The 5',6-oxomethylene transglycosidic tether for conformational restriction of pyrimidine ribonucleosides. Investigation of 6-formyl- and 6-(hydroxymethyl)uridine 5'-carboxaldehydes

Article abstract of DOI:10.1016/S0040-4020(00)00970-4

In an effort to develop a new motif for the transglycosidic tethering of the pyrimidine nucleoside framework, the 2',3'-O-isopropylidenated and unprotected versions of 6-formyl- and 6-(hydroxymethyl)uridine 5'-carboxaldehyde were prepared and these were e

Full text of DOI:10.1016/S0040-4020(00)00970-4

TETRAHEDRON
Pergamon
Tetrahedron 56 (2000) 9885±9893
The 50,6-Oxomethylene Transglycosidic Tether for
Conformational Restriction of Pyrimidine Ribonucleosides.
Investigation of 6-Formyl- and 6-(Hydroxymethyl)uridine
50-Carboxaldehydes
b
Michael P. Groziaka, and Ronghui Lin
*
aPharmaceutical Discovery Division, SRI International, 333 Ravenswood Avenue, Menlo Park, CA 94025-3493, USA
bDrug Discovery Research, R. W. Johnson Pharmaceutical Research Institute, 920 U.S. Route 202, P.O. Box 300, Raritan,
NJ 08869-0602, USA
Received 11 July 2000; accepted 13 October 2000
AbstractÐIn an effort to develop a new motif for the transglycosidic tethering of the pyrimidine nucleoside framework, the 20,30-O-
isopropylidenated and unprotected versions of 6-formyl- and 6-(hydroxymethyl)uridine 50-carboxaldehyde were prepared and these were
examined for their ability to adopt 50,6-oxomethylene tethered solution structures. In aqueous solution, the 20,30-O-isopropylidenated
nucleosides readily generated spiro-dihydrouridines via proximity-induced transglycosidic intramolecular reactions. In stark contrast,
their unprotected counterparts existed mainly as the untethered aldehyde hydrates. Based on these ®ndings, the 50,6-oxomethylene trans-
glycosidic tether appears to constitute a useful conformational restriction motif for the pyrimidine ribonucleoside framework, but only when
the 50-OH group is functionalized. q 2000 Elsevier Science Ltd. All rights reserved.
Introduction
carbohydrate hydroxyl groups, and (4) a design that is
readily applicable to both the pyrimidine and the purine
nucleoside frameworks. We have been examining new
tethering motifs that could meet all of these criteria.3 In
the present work, we focused attention on the 50,6-
oxomethylene tethering of pyrimidine ribonucleosides
and speci®cally sought to determine whether this
tether was compatible with an unprotected 50-hydroxyl
functionality.
Conformational restriction is a powerful means of selecting
for a subset of biologically active rotamers at the expense of
inactive ones to increase the effectiveness of a biochemical
probe or medicinal agent. In the case of the nucleosides, of
which uridine and its nucleotides constitute a biologically
important subgroup,1 there are just two freely rotating bonds
which determine much of the overall molecular topography.
These are the C40 ±C50 bond and the glycosidic one (C10 ±
N1 for pyrimidines and C10 ±N9 for purines). Transglyco-
sidic tethers that simultaneously restrict both of these have
been developed in the past, but efforts to date have afforded
compounds that are not very biomimetic and therefore are of
limited utility as biochemical probes.2 The desired features
of a truly useful transglycosidic nucleoside tethering motif
include (1) a restriction of the glycosidic bond rotation in its
most often biologically relevant anti conformation, (2) a
preservation of all of the natural, biorecognition-critical
hydrogen bonding capabilities of the heterocyclic aglycon,
(3) a retention of all of the natural, bio-functionalizable
Our previous investigations into the unusual structural
properties of the 6-formyluracil-based glycofuranosides4±6
led to the discovery that their formyl group is so highly
electrophilic that it imbues these nucleosides with certain
carbohydrate-like properties. In speci®c, several new trans-
glycosidically tethered cyclic hemiacetals were found to be
dominant structural forms in solution as well as in the solid
state. For example, the `free' aldehyde form of the parent
uridine-6-carboxaldehyde accounts for only 10% of the
material in (CD3)2SO solutionÐthe remainder is a single
50-cyclic hemiacetal diastereomer. In D2O, the aldehyde
cannot be detected at all by 1H NMR. Instead, in this
solvent this nucleoside exists as a 2:1 mixture of its hydrate
and the same 50-cyclic hemiacetal diastereomer that pre-
dominates in (CD3)2SO. Finally, 6-formyluridine crystal-
lizes out of water exclusively as a 50-cyclic hemiacetal
diastereomer.5
Keywords: nucleosides; uridines; aldehydes; cyclisation; acylals conforma-
tion; transglycosidic; tethered.
* Corresponding author. Tel.: 1650-859-6299; fax: 1650-859-3153;
e-mail: michael.groziak@sri.com
0040±4020/00/$ - see front matter q 2000 Elsevier Science Ltd. All rights reserved.
PII: S0040-4020(00)00970-4

9886
M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
Scheme 1. Reagents and conditions: (i) 2.5 equiv. LDA, 2788C, then HCO2Et, 2788C; (ii) TBAF, THF; (iii) NaBH4, EtOH; (iv) Dess±Martin periodinane.
by inducing a proximity effect.8 Indeed, we encountered the
unusual spiro-fused dihydrouridine 2a very early on in our
investigation of 1a,9 and can now with certainty attribute its
ready formation to the presence of the 20,30-O-isopropyl-
idene group.
The objectives of the present investigation, then, were to
develop a reasonable synthesis of the carbohydrate pro-
tected 6-formyluridine 50-carboxaldehyde 1a, to deprotect
it to 1b, and to characterize the solution structures
displayed by these two dialdehydesÐwhether they be the
spiro-dihydrouridines 2a,b, the transglycosidically tethered
monohydrates 3a,b, or simply the untethered dihydrates.
An additional objective was to prepare and study the
The preponderance of transglycosidically tethered struc-
analogously protected and unprotected versions of
tures displayed by the 6-formyluridines reveals that the
6-(hydroxymethyl)uridine
50-carboxaldehyde,
again
50,6-oxomethylene tether is energetically favored when
attached to a uridine nucleoside platform. With this as our
starting point, we began to explore variations on this tether
motif that could lead to more biochemically relevant
structures. In particular, we were interested in regaining
the 50-OH group that had been rendered `absent' during
the facile 50-cyclic hemiacetal formations encountered in
our earlier studies. Interestingly, the same facile covalent
hydrate formation that is endemic to the 6-formyluridines
has long been known to be a property of the nucleoside
50-carboxaldehydes as well,7 albeit to a lesser degree. We
reasoned that if both of these hydration-prone aldehydes
were present in the same molecule, then a single bridging
hydration might occur to produce a new transglycosidically
tethered structure possessing the desired 50-OH group (i.e. 3,
shown below). To explore this possibility, we chose to
prepare the 6-formyluridine 50-carboxaldehydes 1a,b. It
was essential that the 20,30-O-isopropylidene-equipped 1a
be studied in addition to its unprotected counterpart 1b,
since 20,30-O-methylidene-based protecting groups are
known to enhance transglycosidic reactivity in nucleosides
searching for 50-hydroxyl-containing structural forms (i.e.
hemiacetals 4a,b) among the solution structures.
Results and Discussion
Our ®rst synthesis of dialdehyde 1a was short but not
inef®cient. The previously reported7,10 20,30-O-isopropyli-
deneuridine 50-carboxaldehyde was converted to its DPI
(1,3-diphenylimidazolidin-2-yl) derivative, and this was
then subjected to Miyasaka's conditions for uridin-3,6-diyl
dilithium generation.11 Perhaps due to the bulk of the DPI
group, this dianion formation was problematic and a quench
with HCO2Et gave the 6-carboxaldehyde only in a very poor
yield. We did manage to obtain 1a by subsequent DPI group
removal, but we immediately sought other routes.
Our second route to 1a (Scheme 1) was based on the
expectation that each of the hydroxymethyl groups in
6-(hydroxymethyl)-20,30-O-isopropylideneuridine
(9)
could be oxidized to aldehydes in a single operation.
When an attempt to formylate the purported12 trianion of
20,30-O-isopropylideneuridine failed to provide 8, we
prepared it instead by desilylating (49%) 6-formyl-50-
O-TBDMS-20,30-O-isopropylideneuridine
(6),12
itself
obtained from the known6 protected uridine 5 by a Miyasaka
dianion-based formylation. The hydroxy-aldehyde 8,
accessed by others along a different route,13 is carbo-
hydrate-like in that it exists almost exclusively in cyclic
hemiacetal form in (CD3)2SO. Thus, not surprisingly, our
attempts to oxidize 8 directly to 1a were unsuccessful.
