Toward l-homo-DNA: Stereoselective de novo synthesis of β-L-erythro-hexopyranosyl nucleosides

Article abstract of DOI:10.1021/jo100691y

A novel route to 2′,3′-dideoxy-β-l-erythro-hexopyranosyl nucleosides equipped with a 1′-(N⁶-benzoyladenin-9-yl) or a 1′-(thymin-1-yl) moiety has been developed. Synthesis of the enantiopure sugar moiety was carried out by a de novo approach based on a domino reaction as the key step. N-Glycosidation was explored via either nucleobase-transfer mechanism (B = T) or in situ anomerization (B = A or T), affording target nucleosides with high overall stereoselectivity.

Full text of DOI:10.1021/jo100691y

pubs.acs.org/joc
Toward L-Homo-DNA: Stereoselective de Novo Synthesis of
β-L-erythro-Hexopyranosyl Nucleosides
Daniele D’Alonzo,*,† Annalisa Guaragna,† Arthur Van Aerschot,‡ Piet Herdewijn,‡ and
Giovanni Palumbo†
†
ꢀ
Dipartimento di Chimica Organica e Biochimica, Universita di Napoli Federico II, via Cinthia 4, I-80126
Napoli, Italy, and ‡Katholieke Universiteit Leuven, Rega Institute for Medical Research,
Minderbroederstraat 10, B-3000 Leuven, Belgium
dandalonzo@unina.it
Received April 9, 2010
A novel route to 20,30-dideoxy-β-L-erythro-hexopyranosyl nucleosides equipped with a 10-(N6-benzoyl-
adenin-9-yl) or a 10-(thymin-1-yl) moiety has been developed. Synthesis of the enantiopure sugar moiety
was carried out by a de novo approach based on a domino reaction as the key step. N-Glycosidation was
explored via either nucleobase-transfer mechanism (B = T) or in situ anomerization (B = A or T),
affording target nucleosides with high overall stereoselectivity.
Introduction
with natural and/or modified complements.4 The strong self-
pairing aptitudes, combined with the natural abundance of
pento- and hexopyranoses, have stimulated questions on
Nature’s choice for pentose over hexose nucleic acids.3,5
Among the many nucleic acid alternatives with prebiotic
relevance, the so-called “β-HomoDNA” (a DNA analogue
comprised of (40f60)-linked 20,30-dideoxy-β-D-erythro-hexo-
pyranosyl nucleotides, Figure 1) represented a long inves-
tigated synthetic model system, because of its intriguing
structure6 and pairing properties.7 On the other hand, depend-
ing on the ability to cross-communicate with natural/modified
complements, applications by six-membered oligonucleotides in
gene expression inhibition (D-HNA,8 D-ANA,9 D-CeNA10),
In the search for nucleic acid analogues to be used in
therapy, diagnostics, synthetic biology, and etiology-
oriented research, a wide variety of oligonucleotide architec-
tures have been devised over the years, exploiting the well-
known concept of conformational restriction of the corre-
sponding monomeric units to improve binding efficiencies.1
Among the numerous examples of constrained oligonucleo-
tides endowed with hybridization potential, those bearing a
D- or L-six-membered ring as sugar moiety (Figure 1) cover a
privileged position. Owing to the higher conformational
rigidity compared to furanoses, oligonucleotides with a six-
membered “carbohydrate” mimic in the backbone1,2 have
displayed in most cases either excellent self-hybridization
properties3 or a notable capacity for cross-communication
(5) Eschenmoser, A. Science 1999, 284, 2118–2124.
(6) (a) Egli, M.; Lubini, P.; Pallan, P. S. Chem. Soc. Rev. 2007, 36, 31–45.
(b) Egli, M.; Pallan, P. S. Annu. Rev. Biophys. Biomol. Struct. 2007, 36, 281–
305.
*To whom correspondence should be addressed. Phone/Fax: þ39 081
674119.
(1) (a) Leumann, C. J. Bioorg. Med. Chem. 2002, 10, 841–854. (b) Zhou,
C.; Chattopadhyaya, J. Curr. Opin. Drug Discovery Dev. 2009, 12, 876–898.
(c) Herdewijn, P. Chem. Biodiversity 2010, 7, 1–59.
(2) Lescrinier, E.; Froeyen, M.; Herdewijn, P. Nucleic Acids Res. 2003, 31,
2975–2989.
(3) Beier, M.; Reck, F.; Wagner, T.; Krishnamurthy, R.; Eschenmoser, A.
Science 1999, 283, 699–703.
(4) Kerremans, L.; Schepers, G.; Rozenski, J.; Busson, R.; Van Aerschot,
A.; Herdewijn, P. Org. Lett. 2001, 3, 4129–4132.
(7) Eschenmoser, A. Chimia 2005, 59, 836–850.
(8) (a) Hendrix, C.; Rosemeyer, H.; Verheggen, I.; Seela, F.; Van Aerschot,
A.; Herdewijn, P. Chem.;Eur. J. 1997, 3, 110–120. (b) Kang, H.; Fisher, M. H.;
Xu, D.; Miyamoto, Y. J.; Marchand, A.; Van Aerschot, A.; Herdewijn, P.;
Juliano, R. L. Nucleic Acids Res. 2004, 32, 4411–4419.
(9) (a) Allart, B.; Khan, K.; Rosemeyer, H.; Schepers, G.; Hendrix, C.;
Rothenbacher, K.; Seela, F.; Van Aerschot, A.; Herdewijn, P. Chem.;Eur.
J. 1999, 5, 2424–2431. (b) Fisher, M.; Abramov, M.; Van Aerschot, A.; Xu,
D.; Juliano, R. L.; Herdewijn, P. Nucleic Acids Res. 2007, 35, 1064–1074.
(c) Fisher, M.; Abramov, M.; Van Aerschot, A.; Rozenski, J.; Dixit, V.;
Juliano, R. L.; Herdewijn, P. Eur. J. Pharmacol. 2009, 606, 38–44.
6402 J. Org. Chem. 2010, 75, 6402–6410
Published on Web 09/08/2010
DOI: 10.1021/jo100691y
r
2010 American Chemical Society

D’Alonzo et al.
JOCArticle
SCHEME 1. Access to β-L-erythro-Hexopyranosyl Nucleo-
sides: Retrosynthetic Analysis
halides17 and sulfides18 or the participation of a C-30 group19
have been employed to control anomeric β-stereochemistry.
On the other hand, a few stereoselective approaches have
been devised for the preparation of pyranosyl nucleosides,
which are mainly restricted to those belonging to the D-series,
owing to the wide availability of the corresponding starting
materials.20 To the best of our knowledge, the only asym-
metric route to deoxy-hexopyranosyl nucleosides belonging
to the L-series, limited to adenine and 6-chloropurine analo-
gues, has been recently achieved by Pd-catalyzed N-glycosy-
lation reaction carried out on an unsaturated L-pyranone.21
The lack of methodologies toward 20-deoxy- and 20,30-di-
deoxy-L-hexopyranosyl nucleosides are mainly due to the
synthetic difficulty to have easy access to deoxy-hexoses
belonging to the rare L-series in enantiomeric pure form. In
this paper, the stereoselective synthesis of nucleosides 1 and
2, as building blocks for β-L-HomoDNA synthesis, was
envisioned through the use of bicyclic intermediate 4
(Scheme 1). Reactivity of the latter has long been at the core
of our investigations on the total synthesis of L-hexoses.22
FIGURE 1. Six-membered D- and L-oligonucleotide analogues.
DNA-based diagnostics (D-HNA and D-ANA,11 β-D-Homo-
DNA12), and chiral and conformational selection of oligonu-
cleotides (D- and L-CNA13) have also been found. Recently,
incorporation of D-hexopyranosyl nucleotides into a natural
DNA duplex mediated by vent-DNA polymerase was
reported,14 demonstrating that the six-membered sugar ring of
hexopyranosyl nucleosides can be recognized by the binding
cavity of some enzymes. Also, the capability of mirror images of
hexitol nucleic acids (L-HNA) to form heterochiral duplexes
with enantiomeric D-complements was pointed out.15 On the
basis of these remarks, our ongoing interest in sugar-modified
oligonucleotides prompted us to examine other potential base-
pairing systems, such as the 20,30-dideoxy-β-L-erythro-hexopyr-
anosyl nucleic acids (β-L-HomoDNA) (Figure 1). In this paper,
attention was primarily given to the synthesis of the correspond-
ing nucleoside building blocks, for which a highly stereoselective
de novo synthetic approach has been described.
