Stereoselective synthesis of 2-O-MEM-2,3-unsaturated-β-O-glycosides and elaboration to useful synthetic tools

Article abstract of DOI:10.1016/j.tetlet.2010.07.061

The glycosylation of alcohols by the new 2-O-MEM-substituted D-galactal-derived allyl epoxide affords the corresponding alkyl 2-O-MEM-3-deoxy-β-D-threo-hex-2-enopyranosides through a completely 1,4-regio- and a highly to completely substrate-dependent stereoselective glycosylation processes. The glycosides obtained can be regioselectively transformed into corresponding 3-deoxy-β-O-glycosides, 3-deoxy-β-D- threo-hexopyranosid-2-uloses, and 3,4-dideoxy-β-D-glycero-hex-3- enopyranosid-2-uloses, which are useful synthetic tools for further transformations.

Full text of DOI:10.1016/j.tetlet.2010.07.061

Tetrahedron Letters 51 (2010) 4937–4941
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Stereoselective synthesis of 2-O-MEM-2,3-unsaturated-b-O-glycosides
and elaboration to useful synthetic tools
*
*
Valeria Di Bussolo , Annalisa Fiasella, Ileana Frau, Lucilla Favero, Paolo Crotti
Dipartimento di Scienze Farmaceutiche, Sede Chimica Bioorganica e Biofarmacia, Università di Pisa, Via Bonanno 33, I-56126 Pisa, Italy
a r t i c l e i n f o
a b s t r a c t
Article history:
The glycosylation of alcohols by the new 2-O-MEM-substituted
D-galactal-derived allyl epoxide affords
Received 10 June 2010
Revised 8 July 2010
Accepted 9 July 2010
Available online 15 July 2010
the corresponding alkyl 2-O-MEM-3-deoxy-b-D-threo-hex-2-enopyranosides through a completely
1,4-regio- and a highly to completely substrate-dependent stereoselective glycosylation processes. The
glycosides obtained can be regioselectively transformed into corresponding 3-deoxy-b-O-glycosides,
3-deoxy-b-
D-threo-hexopyranosid-2-uloses, and 3,4-dideoxy-b-D-glycero-hex-3-enopyranosid-2-uloses,
which are useful synthetic tools for further transformations.
Keywords:
Ó 2010 Elsevier Ltd. All rights reserved.
Glycals
Allyl epoxides
Glycosylation
3-Deoxy sugars
2-Keto sugars
Alkyl 2-O-protected-3-deoxy-hex-2-enopyranosides (1) are use-
We recently found a new O-glycosylation protocol consisting of
the addition of O-nucleophiles (alcohols and partially protected
ful synthetic intermediates,1 and, starting from 2-acetoxy-3,4,6-tri-
O-acetyl-
D
-glucal as the chiral precursor, have been prepared and
monosaccharides) to
epoxide (4, X = H): corresponding alkyl 2,3-dideoxy-
D
-allal- (3, X = H) and
D
-galactal-derived allyl
- (5, X = H)
used in the synthesis of natural products such as tetrodotoxin and
okadaic acid, valuable antitumor agents,1a,b and Melillo’s lactone.1c
Moreover, compounds 1 can reasonably be transformed into corre-
sponding alkyl 3-deoxy-hexopyranosid-2-uloses (2), which are
useful for further functionalizations (Scheme 1).
a
and -b-hex-2-enopyranosides (6, X = H) were obtained, respec-
tively, in a completely 1,4-regio- and stereoselective fashion
(Scheme 2).2 In this connection, the possibility of an interesting
application of this original protocol to the regio- and stereoselec-
tive synthesis of the title compounds was envisaged: if the corre-
sponding 2-O-protected-epoxides 3 and 4 (X = OR2) showed a
behavior similar to 3 and 4 (X = H), in the addition reaction of alco-
hols, a stereoselective synthesis of alkyl 2-O-protected-3-deoxy-
O
OR1
O
O
OR1
OR2
R3O
HO
R3O
R1, R3 = alkyl
R2 = alkyl or acyl
hex-2-enopyranosides
(Scheme 2).
