried over
magnesium sulfate, filtered, and concentrated. The dark red
oil was chromatographed (30% ethyl acetate/hexane, 4 × 20
cm) to yield an oil which crystallized upon the addition of
ethanol. This crude product was recrystallized from ethanol
to yield 0.30 g (21%) of 13d : mp 146.5-147 °C; Rf 0.36 (30%
ethyl acetate/hexane); 1H NMR (300 MHz, CDCl3) δ 8.11 (dd,
2H), 7.94 (m, 4H), 7.59-7.22 (m, 14H), 6.30 (d, 6.6H, H-1′),
5.89 (dd, 1H), 5.73 (t, 1H), 4.84-4.63 (m, 3H), 4.44 (dd, 2H),
2.59 (s, 3H). Anal. Calcd for C38H31N3O7S2: C, H, N.
3-Met h yl-5-(m et h ylt h io)-6-â-D-r ib ofu r a n osylim id a zo-
[4,5-d ]isoth ia zole (15b). A suspension of 13b (0.98 g, 1.5
mmol) in 40 mL of methanolic ammonia was stirred at room
temperature in a pressure tube for 24 h. The solvent was
removed, and the residue was partitioned between water (40
mL) and chloroform (20 mL). The organic layer was removed,
and the aqueous layer was washed with chloroform (3 × 20
mL). The aqueous phase was concentrated in vacuo, and the
residue was recrystallized from methanol to yield 0.30 g of 15b
(61%): mp 119-121 °C; [R]D -34.5° (c 1.0, methanol); Rf 0.31
3-Meth yl-5-[(ph en ylm eth yl)th io]-6-â-D-r ibofu r an osylim -
id a zo[4,5-d ]isoth ia zole (15d ). A suspension of 13d (280 mg,
0.40 mmol) in 10 mL of methanolic ammonia was stirred at
room temperature in a pressure tube for 48 h. The solvent
was removed and the residue chromatographed (5% methanol/
chloroform to 10% methanol/chloroform, 2 × 15 cm) to yield a
glass. This residue was dissolved in a small amount of
methanol, diluted with excess water, and then lyophilized to
yield 0.16 g (98%) of 15d : mp 108.5-110 °C; [R]D -72.6° (c
1
(10% methanol/chloroform); H NMR (300 MHz, DMSO-d6) δ
1
5.70 (d, 1H, D2O exchangeable), 5.62 (d, J ) 6.3 Hz, 1H, H-1′),
5.38 (d, 1H, D2O exchangeable), 5.05 (t, 1H, D2O exchange-
able), 4.18 (dd, 1H), 4.01 (m, 1H), 3.96 (m, 1H), 3.55-3.48 (m,
2H), 2.67 (s, 3H), 2.48 (s, 3H); 13C NMR (90 MHz, DMSO-d6)
δ 152.78 (C5), 152.03 (C3), 147.54 (C3a), 147.19 (C6a), 87.91,
86.02, 73.62, 70.82, 61.70, 16.32, 15.71; UV λmax (methanol)
221 (15 600), 268 (7700) nm. Anal. Calcd for C11H15N3O4S2‚
1/2H2O: C, H, N.
1.2, methanol); Rf 0.34 (10% methanol/chloroform); H NMR
(360 MHz, DMSO-d6) δ 7.44-7.26 (m, 5H), 5.74 (d, J ) 6.84
Hz, 1H, H-1′), 5.61 (d, 1H, D2O exchangeable), 5.37 (d, 1H,
D2O exchangeable), 5.04 (t, 2H, D2O exchangeable), 4.49 (s,
2H), 4.17 (m, 1H), 4.00 (m, 1H), 3.95 (m, 1H), 3.53 (m, 2H),
2.50 (s, 3H); 13C NMR (90 MHz, DMSO-d6) δ 152.19 (C3),
150.84 (C5), 147.54 (C3a), 147.08 (C6a), 136.84, 129.03, 128.51,
127.51, 87.98, 86.04, 73.67, 70.79, 61.67, 37.44, 16.29; UV λmax
(methanol) 221 (21 730), 271 (9660) nm. Anal. Calcd for C17
-
5-(Met h ylt h io)-4-â-D-r ib ofu r a n osylim id a zo[4,5-d ]iso-
th ia zole (16c). 5-(Methylthio)imidazo[4,5-d]isothiazole (1c,
684 mg, 4 mmol) was suspended in dry acetonitrile (40 mL),
to which was added N,O-bis(trimethylsilyl)acetamide (1.04 mL,
4.2 mmol) in one portion. The suspension was heated to 80
°C for 0.5 h, and the resulting solution was cooled to 50 °C.