The stereocontrolled synthesis of 20-deoxy- or 20,30-di-
deoxy-β-D- and β-L-nucleosides is a difficult task and a
challenging synthetic target, due to the lack of control
elements at the C-20 position. Investigations have mainly
been focused on 20-deoxy-furanosyl nucleosides and their
analogues, as a consequence of the great interest aroused by
antiviral research.16 For such a purpose, use of glycosyl
Results and Discussion
The synthesis began with the three-carbon homologation
of methyl R,β-isopropylidene-L-glycerate (6) (Scheme 2). As
the coupling reaction between 5 and 6 was accomplished,
acetate anti-8b was obtained in a stereoselective manner, as
previously described.22b Treatment of the latter with DDQ in
CH2Cl2/MeOH smoothly afforded bicycle 4a as an anomeric
mixture (R/β = 85/15). This reaction can be regarded as a
domino process,23,24 since several synthetic transformations
(PMB group removal, oxidation of the resulting primary
alcohol, isopropylidene group removal, and subsequent
(10) (a) Wang, J.; Verbeure, B.; Luyten, I.; Lescrinier, E.; Froeyen, M.;
Hendrix, C.; Rosemeyer, H.; Seela, F.; Van Aerschot, A.; Herdewijn, P. J.
Am. Chem. Soc. 2000, 122, 8595–8602. (b) Nauwelaerts, K.; Fisher, M.;
Froeyen, M.; Lescrinier, E.; Van Aerschot, A.; Xu, D.; DeLong, R.; Kang,
H.; Juliano, R. L.; Herdewijn, P. J. Am. Chem. Soc. 2007, 129, 9340–9348.
(11) (a) Abramov, M.; Schepers, G.; Van Aerschot, A.; Van Hummelen,
P.; Herdewijn, P. Biosens. Bioelectron. 2008, 23, 1728–1732. (b) Wang, G.;
Bobkov, G. V.; Mikhailov, S. N.; Schepers, G.; Van Aerschot, A.; Rozenski,
J.; Van der Auweraer, M.; Herdewijn, P.; De Feyter, S. ChemBioChem 2009,
10, 1175–1185.
(12) Crey-Desbiolles, C.; Ahn, D.-R.; Leumann, C. J. Nucleic Acids Res.
2005, 33, e77.
(18) Sugimura, H.; Osumi, K.; Kodaka, Y.; Sujino, K. J. Org. Chem.
1994, 59, 7653–7660.
(13) (a) Maurinsh, Y.; Rosemeyer, H.; Esnouf, R.; Medvedovici, A.;
Wang, J.; Ceulemans, G.; Lescrinier, E.; Hendrix, C.; Busson, R.; Sandra, P.;
Seela, F.; Van Aerschot, A.; Herdewijn, P. Chem.;Eur. J. 1999, 5, 2139–
2150. (b) Froeyen, M.; Morvan, F.; Vasseur, J.-J.; Nielsen, P.; Van Aerschot,
A.; Rosemeyer, H.; Herdewijn, P. Chem. Biodiversity 2007, 4, 803–817.
(14) Renders, M.; Abramov, M.; Froeyen, M.; Herdewijn, P. Chem.;
Eur. J. 2009, 15, 5463–5470.
(19) (a) Young, R. J.; Shaw-Ponter, S.; Hardy, G. W.; Mills, G. Tetra-
hedron Lett. 1994, 35, 8687–8690. (b) Choi, W.-B.; Wilson, L. J.; Yeola, S.;
Liotta, D. C.; Schinazi, R. F. J. Am. Chem. Soc. 1991, 113, 9377–9379.
(20) For early results, see for example: Nord, L. D.; Dalley, N. K.;
McKerman, P. A.; Robins, R. K. J. Med. Chem. 1987, 30, 1044–1054.
(21) Guppi, S. R.; Zhou, M.; O’Doherty, G. A. Org. Lett. 2006, 8, 293–
296.
(15) (a) D’Alonzo, D.; Guaragna, A.; Van Aerschot, A.; Herdewijn, P.;
Palumbo, G. Tetrahedron Lett. 2008, 49, 6068–6070. (b) D’Alonzo, D.; Van
Aerschot, A.; Guaragna, A.; Palumbo, G.; Schepers, G.; Capone, S.;
Rozenski, J.; Herdewijn, P. Chem.;Eur. J. 2009, 15, 10121–10131.
(16) Mehellou, Y.; De Clercq, E. J. Med. Chem. 2010, 53, 521–538.
(17) For a recent example, see: Arico, J. W.; Calhoun, A. K.; McLaughlin,
L. W. J. Org. Chem. 2010, 12, 120–122.
(22) (a) D’Alonzo, D.; Guaragna, A.; Palumbo, G. Curr. Org. Chem.
2009, 13, 71–98. (b) Guaragna, A.; D’Alonzo, D.; Paolella, C.; Napolitano,
C.; Palumbo, G. J. Org. Chem. 2010, 75, 3558–3568.
(23) See: Domino Reactions in Organic Synthesis; Tietze, L. F., Brasche,
G., Gericke, K. M., Eds.; Wiley-VCH: Weinheim, Germany, 2006.
(24) D’Alonzo, D.; Guaragna, A.; Napolitano, C.; Palumbo, G. J. Org.
Chem. 2008, 73, 5636–5639.
J. Org. Chem. Vol. 75, No. 19, 2010 6403

D’Alonzo et al.
JOCArticle
SCHEME 2. De Novo Synthesis of Intermediate 9 from Thioe-
nol Ether 5
SCHEME 3. Synthesis of Thymidine Analogue 1 by Intramo-
lecular N-Glycosidation
intramolecular N-glycosidation. Differently from previous ap-
proaches, which did not allow the isolation of the bridged
nucleoside intermediate,26a in this case the presence of a 6-O-
thymine moiety in 12 enabled cyclonucleoside 13 to be obtained
smoothly in very good yield (89%).
cyclization) took place in a single step. Bicycle 4a was directly
subjected to acetylation (to avoid any further cyclization),24
obtaining diacetate 4b. Treatment of the latter with Ni/Ra
afforded methyl glycosides 9r/β (84% yield), the major
R-anomer being easily separated from the corresponding
β-component.
With the aim to find the best conditions for the stereo-
selective β-insertion of the nucleobase in the absence of
anchimeric assistance by C-2 and C-3 groups, N-glycosida-
tion reaction starting from methyl glycoside 9 was next
studied. To test the breadth of our methodology, insertion
of thymine (T) and N6-benzoyladenine (ABz) as nucleobases
was preliminarily considered.
As far as it concerns T nucleoside synthesis, the opportunity to
reach complete stereocontrol by C-10 insertion of the nucleobase
through an intramolecular base-transfer glycosidation25 was
initially explored (Scheme 3). Sugimura et al. already reported
an application of this methodology, using a D-pyranosyl thio-
glycoside intermediate.26 To examine a convenient alternative to
such conditions, we studied this reaction directly on methyl 2,3-
dideoxy-R-L-erythro-hexopyranoside 9r, using 2-chlorothymine
(11) as the nucleobase precursor (Scheme 3). Thus, methyl
glycoside 9r was first deacetylated (MeONa), then the resulting
diol was treated at 0 °C with NaH and 11 (in turn obtained from
2,4-dichloro-5-methylpyrimidine, 10). The resulting crude resi-
due was directly acetylated (Ac2O/Py), achieving 60-O-(thymin-
2-yl)-40-O-acetyl-20,30-dideoxy-R-L-erythro-hexopyranoside (12)
as the major product (40% yield over three steps). Then,
treatment of 12 with SnCl4 and BSA in CH3CN at 0 °C enabled
Unexpectedly, tricycle 13 displayed poor reactivity to the
common conditions required for 60-OH function release.27
Indeed, treatment of the latter with 1 M NaOH or 0.5 M
BaOH furnished target T nucleoside 1 (via intermediate 14)
only after prolonged reaction times (24 h at reflux) and in low
yield (Scheme 3). Acidic treatment (aq HCl or aq AcOH) was
also ineffective at rt (refluxing conditions only resulted in
hydrolysis at the C-10 position). This behavior strongly
disagrees with the reactivity of cyclonucleosides 15 and 16
(Figure 2), in which the primary hydroxyl group release has
been reported to occur by OH- attack at the C-2 position with
moderate to excellent yields even at low temperatures.26a,28 With
the aim to shed light on such contrasting experimental data, a
chemical reactivity analysis based on the FMO theory was
performed, comparing the energies of LUMO orbitals of
compounds 14, 15, and 16 calculated by the semiempirical
method PM3 (Hyperchem 8.0). As cyclonucleoside 14 bears a
different pyrimidine nucleobase (T vs U) compared to 15 and
16, the effect of the Me group at the C-5 position was also
calculated.29 Interestingly, lower LUMO energy values related
to compounds 16a,bwere found (R = H, -0.352 eV; R = CH3,
-0.312 eV) than those associated to tricycles 14a,b (R = H,
-0.297 eV; R = CH3, -0.259 eV). Far from them, the energy
values related to the LUMO of compounds 15a,b were the
lowest ones (R = H, -5.388 eV; R = CH3, -5.284 eV)
(Figure 2). On this matter, it must be considered that the total
(27) Mieczkowski, A.; Roy, V.; Agrofoglio, L. A. Chem. Rev. 2010, 110,
1828–1856.