Aldehyde 8 gave the lactone 10 (52%) under Moffatt con-
ditions,14 but it was found to be inert to the Dess±Martin
periodinane oxidant.15
Therefore, we reduced nucleoside 8 to the diol 9, which we
subsequently found could also be prepared by reducing

M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
9887
Scheme 2. Reagents and conditions: (i) (PhNHCH2)2, cat. AcOH; (ii)
TBAF, THF; (iii) Dess±Martin periodinane; (iv) TsOH, Me2CO/CH2Cl2.
Scheme 3. Reagents and conditions: (i) NaBH4, EtOH; (ii) TBDPS-Cl,
imidazole, DMF; (iii) PPTS, EtOH; (iv) Dess±Martin periodinane, then
Na2S2O3; (v) TBAF, THF.
aldehyde 6 and then desilylating the product alcohol 7 (70%
overall yield). Diol 9 was oxidized with excess Dess±
Martin periodinane, and the reaction mixture was immedi-
ately separated by a silica gel chromatography employing
an alcohol-free eluent (Me2CO/CH2Cl2) to protect the
assured alcohol-reactive 1a. This gave a chromatograph-
ically homogeneous ternary mixture of 1a, 8 (as the
hemiacetal), and an unknown periodinane-functionalized
glycosidic reactions leading to 2a. Likely, it facilitates
these by rendering the C10 ±N1 and C40 ±C50 bonds closer
to coparallel from within an O40-exo (4E) ribosyl conforma-
tional form.16
The fact that dialdehyde 1b does not exist even partly in
bridged monohydrate form (3b) is in stark contrast to the
behavior of uridine-6-carboxaldehyde itself, which readily
adopts a 50-cyclic hemiacetal form in both solution and in
the solid state.4±6 We suspect that it is the 50-hemiacetal
portion of 3b which is particularly labile.
nucleoside, by 1H NMR. Exposure of this mixture to
9,10
Na2S2O3 in saturated aqueous NaHCO3
removed the
latter component, and another chromatographic separation
afforded a mixture of 1a and 8 (18% combined yield) and the
same C6 spiro-fused nucleoside 2a (7%) later discovered to
spontaneously arise from 1a (see below).
Turning to the 6-(hydroxymethyl)uridine 50-carboxalde-
hydes that might possibly exist as the 50-OH-equipped
cyclic hemiacetals 4a,b, we synthesized 6-(hydroxy-
methyl)-20,30-O-isopropylideneuridine 50-carboxaldehyde
(18) by the route shown in Scheme 3. Aldehyde 6 was
reduced with NaBH4 to the alcohol 14 (76%), and this
was then protected with a TBDPS group (86%) to give 15,
which in turn was regioselectively deprotected (97%) to the
alcohol 16 for a Dess±Martin periodinane-mediated oxida-
tion (80%) to aldehyde 17. Deprotection of 17 with TBAF
gave the 2,4-DNP-positive aldehyde 18 by TLC, but this
proved to have a ¯eeting existence.
Scheme 2 shows our third and best route to 1a. The aldehyde
6 from above was converted to its DPI derivative (73%) and
the resulting protected nucleoside 11 was desilylated to the
alcohol 12 (94%). Oxidation of 12 with the Dess±Martin
periodinane gave the aldehyde 13 (94%), which was
deprotected to 1a (60%).
Dialdehyde 1a immediately and exclusively forms a
1
1
dihydrate when dissolved water, by H NMR. By both H
NMR and UV, this dihydrate gives rise to the structurally-
unusual C6 spiro-fused dihydrouridine 2a9 in a moderately
facile transformation (t1/2,2 h at 238C) that most likely
proceeds through the bridged hydrate 3a as a transient inter-
mediate. Deprotection of 1a with aqueous TFA gave 1b, but
this too failed to form a bridged monohydrate structure (3b).
Interestingly, it also did not undergo transformation to the
unprotected spiro-fused nucleoside 2b, which we obtained
separately by deprotecting 2a. The striking difference in
reactivity between 1a and 1b clearly shows that the 20,30-
O-isopropylidene group in 1a is essential for the trans-
Because the TBAF deprotection conditions had produced
the alcohol as its alkoxide, 18 underwent a very rapid
spiro dioxolane ring assembly (t1/2,a few minutes). The
spiro-fused product 19, the 7-deoxy congener of 2a, was
isolated in a 55% yield, and this was deprotected (aq.
TFA, 62%) to give the stable spiro dihydrouridine 20
(Scheme 4). Under mild acetylating conditions (Ac2O,
pyridine, 08C for 15 min, 258C for 1.5 h), 20 gave the
diacetate 21, and under slightly more forcing conditions

9888
M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
Scheme 4. Reagents and conditions: (i) aq. TFA; (ii) Ac2O, pyridine, 238C; (iii) Ac2O, pyridine, 658C; (iv) MeOH, cat. KCN, 238C.
(658C, 6.5 h) 21 in turn underwent an acylative ring opening
to give the stable triacetate 22. Nucleoside 22, the 50,6-
oxomethylene tethered version of 20,30,50-tri-O-acetyl-
uridine, was isolated as a 7:3 mixture of diastereomers
(90% yield from 20).
Lypho-Lock 4.5 L bench-top freeze-dryer. 1H and 13C NMR
spectra were recorded on a Varian VXR-300 (300 and
75 MHz) or Varian VXR-500 (500 and 125 MHz) instru-
ment. These spectra were recorded with (CH3)4Si or 2,2-
dimethyl-2-silapentane-5-sulfonic acid, sodium salt (DSS)
1
(dˆ0.0 for H), and CDCl3 (dˆ77.0 for 13C), (CD3)2SO
Applying the mild conditions developed by Nudelman's
group for the deacylation of carbohydrates (abs. MeOH,
cat. KCN),17 triacetate 22 was smoothly deprotected
(88%) within minutes directly to the aldehyde 23, which
was generated under these conditions in its 50-methyl hemi-
acetal form. Just as with 1b, 23 exhibited no tendency what-
soever toward existing in cyclic hemiacetal form. Only the
aldehyde hydrate was detected in D2O solution, by 1H
NMR.
(dˆ39.5 for 13C), or 1,4-dioxane (dˆ66.5 for 13C in D2O)
as internal reference. 1H and 13C NMR spectral assignments
were made with the assistance of 1H±1H homonuclear shift
correlation (COSY) and 1H±13C heteronuclear shift correla-
tion (HETCOR) 2D NMR spectroscopic analyses. Except
where noted, the purity of compounds was shown to be
1
.95% by TLC and high-®eld H NMR. UV spectra were
recorded on a Shimadzu UV-160U spectrophotometer.
BuLi, iPr2NH, HCO2Et, (PhNHCH2)2, 2-iodobenzoic acid,
KBrO3, TsOH, NaBH4, 2,20-biquinoline, anhydrous iBuOH,
and 1 M TBAF in THF solution were purchased from the
Aldrich Chemical Co. TBDMS-Cl and TBDPS-Cl were
obtained from HuÈls America, Inc. The BuLi was titrated
by the modi®ed Watson-Eastham procedure.18 THF and
Et2O were dried by distillation from Na-benzophenone
ketyl under argon. Pyridine and iPr2NH were dried by
distillation from CaH2 under argon. The HCO2Et was
dried by distillation from P2O5 under argon. Dowex-50
(H1 form) was obtained from the Sigma Chemical Co.,
and before use was washed with 1N HCl and then rinsed
with distilled water until pH neutral. The Dess±Martin
periodinane reagent was prepared according to the literature
procedure.15 Elemental microanalyses and mass spectral
analyses were obtained from the University of Illinois.
Conclusions
Our results reveal some of the consequences of attaching a
50,6-oxomethylene transglycosidic tether to the pyrimidine
ribonucleoside framework. When the 50-hydroxyl group is
functionalized as it is in 22, this type of tether can be used to
create stable conformationally restricted uridines. Once an
unprotected 50-OH functionality is revealed, though, the
formation of a spirodihydrouridine is energetically favored
if a 20,30-O-methylidene-based protecting group is present,
as was seen in the facile transformations 1a!2a and
18!19. Finally, without such a 20,30-O-cyclic protecting
group to facilitate spirodihydrouridine formation, it is
simply the untethered, `open' hydroxy-aldehyde that
becomes favored, as was seen with 1b and 23. Thus,
while an inherent lability renders the 50,6-oxomethylene
tethered uridine ribonucleoside derivatives of little practical
value, the corresponding 50-ribonucleotide derivatives
should be nicely stable. Work along these lines is in
progress.
