Glycosynthases: Mutant glycosidases for oligosaccharide synthesis [5]

Full text of DOI:10.1021/ja980833d

J. Am. Chem. Soc. 1998, 120, 5583-5584
5583
Glycosynthases: Mutant Glycosidases for
Oligosaccharide Synthesis
Lloyd F. Mackenzie,† Qingping Wang,†
R. Antony J. Warren,‡ and Stephen G. Withers*,†
Protein Engineering Network of Centres of
Excellence and Departments of Chemistry
and Microbiology, UniVersity of British Columbia
VancouVer, B.C., Canada, V6T 1Z1
ReceiVed March 13, 1998
Efficient methods for the synthesis of oligosaccharides, par-
ticularly on the large scale, are unavailable despite the enormous
importance of this class of molecules in a range of biological
processes1 and their potential as new therapeutics.2 The synthetic
problems reside in the control of both the stereochemistry and
the regiochemistry of bond formation, requiring complex protec-
tion and activation strategies. While significant advances in
methodologies for chemical synthesis have been made in recent
years,3 particularly in the use of polymer-supported approaches,4
these remain complex and unsuitable for large-scale production.
The alternative is enzymatic synthesis, and glycosyl transferases
have indeed been used for this extensively in recent years, even
on a large scale.5 However, the poor availability of these enzymes
and the high substrate cost6 have limited their application.
Retaining glycosidases, run in the transglycosylation mode, are
the other choice.7 These enzymes function through the mecha-
nism shown in Figure 1 (lower path) in which a covalent
glycosyl-enzyme intermediate is formed and hydrolyzed with
general acid/base catalytic assistance.8 Synthesis can be achieved,
as shown in Figure 1 (upper path), by intercepting the reactive
glycosyl-enzyme intermediate with an added acceptor sugar.
While this approach offers the use of a cheap donor sugar9 and
relatively easily available enzymes, it has the substantial disad-
vantage that the product is necessarily a substrate for the enzyme
and is subsequently hydrolyzed, resulting in poor yields, unless
the equilibrium can somehow be displaced.10 We report here a
new approach to oligosaccharide synthesis involving the use of
a specifically mutated glycosidase (Glycosynthase) which, in
Figure 1. Mechanisms of hydrolysis and transglycosylation catalyzed
by Agrobacterium sp. â-glucosidase.
conjunction with activated glycosyl donors of the opposite
anomeric configuration to that of the normal substrate and suitable
acceptors, can efficiently synthesize oligosaccharides, but does
not hydrolyze them.
Earlier studies on â-glycosidases revealed11,12 that replacement
of the active site carboxylate nucleophile with a nonnucleophilic
amino acid side chain results in a correctly folded enzyme, which
is catalytically inactive13 since it cannot form the requisite
R-glycosyl-enzyme intermediate. However, since the rest of the
active site is intact14 it might be expected to catalyze the ligation
of an activated R-glycosyl derivative, bound at the active site in
place of the normal glycosyl-enzyme intermediate, to a suitable
acceptor sugar bound in the aglycon pocket. The oligosaccharide,
once formed, could not then be hydrolyzed by the enzyme. This
process would be assisted by general base catalysis from the
deprotonated acid/base residue.15 The Glu358Ala mutant of the
Agrobacterium sp. â-glucosidase/galactosidase (AbgGlu358Ala)
was chosen to test this approach because the wild-type enzyme
ordinarily catalyses efficient transglycosylation16 and the mutant
has no detectable hydrolytic activity.12 R-Glycosyl fluorides were
chosen as readily synthesized17 activated R-glycosyl donors with
† Department of Chemistry.
‡ Department of Microbiology.
(1) Kobata, A. Acc. Chem. Res. 1993, 26, 319. Varki, A. Glycobiology
1993, 3, 97.
(2) Wong, C.-H.; Halcomb, R. L.; Ichikawa, Y.; Kajimoto, T. Angew.
Chem., Int. Ed. Engl. 1995, 34, 521. Sears, P.; Wong, C.-H. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 12086. Zopf, D.; Roth, S. The Lancet 1996, 347, 1017.
Simon, P. M. Drug DiscoVery Today 1996, 1, 522.
(3) Danishefsky, S. J.; Bilodeau, M. Angew. Chem., Int. Ed. 1996 35, 1380.
Boons, G.-J. Tetrahedron 1996, 52, 1095.
