Pyridine-sugar conjugates as potent inhibitors of enzyme-catalysed glycoside hydrolysis

Article abstract of DOI:10.1039/b211399f

A series of O- and N-linked pseudo-disaccharides incorporating simple functionalised pyridines were synthesized and demonstrated potent inhibition of the glucoamylase-catalysed reaction.

Full text of DOI:10.1039/b211399f

Pyridine–sugar conjugates as potent inhibitors of enzyme-catalysed
glycoside hydrolysis
Peter A. Nkansah†, Alan H. Haines and N. Patrick J. Stamford*
School of Chemical Sciences, University of East Anglia, Norwich, UK NR4 7TJ
Received (in Cambridge, UK) 21st November 2002, Accepted 6th February 2003
First published as an Advance Article on the web 20th February 2003
A series of O- and N-linked pseudo-disaccharides incorpo-
rating simple functionalised pyridines were synthesized and
demonstrated potent inhibition of the glucoamylase-cata-
lysed reaction.
It is generally recognized1 that glycosides hydrolyse chemically
via a cyclic oxocarbenium ion and that a similar transition state
Scheme 1 Reagents and conditions: i, I2, Na2CO3, 26%; ii, BnBr, NaH,
is almost certainly involved in the enzyme-catalysed process.
DMF, 80%.
Potent known glycosidase inhibitors mimic features of either
the enzymeAs substrate (e.g. 1) and/or the transition state
involved in the enzyme-catalysed process 2.2 Unfortunately, the
complexity of these molecules often makes their syntheses and
that of similar compounds a time consuming and difficult task.
Moreover, this synthetic problem is magnified considerably in
the development of glycosidase-specific inhibitors by the level
of structural and chemical complexity in heterogeneous carbo-
hydrate chains.
2-pyridylthio pseudo-sugars and suggests that their effective-
ness may largely be due to the pyridine moiety alone.
In the study of 2-pyridylthio pseudo-sugars,6 as with that
where cyclic guanidinium ions were utilised in the formation of
pseudo-disaccharides,5 the enzymes investigated were glycosi-
dases in which it is likely that the majority of bonding
interactions occurred solely with the terminal sugar unit. A
more pragmatic approach would be to incorporate these simple
inhibitor units into pseudo-saccharides as inhibitors of glycosi-
dase reactions where the substrate recognition event is much
more dependent on co-operative bonding interactions between
the enzyme and more than one sugar unit of the substrate
oligosaccharide. Towards these ends we synthesized a series of
O- and N-linked pseudo-disaccharides incorporating simple
functionalised pyridines such as 3 as potential inhibitors of well
characterized amylase reactions.
To circumvent these problems some research has recently
been directed towards identifying new glycomimetic glycosi-
dase inhibitors which are based on very simple molecular
designs. For example, it is known that acyclic molecules such as
trihydroxymethylaminomethane (tris)3 and simple heterocyclic
compounds like the cyclic guanidinium ion4 are competitive
inhibitors of glycosidases. Cyclic guanidinium ions have also
been employed in the formation of pseudo-disaccharides in an
attempt to improve inhibitor efficacy and selectivity.5 Inter-
estingly, these disaccharide analogues were shown to bind less
effectively than simple derivatives of the cyclic guanidinium
moiety.
Preparation of amino and hydroxy pyridines can be achieved
through nucleophilic substitution of a pyridine substrate. For the
manufacture of more complex conjugates incorporating pyr-
idine with a spatial arrangement of peripheral hydroxyl groups
similar to that observed for 3, an appropriately substituted
pyridine first had to manufactured. In this case an adaptation of
the procedure described for the synthesis of 3-hydroxy-
2-iodopyridine9 was utilised to afford 3-hydroxy-2-hydrox-
ymethyl-6-iodopyridine (4).‡ Subsequent O-benzylation then
afforded a substrate (5) that could be used in SNAr reactions.
Nucleophilic substitution worked tolerably well for the
preparation of the desired O-linked conjugate methyl 2,3,6-tri-
A key, seminal observation by Withers6 was that carbohy-
drate–pyridine conjugates also possessed useful inhibitory
properties towards glycosidases but, unfortunately, in this case
no comparison was made to the effectiveness of the isolated
pyridine moiety as an inhibitor of the enzyme. This suggested
that a detailed examination of suitably substituted pyridines
might be fruitful in the development of a simple yet effective
class of new glycosidase inhibitors. An investigation that
seemed especially relevant considering the benzyl ring has been
successfully mobilised as an effective mimetic of the flat
pyranosyl donor sugar proposed for the transition state in a
glycosyltransferase reaction.7
O-benzyl-4-O-(2A-pyridyl)-a-
methyl 2,3,6-tri-O-benzyl-a-
2-chloropyridine (Scheme 2). The methyl 2,3,6-tri-O-benzyl-a-
-glucopyranoside (6) used in this reaction was prepared from
methyl 4,6-O-benzylidene-a-
-glucopyranoside10 via benzyl
D
-glucopyranoside (7a) from
-glucopyranoside (6) and
D
D
D
Work recently performed in this laboratory8 identified
3-hydroxy-2-hydroxymethylpyridine (Scheme 1; 3) as a moder-
ate inhibitor of the b-glucosidase [EC 3.2.1.21] from sweet
almonds (Ki = 1.9 3 1023 M). This observation compares
favourably with the study by Withers and coworkers6 where
kinetic data was obtained using a b-glucosidase and a range of
Scheme 2 Reagents and conditions: i, 2-Chloropyridine, NaH, NMP, 140
°C for 12 h, 7a = 50%; ii, 5, NaH, NMP, Cu(I)Br, 140 °C for 18 h, 7b =
30%; iii, EtOH–cyclohexene (2+1), Pd black, reflux, 24 h, 8a = 82%, 8b =
46%.
† Present address: Faculty of Pharmacy, Kwame Nkrumah University of
Science and Technology, Kumasi, Ghana, West Africa.
784
CHEM. COMMUN., 2003, 784–785
This journal is © The Royal Society of Chemistry 2003