50-O-(tert-Butyldimethylsilyl)-6-(hydroxymethyl)-20,30-
O-isopropylideneuridine (7). A solution of 66 (3.04 g,
7.13 mmol) in 90 mL of THF was added in portions to a
stirred suspension of NaBH4 in 75 mL of EtOH at 238C. The
reaction mixture was stirred for 40 min, and then it was
concentrated in vacuo to a residue that was separated by
column chromatography (5% MeOH/CH2Cl2 as eluent) to
1
give 2.29 g (76%) of 7 as a foam: mp 91±938C. H NMR
(CDCl3) d 10.2 (bs, 1), 5.82 (s, 1), 5.80 (s, 1), 5.19 (d, 1),
4.80 (dd, 1), 4.53 (bs, 2), 4.29 (bs, 1), 4.15 (m, 1), 3.87 (m,
2), 1.55 (s, 3), 1.34 (s, 3), 0.90 (s, 9), 0.08 (s, 6). 13C NMR
(CDCl3) d 164.1, 155.5, 150.4, 113.9, 101.2, 91.1, 89.4,
84.2, 81.6, 64.2, 60.5, 27.1, 25.2, 25.9, 18.4. Low-resolution
EI-mass spectrum, m/e 413.2 (M12Me), 371.2 [100%,
(M12CMe3)]. Low-resolution CI-mass spectrum, m/e
430.2 (M112), 413.2 (M12Me), 371.2 [95%,
(M12CMe3)].
Experimental
General procedures
Melting points were determined on a Thomas±Hoover
UniMelt capillary apparatus and are uncorrected. Radial
preparative-layer chromatography was performed on a
Chromatotron instrument (Harrison Research, Inc., Palo
Alto, CA) using Merck silica gel-60 PF254 as the adsorbent,
¯ash column chromatography was performed using 230±
400 mesh ASTM Merck silica gel-60, and TLC analyses
were performed on Analtech 250 mm silica gel GF
Uniplates. Lyophilizations were conducted on a Labconco
6-Formyl-20,30-O-isopropylideneuridine (8). A solution
of 7 (427 mg, 1.00 mmol) in 5 mL of anhydrous THF was
treated dropwise with 1.1 mL of a 1.0 M solution of TBAF

M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
9889
in THF and the reaction mixture was stirred at 238C for 24 h.
The white precipitate produced exhibited a 1H NMR
spectrum consistent with that expected for the O50-tetra-
butylammonium salt of 8, it was collected by ®ltration
and was rinsed with fresh anhydrous THF and CH2Cl2.
The combined ®ltrate and washings were rotary evaporated
at 40±508C in vacuo to an oil. This oil was combined with
the solid from above and the whole was dissolved in a small
amount of 5% MeOH/CH2Cl2. Radial chromatographic
separation (same solvent system as eluent) gave 153 mg
(49%) of a material that was found to exist as a mixture
graphy failed, the periodinane-derived species was removed
by dissolving the mixture in excess satd. aq. NaHCO3
containing Na2S2O3,15,19 followed by a chromatographic
separation as described above. This gave a 30 mg sample
of a binary mixture of 1a and 8, and a 12 mg sample of a
nucleoside subsequently identi®ed as spiro 2a by 1H, 1H±1H
1
COSY, H-coupled and -decoupled 13C, and short-range
1H±13C HETCOR NMR, low and high resolution mass
spectral, and X-ray crystallographic analyses.
20,30-O-Isopropylideneorotidine 50-lactone (10). A solu-
tion of 8 (312 mg, 1.0 mmol) and DCC (0.8 g, 3.9 mmol)
in 10 mL of anhydrous DMSO was treated with dry pyridine
(0.1 mL) and TFA (0.05 mL), and the resulting mixture was
stirred at 238C for 50 h. Water (1 mL) was then added and
the mixture was stirred for additional 0.5 h. The precipitated
dicyclohexylurea was removed by suction ®ltration, and the
®ltrate was evaporated to dryness at 508C in vacuo. Radial
chromatography (5% MeOH/CH2Cl2 as eluent) gave
161 mg (52%) of lactone 10 as a white solid: mp 270±
of
(7R)-7-hydroxy-20,30-O-isopropylidene-O50,6-metha-
nouridine [8, 7(R),O50-hemiacetal], its (7S)-diastereomer
counterpart, and aldehyde 8 by H NMR analysis of a
1
freshly prepared (CD3)2SO solution. After 3 d, the minor
(7S)-diastereomer was absent, by 1H NMR. 7(R),O50-Hemi-
acetal of 8: 1H NMR [(CD3)2SO] d 11.4 (bs, 1H, NH), 7.32
(d, Jˆ6.6 Hz, 1H, 7-OH), 6.28 (d, Jˆ1.2 Hz, 1H, H10), 5.82
(s, 1H, H5), 5.74 (d, Jˆ6.6 Hz, 1H, H7), 4.74 (d, Jˆ6.3 Hz,
1H, H30), 4.68 (dd, Jˆ6.3, 1.2 Hz, 1H, H20), 4.60 (d,
Jˆ3.0 Hz, 1H, H40), 3.99 (d, Jˆ12.9 Hz, 1H, H50R), 3.83
(dd, Jˆ12.9, 3.0 Hz, 1H, H50S), 1.45 and 1.27 (each s, each
3H, each Me). 7(S),O50-Hemiacetal of 8: 1H NMR (CDCl3)
1
2758C (dec.). H NMR (CDCl3) d 8.32 (bs, 1H, NH), 6.08
(s, 1H, H5), 5.98 (d, 1H, H10), 5.02 (m, 1H, H50), 4.97 (dd,
1H, H20), 4.78 (d, 1H, H30), 4.72 (m, 1H, H40), 4.25 (m, 1H,
1
1
d 7.65 (d, 1H, 7-OH), 6.36 (d, 1H, H10). Aldehyde 8: H
H50), 1.58 and 1.37 (each s, each 3H, each Me). H NMR
NMR (CDCl3) d 9.52 (s, 1H, CHO).
[(CD3)2SO] d 11.74 (bs, 1H, NH), 5.76 (s, 1H, H5), 5.73 (d,
1H, H10), 5.11 (d, 1H, H50), 4.68 (t, 1H, H30), 4.63 (t, 1H,
H20),4.55 (dd, 1H, H40), 4.17 (t, 1H, H500), 1.45 (s, 3H, Me),
1.28 (s, 3H, Me). 13C NMR [(CD3)2SO] d 165.5 (C7), 162.8
(C4), 148.7 (C2), 143.1 (C6), 112.0 (C5 or CMe2), 103.0
(C5 or CMe2), 95.0, 86.4, 85.4, and 79.7 (each C10, C20,
C30, or C40), 66.2 (C50), 26.3, and 24.7 (CMe2). Low-
resolution ACE-mass spectrum, m/e 310.2 (M1), 311.2
(MH1).
6-(Hydroxymethyl)-20,30-O-isopropylideneuridine (9). A
solution of 7 (171 mg, 0.40 mmol) in 2 mL of THF was
treated with 0.40 mL of a 1.0 M solution of TBAF
(0.40 mmol) in THF. The reaction mixture was stirred at
238C for 6 h and then was concentrated in vacuo. Radial
chromatographic separation of the residue (10% MeOH/
CH2Cl2) afforded 112 mg (89%) of 9 as a hygroscopic
foam identical by 1H NMR to that obtained in the following
procedure.
50-O-(tert-Butyldimethylsilyl)-6-(1,3-diphenylimidazoli-
din-2-yl)-20,30-O-isopropylideneuridine (11). A solution
of 6 (427 mg, 1.0 mmol) in 10 mL of dry CH2Cl2 was
A suspension of 8 (94 mg, 0.3 mmol) in 3 mL of anhydrous
EtOH was treated with NaBH4 (39 mg, 1 mmol). After a few
min, the suspension was replaced by a clear solution. The
mixture was then concentrated by rotary evaporation in
vacuo, and the residue was puri®ed by radial chromato-
graphy (10% MeOH/CH2Cl2 as eluent) to give 66 mg
treated with
a
solution of (PhNHCH2)2 (910 mg,
4.3 mmol) in dry diethyl ether (5 mL) and 0.12 mL of
glacial acetic acid. The reaction mixture was stirred at
238C for 3 d and then was partitioned between saturated
aqueous NaHCO3 and CH2Cl2. The layers were separated,
and the aqueous phase was extracted with fresh CH2Cl2. The
organic solutions were combined, dried over MgSO4, and
then rotary evaporated to dryness. Column chromatography
(2.5% MeOH/CH2Cl2 as eluent) gave 453 mg (73%, 98%
1
(70%) of 9 as a hygroscopic foam: H NMR [(CD3)2SO] d
11.40 (bs, exchanges upon addition of D2O, 1H, NH), 5.85
(t, exchanges upon addition of D2O, 1H, 7-OH), 5.74 (s, 1H,
H10 or H5), 5.69 (s, 1H, H10 or H5), 5.19 (d, 1H, H20), 4.81
(t, exchanges upon addition of D2O, 1H, 50-OH), 4.74 (dd,
1H, H30), 4.34 (m, 2H, 6-CH2OH), 4.94 (dd, 1H, H40),
3.56±3.33 (m, 2H, 50-CH2), 1.48 and 1.25 (each s, each
1
based upon unrecovered starting material) of 11: H NMR
(CDCl3) d 9.82 (bs, 1H, exchanges upon addition of D2O,
NH), 7.31±6.60 (m, 10H, two Ph), 5.86 (d, 1H, H10), 5.85 (s,
1H, H5 or H7), 5.83 (s, 1H, H5 or H7), 5.27 (dd, 1H, H20),
5.72 (dd, 1H, H30), 4.00 (m, 1H, H40), 3.81 (m, 2H, 50-CH2),
3.73±3.44 (m, 4H, NCH2CH2N), 1.25 and 1.02 (each s,
each 3H, each Me), 0.89 (s, 9H, CMe3), 0.07 and 0.50
3
3
3
0
0
0
0
0
0
3H, Me); J1 ±2 ˆ0.0 Hz, J2 ±3 ˆ6.4 Hz, J3 ±4 ˆ0.0 Hz,
3
3
0
0
J5 ±5 -OHˆ5.8 Hz, J7±7-OHˆ5.6 Hz.