(4) Douglas, S. P.; Whitfield, D. M.; Krepinsky, J. J. J. Am. Chem. Soc.
1991, 113, 3, 5095. Roberge, J. Y.; Beebe, X.; Danishefsky, S. J. Science
1995, 269, 202. Liang, R.; Yan, L.; Loebach, J.; Ge, M.; Uozumi, Y.; Sekanina,
K.; Horan, N.; Gildersleeve, J.; Thompson, C.; Smith, A.; Biswas, K.; Still,
W. C.; Kahne, D. Science 1996, 274, 1520.
(5) Nilsson, J. G. I. TIBTECH 1988, 6, 256. Thiem, J. FEMS Microbiol.
ReV. 1995, 16, 193.
(6) Complex, but in many cases effective, recycling schemes can be used
to regenerate NDPsugars in some cases (Ichikawa, Y.; Lin, Y.-C.; Dumas, D.
P.; Shen, G.-J.; Garcia-Junceda, E.; Williams, M. A.; Bayer, R.; Ketcham,
C.; Walker, L. E.; Paulson, J. C.; Wong, C.-H. J. Am. Chem. Soc. 1992, 114,
9283).
(7) Retaining glycosidases hydrolyze glycosides with net retention of
anomeric configuration. Synthesis can also be carried out under equilibrium
conditions, using high concentrations of the free sugars.
(8) McCarter, J. D.; Withers, S. G. Curr. Opin. Struct. Biol. 1994, 4, 885-
892. Sinnott, M. L. Chem. ReV. 1990, 90, 1171.
(9) Typical donors are nitrophenyl glycosides or glycosyl fluorides. Having
high kcat/Km values they are used by the enzyme in preference to the
accumulating products.
(11) Wakarchuk, W. W.; Campbell, R. L.; Sung, W. L.; Davoodi, J.;
Yaguchi, M. Protein Sci. 1994, 3, 467. Yuan, J.; Martinez-Bilbao, M.; Huber,
R. E. Biochem. J. 1994, 299, 527. Withers, S. G.; Rupitz, K.; Trimbur, D.;
Warren, R. A. J. Biochemistry 1992, 31, 9979.
(12) Wang, Q.; Graham, R. W.; Trimbur, D.; Warren, R. A. J.; Withers,
S. G. J. Am. Chem. Soc. 1994, 116, 11594.
(13) Utmost care must be exercised to avoid contamination with wild-type
enzyme including use of new column packings for mutant purifications. A
Glu358Gly mutant of the Agrobacterium sp. â-glucosidase was contaminated
(1 in 10 000) due to translational misincorporation.12
(14) Additional evidence for correct folding of this mutant resides in its
substantial activity as an inverting glycosidase if azide is added as an alternate
nucleophile.12
(15) Conversion of the nucleophilic Glu to Gln in the mechanistically
analogous Bacillus circulans xylanase dropped the pKa of the acid/base catalyst
2.5 units, to a value below the optimum pH (McIntosh, L. P.; Hand, G.;
Johnson, P. E.; Joshi, M. D.; Korner, M.; Plesniak, L. A.; Ziser, L.; Wakarchuk,
W. W.; Withers, S. G. Biochemistry 1996, 35, 9958).
(16) Kempton, J. B.; Withers, S. G. Biochemistry 1992, 31, 9961. Prade,
H.; Mackenzie, L. F.; Withers, S. G. Carbohydr. Res. 1998. In press.
(17) R-Glycosyl fluorides are synthesised by reaction of the per-O-acetylated
sugar with HF/pyridine followed by deprotection with methoxide in methanol.
(10) Equilibria have been displaced using a coupled enzyme to selectively
convert product formed by adsorption of product onto charcoal and by selective
crystallization of product or use of organic cosolvents.
S0002-7863(98)00833-6 CCC: $15.00 © 1998 American Chemical Society
Published on Web 05/23/1998

5584 J. Am. Chem. Soc., Vol. 120, No. 22, 1998
Communications to the Editor
Table 2. Glycosynthase-Catalyzed Transglycosylation Reactions
Using R-Glucosyl Fluoride as Donora
Figure 2. Proposed mechanism of Glycosynthase AbgGlu358Ala.