(28) Jung, M. E.; Castro, C. J. Org. Chem. 1993, 58, 807–808.
(29) Moreover, even though cyclonucleosides 14, 15, and 16 differ in the
nature of the protective groups at the C-30/C-40 positions and in the
configuration of the stereogenic centers, it was calculated that these struc-
tural differences did not significantly influence reactivity of the resulting
nucleosides (data not shown).
(25) Sekine, M. In Glycoscience; Fraser-Reid, B. O., Tatsuta, K.; Thiem,
J., Eds.; Springer: Berlin, Germany, 2001; Chapter 3.6, pp 673-689.
(26) See for example: (a) Sugimura, H. Nucleosides, Nucleotides Nucleic
Acids 2000, 19, 629–635. (b) Watanabe, R.; Sugimura, H. Nucleic Acids
Symp. Ser. 2004, 48, 51–52.
6404 J. Org. Chem. Vol. 75, No. 19, 2010

D’Alonzo et al.
JOCArticle
SCHEME 4. Preparation of Dideoxy-L-Glycosides 17, 18, and 20
FIGURE 2. LUMO energy values of cyclonucleoside analogues
14-16.
charge distribution in tricycles 15a,b is such that C-2 bears a
higher positive charge;30 therefore, a double effect, ruled by both
Coulombic and frontier orbital interactions, favors the attack at
the C-2 position. On the other hand, Coulombic forces asso-
ciated to both cyclonucleosides 14 and 16 are relatively small,30
therefore frontier orbital interactions must be considered to only
explain their different reactivity. As for compounds 14a,b this
interaction is much less effective than that for compounds
16a,b, we can overall conclude that theoretical data are in
agreement with the experimental reactivity of the above cyclo-
nucleosides toward alkaline hydrolysis, which follows the order
15>16>14.
Because of the poor reactivity displayed by tricycle 13 to
common hydrolysis conditions, intramolecular glycosida-
tion as reported in Scheme 3 was not considered of practical
interest for the synthesis of thymidine analogue 1. Therefore,
we turned our attention to investigate other stereoselective
methods for direct β-N-glycosidation starting from methyl
glycoside 9r (Scheme 4). Such approaches resumed previous
work on the synthesis of D-hexose oligonucleotides31 the
above mentioned (“β-Homo-DNA”), which made use of the
D-enantiomer of 9. In that case, target D-nucleosides ob-
Our initial attempts at stereoselective N-glycosidation
were in fact not encouraging. Use of glycosyl iodide 18 under
alkaline conditions (NaH/CH3CN) was primarily examined,
with the aim to achieve a direct SN2 displacement by the
nucleobase. To our regret, the anomeric mixture of T and A
nucleosides 21 and 22 was formed at 0 °C without selectivity
(R:β = 1:1, 38-40% yield; entries 1 and 2). Use of THF in
place of CH3CN (entry 3) did not improve selectivity. In the
reaction with ABz, N-isomer nucleosides were also detected
(18-20% yield). A lower temperature (-20 °C) led even to a
slight preference for the undesired R-anomer (entry 4). In all
cases, a large amount of L-glycal31a 23 was found (40-47%
yield). Importantly, while T nucleoside anomers 21 were
easily separable by chromatography, the corresponding
anomeric ABz nucleoside mixture 22 was nearly inseparable
under the same conditions. Reactivity of glycosyl sulfoxide
20 was then studied. The anomeric mixture of 20 was
envisaged to undergo an SN2 attack by the nucleobase onto
the R-triflate intermediate 24 (Scheme 5), which is usually
reported to be formed at low temperatures on similar sub-
strates under common activating conditions (Tf2O34 or
TMSOTf35). In our hands, treatment of T nucleobase with
glycoside 20 and BSA/TfOTMS36 at -100 °C (entry 5)
proceeded with fairly high stereoselectivity (R:β = 1:6), even
though in low yield (38%). On the other hand, reactions at
higher temperatures (-40 or -20 °C) gave better yields
(75-81%), but without stereoselectivity (entries 6 and 7).
Such contrasting stereoselective outcome could be explained
assuming that, at lower temperatures, glycosyl triflate 24,
likely along with the corresponding transient contact ion
pair (CIP)37,34b 25, are the most active species, resulting in
the high observed selectivity after SN2 displacement by
the nucleobase; conversely, at higher temperatures, the
€
tained under common Vorbruggen conditions (HMDS/
TMSCl/SnCl4) were endowed with low stereoselectivity31a
at the anomeric center (B = T, R:β = 18:65; B = ABz, R:β =
43:57). In an effort to improve the stereoselective outcome of
this reaction using our L-hexose derivative, we studied the
reactivity of some other glycosyl donors, such as glycosides
17, 18, and 20, prepared from compound 9r (Scheme 4).
Particularly, methyl glycoside 9r was first converted into the
corresponding acetate 17 by mild acetolysis (PTSA/Ac2O/
AcOH), which occurred stereoselectively (R:β = 5:1). Then,
treatment of triacetate 17 with Et3SiH/I2 in CH2Cl2 at 0 °C
quantitatively afforded iodide 18 as a single anomer (1H NMR
analysis).32 Reaction of the crude 18 with Na(BH3)SPh, in turn
obtained by treatment of PhSSPh with NaBH4,33 furnished
thioglycosides 19 (R:β = 2:3; 65% yield from acetate 17).
Finally, thioether oxidation under common conditions (mCPBA)
smoothly gave sulfoxides 20 (86%) as a mixture of four
diastereomers.
(34) (a) Crich, D.; Sun, S. J. Am. Chem. Soc. 1997, 119, 11217–11223.
(b) Crich, D. Acc. Chem. Res. 2010, 43, 1144–1153.
(35) Sliedregt, L. A. J. M.; van der Marel, G. A.; van Boom, J. H.
Tetrahedron Lett. 1994, 35, 4015–4018. For an application of glycosyl
sulfoxides in N-glycosidation reactions, see: Chanteloup, L.; Beau, J.-M.
Tetrahedron Lett. 1992, 33, 5347.
(30) See the Supporting Information.
(31) (a) Bohringer, M.; Roth, H. J.; Hunziker, J.; Gobel, M.; Krishnan,
R.; Giger, A.; Schweizer, B.; Schreiber, J.; Leumann, C.; Eschenmoser, A.
Helv. Chim. Acta 1992, 75, 1416–1477. (b) Augustyns, K.; Vandendriessche,
F.; Van Aerschot, A.; Busson, R.; Urbanke, C.; Herdewijn, P. Nucleic Acids
Res. 1992, 20, 4711–4716.
(36) Early attempts with Tf2O gave similar results.
(37) Crich, D.; Chandrasekera, N. S. Angew. Chem., Int. Ed. 2004, 43,
5386–5389.
(38) The authors did not further investigate into the reaction mechanism
of the sulfoxide glycosidation at low temperatures, for example, analyzing
the formation of the glycosyl sulfenate intermediates (see for example:
Gildersleeve, J.; Pascal, R. A., Jr.; Kahne, D. J. Am. Chem. Soc. 1998,
120, 5961–5969), as the low yield achieved under the best stereoselective
conditions (entry 5) does not render the reaction of practical interest for our
synthetic purposes.
