Studies on the stereoselective synthesis of deuterated D-ribose derivatives

Article abstract of DOI:10.1002/hlca.200390062

In view of the importance of the site-specific substitution of the H-atom by its stable isotope ²H in a stereoselective/stereospecific manner at the pentose sugar residue, decreasing the spectral overcrowding in various regions of 1D and 2D hom

Full text of DOI:10.1002/hlca.200390062

Helvetica Chimica Acta Vol. 86 (2003)
633
Studies on the Stereoselective Synthesis of Deuterated d-Ribose
Derivatives
by Mrinal K. Kundu*a)b), Andras Fˆldesib), and Jyoti Chattopadhyayab)
a) Institute of Organic Chemistry, University of Basel, St. Johanns Ring 19, CH-4056 Basel
b) Department of Bioorganic Chemistry, Box-581, BMC, University of Uppsala, S-75123 Uppsala
In view of the importance of the site-specific substitution of the H-atom by its stable isotope 2H in a
stereoselective/stereospecific manner at the pentose sugar residue, decreasing the spectral overcrowding in
various regions of 1D and 2D homo- and heteronuclear correlation spectra of oligo-DNA and -RNA, there is
always a need for the development of new methods for the incorporation of 2H at different sites of a ribose.
High-yielding multistep syntheses of C(2)-, and (5R)- and (5S)-3,5-deuterated ribose derivatives have been
envisaged for the application of site-specific incorporation of multilabeled nucleosides into oligomers to
facilitate their structure elucidation by NMR spectroscopy. All syntheses started from d-glucose after proper
derivatization. In the case of C(2), > 97 atom-% isotope was incorporated, employing an inversion of the
configuration at C(2) as the key reaction. For C(5), two different routes were envisaged: on the one hand,
deuterated achiral reagent was treated with a conformationally locked sugar moiety 15, while, on the other,
chiral protonated sources were used to transfer the H-atom to a C(5)-deuterated aldehyde 18. In all cases,
enantiomeric and isotopic purities were found to be as high as > 97% as determined by NMR spectroscopy.
Introduction. The importance of the structure biological activity relationship of
an oligo-DNA and oligo-RNA has been investigated by different physicochemical
techniques, amongst which NMR spectroscopy was found to be the most powerful tool
as it provides the conformational data under quasi-physiological conditions [1].
Although, with the increasing chain length, the usefulness of NMR spectroscopy
becomes restricted, site-specific isotope labeling has been proven to overcome this
problem in the recent past (for a review, see [2]). While, the use of 13C/15N isotopes
2
increases the number of observable resonances [3][4][5 7], H labeling of oligonu-
cleotides is based on the primary idea of suppressing part(s) of the 1H-NMR spectrum
[8].
Partial or complete substitution of 1H by 2H in the sugar residue of a nucleoside and
incorporating them into an oligomer in a sequence specific manner has helped to solve
the above problem as described in −Uppsala NMR Window× concept [9]. Recently, the
results have been reported on the chemical synthesis of (5R)- and (5S)-[3',5'-2H2]nu-
cleosides [9] and [3',4',5',5''-2H4]nucleosides [10a] for the use in the preperation of non-
uniformly labeled DNA and RNA, and in their NMR measurements. In continuation of
the studies, it was envisioned that whether we can make a new window where the (5'R)-
[3',5'-2H2]- or (5'S)-[3',5'-2H2]ribonucleoside would be incorporated inside the NMR
window, whereas [2',3',4',5',5''-2H5]ribonucleosides would be placed outside the NMR
window so that the spectral assignments become clear, e.g., i) to enhance the
structurally important NOE intensities with diminished spin diffusion while removing
1
the insignificant ones, ii) to determine the exact H,31P coupling constants, provided

634
Helvetica Chimica Acta Vol. 86 (2003)
suitable general methods for 2H incorporation at C(2) and C(5) ((R) and (S)) could be
found. Nevertheless, this can also be implemented to prepare [2',3',5'(R and S)-2H3]-2'-
deoxyribonucleosides [10b] to incorporate into the NMR-window region of a large
oligo-DNA to reduce the spectral overcrowding for NMR structural determination
(Fig. 1). In this report, detail studies on the high-yielding, efficient general method-
ologies towards deuteration at C(2) and C(5)(R and S) at the sugar level starting from
d-glucose (considering that these protocols would be implemented in the case of the
synthesis of both 13C,2H-labeled nucleosides) have been sought. The strategies have
been envisioned in such a way that one protocol can be switched to another for further
labeling in different position. Two syntheses were carried out parallel; for simplicity,
C(2) labeling was performed on standard d-glucose, whereas C(5) was deuterated on a
C(3)-isotope-enriched sugar derivative to check the applicability of the method to
produce multi-labeled pentoses.
Fig. 1. Schematic representation of a typical NMR window
Results and Discussion. 1. Synthesis of [2-2H1]Ribose Derivative 9 (Scheme 1).