Table 1 Glucoamylase inhibitory activity of 2-pyridyl pseudo-sugars
Compound
3
8a
8b
11a
11b
Ki (M)
No inhibition at 1 3 1022
No inhibition at 1 3 1023
1.1 3 1023
2.2 3 1024
1.8 3 1026
protection11 and regioselective ring opening12 of the 4,6-O-
benzylidene acetal moiety.
structure of the mimetic may therefore indicate an appropriate
match for the topological and spatial arrangement of the all-
equatorial functionalities at C-4 and C-5 of the non-reducing
glucose unit of the substrate to enable these important hydrogen
bond formations. Furthermore, these results, in conjunction
with previous studies,20 also enforce the importance of having a
basic exocyclic amino group adjacent to C-1 of a sugar for
effective binding in the amylase active site.
In conclusion, the preparation of O- and N-linked pseudo-
disaccharides incorporating simple functionalized pyridines has
been achieved. Moreover, their associated biological activity
demonstrates that this approach represents a viable new
alternative for the development of simple non-carbohydrate
carbohydrate mimetics for the modification or interception of
specific key events in carbohydrate metabolism.
Unfortunately, when 3-benzyloxy-2-benzyloxymethyl-6-io-
dopyridine (5) was used as substrate for SNAr reactions, even
under harsh reaction conditions (NaH, NMP, 80 °C), formation
of the desired pyridine conjugate was not observed. Despite this
unreactivity, synthesis of methyl 2,3,6-tri-O-benzyl-4-O-(5A-
benzyloxy-6A-benzyloxymethyl-2A-pyridyl)-a- -glucopyrano-
D
side (7b) was achieved via copper(I) halide-catalysed alkoxyla-
tion13 of 5. The synthesis of the equivalent N-linked conjugates
through straightforward SNAr reactions of methyl 2,3,6-tri-O-
benzyl-4-amino-4-deoxy-a-
-glucopyranoside11 (9; Scheme 3)
D
with 2-chloropyridine, 2-chloropyridine N-oxide, 2-bromopyr-
idine or 5 also proved unsuccessful, even in the presence of
copper(I) halide.
Grateful acknowledgement is made to A. W. R. Saunders for
microanalysis, the Mass Spectrometry Service Centre at
Swansea for high resolution mass spectra and the Com-
monwealth Scholarship Commission for an award to P. A. N.
Notes and references
‡ This procedure also led to the formation of 3-hydroxy-2-hydroxymethyl-
4,6-diiodopyridine which was removed from the cooled reaction mix by
precipitation and filtration following acidification (pH 6.0) through the
dropwise addition of HCl (2 M). The resultant crude 3-hydroxy-
2-hydroxymethyl-6-iodopyridine was then purified by successive re-
crystallization from ethanol. Although the spectroscopic and analytical data
for 3-hydroxy-2-hydroxymethyl-6-iodopyridine was in full agreement with
the proposed structure, the structure was confirmed by X-ray analysis of its
crystal structure.
Scheme 3 Reagents and conditions: i, 2-Bromopyridine or 5, Pd(OAc)2,
(±)-BINAP, NaOt-Bu, toluene, 100 °C under argon for 20 h, 10a = 22%,
10b = 29%; iii, EtOH–cyclohexene (2+1), Pd black, reflux, 48 h, 11a =
80%, 11b = 25%.
§ All pseudo-disaccharides were purified by silica-gel chromatography and
gave satisfactory elemental analysis or HRMS data. NMR spectra were in
full accord with the proposed structures.
However, the successful palladium-catalysed cross-coupling
of pyridyl bromides and primary amines employing chelating
bis-(phosphine) ligands has been reported.14 Utilising similar
reaction conditions15 we found it was possible to synthesize
both methyl 2,3,6-tri-O-benzyl-4-amino-4-deoxy-4-N-(2A-pyr-
1 M. L. Sinnot, Chem. Rev., 1990, 90, 1171.
2 B. Ganem, Acc. Chem. Res., 1996, 29, 340.