(32) Compare: (a) Lam, S. N.; Gervay-Hague, J. Org. Lett. 2003, 5, 4219–
ꢀ
4222. (b) Adinolfi, M.; Iadonisi, A.; Ravida, A.; Schiattarella, M. Synlett
2002, 269–270.
(33) Valerio, S.; Iadonisi, A.; Adinolfi, M.; Ravida, A. J. Org. Chem.
2007, 72, 6097–6106.
ꢀ
J. Org. Chem. Vol. 75, No. 19, 2010 6405

D’Alonzo et al.
JOCArticle
SCHEME 5. Working Hypothesis for the Sulfoxide N-Glyco-
sidation
TABLE 1. N-Glycosidation Reactiona
yield (%)
entry compd
B
T (°C) t (h) R/β ratio 21
22
23
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
18b
18b
18b,d
18b
20e
20f
20
T
0
0
0
0.5
0.5
0.5
0.5
0.5
0.5
0.5
1:1
1:1
1:1
2:1
1:6
1:1
1:1
1:4
4:6
1:4
1:3
1:6
1:11
1:6
1:9
1:1
1:1
1:20
1:11
38
40
45
44
47
10
<5
7
11
15
15
22
22
25
47
11
9
ABz
ABz
40c
38c
37
ABz -20
T
T
T
T
-100
-40
-20
40
38
81
75
75
9rg
9rg
17
40
ABz 40
40
40
40
40
40
40
3
45c
51c
T
rt
70
17
17
17
ABz rt
T
T
45
60
51
60
reaction proceeds with the substantial development of a
solvent-separated ion pair (SSIP)34b 26, with evident loss of
stereoselectivity.38
17
ABz 60
ABz 30
40
65
17h
20
T
rt
ABz rt
rtf60 20
ABz rtf60 48
0.5
0.5
79
65
With the aim to improve the anomeric β-selectivity,
N-glycosidation reactions were also carried out at higher
temperatures, providing more promising results. We rea-
soned that absence of electron-withdrawing substituents on
the sugar moiety could promote anomerization reactions,
which would be driven toward the thermodynamically more
stable β-nucleosides. As a result, an attempt carried out on
methyl glycoside 9r39 and T with SnCl4/BSA at 40 °C (entry 8)
led to an R:β = 1:4 mixture of nucleosides 21 (75% yield). On
the other hand, reaction of 9r with ABz (entry 9) afforded a
mixture of anomers 22 (45% yield) with lower selectivity (R:β =
4:6). These results are substantially in line with those found for
the corresponding D-nucleosides.31a Use of acetate 17 as glycosyl
donor was then considered. Treatment of the latter with T
nucleobase at rt (entry 10) led to nucleosides 21 with identical
selectivity (R:β = 1:4) as observed for 9r. The same reaction,
carried out with ABz (entry 11), led to nucleosides 22 with R:β =
1:3. To our delight, we observed that for both A and T nucleo-
sides anomeric selectivity increased time-dependently when the
reaction was warmed to 45-60 °C, which was strongly indica-
tive for a thermodynamically controlled anomerization reaction
(entries 12-14).40 With regards to T nucleosides 21, after 40 h at
60 °C (entry 13) the anomeric ratio was significantly high (R/β
= 1:11, 60% yield), along with a relatively small amount of 23
(25%). On the other hand, in the case of A nucleoside 22, after
40 h at 60 °C a R:β = 1:6 mixture of anomers was isolated in
40% yield (entry 14); only a few traces of N-regioisomer
nucleosides were detected. However, regardless of the good
selectivity achieved, the elimination reaction was also fostered,
to the extent that glycal 23 was isolated as the main product
(47%). In the search for more convenient conditions, we noticed
that, when previously silylated base was used, reaction times and
temperature could be drastically reduced, hampering formation
of 23. Hence, treatment of acetate 17 with bis-silylated ABz and
TMSOTf in CH3CN at 30 °C for 3 h (entry 15) afforded 22r/β
20
20
20
81
7
11
19
T
70
aConditions: BSA/TfOTMS/CH3CN (unless otherwise specified).
bNucleobase activation under alkaline conditions (NaH). cRegioiso-
meric ABz nucleosides also detected. dTHF used as solvent. eCH2Cl2 used
as solvent. fUse of CH3CN or CH2Cl2 led to similar results. gSnCl4 was
used in place of TfOTMS. hPreviously silylated nucleobase was used.
SCHEME 6. Anomerization of Nucleosides 21 and 22
in a satisfying yield (65%) and with a considerably high
anomeric ratio (R:β = 1:9); also, formation of 23 was remark-
ably reduced (11%). Eventually, use of sulfoxide 20 was in-
vestigated. While reactions at rt with T and ABz (entries 16 and
17) run very quickly (30 min) but without selectivity, the same
mixtures, heated to 60 °C (entries 18-19), led, after 20-48 h, to
the corresponding nucleosides 21 and 22 with the most prefer-
able selectivities (B = T, R:β= 1:20, 65% yield; B = ABz,R:β=
1:11, 70% yield). A relatively small amount of 23 was detected
(11-19%); moreover, in the case of ABz nucleosides 22, no
regioisomer nucleoside was found.
To further test the anomerization reaction occurring
under these conditions, an anomeric mixture of T and A
nucleosides 21 and 22 (R:β ≈ 1:1) was treated with BSA/
TfOTMS in CH3CN at 60 °C (Scheme 6). A substoichio-
metric amount of the nucleobase was also added, as it proved
to speed up the reaction. After 20-48 h, an improvement of
(39) Reaction with the pure 9β gave nearly the same results.
(40) This reaction augments previous results carried out on similar
substrates: Augustyns, K.; Rozenski, J.; Van Aerschot, A.; Busson, R.;
Claes, P.; Herdewijn, P. Tetrahedron 1994, 50, 1189–1198 and references
cited therein.
6406 J. Org. Chem. Vol. 75, No. 19, 2010

D’Alonzo et al.
JOCArticle
SCHEME 7. Deprotection of Nucleosides 21β and 22β
gram-scale preparation of our target nucleosides, which is a
difficult task to reach by common sugar-based approaches,
owing to the limited availability of the corresponding L-hexoses
as starting materials. Further experiments aimed to incorporate
our L-nucleoside building blocks 1 and 2 into natural oligonu-
cleotide sequences as well as to construct fully modified
L-oligonucleotide analogues (“β-L-HomoDNA”) for hybridiza-
tion studies with natural and unnatural complements are on-
going and will be published in due course.
Experimental Section
the R/β ratio (B = T, R:β>1:20; B = ABz, R:β = 1:11) was
observed (75-85% yield). It is worthy to note that results
obtained by N-glycosidation of sulfoxide 20 with T/ABz (Table
1) and those obtained by anomerization of T/ABz nucleosides
21-22r/β (Scheme 6) are exactly in line, as reactions with
sulfoxide 20 at rt immediately furnished an equimolar R/β
mixture of anomeric nucleosides (TLC analysis), which were
then subjected to in situ anomerization as temperature was
raised to 60 °C.
It is also noteworthy to underline that glycal 23 could be
recovered to provide starting material 17, which could be
resubjected to N-glycosidation reaction under anomeriza-
tion conditions to produce more β-nucleosides 21 and 22,
with the advantage of increasing the overall efficiency of the
process. As already reported for similar substrates,32a treat-
ment of 23 with HBr/AcOH/Ac2O at 0 °C gave back
triacetate 17 with moderate stereoselectivity (R:β = 4:1)
(Scheme 6).
Finally, removal of protective groups from β-nucleosides
21 and 22 provided our target nucleosides 1 and 2 (Scheme 7).
Particularly, treatment of 21β with saturated methanolic
NH3 solution easily gave 10-(thymin-1-yl)-20,30-dideoxy-
β-L-erythro-hexopyranoside (1) in 93% yield. Similarly, treat-
ment of 22β with 0.1 M MeONa led to pure 10-(N6-benzoyla-
denin-9-yl)-20,30-dideoxy-β-L-erythro-hexopyranoside (2) in
95% yield.