The high degree of isotope substitution at C(2) has been found as the most difficult task
either at the sugar level or at the nucleoside level as reported by Serianni et al. decades
ago [11]. Although some selective isotope labelings to produce epimeric mixtures have
already been reported, there were several limitations of those studies, e.g., the use of
highly toxic reagent such as KCN in the reactions usually carried out at relatively high
pH. Moreover, the cyanohydrin products were only stable at pH 4.0, so that the
chromatographic purifications were also performed under this condition. Considering
the acid- and base-labile protecting groups present in a pentose sugar (as in our case),
the application of these procedures becomes restricted. Nevertheless, for multiple
labeling at different sites of a ribose sugar (e.g., C(4) and C(5) together with C(3) and
C(2)), the methodologies were turned out to be certainly disadvantageous. The
oxidation followed by the reduction of 1,3,5-O-tribenzoylribofuranose has been less
2
attractive because of the low-level incorporation of H at C(2) as well as the presence
of the base-labile protecting groups in the ribofuranose moiety [12]. Although the

Helvetica Chimica Acta Vol. 86 (2003)
635
Raney NiD2O exchange reaction [9] of an anomeric mixture of methyl a/b-d-
ribofuranose offers an alternative route, the slow exchange rate at specific positions
(especially at the secondary C-atoms; hence, 3 4 cycles of isotope exchange are
necessary) eliminates the possibility of being noteworthy. On the other hand, the highly
stereoselective reduction of benzyl arabinopyranoside [13] to the corresponding
ribopyranoside is proved to be an effective but lengthy procedure. Therefore, there was
indeed a need for an efficient alternative synthesis that would overcome all limitations
mentioned above and could be used in multi-gram-scale preparation of multilabeled
deuterated nucleosides.
Scheme 1. Synthesis of 1-O-Methyl-b-d-[2-2H1]ribofuranose (9)
i) (COCl)2, DMSO, CH2Cl2, À 708; ii) LiAlD4/NaBD4, Et2O/EtOH, r.t. iii) p-TolCl, pyridine, r.t. iv) NH3/
MeOH, r.t. v) Tf2O, DMAP, pyridine, 08, CH2Cl2. vi) Cesium propionate, DMF, r.t. vii) NH3/MeOH. viii) Pd-C/
H2, EtOH, r.t.
Our methodology outlined in Scheme 1 is based on the oxidation reduction
process of methylribofuranoside derivative 1 (a/b 7:3) [14] to form C(2)-deuterated
compound 2, followed by inversion of configuration at C(2) of the arabino compound
2b to afford the sugar derivative 7. Thus, 1 was subjected to Swern oxidation at À 708,
followed by subsequent reduction with either LiAlD4 in dry Et2O or NaBD4 in EtOH at
room temperature to give a unseparable epimeric mixture (7:3 in favor of arabino) of
2. However, it was necessary to separate the epimers at this stage. Compound 2 was
treated with p-toluoyl chloride (1.15 equiv.) in dry pyridine overnight at room
temperature, and the toluoylated ribo and arabino derivatives 3 and 41), respectively,
were successfully separated over a silica-gel column with petroleum ether/AcOEt as eluent.
26
D
1
)
Data of 3: 26%. ‰aŠ ˆ ‡97.9 (c ˆ 0.67, CHCl3). 1H-NMR (CDCl3): 8.03 (d, J ˆ 8.1, 2 H, Tol); 7.33 7.23
(m, 12 H, Tol, Bn); 5.21 (s, HÀC(1)); 4.73 4.43 (m, PhCH2); 4.30 4.24 (m, HÀC(4)); 4.10 (d, J ˆ 4.45,
26
D
HÀC(3)); 3.52 3.33 (m, CH2(5), MeO); 2.42 (s, Me). Data of 4: 62%. ‰aŠ ˆ À74 (c ˆ 0.246, CHCl3).
1H-NMR (CDCl3): 7.92 (d, J ˆ 8.1, 2 arom. H); 7.36 7.23 (m, 12 arom. H); 5.21 (s, HÀC(1)); 4.66 4.59
(m, 2 CH2); 4.34 (d, J ˆ 5.5, HÀC(3)); 4.25 4.19( m, HÀC(4)); 3.65 3.54 (m, CH2(5)); 3.28 (s, MeO); 2.41
(s, Me).

