Synthesis of new transglycosidically tethered 5′-nucleotides constrained to a highly biologically relevant profile

Article abstract of DOI:10.1021/jo0110045

A new motif for restricting 5′-nucleotides to highly biologically relevant conformations has been developed. The 5′,6-oxomethylene transglycosidically tethered versions of uridine 5′-monophosphate and 2′-deoxyuridine 5′-monophosphate (1 and 2, respectively) were synthesized in 10-11 steps from their respective natural nucleoside precursors along routes general to the preparation of tethered versions of a wide variety of 5′-nucleotide-based compounds. In both routes, a shelf-stable 6-hydroxymethyl pyrimidine nucleoside 5′-carboxaldehyde is the key intermediate. It exists in a carbohydrate-like fashion in a cyclic hemiacetal form under aprotic conditions. The phosphorylated cyclic hemiacetals 1 and 2 were isolated as binary mixtures of 5′-diastereomers differing principally in the trajectory of the phosphate group with respect to the carbohydrate. By ¹H NMR, both 1 and 2 were demonstrated to be stable to hydrolysis at ambient temperature in D₂0 solution for at least 2 months. The oxomethylene transglycosidic tether as deployed in 1 and 2 leaves all of the native 5′-nucleotide molecular recognition Sites intact while it restricts the framework to a low-energy anti glycosyl conformation and an extended phosphate disposition. This provides a spatial presence that approximates nearly three-quarters of the protein-bound 5′-nucleotide ligands described in the Protein Data Bank. The tether has a low structural and electronic impact, occupies a region of space (over the β-face of the furan ring) seldom penetrated by proteins, and should be accommodated as readily on purine-based 5′-nucleotide frameworks as on pyrimidine-based ones. Because of its unique and attractive features, this new motif for the conformational restriction of 5′-nucleotides is expected to be useful for producing probes of structure/function relationships and in assessing the conformational binding requirements that enzymes and receptor sites have for their natural 5′-nucleotide-based ligands.

Full text of DOI:10.1021/jo0110045

2152
J . Org. Chem. 2002, 67, 2152-2159
Syn th esis of New Tr a n sglycosid ica lly Teth er ed 5′-Nu cleotid es
Con str a in ed to a High ly Biologica lly Releva n t P r ofile
Michael P. Groziak* and David W. Thomas†
SRI International, 333 Ravenswood Ave., Menlo Park, California 94025-3493
michael.groziak@sri.com
Received October 15, 2001
A new motif for restricting 5′-nucleotides to highly biologically relevant conformations has been
developed. The 5′,6-oxomethylene transglycosidically tethered versions of uridine 5′-monophosphate
and 2′-deoxyuridine 5′-monophosphate (1 and 2, respectively) were synthesized in 10-11 steps
from their respective natural nucleoside precursors along routes general to the preparation of
tethered versions of a wide variety of 5′-nucleotide-based compounds. In both routes, a shelf-stable
6-hydroxymethyl pyrimidine nucleoside 5′-carboxaldehyde is the key intermediate. It exists in a
carbohydrate-like fashion in a cyclic hemiacetal form under aprotic conditions. The phosphorylated
cyclic hemiacetals 1 and 2 were isolated as binary mixtures of 5′-diastereomers differing principally
1
in the trajectory of the phosphate group with respect to the carbohydrate. By H NMR, both 1 and
2 were demonstrated to be stable to hydrolysis at ambient temperature in D2O solution for at least
2 months. The oxomethylene transglycosidic tether as deployed in 1 and 2 leaves all of the native
5′-nucleotide molecular recognition sites intact while it restricts the framework to a low-energy
anti glycosyl conformation and an extended phosphate disposition. This provides a spatial presence
that approximates nearly three-quarters of the protein-bound 5′-nucleotide ligands described in
the Protein Data Bank. The tether has a low structural and electronic impact, occupies a region of
space (over the â-face of the furan ring) seldom penetrated by proteins, and should be accommodated
as readily on purine-based 5′-nucleotide frameworks as on pyrimidine-based ones. Because of its
unique and attractive features, this new motif for the conformational restriction of 5′-nucleotides
is expected to be useful for producing probes of structure/function relationships and in assessing
the conformational binding requirements that enzymes and receptor sites have for their natural
5′-nucleotide-based ligands.
In tr od u ction
puckering so that when incorporated into DNA or RNA
strands, they preorganize those oligomers for an en-
hanced molecular recognition by other nucleic acids.2
Tethers that have been attached to the glycon portion of
nucleotide units include the â-2′,3′-oxoethylene,3 the
R-3′,4′-propylene,4 the R-3′,5′-azaethylene,5 the R-2′,4′-(2-
oxapropano),6 and the R-3′,1′-azamethylene tether.7 Oth-
ers more extensively developed are the â-3′,5′-ethylene
tether of bicyclo-DNA8 and its cyclopropane-fused version
in tricyclo-DNA9 and the R-2′,4′-oxomethylene tether of
LNA (locked nucleic acids).10
The conformational restriction of nucleotides with
tethers designed to limit primarily the flexibility of the
carbohydrate moiety has received an enormous amount
of attention recently.1-10 Most of these carbohydrate-
tethered nucleotide units are restricted only in furanose
† Director, Mass Spectrometry Laboratory.
(1) Meldgaard, M.; Wengel, J . J . Chem. Soc., Perkin Trans. 1 2000,
3539-3554. Herdewijn, P. Biochim. Biophys. Acta 1999, 1489, 167-
179.
(2) Kool, E. T. Chem. Rev. 1997, 97, 1473-1487.
Far less attention has been paid to the design and use
of tethered nucleotides as biochemical probes for char-
acterizing proteins that have 5′-nucleotide monomers or
small molecules containing these units as their ligands.
This is surprising, given the plethora of NMPs, NDPs,
NTPs, and other biologically important small molecules
that contain these units such as the NDP-carbohydrates
and enzyme cofactors such as NAD, FAD, and coenzyme
A. Cyclopropanated carbocyclic bicyclo[3.1.0]hexane ver-
sions of the antiviral agent AZT and its 5′-monophos-
phate have been prepared as carbohydrate conforma-
tionally locked probes of the kinase and HIV-1 reverse
transcriptase enzymes key to its therapeutic efficacy,11
and similarly tethered versions of 2′-deoxyadenosine were
used as probes of the enzyme adenosine deaminase.12 The
(3) Nielsen, P.; Pfundheller, H. M.; Wengel, J . J . Chem. Soc., Chem.
Commun. 1997, 825-826.
(4) Bjo¨rsne, M.; Szabo´, T.; Samuelsson, B.; Classon, B. Bioorg. Med.
Chem. 1995, 3, 397-402. Bar, N. C.; Patra, R.; Achari, B.; Mandal, S.
B. Tetrahedron 1997, 13, 4727-4738.
(5) Wang, G. Y. Tetrahedron Lett. 1999, 40, 6343-6346.
(6) Wang, G. Y.; Gunic, E. Nucleosides Nucleotides 1999, 18, 531-
536. Wang, G. Y.; Girardet, J . L.; Gunic, E. Tetrahedron 1999, 55,
7707-7724. Wang, G. Y.; Gunic, E.; Girardet, J . L.; Stoisavljevic, V.
Bioorg. Med. Chem. Lett. 1999, 9, 1147-1150.
(7) Kværnø, L.; Wightman, R. H.; Wengel, J . J . Org. Chem. 2001,
66, 5106-5112.
(8) Bolli, M.; Lubini, P.; Leumann, C. Helv. Chim. Acta 1995, 78,
2077-2096. Meier, R.; Grueschow, S.; Leumann, C. Helv. Chim. Acta
1999, 82, 1813-1828.
(9) Steffens, R.; Leumann, C. Helv. Chim. Acta 1997, 80, 2426-2439.
(10) Nielsen, C. B.; Singh, S. K.; Wengel, J .; J acobsen, J . P. J .
Biomol. Struct. Dyn. 1999, 17, 175-191. Kværnø, L.; Wengel, J . J .
Chem. Soc., Chem. Commun 1999, 657-658. Nielsen, K. E.; Singh, S.
K.; Wengel, J .; J acobsen, J . P. Bioconjugate Chem. 2000, 11, 228-
238. Petersen, M.; Nielsen, C. B.; Nielsen, K. E.; J ensen, G. A.;
Bondensgaard, K.; Singh, S. K.; Rajwanshi, V. K.; Koshkin, A. A.; Dahl,
B. M.; Wengel, J .; J acobsen, J . P. J . Mol. Recognit. 2000, 13, 44-53.
(11) Marquez, V. E.; Ezzitouni, A.; Russ, P.; Siddiqui, M. A.; Ford,
H. J .; Feldman, R. J .; Mitsuya, H.; George, C.; Barchi, J . J ., J r. J . Am.
Chem. Soc. 1998, 120, 2780-2789.
10.1021/jo0110045 CCC: $22.00 © 2002 American Chemical Society
Published on Web 03/03/2002

Synthesis of New Transglycosidically Tethered 5′-Nucleotides
J . Org. Chem., Vol. 67, No. 7, 2002 2153
LNA13 and â-2′,3′-oxomethylene-tethered14 versions of
AZT were also reported recently, but they were devoid
of anti-HIV activity. There are very few examples of
transglycosidically constrained nucleotide analogues.
Among them are the 8,2′-thio-tethered version of ATP
and 2,2′-anhydro-UTP synthesized as potential agonists
for the P2X2-purinoceptors.15
Resu lts a n d Discu ssion
Our initial efforts focused on the uridine nucleotides,
an important subclass of endogenous nucleotides.20 Al-
though the zero-atom-tethered 6,5′-cyclouridines21 and
the one-atom-tethered 6,5′-methanouridines22 are known,
the electron-withdrawing enone portion of the pyrimidin-
6-yl moiety could labilize a 5′-phosphate group installed
on these strained platforms under certain conditions. A
two-atom-saturated tether would insulate the phosphate
from such lability and, further, would not introduce much
ring strain, but a simple 6,5′-ethylene tether appeared
to be too difficult to construct. From our earlier investi-
gations of the 6-formyluridines23,24 and especially our
recent examination of 6-formyl- and 6-(hydroxymethyl)-
uridine 5′-carboxaldehydes,25 we came to expect that the
5′,6-oxomethylene tethered uridine and 2′-deoxyuridine
5′-monophosphates (1 and 2, respectively) would be
readily accessed and hydrolytically stable.
