Synthesis of 3-azido-3-deoxy-β-d-galactopyranosides

Article abstract of DOI:10.1016/j.carres.2009.05.005

Three efficient routes to 3-azido-3-deoxy-β-d-galactopyranosides were developed relying on a double inversion protocol at C3. Two of the routes were demonstrated to work with both O- and S-glycosides. In all three routes, the 2-O-acetyl-3-azido-4,6-O-benz

Full text of DOI:10.1016/j.carres.2009.05.005

Carbohydrate Research 344 (2009) 1282–1284
Contents lists available at ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Synthesis of 3-azido-3-deoxy-b-D-galactopyranosides
Christopher T. Öberg, Ann-Louise Noresson, Tamara Delaine, Amaia Larumbe, Johan Tejler,
*
Henrik von Wachenfeldt, Ulf J. Nilsson
Organic Chemistry, Lund University, PO Box 124, SE-22100 Lund, Sweden
a r t i c l e i n f o
a b s t r a c t
Article history:
Three efficient routes to 3-azido-3-deoxy-b-D-galactopyranosides were developed relying on a double
Received 30 March 2009
Accepted 7 May 2009
Available online 9 May 2009
inversion protocol at C3. Two of the routes were demonstrated to work with both O- and S-glycosides.
In all three routes, the 2-O-acetyl-3-azido-4,6-O-benzylidene-3-deoxy-b- -galactopyranosides were
obtained by an azide inversion of the key intermediates 2-O-acetyl-4,6-O-benzylidene-3-O-trifluoro-
methanesulfonyl-b- -gulopyranosides. The intermediate gulopyranosides were in turn obtained from
2-O-acetyl-4,6-O-benzylidene-3-O-trifluoromethanesulfonyl-b- -galactopyranosides, installed in one
pot from the 4,6-O-benzylidene-b- -galactopyranosides, by inversion with nitrite or acetate. For O-glyco-
sides, the gulopyranoside configuration could alternatively be obtained from the 4,6-O-benzylidene-b-
D
D
Keywords:
Galactose
Gulose
Epimerization
Azide
D
D
D-
galactopyranoside by elimination to give the 2,3-dianhydro derivative followed by a highly stereoselec-
tive cis-dihydroxylation.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
furanose in which C3 is doubly inverted together with a single
inversion at C4. This route has been used, for example, in the syn-
D
-Galactopyranose is ubiquitous to mammalian glycoconju-
thesis of 3-amino-gal-UDP, which still is one of the most potent
GalT inhibitors reported to date.3 Synthesis strategies toward gan-
glioside GM3 analogs incorporating a galactose 3-thio linkage were
developed based on the discovery that 4,6-O-benzylidene-2-O-
gates, where it can be glycosidically linked via O1–O4 and O6 oxy-
gen atoms. The HO2 is fucosylated in blood group determinants,
the HO3 is glycosylated with various monosaccharides, including
sialic acid, as well as sulfated, HO4 is glycosylated with b-
D-GalNAc
acyl-b-D-galactopyranoside derivatives readily undergo double
or -Gal in the ganglio- and globo-series of glycolipids, HO6 is
a
-
D
inversions at C3 and thus, via the gulo-epimer, provide access to
the galactose 3-thio analogs that either act as neuraminidase
inhibitors4,5 or can be used as vaccine constituents with improved
hydrolytic stability.6 More recently, the Ramström group has made
substantial contributions to our general understanding of the influ-
ence of protecting group pattern, relative configuration, solvent,
and nucleophile on carbohydrate hydroxyl inversion efficiency,
which in turn has led to the development of highly efficient and
practical protocols for carbohydrate epimerizations including
C4.7–9
We have reported on efficient galectin inhibition with galactose
3C-amides or 3C-triazoles.10–16 These inhibitors have all been
derived from the 3-azido-3-deoxy-galactopyranose originally
described by Lowary and Hindsgaul3 and later developed into a
thioglycoside donor by us.10 However, the synthesis of 3-azido-3-
sialylated or sulfated in glycolipid or glycoprotein glycans. The
diversity of galactose glycosylation and modification in mamma-
lian glycoconjugates most often play critical roles in controlling
the function of the glycoconjugate. The chemical properties of
the glycosidic atom linked to galactopyranose can be expected to
influence conformational preferences, as well as interactions with,
for example, proteins, which in turn affects biological activities.