636
Helvetica Chimica Acta Vol. 86 (2003)
As mentioned, the strategy of this study was to invert the configuration at C(2) of
the arabino compound 4 to that of the ribo counterpart 7, and to use 3 and 7 (after
proper derivatization) for base-coupling reaction to prepare the deuterated nucleo-
sides. Here, we made use of cesium-propionate, which acts as an effective nucleophile
in a Sn2 reaction provided there is a good leaving group such as triflate at the reaction
center. Therefore, the sugar derivative 4 was detoluoylated by methanolic NH3 at room
temperature in 2 days, followed by the removal of solvent and chromatographic
purification over silica gel, to afford methyl b-3,5-di-O-benzyl[2-2H1]arabinofuranoside
(5) in 95% yield2). The 2-O-triflate 6 was then obtained from 5 on treatment with triflic
anhydride in the presence of 4-(dimethylamino)pyridine (DMAP) and pyridine in
CH2Cl2. (It should be noted that, in the first instance, this reaction was carried out with
2 itself. Although, the ribo and arbino derivatives were also separated and the arabino-
triflate compound 6 could be obtained directly for nucleophilic substitution reaction,
we prefered the toluoylation and detoluoylation steps, as the ribo triflate could not be
deprotected cleanly for further use, leading to a low yield of the [2-2H1]sugar. On the
other hand, detoluoylation gave almost quantitative conversion.) Then, the arabino-
triflate was converted to ribo propionate 7 with cesium-propionate in DMF [15] at
room temperature under inert atmosphere. The change of the chemical shift from 5.0 to
4.87 for the anomeric H-atom confirmed that the displacement was indeed Sn2. Also,
the change of the peak positions of C(3) and C(4) (from 80.8 (C(3)) and 79.6 (C(4)) of
compound 6 to 80.3 (C(4)) and 77.7 (C(3)) of compound 7 as observed in the 13C-NMR
spectrum and confirmed by H,C-correlation experiment) showed that the inversion of
configuration at C(2) was very successful. Considering that 1-O-acetate provides better
coupling with the silylated base, and 2-O-toluoyl also helps to afford exclusively the b-
anomer, the deuterated ribo-propionate 7 was converted to 1-O-methyl-b-d-[2-2H1]-
ribose 9 as a precursor of the acetate. The appearance of a clean doublet (J(3,4)) at
4.08 ppm corresponding to HÀC(3) demonstrated the high-level (> 97 atom-%)
incorporation of 2H at C(2). The compound 9 can be easily converted to the
corresponding acetate, followed by coupling with silylated bases according to well-
established methods [9], to give [2-2H1]ribonucleosides.
2. Synthesis of (5R)- and (5S)-[3,5-2H2]Ribose Derivatives (Scheme 2). Stereo-
selective (5'S)- or (5'R)-2H1 incorporation provides means to determine the exact
3J(5',P) and 3J(5'',P) coupling constants and unambiguous NOE assignments, which are
essential to elucidate the conformation of sugar phosphate backbone of DNA
oligomers [16 19]. They are used to probe the internal dynamics of oligonucleotides
2
by solid-phase H-NMR spectroscopy [20][21]. C(5)-Deuterated compounds are also
used in conformational and mechanistic studies [22 27]. We herein report our studies
on the synthesis of (5R)- and (5S)-[3,5-2H2]ribose derivatives and development of the
procedures on C(5)-deuteration at sugar level.
The synthesis was carried out starting from 1,2 :5,6-O-diisopropylidene-[3-2H1]-d-
glucose (10), which was converted to compound 11 by a sequence of steps [9]
(Scheme 3). Towards our aim in synthesizing both (5R)- and (5S)-3,5-dideuterated
ribose, two different synthetic routes have been envisaged. In the first instance, achiral
1
2
)
Data of 5: H-NMR (CDCl3): 7.34 7.25 (m, 10 arom. H); 4.85 (s, HÀC(1)); 4.77 4.56 (m, 2 CH2); 4.26
4.06 (m, HÀC(4)); 3.84 (d, J ˆ 5.5, HÀC(3)); 3.55 3.52 (m, CH2(5)); 3.41 (s, MeO).

Helvetica Chimica Acta Vol. 86 (2003)
637
Scheme 2. Strategies towards the Derivatives of (5R)- and (5S)-[3,5-2H2]Ribose
Scheme 3. Synthesis of (5R)-1,5-Anhydro-2,3-O-isopropylidene-d-[3,5-2H2]ribose (16)
deuterated reagent was used to introduce 2H at C(5) of a conformationally locked sugar
moiety, while in the other, a H-atom was transferred from external protonated chiral
reagents to a C(5)-deutero aldehyde 18 (cf. Scheme 4).
i) Usual reduction with NaBD4 or LiAlD4 of the corresponding C(5)-aldehyde
produced a ca. 1:1 mixture of the unseparable (5R)- and (5S)-isomers [28] (Scheme 4).
The use of LiI and tert-pentyl alcohol in the presence of (À)-isoborneol improved the
selectivity; however, the best result provided only a ca. 4 :1 mixture in favor of the
(5R)-isomer [29]. In a recent report, it has been shown that 2H can be transfered easily
from (À)-[2-2H1]isobornyloxymagnesium bromide in ca. 100% enantiomer purity but
with concomittant nonlabeled ribose derivative up to 15% [30].