Conformational restriction is a powerful technique that
selects a predetermined subset of native rotamers for
probing the binding requirements of a biomacromolecule
for its natural ligand. In the case of the ubiquitous 5′-
nucleotide unit, the rotational disposition of just two
bonds, the C4′-C5′ bond and the glycosyl bond (C1′-N1
for pyrimidines and C1′-N9 for purines), largely deter-
mines the molecule’s spatial presentation, which is fine-
tuned by the furanose puckering. A search of the Protein
Data Bank16 using ReLiBase17 reveals 981 instances of
a protein-bound 6-unsubstituted pyrimidine or 8-unsub-
stituted purine 5′-nucleotide-based ligand. The resolution
of these structures notwithstanding, nearly three-quar-
ters of them (711) have a C5′-to-C6 (pyrimidine) or C5′-
to-C8 (purine) separation of e4.5 Å indicative of an anti
glycosyl disposition (pyrimidine C6 or purine C8 over the
furanose ring), and most have an extended phosphate
disposition. Interestingly, the region of space over the
â-face of the furan ring is not utilized by proteins when
they bind 5′-nucleotide ligands: only 4 of those 711 PDB
entries have a protein heavy atom located within 3.5 Å
of both the C5′ and the pyrimidine C6 or purine C8.
Therefore, a transglycosidic tether in this region of space
should not only impose a useful rigidity on the 5′-
nucleotide framework by simultaneously restricting the
C4′-C5′ and the glycosyl bond rotations, but importantly,
it should not encumber protein binding based on steric
considerations.
In this context, the transglycosidic tethering of pyri-
midine 5′-nucleotides between C5′ and C6 and purine 5′-
nucleotides between C5′ and C8 is particularly attractive.
5′-Nucleotide biochemical probes restricted in this man-
ner would be anti glycosyl rotamers with laterally
extended phosphate groups, a disposition known to be
favored by isolated 5′-nucleotides18 and very well repre-
sented by those found bound to proteins by X-ray
methods. Unfortunately, no such highly biomimetic and
thus broadly useful set of compounds has yet been
reported.19 We now describe the synthesis and aqueous
solution stability of the first two members of a new class
of highly biomimetic transglycosidically tethered 5′-
nucleotides that should provide valuable probes of the
conformational binding requirements that proteins have
for 5′-nucleotide ligands.
The synthesis of the 5′,6-oxomethylene-tethered 5′-
UMP 1 is shown in Scheme 1. 2′,3′-O-(Methoxymeth-
ylidene)uridine (3),26 consistently obtained only as a 55:
45 mixture of diastereomers,27 was 5′-O-TBDMS-protected
to 4 (90%), which was then C6-formylated to 5 in
excellent yield via its Miyasaka dianion.28 Aldehyde 5 was
reduced in near quantitative yield to give alcohol 6, which
was then O7-protected with an Fmoc group (90%). The
choice of this group was critical because it could be
removed without producing a spiro-dihydrouridine like
the one we obtained when removing (TBAF) a TBDPS
(19) Cyclouridylic acids featuring an ethylene or propylene tether
connecting the uracil C5 to the 5′-phosphate are highly biomimetic
except that they mask one of the two native dissociable P-OH
groups: Seio, K.; Wada, T.; Sakamoto, K.; Yokoyama, S.; Sekine, M.
J . Org. Chem. 1996, 61, 1500-1504. Seio, K.; Wada, T.; Sakamoto,
K.; Yokoyama, S.; Sekine, M. J . Org. Chem. 1998, 63, 1429-1443.
Sekine, M.; Kurasawa, O.; Shohda, K.-i.; Seio, K.; Wada, T. J . Org.
Chem. 2000, 65, 3571-3578. Sekine, M.; Kurasawa, O.; Shohda, K.;
Seio, K.; Wada, T. J . Org. Chem. 2000, 65, 6515-6524.
(20) Connolly, G. P.; Duley, J . A. Trends Pharmacol. Sci. 1999, 20,
218-225.
(21) Rabi, J . A.; Fox, J . J . J . Org. Chem. 1972, 37, 3898-3901. Otter,
B. A.; Falco, E. A.; Fox, J . J . J . Org. Chem. 1976, 41, 3133-3137. Otter,
B. A.; Falco, E. A.; Fox, J . J . J . Org. Chem. 1978, 43, 481-486.
(22) Otter, B. A.; Falco, E. A. Tetrahedron Lett. 1978, 45, 4383-
4386. Ueda, T. Nucleosides Nucleotides 1985, 4, 67-75. Yamagata, Y.;
Tomita, K.; Usui, H.; Sano, T.; Ueda, T. Chem. Pharm. Bull. 1989, 37,
1971-1976.
(12) Marquez, V. E.; Russ, P.; Alonso, R.; Siddiqui, M. A.; Shin, K.
J .; George, C.; Nicklaus, M. C.; Dai, F.; Ford, H. J . Nucleosides
Nucleotides 1999, 18, 521-530.
(13) Olsen, A. G.; Rajwanshi, V. K.; Nielsen, C.; Wengel, J . J . Chem.
Soc., Perkin Trans. 1 2000, 21, 3610-3614. Obika, S.; Andoh, J .-i.;
Sugimoto, T.; Miyashita, K.; Imanishi, T. Tetrahedron Lett. 1999, 40,
6465-6468.
(14) Sørensen, M. H.; Nielsen, C.; Nielsen, P. J . Org. Chem. 2001,
66, 4878-4886.
(15) Tusa, G.; Reed, J . K. Nucleosides, Nucleotides, Nucleic Acids
2000, 19, 805-813.
(16) Berman, H. M.; Westbrook, J .; Feng, Z.; Gilliland, G.; Bhat, T.
N.; Weissig, H.; Shindyalov, I. N.; Bourne, P. E. Nucleic Acids Res.
2000, 28, 235-242 (via the Internet at http://rutgers.rcsb.org/pdb).
(17) Hendlich, M. Acta Crystallogr. 1998, D54, 1178-1182 (via the
Internet at http://relibase.rutgers.edu).
(18) Gelbin, A.; Schneider, B.; Clowney, L.; Hsieh, S. H.; Olson, W.
K.; Berman, H. M. J . Am. Chem. Soc. 1996, 118, 519-529.
(23) Groziak, M. P.; Koohang, A. J . Org. Chem. 1992, 57, 940-944.
Groziak, M. P.; Lin, R.; Stevens, W. C.; Wotring, L. L.; Townsend, L.
B.; Balzarini, J .; Witvrouw, M.; De Clercq, E. Nucleosides Nucleotides
1996, 15, 1041-1058.
(24) Groziak, M. P.; Koohang, A.; Stevens, W. C.; Robinson, P. D. J .
Org. Chem. 1993, 58, 4054-4060.
(25) Groziak, M. P.; Lin, R. Tetrahedron 2000, 56, 9885-9893.
(26) Griffin, B. E.; J arman, M.; Reese, C. B.; Sulton, J . E. Tetrahe-
dron 1967, 23, 2301-2313.
(27) For a report of 3 obtained as a single diastereomer, see:
Shiragami, H.; Irie, Y.; Shirae, H.; Yokozeki, K.; Yasuda, N. J . Org.
Chem. 1988, 53, 5170-5173.
(28) Tanaka, H.; Nasu, I.; Miyasaka, T. Tetrahedron Lett. 1979, 19,
4755-4758. Tanaka, H.; Hayakawa, H.; Miyasaka, T. Chem. Pharm.
Bull. 1981, 29, 3565-3572. Tanaka, H.; Hayakawa, H.; Miyasaka, T.
Tetrahedron 1982, 38, 2635-2642.

2154 J . Org. Chem., Vol. 67, No. 7, 2002
Groziak and Thomas
Sch em e 1. Syn th esis of th e Teth er ed 5′-UMP 1a
Sch em e 2. Syn th esis of th e Teth er ed 5′-d UMP 2a
a
Conditions: (a) LDA (5 equiv), THF, -78 °C, 3.5 h, then
HCO2Me, THF, -78 °C, 5 h (45%). (b) NaBH4, MeOH, 23 °C, 30
min (99%). (c) Fmoc-Cl, C5H5N, 23 °C, 16 h (62%). (d) catalytic
PPTS, EtOH, 23 °C, 16 h (77%). (e) Dess-Martin periodinane,
CH2Cl2, 23 °C, 2 h (98%). (f) NH3, EtOH, 4 °C, 12 h (86%). (g)
BuLi (2 equiv), THF, -78 °C, 5 min, then (Cl3CCH2O)2P(O)Cl, -78
°C, 15 min (15-20%). (h) Zn(Cu), DMF, 23 °C, 6 h, then Chelex
(99%). (i) 10% aq AcOH, 23 °C, 16 h (90%).
a
Conditions: (a) TBDMS-Cl, C3H4N2, 45 °C, 16 h (90%). (b)
LDA (2.5 equiv), THF, -78 °C, 1.5 h, then HCO2Me, THF, -78
°C, 30 min (97%). (c) NaBH4, MeOH, 23 °C, 30 min (99%). (d)
Fmoc-Cl, C5H5N, 23 °C, 16 h (90%). (e) NH4F (5 equiv), MeOH, 65
°C, 1 h (57%). (f) Dess-Martin periodinane, CH2Cl2, 23 °C, 2 h
(88%). (g) NH3, EtOH, 4 °C, 16 h (83%). (h) BuLi (2 equiv), THF,
-78 °C, 5 min, then (Cl3CCH2O)2P(O)Cl, -78 °C, 15 min (15-
20%). (i) Zn(Cu), DMF, 23 °C, 6 h, then Chelex (99%). (j) 10% aq
AcOH, 23 °C, 1.5 h, then MeOH, 65 °C, 0.5 h (90%).
phate protection strategy35,36 because it permitted the
removal of the phosphate protecting groups under non-
hydrolytic conditions. The low yield of 11 (15-20%) from
Shiba phosphorylation of 10 with the appropriate reagent
is likely due solely to the poor solubility and, hence,
sluggish reactivity of the dianion.37 Conditions that give
a higher yield of phosphorylated product are being
sought. The 1H NMR spectrum of the phosphorylated
hemiacetal 11 revealed it to be a mixture of four diaster-
eomers. Particularly diagnostic was the downfield-shifted
(ca. 6.6 ppm) NMR signal for the anomeric H1′ located
in the range (6.6-6.8 ppm) of the diastereomeric 5′,6-
oxomethylene-tethered 2′,3′,5′-tri-O-acetyluridine we re-
ported recently.25 The stable protected tethered nucleo-
tide 11 was easily purified by chromatography.