5 and 6
(X = OR2) could be obtained
HO
2
1
Considering that among the different possibilities attempted,3 the
2-O-methoxyethoxymethyl (2-O-MEM) functionality turned out to
be the best protective group for our purposes, the experimental
Scheme 1. Alkyl 2-O-protected-3-deoxy-hex-2-enopyranosides (1) and corre-
sponding alkyl 3-deoxy-hexopyranosid-2-uloses (2).
OR1
OR1
O
O
O
O
R OH
R1OH
X
1
BnO
BnO
HO
BnO
BnO
HO
β
α
X
X
X
O
O
β
α
5
4
3
6
X, = H, OR2; R1 = Me, Et, i-Pr, t-Bu; R2 = Me, Bn, Ac
Scheme 2. Glycosylation of alcohols by allyl epoxides 3 and 4.
* Corresponding authors. +39 0502219690; fax: +39 0502219660 (P.C.).
E-mail address: paolo.crotti@farm.unipi.it (P. Crotti).
0040-4039/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2010.07.061

4938
V. Di Bussolo et al. / Tetrahedron Letters 51 (2010) 4937–4941
lead to what we needed. Actually, the application of the direct oxi-
dative glycosylation to p-thiocresol and PhSeH by the use of our
glycal 9 afforded, through the intermediate formation of glucal
O
O
O
BnO
AcO
BnO
BnO
MsO
OMEM
OMEM
O
OH
OTBS
a
-epoxide 12, p-tolyl trans-2-hydroxy-b-thioglycoside 13 and phe-
nyl trans-2-hydroxy-b-selenoglycoside 14,4 respectively, in a com-
pletely stereoselective fashion, as desired (Scheme 4). It is
interesting to note that the complete anti-b-stereoselectivity found
in the glycosylation of PhSeH, under these conditions,4 is opposite
7
8
9
Scheme 3. Retrosynthetic analysis for epoxide 7.
to the complete syn-
with the same nucleophile, in the glycosylation of tri-O-benzyl-
-glucal under his original reaction conditions, which similarly
proceed through a corresponding glucal
-epoxide.7
The decidedly simpler protocol related to the synthesis of the
O-MEM selenoxide 16 and subsequent syn-elimination process
(one-pot reaction, 0 °C to room temperature, 93% yield, Scheme
5) compared with the alternative protocol by means of the
corresponding sulfoxide [two-step reaction, including a scantily
reliable and low yielding thermal elimination (90 °C), 23% yield]
(see Supplementary data) determined the preference of b-phe-
nylselenoglycoside 14 for the pursuit of the synthetic route to
epoxide 7.
a-stereoselectivity found by Danishefsky,
feasibility of the above described O-glycosylation process was
checked by means of the 2-O-MEM-substituted- -galactal-
derived allyl epoxide 7, taken as an appropriate model (Scheme 3).
A retrosynthetic analysis indicated that epoxide 7 could be
obtained by cyclization under basic conditions of the trans-hydro-
xy mesylate 8, and could be prepared on its own from the appro-
D
D
a
priately O-protected-D-glucal 9 (Scheme 3). Our intention was to
introduce the 2-O-MEM-1,2-unsaturated moiety present in 7 and
8 by a glycosylation reaction of 9 with an appropriate nucleophile,
followed by elimination.
Fully protected
residual OH-functionality of the known mono-O-TBS derivative
11 of 6-O-(benzyl)-
-glucal (10)2b (Scheme 4).
D
-glucal 94 was prepared by acetylation of the
D
By refluxing with MEM-NEt3+ClÀ in THF, b-phenylselenoglyco-
side 14 was transformed into the corresponding 2-O-MEM deriva-
tive 15.8 Following a previously described protocol [Ti(O-i-Pr)4/
TBHP in CH2Cl2 in the presence of DIPEA, 0 °C to room temperature],9
2-O-MEM-b-phenylselenoglycoside 15 underwent a Sharpless-type
oxidation with the intermediate formation of selenoxide 16 which
was followed by spontaneous selenoxide b-syn-elimination with
the vicinal H(2) giving the 2-O-MEM-substituted glucal 17. Saponi-
fication (MeONa/MeOH) of glucal 17 afforded secondary alcohol 18
which was transformed (MsCl/Py) into the corresponding mesylate
19. Deprotection of 19 by the TBAF/THF protocol at 0 °C led to trans
hydroxy mesylate 8, the stable ultimate precursor of allyl epoxide 7
(Schemes 5 and 6).