1-O-Acetyl-2,3,5-tri-O-benzoyl-â-D-ribofuranose (12, 2.02 g, 4
mmol) was added in one portion, and then (trimethylsilyl)-
trifluoromethanesulfonate (0.81 mL, 4.2 mmol) was added
dropwise over 5 min. The solution was stirred for 1 h at 50
°C and then heated to 80 °C for 0.5 h. The resulting dark
solution was poured into ethyl acetate (50 mL) and saturated
aqueous sodium bicarbonate (50 mL). The organic phase was
removed, dried over magnesium sulfate, filtered, and concen-
trated. The dark red oil was chromatographed (30% ethyl
acetate/hexane to 50% ethyl acetate/hexane, 4 × 20 cm), and
fractions containing a product having an Rf 0.30 were pooled
and concentrated to yield a dark oil. The crude material was
stirred in 10 mL of methanolic ammonia at room temperature
in a pressure bottle for 48 h. The dark solution was concen-
trated to a gum and then chromatographed (10% methanol/
chloroform, 2 × 15 cm) to afford an oil which crystallized upon
the addition of water to yield 136 mg (11%) of 16c (mp 167-
172 °C). A portion was recrystallized from methanol (80%
recovery) for analysis: mp 167-172 °C dec; [R]D -42.3° (c 1.2,
H19N3O4S2‚H2O: C, H, N.
5-(Met h ylt h io)-6-â-D-r ib ofu r a n osylim id a zo[4,5-d ]iso-
th ia zole (15c). Sodium hydride (80% w/w in mineral oil, 47
mg, 1.56 mmol) was added in one portion to a solution of 1c
in dry acetonitrile at room temperature. The mixture was
stirred at room temperature for 1 h, and a solution of 2,3,5-
tri-O-benzoyl-â-D-ribofuranosyl bromide (17, 6.25 mmol) in 5
mL of acetonitrile was added dropwise over 5 m. The mixture
was stirred for 2 h at room temperature, and then poured into
water (30 mL) and ethyl acetate (50 mL). The organic phase
was washed with saturated sodium bicarbonate and brine,
dried over magnesium sulfate, filtered, and concentrated. The
residue was chromatographed (chloroform to 1% methanol/
chloroform, 4 × 20 cm) to afford a large amount of sugar
byproducts, followed by nucleoside material. The crude resi-
due was dissolved in methanol (15 mL) containing sodium
methoxide (20 mg) and stirred 3 h at room temperature. The
solution was adjusted to pH 5 with 1 N HCl, concentrated,
and chromatographed (2 × 15 cm, 10% methanol/chloroform),
and the residue was recrystallized from methanol to yield 43
mg (11%) of 15c: mp 151-152 °C; [R]D -28.8° (c 1.4,
1
methanol); Rf 0.18 (10% methanol/chloroform); H NMR (300
MHz, DMSO-d6) δ 8.56 (s, 1H), 5.70 (d, J ) 6.9 Hz, 1H, H-1′),
5.61 (d, 1H, D2O exchangeable), 5.36 (d, 1H, D2O exchange-
able), 5.05 (t, 2H, D2O exchangeable), 4.23 (m, 1H), 4.03-3.97
(m, 2H), 3.55 (m, 2H), 2.67 (s, 3H); 13C NMR (90 MHz, DMSO-
d6) δ 153.69 (C5), 149.20 (C3a), 148.00 (C6a), 143.70 (C3),
87.88, 86.11, 73.58, 70.82, 61.66, 15.48; UV λmax (methanol)
221 (12 130), 273 (6050) nm. Anal. Calcd for C10H13N3O4S2:
C, H, N.
