"Fleximers". Design and synthesis of a new class of novel shape-modified nucleosides(1).

Article abstract of DOI:10.1021/jo0255476

A new class of shape-modified nucleosides is introduced. These novel "fleximers" feature the purine ring systems of adenosine, inosine, and guanosine split into their individual imidazole and pyrimidine components (as in 1-3). This structural modification serves to introduce flexibility into the nucleoside while still retaining the elements essential for recognition. As a consequence, these novel fleximers should find use as bioprobes for investigating enzyme-coenzyme binding sites as well as nucleic acid and protein interactions. Their design and synthesis are described.

Full text of DOI:10.1021/jo0255476

J . Org. Chem. 2002, 67, 3365-3373
3365
“F lexim er s”. Design a n d Syn th esis of a New Cla ss of Novel
Sh a p e-Mod ified Nu cleosid es1
Katherine L. Seley,* Liang Zhang, Asmerom Hagos, and Stephen Quirk†
School of Chemistry and Biochemistry, Georgia Institute of Technology, Atlanta, Georgia 30332-0400
katherine.seley@chemistry.gatech.edu
Received J anuary 22, 2002
A new class of shape-modified nucleosides is introduced. These novel “fleximers” feature the purine
ring systems of adenosine, inosine, and guanosine split into their individual imidazole and
pyrimidine components (as in 1-3). This structural modification serves to introduce flexibility into
the nucleoside while still retaining the elements essential for recognition. As a consequence, these
novel fleximers should find use as bioprobes for investigating enzyme-coenzyme binding sites as
well as nucleic acid and protein interactions. Their design and synthesis are described.
In tr od u ction
Understanding the structural interactions between
enzymes and their substrates or cofactors is vital to the
design of more effective medicinal agents. Purines are
one of the most ubiquitous heterocyclic ring systems
found in nature; they are components in numerous
biologically significant molecules and thus present an
excellent scaffold upon which to construct bioprobes.
F igu r e 1. Leonard’s benzo-separated analogues.
Pioneering work in this area began with Leonard’s lin-,
dist-, and prox-benzo-separated adenosine analogues
shown in Figure 1.2-4
phenomenon causes inaccuracies when employing the
structure-based drug design method for a potential
enzyme-substrate complex using crystallographic data;
although the fit may appear to be good, in reality, the
“best guess” drug often proves to be a poor inhibitor.13,14
As a first step toward a potential solution to this
dilemma, we set out to design a flexible substrate to pair
with a flexible binding site. Since the sugar moieties of
nucleosides inherently possess the ability to adapt their
conformations, we instead chose to focus our attention
on the heterobases. Separation of the components of the
purine ring while maintaining some modicum of con-
nectivity would allow the purine ring to become “flexible”,
thereby allowing it to adapt to spatial confines of the
binding site more readily. In addition, it was essential
that the nucleoside still possess the molecular elements
needed for recognition; in essence, the nucleobase moiety
could be “split” but still retain the maximum number of
its key structural features. Then, by allowing both the
enzyme and the substrate to adjust to each other, the
ideal fit would more likely be achieved and maintained,
thereby providing a unique perspective on enzyme-
substrate binding conformations. This information could
then be used to design more effective inhibitors.
One focus for our research has involved the design and
synthesis of novel shape-modified nucleosides to explore
fundamental aspects of nucleic acid structure, function,
and stability, as well as to investigate enzyme binding
site parameters. With the increasing availability of
crystal structures for various enzyme-substrate com-
plexes, it has become apparent that many binding sites
are more flexible than previously thought and can
therefore adjust to fit a wide range of substrates.5-12 This
* To whom correspondence should be addressed. Phone: 404-894-
4013. Fax: 404-894-2295.
† Present Address: Kimberly-Clark Corporation, 1400 Holcomb
Bridge Rd., Roswell, GA 30076-2199.
(1) A preliminary report of this work was recently published: Seley,
K. L.; Zhang, L.; Hagos, A. Org. Lett. 2001, 3, 3209.
(2) Leonard, N. J .; Morrice, A. G.; Sprecker, M. A. J . Org. Chem.
1975, 40, 356.
(3) Leonard, N. J . Acct. Chem. Res. 1982, 15, 128.
(4) Leonard, N. J .; Hiremath, S. P. Tetrahedron 1986, 42, 1917.
(5) Hu, Y.; Yang, X.; Yin, D. H.; Mahadevan, J .; Kuczera, K.;
Schowen, R. L.; Borchardt, R. T. Biochem. 2001, 40, 15143.
(6) Hu, Y.; Komoto, J .; Huang, Y.; Gomi, T.; Ogawa, H.; Takata, Y.;
Fujioka, M.; Takusagawa, F. Biochem. 1999, 38, 8323.
(7) Yin, D.; Yang, X.; Hu, Y.; Kuczera, K.; Schowen, R. L.; Borchardt,
R. T.; Squier, T. C. Biochem. 2000, 39, 9811.
(8) Bugg, C. E.; Carson, W. M.; Montgomery, J . A. Sci. Am. 1993,
269, 92.
(9) Ealick, S. E.; Babu, Y. S.; Bugg, C. E.; Erion, M. D.; Guida, W.
C.; Montgomery, J . A.; Secrist, J . A., III. Proc. Natl. Acad. Sci., USA
1991, 88, 11540.
(10) Ealick, S. E.; Rule, S. A.; Carter, D. C.; Greenhough, T. J .; Babu,
Y. S.; Cook, W. J .; Habash, J .; Helliwell, J . R.; Stoeckler, J . D.; Parks,
R. E., J r.; Chen, S.-f.; Bugg, C. E. J . Biol. Chem. 1990, 265, 1812.
(11) Erion, M. D.; Stoeckler, J . D.; Guida, W. C.; Walter, R. L.;
Ealick, S. E. Biochemistry 1997, 36, 11735.
(12) Erion, M. D.; Takabayashi, K.; Smith, H. B.; Kessi, J .; Wagner,
S.; Ho¨nger, S.; Shames, S. L.; Ealick, S. E. Biochemistry 1997, 36,
11725.
These novel fleximers feature a purine ring split into
the individual components; the imidazole and pyrimidine
rings of adenosine, inosine, and guanosine (as in 1-3,
respectively, Figure 2) are connected by a single carbon-
carbon bond, thereby creating a separation of 1.50 Å
(13) Montgomery, J . A. Med. Res. Rev. 1993, 13, 209.
(14) Montgomery, J . A.; Niwas, S.; Rose, J . D.; Secrist, J . A., III;
Babu, Y. S.; Bugg, C. E.; Erion, M. D.; Guida, W. C.; Ealick, S. E. J .
Med. Chem. 1993, 36, 55.
10.1021/jo0255476 CCC: $22.00 © 2002 American Chemical Society
Published on Web 04/18/2002

3366 J . Org. Chem., Vol. 67, No. 10, 2002
Seley et al.
F igu r e 2. Fleximer targets.
Ta ble 1. Va r iou s P r op er ties of th e
F igu r e 3. Optimized structure of the adenosine fleximer.
Ha r tr ee-F ock -Op tim ized F lexim er Str u ctu r esa
fleximer
1
2
3
fleximer nucleosides all adopt a high anti conformation
with a corresponding dihedral angle in the range of 225°,
and the ribose, in all instances, maintained a C2′ endo
conformation. There are no significant structural or
energetic differences between the final optimized fleximer
structures; in all 48 final structures, the pyrimidine is
rotated out of the plane defined by the imidazole ring.
This rotation dips the N1 portion of the pyrimidine ring
below the imidazole plane. The inosine fleximer shows
an average of -26.6° for the rotation, which is greater
than either the guanosine (-19.1°) or the adenosine
(-12.9°) fleximers. The observation that, at the 6-311G*
level of theory, all three purine fleximers adopt a non-
planar base conformation in the gas phase is encourag-
ing.
Next, to gauge the level of conformational flexibility
in the context of an enzyme active site, the adenosine
fleximer was chosen for further modeling. Automated
molecular docking reveals that the molecule occupies the
S-adenosylhomocysteine hydrolase active site in an ori-
entation that is similar, but not identical, to the position
of the known nucleoside inhibitor in the crystal structure
published by Turner et al.17 The RMS deviation between
the two molecules is 1.1 Å. Most of this difference is
accounted for by a 0.4 Å translation of the adenosine
fleximer toward His 353 and a 12° rotation of the entire
molecule toward Asp 190. Most notably is the observation
that the docked adenosine fleximer pyrimidine ring is
rotated to be nearly perpendicular to the imidazole ring
(87°). Figure 4 shows the conformation the molecule
preferred when docked in the enzyme active site. Al-
though the gas-phase-optimized structure reveals a low-
energy conformation with that angle at -12.9°, the
fleximer can sample a high degree of conformational
space and adopt alternate conformations in order to
maximize intermolecular contacts. The docking search
resulted in two main clusters; the primary group (-60
kcal/mol of intermolecular energy) accounted for 80% of
the solution orientations. The second group (-40 kcal/
mol of intermolecular energy) showed an orientation
closer to the inhibitor in the crystal structure. Full details
of the molecular docking and binding characteristics of
the fleximers will be given elsewhere.
total energy (AU)
dipole (Debye)
dihedral angle (deg)
C2′-C1′-N1-C2
dihedral angle (deg)
N1-C5-C6′-N1′
volume (Å3)
-1035.5 ( 22 -1110.5 ( 37 -1055.4 ( 28
4.9
11.7
9.7
223.2 ( 3.4
225.6 ( 2.1
228.7 ( 3.3
-12.9 ( 4.1
-19.1 ( 6.6
-26.6 ( 5.4
305.9
0.009
317.1
0.004
0.013
302.9
0.007
0.017
bond (Å) rms dev
angles (deg) rms dev 0.012
a
A 6-311G* basis set was utilized for the optimizations. Values
are the average ((SD) of the 16 optimized conformations.
between the rings. Analogous to Leonard’s dist-benzo
analogues, these fleximers are connected at the C-5 of
the imidazole and the C-6 of the pyrimidine. To our
knowledge, there is only one other example of a split
nucleoside in the literature;15 however, it differs signifi-
cantly in its structure. There is also a very recent report
in the literature16 of a biphenyl nucleoside; however, we
believe our novel fleximers more closely resemble the
naturally occurring purine nucleosides, and that as such,
they will provide a unique perspective on enzyme/
substrate interactions.
