Thioglycoligases: Mutant glycosidases for thioglycoside synthesis

Article abstract of DOI:10.1002/anie.200390114

Engineering of the active site of retaining glycosidases in conjunction with the appropriate modification of their substrates has led to the enzymatic synthesis of thioglycosidic linkages in thiooligosaccharides (see scheme).

Full text of DOI:10.1002/anie.200390114

Communications
Glycobiology
glycosidases that lack the catalytic nucleophile, glycosynthas-
es, in conjunction with activated donors of the opposite
23]
anomeric configuration to that of the natural substrate.[21
Thioglycoligases: Mutant Glycosidases for
Thioglycoside Synthesis**
Herein we present a novel strategy for the synthesis of S-
glycosidic linkages in oligosaccharides by the use of retaining
glycosidases that lack the catalytic acid/base amino acid
residue. The choice of appropriate donors and acceptors to be
used in such an approach becomes apparent by considering
the double displacement mechanism of retaining glycosidases
Michael Jahn, Jennifer Marles,
R. Antony J. Warren, and Stephen G. Withers*
Carbohydrate mimetics that are resistant towards enzy-
matic hydrolysis have proven to be useful as competitive
glycosidase inhibitors and therefore have potential as
therapeutics. Thioglycosides, in which the glycosidic
oxygen atom has been replaced by a sulfur atom, have
been especially valuable as stable glycoside analogues in
a range of studies of glycosidases, both as competitive
inhibitors, for example, for a-l-fucosidases,[1] pancreatic
a-amylase,[2] or for cellulases,[3] and recently in the
formation of stable complexes for X-ray crystallograph-
ic analysis, for example, with endoglucanase Cel7B from
Fusarium oxysporum,[4] maize b-glucosidase ZMGlu1,[5]
barley b-d-glucan glucohydrolase,[6] endoglucanase Ce-
l5A,[7] or E. coli maltodextrin phosphorylase (MalP),[8]
and they are gaining increasing interest as targets for the
pharmaceutical industry.[9]
Thioglycosides have been synthesized by glycosyla-
12]
tion of thioacceptors with activated glycosyl donors,[10
Scheme 1. Mechanism of a retaining glycosidase: Glycosylation and deglyco-
sylation with wild-type (a) and acid/base mutant (b). DNP¼dinitrophenyl,
Nu¼nucleophile.
by the SN2reaction of 1-thio sugars with activated
acceptors,[13] and by Michael-like addition of 1-thiosu-
gars to a,b-unsaturated systems;[14] all routes involve
numerous protection and deprotection steps and require
good control of anomeric stereochemistry.
The few naturally occurring thioglycosides belong to the
family of glucosinolates and are found in cruciferous plants.[15]
The enzymes that catalyze the formation of the thioglycosidic
linkages, S-glucosyltransferases, have been purified from
plant extracts,[16] as well as cloned into and expressed from
first proposed by Koshland (Scheme 1).[24] In the glycosyla-
tion step the concerted action of the catalytic acid/base
(protonated, acting as an acid) and the catalytic nucleophile
(deprotonated) leads to the departure of the aglycon group
and to the formation of the covalent glycosyl enzyme
intermediate (a). Use of a mutant glycosidase in which the
catalytic acid/base residue has been replaced by a noncatalyti-
cally active residue necessitates glycosyl donors with good
leaving groups that do not need acid catalysis, for example,
dinitrophenyl groups, to allow formation of the glycosyl en-
zyme intermediate (b). However, turnover of that intermedi-
ate via transglycosylation to an oxygen acceptor is impracti-
cally slow, since general base catalysis is required to speed this
step (a). The acid/base mutant therefore requires strong
nucleophiles as acceptors that do not need general base
catalysis (b).
E. coli.[17] There are a few examples of transfer to simple
20]
thiols.[18
However, no useful enzymatic syntheses of
thioglycosidic linkages in oligosaccharides in vitro have been
reported–such approaches would be valuable.
The action of retaining glycosidases on glycosides is in
most instances mediated by two key active site amino acid
residues, the catalytic nucleophile and the catalytic acid/base.