The absolute stereoselectivity and isotopic purity at C(5) of the ribose moiety at the
sugar level can be increased via photobromination [31] of 1,5-anhydro-2,3-O-
isopropylidene-d-ribose, followed by reductive elimination of bromine with different
deuteride reagents as isotope source amongst which −super deuteride× was found to be
the best. Encouraged by these findings, we thought of using this strategy for C(5)-

638
Helvetica Chimica Acta Vol. 86 (2003)
Scheme 4. Use of Protonated Chiral Reagents: Synthesis of (5R)- and (5S)-3-O-Benzyl-1,2-O-isopropylidene-d-
[3,5-2H2]ribose (21a and 21b, respectively)
deuteration on our [3-2H1]sugar derivative as shown in Scheme 3. Hence, compound
11a was reduced with NaBH4 in EtOH, followed by boiling with 80% AcOH in H2O to
give [3-2H1]ribose derivative 12 [9]. Acetone treatment [10b] of 12 in the presence of
catalytic amount of conc. H2SO4 at room temperature, followed by neutralization with
solid Na2CO3, gave compound 13 in 50% yield3). The acetonide 13 was then tosylated
at À 708 in dry pyridine with 1.1 equiv. of TsCl to give only C(5)-monotosyl compound,
which, after workup, was refluxed in dry MeCN in the presence of 1,8-diazabicy-
clo[5.4.0]undec-7-ene (DBU) for 4 h. Purification by silica-gel column chromatog-
raphy with AcOEt/hexane afforded 1,5-anhydro [3-2H1]compound 14 in 72% yield
(after two steps)4). The photobromination step was achieved according to [31]. The
appearance of signals at 5.76 and 5.75 (2s, HÀC(1), HÀC(5)), 4.90 (s, HÀC(4)), and
2
4.27 (s, HÀC(2)) indicated that photobromination was indeed stereoselective. H was
then introduced at C(5) in dry THF with LiEt3BD (Aldrich; 1.0m in THF, 2 equiv.) as
2H source. High-field NMR spectroscopy revealed that the (R)-16 was formed (62%)
with > 99% chiral purity5). Interestingly, during the deuteration reaction in our hand,
we observed the presence of small amount (3 6 atom-%) of nonlabeled anhydro
compound depending on the reaction conditions.
At this point, it was clear that, although we could obtain compound 16 only as (5R)-
isomer, the level of deuterium incorporation at C(5) was not satisfactory. Nevertheless,
the photobromination step was low-yielding (30 40% (also mentioned in [31]), which
3
)
Data of 13 (b-anomer): 1H-NMR (CDCl3): 5.42 (d, HÀC(1)); 4.58 (s, HÀC(2)); 4.41 (t, HÀC(4)); 3.75
3.71 (m, CH2(5)); 1.49, 1.32 (2s, Me). 13C-NMR (CDCl3): 112.0; 102.9(C(1)); 87.6 (C(4)); 86.7 (C(2));
63.5 (C(5)); 26.3; 24.6.
)
Data of 14: 1H-NMR (CDCl3): 5.44 (s, HÀC(1)); 4.70 (d, J ˆ 3.7, HÀC(4)); 4.28 (s, HÀC(2)); 3.42 (dd, J ˆ
4
5
3.8, 7.2, 1 HÀC(5) (R)); 3.30 (d, J ˆ 7.2, 1 HÀC(5) (S)).
)
1H-NMR (CDCl3): 5.45 (s, HÀC(1)); 4.70 (s, HÀC(4)); 4.28 (s, HÀC(2)); 3.41 (d, J ˆ 3.8, HÀC(5) (R)).
13C-NMR (CDCl3): 112.0; 99.7 (C(1)); 81.1 (C(2)); 77.3 (C(4)); 25.8; 25.1.

Helvetica Chimica Acta Vol. 86 (2003)
639
would result in the reduction of overall yield), and we cannot afford this especially
considering that we were working with valuable C(3)-deuterated compound. Instead, if
we could use C(5)-deuterated aldehyde 11b in lieu of the derivative 11a as substrate,
1
1
and incorporate H from protonated chiral reagents, the H contamination would
definitely vanish. We wish to report our findings on this aspect as below.
ii) Protected [3-2H1]aldehyde 11b (R1 ˆ H, R2 ˆ Bn) was converted to [3,5,5-
2H3]ribose moiety 17 (Scheme 4) according to [10a]. The trideutero derivative 17 was
then subjected to Moffat oxidation [32] to give the [3,5-2H2]aldehyde 18, which was
ready for proton-transfer reactions. We used two different chiral reagents. First, the
Grignard reagent 19 [33] (obtained from (‡)-camphor after reduction with LiAlH4,
followed by the addition to a solution of BuMgBr in dry Et2O) was tried. When the
aldehyde 18 was refluxed with (À)-isobornyloxymagnesium bromide (19) in benzene
overnight, (5R)-3-O-benzyl-1,2-O-isopropylidene[3,5-2H2]ribose (21a; 61% yield after
1
2 steps (oxidation and reduction)) was obtained6). High-field H-NMR spectroscopy
1
revealed that there was no H/2H exchange taking place during the reaction, and
isotopic purity was > 97 atom-% (Fig. 2).