Deprotection of 11 using a freshly prepared Zn(Cu)
couple with a 16% Cu content38 proceeded in excellent
yield (99%). The tethered protected nucleotide 12, ob-
tained after a workup with Chelex resin, was then
exposed to 10% aqueous AcOH followed by boiling MeOH
to remove the methoxymethylidene protecting group. In
this way, the target 5′,6-oxomethylene-tethered 5′-UMP
analogue 1 was obtained in its monosodium salt form as
a 1:1 mixture of diastereomers. Target 1 was as stable
to hydrolysis as we had anticipated. 1H NMR showed that
it remained intact in D2O solution at 23 °C for up to 2
months.
group from a related nucleoside.29 We examined many
options for removing the 5′-O-TBDMS group from 7
without affecting the other protecting groups, but only
Robins’ method (excess NH4F in refluxing MeOH)30
worked satisfactorily. Alcohol 8 was isolated in pure
condition in a yield (57%) that mainly reflects the
difficulty of separating it from the excess NH4F.
Dess-Martin periodinane oxidation31 of 8 smoothly
afforded aldehyde 9 (88%). Although we were able to
remove the Fmoc group of 9 with anhydrous Et3N/
pyridine32 and follow this immediately by an in situ
phosphorylation, this sequence surprisingly captured the
“open” hydroxy-aldehyde form of 10. This product (see
Experimental Section) provided useful spectroscopic data
for comparison to that of the desired phosphorylated
cyclic hemiacetal, constitutional isomer 11.
The Fmoc group of 9 could be removed using ethanolic
NH3, and the Fmoc-NH2 byproduct (identified by MS) was
removed by rapid filtration through SiO2 to give pure 10
in 83% yield. As was expected on the basis of its
carbohydrate-like structure, 10 exists exclusively in a
cyclic hemiacetal form in aprotic solvents. As a 1:1
mixture of epimeric hemiacetals, the mixture of four
1
diastereomers in 10 gave a complex H NMR spectrum.
Countering this difficulty was the advantage that 10 was
readily desiccated to a nonhygroscopic, shelf-stable pow-
der, ideal for preparing and storing this key intermediate
in large quantities.
Phosphorylation of the cyclic hemiacetal form of 10 was
successful when Shiba’s methodology was employed.33
This approach involves the generation of a hemiacetal
anion with BuLi at -78 °C followed by the rapid intro-
duction of a dialkylphosphorochloridate.34 For our pur-
poses, we selected Scheit’s bis-(2,2,2-trichloroethyl) phos-
The synthesis (Scheme 2) that culminated in the 5′,6-
oxomethylene-tethered 5′-dUMP (2) was similar, but with
other noteworthy features. Others have observed no
dianion formation when 13 is treated with LDA,39 but
we successfully generated its Miyasaka 3,6-diyl dilithium
species with 5 equiv of LDA. Treatment with HCO2Me
gave the desired 6-carboxaldehyde 14 in 45% yield. The
1
structure of 14 was established by H and 13C NMR and
MS and also by using it to prepare target 2.
The reduction of 14 to 15 (99%) and the Fmoc protec-
tion of 15 to 16 (62%) each proceeded without incident.
Because of the greater lability of a TBDMS group at a
primary alcohol position, we could regioselectively O5′-
(29) Groziak, M. P.; Lin, R.; Robinson, P. D. Acta Crystallogr. 1995,
C51, 1204-1207.
(30) Zhang, W.; Robins, M. J . Tetrahedron Lett. 1992, 33, 1177-
1180.
(31) Dess, D. B.; Martin, J . C. J . Org. Chem. 1983, 48, 4155-4156.
(32) Ben-Hattar, J .; J iricny, J . J . Org. Chem. 1986, 51, 3211-3213.
(33) Inage, M.; Chaki, H.; Kusumoto, S.; Shiba, T. Chem. Lett. 1982,
1281-1284.
(34) Boger, D. L.; Honda, T. J . Am. Chem. Soc. 1994, 116, 5647-
5656.
(35) Eckstein, F.; Scheit, K. H. Angew. Chem. 1967, 79, 317.
(36) Scheit, K. H. Biochim. Biophys. Acta 1968, 157, 632-633.
(37) The major product was 11, but the mass recovery was low.
(38) Imai, J .; Torrence, P. F. J . Org. Chem. 1981, 46, 4015-4021.
(39) Armstrong, R. W.; Gupta, S.; Whelihan, F. Tetrahedron Lett.
1989, 30, 2057-2060.

Synthesis of New Transglycosidically Tethered 5′-Nucleotides
J . Org. Chem., Vol. 67, No. 7, 2002 2155
F igu r e 1. MM conformational analysis of 5′-(R)-1 and 5′-(S)-1. Definitions: γ ) O5′-C5′-C4′-C3, δ ) C5′-C4′-C3′-O3′, ø )
O4′-C1′-N1-C2, P ) arctan[0.3249(ν4 + ν1 - ν3 - ν0)ν2 - 1], νmax ) ν2(cosP) - 1, ν0 ) C4′-O4′-C1′-C2, ν1 ) O4′-C1′-
C2′-C3, ν2 ) C1′-C2′-C3′-C4′, ν3 ) C2′-C3′-C4′-O4′, ν4 ) C3′-C4′-O4′-C1′. Where ν2 is negative, 180 is added to P.
desilylate 16 in EtOH containing catalytic PPTS (77%).
Dess-Martin periodinane oxidation of the resulting 17
(98%) followed by removal of the Fmoc group in aldehyde
18 (86%) then gave 19. Like 10, key intermediate 19
exists exclusively in a cyclic hemiacetal form in anhy-
drous aprotic solvents. It was desiccated to a nonhygro-
scopic, shelf-stable powder and then recrystallized from
absolute EtOH. Shiba phosphorylation of 17 did proceed,
but in a low yield, similar to that with 10. The protected
phosphate 20 was deblocked to 21 (99%), and the 3′-O-
TBDMS group from 21 was then removed by hydrolysis
under conditions (10% aqueous AcOH, 23 °C, 16 h) that
clearly show a neighboring group participation by the
phosphate.40 The 5′,6-oxomethylene-tethered 5′-dUMP
target 2 was obtained as a 5:1 mixture of diastereomers.
Importantly, it was found to be just as hydrolytically
stable as 1.
Sp a tia l P r esen ta tion . Even before the diastereomers
of 1 and 2 are examined in great structural detail by
X-ray crystallographic or high-field two-dimensional
NMR spectral methods, their spatial presentation can be
assessed in a reasonably accurate manner. This informa-
tion will be valuable to enzymologists or other biochem-
ists who would use these transglycosidically oxomethylene-
tethered 5′-nucleotides as probes of protein structure/
function relationships. The disposition of the 5′-6-
oxomethylene tether in the crystalline intramolecular
cyclic hemiacetal form of 1-â-D-ribofuranosyluracil-6-
carboxaldehyde as revealed by X-ray crystallography24
was used as a starting point for MM energy minimiza-
tions of 5′-(R)- and 5′-(S)-1 performed to reveal the
topographical aspects of the lowest steric energy forms.
Parameters most useful in comparing these structures
to untethered naturally occurring 5′-nucleotide counter-
parts are shown in Figure 1 for these diastereomers of
1. Other calculations revealed that oxomethylene-teth-
ered purine and pyrimidine nucleotides are quite similar
to each other, as are the ribofuranosyl and 2′-deoxyribo-
furanosyl variants.
have an average γ value of 178°. With these C5′-C4′
bond rotational restrictions, the oxomethylene-tethered
5′-nucleotides cannot achieve the +sc (+gauche) forms
found in A-DNA (γ ) 49-66°), B-DNA (γ ) 33-57°), or
A-RNA (γ ) 0-46°). One-half of the units in Z-DNA (γ )
57-93°) cannot be replicated, but the other half (γ )
159-197°) could be reasonably approximated by a 5′-(S)
oxomethylene-tethered unit. While their utility in the
study of natural DNA and RNA features appears to be
limited, the oxomethylene-tethered 5′-nucleotides could
very well find some utility in replicating such local DNA/
RNA topological features as “hairpin” turns, bulges, and
“flipped” residues.