Resu lts a n d Discu ssion
We initially carried out a series of ab initio quantum
mechanical structure optimization calculations. In the
case of all three fleximers, sixteen initial structures
converged to a similar final conformation. Three-dimen-
sional least squares analysis was employed to fully
analyze the structural deviations of the sixteen optimized
structures for each fleximer. Some structural and ener-
getic properties of the optimized fleximer nucleosides are
shown in Table 1. Bond lengths and angles are similar
among the optimized structures as is evidenced by the
low root-mean-squared deviations. In all cases, there was
little effect of the initial pyrimidine/imidazole rotation
or of the initial syn/anti conformation on the final
optimized structure.
All three of the fleximers proved to be quite similar in
their optimized structures, and Figure 3 shows the final
conformation for the adenosine fleximer 1. The structures
were optimized using RHF SCF gradient optimization at
the 6-311G* level of theory from an initial model. The
Next, construction of the fleximers was undertaken.
The initial approach was envisioned from a Suzuki or
Stille coupling to attach the imidazole and pyrimidine
(15) Van Calenbergh, S.; De Bruyn, A.; Schrami, J .; Blaton, N.;
Peeters, O.; De Keukeleire, D.; Busson, R.; Herdewijn, P. Nucleosides
Nucleotides 1997, 16, 291.
(16) Singh, I.; Hecker, W.; Prasad, A. K.; Parmar, V. S.; Seitz, O.
Chem. Commun. 2002, 500.
(17) Turner, M. A.; Yuan, C.-S.; Borchardt, R. T.; Hershfield, M. S.;
Smith, D. G.; Howell, P. L. Nat. Struct. Biol. 1998, 5, 369.

“Fleximers”: Design and Synthesis
J . Org. Chem., Vol. 67, No. 10, 2002 3367
Sch em e 2
solubility, so we reverted to coupling the thiomethyl
intermediate 7 (as shown in Scheme 2), something
Leonard19 and others20-23 have found necessary in the
synthesis of other lin-benzoadenosine analogues.
Use of a coupling method24 that employs bis(trimeth-
ylsilyl)acetamide (BSA) and trimethylsilyl triflate suc-
cessfully coupled 7 to tetraacetate 5. We expected two
products, 8 and 9, would be formed, resulting from N-1
and N-3 coupling, since both nitrogens of the heterobase
are available for glycolization. This is a common occur-
rence in nucleoside synthesis, but to our surprise, only
one product was formed. Upon structural characteriza-
tion, the product was shown by two-dimensional NMR
to be 9, the N-1 coupled product, rather than the desired
N-3 product (8), so our focus then turned to coupling at
an earlier step in the construction of the base.
F igu r e 4. Conformation of the adenosine fleximer adopted
in the S-adenosylhomocysteine hydrolase active site deter-
mined by automated molecular docking. Shown is the best-fit
conformation, corresponding to -60 kcal/mol in the active site.
Sch em e 1
Several of the heterobase intermediates were investi-
gated, but dibromoimidazole proved to be advantageous
for two reasons: one, 4,5-dibromoimidazole is sym-
metrical, and therefore coupling at either nitrogen would
result in only one product, thereby foregoing the need
for tedious structure proof, and two, it would circumvent
the difficulties we had previously encountered during
deprotection of the imidazole nitrogen during the syn-
thesis of the bases. Coupling was successful, and the
synthesis of the nucleoside resumed with a series of
reactions analogous to that used previously for the
construction of the bases.18
As shown in Scheme 3, coupling of 4,5-dibromoimida-
zole (10) again using BSA and trimethylsilyl triflate
produced 11.24 Removal of the acetate groups was then
accomplished quantitatively with methanolic ammonia
overnight to afford the triol 12.
Reprotection of the hydroxyl groups was then under-
taken. The correct choice of protecting group was critical,
since it would be required to withstand both acidic and
basic conditions, as well as attack from Grignard re-
agents. We initially chose to protect the hydroxyls as
methyl ethers, since methoxy groups are robust to the
various conditions we were facing but not sterically
hindering for adjacent hydroxyls. Using standard me-
thylation procedures, triol 12 was then easily converted
to 13a . Formylation of the C-5 position was then carried
out using Grignard conditions to provide 14a . As we have
previously noted, this reaction gives primarily the desired
C-5 aldehyde, along with a C-4 monobrominated side
product, which is readily separated from the aldehyde.
Despite numerous attempts at optimization of this step,
rings, an approach we chose for the synthesis of the prox-
analogues (manuscript in preparation). Unfortunately,
the requisite pyrimidine rings substituted in the neces-
sary position for coupling were unavailable commercially
or tedious to obtain synthetically. We then turned our
attention toward construction of a tricyclic ring system
that could be preferentially cleaved at only one of the
connecting bonds.
Although the targets resembled Leonard’s dist-benzo
analogues in their attached ring motif, the benzene
spacer ring was, of course, unsuitable for our purposes
since it did not possess a single carbon-carbon bond
connection. It quickly became apparent, however, that
insertion of a thiophene spacer ring offered an exceptional
alternative. The carbon-carbon single-bond connection
of the thiophene spacer ring was as we desired, and
removal of the sulfur would then result in the partial
cleavage of the nucleobase while keeping the two com-
ponents connected. Admittedly, there was some concern
that the extended aromatic system of the tricyclic system
would be resistant to cleavage since loss of aromaticity
would occur, but we felt confident that treatment with
Raney nickel would be successful.
As depicted retrosynthetically in Scheme 1, we initially
intended to follow Leonard’s approach by coupling the
preconstructed heterobase (4) to the commercially avail-
able tetraacetate-protected sugar (5), which, following
deprotection, would provide us with the desired tricyclic
nucleoside intermediate 6. Subsequent removal of the
sulfur would yield the fleximer nucleoside. As we have
previously reported,18 the synthesis of 4 and the corre-
sponding tricyclic guanine heterobase was successful, but
unfortunately, coupling proved to be problematic. Initial
attempts at coupling 4 failed due to problems with
(19) Leonard, N. J .; Laursen, R. A. Biochem. 1965, 4, 354.
(20) Zhang, J .; Nair, V. Nucleosides Nucleotides 1997, 16, 1091.
(21) Zhang, J .; Neamati, N.; Pommier, Y.; Nair, V. Bioorg. Med.
Chem. Lett. 1998, 8, 1887.
(22) Rajappan, V. P.; Schneller, S. W. Tetrahedron 2001, 57, 9049.
(23) Rajappan, V. P.; Schneller, S. W. Nucleosides, Nucleotides and
Nucleic Acids 2001, 20, 1117.
(24) Koshkin, A. A.; Singh, S. K.; Nielsen, P.; Rajwanshi, V. K.;
Kumar, R.; Meldgaard, M.; Olsen, C. E.; Wengel, J . Tetrahedron 1998,
54, 3607.
(18) Seley, K. L.; J anuszczyk, P.; Hagos, A.; Zhang, L.; Dransfield,
D. T. J . Med. Chem. 2000, 43, 4877.

3368 J . Org. Chem., Vol. 67, No. 10, 2002
Seley et al.
Sch em e 3
Sch em e 5
Sch em e 6
a
(i) 1,2,3,5-Tetra-O-acetyl-â-D-ribofuranose, BSA, acetonitrile;
(ii) TMSOTf. NH3, MeOH. c For series a , NaH and then CH3I;
b
d
for series b, Bu4NI, NaH, BnBr. EtMgBr, anhydrous DMF.
e Hydroxylamine hydrochloride. f Ac2O. g NH2C(O)CH2SH, K2CO3.
h
NaOEt, EtOH.
Sch em e 4
at best. Incomplete benzylation of the three hydroxyls
in various combinations occurred, thereby resulting in
an inseparable mixture of products; therefore, we ad-
justed our procedure to utilize an in situ formation of
benzyl iodide with benzyl bromide and tetrabutylammo-
nium iodide.26 This modification proved to be successful,
and 13b was obtained in good yield (87%). In addition,
purification was accomplished with ease, since there were
few side products formed.
Proceeding onward with 13b, we obtained tricyclic
intermediate 22b by way of a series of reactions identical
to that used before (Scheme 4). Grignard addition of the
formyl group gave 14b, which was then converted to the
oxime 15b. Dehydration of 15b with acetic anhydride
then provided 16b. Next, addition of the thioglycolamide
group, followed by base-catalyzed ring closure of the
thiophene ring, was accomplished to give 18b. Closure
of the pyrimidine ring was carried out as before to afford
19b. Treatment with phosphorus pentasulfide, followed
by methylation, provided 21b. Ammonolysis then gave
22b.
Once 22b was in hand, removal of the benzyl groups
was undertaken. Due to the presence of the sulfur, we
suspected that standard conditions employing palladium
might be unsuccessful (which proved to be true); however,
several alternative methods were tried. Once again, we
met with numerous failures as summarized in Scheme
6. Taking a different approach, we then tried removal of
the sulfur to form the protected split nucleoside 23 in
the hopes this would overcome the poisoning of the
palladium deprotection methods, but this failed as well.
The split nucleoside 23 was readily formed, but all
procedures employed failed to remove the benzyl groups.
Finally, as outlined in Scheme 7, a procedure was
found that utilized an excess of boron trifluoride etherate
a
b
d
CH(OEt)3/Ac2O (1:1). P2S5, pyridine. c K2CO3, MeI. NH3,
BuOH.
we have been unable to limit the formation of this side
product. Next, 14a was converted to the oxime 15a and
subsequently dehydrated to yield nitrile 16a . Introduc-
tion of the thiol side chain using thioglycolamide gave
17a , which was then immediately subjected to ring
closure to provide the bicyclic imidazole-thiophene inter-
mediate 18a . We have found this step to be quite facile;
ring closure occurs to a large extent even before treat-
ment with base most likely due to the extended aroma-
ticity that results.