General Methods and Materials. All moisture-sensitive reac-
tions were performed under nitrogen atmosphere, using oven-
dried glassware. Solvents were dried over standard drying
agents and freshly distilled prior to use. Reactions were mon-
itored by TLC (precoated silica gel plate F254, Merck). Column
chromatography: Merck Kieselgel 60 (70-230 mesh); flash
chromatography: Merck Kieselgel 60 (230-400 mesh). Melting
points are uncorrected and were determined with a Gallenkamp
apparatus. Optical rotations were measured at 25 ( 2 °C in the
stated solvent. 1H and 13C NMR spectra were recorded on NMR
spectrometers operating at 200, 300, 400, or 500 MHz and 50, 75,
100, or 125 MHz, respectively. Combustion analyses were per-
formed with a Perkin-Elmer Series II 2400 CHNS analyzer.
(5R,7S,8S)-5-Methoxy-7-[(methylcarbonyloxy)methyl]-2,3,7,
8-tetrahydro-5H-[1,4]dithiino[2,3-c]pyran-8-yl Acetate (4b). To a
stirred 3:1 CH2Cl2/CH3OH solution (15 mL) containing the
PMB ether 8b (2.0 g, 4.6 mmol; prepared according to refs 22b
and 24) was added DDQ (1.5 g, 6.8 mmol) in one portion at
room temperature. The reaction mixture was stirred at the same
temperature for 24 h. Then, H2O was added and the mixture was
extracted with CH2Cl2. The organic layer was dried (Na2SO4)
and the solvent evaporated. The crude residue was directly
acetylated by addition of pyridine (20 mL) and acetic anhydride
(10 mL). The resulting solution was stirred for 4 h. Then solvents
were evaporated under diminished pressure. Chromatography
of the crude residue (hexane/EtOAc = 7:3) gave 4b (1.0 g, 70%
yield). The anomeric ratio as determined by 1H NMR was
1
R/β = 85:15. H NMR (500 MHz, CDCl3): δ 2.10 (s, 2.55H),
2.11 (s, 2.55H), 2.12 (s, 0.45H), 2.13 (s, 0.45H), 3.14-3.34 (m, 4H),
3.46 (s, 2.55H), 3.47 (s, 0.45H), 4.06-4.16 (m, 1H), 4.19-4.27 (m,
1H), 4.29-4.33 (m, 1H), 4.77 (s, 0.85H), 4.90 (s, 0.15H), 5.56 (bd,
J = 9.2 Hz, 1H). 13C NMR (major R-anomer; 50 MHz, CDCl3): δ
20.8 (2C), 27.3, 28.2, 55.6, 62.5, 65.2, 67.3, 97.9, 122.3, 126.3, 169.9,
170.1. Anal. Calcd for C13H18O6S2: C 46.69, H 5.43, S 19.18.
Found: C 46.85, H 5.45, S 19.26.
Conclusions
In summary, a stereoselective procedure for the synthesis
of 20,30-dideoxy-β-L-erythro-hexopyranosyl nucleosides has
been herein developed, starting from our 1,2-bis-thioenol
ether synthon 5, as homologating agent, and methyl R,β-
isopropylidene-L-glycerate (6), as chiral electrophile. Fast
and efficient access to enantiopure sugar scaffold 9 belonging
to the L-series has been achieved by a de novo approach
based on a domino reaction, which deeply reduces the total
number of synthetic steps. N-Glycosylation study between
glycosyl donors 9, 17, 18, and 20 and ABz/T (as models of
purine/pyrimidine nucleobases) revealed that, in both cases,
anomerization reaction toward nucleosides 21β and 22β
occurred with high stereoselectivity, despite the absence of
directing groups at C-2 and C-3 positions. The reaction is of
synthetic relevance especially for A nucleoside 22β, since its
anomeric mixture is not easily separable by common puri-
fication techniques. On the other hand, in the case of T
nucleoside synthesis, complete stereocontrol has also been
explored by intramolecular glycosidation, even though final
hydrolysis was complicated by inefficient HOMO-LUMO
interactions, as highlighted by semiempirical calculations.
Overall, the procedure herein reported enables an effective
Methyl 4,6-Di-O-acetyl-2,3-dideoxy-r-L-erythro-hexopyrano-
side (9R). A solution of 4b (5.0 g, 15 mmol) in THF (120 mL)
was added in one portion to a stirred suspension of Raney-Ni (W2)
(wet, 50 g) in the same solvent (120 mL) at 0 °C. The resulting
suspension was warmed to room temperature and further stirred
for 4 h, then the solid was filtered off and washed with EtOAc. The
filtrate was evaporated under reduced pressure to afford a
crude residue that after chromatography over silica gel (CH2Cl2)
gave pure 9r (2.6 g) as a colorless oil, besides a small amount of the
corresponding β-anomer 9β (0.5 g) (84% overall yield). Data for
9r: oily, [R]25 -132.7 (c 0.8, CHCl3). 1H NMR (500 MHz,
D
CDCl3): δ 1.78-1.86 (m, 3H), 1.95-2.01 (m, 1H), 2.05 (s, 3H),
2.09 (s, 3H), 3.38 (s, 3H), 3.90 (ddd, J = 2.0, 4.9, 9.8 Hz, 1H), 4.11
(dd, J = 2.0, 11.7 Hz, 1H), 4.26 (dd, J = 4.9, 11.7 Hz, 1H),
4.70-4.76 (m, 2H). 13C NMR (125 MHz, CDCl3): δ 20.8, 21.1,
23.9, 28.6, 54.6, 63.2, 67.7, 68.5, 97.5, 170.0, 170.9. Anal. Calcd for
C11H18O6: C 53.65, H 7.37. Found: C 53.83, H 7.64.
2-Chlorothymine (11). To a solution of 2,4-dichloro-5-methyl-
pyrimidine (10) (1.0 g, 6.17 mmol) in 5 mL of THF was added a
1 N aq NaOH solution (6 mL) at rt. The reaction mixture was
J. Org. Chem. Vol. 75, No. 19, 2010 6407

D’Alonzo et al.
JOCArticle
stirred for 16 h at the same temperature; then the reaction
mixture was slightly acidified with 1 N HCl and extracted with
CHCl3. The organic layer was dried (Na2SO4) and the solvent
was removed under reduced pressure. Chromatography of the
crude residue (CH2Cl2) gave the pure 11 (0.53 g, 60% yield). 1H
NMR (200 MHz, CDCl3): δ 2.38 (s, 3H), 8.40 (s, 1H). 13C NMR
(50 MHz, CDCl3): δ 15.7, 129.0, 158.0, 159.9, 162.4. Anal. Calcd
for C5H5ClN2O: C 41.54, H 3.49, N 19.38. Found: C 41.58, H
3.47, N 19.52.
anhydrous CH2Cl2 (6 mL) at 0 °C were added a 30% HBr/
AcOH solution (50 μL, 0.25 mmol of HBr) and acetic anhydride
(4 mL) dropwise. The resulting mixture was stirred for 16 h at
0 °C, then AcONa was added until neutrality. After 15 min, the
resulting suspension was filtered and the solid washed with
AcOEt. The filtrate was then washed with brine (2 ꢀ 150 mL).
The organic layer was dried (Na2SO4) and the solvent was
evaporated under reduced pressure. The crude acetyl glycoside
17 was directly engaged in the following step without further
purification. The anomeric ratio as determined by 1H NMR was
R/β = 4/1. Data for compound 17: 1H NMR (500 MHz,
CDCl3): δ 1.75-2.02 (m, 4H), 2.06 (s, 0.45H), 2.08 (s, 2.55H),
2.10 (s, 2.55H), 2.11 (s, 0.45H), 2.13 (s, 0.45H), 2.14 (s, 2.55H),
3.80-3.86 (m, 0.15H), 3.98-4.04 (m, 0.85H), 4.11 (dd, J = 2.2,
12.2 Hz, 0.85H), 4.14-4.22 (m, 0.30H), 4.28 (dd, J = 4.8, 12.2
Hz, 0.85H), 4.65-4.77 (m, 0.15H), 4.82 (dt, J = 4.9, 10.5 Hz,
Methyl 40-O-Acetyl-60-O-(thymin-2-yl)-20,30-dideoxy-r-L-ery-
thro-hexopyranoside (12). Diacetate 9r (0.5 g, 2.0 mmol) was
dissolved in a 0.1 M MeONa solution in MeOH (10 mL) and the
resulting mixture was stirred at room temperature overnight.