The δ values of 5′-(R)- and 5′-(S)-1 close to 160° are
common ones for natural 5′-nucleotides, and their gly-
cosyl torsion angles (ø) near 200° are clearly in the anti
range, defined to be between 170 and 280°. This is typical
of the majority of nucleosides and nucleotides, both
pyrimidine and purine. The flexibility of their tether
should enable the oxomethylene-tethered 5′-nucleotides
to “sweep out” a range of ø values perhaps as large as
30-50°. The pseudorotation (P) values of 212 and 216°
4
place them in the 3T (C-4′-endo/C-3′-exo twist) section
of the continuum of ring-puckered states.41 Importantly,
though, transglycosidically oxomethylene-tethered 5′-
3
nucleotides should display an equilibrium of E (C-3′-
2
endo) and E (C-2′-endo) conformations. The puckering
amplitude (νmax) values are somewhat smaller than the
average value of 36-37° typically found for nucleosides
and nucleotides, indicating that 1 and 2 have a slightly
more flat furan ring than is common.18,41
Con clu sion s
In summary, we have developed a general synthetic
methodology for new highly biomimetic 5′,6-oxomethyl-
ene-tethered 5′-nucleotides that can be used to probe the
requirements that biomacromolecules have for binding
their 5′-nucleotide ligands. Key features of this new motif
include a low structural and electronic disturbance
caused by the tether, a preservation of all native hydro-
gen bond donating and accepting recognition sites, and
a conformational restriction about the glycosyl linkage
that enforces the most commonly found bioactive anti
form. The fact that diastereomers are generated by the
synthetic attachment of a tether to the normally prochiral
C5′ locus is valuable, since it makes it highly likely that,
once separated, the 5′-epimers will become a useful pair
As expected, the native O5′-C5′-C4′-C3′ (γ) torsion
angle is the major difference between 5′-(R)- and 5′-(S)-
1. Near-classical staggered conformations of rotation
about the C5′-C4′ bond were evident. The γ value of 312°
for 5′-(R)-1 is that of an -sc (-gauche) form. For
comparison, -sc (-gauche) rotamers in the Nucleic Acids
Database (NDB) have an average γ value of 286°. The γ
value of 199° for 5′-(S)-1 is that of an ap (trans) form.
For comparison, ap rotameric nucleotides in the NDB
(40) Kawahara, S.-i.; Wada, T.; Sekine, M. J . Am. Chem. Soc. 1996,
118, 9461-9468.
(41) Altona, C.; Sundaralingam, M. J . Am. Chem. Soc. 1972, 94,
8205-8212.

2156 J . Org. Chem., Vol. 67, No. 7, 2002
Groziak and Thomas
minimizations were performed using CambridgeSoft’s Chem
3D Pro version 5.0.
of biochemical probes used in parallel. The synthesis of
these oxomethylene-tethered 5′-nucleotides is somewhat
long (10-11 steps), but most of the steps give high yields
and do not involve overly complicated reaction conditions
or manipulations. The analogous 5′,8-oxomethylene-
tethered purine 5′-nucleotides should be accessible along
similar routes from suitable 8-substituted purine nucleo-
side precursors,42 and work in this direction is already
under way.
2′,3′-O-(Meth oxym eth ylid en e)u r id in e (3). A suspension
of uridine (2.685 g, 11 mmol) in 7 mL of (MeO)3CH was treated
with pTsOH‚H2O (80 mg, catalytic) and the mixture stirred
at 23 °C for 1.75 h, by which time complete dissolution had
occurred. TLC (5% MeOH/CH2Cl2) revealed that most of the
uridine had been consumed and that 3 was present together
with a higher Rf value byproduct (estimated 25%) subsequently
identified as an O5′-dimethoxymethylated nucleoside by 1H
NMR. The reaction mixture was quenched by the addition of
2 drops of concentrated NH4OH, and it was then diluted to 50
mL with CH2Cl2 and filtered through a pad of Celite in a
Bu¨chner funnel. The Celite pad was washed with 5% MeOH/
CH2Cl2, and the filtrate and washing were combined and
rotary evaporated to a colorless foam/gum. Purification by
radial chromatography using 7% MeOH/CH2Cl2 as the eluent
gave 1.87 g (60%) of 3 as a white powder/foam: 1H NMR
(CDCl3) δ 9.4 (bs, 1), 7.41 (d, J ) 8.2 Hz, 1), 7.36 (d, J ) 8.2
Hz, 1), 5.99 (s, 1), 5.93 (s, 1), 5.76 (s, 1), 5.74 (s, 1), 5.70 (d, J
) 3.0 Hz, 1), 5.59 (d, J ) 3.0 Hz, 1), 5.15-5.02 (m, 4), 4.42
(pseudo-q, 1), 4.26 (pseudo-q, 1), 3.94 (d, J ) 12.4 Hz, 1), 3.83
Exp er im en ta l Section
Ma ter ia ls a n d Gen er a l Meth od s. The Dess-Martin
periodinane reagent obtained from Omega, Inc., of Que´bec,
Canada, was kept under argon at <0 °C and allowed to warm
to 23 °C just prior to use. 9-Fluorenylmethyl chloroformate
was from Avocado, and tert-butyldimethylchlorosilane was
from Spectrum. LDA (2.0 M) in heptane/THF/C6H5Et and 1.6
M BuLi in hexanes were from Aldrich, as were the uridine,
2′-deoxyuridine, bis(2,2,2-trichloroethyl)phosphorochloridate,
NaBH4, and NH4F. Anhydrous DMF, THF, CH2Cl2, C5H5N,
and CH3OH were also from Aldrich. These were used as
received. The CD3OD was from Cambridge Isotope Laborato-
ries. Chelex 100, 100-200 mesh, Na+ form analytical-grade
chelating resin from BioRad was washed first with 0.1 N NH4-
HCO3 and then with water immediately prior to use. Aldrich
98% Cu(OAc)2 and activated43 Eastman practical-grade Zn
dust were used to prepare the Zn(Cu) couple with a 16% Cu
content according to a published procedure.38 The Zn(Cu)
couple was stored at 23 °C under anhydrous Et2O under argon.
Thin-layer chromatography (TLC) was performed on Anal-
tech silica gel GF Uniplates. Radial preparative-layer chro-
matography was performed on a Chromatotron instrument
(Harrison Research, Inc., Palo Alto, CA) using Adsorbosil-Plus
P254 from Alltech Associates. This was mixed with CaSO4
binder (VWR) and zinc silicate phosphor (Sigma) before using
it to prepare the plates. Column chromatography was per-
formed using Merck silica gel-60 (200-430 ASTM), and
ascending preparative-layer chromatography was performed
on 2 mm silica gel-GF plates from Analtech. Visualizations
employed short-wave (254 nm) UV light. Nucleosides 3-11 and
13-20 (i.e., those up to the penultimate compounds 12 and
21) were all obtained as chromatographically homogeneous
foam/solids by chromatography on SiO2. Most of these clung
tenaciously to the organic solvents that were used as eluents
(e.g., CH2Cl2, MeOH, EtOAc), even after drying at 23 °C in
vacuo for up to 48 h, by 1H NMR. Some of the elemental
combustion analytical data reflect this trace solvent content.
1
(d, J ) 12.4 Hz, 1), 3.40 (s, 3), 3.33 (s, 3). An H NMR spectrum
of 3 recorded in (CD3)2SO solution matched the data given in
the first report of this compound.26 It was consistently obtained
as a 55:45 mixture of diastereomers rather than as a single
diastereomer.27
2′,3′-O-(Meth oxym eth ylid en e)-5′-O-(ter t-bu tyld im eth -
ylsilyl)u r id in e (4). The 1.87 g (6.54 mmol) of 3 from above
was rendered MeOH-free by repetitive azeotrope with toluene,
dissolved in 6.5 mL of anhydrous DMF, and treated with 1.08
g (7.1 mmol) of TBDMS-Cl and 556 mg (8.17 mmol) of
imidazole. The reaction mixture was kept at 45 °C for 16 h,
diluted with EtOAc, and extracted eight times with water to
remove the DMF. The organic phase was dried (Na2SO4) and
rotary evaporated to a gum, which was purified by filtering a
concentrated CH2Cl2 solution of it through a pad of SiO2 in a
sintered glass funnel. Elution first with pure CH2Cl2 as the
eluent removed traces of the silylating reagent, and then 10%
MeOH/CH2Cl2 was used to remove the product. Rotary evapo-
ration gave, after drying in vacuo at 23 °C, 2.35 g (90%) of 4
as a colorless oil/gum: 1H NMR (CDCl3) δ 8.2 (bs, 1), 7.65 (d,
J ) 8.2 Hz, 1), 7.58 (d, J ) 8.2 Hz, 1), 6.09 (d, J ) 3.0 Hz, 1),
5.99 (s, 1), 5.94 (d, J ) 3.0 Hz, 1), 5.92 (s, 1), 5.61 (s, 1), 5.58
(s, 1), 4.92-4.64 (m, 4), 4.45 (pseudo-q, 1), 4.31 (pseudo-q, 1),
3.93 (d, J ) 12.4 Hz, 1), 3.83 (d, J ) 12.4 Hz, 1), 3.43 (s, 3),
3.35 (s, 3), 0.91 (s, 9), 0.09 (s, 6). Anal. Calcd for C17H28N2O7-
Si: C, 50.98; H, 7.05; N, 6.99. Found: C, 50.66; H, 6.95; N,
6.98 and C, 50.65; H, 6.88; N, 6.93.
2′,3′-O-(Meth oxym eth ylid en e)-5′-O-(ter t-bu tyld im eth -
ylsilyl)u r id in e-6-ca r boxa ld eh yd e (5). A sample of 4 was
readied by eluting an EtOAc solution of it through a short pad
of SiO2, evaporating the eluate to dryness, and then coevapo-
rating the residue several times with toluene and keeping the
initially gummy white solid under vacuum at 23 °C overnight.