1
methanol); Rf 0.30 (10% methanol/chloroform); H NMR (360
MHz, DMSO-d6) δ 8.72 (s, 1H), 5.65 (d, J ) 7.3 Hz, 1H, H-1′),
5.54 (d, 1H, D2O exchangeable), 5.30 (d, 1H, D2O exchange-
able), 5.18 (t, 2H, D2O exchangeable), 4.37 (dd, 1H), 4.09 (m,
1H), 3.98 (m, 1H), 3.62 (m, 2H), 2.72 (s, 3H); 13C NMR (90
MHz, DMSO-d6) δ 163.29, (C6a) 155.92 (C5), 140.20 (C3),
136.66 (C3a), 88.18, 86.28, 73.61, 70.41, 61.37, 15.25; UV λmax
(methanol) 222 (4430), 255 (10 550), 279 (12 640) nm. Anal.
Calcd for C10H13N3O4S2: C, H, N.
In Vitr o An tip r olifer a tive Stu d ies. The in vitro cyto-
toxicity against L1210 was determined in an L1210 cell growth
assay described previously.37 L1210 murine leukemic cells
(kindly provided by J . R. Bertino, Memorial Sloan Kettering
Center, New York, NY 10021) were grown in Fischer’s medium
supplemented with 10% heat inactivated (56 °C, 30 min) horse
3-Meth yl-5-[(p h en ylm eth yl)th io]-6-(2,3,5-tr i-O-ben zoyl-
â-D-r ibofu r an osyl)im idazo[4,5-d]isoth iazole (13d). 3-Meth-
yl-5-[(phenylmethyl)thio]imidazo[4,5-d]isothiazole (1d , 527 mg,

Imidazo[4,5-d]isothiazole Nucleosides
J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5 783
serum and were subcultured by serial dilution. Growth rates
in the presence of selected concentrations of the test compound
were calculated from determinations of the number of cells at
0, 24, 48, 72, and 96 h. The growth rate was calculated as
the slope of a semilogarithmic plot of these cell numbers
against time for the treated culture as a percent of the slope
obtained for control cultures. The IC50 was defined as the
concentration required to decrease the growth rate to 50% of
that of the untreated control cells (population doubling time
) 12 h).
The G-361 human melanoma, HT-29 human colon carci-
noma, and ZR-75-1 human breast carcinoma cell lines (all from
the ATCC, Rockville, MD 20852-1776) were grown in RPMI
1640 medium supplemented with 5% fetal bovine serum and
were subcultured using 0.5 g trypsin, 0.2g EDTA/L. Inhibition
of cell proliferation was assayed as described previously.38 The
cells were plated at the following densities, compounds were
added the following day, and growth was assayed on the
indicated number of days after plating: G-361, 1000 cells/well,
6 days; HT-29, 2000 cells/well, 5 days; ZR-75-1, 5000 cells/
well, 6 days. The results were expressed as percent of control
optical density (sulforhodamine B) on the assay day.
Cell Cu lt u r e P r oced u r es for An t ivir a l St u d ies. The
routine growth and passage of KB, BSC-1, and HFF cells was
performed in monolayer cultures using minimal essential
medium (MEM) with either Hanks salts [MEM(H)] or Earle
salts [MEM(E)] supplemented with 10% calf serum or 10%
fetal bovine serum (HFF cells). The sodium bicarbonate
concentration was varied to meet the buffering capacity
required. Cells were passaged at 1:2 to 1:10 dilutions accord-
ing to conventional procedures by using 0.05% trypsin plus
0.02% EDTA in a HEPES-buffered salt solution.
Vir ologica l P r oced u r es. The Towne strain, plaque-puri-
fied isolate Po, of HCMV was kindly provided by Dr. Mark
Stinski, University of Iowa. The KOS strain of HSV-1 was
used in most experiments and was provided by Dr. Sandra K.
Weller, University of Connecticut. Stock HCMV was prepared
by infecting HFF cells at a multiplicity of infection (moi) of
<0.01 plaque-forming units (pfu) per cell as detailed previ-
ously.39 High-titer HSV-1 stocks were prepared by infecting
KB cells at an moi of <0.1 also as detailed previously.39 Virus
titers were determined using monolayer cultures of HFF cells
for HCMV and monolayer cultures of BSC-1 cells for HSV-1
as described earlier.40 Briefly, HFF or BSC-1 cells were
planted as described above in 96-well cluster dishes and
incubated overnight at 37 °C. The next day cultures were
inoculated with 0.2 mL of HCMV or HSV-1 stock suspension
and serially diluted 1:3 across the remaining 11 columns of
the 96-well plate. After virus adsorption the inoculum was
replaced with fresh medium, and cultures were incubated for
7 days for HCMV, 2 or 3 days for HSV-1. Plaques were
enumerated under 20-fold magnification in wells having the
dilution which gave 5-20 plaques per well. Virus titers were
calculated according to the following formula: Titer (pfu/mL)
) number of plaques × 5 × 3n; where n represents the nth
dilution of the virus used to infect the well in which plaques
were enumerated.