Ring closure of the pyrimidine ring to form the inosine
intermediate 19a was accomplished with a 1:1 mixture
of acetic anhydride and triethyl orthoformate (Scheme
4).25 Conversion of the carbonyl of 19a was carried out
using standard conditions with phosphorus pentasulfide
to give the thiocarbonyl 20a , which was then immediately
methylated to afford 21a . Standard ammonolysis then
provided 22a . At this point we attempted to deblock the
hydroxyls; however, the methyl protecting groups proved
to be completely recalcitrant to any attempts at removal
(Scheme 5). Several standard methods using a variety
of conditions were tried but all failed to produce 6.
Realizing that the benzyl group, although more sterically
hindering than the methyl group, would be stable to all
of the reaction conditions in the pathway and removal is
typically is more facile than with a methyl group, we were
forced to return to the beginning of the synthetic path-
way, this time using the benzyl protecting group.
Using standard benzylation conditions employing so-
dium hydride and benzyl bromide proved to be difficult
(25) Rousseau, R. J .; Robins, R. K. J . Heterocycl. Chem. 1965, 2,
196.
(26) Czernecki, S.; Georgoulis, C.; Provelenghiou, C. Tetrahedron
Lett. 1976, 17, 3535.

“Fleximers”: Design and Synthesis
J . Org. Chem., Vol. 67, No. 10, 2002 3369
Sch em e 7
Sch em e 9
a
b
BF3‚OEt2, EtSH, CH2Cl2. Raney Ni, MeOH, reflux. c Pd on
a
b
Raney Ni, MeOH, reflux. BF3‚OEt2, EtSH, CH2Cl2. c Pd on
C, ammonium formate, MeOH.
C, ammonium formate, MeOH.
Sch em e 8
carbon disulfide, followed by addition of hydrogen per-
oxide and finally ammonia, yielded 24. Removal of the
protecting groups using the boron trifluoride etherate
method gave the desired guanosine tricyclic intermediate
25, which was then subjected to treatment with Raney
nickel to give fleximer 3.
Once again repeating the reversal of the steps, cleavage
of the sulfur in the spacer ring gave the protected
guanosine fleximer 26, but only deprotection with pal-
ladium and ammonium formate consistently afforded 3.
Finally, the inosine analogue 2 was obtained from 19b
in a manner analogous to that shown in Scheme 9.
Treatment of 19b with the boron trifluoride etherate
method gave the deprotected tricyclic nucleoside 27,
which could then be converted to 2; however, the best
yield once again came from treatment of 19b with Raney
nickel and then deprotection with the palladium method
to provide the inosine fleximer 2.
We have now shown that in all three cases, (i) the
tricyclic nucleosides were completely resistant to the
palladium method of deprotection (most probably due to
poisoning of the catalyst by the sulfur), as well as many
other commonly used standard methods, (ii) the use of
the boron trifluoride etherate deprotection method worked
consistently well for the tricyclics, (iii) the use of the
boron trifluoride etherate method did not produce con-
sistent or acceptable yields for the fleximers, and puri-
fication was difficult at best, and (iv) the protected
fleximers reacted quickly and cleanly and gave excellent
yields when using the palladium method of deprotection.
In summary, we have successfully synthesized the first
examples of a new class of nucleosides that can be used
as dimensional bioprobes of a wide variety of biological
systems. Broad-screen testing to determine medicinal
properties, as well as other biological and biophysical
investigations, for both the tricyclics and the fleximers
has just begun, and reports of these and other studies
will be forthcoming.
a
b
(i) NaOH, MeOH; (ii) CS2, heat; (iii) H2O2; (iv) NH3. Raney
Ni, MeOH, reflux. c BF3‚OEt2, EtSH, CH2Cl2. Pd on C, am-
monium formate, MeOH.
d
and ethanethiol to deblock the hydroxyls.27,28 This proved
to be successful, and tricyclic adenosine analogue 6 was
obtained, albeit in very low yield. Next, removal of the
sulfur in the spacer ring of 6 was accomplished by
treatment with Raney nickel and refluxing in methanol
for 8 h to provide fleximer 1. Since we were curious to
see if we could improve the yield, we reversed the steps
and carried out the desulfurization to give 23 before
removal of the benzyl groups. Again employing the boron
trifluoride etherate/ethanethiol method, we were suc-
cessful in deprotecting 23; however, the yields were again
quite low, and the purification was difficult even employ-
ing HPLC methods.
Once again searching for a new method, we uncovered
a recent paper29 employing palladium on charcoal and
ammonium formate. To our delight, this method proved
to be outstanding: the yields were high, and purification
was accomplished with ease; 1 was obtained from 23 in
94% yield. To obtain the guanosine fleximer, intermediate
18b was subjected to an alternative procedure for closure
of the pyrimidine ring (Scheme 8). Our previously re-
ported18 method employed chloroformamidine hydrochlo-
ride, but this could not be used for these analogues since
the acidic conditions inherent in this procedure resulted
in nearly quantitative cleavage of the glycosidic bond of
the nucleoside; therefore, we turned our attention to an
alternative procedure.30 Sequential treatment of 18b with
sodium hydroxide and methanol and then heating with
Exp er im en ta l Section
Gen er a l. Melting points are uncorrected. Combustion
analyses were performed by Atlantic Microlabs, Inc., Atlanta,
GA. 1H and 13C NMR spectra were recorded at 300 and 75
MHz, respectively, all referenced to internal tetramethylsilane
(TMS) at 0.0 ppm. The spin multiplicities are indicated by the
symbols s (singlet), d (doublet), t (triplet), q(quartet), m
(multiplet), and b (broad). Reactions were monitored by thin-
layer chromatography (TLC) using 0.25 mm silica gel 60-F254
precoated plates. Column chromatography was performed on
silica (200-400 mesh, 60 Å) and elution with the indicated
solvent system. Yields refer to chromatographically and spec-
troscopically (1H and 13C NMR) homogeneous materials.
2,3-Diacetoxy-5-acetoxym eth yl-1-(7-th iom eth ylim idazo-
[4′,5′:4,5]th ien o[3,2-d]pyr im idin -3-yl)-â-D-r ibofu r an ose (8)
(27) Fuji, K.; Ichikawa, K.; Node, M.; Fujita, E. J . Org. Chem. 1979,
44, 1661.
(28) Node, M.; Nishide, K.; Fuji, K.; Fujita, E. J . Org. Chem. 1980,
45, 4275.
(29) Koshkin, A. A.; Fensholdt, J .; Pfundheller, H. M.; Lomholt, C.
J . Org. Chem. 2001, 66, 8506.
(30) Bennett, S. M.; Nguyen-Ba, N.; Ogilvie, K. K. J . Med. Chem.
1990, 33, 2162.

3370 J . Org. Chem., Vol. 67, No. 10, 2002
a n d 2,3-Dia cet oxy-5-a cet oxym et h yl-1-(7-t h iom et h yl-
Seley et al.
chromatography eluting with hexane/ethyl acetate (2:1) gave
13a as a colorless syrup (10.4 g, 89%), which was used directly
in the next step without further purification: 1H NMR (CDCl3)
δ 3.41 (s, 3 H), 3.44 (s, 3 H), 3.53 (s, 3 H), 3.55 (dd, 2.4 Hz,
11.1 Hz, 1 H), 3.73 (dd, 2.7 Hz, 11.1 Hz, 1 H), 3.95-4.01 (m,
2 H), 4.19-4.23 (m, 1 H), 5.75 (d, 3.0 Hz, 1 H), 8.07 (s, 1 H);
13C NMR (CDCl3) δ 58.3, 58.8, 59.2, 71.1, 77.4, 81.3, 83.3, 89.1,
101.4, 117.7, 136.4.
imidazo[4′,5′:4,5]thieno[3,2-d]pyrimidin-1-yl)-â-D-ribofuranose
(9). To a stirred mixture of 718 (178 mg, 0.80 mmol) and 1,2,3,5-
tetra-O-acetate-â-D-ribofuranose (0.28 g, 0.88 mmol) in anhy-
drous acetonitrile (20 mL) was added N,O-bis(trimethylsilyl)-
acetamide (BSA) (1.20 mL, 4.8 mmol). The mixture was stirred
at room temperature under Ar for 6 h. The mixture was cooled
to 0 °C, at which point trimethylsilyltriflate (TMS-OTf) (174
µL, 0.9 mmol) was added. The mixture was further stirred at
60 °C for 18 h. The solvent was removed under reduced
pressure, and the residue was cooled to 0 °C and quenched
with cold, saturated NaHCO3. The mixture was extracted with
CH2Cl2 (30 mL × 3), and the organic extracts were combined
and washed with brine. The solvent was removed under
reduced pressure to give a brown syrup. Column chromatog-
raphy eluting with hexane/ethyl acetate (2:1) gave a single
compound as a colorless syrup (180 mg, 46% yield), which,
following structural characterization with two-dimensional
NMR, proved to be 9: 1H NMR (CDCl3) δ 2.06 (s, 3 H, H-Ac),
2.09 (s, 3 H, H-Ac), 2.19 (s, 3 H, H-Ac), 2.78 (s, 1 H, H-SMe),
3.68 (dd, 3.6 Hz, 11.0 Hz, 1 H, H-5), 4.23 (dd, 3.6 Hz, 11.0 Hz,
1 H, H-5), 4.46 (dd, 3.6 Hz, 8.4 Hz, 1 H), 4.52 (d, 8.4 Hz, 1 H),
4.59-4.64 (m, 1 H), 5.86 (dd, 3.9 Hz, 4.8 Hz, 1 H, H-2), 8.18
(s, 1 H), 8.94 (s, 1 H); 13C NMR (CDCl3) δ 11.9, 20.0, 20.2, 20.4,
62.9, 70.1, 73.0, 80.2, 88.0, 128.9, 143.5, 144.3, 150.7, 152.97,
153.03, 163.9, 168.8, 169.2, 170.1. Anal. Calcd for C19H20N4O7S2‚
0.4EtOAc: C, 47.80; H, 4.41; N, 11.15; S, 12.76. Found: C,
47.98; H, 4.53; N, 10.97; S, 12.76.