Oxidation of 9. A solution of 9 (168 mg, 0.53 mmol) in
10 mL of anhydrous MeCN was added to a suspension of
Dess±Martin periodinane reagent15 (359 mg, 1.28 mmol) in
3 mL of the same solvent. The reaction mixture was stirred
at 238C for 3.5 h, after which time the starting material had
been consumed, by TLC analysis. The resulting mixture was
rotary evaporated to dryness, and the residue puri®ed by
radial chromatography (60% Me2CO/CH2Cl2 as eluent) to
afford an inseparable mixture consisting of dialdehyde 1a,
hemiacetal 8, and an unknown periodinane-derived species.
When attempts at further puri®cation by repeated chromato-
3
3
0
0
0
0
(each s, each 3H, SiMe2); J1 ±2 ˆ1.5 Hz, J2 ±3 ˆ6.6 Hz,
J3 ±4 ˆ4.2 Hz, J4 ±5 ˆ6.9 Hz. 13C NMR (CDCl3) d 163.7
3
3
0
0
0
0
(C4), 153.5, 151.4, 148.6, 145.4, 130.0, 129.8, 123.6, 120.4,
120.0, 114.6, 113.7, 102.4, 92.5, 89.8, 84.2, 82.6, 77.8, 64.7
(C50), 52.1 and 47.1 (NCH2CH2N), 26.9 and 26.3 (CMe2),
26.3 (CMe3), 18.9 (SiMe2). LR-EIMS, m/e 620.5 (M1);
LR-CIMS, m/e 621.4 (MH1).
6-(1,3-Diphenylimidazolidin-2-yl)-20,30-O-isopropylide-
neuridine (12). A solution of 11 (336 mg, 0.54 mmol) in

9890
M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
3
1
0
0
0
0
2 mL of anhydrous THF was treated with 0.6 mL of 1.0 M
TBAF in THF solution. The reaction mixture was stirred at
238C for 36 h, and then the solvents were removed by rotary
evaporation in vacuo. The residue was puri®ed by radial
chromatography (5% MeOH/CH2Cl2 as eluent) to afford
each 3H, each Me); 3J1 ±2 ˆ0.0 Hz, J2 ±3 ˆ6.0 Hz. H NMR
(CDCl3) d 10.5 (bs, exchanges upon addition of D2O, 1H,
NH), 9.60 (s, 1H, CHO), 9.42 (s, 1H, CHO), 9.2 (bs,
exchanges upon addition of D2O, 1H, NH), 6.75 (s, 1H,
H5), 6.34 (s, 1H, H10), 5.22 (dd, 1H, H20), 5.08 (d, 1H,
H30), 5.55 (d, 1H, H40), 1.55 and 1.36 (each s, each 3H,
1
253 mg (94%) of 12 as a foam: H NMR (CDCl3) d 10.38
3
3
3
0
0
0
0
0
0
(bs, exchanges upon addition of D2O, 1H, NH), 7.22±6.68
(m, 10H, two Ph), 5.87 (s, 1H, H5 or H7), 5.79 (s, 1H, H5 or
H7), 5.77 (d, 1H, H10), 5.28 (dd, 1H, H20), 4.97 (dd, 1H,
H30), 4.01 (dd, 1H, H40), 3.92±3.70 (m, 2H, 50-CH2),
each Me); J1 ±2 ˆ1.5 Hz, J2 ±3 ˆ6.3 Hz, J3 ±4 ˆ1.5 Hz.
13C NMR [(CD3)2SO] d 201.1 (CHO), 188.0 (CHO),
162.6 (C4), 151.5 (C2 or C6), 146.7 (C2 or C6), 114.4
(C5), 111.9 (CMe2), 93.3 (C40), 92.0 (C10), 85.0 (C20 or
C30), 84.0 (C20 or C30), 26.3 and 24.6 (CMe2). LR-EIMS,
m/e 310.1 (M1); LR-CIMS, m/e 311.1 (MH1). UV lmax, nm
(e£1023): (H2O) 261 (8.9), 204 (9.2). Anal. Calcd for
C13H14N2O7: C, 50.33; H, 4.55; N, 9.03. Found: C, 50.20;
H, 4.40; N, 8.98. 1a, 50-dimethyl acetal: 1H NMR (CDCl3) d
9.58 (s, 1H, H7), 9.42 (bs, exchanges upon addition of D2O,
1H, NH), 6.60 (s, 1H, H5 or H10), 6.25 (s, 1H, H5 or H10),
5.22 (d, 1H, H20), 5.00 (dd, 1H, H30), 4.67 (d, 1H, H50), 4.25
3.70±3.40 (m, 5H, NCH2CH2N and 50-OH), 1.25 and 1.02
3
(each s, each 3H, each Me); 3J1 ±2 ˆ2.4 Hz, J2 ±3 ˆ6.9 Hz,
0
3
0
0
0
J3 ±4 ˆ3.9 Hz, J4 ±5 ˆ2.7 Hz, J5 ±5 ˆ12.0 Hz. 13C NMR
3
3
0
0
0
0
0
00
(CDCl3) d 163.0 (C4), 152.5, 151.8, 146.9, 146.0, 129.5,
129.3, 121.8, 120.8, 117.8, 116.1, 113.8, 102.5, 92.1, 87.4,
83.0, 80.4, 76.2, 62.5 (C50), 50.1 and 48.1 (NCH2CH2N),
26.4 and 25.3 (CMe2). LR-EIMS, m/e 506.3 (M1);
LR-CIMS, m/e 507.3 (MH1).
(dd, 1H, H40), 3.48 and 3.35 (each s, each 3H, C(OMe)2),
3
6-(1,3-Diphenylimidazolidin-2-yl)-20,30-O-isopropylide-
neuridine 50-carboxaldehyde (13). A solution of 12
(251 mg, 0.50 mmol) in 2 mL of CH2Cl2 was added to a
suspension of the Dess±Martin periodinane (315 mg,
0.75 mmol) in 3 mL of CH2Cl2 at 238C. The mixture was
stirred for 1 h, and then diethyl ether (13 mL) and a solution
of Na2S2O3´5H2O (1.30 g, 5.22 mmol) in 20 mL of saturated
aqueous NaHCO3 were added. The two layers were sepa-
rated, and the aqueous phase was extracted with fresh
CH2Cl2. The organic solutions were combined, dried over
MgSO4, and then rotary evaporated to dryness. The residue
was puri®ed by radial chromatography (2:1 EtOAc/hexanes
as eluent) to afford 235 mg (94%) of 13 as a foam: 1H NMR
(CDCl3) d 10.15 (bs, exchanges upon addition of D2O, 1H,
NH), 9.42 (s, 1H, CHO), 7.35±6.65 (m, 10H, two Ph), 6.05,
5.89, and 5.85 (each s, each 1H, each H10, H5, or H7), 5.10
and 4.16 (each d, each 1H, H20 and H30), 4.42 (s, 1H, H40),
3.75±3.45 (m, 4H, NCH2CH2N), 1.27 and 0.99 (each
1.58 and 1.38 (each s, each 3H, each Me); J1 ±2 ˆ0.0 Hz,
0
0
3
3
3
0
0
0
0
0
0
J2 ±3 ˆ6.6 Hz, J3 ±4 ˆ4.3 Hz, J4 ±5 ˆ7.5 Hz. LR-EIMS,
m/e 341.2 [(M2Me)1]; LR-CIMS, m/e 357.2 (MH1).