Hence, replacement of oxygens of glycosidic bonds involving galac-
topyranose with other atoms is an important strategy for decipher-
ing the biological importance of the glycosidic oxygens in
galactose-containing glycans. Within this context, various oxygen
replacements of HO3, HO4, and HO6 have been reported, while
HO2 oxygen replacements are less frequent. Galactose HO4 have
been readily prepared, for example, by substituting glucose 4-O-
sulfonate leaving groups with nucleophiles.1,2 Galactose HO6 is
maybe even more readily accessible via substitutions of a, for
example, galactose 6-O-sulfonate. In contrast, replacement of the
galactose HO3 oxygen is more challenging. A reliable route for
deoxy galactosides from di-isopropylidene-D-glucofuranose
involves several steps of which a majority requires column chro-
matography for purification of intermediates, PCC oxidations, and
hydrogenations that are laborious on large scale. This prompted
us to investigate the possibility to introduce the 3-azido-function-
replacement of this oxygen starts from di-isopropylidene-D-gluco-
ality by double inversions at C3 of 4,6-O-benzylidene-b-D-galacto-
pyranosides. Apart from being potentially more efficient and easier
to handle on large scale, this route would allow for installation of
* Corresponding author.
E-mail address: ulf.nilsson@organic.lu.se (U.J. Nilsson).
0008-6215/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2009.05.005

C. T. Öberg et al. / Carbohydrate Research 344 (2009) 1282–1284
1283
an O- or S-glycoside aglycon prior to 3-azido introduction. Herein,
we report on the evaluation and optimization of two alternative
double inversion protocols at C3 of 4,6-O-benzylidene-b-D-galacto-
pyranosides that are efficient routes toward 3-azido-3-deoxy
galactopyranosides.
was immediately converted into the 3-azido-3-deoxy-galactoside
5 in a 72% yield from 7. The higher yield in going from 7 to 5, as
compared to going from 3 to 5, is unexpected, because the route
7 to 5 passes through 3 as an intermediate. We believe that the ex-
planation is in the different scale the reactions were performed un-
der. Route 3 to 5 was performed in 0.3 mmol scale and route 7 to 5
in 13.6 mmol scale. Typically, triflation yields increase with the
scale of the reaction.
2. Results and discussion
An alternative third double inversion route was envisioned pos-
sible via the readily accessible 2,3-dianhydro galactoside 9, which
should give gulopyranosides upon dihydroxylation. Hence, 1 was
instead treated with excess triflic anhydride to give the 2,3-di
-triflate 8, which was immediately eliminated in a zinc-free Tip-
son–Cohen18,19 reaction to give the 2,3-dianhydro galactoside 9
in high yield (Scheme 2). Cis-dihydroxylation under Upjohn condi-
tions20 of 9 afforded the expected C3-inverted gulo compound 7
(79 %), as the only observed diastereomer, together with unreacted
9 (21%). Compound 7 was converted into the target azide 5 as de-
scribed above.
2.1. O-Glycosides
First, two closely related routes to methyl 2-acetyl-3-azido-4,6-
O-benzylidene-3-deoxy-b-
gated. Both routes started with activation of HO3 of methyl
4,6-O-benzylidene-b- -galactopyranoside 1 by triflation, followed
D-galactopyranoside 5 were investi-
D
by acylation of HO2, in a high-yielding one-pot reaction: Treat-
ment of 1 with triflic anhydride in dichloromethane/pyridine,17
followed by addition of acetyl chloride provided the triflate 2
(Scheme 1). The crude triflate 2 was immediately treated with tet-
rabutylammonium nitrate to give the gulo-derivative 3 in 60%
from 1. Triflation of 3, followed by immediate treatment of the
crude gulo-triflate 4 with tetrabutylammonium azide gave the
3-azido-3-deoxy-galactoside 5 in 62% from 3.
2.2. S-Glycosides
Although the double inversion route via the 2,3-dianhydro com-
pound 9 (Scheme 2) was deemed more practical and more efficient
than the double inversion via the two triflates 2 and 4 (Scheme 1),
it was not applicable to thioglycosides, which was probably due to
their sensitivity to the formation of thiophilic iodine during the so-
dium iodide-mediated elimination (as described for 8 to give 9).