3 P. A. Fowler, A. H. Haines, R. J. K. Taylor, E. J. T. Chrystal and M. B.
Gravestock, J. Chem. Soc., Perkin 1, 1994, 2229.
4 R. Jiricek, J. Lehmann, B. Rob and M. Scheuring, Carbohydr. Res.,
1993, 250, 31.
5 J. Lehmann, B. Rob and H.-A. Wagenknecht, Carbohydr. Res., 1995,
278, 167.
6 S. Knapp, Y. Dong, K. Rupitz and S. G. Withers, Bioorg. Med. Chem.
Lett., 1995, 5, 763.
7 B. Müller, C. Schaub and R. R. Schmidt, Angew. Chem., Int. Ed., 1998,
37, 1433.
8 P. A. Nkansah, D. Hall, D. Leeson, A. H. Haines and N. P. J. Stamford,
in preparation.
9 V. Koch and S. Schnatterer, Synthesis, 1990, 6, 497.
10 J. J. Patroni, R. V. Stick, B. W. Skelton and A. H. White, Aust. J. Chem.,
1988, 41, 91.
11 Y. Kobayashi and M. Shiozaki, J. Org. Chem., 1995, 60, 2570.
12 M. Ek, P. J. Garegg, H. Hultberg and S. Oscarson, J. Carbohydr. Chem.,
1983, 2, 305.
13 M. A. Keegstra, T. H. A. Peters and L. Brandsma, Tetrahedron, 1992,
48, 3633.
14 S. Wagaw and S. L. Buchwald, J. Org. Chem., 1996, 61, 7240.
15 J. P. Wolf and S. L. Buchwald, J. Org. Chem., 2000, 65, 1158.
16 G. M. Anantharamaiah and K. M. Sivanandaiah, J. Chem. Soc., Perkin
Trans. 1, 1977, 490.
17 M. M. Palcic, T. Skrydstrup, K. Bock, N. Le and R. U. Lemieux,
Carbohydr. Res., 1993, 250, 87.
18 B. Svensson and M. R. Sierks, Carbohydr. Res., 1992, 227, 29.
19 E. M. S. Harris, A. E. Aleshin, L. M. Firsov and R. B. Honzatko,
Biochemistry, 1993, 32, 1618; A. E. Aleshin, L. M. Firsov and R. B.
Honzatko, J. Biol. Chem., 1994, 269, 15631.
20 G. Legler, Biochim. Biophys. Acta., 1978, 524, 94; E. Truscheit, W.
Frommer, B. Junge, L. Müller, D. D. Schmidt and W. Wingender,
Angew. Chem., Int. Ed. Engl., 1981, 20, 744; J. S. Andrews, T. Weimar,
T. P. Frandsen, B. Svensson and B. M. Pinto, J. Am. Chem. Soc., 1995,
117, 10799.
idyl)-a-
zyl-4-amino-4-deoxy-4-N-(5A-benzyloxy-6A-benzyloxymethyl-
2A-pyridyl)-a- -glucopyranoside (10b) from methyl
2,3,6-tri-O-benzyl-4-amino-4-deoxy-a- -glucopyranoside (9)
D-glucopyranoside (10a) and methyl 2,3,6-tri-O-ben-
D
D
and the appropriate pyridyl halide (Scheme 3). Finally, de-O-
benzylation for all benzyl protected pseudo-disaccharides
synthesized (7a, 7b, 10a and 10b) by catalytic transfer
hydrogenation (cyclohexene–Pd black–EtOH)16 at reflux tem-
perature for 24–48 h afforded the desired O- and N-linked
pseudo-disaccharides.§
The biological activity of the prepared pyridylglucoconju-
gates 8a, 8b, 11a and 11b was then examined using a
glucoamylase (GA)-dependent reaction.17 The observed results
(Table 1) clearly define for the first time the potential of these
non-carbohydrate carbohydrate mimetics as potent moderators
of glycoprocessing events. Indeed, the best inhibitor (11b) is
comparable in biological activity to the acarvosine unit of the
potent GA inhibitor acarbose (Ki = 1 3 1026 M).18
What is most strikingly evident with these simple pseudo-
disaccharides is that the pendant polyvalent carbohydrate ligand
(e.g. in 11a) clearly overcomes the bonding limitations of the
simple transition state mimetic (e.g. 3) and confers additional
bonding interactions in the enzyme active site that dramatically
increases their inhibitory effectiveness. It should be noted that
the X-ray crystal structures of GA with inhibitor bound in the
active site clearly show strong bonding interactions between
key residues at the catalytic subsite and the C-4 and C-6
hydroxyls of substrate analogues.19 The associated improve-
ment in the biological activity of those pseudo-disaccharide
inhibitors with peripheral hydroxylation on the flat aromatic
CHEM. COMMUN., 2003, 784–785
785