HCMV P la qu e Red u ction Assa y. HFF cells in 24-well
cluster dishes were infected with approximately 100 pfu of
HCMV per cm2 cell sheet using the procedures detailed
above.39 Following virus adsorption, compounds dissolved in
growth medium were added to duplicate wells in four to eight
selected concentrations. After incubation at 37 °C for 7 days,
cell sheets were fixed and stained with crystal violet, and
microscopic plaques were enumerated as described above.
Drug effects were calculated as a percentage of reduction in
number of plaques in the presence of each drug concentration
compared to the number observed in the absence of drug.
and rinsed, and horseradish peroxidase conjugated rabbit anti-
HSV-1 antibody was added. Following removal of the antibody
containing solution, plates were rinsed and then developed by
adding 150 µL per well of a solution of tetramethylbenzidine
as substrate. The reaction was stopped with H2SO4, and
absorbance was read at 450 and 570 nm. Drug effects were
calculated as a percentage of the reduction in absorbance in
the presence of each drug concentration compared to absor-
bance obtained with virus in the absence of drug.
Cytotoxicity Assa ys in An tivir a l Stu d ies. Two different
assays were used to explore cytotoxicity of selected compounds
using methods we have detailed previously. (i) Cytotoxicity
produced in stationary HFF cells was determined by micro-
scopic inspection of cells not affected by the virus used in
plaque assays.38 (ii) The effect of compounds during two
population doublings of KB cells was determined by crystal
violet staining and spectrophotometric quantitation of dye
eluted from stained cells as described earlier.42 Briefly, 96-
well cluster dishes were planted with KB cells at 3000-5000
cells per well. After overnight incubation at 37 °C, test
compounds were added in quadruplicate at six to eight
concentrations. Plates were incubated at 37 °C for 48 h in a
CO2 incubator, rinsed, fixed with 95% ethanol, and stained
with 0.1% crystal violet. Acidified ethanol was added, and
absorbance of solutions in individual wells was read at 570
nm in a spectrophotometer designed to read 96-well plates.
Da ta An a lysis. Dose-response relationships were con-
structed by linearly regressing the percent inhibition of
parameters derived in the preceding sections against log drug
concentrations. Fifty-percent inhibitory (IC50) concentrations
were calculated from the regression lines. Samples containing
positive controls (acyclovir for HSV-1, ganciclovir for HCMV,
and 2-acetylpyridine thiosemicarbazone for cytotoxicity) were
used in all assays.
Ack n ow led gm en t. We thank J ulie M. Breitenbach,
J ames S. Lee, and Roger G. Ptak for expert performance
of the antiviral and cytotoxicity assays and Ms. Marina
Savic for the preparation of the manuscript. These
studies were supported by Department of Health and
Human Services Research Contract NO1-AI72641 and
Grant UO1-AI31718 from the National Institute of
Allergy and Infectious Diseases, and by American
Cancer Society Research Grant No. DHP-36.
Refer en ces
(1) Brown, G. B.; Weliky, V. S. The Synthesis of 9-â-D-Ribofurano-
sylpurine and the Identity of Nebularine. J . Biol. Chem. 1953,
204, 1019-1024.
(2) Giner-Sorolla, A.; Medrek, L.; Bendich, A. Synthesis and Biologi-
cal Activity of 9-â-D-Ribofuranosyl-6-hydroxylaminopurine. J .
Med. Chem. 1966, 9, 143-144.
(3) Elion, G. B.; Burgi, E.; Hitchings, G. H. Studies on Condensed
Pyrimidine Systems. IX. The Synthesis of Some 6-Substituted
Purines. J . Am. Chem. Soc. 1952, 74, 411-414.
(4) Schabel, F. M., J r.; Montgomery, J . A.; Skipper, H. E.; Laster,
W. R., J r.; Thomson, J . R. Experimental Evaluation of Potential
Anticancer Agents I. Quantitative Therapeutic Evaluation of
Certain Purine Analogs. Cancer Res. 1961, 21, 690-699.