A freshly prepared sample of pure 1a in D2O solution
revealed NMR spectral features consistent with the simple
1
dihydrate: H NMR (D2O) d 6.32 (s, 1H, H10, H5, or H7),
6.07 (s, 1H, H10, H5, or H7), 5.95 (s, 1H, H10, H5, or H7),
5.37 (d, 1H, H20), 5.10 (d, 1H, H50), 5.03 (dd, 1H, H30), 5.93
(dd, 1H, H40), 1.60 and 1.75 (each s, each 3H, each Me);
3
3
3
3
0
0
0
0
0
0
J1 ±2 not well resolved, J2 ±3 ˆ3.9 Hz, J3 ±4 ˆ4.5 Hz,
J4 ±5 ˆ7.5 Hz. 13C NMR (D2O) d 165.7 (C4), 155.3 (C2),
0
0
151.7 (C6), 114.5 (CMe2), 100.2 (C5), 91.4 (C10), 89.9
(C40), 89.7 (C50), 86.3 (C7), 84.2 (C20), 81.7 (C30), 26.0
and 24.3 (CMe2).
6-Formyluridine 50-carboxaldehyde (1b). A solution of
1a (51 mg, 0.17 mmol) in 50% aqueous TFA (1 mL) was
stirred at 238C for 2 h. The resulting solution was evapo-
rated to dryness, and residual TFA was removed from the
residue by repetitive azeotropic coevaporation with water.
Lyophilization afforded essentially pure 1b in quantitative
yield: 210±2208C (dec.). The NMR spectral features of 1b
3
3
0
0
0
0
s, each 3H, each Me); J1 ±2 ˆ0.0 Hz, J2 ±3 ˆ6.9 Hz,
J3 ±4 ˆ0.0 Hz. 13C NMR (CDCl3) d 199.7 (CHO), 163.1
(C4), 152.4, 151.8, 147.5, 145.3, 129.5, 129.3, 122.7,
120.1, 119.0, 114.8, 112.8, 102.6, 94.1, 93.8, 84.8, 84.0,
76.1, 50.9 and 47.4 (NCH2CH2N), 25.5 and 25.3 (CMe2).
LR-EIMS, m/e 504.3 (M1); LR-CIMS, m/e 505.3 (MH1).
3
0
0
in D2O solution were consistent with
a dihydrate
1
structure: H NMR (D2O) d 6.14 (s, 1H, H5), 5.98 (m,
2H, H10 and H7), 5.12 (d, 1H, H50), 4.84 (s, H20 under
HOD), 4.50 (pseudo-t, 1H, H30), 3.80 (pseudo-t, 1H, H40);
6-Formyl-20,30-O-isopropylideneuridine 50-carboxalde-
hyde (1a). A solution of 13 (285 mg, 0.5 mmol) in 17 mL
of CH2Cl2 was treated with a solution of TsOH´H2O in 8 mL
of Me2CO, and the reaction mixture was stirred at 238C for
40 min. NaHCO3 (250 mg, 3.0 mmol) was then added, and
the suspension obtained was stirred for 5 min before the
mixture was ®ltered and the solid was rinsed with fresh
Me2CO and CH2Cl2. The combined ®ltrate and washings
were then dried (MgSO4) and evaporated to dryness. The
residue was dissolved in a small amount of Me2CO and
puri®ed by radial chromatography (3:2 Me2CO/CH2Cl2 as
eluent). The product isolated was nearly pure but contained
3
3
3
3
0
0
0
0
0
0
J1 ±2 not well resolved, J2 ±3 ˆ6.2 Hz, J3 ±4 ˆ5.9 Hz,
J4 ±5 ˆ5.7 Hz. 13C NMR (D2O) d 165.7 (C4), 156.0 (C2),
0
0
151.6 (C6), 100.2 (C5), 92.1 (C10), 89.9 (C50), 85.8 (C7),
85.5 (C40), 71.6 (C20), 70.3 (C30). UV lmax, nm (e£1023):
(H2O) 261 (8.9), 204 (9.2); (pH 1) 262 (7.8), 209 (7.2). LR-
CIMS of 1b, m/e 271.1 (MH1); HR-CIMS, m/e calcd for
C10H11N2O7 271.0566, found 271.0564.
(6R,7R,50S)-6,50:7,50-Dianhydro-5,6-dihydro-6,50-di-
hydroxy-6-[(dihydroxy)methyl]-20,30-O-isopropylide-
neuridine (2a). Slow evaporation of an aqueous solution of
1a within a NaOH-containing desiccator charged with an
argon atmosphere afforded X-ray quality crystals of 2a: mp
205±2108C (dec.). 1H NMR [(CD3)2SO] d 10.7 (bs,
exchanges upon addition of D2O, 1H, NH), 7.60 (d,
exchanges upon addition of D2O, 1H, 7-OH), 6.10 (s, 1H,
H10), 5.83 (d, 1H, H50), 5.44 (d, 1H, H7), 4.82 (d, 1H, H20 or
1
small amounts of hemiacetal and/or hydrate species, by H
NMR spectral analysis. Abderhalden drying (P2O5, 788C) in
vacuo overnight gave 93 mg (60%) of pure dialdehyde 1a:
1H NMR [(CD3)2SO] d 12.0 (bs, exchanges upon addition of
D2O, 1H, NH), 9.53 (s, 1H, CHO), 9.44 (s, 1H, CHO), 6.58
(s, 1H, H10), 6.53 (s, 1H, H5), 5.06 (d, 1H, H20 or H30), 4.97
(d, 1H, H20 or H30), 4.50 (s, 1H, H40), 1.45 and 1.29 (each s,

M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
9891
H30), 4.76 (d, 1H, H20 or H30), 4.28 (d, 1H, H40), 3.16 [d, 1H,
H5(S)], 2.77 [d, 1H, H5(R)], 1.40 and 1.26 (each s, each 3H,
7.32 (m, 10H, two Ph), 5.80 (s, 1H, H10 or H5), 5.71 (d, 1H,
H10 or H5), 5.20 (d, 1H, H20), 4.80 (dd, 1H, H30), 4.58 and
4.40 (each d, each 1H, 6-CH2O), 4.13 (m, 1H, H40), 3.80
(m, 2H, H50), 1.50 and 1.32 (each s, each 3H, each Me), 1.08
(s, 9H, Me3CSiPh2), 0.88 (s, 9H, Me3CSiMe2), 0.04 (s, 6H,
3
3
3
0
0
0
0
0
0
each Me); J1 ±2 ˆ0.0 Hz, J2 ±3 ˆ6.0 Hz, J3 ±4 ˆ0.0 Hz,
J4 ±5 ˆ1.2 Hz, J7±7-OHˆ6.9 Hz, J5R±5Sˆ15.6 Hz. 13C
3
3
2
0
0
NMR [(CD3)2SO] d 166.9 (C4), 150.8 (C2), 111.6
(CMe2), 103.8 (C50), 102.4 (C7), 91.4 (C6), 89.4 (C10),
86.4 (C20 or C30), 78.0 (C20 or C30), 85.7 (C40), 37.3
(C5), 26.1, and 24.3 (CMe2). LR-CIMS, m/e 329.2
(MH1); HR-CIMS, m/e calcd for C13H17N2O8 329.0985,
found 329.0989. UV lmax, nm (e£1023): (H2O) 262 (1.7),
209 (6.8); (pH 1) 261 (1.2), 210 (6.0); (pH 7) 263 (1.3), 210
(6.1); (pH 11) 271 (3.6), 238 (6.2), 215 (4.8), 206 (4.4).
0
0
0
0
0
0
,
SiMe2); J2 ±3 ˆ6.3 Hz, J3 ±4 ˆ4.5 Hz, J7a±7bˆ14.1 Hz, J1 ±2
J4 ±5 not well resolved. 13C NMR (CDCl3) d 163.7 (C4),
154.0 (C2), 150.6 (C6), 135.4, 134.7, 131.8, 130.2, 129.5,
128.0, 127.9, and 127.6 (eight C's's's from two Ph), 113.5
(CMe2), 101.6 (C5), 91.3 (C10), 89.7 (C40), 84.3 (C20), 81.9
(C30), 64.3 (C50), 62.0 (C7), 27.2 and 25.3 (CMe2), 26.5
(Me3CSiPh2), 25.9 (Me3CSiMe2), 19.1 and 18.4 (Me2Si).
LR-EIMS, m/e 609.3 [25%, (M12CMe3)]. LR-CIMS, m/e
667.4 (50%, MH1), 609.3 [70%, (M12CMe3)].