2,3-Dib en zyloxy-5-b en zyloxym et h yl-1-(4,5-d ib r om o-
im id a zol-3-yl)-â-D-r ibofu r a n ose 13b. Sodium hydride
(95%, 2.9 g, 0.115 mol) was added to a stirred solution
of 12 (13.7 g, 38.3 mmol) in anhydrous THF (100 mL)
at 0 °C under Ar. Then, the mixture was stirred at room
temperature for 3 h. Tetrabutylammonium iodide (0.5 g,
1.35 mmol) and benzyl bromide (19.7 g, 0.115 mol) was
added. The mixture was stirred at room temperature for
3 h, followed by quenching with ethanol (20 mL). The
solvent was removed under reduced pressure; H2O (200
mL) was added and the mixture extracted with CH2Cl2
(3 × 100 mL). The organic extracts were combined, washed
with brine (200 mL), and dried over MgSO4. The solvent
was removed under reduced pressure to give a pale brown
syrup. Column chromatography eluting with hexane/ethyl
acetate (3:1) gave 13b as a colorless syrup (20.9 g, 87%
yield): 1H NMR (CDCl3) δ 3.54 (dd, 2.4 Hz, 10.8 Hz, 1
H), 3.75 (dd, 2.7 Hz, 10.8 Hz, 1 H), 4.10-4.16 (m, 2 H),
4.31-4.34 (m, 1 H), 4.43-4.66 (m, 6 H), 5.83 (d, 3.3 Hz,
1 H), 7.22-7.38 (m, 15 H), 7.88 (s, 1 H); 13C NMR (CDCl3)
δ 68.3, 72.5, 72.6, 73.5, 75.5, 80.5, 81.9, 89.5, 101.8, 117.6,
127.8, 127.9, 128.0, 128.2, 128.5, 128.6, 136.2, 136.7, 137.1.
Anal. Calcd for C29H28Br2N2O4: C, 55.43; H, 4.49; N, 4.46;
Br, 25.43. Found: C, 55.33; H, 4.60; N, 4.29; Br, 25.15.
2,3-Dim eth oxy-5-m eth oxym eth yl-1-[(5-br om o-4-ca r ba l-
d eh yd eoxim e)im id a zol-3-yl]-â-D-r ibofu r a n ose 15a . Eth-
ylmagnesium bromide (3.0 M in Et2O, 2.4 mL, 7.2 mmol) was
added to a stirred solution of 13a (2.8 g, 7.0 mmol) in
anhydrous Et2O (50 mL) under Ar at room temperature. The
mixture was then stirred for 5 h. Anhydrous DMF (3.0 mL)
was added, and the mixture was further stirred for 18 h. The
solvent was removed under reduced pressure; saturated NH4-
Cl aqueous solution (100 mL) was added and the mixture
extracted with CH2Cl2 (3 × 60 mL). The organic fractions were
combined, washed with brine (2 × 100 mL), and dried over
MgSO4. The solvent was removed under reduced pressure to
give a pale brown syrup. Column chromatography eluting with
hexane:ethyl acetate (1:1) gave 14a as a tan crystalline solid
(1.4 g, 58% yield), which was used without further purifica-
tion: 1H NMR (CDCl3) δ 3.39 (s, 3 H), 3.45 (s, 3 H), 3.72 (s, 3
H), 3.61 (dd, 2.0 Hz, 11.4 Hz, 1 H), 3.82 (br d, 3.9 Hz, 1 H),
3.93 (dd, 1.8 Hz, 11.4 Hz, 1 H), 3.98 (dd, 4.2 Hz, 9.0 Hz, 1 H),
4.23 (dt, 2.0 Hz, 9.0 Hz, 1 H), 6.38 (br s, 1 H), 8.60 (s, 1 H),
9.75 (d, 1.2 Hz, 1 H); 13C NMR (CDCl3) δ 58.4, 59.1, 69.8, 75.7,
80.7, 82.4, 89.2, 125.7, 141.3, 169.0, 179.3. A solution of 14a
(0.50 g, 1.4 mmol) in ethanol (20 mL) was added to a mixture
of hydroxylamine hydrochloride (0.69 g, 10.0 mmol) and
NaHCO3 (0.84 g, 10.0 mmol) in H2O (5 mL). The mixture was
stirred at room temperature for 2 h. The solvent was removed
under reduced pressure; H2O (30 mL) was added and the
mixture filtered. The solid collected was washed further with
H2O to give 15a as a hygroscopic off-white crystalline solid
(0.27 g, 50% yield), mp 159-161 °C: 1H NMR (DMSO-d6) δ
3.32 (s, 3 H), 3.32 (s, 3 H), 3.36 (s, 3 H), 3.49 (dd, 3.6 Hz, 11.0
Hz, 1 H), 3.59 (dd, 3.6 Hz, 11.0 Hz, 1 H), 3.94 (t, 4.8 Hz, 1 H),
4.09 (dd, 3.6 Hz, 8.4 Hz, 1 H), 4.21 (t, 4.8 Hz, 1 H), 6.28 (d, 4.5
Hz, 1 H), 7.99 (s, 1 H), 8.18 (s, 1 H), 11.64 (s, 1 H); 13C NMR
(DMSO-d6) δ 57.7, 58.2, 58.8, 71.5, 77.3, 80.9, 81.8, 87.8, 119.3,
121.6, 130.5, 138.1. Anal. Calcd for C12H18BrN3O5‚0.125H2O:
C, 39.32; H, 5.02; N, 11.46. Found: C, 39.36; H, 5.05; N, 11.19.
2,3-Diben zyloxy-5-ben zyloxym eth yl-1-[(5-br om o-4-ca r -
baldeh ydeoxim e)im idazol-3-yl]-â-D-r ibofu r an ose 15b. Eth-
ylmagnesium bromide (12.8 mL, 3.0 M in Et2O, 38.4 mmol)
was added to a stirred solution of 13b (24.0 g, 38.2 mmol) in
anhydrous diethyl ether (200 mL) under Ar at room temper-
ature. The mixture was then stirred for 5 h. Anhydrous DMF
(20.0 mL) was added, and the mixture was further stirred for
2,3-Diacetoxy-5-acetoxym eth yl-1-(4,5-dibr om oim idazol-
3-yl)-â-D-r ibofu r a n ose (11). To a stirred solution of 1,2,3,5-
tetra-O-acetate-â-D-ribofuranose (25.4 g, 79.7 mmol) and 4,5-
dibromoimidazole 1018 (18.0 g, 79.7 mmol) in anhydrous
acetonitrile (500 mL) was added BSA (1.5 mL, 6.0 mmol). The
reaction mixture was stirred for 5 h at room temperature and
cooled to 0 °C. TMS-OTf (0.30 mL, 1.6 mmol) was added. The
mixture was further stirred at 60 °C for 8 h, at which point it
was chilled to 0 °C, and saturated aqueous NaHCO3 solution
(15 mL) was added slowly. The mixture was extracted with
CH2Cl2 (3 × 30 mL). The extracts were combined and washed
with brine (50 mL). The organic layer was dried over MgSO4,
and the solvent was removed under reduced pressure to give
a brown syrup. Flash column chromatography eluting with
petroleum ether/EtOAc (1:1) gave 11 as a pale yellow syrup
(23.0 g, 47.5 mmol; 66%), which was used directly in the next
step: 1H NMR (CDCl3) δ 2.06 (s, 9 H), 4.25-4.30 (m, 2 H),
4.33-4.38 (m, 1 H), 5.31 (dt, 2.7 Hz, 5.2 Hz, 1 H), 5.46 (dt, 2.2
Hz, 5.2 Hz, 1 H), 5.77 (dd, 2.6 Hz, 4.6 Hz, 1 H), 7.76 (s, 1 H);
13C NMR (CDCl3) δ 20.2, 20.3, 20.5, 62.3, 69.6, 73.5, 80.0, 88.0,
102.4, 118.2, 135.2, 168.8, 169.3, 170.0.
(4,5-Dibr om oim id a zol-3-yl)-1-â-D-r ibofu r a n ose (12). A
solution of triacetate 11 (13.7 g, 28.3 mmol) in saturated
methanolic ammonia (300 mL) in a steel bomb was heated at
100 °C for 3 h. The reaction was cooled and the solvent
removed under reduced pressure to give a light brown foam.
Column chromatography eluting with CHCl3/MeOH (9:1) gave
12 as a colorless syrup (10.1 g, quantitative): 1H NMR (CD3-
OD) δ 3.72 (dd, 3.2 Hz, 12.2 Hz, 1 H), 3.84 (dd, 3.0 Hz, 12.2
Hz, 1 H), 4.03-4.07 (m, 1 H), 4.22 (t, 5.1 Hz, 1 H), 4.32 (t, 4.8
Hz, 1 H), 5.69 (d, 4.5 Hz, 1 H), 8.24 (s, 1 H); 13C NMR (CD3-
OD) δ 61.9, 71.0, 76.8, 86.6, 92.4, 104.3, 117.5, 138.1. Anal.
Calcd for C8H10Br2N2O4‚0.25MeOH: C, 27.05; H, 3.02; N, 7.64;
Br, 43.68. Found: C, 26.99; H, 2.83; N, 7.62; Br, 43.57.
2,3-Dim eth oxy-5-m eth oxym eth yl-1-(4,5-d ibr om oim id a -
zol-3-yl)-â-D-r ibofu r a n ose 13a . Sodium hydride (95%, 2.12
g, 88.5 mmol) was added to a stirred solution of 12 (10.1 g,
28.2 mmol) in anhydrous THF (200 mL) at 0 °C under Ar.
Then, the mixture was stirred at room temperature for 1 h.