Terpenes, being replenishable source of chiral carbon compounds available both in
cyclic (e.g., (‡)-a-pinene) and acyclic forms, and displaying a plethora of conforma-
tional features, have been recognized as wealth of stereochemical attributes. Decades
ago, Brown et al. have shown the usefulness of pinene in synthesizing various chiral
reagents (e.g., chloro(diisopinocampheyl)borane (Ipc2BCl, 20) for the reductions of
carbonyl compounds with unprecedented stereoselectivity [34]. Thus, by the addition
of Ipc2BCl (2.5 equiv.) to a solution of 18 in dry THF at À 308 under N2, (5S)-isomer
21b was obtained (45%, after 2 steps). The 1H-NMR signals in CDCl3 at 5.73 (d, J ˆ 3.8,
HÀC(1)); 4.76, 4.60 (2d, J ˆ 10.3, PhCH2); 4.56 (d, J ˆ 3.6, HÀC(2)); 4.11 (d, J ˆ 2.8,
HÀC(4)); and 3.61 (br. s, HÀC(5) (S)) were in accordance with the structure (Fig. 2).
The formation of two different enantiomers by using two different chiral reagents could
well be understood from the proposed models as shown (Fig. 3).
In conclusion, we have demonstrated different methodologies for the synthesis of
the [2-2H1]-, and (5R)- and (5S)-[3,5-2H2]ribose derivatives with high isotope enrich-
ment and high optical purity. These derivatives can be easily converted to the
deuterated nucleosides according to already established reaction conditions. It is worth
noting that, by using two different protonated chiral reagents, in the case of C(5)-
labeling, 18 afforded two different isomers, (5R) and (5S), which is certainly
advantageous for us as we would obtain two different sets of informations about
3J(5,P) couplings in our NMR analysis of oligo-deoxyribonucleic acids. The use of these
protocols in large-scale production of (R)- and (S)-3,5-dideutero ribonucleosides and
(5'R)- and (5'S)-2'-deoxy[2',3',5'-2H2]ribonucleosides (in combination of C(2)- and
C(5)-deuteration procedures) and their incorporation into the oligomer together with
their structural studies will be our following target.
Data of 21a: 1H-NMR (CDCl3): 5.73 (d, J ˆ 3.7, HÀC(1)); 4.77, 4.59(2 d, J ˆ 11.9, PhCH2); 4.56 (d, J ˆ 3.7,
HÀC(2)); 4.11 (d, J ˆ 2.3, HÀC(4)); 3.89(br. s, HÀC(5) (R)).
6
)

640
Helvetica Chimica Acta Vol. 86 (2003)
Fig. 2. Comparison of the 270-MHz NMR spectrum (sugar region) of the d-[3,5-2H2]ribose derivatives 17, 21a,
and 21b

Helvetica Chimica Acta Vol. 86 (2003)
641
Fig. 3. Induction of chirality: proposed mechanism of 1H-transfer to the aldehyde 18
Experimental Part
General. All dry solvents were prepared according to standard procedures and stored over molecular sieves.
The chromatographic separations were performed on Merck G60 silica gel with AcOEt/petroleum ether. TLC:
Merck pre-coated silica-gel 60 F 254 glass baked plates. 1H- and 13C-NMR spectra: Jeol GX 270 spectrometer at
270 and 67.9MHz, resp., with TMS or MeCN as reference; chemical shifts are reported in ppm. Electron-spray
ionization (ESI) mass spectra: LCTTM oa-TOF mass spectrometer (Micromass, Manchester, UK).
Some Representative Procedures. 3,5-Di-O-benzyl-1-O-methyl-a/b-d-[2-2H1]arabino/ribofuranose (2). In a
two-necked round bottom flask fitted with a dropping funnel and N2-gas inlet was placed oxalyl chloride (3.7 ml,
42.3 mmol) in dry CH2Cl2 (15 ml) at À 708. A soln. of DMSO (6 ml, 84.6 mmol) in CH2Cl2 (15 ml) was added
slowly, followed by the dropwise addition of a soln. of 1 (11.25 g, 32.7 mmol, in 35 ml of CH2Cl2), and the
resultant mixture was stirred for 4 h at the same temp. Et3N (23 ml, 163 mmol) was added to quench the
reaction, and then the mixture was heated slowly to r.t. H2O was added, and the compound was extracted with
CH2Cl2. The org. layer was successively washed with brine and H2O, dried (MgSO4), and concentrated. The
residue was dissolved in dry Et2O (100 ml), LiAlD4 (800 mg, 19mmol) was added in portions at 0 8, and the
mixture was stirred overnight at r.t. H2O was added slowly to quench the reaction, and the mixture was extracted
with CH2Cl2. Removal of the solvent and purification by column chromatography (CC) gave compound 2 (7.2 g,
64%, after two steps). 1H-NMR (CDCl3): 7.34 7.27 (m, 10 arom. H); 4.88, 4.85 (2s, HÀC(1) (a and b)); 4.77
4.56 (m, PhCH2); 4.26 4.22, 4.16 4.06 (2m, HÀC(4) (a and b)); 3.84, 3.78 (2d, HÀC(3) (a and b)); 3.54 3.52
(m, CH2(5)); 3.47, 3.41 (s, Me (a and b)). 13C-NMR (CDCl3): 137.9; 129.7; 129.6; 128.2; 127.8; 127.6; 127.5; 102.5
(C(1)); 84.5 (C(3)); 80.7 (C(4)); 73.2 (CH2); 71.9(C(5)); 71.7 (CH 2); 55.3 (Me).