We have previously reported on the efficient synthesis of O-
glycosidic linkages in oligosaccharides by the use of retaining
[*] Prof. Dr. S. G. Withers, Dr. M. Jahn
Department of Chemistry
Earlier experiments revealed that small anionic mole-
cules, such as N3 , AcOꢀ, HCO2 , and Fꢀ,[26] rescue the
reaction by enhancing the rate of the deglycosylation step. In
the present study we exploit this concept to allow the
synthesis of thiooligosaccharides by use of deoxythio sugars
as nucleophilic acceptors that do not need base catalysis in the
deglycosylation step. In summary, our approach requires
activated donor glycosides, such as dinitrophenyl glycosides,
and chemically synthesized deoxythio sugars as acceptors in
conjunction with mutant glycosidases modified at the acid/
base position.
ꢀ
ꢀ [25]
University of British Columbia
2036 Main Mall, Vancouver, B.C. V6T 1Z1 (Canada)
Fax: (þ1)604-822-2847
E-mail: withers@chem.ubc.ca
J. Marles, Prof. Dr. R. A. J. Warren
Department of Microbiology
University of British Columbia (Canada)
[**] M. J. was financed by a DFG fellowship.
Supporting information for this article is available on the WWW
under http://www.angewandte.org or from the author.
352
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0352 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 3

Angewandte
Chemie
We have probed our strategy using the alanine acid/base
mutants of two retaining b-glycosidases, the b-glucosidase
from Agrobacterium sp. Abg E171A and the b-mannosidase
from Cellulomonas fimi Man2A E429A. The mutant Abg
E171A was generated by site-directed mutagenesis by a
™megaprimer∫ PCR method by using three oligonucleotide
primers, one containing the mutation, as described for the
Abg E358S mutant.[22] The purified product of the first PCR
reaction served as a megaprimer in the second PCR reaction.
The purified gene was subcloned into an expression vector,
and after expression the protein was purified in a single step
by Ni2þ-chelation chromatography. The mutant Man2A
E429A was the same protein sample described by Zechel
et al.[26]
As donors we chose the readily available dinitrophenyl
glycosides: 2,4-dinitrophenyl b-d-glucopyranoside (DNP-
Glc) for studies with the mutant glucosidase and 2,5-
dinitrophenyl b-d-mannopyranoside (DNP-Man) for experi-
ments involving the mutant mannosidase.[26] The low pKa
values of 3.96 for 2,4-dinitrophenol and 5.15 for 2,5-dinitro-
phenol render acid catalysis for the glycosylation step
unnecessary.[27] As an acceptor we initially chose para-nitro-
phenyl 4-deoxy-4-thio-b-d-glucopyranoside (1) since the
parent sugar para-nitrophenyl b-d-glucopyranoside (pNP-
Glc) acts as an excellent acceptor for both wild type and
nucleophile mutant forms of Abg[21] and Man2A[23] under-
going transfer preferentially to the 4-hydroxy group. The
deoxythio sugar 1 was readily synthesized by regioselective
protection of the galacto sugar, activation of the unprotected
axial alcohol by formation of the triflate, nucleophilic
substitution with thioacetate with inversion of configuration,
and finally Zemplen deprotection under exclusion of oxygen
(in the presence of DTT). Anaerobic conditions, even during
the enzymatic reactions, are compulsory to prevent oxidation
of the thiols to the corresponding disulfides.
was occurring involving specific thiolinkage formation. This
was confirmed by isolation of the products of the enzymatic
reactions by silica gel chromatography after acetylation.
1
Analysis of the products by H and 13C NMR spectroscopy
and mass spectrometry revealed that they were indeed sulfur-
linked disaccharides (Scheme 2).
These extremely encouraging results led us to question
whether the regiochemical outcome of the transglycosylation
reaction could be controlled by virtue of the location of the
thiol group within the acceptor sugar. We therefore synthe-
sized and tested para-nitrophenyl 4-deoxy-4-thio-b-d-xylo-
pyranoside (2) as an acceptor for the two enzymes. Previous
studies with both the wild-type Abg and its nucleophile
mutant had revealed that para-nitrophenyl b-d-xylopyrano-
side (pNP-Xyl) is an excellent acceptor substrate, but that
transfer, surprisingly, occurred exclusively to the 3-position.[21]
Use of 2 as an acceptor would reveal whether the greater
nucleophilicity of the thiol group controls the reaction
outcome. Indeed, incubation of 2 with Abg E171A and
DNP-Glc resulted in an excellent yield (79%) of the sulfur-
linked disaccharide, as revealed by 1H and 13C NMR spectro-
scopy and ESI-MS. Similarly, incubation of 2 with Man2A
E429A and DNP-Man produced, in 82% yield, the b(1 4)-
linked thiodisaccharide product (Scheme 2).