Then, a few drops of AcOH were added until neutrality. The
crude residue was dissolved in CH2Cl2 and filtered through a
short pad of silica gel; the resulting filtrate was then concen-
trated to dryness. The residue was dissolved in anhydrous DMF
(10 mL), the mixture was cooled to 0 °C, then NaH (0.15 g, 6.5
mmol) was added in one portion. After 30 min, pyrimidine 11
(0.35 g, 2.4 mmol) was added. The resulting brown suspension
was warmed to room temperature and further stirred for 16 h.
The mixture was diluted with AcOEt and washed with a
saturated NaHCO3 solution and brine. The organic layer was
dried (Na2SO4) and the solvent was evaporated under reduced
pressure. Chromatography of the crude residue over silica gel
(CH2Cl2/MeOH = 97/3) furnished the pure 12 (0.25 g, 40%
overall yield from compound 9r): [R]25D -42.3 (c 0.7, CHCl3).
1H NMR (500 MHz, CDCl3): δ 1.78-1.86 (m, 3H), 2.00-2.10
(m, 4H), 2.18 (s, 3H), 3.38 (s, 3H), 4.07 (ddd, J = 2.4, 5.4, 10.3
Hz, 1H), 4.47 (dd, J = 5.4, 11.7 Hz, 1H), 4.51 (dd, J = 2.4, 11.7
Hz, 1H), 4.73 (br s, 1H), 4.83 (dt, J = 4.4, 10.2 Hz, 1H), 8.15 (s,
1H). 13C NMR (125 MHz, CDCl3): δ 12.0, 21.0, 23.8, 28.5, 54.5,
66.2, 67.7, 68.3, 97.3, 116.9, 157.2, 157.7, 168.5, 169.8. Anal.
Calcd for C14H19N2O6: C 54.01, H 6.15, N 9.00. Found: C 54.19,
H 6.14, N 8.96.
Compound 13. To a stirring solution of methyl glycoside 12
(0.15 g, 0.48 mmol) in anhydrous CH3CN (2.5 mL) under
nitrogen stream was added BSA (0.50 mL, 2.0 mmol) at room
temperature. After 15 min, the mixture was cooled to 0 °C, and
SnCl4 (0.07 mL, 0.58 mmol) was added dropwise. The resulting
solution was stirred for 2 h at the same temperature, then the
mixture was diluted with AcOEt and washed with a saturated
NaHCO3 solution (2 ꢀ 50 mL) and brine (2 ꢀ 50 mL). The
collected organic layers were dried (Na2SO4), and the solvent was
evaporated under reduced pressure. Chromatography of the crude
residue (CH2Cl2/MeOH = 97/3) gave the pure 13 (0.12 g, 89%
yield). [R]25D -96.0 (c 0.5, CHCl3). 1H NMR (500 MHz, CDCl3): δ
1.83-2.16 (m, 5H), 2.18 (s, 3H), 2.52 (dt, J = 4.4, 17.1 Hz, 1H),
4.18-4.25 (m, 1H), 4.58 (dd, J = 2.9, 12.0 Hz, 1H), 4.61 (dd, J =
4.9, 12.0 Hz, 1H), 4.71-4.78 (m, 1H), 5.15 (q, J = 7.3 Hz, 1H), 6.38
(d, J = 6.3 Hz, 1H), 8.15 (s, 1H). 13C NMR (125 MHz, CDCl3): δ
12.1, 21.1, 25.6, 29.7, 65.3, 65.6, 73.8, 97.7, 117.1, 142.7, 158.0,
168.4, 169.9. Anal. Calcd for C13H16N2O5: C 55.71, H 5.75, N 9.99.
Found: C 55.53, H 5.77, N 10.02.
1H), 5.80 (dd, J = 2.4, 8.4 Hz, 0.15H), 6.15 (br s, 0.85H). 13
C
NMR (major R-anomer; 100 MHz, CDCl3): δ 20.8, 21.0, 21.1,
23.6, 27.6, 62.7, 67.1, 70.8, 90.8, 169.3, 169.8, 170.8. Anal. Calcd
for C12H18O7: C 52.55, H 6.62. Found: C 52.75, H 6.65.
4,6-Di-O-acetyl-2,3-dideoxy-r-L-erythro-hexopyranosyl Iodide (18).
To a stirring solution of triacetate 17 (0.5 g, 1.8 mmol) in
anhydrous CH2Cl2 (5 mL) at 0 °C was added a solution of iodine
(0.6 g, 2.2 mmol) in anhydrous CH2Cl2 (5 mL) in one portion.
After a few minutes, Et3SiH (0.3 mL, 2.2 mmol) was added at the
same temperature. The resulting mixture was stirred for 20 min,
then solvent was removed under reduced pressure. Given its low
stability to humidity, the crude glycosyl iodide 18 was directly
engaged in the following step (i.e., N-glycosidation or preparation
of thioglycoside 19) without further purification. Data for com-
1
pound 18: H NMR (300 MHz, CDCl3): δ 1.86-1.92 (m, 1H),
2.02-2.12 (m, 8H), 2.28-2.34 (m, 1H), 3.83 (ddd, J = 2.4, 4.9,
10.1 Hz, 1H), 4.13 (dd, J = 1.9, 12.3 Hz, 1H), 4.37 (dd, J = 4.6,
12.3 Hz, 1H), 4.83 (dt, J = 4.9, 10.1 Hz, 1H), 7.14 (d, J = 3.0 Hz,
1H). 13C NMR (100 MHz, CDCl3): δ 20.7, 21.0, 26.0, 37.1, 61.8,
66.3, 68.2, 75.1, 169.9, 170.7. ESI-MS calcd for [C10H15IO5Na]þ
364.99, found 364.83. Anal. Calcd for C10H15IO5: C 35.11, H 4.42.
Found: C 35.18, H 4.44.
Thiophenyl 4,6-Di-O-acetyl-2,3-dideoxy-L-erythro-hexopyra-
noside (19). To a stirring solution of phenyl disulfide (0.05 g,
0.2 mmol) in anhydrous CH3CN (3 mL) was added NaBH4
(0.02 g, 0.4 mmol). The mixture was stirred until evolution of
hydrogen ceased and then was shortly heated at 50 °C to ensure
completion of the reaction. The mixture was then added to a
solution of the crude glycosyl iodide 18 (0.1 g, 0.3 mmol) in
CH3CN (2 mL). The resulting solution was stirred at 0 °C for
10 min, then acetic acid was added until neutrality. The mixture was
diluted with CH2Cl2 and washed with water; then the organic phase
was dried and evaporated under vacuum. Chromatography of the
crude residue (hexane/EtOAc = 85:15) gave sulfide 19 (0.06 g,
65% yield calculated from triacetate 17) as a mixture of anomers
(R:β = 2:3). Data for compound 19: 1H NMR (400 MHz, CDCl3):
δ 1.77-1.89 (m, 2H), 1.99 (s, 1.8H), 2.01 (s, 1.2H), 2.02-2.18 (m,
4H), 2.19-2.35 (m, 1H), 3.65 (ddd, J = 4.3, 4.6, 9.7 Hz, 0.6H), 4.08
(dd, J = 2.1, 12.0 Hz, 0.4H), 4.18 (d, J = 4.3 Hz, 1.2H), 4.27 (dd,
J = 5.8, 12.0 Hz, 0.4H), 4.43-4.48 (m, 0.4H), 4.68 (ddd, J = 4.8,
9.7, 10.4 Hz, 0.6H), 4.69-4.77 (m, 0.4H), 4.78 (dd, J = 2.0, 11.5
Hz, 0.6H), 5.58 (d, J = 5.2 Hz, 0.4H), 7.20-7.37 (m, 3H),
7.43-7.57 (m, 2H). 13C NMR (100 MHz, CDCl3): δ 20.6 (2C),
20.8 (2C), 25.3, 29.4 (2C), 30.3, 62.9, 63.3, 67.3, 67.8, 69.0, 77.8,
84.0, 84.4, 127.0, 127.2, 128.6 (2C), 128.7 (2C), 131.2 (4C), 133.9,
134.3, 169.7 (2C), 170.6 (2C). ESI-MS calcd for [C16H20O5SNa]þ
347.09, found 346.90. Anal. Calcd for C16H20O5S: C 59.24, H 6.21,
S 9.88. Found: C 59.46, H 6.23, S 9.92.
1,4,6-Tri-O-acetyl-2,3-dideoxy-L-erythro-hexopyranoside (17).