0
0
(6R,7R,50S)-6,50:7,50-Dianhydro-5,6-dihydro-6,50-di-
hydroxy-6-[(dihydroxy)methyl]uridine (2b). A solution
of 2a (20 mg, 0.06 mmol) in 0.5 mL of 50% aqueous TFA
was stirred at 238C for 36 h. The solution was evaporated to
dryness in vacuo, and any residual TFA was removed azeo-
tropically by repetitive coevaporation with water under
reduced pressure. Recrystallization of the resulting crude
product from water afforded 13 mg (74%) of 2b as a
6-[(tert-Butyldiphenylsilanoxy)methyl]-20,30-O-isopro-
pylideneuridine (16). A solution of 15 (3.08 g, 82% pure,
3.79 mmol) in absolute ethanol (19 mL) was treated with
pyridinium p-toluenesulfonate (PPTS, 284 mg, 1.13 mmol)
and the reaction mixture was stirred at 238C for 24 h. The
solvent was removed in vacuo and the residue was dissolved
in CH2Cl2. The organic solution was washed with saturated
aqueous brine and then dried over MgSO4. The solvent was
removed in vacuo and the crude product was puri®ed by
radial chromatography using 5% MeOH/CH2Cl2 as eluent
to give 16 as a white foam (2.03 g, 97%): 1H NMR (CDCl3)
d 10.0 (bs, 1H, NH), 7.70±7.38 (m, 10H, two Ph), 5.74 (d,
1H, H10), 5.70 (s, 1H, H5), 5.24 (dd, 1H, H20), 5.02 (dd, 1H,
H30), 4.52 and 4.42 (each d, each 1H, 6-CH2O), 4.20 (m, 1H,
H40), 3.89±3.78 (m, 2H, H50), 3.2 (bs, 1H, 50-OH), 1.47
and 1.32 (each s, each 3H, each Me), 1.08 (9H, s,
1
white solid: H NMR [(CD3)2SO] d 10.6 (bs, 1), 7.51 (d,
1), 6.00 (d, 1), 5.78 (d, 1), 5.42 (d, 1), 5.24 (d, 1), 5.08 (d, 1),
4.30 (m, 1), 4.12 (m, 1), 4.05 (d, 1), 3.16 (d, 1), 2.76 (d, 1).
13C NMR [(CD3)2SO] d 167.0, 150.9, 104.0, 102.4, 91.6,
91.4, 88.9, 78.0, 69.5, 38.4. UV lmax, nm (e£1023): (H2O)
271 (1.0), 212 (7.9); (pH 1) 261 (2.0), 203 (9.0); (pH 11) 274
(13.1). LR-CIMS, m/e 289.1 (100, MH1), 271.1 (70%,
MH12H2O), 253.1 (15%, MH122H2O); HR-CIMS, m/e
calcd for C10H13N2O8 289.0672, found 329.0670.
50-O-(tert-Butyldimethylsilyl)-6-(hydroxymethyl)-20,30-
O-isopropylideneuridine (14). A solution of 66 (3.04 g,
7.13 mmol) in 90 mL of THF was added in portions to a
stirred suspension of NaBH4 in 75 mL of EtOH at 238C. The
reaction mixture was stirred for 40 min, and then it was
concentrated in vacuo to a residue that was separated by
column chromatography (5% MeOH/CH2Cl2 as eluent) to
Me3CSiPh2); J1 ±2 ˆ2.5 Hz, J2 ±3 ˆ6.5 Hz, J3 ±4 ˆ4.1 Hz,
0
0
0
0
0
0
J7a±7bˆ14.2 Hz, J4 ±5 not well resolved. 13C NMR
(CDCl3) d 162.9 (C4), 153.6 (C2), 151.3 (C6), 135.5,
131.7, 130.3 and 128.0 (four C's's's from two Ph), 114.0
(CMe2), 102.1 (C5), 91.8 (C10), 87.5 (C40), 83.3 (C20), 80.4
(C30), 62.7 (C50), 62.1 (C7), 27.2 and 25.2 (CMe2), 26.5
(Me3CSiPh2), 19.1 (Me3CSi). LR-EIMS, m/e 537.2 (20%,
M12Me), 495.2 [15%, (M12CMe3)], 437.1 [70%,
(M12CMe3±Me2CO)]. LR-CIMS, m/e 553.3 (15%,
MH1), 495.2 [20%, (M12CMe3)].
0
0
1
give 2.29 g (76%) of 14 as a foam: mp 91±938C. H NMR
(CDCl3) d 10.2 (bs, 1), 5.82 (s, 1), 5.80 (s, 1), 5.19 (d, 1),
4.80 (dd, 1), 4.53 (bs, 2), 4.29 (bs, 1), 4.15 (m, 1), 3.87 (m,
2), 1.55 (s, 3), 1.34 (s, 3), 0.90 (s, 9), 0.08 (s, 6). 13C NMR
(CDCl3) d 164.1, 155.5, 150.4, 113.9, 101.2, 91.1, 89.4,
84.2, 81.6, 64.2, 60.5, 27.1, 25.2, 25.9, 18.4. Low-resolution
EI-mass spectrum, m/e 413.2 (M12Me), 371.2 [100%,
(M12CMe3)]. Low-resolution CI-mass spectrum, m/e
430.2 (M112), 413.2 (M12Me), 371.2 [95%,
(M12CMe3)].
6-[(tert-Butyldiphenylsilanoxy)methyl]-20,30-O-isopro-
pylideneuridine 50-carboxaldehyde (17). A solution of 16
(1.11 g, 2.0 mmol) in 10 mL of CH2Cl2 was added to a
suspension of Dess±Martin periodinane (1.70 g,
4.0 mmol) in 20 mL of anhydrous CH2Cl2. The reaction
mixture was stirred at 238C for 4 h, at which time the start-
ing material had been consumed, by TLC analysis. The
resulting mixture was ®rst treated with Na2S2O3´5H2O
(6.97 g, 28.0 mmol) in saturated aq NaHCO3 (107 mL),
and extracted with CH2Cl2 (5£100 mL). The combined
extracts were dried (MgSO4), concentrated and puri®ed by
chromatography (50% EtOAc/hexanes). The expected
product 17 was ®nally isolated as a foam (885 mg, 80%)
and was found to be a pure 40-b epimer. Nucleoside 17 in
CDCl3 solution was found to be susceptible to slow air
oxidation, giving a 2:1 mixture of 17 and the corresponding
6-[(tert-Butyldiphenylsilanoxy)methyl]-20,30-O-isopro-
pylidene-50-O-(tert-butyldimethylsilyl)uridine (15).
A
solution of 14 (2.14 g, 5.0 mmol) and imidazole (0.749 g,
11.0 mmol) in 5 mL of anhydrous DMF was treated drop-
wise with TBDPS-Cl (1.51 g, 1.43 mL, 5.5 mmol). The
reaction mixture was stirred at 238C for 4.5 h and then it
was evaporated to dryness in vacuo. The residue was
dissolved in CH2Cl2 and the mixture was separated by radial
chromatography using 5% MeOH/CH2Cl2 as eluent. The
isolated product was dried in vacuo in an Abderhalden
chamber over P2O5 at 568C, giving 15 as a pale yellow
glassy solid (3.50 g, 82% purity, 86% calculated yield):
mp 55±628C. This sample contained 0.12 equiv. of DMF,
1
orotidine within 10 d. H NMR (CDCl3) d 9.42 (s, 1H,
CHO), 8.76 (bs, 1H, NH), 7.72±7.35 (m, 10H, two Ph),
6.09 (s, 1H, H10), 5.62 (d, 1H, H5), 5.22 (dd, 1H, H20),
5.12 (d, 1H, H30), 4.58 and 4.43 (each d, each 1H,
6-CH2O), 4.51 (s, 1H, H40), 1.52 and 1.35 (each s, each
1
1
by H NMR. H NMR (CDCl3) d 9.7 (bs, 1H, NH), 7.75±

9892
M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
0
0
3H, each Me), 1.08 (9H, s, Me3CSiPh2); J2 ±3 ˆ6.3 Hz,
(e£1023): (H2O) 261 (2.1), 210 (5.8); (pH 1) 263 (5.5), 210
(5.6); (pH 11) 264 (9.7); 227 (7.4). Anal. Calcd for
C10H12N2O7: C, 44.10; H, 4.44; N, 10.29. Found: C,
44.01; H, 4.47; N, 10.08.
0
0
0
0
0
0
J3 ±4 ˆ0 Hz, J4 ±5 ˆ0 Hz, J7a±7bˆ13.8 Hz, J1 ±2 not well
13
resolved. C NMR (CDCl3) d 199.6 (C50), 163.1 (C4),
153.3 (C2), 151.5 (C6), 135.5, 131.7, 130.4 and 128.0
(four C's's's from two Ph), 113.2 (CMe2), 102.5 (C5),
94.2 (C40), 94.0 (C10), 85.3 (C20 or C30), 84.1 (C20 or
C30), 62.2 (C7), 26.5 (Me3CSiPh2), 26.4 and 24.7 (CMe2),
19.2 (Me3CSi). LR-EIMS, m/e 535.2 (10%, M12Me),
493.2 [85%, (M12CMe3)], 435.1 [100%, (M12CMe3±
Me2CO)]. LR-CIMS, m/e 551.2 (100%, MH1).