A solution of this now white powder (4.4 g, 11 mmol) in 100
mL of anhydrous THF under argon was cooled to -78 °C and
then transferred dropwise via cannula under positive argon
pressure to a -78 °C solution of 13.7 mL (27.5 mmol, 2.5 equiv)
of 2 M LDA in 86 mL of THF under argon. The dianion mixture
was stirred for 1.5 h at -78 °C, and then HCO2Me (1.7 mL,
1.65 g, 27.5 mmol, 2.5 equiv) was introduced dropwise. Stirring
was continued for 30 min at -78 °C, and then AcOH (1.57
mL, 1.65 g, 27.5 mmol, 2.5 equiv) was added followed im-
mediately by 30 mL of 1 M pH 7 phosphate buffer. The mixture
was allowed to warm to 23 °C, and then most of the THF was
removed by rotary evaporation at 23 °C. EtOAc (400 mL) and
water (50 mL) were added, and the layers were separated; the
organic phase was washed twice with water (50 mL) and then
dried (Na2SO4). TLC (5% MeOH/CH2Cl2) revealed essentially
pure 2,4-DNP-positive 5 (Rf 0.29) together with only a very
slight trace of 4 (Rf 0.42). Rotary evaporation gave 5 as a yellow
foam, and this was purified by applying a concentrated CH2-
Cl2 solution of it directly to a short pad of SiO2 (ca. 30 g) in a
Bu¨chner funnel and first eluting with pure CH2Cl2 and then
1H and 13C NMR spectra were recorded at 300 and 75 MHz,
respectively, using internal tetramethylsilane or external
3-(trimethylsilyl)-1-propanesulfonic acid, sodium salt (DSS) (δ
) 0.0 for 1H), CHCl3 (δ ) 77.0 for 13C), or (CD3)2SO (δ ) 39.5
for 13C) as a reference. The intermediates on the route to the
5′,6-oxomethylene-tethered 5′-UMP target 1 were diastereo-
meric 2′,3′-O-methoxymethylidenated nucleosides with com-
plex 1H NMR spectra. Therefore, all intermediates on both
routes were additionally characterized by low-resolution de-
sorption chemical ionization mass spectrometry (DCI MS).
These spectra (data summarized in Tables 1 and 2 in Sup-
porting Information) were recorded using ammonia as the
chemical reagent gas. For DCI MS analysis of 1, 2, 12, and
21, derivatization using BSTFA containing 1% TMSCl (Regis
Chemical Co.) and pyridine was employed. Fourier transform
high-resolution mass spectra were recorded by matrix-assisted
laser desorption/ionization (MALDI HRMS) at The Scripps
Research Institute Center for Mass Spectrometry, La J olla,
CA. Elemental microanalyses were obtained from Atlantic
Microlab, Inc., Atlanta, GA. Molecular mechanics energy
(42) Hayakawa, H.; Haraguchi, K.; Tanaka, H.; Miyasaka, T. Chem.
Pharm. Bull. 1987, 35, 72-79.
(43) Tsuda, K.; Ohki, E.; Nozoe, S. J . Org. Chem. 1963, 28, 783-
785.

Synthesis of New Transglycosidically Tethered 5′-Nucleotides
J . Org. Chem., Vol. 67, No. 7, 2002 2157
with 10% MeOH/CH2Cl2. Rotary evaporation followed by
coevaporation with toluene gave, after drying in vacuo at 23
°C overnight, 4.57 g (97%) of pure 5 as a yellow gum. 1H NMR
(CDCl3) δ 9.65 (s, 1), 9.63 (s, 1), 8.21 (s, 1), 6.58 (s, 1), 6.47 (s,
1), 6.23 (two s, both 1), 6.00 (d, J ) 3.0 Hz, 1), 6.94 (d, J ) 3.0
Hz, 1), 5.19 (m, 1), 5.01 (m, 1), 4.89 (m, 1), 4.33 (pseudo-q, 1),
4.11 (pseudo-q, 1), 4.20 (d, J ) 12.4 Hz, 1), 3.41 (d, J ) 12.4
Hz, 1), 3.42 (s, 3), 3.33 (s, 3), 0.91 (s, 9), 0.09 (s, 6). Anal. Calcd
for C18H28N2O8Si: C, 50.45; H, 6.59; N, 6.54. Found: C, 52.25;
H, 7.35; N, 7.40 and C, 52.14; H, 7.27; N, 7.38.
6-Hyd r oxym eth yl-2′,3′-O-(m eth oxym eth ylid en e)-5′-O-
(ter t-bu tyld im eth ylsilyl)u r id in e (6). A solution of 3.965 g
(9.25 mmol) of 5 in 275 mL of MeOH was treated portionwise
with 770 mg (20 mmol, 2.2 equiv) of NaBH4, and the reaction
mixture was stirred for 1 h at 23 °C. A 5.35 g sample of NH4-
Cl was added, and after it had dissolved, the reaction mixture
was evaporated to near dryness at 35 °C. The residue was
treated with 250 mL of EtOAc and 20 mL of water; the layers
were separated, and the aqueous phase was washed with an
additional 20 mL of EtOAc. The combined organic phases were
dried (Na2SO4) and rotary evaporated to give 3.94 g (99%) of
6 as a yellow powder/foam: 1H NMR (CDCl3) δ 5.96 (s, 1), 5.92
(d, J ) 3.0 Hz, 1), 5.89 (s, 2), 5.82 (s, 1), 5.80 (d, J ) 3.0 Hz,
1), 5.38 (m, 2), 5.00 (m, 1), 4.88 (m, 1), 4.52 (s, 2), 4.32 (pseudo-
q, 1), 4.10 (pseudo-q, 1), 3.84 (m, 2), 3.39 (s, 3), 3.31 (s, 3), 0.90
(s, 9), 0.08 (s, 6). Anal. Calcd for C18H30N2O8Si: C, 50.22; H,
7.02; N, 6.51. Found: C, 52.20; H, 7.49; N, 7.32 and C, 52.08;
H, 7.61; N, 7.32.
6-[(9-F lu or en ylm eth yloxyca r bon yl)oxym eth yl]-2′,3′-O-
(m eth oxym eth ylid en e)-5′-O-(ter t-bu tyld im eth ylsilyl)u r i-
d in e (7). A solution of 3.38 g (7.5 mmol) of 6 in 30 mL of
anhydrous C5H5N under argon was treated with 2.6 g (10
mmol) of 9-fluorenylmethyl chloroformate and the reaction
mixture kept at 23 °C overnight. The bulk of the C5H5N was
removed by rotary evaporation in vacuo at <35 °C, and the
residue was twice treated with 50 mL portions of toluene and
rotary evaporated. The residue was then dissolved in 300 mL
of CH2Cl2, and the organic solution was extracted five times
with 50 mL portions of water and then dried (Na2SO4). The
solution was concentrated to about 30 mL by rotary evapora-
tion, and then it was applied to a short pad of SiO2 that had
been preequilibrated with CH2Cl2 in a sintered glass funnel.
Elution with CH2Cl2 removed the excess Fmoc reagent, and
then elution with 5% MeOH/CH2Cl2 gave 4.37 g (90%) of 7 as
a yellow foam/powder: 1H NMR (CDCl3) δ 7.76 (d, J ) 7.7
Hz, 2), 7.60 (d, J ) 7.7 Hz, 2), 7.39 (pseudo-t, 2), 7.31 (pseudo-
t, 2), 5.98 (s, 1), 5.86 (s, 1), 5.83 (s, 2), 5.77 (s, 1), 5.62 (s, 1),
5.31 (t, J ) 5 Hz, 1), 5.05 (m, 2), 4.89 (m, 1), 4.46 (d, J ) 5 Hz,
2), 4.32 (pseudo-q, 1), 4.25 (pseudo-q, 1), 3.82 (m, 2), 3.31 (s,
3), 3.29 (s, 3), 0.90 (s, 9), 0.06 (s, 6). Anal. Calcd for C33H40N2O10-
Si: C, 60.72; H, 6.18; N, 4.29. Found: C, 59.84; H, 6.71; N,
5.38 and C, 59.93; H, 6.62; N, 5.34.
6-[(9-F lu or en ylm eth yloxyca r bon yl)oxym eth yl]-2′,3′-O-
(m eth oxym eth ylid en e)u r id in e (8). A solution of 2.47 g (3.8
mmol) of 7 in 100 mL of MeOH was treated with 2.5 g (67
mmol) of NH4F and then heated at reflux for 1 h. TLC (5%
MeOH/CH2Cl2) indicated that 7 had been consumed, so the
reaction mixture was allowed to cool to 23 °C and then rotary
evaporated onto 35 g of SiO2 at <23 °C. This was pumped dry
in vacuo and then placed on top of 35 g of SiO2 in a sintered
glass funnel. The SiO2 was eluted first with pure CH2Cl2 to
remove the trace of an Fmoc-related compound that had
formed and then with 5% MeOH/CH2Cl2 to give, after drying
in vacuo overnight, 1.17 g (57%) of 8 as a pale orange solid/
foam: 1H NMR (CDCl3) δ 7.77 (d, J ) 7.7 Hz, 2), 7.61 (d, J )
7.7 Hz, 2), 7.42 (pseudo-t, 2), 7.32 (pseudo-t, 2), 5.99 (s, 1),
5.90 (s, 1), 5.84 (s, 1), 5.72 (d, J ) 3.0 Hz, 1), 5.67 (d, J ) 3.0
Hz, 1), 5.55 (d, J ) 5.0 Hz, 1), 5.34 (m, 1), 5.05 (s, 2), 4.48 (m,
2), 4.38 (pseudo-q, 1), 4.27 (t, J ) 5.0 Hz, 1), 4.18 (pseudo-q,
1), 3.90 (d, J ) 12 Hz, 1), 3.82 (d, J ) 12 Hz, 1), 3.30 (s, 3),
3.29 (s, 3). Anal. Calcd for C27H26N2O10: C, 60.22; H, 4.87; N,
5.20. Found: C, 61.16; H, 5.29; N, 5.37 and C, 61.31; H, 5.35;
N, 5.35.