Furthermore, the oxidative conditions required to generate the
guloside 7 from 9 are generally not compatible with thio-glyco-
In the second alternative route, the first inversion with the ga-
lacto-triflate was done with cesium acetate to give the diacetate
6 (Scheme 1). The inversion with cesium acetate to give 6 was
somewhat higher yielding than the corresponding inversion with
sodium nitrite to give 3. However, this advantage was to some ex-
tent counter-balanced by the need for two extra steps to reach the
gulo-triflate 4 necessary for azide introduction. Nevertheless, these
two extra steps, deacetylation to give the dihydroxy compound 7
and then selective HO2-acetylation and HO3-triflation toward 4,
were indeed uneventful and proceeded in high yields. Acetylation
of 7 with one equivalent of acetyl chloride at ambient temperature
proceeded with high selectivity for HO2 and addition of triflic
anhydride at À10 °C provided the triflate 4. The crude triflate 4
sides. Hence, synthesis of 3-azido-3-deoxy-1-thio-b-D-galactopyr-
anosides, for example, for use as galactosyl donors, was only
possible via sequential inversion of O3 galacto and gulo triflates
according to routes 1 and 2 described above for the corresponding
methyl galactoside 5. Applicability of routes 1 and 2 was demon-
Ph
O
O
O
OMe
Ph
OAc
Ph
Ph
HO
3
O
O
O
O
O
c
O
Ph
b
e
O
O
O
d
R2O
OMe
N3
OMe
OMe
O
OR1
OAc
O
OAc
g
TfO
O
1 R1=R2=H
5
4
OMe
a
2 R1=Ac, R2=OTf
OR1
R2O
6 R1=R2=Ac
f
7 R1=R2=H
Scheme 1. Reagents and conditions for routes 1 and 2: (a) i. Tf2O, pyridine, CH2Cl2, À10 °C, ii. AcCl; (b) QNO2, DMF, 50 °C (60% from 1); (c) Tf2O, pyridine, CH2Cl2, À10 °C; (d)
i
QN3, DMF, 50 °C (62% from 3, 72% from 7); (e) CsOAc, DMF (81% from 1); (f) NaOMe, MeOH (92%); (g) AcCl, pyridine, CH2Cl2, rt, iiTf2O, À10 °C.
Ph
Ph
Ph
Ph
O
O
O
O
O
O
O
O
O
O
O
O
d-e
c
b
RO
OMe
OMe
OMe
N3
OMe
OR
OH
OAc
OH
9
7
5
1 R=H
a
8 R=OTf
Scheme 2. Reagents and conditions for route 3: (a) Tf2O, pyridine, CH2Cl2, À10 °C (95%); (b) NaI, DMF, imidazole (95%); (c) OsO4, NMO, DABCO, MsNH2, CH2Cl2 (79%); (d) i.
AcCl, pyridine, CH2Cl2, rt, ii. Tf2O, À10 °C; (e) QN3, DMF, 50 °C (72% from 7).

1284
C. T. Öberg et al. / Carbohydrate Research 344 (2009) 1282–1284
Ph
O
O
O
SPh
Ph
Ph
Ph
OAc
HO
O
O
O
12
O
c
O
O
Ph
O
O
O
b
e
d or h
R2O
SPh
SPh
N3
SPh
O
OR1
OAc
OAc
O
TfO
g
14
O
10 R1=R2=H
13
SPh
a
11 R1=Ac, R2=OTf
OR1
R2O
15 R1=R2=Ac
f
16 R1=R2=H
Scheme 3. Reagents and conditions for routes 1 and 2 applied onto phenyl 1-thio-b-
D
-galactopyranoside: (a) i. Tf2O, pyridine, CH2Cl2, À10 °C, ii. AcCl; (b) QNO2, DMF, 50 °C
(60% from 10); (c) Tf2O, pyridine, CH2Cl2, À20 °C; (d) QN3, DMF, 50 °C (59% from 12); (e) CsOAc, DMF (77% from 10); (f) NaOMe, MeOH (98%); (g) i. AcCl, pyridine, CH2Cl2, rt, ii.
Tf2O, À10 °C; (h) NaN3, DMF, 50 °C (42% from 16).
Table 1
herein are demonstrated with azide ion as the second nucleophile,
we envision that reaction of 4 and 13 using other nucleophiles, such
as halides, sulfur or carbon-based nucleophiles, will allow for effi-
cient preparation of other 3-deoxy-galactoside derivatives.
Summary of routes to 3-azido-galactose derivatives
Route
No of rxns
No of chrom.
Total yield (%)
Literature routea
Route 1, Scheme 1
Route 2, Scheme 1
Route 3, Scheme 2
Route 1, Scheme 3
Route 2, Scheme 3
1121–23,3,10,15
9
2
3
2
2
2
36b/19c
36
53
4
5
5
4
5
Acknowledgments
51
25d/35e
32
We thank Lund University, the Swedish Research Council, the
programs ‘Glycoconjugates in Biological Systems’ (GLIBS), and
‘Chemistry for Life Sciences’ sponsored by the Swedish Strategic
Research Foundation for financial support.
a
From di-isopropylidene-
-galactopyranoside.21–23,3,10,15
The best overall yield possible by combining the literature reports and our own
D-glucofuranose to methyl 3-azido-4,6-O-benzylidene-
3-deoxy-1-thio-b-
D
.
b
results.
c
d
e
The best overall yield in our hands.
NaN3 was used in step h.