Interestingly, control reactions, in which the parent sugars
para-nitrophenyl b-d-xylopyranoside (pNP-Xyl) and para-
nitrophenyl b-d-glucopyranoside (pNP-Glc) were used as
acceptors for Man2A E429A in the presence of DNP-Man,
showed the formation of disaccharide products, but only very
slowly. This is in contrast to Abg E171A where no formation
of disaccharides was observed when using pNP-Glc or pNP-
Xyl in the presence of DNP-Glc as donor. To assess the
relative rates of transfer to the oxygen and sulfur nucleophiles
in the case of Man 2A E429A we established a competition
reaction in which equimolar amounts of pNP-Xyl and its 4-
thio analogue 2 were incubated together with DNP-Man as
donor and Man2A E429A (Figure 1). When the concentra-
tion of donor was limiting (one third of the concentration of
acceptors) we saw exclusive formation of the S-linked
disaccharide 6 (a). However, upon addition of further donor
to a final concentration twice that of the acceptors the O-
linked disaccharides (indicated by arrows) were formed (b).
Isolation and characterization after acetylation showed that
the b(1!3)- and b(1!4)-linked disaccharides were
formed.[28] Integration of HPLC peaks indicates that transfer
Upon incubation of 1 (20 mm) with the appropriate donor
(45 mm) and mutant enzyme (~ 1 mgmLꢀ1) in phosphate
buffered solutions at pH 6.8 TLC analysis revealed that a new
product with the expected mobility of a disaccharide was
formed quite rapidly in each case. These products were stable
towards hydrolysis by both the wild-type and mutant en-
zymes. By contrast, incubation of the wild-type enzymes with
the appropriate DNP sugar donors and thiosugar acceptor 1
resulted in no detectable formation of disaccharide. It was
therefore suspected that a novel enzymatic coupling reaction
Scheme 2. Enzymatic synthesis of thioglycosides 3 6. Experimental details and spectroscopic data are provided in the Supporting Information.
DNP¼dinitrophenyl, pNP¼para-nitrophenyl.
Angew. Chem. Int. Ed. 2003, 42, No. 3
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0353 $ 20.00+.50/0
353

Communications
Figure 1. HPLC traces of the Man2A E429A-catalyzed reaction of the two competing acceptors pNP-Xyl (5.5 mm) and its thio analogue 2 (5.5 mm)
with varying concentrations of DNP-Man: a) 3.75 mm, b ) 22 mm. Arrows indicate the elution of the b(1!3)- and b(1!4)-O-linked disaccharides,
6 is the S-linked disaccharide. Experimental details are provided in the Supporting Information.
to the thiosugar acceptor is at least 100-fold faster than
transfer to its oxygen analogue. These results were confirmed
by ESI-MS analysis.
[15] J. W. Fahey, A. T. Zalcmann, P. Talalay, Phytochemistry 2001, 56,
51.
[16] J. W. D. GrootWassink, D. W. Reed, A. D. Kolenovsky, Plant
Physiol. 1994, 105, 425 433.
5
The glycosynthases have proven to be powerful tools for
the synthesis of O-linked oligosaccharides. Herein we have
shown again that the concept of engineering glycosidases on
mechanistic grounds in conjunction with use of appropriately
modified substrates can lead to novel synthetic method-
ologies. We anticipate that the outlined approach is general,
as demonstrated with two different enzymes from two
separate glycosidase families, and will prove to be of
considerable value in the specific assembly of thiooligosac-
charides for mechanistic and possibly therapeutic use.
[17] E. F. Marillia, J.M. MacPherson, E. W. T. Tsang, K. Van
Audenhove, W. A. Keller, J. W. D. GrootWassink, Physiol. Plant.
2001, 113, 176 184.