1.44 g (3.4 mmol) sample of the Dess-Martin periodinane was
added to 12.5 mL of anhydrous CH2Cl2 under argon. After
stirring for 30 min, the mixture was first treated with 167 µL
(2.1 mmol) of anhydrous pyridine and then treated dropwise
with a solution of 1.17 g (2.17 mmol) of 8 in 12.5 mL of
anhydrous CH2Cl2. Stirring was continued for 2 h at 23 °C, at
which time TLC (5% MeOH/CH2Cl2) revealed that 9, identical
in Rf value to 8 but 2,4-DNP-positive, had formed. The mixture
was diluted to 125 mL with CH2Cl2 and treated with 60 mL
of saturated aqueous NaHCO3 and 12.5 mL of saturated
aqueous Na2S2O3. After the mixture was stirred rapidly for
30 min, the layers were separated and the aqueous layer was
extracted twice with CH2Cl2 (25 mL). The combined organic
phases were back-extracted with water and then dried (Na2-
SO4). Rotary evaporation gave 1.04 g (89%) of 9 as a pale
yellow foam/powder: 1H NMR (CDCl3) δ 9.42 (s, 1), 9.40 (s,
1), 7.77 (d, J ) 7.7 Hz, 2), 7.61 (d, J ) 7.7 Hz, 2), 7.42 (pseudo-
t, 2), 7.32 (pseudo-t, 2), 5.91 (s, 1), 5.90 (s, 1), 5.86 (s, 1), 5.84
(s, 1), 5.82 (s, 1), 5.75 (m, 2), 5.35 (m, 1), 5.10 (m, 1), 5.05 (m,
1), 4.48 (m, 2), 4.34 (pseudo-q, 1), 4.26 (m, 2), 3.31 (s, 3), 3.28
(s, 3). Anal. Calcd for C27H24N2O10: C, 60.45; H, 4.51; N, 5.22.
Found: C, 59.57; H, 4.88; N, 5.44 and C, 59.50; H, 4.82; N,
5.39.
6-(Hyd r oxym eth yl)-2′,3′-O-(m eth oxym eth ylid en e)u r i-
d in e 5′-Ca r b oxa ld eh yd e (10). A solution of 420 mg (0.78
mmol) of 9 in 40 mL of absolute EtOH was cooled to 0 °C, and
then anhydrous NH3 was bubbled in until an excess had
clearly been delivered. The solution was kept at 4 °C overnight
and then was allowed to warm to 23 °C for 1.5 h. Rotary
evaporation at 23 °C gave a residue that was treated with 25
mL of toluene and rotary evaporated again to dryness at 23
°C. The residue (357.5 mg after it was placed in vacuo at 23
°C) was triturated three times with 25 mL portions of
anhydrous Et2O to remove the Fmoc-NH2 byproduct. This solid
remaining was placed in vacuo next to P2O5 over several days,
giving 203 mg (83%) of 10a as a white nonhygroscopic powder.
A 1.0 g sample of 9 treated similarly but purified by radial
chromatography (10 then 12.5% MeOH/CH2Cl2 as eluents)
after Et2O trituration gave 447 mg (76%) of 10, which exists
as a mixture of diastereomeric cyclic hemiacetals in anhydrous
aprotic solvents, but as a mixture of diastereomeric methyl
hemiacetals in MeOH: 1H NMR (CD3OD) δ 5.99 (d, J ) 4 Hz),
5.93 (s, 1), 5.89 (d, J ) 4 Hz, 1), 5.82 (s, 1), 5.28 (m, 1), 5.00
(m, 1), 4.65 (m, 1), 4.52 (d, J ) 12 Hz, 1), 4.22 (d, J ) 12 Hz,
1), 3.42 (s, 3), 3.28 (s, 3); 13C NMR (CD3OD) δ 166, 157, 152,
120, 119, 102, 99, 97, 94, 92, 86, 85, 83, 64, 61, 55, 52, 51, 18,
15; MALDI HRMS calcd for C12H15N2O8 (MH)+ 315.0823, found
315.0823. Anal. Calcd for C12H14N2O8: C, 45.86; H, 4.49; N,
8.91. Found: C, 46.46; H, 5.02; N, 8.29 and C, 46.42; H, 5.01;
N, 8.35.
6-[Di-(2,2,2-tr ich lor oeth yl)p h osp h or yloxym eth yl]-2′,3′-
O-(m eth oxym eth yliden e)u r idin e 5′-Car boxaldeh yde. This
untethered constitutional isomer of 11 was obtained when we
removed the Fmoc group of 10 under strictly anhydrous,
aprotic conditions32 and then immediately phosphorylated the
nucleoside produced. It gave valuable 1H NMR and mass
spectral data for comparison to those of the desired tethered
isomer 11. A solution of 268 mg (0.5 mmol) of 10 in 2.5 mL of
anhydrous pyridine under argon was treated with 70 µL (51
mg, 0.5 mmol) of CaH2-dried Et3N, and the reaction mixture
was stirred at 23 °C for 1 h. It was cooled to 4 °C in an ice
bath and then treated with 284 mg (0.75 mmol) of bis(2,2,2-
trichloroethyl)phosphorochloridate. The reaction mixture was
kept at 0 °C for 20 h, and then it was diluted with 25 mL of
CH2Cl2 and vigorously stirred with 10 mL of water for 2 h.
The layers were separated, and the organic phase was
extracted four times with water (10 mL) and dried (Na2SO4).
Rotary evaporation gave a residue that was twice coevaporated
with 50 mL of toluene to remove residual pyridine. The residue
was pumped free of volatiles at 23 °C for 1 h, and then it was
dissolved in ca. 3 mL of CH2Cl2. This solution was added to 5
g of SiO2 in a sintered glass funnel, and this was then eluted
first with pure CH2Cl2 to remove dibenzofulvene and then with
5% MeOH/CH2Cl2 to give a mixture of nucleoside products.
Rotary evaporation gave 170 mg of a pale yellow foam that
6-[(9-F lu or en ylm eth yloxyca r bon yl)oxym eth yl]-2′,3′-O-
(m eth oxym eth ylid en e)u r id in e 5′-Ca r boxa ld eh yd e (9). A

2158 J . Org. Chem., Vol. 67, No. 7, 2002
Groziak and Thomas
was pumped dry in vacuo at 23 °C overnight. TLC (5% MeOH/
CH2Cl2) revealed three components, and separation by radial
chromatography (2.5% MeOH/CH2Cl2 as the eluent) followed
by ascending preparative chromatography (4% MeOH/CH2Cl2)
gave 36 mg of 6-[di(2,2,2-trichloroethyl)phosphoryloxymethyl]-
2′,3′-O-(methoxymethylidene)uridine 5′-carboxaldehyde, char-
to a 10% concentration, and the 1H NMR spectrum was
recorded every 15 min to monitor the hydrolysis of the
methoxymethylidene group. Within 1.5 h, the hydrolysis had
proceeded to completion, giving a 2′(3′)-O-formylated form of
1, by NMR. The solution was diluted with water and rotary
evaporated, and the residue was twice coevaporated with
toluene to remove water and acetic acid. A solution of this
residue in 5 mL of MeOH was heated on a steam bath for 30
min to effect the solvolysis of the formate ester. Rotary
evaporation afforded 13.4 mg (estimated 90%) of 1 as a white
solid contaminated only by NH4OAc. By 1H NMR, 1 was
obtained as a 1:1 mixture of diastereomers (epimeric phos-
phorylated hemiacetals): 1H NMR (D2O) δ 6.54 (d, J ) 4.6
Hz, 1), 6.28 (d of d, J ) 5, 2.7 Hz, 1), 6.02 (pseudo-t, 1), 5.91
(s, 1), 5.86 (s, 1), 5.75 (m, 1), 5.58 (d, J ) 7.7 Hz, 1), 5.49 (m,
1), 5.13 (m, 1), 4.68-4.48 (m, 4), 4.40 (m, 1), 4.30 (m, 1). In
D2O solution at 23 °C, the 5′,6-oxomethylene-tethered 5′-UMP
1 remains intact for at least 2 months, by 1H NMR (see
Supporting Information). It also remains intact after adsorp-
tion onto activated charcoal for 3 days followed by removal
with 50% aqueous EtOH containing 4% NH4OH, by DCI MS.
MALDI HRMS calcd for C10H14N2O11P (M - H)- 351.0235,
found 351.0228.
3′,5′-Di-O-(ter t-bu tyld im eth ylsilyl)-2′-d eoxyu r id in e-6-
ca r boxa ld eh yd e (14). Although 3′,5′-di-O-(tert-butyldimeth-
ylsilyl)-2′-deoxyuridine (13) has been reported to resist double
deprotonation by LDA,39 we found that a Miyasaka uridin-
3,6-diyl dilithium species of this nucleoside could be generated
by exposure to 5 equiv of this base: a solution of 3′,5′-di-O-
(tert-butyldimethylsilyl)uridine (7.17 g, 15.7 mmol, prepared
in 80% purified yield from uridine) in 150 mL of anhydrous
THF at -78 °C under argon was transferred dropwise over
30 min via positive argon pressure through a cannula into a
solution of 39 mL of 2 M LDA (78 mmol, 5 equiv) in 150 mL of
anhydrous THF at -78 °C under argon. The reaction mixture
was stirred at -78 °C for 3.5 h, and then 4.85 mL (78 mmol,
5 equiv) of HCO2Me was added dropwise. This was slightly
exothermic, and some gas evolution was observed. This was
likely CdO generated by LDA deprotonation of the excess
HCO2Me. The mixture was stirred at -78 °C for 5 h, and then
it was quenched by the dropwise addition of 4.5 mL (78 mmol,
5 equiv) of AcOH. The cooling bath was removed and the
mixture allowed to warm to 23 °C under argon overnight. The
volatiles were removed by rotary evaporation at less than 30
°C, and the residue was treated with a mixture of EtOAc (300
mL) and water (100 mL). The layers were separated, and the
organic layer was dried (Na2SO4). TLC (5% MeOH/CH2Cl2)
showed the presence of nearly equal amounts of the starting
material (Rf 0.51) and the 2,4-DNP-positive 14 (Rf 0.38). Rotary
evaporation gave 11 g of the mixture, which was separated
by careful column chromatography (5% MeOH/CH2Cl2), giving
3.42 g (45%) of 14 as a white foam/solid. The identity of 14 as
the 6-formyl regioisomer is supported by its NMR spectral
characteristics and by its utility in obtaining the final target:
1H NMR (CDCl3) δ 10.0 (s, 1), 9.4 (bs, 1), 6.37 (t, J ) 7.3 Hz,
1), 6.06 (s, 1), 4.48 (m, 1), 3.90 (m, 1), 3.83-3.70 (m, 2), 2.37
(m, 1), 2.28 (m, 1), 0.90 (s, 18), 0.08 (s, 12); 13C NMR (CDCl3)
δ 184.3, 162.4, 150.0, 148.6, 106.1, 88.2, 86.2, 71.1, 62.0, 41.4,
25.8, 18.3, 17.9, -4.8, -5.7. Anal. Calcd for C22H40N2O6Si2: C,
54.51; H, 8.32; N, 5.78. Found: C, 54.97; H, 8.24; N, 6.50 and
C, 55.08; H, 8.36; N, 6.39.