Method A: To a stirring solution of methyl glycoside 9r (5.0 g,
20.5 mmol) in Ac2O/AcOH (2:1 v/v, 150 mL) at 0 °C was added
p-toluenesulfonic acid (7.5 g, 41.0 mmol) in one portion. The
resulting mixture was stirred for 4 h, keeping the temperature at
0 °C. Then the solution was diluted with AcOEt, warmed to
room temperature, and washed with NaHCO3 solution and
brine. The organic layer was dried (Na2SO4) and the solvent
evaporated under reduced pressure. The crude acetyl glycoside
17 was directly engaged in the following step without further
purification. The anomeric ratio as determined by 1H NMR was
R/β = 5/1. Method B: To a stirring solution of olefin 23 (0.20 g,
0.9 mmol; obtained in significant amount as byproduct during
preparation of nucleosides 21β and/or 22β, see farther on) in
4,6-Di-O-acetyl-2,3-dideoxy-1-(phenylsulfinyl)-L-erythro-hexo-
pyranose (20). To a stirring solution of thioglycoside 19 (0.1 g, 0.3
mmol) inanhydrousCH2Cl2 (4 mL) at-78 °Cwas slowly addeda
solution of mCPBA (0.05 g, 0.3 mmol) at the same temperature.
6408 J. Org. Chem. Vol. 75, No. 19, 2010

D’Alonzo et al.
JOCArticle
After 10 min, the resulting mixture was further diluted with CH2Cl2
(50 mL), warmed to rt, and washed with sat NaHCO3 and brine.
The organic phase was dried (Na2SO4) and evaporated under
reduced pressure. Flash chromatography of the crude residue
(hexane/EtOAc = 85:15) gave sulfoxide 20 (0.09 g, 86% yield) as
a mixture of four diastereomers. Data for compound 20: 1H NMR
(major two diastereomers; 500 MHz, C6D6): δ 0.86-0.98 (m,
0.5H), 1.31-1.42 (m, 0.5H), 1.52 (s, 1.5H), 1.61 (s, 1.5H), 1.67
(s, 1.5H), 1.68 (s, 1.5H), 1.72-1.97 (m, 2H), 2.43-2.50 (m, 0.5H),
3.01-3.04 (m, 0.5H), 3.78 (dd, J = 3.5, 10.4 Hz, 0.5H), 4.02 (dd,
J = 2.2, 11.8 Hz, 0.5H), 4.05-4.21 (m, 2.5H), 4.66-4.78 (m,
1.5H), 7.01-7.12 (m, 3H), 7.58-7.62 (m, 2H). 13C NMR (major
two diastereomers; 125 MHz, C6D6): δ 20.3, 20.4, 20.5, 21.4, 23.3,
24.9, 27.9, 30.1, 62.8, 62.9, 67.0, 67.1, 74.5, 78.7, 94.2, 94.4, 124.8
(2C), 125.0 (2C), 128.7, 128.9 (2C), 129.0 (2C), 130.9, 142.9, 143.7,
169.0, 169.2, 169.9, 170.0. ESI-MS calcd for [C16H20O6SNa]þ
363.09, found 362.85. Anal. Calcd for C16H20O6S: C 56.46, H
5.92, S 9.42. Found: C 56.30, H 5.94, S 9.46.
stirred for 1 h; after a few minutes, the solution became homo-
geneous. The solvent was removed under reduced pressure, and the
crude residue was coevaporated three times with anhydrous
CH3CN (40 mL). Then, a solution of acetyl glycoside 17 (1.5 g,
5.5 mmol) in anhydrous CH3CN (50 mL) was added in one portion.
Hence, TfOTMS (1.2 mL, 6.6 mmol) was added dropwise in
20 min. The resulting mixture was further stirred for 3 h at 30 °C.
Then the solution was diluted with AcOEt and washed with a
saturated NaHCO3 solution and brine. The organic layer was dried
(Na2SO4) and the solvent was evaporated under reduced pressure.
Flash chromatography of the crude residue (CH2Cl2/MeOH =
97/3) gave the pure 22β (1.5 g, 65% yield); a minor amount of
22r (0.2 g) was also recovered. Moreover, a small quantity of
L-glycal 23 (11%) was isolated. Method B: To a stirring suspension
of sulfoxide 20 (0.2 g, 0.5 mmol) and N6-benzoyladenine (0.2 g,
0.6 mmol) in anhydrous CH3CN (5 mL) under nitrogen atmo-
sphere was added BSA (0.5 mL, 2.0 mmol). The resulting mixture
was warmed to 60 °C and stirred for 30 min; after a few minutes the
solution became homogeneous. Then the solution was cooled to
room temperature, and TfOTMS (0.1 mL, 0.6 mmol) was added
dropwise in 10 min. The resulting mixture was further stirred until
disappearance of sulfoxide 20 (<30 min). Then the mixture was
warmed to 60 °C and stirred for 48 h at the same temperature. The
solution was then diluted with AcOEt and washed with a saturated
NaHCO3 solution and brine. The organic layer was dried (Na2SO4)
and the solvent was evaporated under reduced pressure. Flash
chromatography of the crude residue (CH2Cl2/MeOH = 97/3)
gave the pure 22β (0.2 g, 70% yield); a minor amount of ano-
meric 22r (0.02 g) was also recovered. In addition, L-glycal 23
(19%) was isolated. Method C (anomerization of dideoxyadeno-
sine analogue 22): To a stirring suspension of nucleoside 22 (1.0 g,
2.2 mmol, R/β = 1/1) and N6-benzoyladenine (0.3 g, 1.1 mmol) in
anhydrous CH3CN (40 mL) and under nitrogen atmosphere was
added BSA (2.0 mL, 8.0 mmol). The resulting suspension was
stirred for 30 min at 60 °C; after a few minutes the solution became
homogeneous. Hence, the mixture was cooled to rt and TfOTMS
(0.5 mL, 2.6 mmol) was added dropwise in 10 min. The reaction
mixture was warmed again to 60 °C and further stirred for 48 h at
the same temperature. Then the solution was diluted with AcOEt
and washed with a saturated NaHCO3 solution and brine. The
organic layer was dried (Na2SO4) and the solvent was evaporated
under reduced pressure. Flash chromatography of the crude residue
(CH2Cl2/MeOH = 97/3) gave the pure 22β (1.5 g, 75% yield),
besides a minor amount of the corresponding R-anomer (0.1 g).
In addition, L-glycal 23 (16%) was isolated. Data for compound
22β: white foam; [R]25D -23.6 (c 0.8, CHCl3). 1H NMR (500 MHz,
CDCl3):δ1.82 (dq, J = 4.8, 13.0 Hz, 1H), 2.06 (s, 3H), 2.10 (s, 3H),
2.26 (dq, J = 3.9, 13.2 Hz, 1H), 2.28-2.36 (m, 1H), 2.43-2.50 (m,
1H), 3.94-3.99 (m, 1H), 4.20 (dd, J = 1.5, 12.2 Hz, 1H), 4.28 (dd,
J = 5.4, 12.2 Hz, 1H), 4.88 (dt, J = 4.8, 10.8 Hz, 1H), 5.94 (dd, J =
2.2, 11.0 Hz, 1H), 7.51 (t, J = 7.8 Hz, 2H), 7.60 (t, J = 7.3 Hz, 2H),
8.03 (d, J = 7.3 Hz, 1H), 8.22 (s, 1H), 8.79 (s, 1H), 9.26 (br s, 1H).
13C NMR (75 MHz, CDCl3): δ 20.6, 20.8, 28.1, 30.4, 62.7, 66.7,
78.0, 81.6, 123.2, 127.9 (2C), 128.8 (2C), 132.7, 133.8, 140.4, 149.7,
151.4, 152.7, 164.5, 169.7, 170.5. Anal. Calcd for C22H23N5O6: C
58.27, H 5.11, N 15.44. Found: C 58.03, H 5.09, N 15.50.
(Thymin-1-yl) 40,60-Di-O-acetyl-20,30-dideoxy-β-L-erythro-hexo-
pyranoside (21β). Method A: To a stirring suspension of acetyl
glycoside 17 (3.0 g, 11.0 mmol, R/β = 5/1) and thymine (1.7 g, 13.2
mmol) in anhydrous CH3CN (150 mL) and under nitrogen atmo-
sphere was added N,N-bis-trimethylsilylacetamide (BSA, 9.8 mL,
39.6 mmol) and the resulting solution was stirred for 30 min at
60 °C; after a few minutes the solution became homogeneous.