At this point, the synthetic protocol required a glycosylation
process which is able to introduce contemporaneously, in an
anti-stereoselective fashion, a hydroxy group on C(2) and a glyco-
syl acceptor on C(1), based on an element, such as S or Se, capable
to subsequently give, through the corresponding oxide (sulfoxide
or selenoxide), a b-syn-elimination.5
Gin’s direct oxidative glycosylation of alcohols by glycals is
known as a very effective protocol for the regio- and stereoselec-
tive synthesis, through the formation of an intermediate glycal
a
-epoxide, of corresponding trans-2-hydroxy-b-O-glycosides.6 We
thought that the replacement of the alcohol-nucleophile, as de-
scribed in the original procedure, by an S- or Se-nucleophile might
O
O
TBSCl
DMF
imidazole
84% (ref. 2)
BnO
BnO
HO
HO
OH
OTBS
Ac2O
10
11
82% Py
(ref. 4)
1) Ph2SO,
O
Y
O
O
O
BnO
AcO
Tf2O, TTBP,
(MeOH), Et3N
BnO
AcO
BnO
OH
AcO
2) p-Me-C6H4SH
or PhSeH
OTBS
OTBS
OTBS
13, Y = p-Me-C6H4S- (59%)
14, Y = PhSe- (78%, ref. 4)
9
12
Scheme 4. Synthesis of b-thioglycoside 13 and b-selenoglycoside 14.
O
O
Se
Ti(O-i-Pr)4
TBHP
O
SePh
OH
O
SePh
BnO
Ph
BnO
AcO
BnO
MEM-NEt3+Cl-
DIPEA
MeOH
CH2Cl2
MeCN
reflux
73%
AcO
OMEM
AcO
OMEM
OTBS
OTBS
OTBS
15
14
16
93%
O
O
O
BnO
MsO
BnO
BnO
TBAF
THF
74%
MeONa
OMEM
MeOH
89%
RO
OMEM
AcO
OMEM
OH
OTBS
OTBS
18, R = H
19, R = Ms
MsCl/Py
76%
8
17
Scheme 5. Synthesis of trans-hydroxy mesylate 8.

V. Di Bussolo et al. / Tetrahedron Letters 51 (2010) 4937–4941
4939
Scheme 6. Regio- and stereoselectivity of the glycosylation of alcohols and 1,2:3,4-di-O-isopropylidene-a-D-galactopyranose by epoxide 7 under protocol A and/or B.
In the examination of the regio- and stereochemical behavior
of the new 2-O-MEM-substituted- -galactal-derived allyl epoxide
result was obtained with the exclusive formation of the b-linked
disaccharide 24b (Scheme 6).
D
7 with O-nucleophiles, the addition reactions with simple, low-
boiling alcohols, such as MeOH, EtOH, i-PrOH, and t-BuOH were
examined under protocol A reaction conditions (Scheme 6). The
reactions turned out to be completely 1,4-regioselective, with
the exclusive nucleophilic attack on the C(1) carbon of the allyl
system. No traces of the corresponding 1,2-addition products
were ever observed.
As for the stereoselectivity, the reactions with the less nucleo-
philic and more sterically hindered i-PrOH and t-BuOH, in spite
of the presence of a large amount of nucleophilic molecules (alco-
hol as the solvent) are completely b-stereoselective with the exclu-
sive formation of the corresponding alkyl b-O-glycosides 22b and
23b, respectively, with the same configuration (b) as the starting
epoxide. On the contrary, the reactions with the more nucleophilic
MeOH and EtOH are not stereoselective, and mixtures of the corre-
Analogously to the same reaction with the related epoxides 3
and 4 (X = H), the glycosylation of alcohols by epoxide 7 was ratio-
nalized by a coordination (hydrogen bond) between the oxirane
oxygen and the O-nucleophile. In this way, the nucleophile (alco-
hol) is guided to the b-face of 7, reacting through conformer
7A,10 which is suitably arranged for an entropically favored attack
on the C(1) carbon from the same side as the oxirane ring (route a,
Scheme 7). The corresponding 2-O-MEM-2,3-unsaturated-b-O-gly-
cosides 20–24b, with the same configuration as the starting epox-
ide, are consequently obtained in a directly substrate-dependent
stereoselective process.11
Procedures were tried in order to deprotect the 2-O-MEM func-
tionality and transform b-O-glycosides 20–24b into the corre-
sponding 3-deoxy-b-hexopyranosid-2-uloses: methyl (20b) and
i-propyl b-O-glycoside 22b were taken as suitable models for
checking the possibility of this transformation.