(6R,50S)-6,50:7,50-Dianhydro-5,6-dihydro-6,50-dihydroxy-
6-(hydroxymethyl)-20,30-di-O-acetyluridine (21). A solu-
tion of 20 (10 mg) in dry pyridine (0.2 mL) at 08C under Ar
was treated dropwise with Ac2O (0.1 mL) and was stirred at
08C for 15 min and then at 238C for 1.5 h. The solution was
evaporated to dryness in vacuo, giving 21 in a quantitative
yield: 1H NMR (CDCl3) d 8.5 (bs, 1H, NH, exchanges with
D2O), 6.51 (s, 1H, H10), 5.70 (s, 1H, H50), 5.45 and 5.40
(each s, each 1H, H20 and H30), 4.44 and 4.06 (each d, each
1H, H7a and H7b), 4.40 (s, 1H, H40), 3.28 and 2.85 (each d,
each 1H, H5a and H5b), 2.13 and 2.12 (each 3H, each s,
(6R,50S)-6,50:7,50-Dianhydro-5,6-dihydro-6,50-dihydroxy-
6-(hydroxymethyl)-20,30-O-isopropylideneuridine (19).
A solution of 785 mg (1.42 mmol) of 17 in 5 mL of anhy-
drous THF was treated dropwise with 1.42 mL of a 1.0 M
solution of TBAF in THF. By TLC, a 2,4-DNP-positive
compound, likely 6-hydroxymethyl-20,30-O-isopropylide-
neuridine 50-carboxaldehyde (18), formed rapidly but then
gradually diminished. After 30 min, the mixture was
concentrated in vacuo and the residue was puri®ed by radial
chromatography using 10% MeOH/CH2Cl2 as eluent. No
trace of 18 was isolated, but instead, the spiro nucleoside
19 was obtained as a white powder (222 mg, 55%) after
3
3
3
0
0
0
0
each CH3); J7a±7bˆ7.8 Hz, J1 ±2 ˆ0 Hz, J2 ±3 ˆ6.5 Hz,
3
3
0
0
J3 ±4 ˆ0 Hz, J5a±5bˆ16.4 Hz.
20,30,50-Tri-O-acetyl-50,6-(oxomethylene)uridine (22). A
solution of 21 from above in pyridine (0.2 mL) under Ar
was treated with Ac2O (0.1 mL) was stirred at 608C for 6.5 h
and then was concentrated to dryness in vacuo. The product
22 was isolated (13.2 mg, 90% based on 20) by successive
preparative chromatographic separations on SiO2 (5%
MeOH/CH2Cl2 then 2:1 EtOAc/hexanes as eluent). 1H
NMR analysis of 22 in CDCl3 revealed that it exists as a
mixture (7:3) of 50-acetal diastereomers. LR-FABMS, m/e
399.1 (60%, MH1), 118.9 (100%). HR-CIMS for
C26H30N4O7 (MH1): calcd 399.1039, found 399.1038. UV
1
recrystallization from acetone: mp 205±2328C (dec.). H
NMR [(CD3)2SO] d 10.7 (bs, 1H, NH), 6.13 (s, 1H, H10),
5.75 (s, 1H, H50), 4.82 (d, 1H, H20 or H30), 4.77 (d, 1H, H20
or H30), 4.35 (d, 1H, H40), 4.18 and 4.03 (each d, each
1H, H7a and H7b), 3.38 and 2.76 (each d, each 1H,
H5a and H5b), 1.41 and 1.27 (each s, each 3H, each Me);
0
0
0
0
0
0
J2 ±3 ˆ5.6 Hz, J1 ±2 ˆ0 Hz, J4 ±5 ˆ0 Hz, J5a±5bˆ15.8 Hz,
J7a±7bˆ7.7 Hz, J3 ±4 not well resolved. 13C NMR
[(CD3)2SO] d 166.9 (C4), 150.5 (C2), 134.5 (C6), 111.7
(CMe2), 104.7 (C50), 89.8 (C10), 86.5 (C20 or C30), 86.3
(C40), 79.8 (C20 or C30), 78.7 (C7), 41.3 (C5), 26.2 and
24.3 (CMe2). LR-EIMS, m/e 297.1 (60%, M12Me), 267.1
[100%, (M12Me±CH2O)]. LR-CIMS, m/e 313.1 (100%,
MH1). HR-CIMS, m/e calcd for C13H17N2O7 313.10358,
found 313.10360. UV lmax, nm (e£1023): (H2O) 262
(10.1); (pH 1) 260 (15.7), 207 (14.7); (pH 11) 261 (13.7);
227.8 (14.2). Anal. Calcd for C13H16N2O7: C, 50.00; H,
5.16; N, 8.97. Found: C, 50.30; H, 5.20; N, 9.05.
l
max, nm (e£1023): (CH3OH) 265 (8.5), 210 (7.0).
0
0
1
Major diastereomer of 22. H NMR (CDCl3) d 9.02 (bs,
1H, NH, exchanges with D2O), 6.84 (d, 1H, H10), 5.75 (d,
1H, H20) 5.74 (s, 1H, H50), 5.71 (dd, 1H, H30), 5.63 (s, 1H,
H5), 4.93 and 4.75 (each d, each 1H, each H7), 4.42 (d,
3
3
3
1H, H40); J7a±7bˆ14.4 Hz, J1 ±2 ˆ5.4 Hz, J2 ±3 ˆ5.7 Hz,
0
0
0
0
J3 ±4 ˆ3.3 Hz. 13C NMR (CDCl3) d 169.8, 169.7, and
168.8 (3£CO), 161.4 (C4), 151.3 and 149.8 (C2/C6),
104.6 (C5), 94.6 (C10), 90.1 (C50), 84.3 (C40), 76.4 and
71.3 (C20/C30), 71.5 (C7), 20.8, 20.7, and 20.4 (3£CH3).
3
0
0
1
(6R,50S)-6,50:7,50-Dianhydro-5,6-dihydro-6,50-dihydroxy-
Minor diastereomer of 22. H NMR (CDCl3) d 9.05 (bs,
6-(hydroxymethyl)uridine (20).
A
solution 60 mg
1H, NH, exchanges with D2O), 6.61 (d, 1H, H10), 6.02 (s,
1H, H50), 5.73 (d, 1H, H20), 5.65 (s, 1H, H5), 5.53 (d, 1H,
H7a), 5.45 (dd, 1H, H30), 4.44 (d, 1H, H40), 4.13 (d,
1H, H7b), 2.19, 2.16, and 2.10 (each s, each 3H, each
(0.19 mmol) of 19 in 2.0 mL of 50% aqueous TFA was
kept at 238C for 48 h and then was rotary evaporated to
dryness in vacuo. The residual TFA was removed by
repetitive azeotropic coevaporation with water in vacuo at
238C. Recrystallization of the residue from acetone gave
33 mg (62%) of the deprotected spiro nucleoside 20 as a
white powder: mp 210±2208C (dec.). 1H NMR
[(CD3)2SO] d 10.7 (bs, 1H, NH), 6.03 (s, 1H, H10), 5.71
(s, 1H, H50), 5.25 (d, 1H, 20-OH), 5.07 (d, 1H, 30-OH), 4.32
(t, 1H, H20), 4.17±4.10 (m, 3H, H7a, H40, and H30), 4.00 (d,
1H, H7b), 3.40 and 2.73 (each d, each 1H, H5a and H5b);
3
3
3
0
0
0
0
CH3CO); J7a±7bˆ14.0 Hz, J1 ±2 ˆ4.8 Hz, J3 ±4 ˆ1.2 Hz,
J4 ±5 ˆ0 Hz. 13C NMR (CDCl3) d 169.6, 169.5, and 168.4
(3£CO), 161.6 (C4), 151.0 and 149.4 (C2/C6), 105.5 (C5),
90.1 (C10), 89.6 (C50), 85.9 (C40), 75.4 (C20), 73.2 (C30),
64.1 (C7), 20.8, 20.6, and 20.4 (3£CH3).
3
0
0
Deprotection of 22. A solution of 22 (5 mg, 0.013 mmol) in
CD3OD (1.0 mL) at 238C was treated with KCN (0.4 mg).
After 15 min, a 1:1 mixture of methyl hemiacetal dia-
0
0
0
0
0
0
0
0
J1 ±2 ˆ0 Hz, J2 ±2 OHˆ9.3 Hz, J3 ±3 OHˆ4.2 Hz, J4 ±5 ˆ0 Hz,
1
0
0
0
0
J5a±5bˆ16.2 Hz, J7a±7bˆ7.8 Hz; J2 ±3 and J3 ±4 not well
resolved. 13C NMR [(CD3)2SO] d 167.0 (C4), 150.6 (C2),
104.7 (C50), 91.7, 90.0 and 89.4 (C6, C10 and C40), 78.7,
78.1 and 69.3 (C20, C30, and C7), 41.6 (C5). LR-EIMS, m/e
297.1 (60%, M12Me), 267.1 (100%, [M12Me±CH2O)].