1
acterized by H NMR and DCI MS. Similar treatment of a 72
mg sample of 10 gave, after radial chromatographic separation,
30 mg of this same nucleoside: 1H NMR (CDCl3) δ 9.43 (s, 1),
9.41 (s, 1), 6.00-5.82 (m, 2), 5.30 (m, 1), 5.15 (m, 1), 4.72 (m,
2), 3.49 (s, 3), 3.39 (s, 3). The DCI mass spectrum (data
provided in Table 1 in Supporting Information) revealed a
diagnostic pair of peaks for the [B + H+] and [B + NH4
]
+
species at 483 and 500 amu, respectively, consistent with the
untethered nature of this undesired constitutional isomer.
Base peaks like these were absent, as anticipated, in the DCI
mass spectrum of the desired, tethered regioisomer 11 as well
as in those of 12 and 1 (derived from 11) and in those of their
counterparts 21 and 2 in the tethered 5′-dUMP synthesis.
5′,6-Oxom et h ylen e-2′,3′-O-(m et h oxym et h ylid en e)u r i-
d in e 5′-Mon op h osp h a t e, Di(2,2,2-t r ich lor oet h yl) E st er
(11). The Shiba procedure33 was employed: a solution of 72
mg (0.225 mmol) of rigorously dried 10 in 2 mL of anhydrous
THF under argon was cooled to -78 °C and treated dropwise
with 312 µL of 1.6 M BuLi in hexanes (0.5 mmol, 2.2 equiv).
The orange suspension of dianion was stirred at -78 °C for 5
min, and then 128 mg (0.338 mmol, 1.5 equiv) of bis(2,2,2-
trichloroethyl)phosphorochloridate was added all at once. The
reaction mixture was stirred for 15 min at -78 °C, during
which time the suspension became less yellow and cloudy. The
cooling bath was removed, and the reaction was quenched by
the addition of a mixture of 11 mL of EtOAc and 5.5 mL of
saturated aqueous NaHCO3. The mixture was stirred rapidly;
then, the layers were separated, and the organic phase was
washed with saturated aqueous NaCl and dried (Na2SO4). TLC
(5% MeOH/CH2Cl2) indicated a trace of 10 (Rf 0.22) together
with four higher Rf value components, one of them major (11,
Rf 0.57). Separation by radial chromatography (5% MeOH/CH2-
Cl2 as the eluent) gave 23 mg (15%) of 11 as a pale yellow
oil/foam. Similar treatment of a 63 mg sample of 10 gave, again
after radial chromatographic separation, 26 mg (20%) of 11:
1H NMR (CDCl3) δ 9.27 (bs, 1), 9.04 (bs, 1), 6.77 (s, 1), 6.67 (s,
1), 6.52 (s, 1), 6.43 (s, 1), 5.99 (s, 1), 5.90 (s, 1), 5.58 (m, 2),
5.09 (m, 1), 4.87 (m, 1), 4.71 (m, 2), 4.45 (m, 1), 3.44 (s, 3),
3.38 (s, 3), 3.33 (s, 3), 3.30 (s, 3).
5′,6-Oxom et h ylen e-2′,3′-O-(m et h oxym et h ylid en e)u r i-
d in e 5′-Mon op h osp h a te (12). A slight modification of the
Imai-Torrence refinement38 of the Eckstein-Scheit proce-
dure35 was employed: a solution of the 19 mg (0.03 mmol) of
11 in 0.6 mL of anhydrous DMF was treated with 30 µL of
acetylacetone and 39 mg (0.6 mmol) of Zn(Cu) couple that had
been freshly prepared with a 16% Cu content. The mixture
was stirred vigorously for 6 h at 23 °C during which time the
Zn(Cu) couple progressively disappeared. No green color
developed, but TLC (5% MeOH/CH2Cl2) showed that the
deprotection had proceeded smoothly. The solution was then
diluted with 18 mL of MeOH and 9 mL of water; Chelex resin
(washed first with 0.1 N NH4HCO3 and then with water; 15
mL settled volume) was added, and the mixture was stirred
for 1 h. The resin was removed by suction filtration through
Celite, and the Celite was washed with a small amount of
water. The filtrate and washing were combined and rotary
evaporated to dryness in vacuo at <40 °C. The residue was
coevaporated from absolute EtOH several times, and then it
was dried in vacuo overnight to give 13 mg (99%) of 12 as an
oil: 1H NMR (D2O) δ 6.69 (s, 1), 6.59 (s, 1), 6.54 (s, 1), 6.44 (s,
1), 6.20 (s, 1), 6.09 (s, 1), 5.87 (s, 1), 5.82 (s, 1), 5.69 (m, 1),
5.55 (m, 1), 5.13 (m, 1), 5.07 (m, 1), 4.33 (m, 1), 4.27 (m, 1),
4.21 (m, 1), 3.52 (s, 3), 3.50 (s, 3), 3.44 (s, 3), 3.43 (s, 3). MALDI
HRMS calcd for C12H14N2O11P (M - H)- 393.0341, found
393.0349.
6-Hyd r oxym et h yl-3′,5′-d i-O-(ter t-b u t yld im et h ylsilyl)-
2′-d eoxyu r id in e (15). Use of the procedure for the reduction
of 5 to 6 allowed reduction of 14, affording 15 in 99% yield as
an off-white foam/solid: 1H NMR (CDCl3) δ 9.11 (bs, 1), 6.26
(t, J ) 7.2 Hz, 1), 5.79 (s, 1), 4.51 (s, 2), 3.90 (m, 1), 3.81 (m,
2), 2.64 (m, 1), 2.17 (m, 1), 0.92 (s, 9), 0.89 (s, 9), 0.09 (s, 6),
0.07 (s, 6). Anal. Calcd for C22H42N2O6Si2: C, 54.29; H, 8.70;
N, 5.76. Found: C, 55.97; H, 8.99; N, 6.32 and C, 55.85; H,
8.89; N, 6.33.
6-[(9-F lu or en ylm eth yloxyca r bon yl)oxym eth yl]-3′,5′-d i-
O-(ter t-bu tyld im eth ylsilyl)-2′-d eoxyu r id in e (16). A solu-
tion of 1.79 g (4.18 mmol) of 15 in 15 mL of anhydrous pyridine
was treated with 1.24 g (4.78 mmol) of 9-fluorenylmethyl
chloroformate, and the reaction mixture was allowed to stand
5′,6-Oxom eth ylen eu r idin e 5′-Mon oph osph ate (5′,6-Oxo-
m eth ylen e-Teth er ed 5′-UMP , 1). The 13 mg (0.03 mmol)
sample of 12 from above was dissolved in 0.5 mL of D2O and
its 1H NMR spectrum recorded. Then, CD3CO2D was added

Synthesis of New Transglycosidically Tethered 5′-Nucleotides
J . Org. Chem., Vol. 67, No. 7, 2002 2159
at 23 °C overnight. Toluene was added and the solution rotary
evaporated in vacuo at <35 °C. The residue was coevaporated
twice more using 90 mL of toluene; then, it was dissolved in
120 mL of EtOAc, and the solution was extracted three times
with water (15 mL) and then dried (Na2SO4). Rotary evapora-
tion gave a residue that was purified by applying a concen-
trated solution of it in 50% CH2Cl2/hexanes to a short pad of
SiO2 and eluting first with 50% CH2Cl2/hexanes to remove
residual Fmoc material and then with pure CH2Cl2 to afford
1.84 g (62%) of 16 as a pale yellow foam/solid: 1H NMR (CDCl3)
δ 8.13 (bs, 1), 7.78 (d, J ) 7.5 Hz, 2), 7.60 (d, J ) 7.4 Hz, 2),
7.43 (pseudo-t, 2), 7.33 (pseudo-t, 2), 6.18 (t, J ) 7.2 Hz, 1),
5.86 (s, 1), 5.17 (d, J ) 15 Hz, 1), 5.05 (d, J ) 15 Hz, 1), 4.54-
4.46 (m, 3), 4.26 (t, J ) 7.5 Hz, 1), 3.90 (m, 1), 3.77 (m, 3),
2.74 (m, 1), 2.17 (m, 1), 0.89 (s, 9), 0.88 (s, 9), 0.08 (s, 6), 0.06
(s, 6). Anal. Calcd for C37H52N2O8Si2: C, 62.68; H, 7.39; N, 3.95.
Found: C, 62.55; H, 7.44; N, 4.13.
6-[(9-F lu or en ylm et h yloxyca r b on yl)oxym et h yl]-3′-O-
(ter t-bu tyld im eth ylsilyl)-2′-d eoxyu r id in e (17). A solution
of 1.35 g (1.9 mmol) of 16 in 7 mL of absolute EtOH was
treated with 108 mg (catalytic) of PPTS, and the reaction
mixture was allowed to stand at 23 °C overnight. The mixture
was kept at 23 °C for 16 h, after which time TLC (5% MeOH/
CH2Cl2) showed that most of the 16 had reacted. The mixture
was rotary evaporated, and the residue was purified first by
column chromatography (5% MeOH/CH2Cl2) in a sintered glass
funnel and then by radial chromatography (4% MeOH/CH2-
Cl2 as the eluent) to afford 875 mg (77%) of 17 as a pale yellow
foam/solid: 1H NMR (CDCl3) δ 8.62 (bs, 1), 7.77 (d, J ) 7.5
Hz, 2), 7.59 (d, J ) 7.4 Hz, 2), 7.43 (pseudo-t, 2), 7.33 (pseudo-
t, 2), 5.87 (s and t, J ) 7.2 Hz, each 1), 5.05 (s, 2), 4.67 (s, 1),
4.55-4.44 (m, 3), 4.26 (t, J ) 7.5 Hz, 1), 3.95-3.65 (m, 4), 3.29
(m, 1), 2.93 (m, 1), 2.08 (m, 1), 0.88 (s, 9), 0.08 (s, 6). Anal.