Methyl iodide (12.57 g, 88.5 mmol) was added dropwise. The
mixture was stirred at room temperature for 18 h and poured
over ice to quench the reaction. The volatiles were removed
under reduced pressure, and H2O was added (200 mL). The
resulting mixture was extracted with CH2Cl2 (3 × 200 mL).
The organic extracts were combined, washed with brine (300
mL), and dried over MgSO4. The solvent was removed under
reduced pressure to give a pale brown syrup. Flash column

“Fleximers”: Design and Synthesis
J . Org. Chem., Vol. 67, No. 10, 2002 3371
18 h. The solvent was removed under reduced pressure.
Saturated NH4Cl solution (100 mL) was added and the mixture
extracted with CH2Cl2 (3 × 200 mL). The organic extracts were
combined, washed with brine (2 × 200 mL), and dried over
MgSO4. The solvent was removed under reduced pressure to
give a pale brown syrup. Column chromatography eluting with
hexane/ethyl acetate (3:1) gave a colorless syrup whose 1H
NMR showed that it was a mixture of 14b and the starting
material 13b, which proved to be inseparable, so the mixture
was used directly without further purification. A solution of
the mixture containing 13b and 14b (2.0 g) in ethanol (200
mL) was added to a mixture of hydroxylamine hydrochloride
(6.9 g, 0.10 mol) and NaHCO3 (8.4 g, 0.10 mol) in H2O (30 mL).
The mixture was stirred at room temperature for 2 h. The
solvent was removed under reduced pressure; the residue was
treated with H2O (200 mL) and the mixture extracted with
CH2Cl2 (3 × 200 mL). The organic extracts were combined,
washed with brine (300 mL), and dried over MgSO4. The
solvent was removed under reduced pressure to give a pale
brown syrup, which was purified by column chromatography
eluting with hexane/ethyl acetate (3:1) to afford 15b as a white
crystalline solid (14.6 g, 65% yield in two steps), mp 178-180
°C: 1H NMR (CDCl3) δ 3.59 (dd, 2.4 Hz, 11.1 Hz, 1 H), 3.86
(dd, 2.4 Hz, 11.1 Hz, 1 H), 4.03 (dd, 2.0 Hz, 4.3 Hz, 1 H), 4.22
(dd, 2.0 Hz, 4.5 Hz, 1 H), 4.34 (dt, 2.0 Hz, 7.5 Hz, 1 H), 4.38-
4.79 (m, 6 H), 6.41 (d, 2.1 Hz, 1 H), 7.22-7.39 (m, 15 H), 7.56
(s, 1 H), 8.12 (s, 1 H); 13C NMR (CDCl3) δ 67.5, 68.3, 72.7, 72.9,
73.4, 75.0, 80.8, 81.0, 89.6, 90.3, 89.6, 90.3, 120.7, 121.7, 127.7,
127.8, 128.0, 128.5, 128.6, 137.3, 137.8, 140.3. Anal. Calcd for
was heated under reflux for 3.0 h. The solvent was removed
under reduced pressure to give a brown solid. Purification via
column chromatography eluting with CH2Cl2/MeOH (15:1)
afforded 18a as a white crystalline solid following recrystal-
lization from MeOH (2.94 g, 92%), mp 57 °C: 1H NMR (DMSO-
d6) δ 3.29 (s, 3 H), 3.32 (s, 3 H), 3.38 (s, 3 H), 3.55 (dd, 3.3 Hz,
9.0 Hz, 1 H), 3.61 (dd, 3.0 Hz, 9.0 Hz, 1 H), 4.00-4.05 (m, 2
H), 4.20 (br d, 1 H), 5.94 (d, 3.0 Hz, 1 H), 6.68 (br s, 1H), 6.88
(br s, 1 H), 8.23 (s, 1 H); 13C NMR (DMSO-d6) δ 58.6, 59.1,
59.3, 72.1, 77.5, 83.0, 84.1, 88.8, 98.9, 128.3, 139.8, 143.4, 146.5,
168.7. Anal. Calcd for C14H20N4O5S‚0.5MeOH: C, 46.76; H,
5.95; N, 15.09; S, 8.61. Found: C, 46.43; H, 5.67; N, 15.11; S,
8.80.
2,3-Diben zyloxy-5-ben zyloxym eth yl-1-[(5-car boxam ide)-
th ien o[2,3-d ]im id a zol-3-yl]-â-D-r ibofu r a n ose (18b). In a
procedure analogous to that for 18a , use of 16b (16.0 g, 27.3
mmol), thioglycolamide (3.7 g, 41.0 mmol) and potassium
carbonate (3.7 g, 27 mmol) gave 17b, which was shown by TLC
to be a mixture of 17b and 18b (7.8 g). The mixture was then
treated with sodium ethoxide (21 wt % in ethanol, 1.0 g, 3.1
mmol) in anhydrous ethanol (100 mL) and refluxed for 4 h.
The solvent was removed in vacuo to provide 18b as a colorless
syrup (5.0 g, 88%): 1H NMR (DMSO-d6) δ 3.59 (dd, 3.0 Hz,
11.1 Hz, 1H), 3.74 (dd, 3.9 Hz, 11.1 Hz, 1H), 4.24-4.34 (m,
3H), 4.45-4.66 (m, 6H), 6.09 (d, 2.7 Hz, 1 H), 6.52 (br s, 1H),
6.84 (br s, 1 H), 7.15-7.35 (m, 15H), 8.22 (s, 1 H); 13C NMR
(DMSO-d6) δ 36.2, 58.0, 58.2, 58.3, 59.2, 72.1, 78.1, 82.2, 88.7,
101.3, 111.5, 141.0, 148.5, 169.7. Anal. Calcd for C32H32
-
N4O5S: C, 65.74; H, 5.52; N, 9.58; S, 5.48. Found: C, 65.65;
H, 5.62; N, 9.47; S, 5.43.
C
30H30BrN3O5: C, 60.80; H, 5.10; N, 7.09; Br, 13.49. Found:
C, 60.76; H, 5.12; N, 7.12; Br, 13.74.
2,3-Dim eth oxy-5-m eth oxym eth yl-1-(im id a zo[4′,5′:4,5]-
th ien o[3,2-d]pyr im idin -3-yl-7-on e)-â-D-r ibofu r an ose (19a).
A mixture of 18a (2.5 g, 6.9 mmol) and diethoxymethyl acetate
(3.45 mL, 20.7 mmol) was refluxed for 3 h. The excess
diethoxymethyl acetate was then evaporated and the residue
purified by chromatography eluting with CH2Cl2/MeOH (8:1)
to afford 19a as a white solid (2.07, 81%), mp 82 °C: 1H NMR
(DMSO-d6) δ 3.31 (s, 3 H), 3.34 (s, 3 H), 3.38 (s, 3 H), 3.55 (dd,
4.8 Hz, 10.8 Hz, 1H), 3.66 (dd, 4.8 Hz, 10.8 Hz, 1H), 4.04 (br
t, 4.2 Hz, 1 H), 4.18 (br t, 3.9 Hz, 1 H), 4.69 (br t, 5.5 Hz, 1 H),
6.25 (d, 3.0 Hz, H-1), 8.30 (s, 1 H), 8.53 (s, 1 H); 13C NMR
(DMSO-d6) δ 57.9, 58.4, 59.3, 72.6, 78.5, 81.8, 82.3, 88.0, 121.8,
128.6, 141.3, 143.5, 147.9, 149.3, 158.4. Anal. Calcd for
2,3-Dim eth oxy-5-m eth oxym eth yl-1-[(5-br om o-4-ca r bo-
n itr ile)im id a zol-3-yl]-â-D-r ibofu r a n ose 16a . A mixture of
oxime 15a (2.10 g, 5.76 mmol) in acetic anhydride (200 mL)
was heated under reflux for 3 h. The volatiles were removed
under reduced pressure to give a brown syrup. Column
chromatography eluting with CH2Cl2/MeOH (25:1) gave 16a
as a light brown syrup (1.84 g, 92%): 1H NMR (CDCl3) δ 3.37
(s, 3 H), 3.38 (s, 3 H), 3.47 (s, 3 H), 3.51 (dd, 2.7 Hz, 10.8 Hz,
1 H), 3.68 (dd, 2.7 Hz, 10.8 Hz, 1 H), 3.92 (br t, 2.1 Hz, 1 H),
3.99 (br t, 4.2 Hz, 1 H), 4.19-4.23 (m, 1 H), 5.74 (d, 3.9 Hz, 1
H), 8.00 (s, 1 H); 13C NMR (CDCl3) δ 57.9, 58.4, 59.0, 71.7,
77.0, 80.6, 81.5, 119.0, 119.9, 120.9, 130.2. Anal. Calcd for
C
12H16BrN3O4: C, 41.63; H, 4.67; N, 12.14; Br, 23.08. Found:
C15H18N4O5S: C, 49.17; H, 4.95; N, 15.29; S, 8.75. Found: C,
C, 41.84; H, 4.76; N, 12.30; Br, 23.31.
49.00; H, 4.95; N, 15.26; S, 8.74.
2,3-Diben zyloxy-5-ben zyloxym eth yl-1-[(5-br om o-4-ca r -
bon itr ile)im id a zol-3-yl]-â-D-r ibofu r a n ose (16b). Using a
procedure identical to that used for 16a , we obtained 16b as
a colorless syrup (16.0 g, 90%) from 15b (18.2 g, 30.6 mmol):
1H NMR (CDCl3) δ 3.50 (dd, 2.7 Hz, 10.8 Hz, 1 H), 3.75 (dd,
3.3 Hz, 10.8 Hz, 1 H), 4.07-4.14 (m, 2H), 4.36 (dd, 3.0 Hz, 7.5
Hz, 1 H), 4.43-4.66 (m, 6H), 5.82 (d, 4.2 Hz, 1 H), 7.20-7.38
(m, 15 H), 7.80 (s, 1 H); 13C NMR (CDCl3) δ 67.9, 68.3, 71.6,
71.7, 71.8, 71.9, 72.7, 74.9, 80.1, 82.2, 88.9, 90.3, 103.1, 109.3,
126.7, 126.8, 127.3, 127.5, 127.8, 136.1, 136.7, 138.3, 147.3,
165.7. Anal. Calcd for C30H28Br2N2O4: C, 62.72; H, 4.91; N,
7.32; Br, 13.91. Found: C, 62.82; H, 4.91; N, 7.17; Br, 13.66.