sponding anomeric methyl O-glycosides 20a and 20b (38:62) and
ethyl O-glycosides 21 and 21b (23:77) are obtained, respectively.
a
After observing that the usual deprotection protocol for the
O-MEM protective group (ZnCl2 in Et2O for 12 h at room
temperature) led to an extensive decomposition of the pseudogly-
cal system of 22b, a completely different approach was tried
(Scheme 8).
Even if not completely stereoselective, it is worth noting that, in
each mixture, the corresponding b-anomer 20b (62%) and 21b
(77%) with the same configuration as the starting epoxide is the
main product. However, in both cases, when the reactions are re-
peated under protocol B reaction conditions (Scheme 6), a decided
increase is observed in the b-stereoselectivity (90% with MeOH and
95% with EtOH), even if a completely b-stereoselective result is not
obtained, as it is with the structurally related allyl epoxide 4
(X = H) (Scheme 2).2b
R1
δ+
O
H
OR1
O
O
BnO
HO
route a
OBn
O
β
a
OMEM
1
Protocol B reaction conditions were used also for the reaction
with a partially protected monosaccharide such as 1,2:3,4-di-O-
β-O-glycoside
O
O
7A
O
isopropylidene-a-D-galactopyranose, for which protocol A reaction
conditions were not clearly possible. A complete b-stereoselective
Scheme 7. Rationalization of the substrate-dependent stereoselectivity in the
glycosylation of O-nucleophiles by allyl epoxide 7.
OR1
O
O
OR1
O
OR1
O
H2O/THF
Hg(OAc)2
BnO
HO
BnO
HO
AcOHg
25
BnO
HO
OMEM
OH
OMEM
AcOHg
20β, R1 = Me
22β, R1 = i-Pr
26
NaBH4
4N NaOH
Et3SiH
80%
71%
O
O-i-Pr
OH
O
OMe
O
O
O-i-Pr
BnO
HO
BnO
BnO
HO
+
HO
OH
29
1 : 1
28
27
Scheme 8. Deprotection of 2-O-MEM functionality of 20b and 22b and synthesis of hydroxy ketone 29 and 3-deoxy-O-glycosides 27 and 28.

4940
V. Di Bussolo et al. / Tetrahedron Letters 51 (2010) 4937–4941
O
O-i-Pr
O
O-i-Pr
O
O-i-Pr
MeOH
BnO
HO
BnO
NaBH4 BnO
OMEM 76%
Hg(OAc)2
OMEM
OMe
OMEM
HO
HO
OMe
AcOHg
30
22β
31
NaBH4
4N NaOH
61%
0.01 N aq. HCl
acetone
72%
O
O-i-Pr
O
O-i-Pr
O
O-i-Pr
BnO
HO
BnO
BnO
+
OMEM
OMe
OMEM
OMe
O
1 : 1
31
33
32
Scheme 9. Deprotection of 2-O-MEM functionality of 22b and synthesis of enone 33.
The oxymercuration of i-propyl O-glycoside 22b (an enol ether)
with Hg(OAc)2 in THF/H2O indicated a very fast consumption of the
starting material (TLC), and the possible formation of 3-acetoxy-
mercuryl hemiacetal 25 (R1 = i-Pr) in a reasonable equilibrium with
the corresponding 3-acetoxymercuryl ketone 26 (R1 = i-Pr). Actu-
ally, the subsequent addition of NaBH4 and aqueous 4 N NaOH
determined the formation of an almost 1:1 mixture of the diaste-
Laboratories for the financial support derived from the 2005 ADP
Chemistry Award.