Hence, TfOTMS (2.4 mL, 13.2 mmol) was added dropwise in
10 min. The reaction mixture was further stirred for 40 h at 60 °C.
Then the solution was diluted with AcOEt and washed with a
saturated NaHCO3 solution and brine. The organic layer was dried
(Na2SO4) and the solvent evaporated under reduced pressure.
Chromatography of the crude residue over silica gel (CH2Cl2/
MeOH = 95/5) gave the pure 21β (2.1 g, 60% yield); a minor
amount of anomeric 21r (0.2 g) was also recovered. In addi-
tion, L-glycal 23 (25% yield) was isolated. Method B
(anomerization of dideoxythymidine analogue 21): To a stirring
solution of nucleoside 21 (0.1 g, 0.3 mmol, R/β = 1/1) and thymine
(0.1 mmol) in anhydrous CH3CN (4 mL) and under nitrogen
atmosphere was added BSA (0.4 mL, 1.0 mmol). The resulting
solution was stirred for 30 min at rt. Hence, TfOTMS (0.1 mL, 0.4
mmol) was added in a few minutes. The reaction mixture was
warmed to 60 °C and further stirred for 20 h at the same
temperature. Then the solution was diluted with AcOEt and
washed with a saturated NaHCO3 solution and brine. The organic
layer was dried (Na2SO4) and the solvent evaporated under
reduced pressure. Flash chromatography of the crude residue
(CH2Cl2/MeOH = 97/3) gave the pure 21β (0.7 g, 85% yield).
In addition, L-glycal 23 (11%) was isolated. Data for 21β: white
foam, [R]25D -27.9 (c 1.0, CHCl3). 1H NMR (300 MHz, CDCl3):δ
1.68-1.89 (m, 2H), 1,97 (s, 3H), 1.70-2.06 (m, 1H), 2.09 (s, 3H),
2.10 (s, 3H), 2.33-2.42 (m, 1H), 3.85 (ddd, J = 2.3, 6.4, 10.3 Hz,
1H), 4.17 (dd, J = 2.2, 12.2 Hz, 1H), 4.25 (dd, J = 6.2, 12.2 Hz,
1H), 4.75 (dt, J = 4.6, 10.3 Hz, 1H), 5.78 (dd, J = 2.5, 10.5 Hz,
1H), 7.19 (s, 1H), 8.82 (br s, 1H). 13C NMR (75 MHz, CDCl3): δ
12.6, 20.8, 21.0, 27.9, 29.3, 62.9, 66.5, 77.9, 81.3, 111.5, 134.9, 150.0,
163.3, 170.0, 170.8. Anal. Calcd for C15H20N2O7: C 52.94, H 5.92,
1
N 8.23. Found: C 53.12, H 5.90, N, 8.20. Data for 23: oily. H
NMR (300 MHz, CDCl3): δ 2.01-2.16 (m, 7H), 2,45 (dddd, J =
2.1, 4.8, 6.3, 17.2 Hz, 1H), 4.05 (ddd, J = 3.3, 5.7, 8.7 Hz, 1H), 4.22
(dd, J = 3.3, 12.0 Hz, 1H), 4.31 (dd, J = 5.7, 12.0 Hz, 1H),
4.66-4.76 (m, 1H), 5.02 (ddd, J= 6.0, 7.8, 13.8 Hz, 1H), 6.35 (dt, J
= 2.1, 6.0 Hz, 1H). 13C NMR (75 MHz, CDCl3):δ 20.8, 21.2, 25.8,
62.3, 65.7, 74.0, 97.7, 142.6, 169.9, 170.2. Anal. Calcd for
C10H14O5: C 56.07, H 6.59. Found: C 56.23, H 6.56.
(Thymin-1-yl) 20,30-dideoxy-β-L-erythro-hexopyranoside (1).
Deacetylation of nucleoside 21β (2.0 g, 5.8 mmol) was accom-
plished by treatment with a saturated metanolic NH3 solution
(50 mL) at 0 °C. After stirring for 24 h at room temperature, the
solvent was removed. Chromatography of the crude residue
(CH2Cl2/MeOH = 8:2) gave the pure 1 (1.4 g, 93% yield) as a
white solid: mp 199.0-199.8 °C (acetone); [R]25D -22.0 (c 1.0,
H2O). 1H NMR (500 MHz, CD3OD): δ 1.66 (dq, J = 4.4, 12.7
Hz, 1H), 1.83 (dq, J = 3.9, 12.7 Hz, 1H), 1.87-1.93 (m, 4H),
2.15-2.22 (m, 1H), 3.38-3.44 (m, 1H), 3.56 (dt, J = 4.9, 10.0
Hz, 1H), 3.75 (dd, J = 5.4, 12.0 Hz, 1H), 3.85 (dd, J = 1.5, 12.0
Hz, 1H), 5.66 (dd, J = 2.2, 10.7 Hz, 1H), 7.62 (s, 1H). 13C NMR
(N6-Benzoyladenin-9-yl) 40,60-Di-O-acetyl-20,30-dideoxy-β-L-
erythro-hexopyranoside (22β). Method A: To a stirring suspen-
sion of N6-benzoyladenine (1.6 g, 6.6 mmol) in anhydrous
CH3CN (50 mL) under nitrogen stream was added BSA (5.0 mL,
20 mmol). The resulting suspension was warmed to reflux and
J. Org. Chem. Vol. 75, No. 19, 2010 6409

D’Alonzo et al.
JOCArticle
(75 MHz, CD3OD): δ 10.9, 29.0, 31.0, 61.3, 64.3, 81.6, 83.1,
110.0, 136.8, 150.7, 164.9. Anal. Calcd for C11H16N2O5: C 51.56,
H 6.29, N 10.93. Found: C 51.40, H 6.27, N 10.97.
61.2, 64.5, 81.1, 83.7, 125.8, 127.7, 128.8, 132.8, 133.6, 143.1, 150.7,
152.0, 152.2, 166.3. Anal. Calcd for C18H19N5O4: C 58.53, H 5.18,
N 18.96. Found: C 58.73, H 5.20, N 19.02.
(N6-Benzoyladenin-9-yl) 20,30-Dideoxy-β-L-erythro-hexopyra-
noside (2). Nucleoside 22β (2.0 g, 4.4 mmol) was treated with a 0.1
M NaOMe solution in MeOH (60 mL) and the mixture was stirred
for 2 h at 0 °C. Then, a few drops of acetic acid were added until
neutrality. The solvent was evaporated under reduced pressure.
Chromatography of the crude residue (CH2Cl2/MeOH = 8:2)
Acknowledgment. D.D. is grateful to Prof. B. Doboszewski
and Dr. A. Iadonisi for helpful discussions on the N-glycosida-
tion reaction. Moreover, the authors wish to thank a reviewer
who suggested the use of sulfoxide 20 for nucleoside synthesis.
1H and 13C NMR experiments were performed at the “Centro
Interdipartimentale di Metodologie Chimico-Fisiche” (CIMCF),
University of Napoli Federico II.
afforded the pure 2 (1.6 g, 95% yield) as a white solid: mp 202.1-
D
1
202.9 °C (MeOH); [R]25 -21.2 (c 1.0, DMSO). H NMR (500
MHz, DMSO-d6): δ 1.65-1.75 (m, 1H), 2.08-2.20 (m, 2H),
2.42-2.52 (m, 1H), 3.40-3.51 (m, 3H), 3.71 (appdd, J = 6.3,
11.2 Hz, 1H), 4.58 (t, J = 5.9 Hz, 1H), 4.99 (d, J = 4.8 Hz, 1H),
5.88 (dd, J = 1.5, 10.8 Hz, 1H), 7.62 (t, J = 7.8 Hz, 2H), 7.72 (t,
J= 7.8 Hz, 1H), 8.07 (d, J= 7.8 Hz, 1H), 8.72 (s, 1H), 8.78 (s, 1H),
11.20 (br s, 1H). 13C NMR (100 MHz, DMSO-d6): δ 29.4, 31.9,
1
Supporting Information Available: Copies of H and 13C
NMR spectra for all new compounds. This material is available
free of charge via the Internet at http://pubs.acs.org.
6410 J. Org. Chem. Vol. 75, No. 19, 2010