3,5-Di-O-benzyl-1-O-methyl-2-O-(trifluoromethylsulfonyl)-b-d-[2-2H1]arabinofuranose (6). The ribose
derivative 5 (4.18 g, 12.11 mmol) was co-evaporated with dry pyridine and dissolved in dry CH2Cl2 (90 ml),
followed by the addition of 4-(dimethylamino)pyridine (DMAP; 5.18 g, 42.4 mmol) and pyridine (9ml). The
mixture was cooled to 08 and triflic anhydride (2.8 ml, 16.96 mmol) was slowly added, and the resultant mixture
was stirred for 3 h at the same temp. The reaction mixture was then poured into cold sat. NaHCO3 soln., and
CH2Cl2 layer was separated. The H2O phase was extracted again with CH2Cl2, and the combined CH2Cl2 layer
27
D
was dried (MgSO4) and concentrated. Chromatographic purification yielded 6 (5.18 g, 89.5%). ‰aŠ ˆ À64.2
(c ˆ 0.74, CHCl3). 1H-NMR (CDCl3): 7.35 7.20 (m, 10 arom. H); 5.00 (s, HÀC(1)); 4.76 4.47 (m, 2 CH2); 4.29
(d, J ˆ 5.4, HÀC(3)); 4.17 4.11 (m, HÀC(4)); 3.58 3.44 (m, CH2(5)); 3.38 (s, Me). 13C-NMR (CDCl3): 137.6;
136.8; 128.4; 128.3; 127.7; 127.6 (arom. C); 100.3 (C(1)); 80.8 (C(3)); 79.6 (C(4)); 73.4 (CH2); 72.5 (CH2); 71.3
‡
(C(5)); 55.4 (Me). HR-EI-MS: 477.1184 (M ; calc. 477.1179).
3,5-Di-O-benzyl-1-O-methyl-2-O-propanoyl-b-d-[2-2H1]ribofuranose (7). Cesium propanoate (2.9g,
14.12 mmol) was added to a soln. of 6 (5.18 g, 10.86 mmol) in dry DMF (60 ml), and the mixture was stirred
for 36 h at r.t. DMF was removed under reduced pressure, H2O was added, and the compound was extracted
with CH2Cl2. After removal of CH2Cl2, the residue was chromatographed over silica gel to give 7 (3.2 g, 73.5%).
26
‰aŠ ˆ ‡14 (c ˆ 0.713, CHCl3). 1H-NMR (CDCl3): 7.33 7.25 (m, 10 arom. H); 4.87 (s, HÀC(1)); 4.61 4.38
D
(m, 2 CH2); 4.25 4.19( m, HÀC(4)); 4.12 (d, J ˆ 7.6, HÀC(3)); 3.63 3.59, 3.53 3.47 (2dd, CH2(5)); 3.33
(s, MeO); 2.44 2.36 (q, CH2O); 1.13 (t, Me). 13C NMR (CDCl3): 173.0 (CO); 138.0; 137.4; 128.2; 127.8; 127.7;
127.5; 106.2 (C(1)); 80.3 (C(4)); 77.7 (C(3)); 73.1, 72.9(CH 2); 71.0 (C(5)); 54.9; 27.3; 8.9. HR-EI-MS: 401.1955
‡
(M ; calc. 401.1949).
3,5-Di-O-benzyl-1-O-methyl-b-d-[2-2H1]ribofuranose (8): Compound 7 (2.7 g, 6.73 mmol) was treated with
NH3/MeOH under stirring at r.t. for 30 h. Usual workup, followed by removal of the solvent and washing with
26
H2O, gave chromatographically homogeneous 8 (2.3 g, 99%). ‰aŠ ˆ À28.7 (c ˆ 0.71, CHCl3). 1H-NMR
D

642
Helvetica Chimica Acta Vol. 86 (2003)
(CDCl3): 7.37 7.25 (m, 10 arom H); 4.86 (s, HÀC(1)); 4.57 (s, 2 CH2); 4.26 4.20 (q, HÀC(4)); 4.07 (d, J ˆ 6.2);
3.54 (d, J ˆ 5.3, CH2(5)); 3.31 (s, Me). 13C-NMR (CDCl3): 138.0; 137.0; 128.5; 128.2; 127.5; 108.4 (C(1)); 80.5
(C(4)); 79.4 (C(3)); 73.2 (PhCH2); 72.7 (PhCH2); 71.5 (C(5)); 54.9(Me).