LR-CIMS, m/e 313.1 (100%, MH1). HR-CIMS, m/e calcd
for C13H17N2O7 313.10358, found 313.10360. UV lmax, nm
stereomers of 23 had formed, by H NMR. The solution
was rotary evaporated and the residue was separated by
preparative chromatography on SiO2 (20% MeOH/CH2Cl2
as eluent). After Abderhalden drying (P2O5, 788C) in vacuo
overnight, 3 mg (88%) of 23 was obtained: 1H NMR
(CD3OD) d 5.81 and 5.80 (each 1H, each s, each H5),
0
3
5.63 (3J1 ±2 ˆ4.8 Hz) and 5.59 ( J1 ±2 ˆ4.5 Hz) (each 1H,
0
0
0

M. P. Groziak, R. Lin / Tetrahedron 56 (2000) 9885±9893
9893
each d, each H10), 5.98±4.73 (m, H50, H20, and H30), 4.47
and 4.46 (each s, each 2H, CH2), 4.07 (dd, 1H, H40, Jˆ3.3,
3.9 Hz), 4.00 (dd, 1H, H40, Jˆ4.2, 6.3 Hz). Nucleoside 23
exists only as the hydrate in D2O: 1H NMR (D2O) d 5.97 (s,
1H, H5), 5.58 (d, 1H, H10), 5.11 (d, 1H, H50), 4.90 and 4.46
(each m, each 1H, H20/H30), 4.59 (s, 2H, CH2), 3.82
5. Groziak, M. P.; Koohang, A.; Stevens, W. C.; Robinson, P. D.
J. Org. Chem. 1993, 58, 4054±4060.
6. 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±1057.
7. (a) Jones, G. H.; Moffatt, J. G. J. Am. Chem. Soc. 1968, 90,
5337±5338. (b) Jones, G. H.; Taniguchi, M.; Tegg, D.; Moffatt,
J. G. J. Org. Chem. 1979, 44, 1309±1317.
3
3
(pseudo-t, 1H, H40); J1 ±2 ˆ4.2 Hz, J4 ±5 ˆ5.7 Hz. 13C
NMR (D2O) d 165.9 (C4), 157.0 (C2), 151.8 (C6), 101.2
(C5), 91.6, 89.9, and 85.9 (C50/C10/C40), 74.5 and 70.4 (C20
and C30), 59.6 (C7).
0
0
0
0
8. (a) Bannister, B.; Kagan, F. J. Am. Chem. Soc. 1960, 82, 3363±
3368. (b) Chambers, R. W.; Kurkov, V. J. Am. Chem. Soc. 1963,
85, 2160±2164. (c) Reist, E. J.; Benitez, A.; Goodman, L. J. Org.
Chem. 1964, 29, 554±558. (d) Santi, D. V.; Brewer, C. F. J. Am.
Chem. Soc. 1968, 90, 6236±6238. (e) Otter, B. A.; Falco, E. A.;
Fox, J. J. J. Org. Chem. 1969, 34, 1390±1396. (f) Kondo, Y.;
Fourrey, J.-L.; Witkop, B. J. Am. Chem. Soc. 1971, 93, 3527±3529.
9. Groziak, M. P.; Lin, R.; Robinson, P. D. Acta Crystallogr. 1995,
C51, 1204±1207.
Acknowledgements
This work was performed at Southern Illinois University
(SIU) where it was supported in part by grants from SIU's
Of®ce of Research Development and Administration.
Current support from NIH (NIGMS 56878) is gratefully
acknowledged.
10. Myers, A. G.; Gin, D. Y.; Rogers, D. H. J. Am. Chem. Soc.
1994, 116, 4697±4718.
11. (a) Tanaka, H.; Hayakawa, H.; Miyasaka, T. Chem. Pharm.
Bull. 1981, 29, 3565±3572. (b) Tanaka, H.; Hayakawa, H.;
Miyasaka, T. Tetrahedron 1982, 38, 2635±2642. (c) Tanaka, H.;
Hayakawa, H.; Iijima, S.; Haraguchi, K.; Miyasaka, T.
Tetrahedron 1985, 41, 861±866. (d) Kittaka, A.; Tanaka, H.;
Odanaka, Y.; Ohnuki, K.; Yamaguchi, K.; Miyasaka, T. J. Org.
Chem. 1994, 59, 3636±3641. (e) Kittaka, A.; Kato, H.; Tanaka, H.;
Nonaka, Y.; Amano, M.; Nakamura, K. T.; Miyasaka, T.
Tetrahedron 1999, 55, 5319±5344.
References
1. Connolly, G. P.; Duley, J. A. Trends Pharmacol. Sci. 1999, 20,
218±225.
2. (a) Rabi, J. A.; Fox, J. J. J. Org. Chem. 1972, 37, 3898±3901.
(b) David, S.; de Sennyey, G.; Pascard, C.; Guilhem, J. J. Chem.
Soc., Chem. Commun. 1981, 780±781. (c) Maruyama, T.; Kimura,
S.; Sato, Y.; Honjo, M. J. Org. Chem. 1983, 48, 2719±2723.
(d) Ueda, T. Nucleosides Nucleotides 1985, 4, 67±75. (e) Sako,
M.; Shimada, K.; Hirota, K.; Maki, Y. J. Chem. Soc., Chem.
Commun. 1986, 1704±1706. (f) Maruyama, T.; Kimura, S.; Sato,
Y.; Honjo, M. Chem. Pharm. Bull. 1986, 34, 3623±3630.
(g) Yamagata, Y.; Tomita, K.; Usui, H.; Sano, T.; Ueda, T.
Chem. Pharm. Bull. 1989, 37, 1971±1976. (h) Chabbi, M.;
Vanlemmens, P.; Ronco, G.; Goethals, G.; Villa, P.; Ewing,
D. F.; Mackenzie, G. Collect. Czech. Chem. Commun. 1996, 61,
S12±S15. (i) Kittaka, A.; Kato, H.; Tanaka, H.; Nonaka, Y.;
Amano, M.; Nakamura, K. T.; Miyasaka, T. Tetrahedron 1999,
55, 5319±5344. (j) Tusa, G.; Reed, J. K. Nucleosides Nucleotides
Nucleic Acids 2000, 19, 805±813. An intramolecularly cyclized
50-uridylic acid featuring a tether between the pyrimidine C5 and
the phosphate group is exceedingly biomimetic, lacking only one
of the ionizing P±OH groups: (k) Seio, K.; Wada, T.; Sakamoto,
K.; Yokoyama, S.; Sekine, M. J. Org. Chem. 1996, 61, 1500±1504.
(l) Sekine, M.; Kurasawa, O.; Shohda, K.-i.; Seio, K.; Wada, T.
J. Org. Chem. 2000, 65, 3571±3578.
12. Tanaka, H.; Nasu, I.; Miyasaka, T. Tetrahedron Lett. 1979, 19,
4755±4758.
13. (a) Otter, B. A.; Patil, S. A.; Spada, M. R.; Jelicks, L. A.;
Yoshimura, Y.; Matsuda, A.; Klein, R. S. Nucleosides Nucleotides
1992, 11, 615±635. (b) Megati, S.; Ealick, S. E.; Naguib, F. N. M.;
el Kouni, M. H.; Klein, R. S.; Otter, B. A. Nucleosides Nucleotides
1994, 13, 2151±2168. (c) Sarma, M. S. P.; Klein, R. S.; Otter, B. A.
Nucleosides Nucleotides 1994, 13, 369±378.
14. (a) P®tzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc. 1965, 87,
5661±5670. (b) P®tzner, K. E.; Moffatt, J. G. J. Am. Chem. Soc.
1965, 87, 5670±5678.
15. Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155±4156.
16. (a) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc. 1972, 94,
8205±8212. (b) Altona, C.; Sundaralingam, M. J. Am. Chem. Soc.
1973, 95, 2333±2344.
17. Herzig, J.; Nudelman, A.; Gottlieb, H. E.; Fischer, B. J. Org.
Chem. 1986, 51, 727±730.
18. Watson, S. C.; Eastham, J. F. J. Organomet. Chem. 1967, 9,
165±167.
3. Groziak, M. P.; Lin, R. Arkivoc 2000, 1, 33±45.
4. Groziak, M. P.; Koohang, A. J. Org. Chem. 1992, 57, 940±944.
19. Linderman, R. J.; Graves, D. M. J. Org. Chem. 1989, 54, 661±
668.