Calcd for C31H38N2O8Si: C, 62.61; H, 6.44; N, 4.71. Found: C,
61.80; H, 6.87; N, 4.63.
6-[(9-F lu or en ylm et h yloxyca r b on yl)oxym et h yl]-3′-O-
(ter t-bu tyld im eth ylsilyl)-2′-d eoxyu r id in e 5′-Ca r boxa ld e-
h yd e (18). A 975 mg (2.3 mmol) sample of the Dess-Martin
periodinane was added to 8.5 mL of anhydrous CH2Cl2 under
argon. After stirring for 30 min, the mixture was first was
treated dropwise with a solution of 875 mg (1.47 mmol) of 17
in 8.5 mL of anhydrous CH2Cl2. Stirring was continued for 2
h at 23 °C. The mixture was diluted to 85 mL with CH2Cl2
and treated with 40 mL of saturated aqueous NaHCO3 plus
8.5 mL of saturated aqueous Na2S2O3. After the mixture was
stirred rapidly for 30 min, the layers were separated and the
aqueous layer was extracted twice with CH2Cl2 (17 mL). The
combined organic phases were back-extracted with water and
then dried (Na2SO4). Rotary evaporation gave 850 mg (98%)
of 18 as a pale yellow foam/solid: 1H NMR (CDCl3) δ 9.59 (s,
1), 8.32 (bs, 1), 7.77 (d, J ) 7.5 Hz, 2), 7.60 (d, J ) 7.4 Hz, 2),
7.43 (pseudo-t, 2), 7.33 (pseudo-t, 2), 5.86 (s and t, J ) 7.2 Hz,
each 1), 5.05 (s, 2), 4.67 (s, 1), 4.60-4.45 (m, 3), 4.25 (m, 1),
4.00-3.75 (m, 2), 3.29 (m, 1), 3.05 (m, 1), 2.01 (m, 1), 0.90 (s,
9), 0.12 (s, 6). Anal. Calcd for C31H36N2O8Si: C, 62.82; H, 6.12;
N, 4.73. Found: C, 62.26; H, 6.72; N, 4.53.
6-Hyd r oxym eth yl-3′-O-(ter t-bu tyld im eth ylsilyl)-2′-d e-
oxyu r id in e 5′-Ca r boxa ld eh yd e (19). A suspension of 850
mg (1.43 mmol) of 18 in 75 mL of absolute EtOH was cooled
to 0 °C, and then anhydrous NH3 was bubbled in until an
excess had clearly been delivered. Most of the 18 dissolved
within ca. 20 min, and the solution was kept at 4 °C overnight
and then allowed to warm to 23 °C over 1.5 h. Rotary
evaporation at 23 °C gave a residue that was purified by
column chromatography using first CH2Cl2 to remove Fmoc
byproduct and then 5 and then 10% MeOH/CH2Cl2 as eluents
to give 454 mg (86%) of 19 as a white powder. This was
coevaporated twice with toluene, pumped dry overnight,
triturated with a small amount of anhydrous Et2O, and
recrystallized from a small amount of absolute EtOH: mp
216-219 °C dec. Like 10, nucleoside 19 was found to exist as
an equal mixture of diastereomeric cyclic hemiacetals in
anhydrous aprotic solvents: 1H NMR [(CD3)2SO] δ 11.4 (bs,
1), 7.42 (d, J ) 5.1 Hz, 1), 7.01 (d, J ) 4.8 Hz, 1), 6.58 (t, J )
7.2 Hz, 1), 6.41 (t, J ) 7.2 Hz, 1), 5.64 (s, 1), 5.62 (s, 1), 5.48
(d, J ) 15 Hz, 1), 5.00 (d, J ) 4.5 Hz, 1), 4.95 (t, J ) 5 Hz, 1),
4.85 (d, J ) 16 Hz, 1), 4.56 (s, 1), 4.55 (d, J ) 16 Hz, 1), 4.52
(s, 1), 4.48 (m, 2), 4.32 (m, 2), 4.06 (s, 1), 3.95 (s, 1), 3.92 (d, J
) 15 Hz, 1), 3.43 (m, 2), 2.62 (m, 1), 2.22 (m, 1), 0.87 (s, 9),
0.08 (s, 6). Anal. Calcd for C16H26N2O6Si: C, 51.87; H, 7.07;
N, 7.56. Found: C, 49.26; H, 6.71; N, 7.19 and C, 49.23; H,
6.69; N, 7.25.
6-[Di(2,2,2-Tr ich lor oet h yl)p h osp h or yloxym et h yl]-2′-
d eoxy-3′-O-(ter t-b u t yld im et h ylsilyl)u r id in e 5′-Ca r b ox-
a ld eh yd e (20). A solution 37 mg (0.1 mmol) of rigorously dried
19 in 1.0 mL of anhydrous THF under argon was cooled to
-78 °C and treated dropwise with 134 µL of 1.67 M BuLi in
hexanes (0.22 mmol, 2.2 equiv). The very thick orange dianion
suspension was stirred with difficulty at -78 °C for 5 min,
and then 57 mg (0.15 mmol, 1.5 equiv) of bis(2,2,2-trichloro-
ethyl)phosphorochloridate was added all at once. The reaction
mixture was stirred for 15 min at -78 °C, during which time
the color lightened to yellow and the suspension became
thinner. The cooling bath was removed, and the reaction was
quenched immediately by the addition of a mixture of 5.5 mL
of EtOAc and 2.75 mL of saturated aqueous NaHCO3. The
mixture was stirred rapidly; then, the layers were separated,
and the organic phase was washed once with saturated
aqueous NaCl and dried (Na2SO4). Separation by radial
chromatography (5% MeOH/CH2Cl2 as the eluent) gave 10 mg
of starting 19 and 10 mg (20% on the basis of unrecovered 19)
of a 5:1 mixture of 20 diastereomers as a pale yellow oil/foam:
1H NMR (CDCl3) δ 8.6 (bs, 1), 6.86 (t, J ) 7.2 Hz, 1), 6.60 (t,
J ) 7.2 Hz, 1), 5.70 (m, 1), 5.60-5.50 (m, 2), 5.00 (d, J ) 15
Hz, 1), 4.75-4.60 (m, 6), 4.40-4.20 (m, 2), 2.60 (m, 1), 2.30
(m, 1), 0.95 (s, 9), 0.89 (s, 9), 0.09 (s, 6), 0.08 (s, 6).
5′,6-Oxom et h ylen e-2′-d eoxy-3′-O-(ter t-b u t yld im et h yl-
silyl)u r id in e 5′-Mon op h osp h a te (21). Use of the procedure
for the deprotection of 11 to 12 allowed deprotection of 20,
affording 21 in 99% yield as a pale yellow oil/foam: 1H NMR
(D2O) δ 6.86 (t, J ) 7.2 Hz, 1), 6.68 (t, J ) 7.2 Hz, 1), 5.88 (s,
1), 5.83 (s, 1), 5.38 (m, 1), 5.12 (d, J ) 15.5 Hz, 1), 4.94 (s, 1),
4.93 (d, J ) 15.5 Hz, 1), 4.44 (m, 2), 2.69 (m, 1), 2.47 (m, 2),
0.98 (s, 9), 0.96 (s, 9), 0.26 (s, 6), 0.21 (s, 6). MALDI HRMS
calcd for C16H26N2O9PSi (M - H)- 449.1151, found 449.1169.
5′,6-Oxom eth ylen e-2′-d eoxyu r id in e 5′-Mon op h osp h a te
(5′,6-Oxom eth ylen e-Teth er ed 5′-d UMP , 2). A modification
of the procedure for the hydrolysis of 12 to 1 was used. Thus,
a solution of 21 in 10% aqueous AcOH was allowed to stand
at 23 °C for 16 h and then rotary evaporated to dryness. The
residue was twice coevaporated with toluene to remove water
and acetic acid to afford 10.3 mg of 2 as an off-white foam/
solid contaminated only by NH4OAc. By 1H NMR, 2 was
obtained as a 5:1 mixture of diastereomers (epimeric phos-
phorylated hemiacetals): 1H NMR (D2O) δ 6.82 (t, J ) 7.2 Hz,
1), 6.69 (t, J ) 7.2 Hz, 1), 5.88 (s, 1), 5.84 (s, 1), 5.43 (m, 1),
5.12 (d, J ) 15.5 Hz, 1), 4.61 (d, J ) 15.5 Hz, 1), 4.44 (s, 2),
2.65 (m, 1), 2.48 (m, 1). In D2O solution at 23 °C, the 5′,6-
oxomethylene-tethered 5′-dUMP 2 remains intact for at least
1
2 months, by H NMR (see Supporting Information). MALDI
HRMS calcd for C10H14N2O10P (M - H)- 335.0286, found
335.0286.
Ack n ow led gm en t. This research was supported by
NIH Grant R55-GM56878, a Shannon Director’s Award
from the National Institute of General Medical Sciences,
the National Institutes of Health.
Su p p or tin g In for m a tion Ava ila ble: Tabular DCI mass
spectral data for 1-21; DCI mass spectral plots of 12, 1, 21,
1
and 2; and periodic H NMR spectra of 1 and 2 kept in D2O
solution at 23 °C This material is available free of charge via
the Internet at http://pubs.acs.org.
J O0110045