(5) Montgomery, J . A.; Schabel, F. M., J r.; Skipper, H. E. Experi-
mental Evaluation of Potential Anticancer Agents IX. The
Ribonucleosides and Ribonucleotides of Two Purine Antagonists.
Cancer Res. 1962, 22, 504-509.
(6) Bennett, L. L., J r.; Vail, M. H.; Chumley, S.; Montgomery, J . A.
Activity of Adenosine Analogs Against
a Cell Culture Line
Resistant to 2-Fluoroadenine. Biochem. Pharmacol. 1966, 15,
1719-1728.
(7) Montgomery, J . A.; Clayton, S. J .; Fitzgibbon, W. E., J r. The
Ribonucleosides of 4-Aminopyrazolo[3,4-d]pyrimidine. J . Het-
erocycl. Chem. 1964, 1, 215-216.
HSV-1 ELISA. An ELISA developed by us41 was employed
to detect HSV-1. Ninety-six-well cluster dishes were planted
with 10 000 BSC-1 cells per well in 200 µL per well of MEM-
(E) plus 10% calf serum. After overnight incubation at 37 °C,
selected drug concentrations in quadruplicate and HSV-1 at
a concentration of 100 pfu/well were added. Following a 3-day
incubation at 37 °C, medium was removed, plates were blocked
(8) Mizuno, Y.; Ikehara, M.; Watanabe, K. A.; Suzaki, S.; Itoh, T.
Synthetic Studies of Potential Antimetabolites. IX. The Anomeric
Configuration of Tubercidin. J . Org. Chem. 1963, 28, 3329-3331.
(9) Tolman, R. L.; Robins, R. K.; Townsend, L. B. Pyrrolopyrimidine
Nucleosides III. The Total Synthesis of Toyocamycin, Sangiva-
mycin, Tubercidin, and Related Derivatives. J . Am. Chem. Soc.
1969, 91, 2102-2108.

784 J ournal of Medicinal Chemistry, 1997, Vol. 40, No. 5
Swayze et al.
(10) Townsend, L. B.; Drach, J . C.; Wotring, L. L.; Vittori, S.; Pudlo,
J . S.; Swayze, E. E.; Gupta, P.; Maruyama, T.; Saxena, N.;
Coleman, L. A.; Westerman, A. C.; Spurr, J .; Nassiri, M. R.;
Turk, S. R.; Krawczyk, S. H. Design, Synthesis, and Studies on
the Structure Activity Relationships of Certain Pyrrolo[2,3-d]-
pyrimidine Nucleosides and Structurally Related Analogs as
Potential Antineoplastic and Antiviral Agents. Il Farmaco 1991,
46, 113-139.
(11) Namikoshi, M.; Carmichael, W. W.; Sakai, R.; J ares-Erijman,
E. A.; Kaup, A. M.; Rinehart, K. L. 9-Deazaadenosine and Its
5′-R-D-Glucopyranoside Isolated from the Cyanobacterium ana-
baena affinis Strain VS-1. J . Am. Chem. Soc. 1993, 115, 2504-
2505.
(12) Koyama, G.; Maeda, K.; Umezawa, H.; Iitaka, Y. The Structural
Studies of Formycin and Formycin B. Tetrahedron Lett. 1966,
597-602.
(13) Robins, R. K.; Townsend, L. B.; Cassidy, F.; Gerster, J . F.; Lewis,
A. F.; Miller, R. L. Structure of the Nucleoside Antibiotics
Formycin, Formycin B and Laurusin (1). J . Heterocycl. Chem.
1966, 3, 110-114.
(14) Shimada, N; Yagisawa, N.; Naganawa, H.; Takita, T.; Hamada,
M.; Takeuchi, T.; Umezawa, H. Oxanosine, a Novel Nucleoside
From Actinomycetes. J . Antibiot. 1981, 34, 1216-1218.
(15) Parkanyi, C. Recent Progress in the Quantum Chemistry of
Organic Sulfur Compounds. Mech. React. Sulfur Compd. 1969,
4, 69-94.
(16) Chwang, T. L.; Nemec, J .; Welch, A. D. Synthesis and Charac-
terization of 5-Amino-2-(â-D-ribofuranosyl)-1,2,4-thiadiazoline-
3-one, a Sulfur-Containing Analog of Cytidine. J . Carbohydr.
Nucleosides Nucleotides 1980, 7, 159-166.