1-O-Methyl-b-d-[2-2H1]ribofuranose (9). The Bn groups of 8 (2.1 g, 6.08 mmol) were cleaved with Pd/
CÀH2 (450 mg) in EtOH (40 ml) during 3 h at r.t. The reagent was filtered over Celite, and the filtrate was
25
concentrated to dryness to afford 9 (980 mg, 98%). ‰aŠ ˆ À38 (c ˆ 0.15, H2O). 1H-NMR (D2O): 4.83
D
(s, HÀC(1)); 4.08 (d, J ˆ 6.9, HÀC(3)); 3.97 3.91 (m, HÀC(4)); 3.76 3.70, 3.57 3.50 (2dd, J ˆ 3.3, 12.2, and
6.4, 12.2, CH2(5)); 3.33 (s, Me). 13C-NMR (D2O): 107.7 (C(1)); 82.6 (C(4)); 70.5 (C(3)); 62.6 (C(5)); 54.9(Me).
‡
HR-EI-MS: 165.0748 (M ; calc. 165.0747).
(5R)-3-O-Benzyl-1,2-O-isopropylidene-d-[3,5-2H2]ribofuranose (21a). A soln. of BuBr (2.9ml, 27.7 mmol)
in dry Et2O (10 ml) was added dropwise to Mg (650 mg, 27.2 mmol) in dry Et2O (10 ml) under N2. After
complete dissolution of Mg, a soln. of isoborneol (3.9g, 25.5 mmol) in dry Et 2O (10 ml) was added slowly. A
white precipitate was observed immediately. Benzene (30 ml) was added, and Et2O was distilled under N2.
The trideutero compound 17 (570 mg, 2.0 mmol) was dissolved in dry DMSO (4 ml), followed by the
addition of N,N'-dicyclohexylcarbodiimide (DCC; 1.24 g, 6 mmol) and Cl2CHCOOH (80 ml, 1 mmol). The
mixture was stirred for 2 h at r.t. Oxalic acid (570 mg) in MeOH (7 ml) was added to quench, and precipitate was
formed upon addition of AcOEt, which was filtered off. The filtrate was concentrated, redissolved in AcOEt,
and the org. layer was washed with sat. NaHCO3 soln. and H2O. AcOEt layer was then dried (MgSO4) and
concentrated to give 18, which was used directly without further purification.
The aldehyde 18 (475 mg, 1.7 mmol) was co-evaporated with dry benzene three times, dissolved in 5 ml of
benzene, and added to the Grignard reagent prepared as described above. The mixture was refluxed for 13 h.
After cooling to r.t., 0.1n HCl was added, and the mixture was stirred for 10 min. The org. compounds were
extracted with CHCl3 (200 ml) and washed successively with sat. NaHCO3 (2 Â 100 ml), brine (1 Â 100 ml), and
H2O (1 Â 100 ml). The org. layer was then dried (MgSO4), concentrated, and the residue was subjected to CC to
give 21a (61%).
(5S)-3-O-Benzyl-1,2-O-isopropylidene-d-[3,5-2H2]ribofuranose (21b). A soln. of chloro(diisopinocam-
pheyl)borane (Ipc2BCl; 1.6 g, 5 mmol) in dry THF (5 ml) was added to a soln. of 18 (2 mmol; prepared as
described above for 21a) at À 308 dropwise, and the mixture was stirred for 6 h under inert atmosphere. Then,
the mixture was allowed to warm to r.t. slowly. The volatile materials were removed in vacuo, and the residue
was dissolved in Et2O (10 ml), followed by the addition of 2,2'-iminobis[ethanol] (1.2 ml), and the mixture was
stirred for 2 h again at r.t. After filtering the precipitate formed, the filtrate was concentrated, and the residue
was purified over silica gel to give 21b (45%).
REFERENCES
[1] K. W¸thrich, −NMR of Proteins and Nucleic Acids×, John Wiley & Sons, New York, 1986.
[2] A. Fˆldesi, A. Trifonova, M. K. Kundu, J. Chattopadhyaya, Nucleosides Nucleotides Nucleic Acids 2000, 19,
1615.
[3] E. P. Nikonowicz, A. Sirr, P. Leagault, F. M. Jucker, L. M. Pardi, A. Pardi, Nucleic Acids Res. 1992, 20, 4507.
[4] R. T. Batey, J. L. Battiste, J. R. Williamson, Methods Enzymol. 1995, 261, 300.
[5] J. R. Wyatt, M. Chastain, J. D. Puglisi, Bio Techniques 1991, 11, 764.
[6] J. Milecki, J. Labelled Compd. Radiopharm. 2002, 45, 307.
[7] I. M. Lagoja, P. Herdewijn, Synthesis 2002, 301.
[8] F. S. Ezra, C. H. Lee, N. S. Kondo, S. S. Danyliuk, R. H. Sarma, Biochemistry 1977, 16, 1977.
[9] A. Fˆldesi, F. P. R. Nilson, C. Glemarec, J. Chattopadhyaya, Tetrahedron 1992, 48, 9033.