2,3-Dim eth oxy-5-m eth oxym eth yl-1-[(5-ca r boxa m id e)-
th ien o[2,3-d ]im id a zol-3-yl]-â-D-r ibofu r a n ose (18a ). A mix-
ture of 16a (4.10 g, 11.40 mmol), freshly prepared thiogly-
colamide (4.80 g, 52.10 mmol), and potassium carbonate (6.0
g, 43.2 mmol) in anhydrous DMF (300 mL) was heated at 55
°C for 18 h. The mixture was cooled, and the solid was filtered
off. Then, the solvent was removed under reduced pressure to
give a dark brown syrup. Column chromatography eluting with
CH2Cl2/MeOH (15:1) gave a white solid 17a (3.2 g, 75%), which
was used directly in the next step: 1H NMR (DMSO-d6) δ 3.29
(s, 3 H), 3.32 (s, 3 H), 3.36 (s, 3 H), 3.50 (dd, 4.5 Hz, 11.1 Hz,
1 H), 3.55 (dd, 4.2 Hz, 11.1 Hz, 1 H), 3.77 (s, 2 H), 3.95 (dd,
4.2 Hz, 8.7 Hz, 1 H), 4.18 (dd, 4.2 Hz, 8.1 Hz, 1 H), 4.27 (br t,
8.1 Hz, 1 H), 5.77 (d, 2.7 Hz, 1 H), 7.14 (br s, 1H), 7.54 (br s,
1 H), 8.28 (s, 1 H); 13C NMR (DMSO-d6) δ 36.2, 58.0, 58.3,
59.2, 72.1, 78.1, 82.2, 84.2, 88.7, 101.3, 111.5, 141.0, 148.5,
169.7. A mixture of 17a (3.2 g, 8.9 mmol) and sodium ethoxide
(2 mL, 21 wt % in ethanol) in anhydrous ethanol (400 mL)
2,3-Dib en zyloxy-5-b en zyloxym et h yl-1-(im id a zo[4′,5′:
4,5]th ien o[3,2-d ]p yr im id in -3-yl-7-on e)-â-D-r ibofu r a n ose
(19b). A mixture of 18b (2.0 g, 3.4 mmol), triethyl orthoformate
(20 mL), and acetic anhydride (20 mL) was refluxed for 3 h.
The excess solvent was evaporated and the residue purified
by column chromatography eluting with ethyl acetate/hexane
(3:1) to afford 19b as a hygroscopic white foam (1.7 g, 81%):
1H NMR (CDCl3) δ 3.63 (dd, 4.1 Hz, 10.5 Hz, 1 H), 3.84 (dd,
5.0 Hz, 10.5 Hz, 1 H), 4.22 (t, 4.8 Hz, 1 H), 4.46 (q, 4.4 Hz, 1
H), 4.52-4.74 (m, 6 H), 4.94 (t, 5.5 Hz, 1 H), 6.28 (d, 5.1 Hz,
1 H), 7.02-7.12 (m, 5 H), 7.24-7.34 (m, 10 H), 7.86 (s, 1 H),
8.16 (s, 1 H); 13C NMR (CDCl3) δ 61.2, 61.6, 61.9, 68.5, 71.9,
73.2, 75.1, 79.1, 81.8, 88.7, 121.4, 127.5, 127.6, 127.8, 128.2,
136.9, 137.2, 137.4, 143.2, 144.4, 144.6, 151.2, 160.2. Anal.
Calcd for C33H30N4O5S‚0.33H2O: C, 65.98; H, 5.14; N, 9.33, S,
5.33. Found: C, 65.97; H, 5.05; N, 9.11, S, 5.22.
2,3-Dim eth oxy-5-m eth oxym eth yl-1-(7-th iom eth ylim id -
azo[4′,5′:4,5]th ien o[3,2-d]pyr im idin -3-yl)-â-D-r ibofu r an ose
(21a ). A solution of P2S5 (2.3 g), dry pyridine (250 mL), and
19a (2.1 g, 5.77 mmol) was refluxed for 24 h, cooled, and
evaporated to dryness under vacuum, and the residue was
purified by column chromatography eluting with CH2Cl2/
MeOH (10:1) to afford 20a as a grayish solid (1.33 g, 61%),
which was used without purification. To a stirring solution of
20a (1.1 g, 2.9 mmol) and potassium carbonate (486 mg, 3.5
mmol) in MeOH (100 mL) was added methyl iodide (446 mg,
3.19 mmol). After 15 min, the solvent was removed by rotary
evaporation and the residue purified via column chromatog-
raphy eluting with CH2Cl2/MeOH (15:1) to give 21a as a
colorless syrup (1.09 g, 93%), which was used directly without

3372 J . Org. Chem., Vol. 67, No. 10, 2002
Seley et al.
further purification: 1H NMR (DMSO-d6) δ 2.71 (s, 3H), 3.30
(s, 3 H), 3.35 (s, 3 H), 3.38 (s, 3 H), 3.58 (dd, 5.4 Hz, 10.8 Hz,
1 H), 3.70 (dd, 5.0 Hz, 10.8 Hz, 1 H), 4.08 (br t, 4.5 Hz, 1 H),
4.22 (dd, 4.8 Hz, 8.7 Hz, 1 H), 4.80 (br t, 5.7 Hz, 1 H), 6.31 (d,
2.7 Hz, 1H), 8.62 (s, 1 H), 9.04 (s, 1 H); 13C NMR (DMSO-d6)
δ 55.0, 58.0, 58.3, 59.2, 72.6, 78.4, 81.9, 82.2, 88.4, 116.7, 127.6,
146.3, 147.5, 150.4, 154.9, 164.6.
324.0766, found 324.0764. Anal. Calcd for C12H13N5O4S‚
0.4H2O: C, 43.61; H, 4.21; N, 21.17; S, 9.70. Found: C, 43.93;
H, 4.15; N, 20.77; S, 9.51.
2,3-Dib en zyloxy-5-b en zyloxym et h yl-1-[5-(4-a m in op y-
r im id in -6-yl)im id a zol-1-yl]-1-â-D-r ibofu r a n ose (23). A mix-
ture of 22b (0.25 g, 0.42 mmol) and freshly prepared Raney
Ni (1.0 g) in MeOH/H2O (50:50 mL) was heated under reflux
for 18 h. The mixture was filtered and the filtrate evaporated
under reduced pressure. The residue was purified by column
chromatography eluting with CH2Cl2/MeOH (10:1) to afford
23 as a white hygroscopic foam (0.21 g, 88%): 1H NMR
(DMSO-d6) δ 3.61 (dd, 3.0 Hz, 10.2 Hz, 1 H), 3.76 (dd, 4.5 Hz,
10.2 Hz, 1 H), 4.23 (br s, 2 H), 4.31-4.35 (m, 1 H), 4.49-4.68
(m, 6 H), 6.66 (s, 1 H), 6.87 (d, 3.3 Hz, 1 H), 6.95 (br s, 2 H),
7.19-7.33 (m, 15 H), 7.43 (s, 1 H), 8.12 (s, 1 H), 8.36 (s, 1 H);
13C NMR (CD3OD) δ 68.4, 72.2, 72.5, 73.2, 75.5, 81.7, 101.4,
127.8, 128.1, 128.2, 128.3, 128.6, 137.8, 137.9, 138.0, 158.0,
164.3. Anal. Calcd for C33H33N5O4: C, 70.32; H, 5.90; N, 12.42.
Found: C, 70.27; H, 5.91; N, 12.29.
2,3-Diben zyloxy-5-ben zyloxym eth yl-1-[(5-a m in oim id -
azo-[4′,5′:4,5]thieno[3,2-d]pyrimidin-3-yl-7-one)]-â-D-ribofuran-
ose (24). In a steel bomb, a mixture of 18b (1.35 g, 0.344 mmol)
and NaOH (0.68 g, 17 mmol) in anhydrous MeOH (50 mL) was
stirred for 30 min. at room temperature until the mixture
became homogeneous. Carbon disulfide (1.05 mL, 17.5 mmol)
was added, and the bomb was sealed and heated in an oil bath
at 145 °C for 18 h. The solvent was removed under reduced
pressure and MeOH (80 mL) added to the residue. The mixture
was cooled to 0 °C; hydrogen peroxide (30%, 7.5 mL) was added
dropwise and the mixture stirred at 0 °C for 2 h. The mixture
was then transferred to a steel bomb and anhydrous ammonia
bubbled through for 10 min. The bomb was sealed and heated
in an oil bath at 120 °C for 18 h. The solvent was removed in
vacuo, and the residue was purified by column chromatography
eluting with hexane/ethyl acetate (1:3). The fractions contain-
ing product were combined and concentrated, and a second
column was run eluting with CH2Cl2/EtOH (95:5) to give 24
as a hygroscopic white foam (0.8 g, 57%): 1H NMR (CDCl3) δ
3.64 (dd, 3.0 Hz, 9.9 Hz, 1 H), 3.84 (dd, 4.5 Hz, 9.9 Hz, 1 H),
4.25 (t, 5.0 Hz, 1 H), 4.44 (dd, 4.0, Hz, 8.7 Hz, 1 H), 4.52-4.79
(m, 6 H), 5.26 (br s, 2 H), 6.30 (d, 4.2 Hz, 1 H), 7.14-7.37 (m,
15 H), 8.12 (s, 1 H), 11.69 (s, 1 H); 13C NMR (CDCl3) δ 68.0,
71.5, 71.7, 72.7, 74.7, 79.2, 81.0, 88.3, 110.2, 126.3, 127.0, 127.3,
127.6, 127.8, 136.7, 143.3, 145.5, 150.6, 152.8, 160.0. Anal.