(17) Gordon, M. P.; Weliky, V. S.; Brown, G. B. A Study of the Action
of Acid and Alkali on Certain Purines and Purine Nucleosides.
J . Am. Chem. Soc. 1957, 79, 3245-3250.
(27) Townsend, L. B. Nuclear Magnetic Resonance Spectroscopy in
the Study of Nucleic Acid Components and Certain Related
Derivatives. In Synthetic Procedures in Nucleic Acid Chemistry;
Zorbach, W. W., Tipson, R. S., Eds.; J ohn Wiley & Sons: New
York, 1972; Vol. 2, pp 323-340.
(28) Bhattacharya, A. K.; Ness, R. K.; Fletcher, H. G., J r. 2-Deoxy-
D-erythro-pentose. IX. Some Relationships Among the Rotations
of Acylated Aldopentoses, Aldopentosyl Halides, and Anhydro-
pentitols. J . Org. Chem. 1963, 28, 428-435.
(29) Ramasamy, K.; Imamura, N.; Robins, R. K.; Revankar, G. R. A
Facile and Improved Synthesis of Tubercidin and Certain
Related Pyrrolo[2,3-d]pyrimidine Nucleosides by the Stereospe-
cific Sodium Salt Glycosylation Procedure. J . Heterocycl. Chem.
1988, 25, 1893-1898.
(30) Ramasamy, K.; Robins, R. K.; Revankar, G. R. Direct Synthesis
of Pyrrole Nucleosides by the Stereospecific Sodium Salt Gly-
cosylation Procedure. J . Heterocycl. Chem. 1987, 24, 863-868.
(31) Glaudemans, C. P. J .; Fletcher, H. G., J r. Syntheses with
Partially Benzylated Sugars. III. A Simple Pathway to a “cis-
Nucleoside,” 9-â-D-Arabinofuranosyladenine (Spongoadenosine).
J . Org. Chem. 1963, 28, 3004-3006.
(32) Pudlo, J . S.; Townsend, L. B. A New and Novel Approach
Towards the Synthesis of 3′-Deoxy-3′-hydroxymethyl Ribofura-
nosides. Tetrahedron Lett. 1990, 22, 3101-3104.
(33) Vorbruggen, H.; Krolikiewicz, K.; Bennua, B. Nucleoside Syn-
thesis with Trimethylsilyl Triflate and Perchlorate as Catalysts.
Chem. Ber. 1981, 114, 1234-1255.
(34) Stevens, J . D.; Ness, R. K.; Fletcher, H. G., J r. Syntheses with
Partially Benzylated Sugars. XI. Studies on the Synthesis of the
Anomeric 5,6-Dimethyl-1-D-ribofuranosylbenzimidazoles (Riba-
zoles). Comparison of the Condensation of 2,3,5-Tri-O-benzoyl-
D-ribofuranosyl Bromide and 2,3,5-Tri-O-benzyl-D-ribofuranosyl
Chloride with 5,6-Dimethylbenzimidazole. J . Org. Chem. 1968,
33, 1806-1810.
(35) Tapiero, C.; Imbach, J .-L. The Determination of Anomeric
Configuration of D-Ribofuranosyl Nucleosides by Proton Mag-
netic Resonance Spectroscopy. In Nucleic Acid Chemistry.
Improved and New Synthetic Procedures, Methods, and Tech-
niques; Townsend, L. B., Tipson, R. S., Eds.; J ohn Wiley &
Sons: New York, 1978; Part 2, pp 1055-1059.
(18) Montgomery, J . A.; Hewson, K. Analogs of 6-Methyl-9-â-D-
ribofuranosylpurine. J . Med. Chem. 1968, 11, 48-52.
(19) (a) Swayze, E. E. The Synthesis of Several Members of the New
Imidazo[4,5-d]isothiazole Ring System: Novel 5:5 Fused Analogs
of Purines and Purine Nucleosides. Ph.D. Thesis, 1994, The
University of Michigan, Ann Arbor, MI. (b) Swayze, E. E.;
Townsend, L. B. The Synthesis of Substituted Imidazo[4,5-d]-
isothiazoles via the Ring Annulation of Isothiazole Diamines:
An Investigation of the Chemical, Physical, and Biological
Properties of Several Novel 5:5 Fused Analogs of the Purine Ring
System. J . Org. Chem. 1995, 60, 6309-6317.