[10] a) A. Trifonova, A. Fˆldesi, Z. Dinya, J. Chattopadhyaya, Tetrahedron 1999, 55, 4747; b) M. K. Kundu, A.
Fˆldesi, J. Chattopadhyaya, Collect. Czech. Chem. Commun., Symp. Ser. 1999, 2, 47; c) M. K. Kundu, A.
Trifonova, Z. Dinya, A. Fˆldesi, J. Chattopadhyaya, Nucleosides Nucleotides Nucleic Acids 2001, 20, 1333.
[11] a) A. S. Serianni, R. Barker, Can. J. Chem. 1979, 57, 3160; b) G. D. Wu, A. S. Serianni, R. Baker, J. Org.
Chem. 1983, 48, 1750; c) C. H. Han, L. O. Sillerud, Carbohydr. Res. 1986, 147, 247.
[12] G. P. Cook, M. M. Greenberg, J. Org. Chem. 1994, 59, 4704.
[13] J. C. Wu, H. Bazin, J. Chattopadhyaya, Tetrahedron 1987, 43, 2355.
[14] G. Ritzmann, R. S. Klein, D. H. Hollenberg, J. J. Fox, Carbohydr. Res. 1975, 39, 227.
[15] W. H. Kruizinga, B. Strijtveen, R. M. Kellog, J. Org. Chem. 1981, 46, 4321.
[16] A. Fˆldesi, T. V. Maltseva, Z. Dinya, J. Chattopadhyaya, Tetrahedron 1998, 54, 14487.

Helvetica Chimica Acta Vol. 86 (2003)
643
[17] S.-I. Tate, Y. Kubo, A. Ono, M. Kainosho, J. Am. Chem. Soc. 1995, 117, 7277.
[18] A. Ono, T. Makita, S.-I. Tate, E. Kawashima, Y. Ishido, M. Kainosho, Magn. Reson. Chem. 1996, 34, S40.
[19] C. Kojima, E. Kawashima, K. Toyama, K. Ohshima, Y. Ishido, M. Kainosho, Y. Kyogoku, J. Biomol. NMR
1998, 11, 103.
[20] T. M. Alam, J. Orban, G. P. Drobny, Biochemistry 1991, 30, 9229.
[21] D. Khare, J. Orban, Nucleic Acids Res. 1992, 20, 5131.
[22] R. G. S. Ritchie, A. S. Perlin, Carbohydr. Res. 1977, 55, 121.
[23] G. Puzo, K. H. Schram, J. G. Liehr, J. A. McCloskey, J. Org. Chem. 1978, 43, 767.
[24] A. Kintanar, T. M. Alam, W.-C. Huang, D. C. Schindele, D. E. Wemmer, D. Drobny, J. Am. Chem. Soc.
1988, 110, 6367.
[25] J. J. De Voss, C. A. Townsend, W.-D. Ding, G. O. Morton, G. A. Ellestad, N. Zelin, A. B. Tabor, S. L.
Schreiber, J. Am. Chem. Soc. 1990, 112, 9669.
[26] B. L. Frank, L. J. Worth, D. F. Christner, J. W. Kozarich, J. Stubbe, L. S. Kappen, I. H. Goldberg, J. Am.
Chem. Soc. 1991, 113, 2271.
[27] J. J. Hangeland, J. J. De Voss, J. A. Healf, C. A. Townsend, W.-D. Dung, J. S. Ascroft, G. A. Ellestad. J. Am.
Chem. Soc. 1992, 114, 9200.
[28] A. M. Ono, T. Shiina, A. Ono, M. Kainosho, Tetrahedron Lett. 1998, 39, 2793.
[29] Y. Ooga, A. M. Ono, A. S. Ono, M. Kainosho, Nucleic Acids Symp. Ser. 1996, 35, 77.
[30] A. M. Ono, A. Ono, M. Kainosho, Tetrahedron Lett. 1997, 38, 395.
[31] H. Ohrui, H. Meguro, Nucleic Acids Symp. Ser. 1986, 17, 29 .
[32] K. E. Pfitzner, J. G. Moffat, J. Am. Chem. Soc. 1965, 87, 5661.
[33] a) A. J. Streitwieser, J. R. Wolfe, J. Am. Chem. Soc. 1957, 79, 903; b) A. J. Streitweiser, M. R. Granger, J.
Org. Chem. 1967, 32, 1528.
[34] a) H. C. Brown, P. K. Jadahav, J. Am. Chem. Soc. 1983, 105, 2092; b) J. Chandrasekharan, P. V.
Ramachandran, H. C. Brown, J. Org. Chem. 1985, 50, 5446; c) H. C. Brown, P. V. Ramachandran, J.
Chandrasekharan, Organometallics 1986, 5, 2138; d) H. C. Brown, J. Chandrasekharan, P. V. Ramachan-
dran, J. Am. Chem. Soc. 1988, 110, 1539; e) H. C. Brown, M. Srebnik, P. V. Ramachandran, J. Org. Chem.
1989, 54, 1577.
Received November7, 2002