Calcd for C33H31N5O5S‚1.0H2O: C, 63.14; H, 5.14; N, 11.16; S,
5.11. Found: C, 63.11; H, 5.05; N, 11.13; S, 5.09.
1-[(5-Am in oim id a zo[4′,5′:4,5]th ien o[3,2-d ]p yr im id in -3-
yl-7-on e)]-â-D-r ibofu r a n ose 2,3,5-Tr iol (25). In a procedure
analogous to that used to obtain 6, deprotection of 24 gave 25
as a hygroscopic white crystalline solid (57 mg, 58%), mp 233-
235 °C: 1H NMR (D2O) δ 3.74 (dd, 2.4 Hz, 10.5 Hz, 1H), 3.82
(dd, 1.8 Hz, 10.5 Hz, 1H), 4.08-4.16 (m, 3H), 5.82 (d, 4.5 Hz,
1 H), 8.46 (s, 1H); 13C NMR (CD3OD) δ 62.2, 70.7, 77.0, 88.6,
92.5, 113.6, 129.4, 133.1, 138.8, 149.0, 154.3, 158.4; HRMS
(ESI) calcd for C12H15N5O5S 339.0637, found 339.0637. Anal.
Calcd for C12H13N5O5S‚1.0H2O: C, 40.33; H, 4.23; N, 19.60; S,
8.97. Found: C, 40.52; H, 3.95; N, 19.32; S, 9.27.
2,3-Dib en zyloxy-5-b en zyloxym et h yl-1-[5-(2-a m in op y-
r im id in -6-yl-4-on e)im id a zol-1-yl]-1-â-D-r ibofu r a n ose (26).
A mixture of 24 (0.20 g, 0.33 mmol) and Raney Ni in MeOH
(50 mL) was heated under reflux for 18 h. The mixture was
filtered, and the filtrate was evaporated under reduced pres-
sure. Chromatography eluting with CH2Cl2/MeOH (10:1) gave
the protected fleximer 26 as a sticky white foam (0.17 g,
90%): 1H NMR (CDCl3) δ 3.60 (dd, 3.0 Hz, 9.0 Hz, 1 H), 3.83
(dd, 4.5 Hz, 9.0 Hz, 1 H), 4.10 (br s, 1 H), 4.20-4.23 (m, 1 H),
4.34-4.38 (m, 1 H), 4.45-4.69 (m, 6 H), 5.68 (br s, 2 H), 6.05
(d, 4.2 Hz, 1 H), 6.88 (s, 1 H), 7.20-7.38 (m, 15 H), 7.50 (s, 1
H), 8.23 (s, 1 H); 13C NMR (DMSO-d6) δ 68.3, 72.7, 73.5, 76.0,
81.1, 81.8, 89.3, 98.4, 127.5, 127.9, 128.6, 132.9, 137.3, 137.4,
155.1, 165.4. Anal. Calcd for C33H33N5O5‚0.25MeOH: C, 67.95;
H, 5.83; N, 11.91. Found: C, 67.68; H, 5.78; N, 11.95.
2,3-Dib en zyloxy-5-b en zyloxym et h yl-1-(7-t h iom et h yl-
imidazo-[4′,5′:4,5]thieno[3,2-d]pyrimidin-3-yl)-â-D-ribofuran-
ose (21b). A mixture of 19b (2.8 g, 6.2 mmol) and P2S5 (2.8 g,
6.3 mmol) in anhydrous pyridine (200 mL) was heated under
reflux for 24 h. The solvent was removed under reduced
pressure to give a dark brown syrup. Column chromatography
eluting with ethyl acetate/hexane (2:1) gave 20b as a light
brown foam (2.2 g, 75%), which was used without further
purification. Methyl iodide (142 mg, 1.0 mmol) was added to
a mixture of 20b (292 mg, 0.5 mmol) and K2CO3 (138 mg, 1.0
mmol) in anhydrous MeOH (20 mL) and stirred for 10 min.
The solvent was removed under reduced pressure, and the
residue was purified by column chromatography eluting with
ethyl acetate/hexane (2:1) to provide 21b (1.69 g, 75%) as a
light brown syrup: 1H NMR (DMSO-d6) δ 2.78 (s, 3 H), 3.68
(dd, 4.3 Hz, 10.5 Hz, 1 H), 3.92 (dd, 5.1 Hz, 10.5 Hz, 1 H), 4.25
(t, 4.8 Hz, 1 H), 4.48 (q, 4.5 Hz, 1 H), 4.55-4.87 (m, 6 H), 6.26
(d, 5.1 Hz, 1 H), 7.01-7.06 (m, 5 H), 7.26-7.37 (m, 10 H), 8.17
(s, 1 H), 8.68 (s, 1 H); 13C NMR (DMSO-d6) δ 12.6, 69.0, 72.3,
73.7, 75.5, 79.2, 82.3, 89.3, 126.7, 127.8, 127.9, 128.0, 128.2,
128.7, 137.4, 137.7, 137.9, 144.2, 145.1, 151.2, 153.1, 163.9.
Anal. Calcd for C14H20N4O5S: C, 65.36; H, 5.16; N, 8.97; S,
10.26. Found: C, 65.49; H, 5.24; N, 8.77; S, 10.05.
2,3-Dim eth yloxy-5-m eth yloxym eth yl-1-[(7-a m in o)im id -
azo[4′,5′:4,5]th ien o[3,2-d]pyr im idin -3-yl]-â-D-r ibofu r an ose
(22a ). A solution of 21a (500 mg, 1.26 mmol) in saturated
butanolic ammonia (20 mL) was sealed in a steel bomb and
heated at 160 °C for 90 h. The solvent was removed by rotary
evaporation, and the residue was purified via column chro-
matography eluting with CH2Cl2/MeOH (8:1) to give 22a as a
white solid (426 mg, 92%), mp 84 °C: 1H NMR (300 MHz,
DMSO-d6) δ 3.27 (s, 3H), 3.28 (s, 3H), 3.29 (s, 3H), 3.47 (dd,,
1 H, H-C5), 3.72 (dd, 1 H, H-C5), 4.05 (t, 1H), 4.19 (t, 1H),
4.84 (t, 1H), 6.27 (d, 1H), 7.48 (s, 2H), 8.40 (s, 1H), 8.47 (s,
1H); 13C NMR (75 MHz, DMSO-d6) δ 57.9, 58.3, 59.2, 72.6,
78.5, 81.8, 88.1, 100.1, 113.4, 128.0, 145.2, 145.7, 148.7, 154.9,
159.3.
2,3-Diben zyloxy-5-ben zyloxym eth yl-1-[(7-a m in o)im id -
azo[4′,5′:4,5]th ien o[3,2-d]pyr im idin -3-yl]-â-D-r ibofu r an ose
(22b). A solution of 21b (1.0 g, 1.6 mmol) in saturated
butanolic ammonia (50 mL) was sealed in a steel bomb and
heated at 160 °C for 90 h. The solvent was removed by rotary
evaporation, and the residue was purified via column chro-
matography eluting with CH2Cl2/MeOH (8:1) to give 22 as a
hygroscopic white foam (720 mg, 76%): 1H NMR (CDCl3) δ
3.64 (m, 1 H), 3.96 (m, 1 H), 4.25 (m, 1H), 4.62-4.89 (m, 2 H),
4.44-4.58 (m, 6H), 6.29 (d, 1 H), 7.06-7.09 (m, 5 H), 7.25-
7.32 (m, 10 H), 8.14 (s, 1 H), 8.34, (s, 1H); 13C NMR (CDCl3) δ
68.9, 72.3, 73.6, 79.5, 82.2, 89.4, 114.3, 137.6, 137.7, 137.9,
146.2, 149.5, 154.3, 158.3. Anal. Calcd for
0.33H2O: C, 66.09; H, 5.32; N, 11.68; S, 5.35; Found: C, 66.04;
H, 5.32; N, 11.56; S, 5.27.
C33H31N5O4S‚
1-[(7-Am in oim id a zo[4′,5′:4,5]th ien o[3,2-d ]p yr im id in -3-
yl)]-2,3,5-â-D-r ibofu r a n ose Tr iol (6). A mixture of 22b (84
mg, 0.14 mmol), BF3‚OEt2 (250 µL, 1.2 mmol), and EtSH (1.2
mL, 16.0 mmol) in anhydrous CH2Cl2 (3.0 mL) was stirred at
room temperature for 24 h. The solvent and excess reagents
were removed under reduced pressure, and the residue was
dissolved in H2O (10 mL) and extracted with CH2Cl2 (3 × 5
mL). The aqueous layer was evaporated to dryness to afford 6
as a hygroscopic white powder, following recrystallization in
H2O (26 mg, 58%): 1H NMR (DMSO-d6) δ 3.56 (dd, 3.3 Hz,
10.5 Hz, 1 H), 3.71 (dt, 3.3 Hz, 10.5 Hz, 1H), 3.99 (br s, 1H),
4.15 (br s, 1 H), 4.64 (q, 2.7 Hz, 1H), 5.23 (br s, 1H), 5.45 (br
d, 6.9 Hz, 1H), 5.78 (dd, 3.3 Hz, 7.5 Hz, 1H), 5.98 (d, 7.2 Hz,
1H), 7.51 (s, 2H), 8.34 (s, 1H), 8.49 (s, 1H); 13C NMR (D2O) δ
60.9, 69.7, 75.8, 87.1, 90.7, 113.0, 123.0, 127.6, 129.9, 130.4,
146.8, 147.3; HRMS (FAB) calcd for (M + 1) C12H14N5O4S
1-[5-(P yr im idin -6-yl-4-on e)im idazol-1-yl]-1-â-D-r ibofu r a-
n ose 2,3,5-Tr iol (27). In a procedure analogous to that used
to obtain 6, deprotection of 19b gave 27 as hygroscopic white

“Fleximers”: Design and Synthesis
J . Org. Chem., Vol. 67, No. 10, 2002 3373
needles (142 mg, 64%), mp 242-244 °C: 1H NMR (DMSO-d6)
δ 3.56 (dd, 3.3 Hz, 10.5 Hz, 1 H), 3.71 (dt, 3.3 Hz, 10.5 Hz, 1
H), 3.99 (br s, 1 H), 4.15 (br s, 1 H), 4.64 (q, 2.7 Hz, 1 H), 5.23
(br s, 1 H), 5.45 (br d, 6.9 Hz, 1 H), 5.78 (dd, 3.3 Hz, 7.5 Hz, 1
H), 5.98 (d, 7.2 Hz, 1 H), 7.51 (s, 2 H), 8.34 (s, 1 H), 8.49 (s, 1
H); 13C NMR (DMSO-d6) δ 61.5, 70.2, 74.1, 86.0, 89.0, 107.8,
120.9, 126.3, 128.2, 129.5, 142.7, 157.7; HRMS (ESI) calcd for
(M + H) C12H13N4O5S 325.0607, found 325.0611. Anal. Calcd
for C12H12N4O5S‚1.0H2O: C, 42.10; H, 4.12; N, 16.37; S, 9.36.