Supplementary data
Supplementary data (experimental details for all reaction prod-
ucts and theoretical conformational analysis for epoxide 7-OMe)
associated with this article can be found, in the online version, at
doi:10.1016/j.tetlet.2010.07.061.
reoisomeric 3-deoxy-b-D D-xylo-hexopyrano-
-lyxo- (27)12 and -b-
sides (28),13 as the contemporary demercuration and reduction
product of the intermediate ketone 26 (R1 = i-Pr).
In this way, the clean deprotection of glycoside 22b was
achieved, but the carbonyl function at C(2) was lost. To our delight,
the replacement of NaBH4–4 N NaOH protocol with Et3SiH as the
demercuration agent turned out to be effective for our purposes
and, starting from methyl O-glycoside 20b, the desired hydroxy
ketone 29 was obtained as the only reaction product (Scheme 8).14,15
In a further attempt to affect the deprotection of 22b (or 20b),
References and notes
1. (a) Ohyabu, N.; Nishikawa, T.; Isobe, M. J. Am. Chem. Soc. 2003, 125, 8798–
8805; (b) Ichikawa, Y.; Isobe, M.; Bai, D.-L.; Goto, T. Tetrahedron 1987, 43,
4737–4748; (c) Udodong, U. E.; Fraser-Reid, B. J. Org. Chem. 1989, 54, 2103–
2112.
2. (a) Di Bussolo, V.; Caselli, M.; Romano, M. R.; Pineschi, M.; Crotti, P. J. Org. Chem.
2004, 69, 7383–7386; (b) Di Bussolo, V.; Caselli, M.; Romano, M. R.; Pineschi,
M.; Crotti, P. J. Org. Chem. 2004, 69, 8702–8708.
a
methoxymercuration–demercuration approach was tried
(Scheme 9). In this way, the reaction of enol ether 22b with
MeOH/Hg(OAc)2 followed by demercuration (NaBH4/aqueous 4 N
NaOH) of the intermediate 3-acetoxymercuryl acetal 30 led to an
almost 1:1 mixture of 4-hydroxy acetal 31 and the corresponding
3. The -O-MEM protection was preferred to –OAc, –OMe and –OBn protective
groups in view of the easier introduction, absence of any interference with the
synthetic sequence to epoxide 7 and easy removable protocol.
4. Di Bussolo, V.; Fiasella, A.; Balzano, F.; Uccello-Barretta, G.; Crotti, P. J. Org.
Chem. 2010, 75, 4284–4287.
5. For a recent use of glycosyl sulfoxides in the synthesis of 2-O- and 2-N-
protected glycals, see: Liu, J.; Huang, C.-Y.; Wong, C.-H. Tetrahedron Lett. 2002,
43, 3447–3448.
6. (a) Di Bussolo, V.; Kim, Y.-J.; Gin, D. Y. J. Am. Chem. Soc. 1998, 120, 13515–
13516; (b) Honda, E.; Gin, D. Y. J. Am. Chem. Soc. 2002, 124, 7343–7352.
7. Gordon, D. M.; Danishefsky, S. J. Carbohydr. Res. 1990, 206, 361–366. For a
tentative rationalization of the different stereoselectivity found in the
a
,b-unsaturated derivative 32.16 In order to avoid the elimination
process, no aqueous 4 N NaOH was added in the demercuration
step of intermediate 30, and only NaBH4 was used. Under these
modified conditions, 4-hydroxy acetal 31 turned out to be the only
reaction product. However, once again, the elimination process
could not be completely eliminated. Actually, the subsequent
hydrolysis of the acetal functionality of 31, necessarily carried
out under acid conditions (0.01 N aqueous HCl/acetone), afforded
the desired deprotection which was, also in this case, accompanied
by a simultaneous elimination reaction, with the formation of en-
one 33,17 as the only final product (Scheme 9).18
glycosylation of PhSeH by intermediate
a-epoxide 12 under the modified Gin
protocol and Danishefsky reaction conditions, see Ref. 4.
8. Corey, E. J.; Gras, J.-L.; Ulrich, P. Tetrahedron Lett. 1976, 17, 809–812. The
insertion protocol based on the reaction of 14 with MEM-Cl in the presence of
DIPEA in CH2Cl2 was unsuccessful.