(36) Baker, B. R. In Ciba Foundation Symposium on the Chemistry
and Biology of Purines; Wolstenholme, G. E., O’Connor, C. M.,
Eds.; Little, Brown and Co.: Boston, 1957; p 120.
(37) Wotring, L. L.; Townsend, L. B. Study of the Cytotoxicity and
Metabolism of 4-amino-3-carboxamido-1-(â-D-ribofuranosyl)pyra-
zolo[3,4-d]pyrimidine Using Inhibitors of Adenosine Kinase and
Adenosine Deaminase. Cancer Res. 1979, 39, 3018-3023.
(38) Meade, E. A.; Wotring, L. L.; Drach, J . C.; Townsend, L. B.
Synthesis, Antiproliferative, and Antiviral Activity of 4-Amino-
1-(â-D-ribofuranosyl)pyrrolo[2,3-d]pyridazin-7(6H)-one and Re-
lated Derivatives. J . Med. Chem. 1993, 36, 3834-3842.
(39) Turk, S. R.; Shipman, C., J r.; Nassiri, M. R.; Genzingler, G.;
Krawczyk, S. H.; Townsend, L. B.; Drach, J . C. Pyrrolo[2,3-d]-
pyrimidine Nucleosides as Inhibitors of Human Cytomegalovi-
rus. Antimicrob. Agents Chemother. 1987, 31, 544-550.
(40) Prichard, M. N.; Turk, S. R.; Coleman, L. A.; Engelhardt, S. L.;
Shipman, C., J r.; Drach, J . C. A Microtiter Virus Yield Reduction
Assay for the Evaluation of Antiviral Compounds Against
Human Cytomegalovirus and Herpes Simplex Virus, J . Virol.
Methods 1990, 28, 101-106.
(41) Prichard, M. N.; Shipman, C., J r. A Three-dimensional Model
to Analyze Drug-drug Interactions. Antiviral Res. 1990, 14, 181-
206.
(42) Prichard, M. N.; Prichard, L. E.; Baguley, W. A.; Nassiri, M. R.;
Shipman, C., J r. Three-dimensional Analysis of the Synergistic
Cytotoxicity of Ganciclovir and Zidovudine. Antimicrob. Agents
Chemother. 1991, 35, 1060-1065.
(20) Srivastava, P. C.; Robins, R. K.; Meyer, R. B. Synthesis and
Properties of Purine Nucleosides and Nucleotides. In Chemistry
of Nucleosides and Nucleotides; Townsend, L. B., Ed.; Plenum
Press: New York, 1988; Vol. 1, pp 113-281.
(21) Revankar, G. R.; Robins, R. K. The Synthesis and Chemistry of
Heterocyclic Analogues of Purine Nucleosides and Nucleotides.
In Chemistry of Nucleosides and Nucleotides; Townsend, L. B.,
Ed.; Plenum Press: New York, 1991; Vol. 2, pp 161-398.
(22) Swayze, E. E.; Townsend, L. B. The Synthesis of Several
Imidazo[4,5-d]isothiazoles; Derivatives of a New Ring System.
206th American Chemical Society National Meeting (Org #351),
August 22-27, 1993, Chicago, IL.
(23) Swayze, E. E.; Townsend, L. B. The Synthesis of Several
Imidazo[4,5-d]isothiazoles; Derivatives of a New Ring System.
J . Heterocycl. Chem. 1993, 30, 1379-1383.
(24) Kazimierczuk, Z.; Cottam, H. B.; Revankar, G. R.; Robins, R. K.
Synthesis of 2′-Deoxyadenosine, and Related 2′-Deoxynucleo-
sides via a Novel Direct Stereospecific Sodium Salt Glycosylation
Procedure. J . Am. Chem. Soc. 1984, 106, 6379-6382.
(25) Hoffer, M. R-Thymidin. Chem. Ber. 1960, 93, 2777-2781.
(26) Chenon, M.-T.; Pugmire, R. J .; Grant, D. M.; Panzica, R. P.;
Townsend, L. B. Carbon-13 Magnetic Resonance. XXV. A Basic
Set of Parameters for the Investigation of Tautomerism in
Purines Established from Carbon-13 Magnetic Resonance Stud-
ies Using Certain Purines and Pyrrolo[2,3-d]pyrimidines. J . Am.
Chem. Soc. 1975, 97, 4627-4636.
J M960605Z