456
C. J. Carroux et al. / Bioorg. Med. Chem. Lett. 23 (2013) 455–459
Copper-catalysed azide–alkyne cycloaddition (CuAAC) with car-
unreactive), Scheme 1, extensive chromatography was however
required for purification. The second approach employed peracet-
ylated b-glycopyranosyl azides 1 and 6 as indirect precursors for
the remaining target azides 2–5 and 7–10, Scheme 2. Deprotection
of 1 and 6 using Zemplén conditions gave the b-glycopyranosyl
azides 13 and 14, respectively, in an almost quantitative yield.29
Next acylation of 13 and 14 with the corresponding propionic,
butyric, pentanoic or 3-methyl butyric acid anhydride in pyridine
gave the target per-O-acylated b-glycopyranosyl azides 2–5 and
7–10 in good to high yields. This second approach was preferred
to the first as it provided good yields from accessible starting
materials and circumvented the purification difficulty caused
by the formation and use of anomeric mixtures. The 1H and
13C NMR spectra of azide compounds 2–5 and 7–10 were consis-
tent with their expected structures.
With the panel of anomeric glycopyranosyl azides 1–10 in hand
next the 1,4-disubstituted-1,2,3-triazole glycoconjugate target
compounds 15–24 were synthesised from these azides and phenyl-
acetylene by CuAAC, Scheme 2. The 1H and 13C NMR spectra of 15–
24 were consistent with their expected structures, the H-4 proton
of the 1,4-disubstituted triazoles resonated at d 8.95–8.82 ppm,
while the triazole CH carbon resonated at d 120.6–120.3 ppm.
These chemical shifts are in agreement with those reported for
other 1,4-disubstituted-1,2,3-triazoles.30
bohydrate-based azide substrates is now routinely applied in drug
discovery and has proven extremely versatile both in the prepara-
tion of glycoconjugate small molecule inhibitors against a range of
enzymes and to append carbohydrates to biomolecules.16–18 Sev-
eral reactions fulfill the criteria that defines click chemistry,19 how-
ever it is CuAAC to regioselectively form 1,2,3-triazoles that has
proven the most accomplished click chemistry reaction for the
development of new molecules with useful chemical properties,
delivering an impressive volume of diverse applications within a
short period.20 In drug development some functional groups, while
synthetically attractive, can be a liability to biopharmaceutical per-
formance of a compound in vivo. The early applications of glyco-
pyranosyl azide building blocks in click chemistry demonstrated
that the formed triazole was compatible with commonly employed
carbohydrate protecting group approaches21 and that peracetylat-
ed glycopyranosyl triazoles were robust molecules.21,22 Glycopyr-
anosyl azides are stable and are inert towards a wide range of
reaction conditions and are readily synthesised anomerically
pure.16 In addition a range of bioactive glycopyranosyl triazoles
have been reported.4,6,17,22–25
The azide precursors of the glycoconjugate target compounds of
this study, compounds 1–10, are O-peracylated b-D-glucopyranosyl
and b- -galactopyranosyl azides containing acetyl (1 and 6), propi-
D
onyl (2 and 7), butanoyl (3 and 8), pentanoyl (4 and 9) and 3-meth-
ylbutanoyl (5 and 10) moieties, Figure 1. Compounds 1 and 6 are
well known and have been widely used,16 however the remaining
acyl analogues are novel compounds. To synthesise compounds 1
and 6 we employed the Lewis acid catalysed synthesis directly
Compound 25 is the fully deacylated glycoconjugate that would
result following the complete hydrolysis of acyl groups of the par-
ent compounds 15–19. Compound 25 was prepared from 15 using
Zemplén conditions and was used as the control compound for the
biopharmaceutical property assays with the glucose series com-
pounds 15–19 described next. Esterase hydrolysis of compounds
15–19 may take place sequentially, to form multiple tri-acylated,
di-acylated and mono-acylated glycoconjugates as intermediates,
Scheme 3. The characteristic molecular weights of the potential
intermediates allow monitoring by mass spectrometry.
In order to assess the susceptibility of these compounds to ester
hydrolysis within plasma, we determined the in vitro stability of
compounds 25 and 15–19 in human plasma at 37 °C, Table 1.
The pentanoyl (18) and 3-methylbutanoyl (19) esters, together
with the corresponding fully deacylated compound 25 were stable
in human plasma. In contrast, the shorter chain acetyl, propionyl
and butanoyl analogues (15, 16 and 17) degraded with estimated
half lives of 55, 26, and 555 min, respectively, the order of stability
is butanoyl > acetyl > propionyl. Multiple mono-, di- and tri-acyl-
ated hydrolysis products as well as the fully hydrolysed compound
25 were detected for each labile compound consistent with serial
hydrolysis of the ester moieties.
from commercially available per-O-acetylated
D-glucose and
per-O-acetylated
D
-galactose using trimethylsilyl azide16 to stere-
oselectively introduce the azide functional group at the anomeric
centre. We then considered two approaches for the synthesis of
azides 2–5 and 7–10. The first approach operated similarly to the
synthesis of 1 and 6, with the intention to directly introduce the
azide at the anomeric centre of the corresponding per-O-acylated
glucose and galactose derivative. The per-O-acylated precursors
needed to be synthesised and acylation of unprotected
with an acid anhydride using typical monosaccharide acylation
conditions was investigated. The reaction of -glucose with propi-
D-glucose
D
onic anhydride and butyric anhydride in either refluxing pyridine
or using sodium propanoate or sodium butanoate, respectively,
generated an anomeric mixture of the peracylated sugars 11 and
12, Scheme 1.26–28 While the target b-anomer predominated
(typical ratio b:
the -anomer by chromatography or crystallization. We continued
with compounds 11 and 12 as anomeric mixtures and successfully
a
ꢀ80:20) it could not be readily separated from
a
Given the broad range of calculated LogP (cLogP) values for the
acylated glycoconjugates (+2.25 to +8.60, Table 1) and the likely
converted the b-anomer to azides 2 and 3 (the
a-anomer is
Compd
R
O
1 (Glc), 6 (Gal)
O
O
O
2 (Glc), 7 (Gal)
3 (Glc), 8 (Gal)
4 (Glc), 9 (Gal)
5 (Glc), 10 (Gal)
OR
OR
OR
O
O
RO
N3
RO
N3
RO
OR
1-5 (Glc)
OR
6-10 (Gal)
O
Figure 1. Target b-glucosyl azides 1–5 and b-galactosyl azides 6–10 each comprising four acetyl, propionyl, butanoyl, pentanoyl and 3-methylbutanoyl acyl groups.
Carroux, Cindy J.