[18] T. Dintinger, D. Dutheil-Bouedec, M. Bouchonneau, C. Tellier,
Biotechnol. Lett. 1994, 16, 689 692.
[19] G. Vic, D. H. G. Crout, Tetrahedron: Asymmetry 1994, 5, 2513
2516.
[20] J. B. Kempton, S. G. Withers, Biochemistry 1992, 31, 9961 9969.
[21] L. F. Mackenzie, Q. P. Wang, R. A. J. Warren, S. G. Withers, J.
Am. Chem. Soc. 1998, 120, 5583 5584.
[22] C. Mayer, D. L. Zechel, S. P. Reid, R. A. J. Warren, S. G. Withers,
FEBS Lett. 2000, 466, 40 44.
Received: September 13, 2002 [Z50153]
[23] O. Nashiru, D. L. Zechel, D. Stoll, T. Mohammadzadeh, R. A. J.
Warren, S. G. Withers, Angew. Chem. 2001, 113, 431 434;
Angew. Chem. Int. Ed. 2001, 40, 417 420.
[24] D. E. Koshland, Biol. Rev. 1953, 28, 416 436.
[25] Q. Wang, D. Trimbur, R. Graham, R. A. J. Warren, S. G. Withers,
Biochemistry 1995, 34, 14554 14562.
[26] D. L. Zechel, S. P. Reid, O. Nashiru, C. Mayer, D. Stoll, D. L.
Jakeman, R. A. J. Warren, S. G. Withers, J. Am. Chem. Soc. 2001,
123, 4350 4351.
[1] Z. J. Witczak, D. Boryczewski, Bioorg. Med. Chem. Lett. 1998, 8,
3265 3268.
[2] M. Blanc-Meusser, L. Vigne, H. Driguez, J. Lehmann, J. Steck,
K. Urbahns, Carbohydr. Res. 1992, 224, 59 71.
[3] C. Schou, G. Rasmussen, M. Schuelein, B. Henrissat, H. Driguez,
J. Carbohydr. Chem. 1993, 12, 743 752.
[4] G. Sulzenbacher, H. Driguez, B. Henrissat, M. Schuelein, G. J.
Davies, Biochemistry 1996, 35, 15280 15287.
[5] M. Czjzek, M. Cicek, V. Zamboni, W. P. Burmeister, D. R.
Bevan, B. Henrissat, A. Esen, Biochem. J. 2001, 354, 37 46.
[6] M. Hrmova, J. N. Varghese, R. De Gori, B. J. Smith, H. Driguez,
G. B. Fincher, Structure 2001, 9, 1005 1016.
[27] We used 2,5-dinitrophenyl b-d-mannopyranoside instead of 2,4-
dinitrophenyl b-d-mannopyranoside as the latter is impractically
unstable due to aryl migration. We also tested the appropriate
pNP-glycosides as donors, but prolonged reaction times make
those less attractive.
[28] The physical data of the compounds were identical to those
described by Nashiru et al.[23]
[7] A. Varrot, M. Schulein, S. Fruchard, H. Driguez, G. J. Davies,
Acta Crystallogr. Sect. D 2001, 57, 1739 1742.
[8] K. A. Watson, C. McCleverty, S. Geremia, S. Cottaz, H. Driguez,
L. N. Johnson, EMBO J. 1999, 18, 4619 4632.
[9] Z. J. Witczak, Curr. Med. Chem. 1999, 6, 165 178.
[10] J. S. Andrews, B. M. Pinto, Carbohydr. Res. 1995, 270, 51 62.
[11] L. X. Wang, N. Sakairi, H. Kuzuhara, J. Chem. Soc. Perkin Trans.
1 1990, 1677 1682.
[12] M. Blanc-Muesser, J. Defaye, H. Driguez, Carbohydr. Res. 1978,
67, 305 328.
[13] S. Fort, A. Varrot, M. Schulein, S. Cottaz, H. Driguez, G. J.
Davies, ChemBioChem 2001, 2, 319 325.
[14] Z. J. Witczak, J.M. Sun, R. Mielguj, Bioorg. Med. Chem. Lett.
1995, 5, 2169 2174.
354
¹ 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1433-7851/03/4203-0354 $ 20.00+.50/0
Angew. Chem. Int. Ed. 2003, 42, No. 3