Found: C, 42.25; H, 4.01; N, 16.03; S, 9.10.
2,3-Diben zyloxy-5-ben zyloxym eth yl-1-[5-(p yr im id in -6-
yl-4-on e)im id a zol-1-yl]-1-â-D-r ibofu r a n ose (28). In a pro-
cedure analogous to that used to obtain 26, cleavage of 19b
gave 28 as a hygroscopic foam (180 mg, 92%): 1H NMR (CDCl3)
δ 3.59 (dd, 2.2 Hz, 10.8 Hz, 1 H), 3.80 (dd, 2.3 Hz, 10.8 Hz, 1
H), 4.18-4.25 (m, 2 H), 4.36-4.39 (m, 1 H), 4.45-4.67 (m, 6
H), 6.71 (s, 1 H), 6.73 (d, 3.0 Hz, 1 H), 7.18-7.36 (m, 15 H),
7.54 (s, 1 H), 8.00 (s, 1 H), 8.22 (s, 1 H); 13C NMR (CDCl3) δ
68.3, 72.4, 72.7, 73.4, 75.7, 81.2, 81.5, 88.9, 110.2, 127.6, 127.7,
127.8, 127.9, 128.4, 133.0, 137.2, 139.5, 148.6, 155.1, 163.7.
Anal. Calcd for C33H32N4O5‚1.25H2O: C, 67.50; H, 5.88; N,
9.54. Found: C, 67.41; H, 5.57; N, 9.80.
of eight syn and eight anti initial structures were generated
that differed from one another by a 45° rotation of the
pyrimidine moiety about the imidazole ring. This family of
multiple conformations was used to assay the energetics of
pyrimidine ring conformational freedom with respect to the
rest of the fleximer nucleoside structure. The sixteen starting
structures therefore served to minimize the uncertainty in the
energetically favored values of the two dihedral angles. Specif-
ically, the dihedral angle defined as C2′-C1′-N1-C2 (the
glycosidic bond that defines the syn-anti conformation of the
fleximer pyrimidine base about the ribose) and the dihedral
defined by N1-C5-C6′-N1′ (the fleximer bond that defines
the pyrimidine ring rotation about the imidazole ring) were
initially altered and ultimately analyzed. Thus, the set of
structures was also used to assess any effects that the starting
geometry had on the final Hartree-Fock-optimized conforma-
tion. All furanose conformations were initially set to C2′ endo.
The sixteen 3-21G-optimized structures were subjected to
self-consistent field (SCF) geometry optimizations using re-
stricted Hartree-Fock (RHF) SCF gradient optimization at
the 6-311G* level of theory. All internal degrees of freedom
were optimized without constraints or parameterization. For
the simulation, the gradient convergence was set to <0.01
Hartree/Bohr for the change in molecular energy as a function
of the displacement forces. Electric dipole moments and
molecular energies were calculated by single-point SCF opti-
mization using an RHF 6-311G* basis set. All calculations
were performed using SPARTAN software.
Au tom a ted Molecu la r -Dock in g Meth od s. The crystal
structure of S-adenosylhomocysteine hydrolase complexed with
the inhibitor (1′R, 2′S, 3′R)-9-(2′,3′-dihydroxycyclopentan-1′-
yl)adenine (PDB code 1A7A) was the source of protein atomic
coordinates. Prior to docking computations, the coordinate file
was edited to remove all water molecules and the inhibitor.
Docking was carried out using the program suite LigandFit
from MSI, Inc. A search grid of 35 × 35 × 35 Å was constructed
in the enzyme active site. Grid points were separated by 0.2
Å. All rotatable dihedrals of the adenosine fleximer were
allowed to move freely, and the molecule was docked via a
Monte Carlo configuration search. The docking run was
allowed to save up to 1000 possible solutions, which were then
automatically clustered into groups. A group was defined by
a root-mean-squared (rms) deviation of less than or equal to
0.3 Å. The clusters were then scored on the basis of the
calculated intermolecular energy of the solution.
1-[5-(4-Am in op yr im id in -6-yl)im id a zol-1-yl]-1-â-D-r ibo-
fu r a n ose 2,3,5-Tr iol (1). A mixture of 23 (0.20 g, 0.62 mmol),
Pd/C (10% Pd on C, 0.20 g), and ammonium formate (0.20 g,
3.2 mmol) in MeOH (50 mL) was heated under reflux for 18
h. The mixture was filtered over Celite and the filtrate
evaporated under reduced pressure. The residue was purified
via column chromatography eluting with EtOAc/EtOH/acetone/
H2O (4:1:1:0.5) to give hygroscopic white needles (98 mg, 94%),
1
mp 186 °C: H NMR (D2O) δ 3.65 (dd, 3.3 Hz, 12.9 Hz, 1 H),
3.82 (dd, 2.4 Hz, 12.9 Hz, 1 H), 4.00-4.13 (m, 2 H), 4.28 (dd,
1.8 Hz, 4.5 Hz, 1 H), 6.29 (d, 2.1 Hz, 1 H), 7.12 (s, 1 H), 8.00
(s, 1 H), 8.46 (s, 1 H), 9.19 (s 1 H); 13C NMR (D2O) δ 59.7,
68.1, 75.9, 84.6, 92.5, 106.8, 123.9, 127.5, 136.4, 150.0, 157.8,
161.8; HRMS (EI) calcd for
C12H15N5O4 293.1124, found
293.1127. Anal. Calcd for C12H15N5O4‚0.5H2O: C, 47.68; H,
5.33; N, 23.17. Found: C, 47.87; H, 5.37; N, 23.14.
1-[5-(2-Am in op yr im id in -6-yl-4-on e)im id a zol-1-yl]-1-â-
D-r ibofu r a n ose 2,3,5-Tr iol (2): hygroscopic white needles (53
mg, 92%), mp > 212 °C dec; 1H NMR (D2O) δ 3.71 (dd, 3.3 Hz,
13.2 Hz, 1 H), 3.88 (br d, 13.2 Hz, 1 H), 4.04-4.16 (m, 2 H),
4.27-4.30 (m, 1 H), 6.42 (s, 1 H), 6.80 (s, 1 H), 7.90 (s, 1 H),
8.28 (s, 1 H), 9.17 (s, 1 H); 13C NMR (D2O) δ 59.6, 68.0, 75.8,
84.2, 92.5, 113.8, 122.1, 129.6, 135.5, 150.5, 152.1, 163.9. Anal.
Calcd for C12H14N4O5‚1.0 H2O: C, 46.15; H, 5.16; N, 17.94.
Found: C, 46.33; H, 5.09; N, 17.85.
Ack n ow led gm en t. Project support provided in part
by the Georgia Tech/Emory Biomedical Research pro-
gram and the Georgia Institute of Technology is greatly
appreciated. The authors thank Professors Piet Her-
dewijn (Rega Institute, Leuven), Stewart W. Schneller
(Auburn University), Donald Doyle (Georgia Tech), as
well as Dr. Bal Bhat (ISIS Pharmaceuticals) and Drs.
Kyo Watanabe, Krystoff Pankiewicz, and Steve Patter-
son (Pharmasset Inc.) for many helpful discussions. The
efforts of Ms. Elizabeth Gooding, Ms. Marion Go¨tz, and
Ms. J ulia Sponaugle in making starting materials are
appreciated.
1-[5-(2-Am in op yr im id in -6-yl-4-on e)im id a zol-1-yl]-1-â-
D-r ibofu r a n ose 2,3,5-Tr iol (3): hygroscopic white needles
1
(0.155 g, 91%), mp > 204 °C dec; H NMR (D2O) δ 3.70 (dd,
3.3 Hz, 10.2 Hz, 1H), 3.87 (dd, 2.4 Hz, 10.2 Hz, 1H), 4.07-
4.16 (m, 2H), 4.29 (dd, 2.4 Hz, 4.5 Hz, 1H), 6.11 (s, 1 H), 6.38
(d, 5.7 Hz, 1 H), 7.84 (d, 1.2 Hz, 1H), 9.14 (dd, 0.6 Hz, 1.5 Hz,
1 H); 13C NMR (D2O) δ 60.1, 69.0, 75.9, 85.4, 91.6, 105.0, 122.5,
130.0, 135.8, 145.3, 154.0, 163.4; HRMS (ESI) calcd for
C
12H15N5O5 309.1073, found 309.1068. Anal. Calcd for C12H15-
N5O5‚0.75H2O: C, 44.65; H, 5.15; N, 21.71. Found: C, 44.53;
H, 5.09; N, 21.57.
Ab In itio Qu a n tu m Mech a n ica l Ca lcu la tion s. The
starting geometries for the three fleximer nucleosides (1-3)
were based on standard bond lengths and angles for purine
nucleosides.31 Each fleximer structure was initially optimized
as a series of syn and anti conformations via unconstrained
gradient geometry optimization using a 3-21G basis set. A total
Su p p or tin g In for m a tion Ava ila ble: Two-dimensional
NMR spectra and structural analysis for compound 9. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(31) Saenger, W. Principles of Nucleic Acids; Springer-Verlag: New
York, 1984.
J O0255476