9. Chambers, D. J.; Evans, G. R.; Fairbanks, A. J. Tetrahedron Lett. 2003, 44, 5221–
5223.
10. A theoretical conformational analysis carried out on an appropriate, simplified
In conclusion, we have demonstrated the possibility of synthesiz-
model (epoxide 7-OMe) indicated that epoxide
7 exists as the only
ing stereoselectively 2-O-MEM-3-deoxy-b-
anosides, such as 20–24b, by the application of the directly
substrate-dependent glycosylation process to the new 2-O-MEM-
D-threo-hex-2-enopyr-
corresponding conformer 7A, with the –CH2OBn group equatorial (see
Supplementary data).
11. The non-completely b-stereoselective glycosylation process observed with
MeOH and EtOH under protocol A and B reaction conditions could be simply
ascribed to their more nucleophilic character compared to the other alcohols
tried. Accordingly, the b-selectivity increases when less nucleophile is used
(protocol B).
12. For compounds structurally related to 27, see: (L-series) (a) Lei, P.-S.;
Duchaussoy, P.; Sizun, P.; Mallet, J.-M.; Petitou, M.; Sinay, P. Bioorg. Med. Chem.
1998, 6, 1337–1346; (b) Huber, H.; Reichstein, T. Helv. Chim. Acta 1948, 31, 1645–
1655.
13. For compounds structurally related to 28, see: (a) Kovac, P.; Edgar, K. E.
Carbohydr. Res. 1990, 201, 79–93; (b) Chiu, T. M. K.; Watanabe, K. A.; Fox, J. J.
Carbohydr. Res. 1974, 32, 211–216; (c) Zinner, H.; Wulf, G. J. Prakt. Chem. 1970,
312, 635–640; (d) Antonakis, K. Bull. Soc. Chim. Fr. 1969, 122–126.
14. Although new in the present application, the use of the oxymercuration–
demercuration protocol for the hydrolysis of a vinyl-O-MEM ether stems from
a previously described original result: Ireland, R. E.; Norbeck, D. W. J. Am. Chem.
Soc. 1985, 107, 3279–3285. To the best of our knowledge, the use of Et3SiH in
the demercuration step is original.
D
-
galactal-derived allyl epoxide 7. The 2-O-MEM-b-O-glycosides
obtained can be successfully transformed into corresponding
27- and 28-type 3-deoxy-b-O-glycosides, 29-type 3-deoxy-b-hexo-
pyranosid-2-uloses, and 33-type 3,4-dideoxy-b-hex-3-enopyrano-
sid-2-uloses, which are useful synthetic intermediates for further
transformations, for which a synthetic procedure, starting from a
common precursor, is not presently available. A new mild protocol
for the deprotection of the vinyl-O-MEM functionality and the
formation of the corresponding ketone is also described.
Acknowledgments
This work was supported by the University of Pisa and MIUR
(PRIN 2008), Roma. P.C. gratefully acknowledges Merck Research
15. For compounds structurally related to 29, see Ref. 13d and Zunszain, P.; Varela,
O. Tetrahedron: Asymmetry 1998, 9, 1269–1276.

V. Di Bussolo et al. / Tetrahedron Letters 51 (2010) 4937–4941
4941
16. The axial position of C(4)–OH group and the slight acidity of axial C(3)–H
a
to
Holden, N. L.; Fraser-Reid, B. Can. J. Chem. 1973, 51, 3357–3365; (d) Bock, K.;
Pedersen, C. Tetrahedron Lett. 1969, 10, 2983–2986.
the acetal group can reasonably be the cause of the elimination process of 31 to
32 under basic conditions.
17. For compounds structurally related to 33, see Ref. 15 and: (a) Köll, P.; Klenke,
K.; Eisermann, D. J. Carbohydr. Chem. 1984, 3, 403–415; (b) Brimacombe, J. S.;
Hunedy, F.; Mather, A. M.; Tucker, L. C. N. Carbohydr. Res. 1979, 68, 231–238; (c)
18. The use of more diluted aqueous HCl/acetone solution (10À5–10À3 N) in the
acetal hydrolysis step was not effective, and the starting acetal 31 was
recovered completely unreacted.