QN3 was used in step d.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.carres.2009.05.005.
strated by conversion of the phenyl 4,6-benzylidene-1-thio-b-
galactopyranoside 10 into the phenyl 2-O-acetyl-3-azido-4,6-ben-
zylidene-3-deoxy-1-thio-b- -galactopyranoside 14 in good overall
yields (Scheme 3). The yield of 14 could be improved by using tet-
rabutylammonium azide (compare Scheme 1), however for larger
scale we prefer using sodium azide for safety and cost reasons de-
spite the slightly lower yield.
D-
References
D
1. Nilsson, U.; Johansson, R.; Magnusson, G. Chem. Eur. J. 1996, 2, 295–302.
2. Sandström, C.; Nilsson, U. J.; Magnusson, G.; Kenne, L. Carbohydr. Res. 1999, 322,
46–56.
3. Lowary, T. L.; Hindsgaul, O. Carbohydr. Res. 1994, 251, 33–67.
4. Liakatos, A.; Kiefel, M. J.; von Itzstein, M. Org. Lett. 2003, 5, 4365–4368.
5. Liakatos, A.; Kiefel, M. J.; Fleming, F.; Coulson, B.; von Itzstein, M. Bioorg. Med.
Chem. 2006, 14, 739–757.
6. Rich, J. R.; Bundle, D. R. Org. Lett. 2004, 6, 897–900.
7. Dong, H.; Pei, Z.; Ramström, O. J. Org. Chem. 2006, 71, 3306–3309.
8. Pei, Z.; Dong, H.; Ramström, O. J. Org. Chem. 2005, 70, 6952–6955.
9. Dong, H.; Rahm, M.; Brinck, T.; Ramström, O. J. Am. Chem. Soc. 2008, 130,
15270–15271.
3. Summary
A comparison of the original route from the commercial di-iso-
propylidene-glucofuranose with the herein presented routes from
10. Sörme, P.; Qian, Y.; Nyholm, P.-G.; Leffler, H.; Nilsson, U. J. ChemBioChem 2002,
3, 183–189.
11. Sörme, P.; Arnoux, P.; Kahl-Knutsson, B.; Leffler, H.; Rini, J. M.; Nilsson, U. J. J.
Am. Chem. Soc. 2005, 127, 1737–1743.
commercial methyl 4,6-benzylidene-b-
D
-galactopyranoside 1 and
phenyl 4,6-benzylidene-1-thio-b-
D-galactopyranoside 10 is pre-
sented in Table 1.
12. Cumpstey, I.; Sundin, A.; Leffler, H.; Nilsson, U. J. Angew. Chem., Int. Ed. 2005, 44,
5110–5112.
13. Salameh, B. A.; Leffler, H.; Nilsson, U. J. Bioorg. Med. Chem. Lett. 2005, 15, 3344–
3346.
4. Concluding remarks
14. Salameh, B. A.; Sundin, A.; Leffler, H.; Nilsson, U. J. Bioorg. Med. Chem. 2006, 14,
1215–1220.
15. Öberg, C. T.; Leffler, H.; Nilsson, U. J. J. Med. Chem. 2008, 51, 2297–2301.
16. Cumpstey, I.; Salomonsson, E.; Sundin, A.; Leffler, H.; Nilsson, U. J. Chem. Eur. J.
2008, 14, 4233–4245.
17. Knapp, S.; Kukkola, P. J.; Sharma, S.; Murali Dhar, T. G.; Naughton, A. B. J. J. Org.
Chem. 1990, 55, 5700–5710.
18. Tipson, R. S.; Cohen, A. Carbohydr. Res. 1965, 1, 338–340.
19. Kobayashi, Y.; Tsuchiya, T.; Umezawa, S.; Yoneta, T.; Fukatsu, S.; Umezawa, H.
Bull. Chem. Soc. Jpn. 1987, 60, 713–720.
In conclusion, 3-azido-3-deoxy-b-
synthesized in few steps and in high yields via double inversion of
4,6-O-benzylidene-b- -galactopyranosides. The route based on cis-
hydroxylation of 2,3-dianhydro-b- -galactopyranosides is faster,
more practical, and preferable for the preparation of 3-azido-3-
deoxy-b- -galactopyranosides. However, another route based on
sequential inversion of galacto and gulo 3-O-triflates is preferred if
1-thio-b- -galactopyranosides are the desired product. Hence, the
D-galactopyranosides can be
D
D
D
D
20. Ahrngren, L.; Sutin, L. Org. Process Res. Dev. 1997, 1, 425–427.
21. Lankin, D. C.; Nugent, S. T.; Rao, S. N. Carbohydr. Res. 1993, 244, 49–68.
22. Slessor, K. N.; Tracey, A. S. Can. J. Chem. 1969, 47, 3990–3995.
23. Lemieux, R. U.; Stick, R. V. Aust. J. Chem. 1975, 28, 1799–1801.
two routes complement each other and both serve as attractive
alternatives to the more common literature procedure from di-iso-
propylidene-D-glucose. Although, the double inversion protocols