Table 1. Glycosynthase-Catalyzed Transglycosylation Reactions
Using R-Galactosyl Fluoride as Donor
a A hyphen (-) indicates reactions performed with 1-1.4 equiv and
an asterisk (*) indicates those with 2.2-3.0 equiv of glycosyl donor.
acceptor, resulting in formation of longer oligosaccharides. As
is shown in Table 2, use of approximately 1 equiv of R-glucosyl
fluoride yielded a mixture of di- and trisaccharide products since
disaccharides, once formed, are preferred acceptors for this
enzyme.21 However, when a disaccharide acceptor such as
p-nitrophenyl cellobioside was employed initially, reaction was
essentially stoicheometric, 1 equiv of R-glucosyl fluoride giving
a 79% yield of trisaccharide, while 2 equiv gave a 64% yield of
the tetrasaccharide directly [#15].
The utility of this Glycosynthase approach was well illustrated
by the gram-scale syntheses of several valuable substrates and
inactivators for cellulases as shown in Table 2. Methylumbel-
liferyl cellotrioside and cellotetraoside [#11a,b], chromogenic
substrates for cellulases, were readily synthesized from the easily
accessible monosaccharide glycosides by successive transfers from
R-glucosyl fluoride. Similarly, two mechanism-based inactivators
for cellulases, 2,4-dinitrophenyl 2-deoxy-2-fluoro-â-cellobioside
and -cellotrioside [#10, see also #3] were made, as an intentional
mixture, by reaction of 2,4-dinitrophenyl 2-deoxy-2-fluoro-â-
glucoside with 1.5 equiv of R-glucosyl fluoride in the presence
of the mutant. The latter synthesis would be impossible using
conventional glycosidase-catalyzed transglycosylation since the
2-deoxy-2-fluoro glycoside rapidly inactivates the wild-type
enzyme by trapping of the intermediate.
a small aglycon/leaving group.18 Acceptors chosen initially were
aryl glycosides, which are known to bind well to the aglycon
site16,19 but cannot be cleaved by the mutant. The proposed
reaction mechanism is shown in Figure 2.
The approach is illustrated with the synthesis of a trisaccharide
Gal-â(1,4)-Glc-â(1,4)-Glc-p-nitrophenyl. A 2 mL reaction mix-
ture containing R-galactosyl fluoride (14 mg, 40 mM) and
p-nitrophenyl â-cellobioside (28 mg, 30 mM) plus AbgGlu358Ala
(0.01 mg) in 150 mM ammonium bicarbonate buffer (to neutralize
released HF) was incubated for 8 h at room temperature, when
thin-layer chromatographic analysis (silica; ethyl acetate/methanol/
water, 7:2:1) indicated that reaction was complete. Workup
involved removal of the enzyme by ultrafiltration followed by
lyophilization of the volatile buffer and purification of the product
by HPLC to yield 34 mg (92%) of the trisaccharide product which
was characterized by NMR, MS, and elemental analysis.
The scope of this new enzymatic reaction was explored using
both R-galactosyl fluoride and R-glucosyl fluoride as donors in
the presence of a range of acceptors. Products from galactose
transfer are summarized in Table 1. The linkage formed was
â(1,4) in all cases shown with the exception of transfer to
â-xylosides, which occurred via a â(1,3) linkage [#2, #13].20
Transfer to an acceptor with an axial 4-hydroxyl group was very
inefficient, thus ensuring that only a single galactosyl transfer
occurred, the product itself being a poor acceptor. By contrast,
transfer of glucose yielded a product which was itself an excellent
This new Glycosynthase approach to oligosaccharide synthesis
has the twin advantages of utilizing inexpensive glycosyl donors
and of obviating hydrolysis of product. We believe this approach
has considerable potential as an alternative strategy for oligosac-
charide synthesis, particularly on a large scale.
(18) R-Glucosyl fluoride does not function as a substrate for, or even bind
to, the wild-type enzyme. The axial fluorine should also disfavor binding to
the aglycon site, decreasing competition between the donor and acceptor sugars.
(19) Day, A. G.; Withers, S. G. Can. J. Biochem. 1986, 64, 914.
(20) Such â(1,3) transfer to xylose has been seen with other glycosidases
(Lopez, R.; Fernandez-Mayoralas A. J. Org. Chem. 1994, 59, 737.).
(21) The yield of the disaccharide was maximized by adding only 0.5-
0.75 equiv of the glycosyl donor.
Acknowledgment. We thank the Protein Engineering Network of
Centres of Excellence of Canada and the Natural Sciences and Engineering
Research Council of Canada for financial support.
JA980833D