Glycosyl azide - A novel substrate for enzymatic transglycosylations

Article abstract of DOI:10.1016/j.tetlet.2005.10.040

Glycosyl azides are new efficient donors for glycosidases. Their high water solubility facilitates transglycosylations with comparable or better yields than common O-glycosides. The azido group totally changes the β-GalNAc-ase/β-GlcNAc-ase ratio in β-N-ac

Full text of DOI:10.1016/j.tetlet.2005.10.040

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 8715–8718
Glycosyl azide—a novel substrate for enzymatic transglycosylations
a
b
b
c
a
´
Pavla Fialova, Ana T. Carmona, Inmaculada Robina, Rudiger Ettrich, Petr Sedmera,
¨
a
a
a,
*
´
´
´
´
´
´
Veˇra Prˇikrylova, Lucie Petraskova-Husˇakova and Vladimır Krˇen
aInstitute of Microbiology, Academy of Sciences of the Czech Republic, Vı´denˇska´ 1083, CZ-142 20 Prague 4, Czech Republic
bDepartment of Organic Chemistry, Faculty of Chemistry, University of Seville, Aptdo Correros 553, E-41071 Seville, Spain
cLaboratory of High Performance Computing, Institute of Physical Biology USB and Institute of Systems Biology and Ecology AS CR,
Za´mek136, CZ-373 33 Nove´ Hrady, Czech Republic
Received 1 September 2005; revised 4 October 2005; accepted 12 October 2005
Abstract—Glycosyl azides are new efficient donors for glycosidases. Their high water solubility facilitates transglycosylations with
comparable or better yields than common O-glycosides. The azido group totally changes the b-GalNAc-ase/b-GlcNAc-ase ratio in
b-N-acetylhexosaminidases (from the usual 0.3–1.0 to <0.01), contrary to all known aglycons.
Ó 2005 Elsevier Ltd. All rights reserved.
Enzymatic transglycosylation catalyzed by glycosidases
is a respected method in carbohydrate synthesis.1 The
spectrum of acceptors is practically infinite, contrary
to glycosyl donors, which are usually nitrophenyl glyco-
sides.2 Their major advantage is good reactivity, how-
ever, their low water solubility causes difficulties as
high concentrations of substrates promote transglycosyl-
ation at the expense of donor hydrolysis. This is espe-
cially the case with modified glycosides.3 Furthermore,
commercial nitrophenyl glycosides are expensive and
the yields of their syntheses may be unsatisfactory.4
Therefore, alternative glycosyl donors are emerging,
such as 3-nitro- and 5-nitro-2-pyridyl glycosides,5 vinyl
glycosides6 and glycosyl fluorides.7
Here, we tested 2-acetamido-2-deoxy-b-D-glucopyrano-
syl azide8a (1) and 2-acetamido-2-deoxy-b-D-galacto-
pyranosyl azide8b (2) as glycosyl donors for a wide
range of fungal b-N-acetylhexosaminidases from our
enzymatic library.10 The cleavage and transglycosyla-
tion potentials of these substrates were examined and
the experimental data were compared with the results
from molecular modelling. Three novel oligosaccharides
were synthesized by transglycosylation with compound
1 as the glycosyl donor.
Substrates 1 and 2 were screened for cleavage by 20 b-N-
acetylhexosaminidases selected on the basis of our previ-
ous chemoenzymatic studies3b with modified substrates.
The results were compared to the standard substrates,
4-nitrophenyl 2-acetamido-2-deoxy-b-D-glucopyrano-
side (3, Table 1, for its structure see Scheme 2: R =
4-nitrophenyl) and 4-nitrophenyl 2-acetamido-2-deoxy-
b-D-galactopyranoside (4, for its structure see Scheme
2: R = 4-nitrophenyl). Due to the lackof a suitable
chromophore in the glycosyl azides, the standard hydro-
lytic activity determination was no longer applicable11
and, therefore, a new method of HPLC determination
was developed (Polymer IEX H+ column).
Glycosyl azides are easily prepared in high yield8 and are
highly water soluble (e.g., azide 1 is ca. 60 times better
soluble than 4-nitrophenyl glycoside 3). Moreover, the
azide ion is easily removable, which strongly facilitates
the purification of transglycosylation reaction mixtures.
The only hint on cleavage of one of these substrates was
published by Day and Withers (b-glucosidase)9 and to
the best of our knowledge, they have never been used
as glycosyl donors in transglycosylation reactions.
Substrate 1 (gluco-) was accepted by all the tested b-N-
acetylhexosaminidases whereas substrate 2 (galacto-)
was not hydrolyzed by any of the enzymes. It is a rather
atypical phenomenon since the tolerance of b-N-acetyl-
hexosaminidases towards the change of configuration
at C-4 is one of their typical features.12 This unexpected
Keywords: b-N-Acetylhexosaminidase; Enzyme catalysis; Glycosyl
azide; Molecular modelling.
*
Corresponding author. Tel.: +42 0296 442510; fax: +42 0296
442509; e-mail: kren@biomed.cas.cz
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.10.040

8716
P. Fialova´ et al. / Tetrahedron Letters 46 (2005) 8715–8718
Table 1. Screening for the hydrolysis of substrate 1 by fungal b-N-
acetylhexosaminidases
Source of enzyme
1a
Acremonium persicinum CCF 1850
Aspergillus awamori CCF 763
A. flavipes CCF 1895
+
+
+
A. flavus CCF 3056
+
A. oryzae CCF 1066
A. parasiticus CCF 1298
A. sojae CCF 3060
++
+
++
++
++
+
++
+++
+++
++
++
++
++
+++
++
++
A. terreus CCF 2539
Fusarium oxysporum CCF 377
Hamigera avellanea CCF 2923
Penicillium oxalicum CCF 2315
P. chrysogenum CCF 1269
P. funiculosum CCF 2985
P. multicolor CCF 2244
P. brasilianum CCF 2155
P. pittii CCF 2277
Figure 1. Substrates 1 (A), 2 (B), 3 (C) and 4 (D) docked into the active
site of the b-N-acetylhexosaminidase from A. oryzae CCF 1066.
Hydrogen bonding (in green) to glutamic acid 519, arginine 193 and
tryptophan 482 fixes the glycosides. The azido group (substrates 1 and
2) is depicted as a blue stick.
P. spinulosum CCF 2159
Talaromyces flavus CCF 2686
T. ohiensis CCF 2229
Trichoderma harzianum CCF 2687
a Enzymes were classified according to the ratio of specific activities
(U mLÀ1) towards substrate 1 and the standard substrate 3: 16–11%
(+++), 10–5% (++) or 4–1% (+); CCF: culture collection of fungi.
(Fig. 1) shows rather similar total interaction energies
for
both
resonance
structures
(À266
and
À257 kJ molÀ1), which are comparable to the standard
substrate 3 (À300 kJ molÀ1). The minor difference
is mainly caused by the steric contribution. On the
contrary, docking of both resonance structures of the
galacto-substrate 2 exhibited a more significant drop in
total interaction energies (À167 and À164 kJ molÀ1
compared to À221 kJ molÀ1 for the standard substrate
4), which is essentially caused by the difference in elec-
trostatic energy contributions. A closer lookat the bind-
ing pocket arrangement shows that the hydrogen
bonding with the carboxyl group of glutamic acid 519
with the standard substrate 4 cannot occur with sub-
strate 2 due to the distance between the respective moi-
eties of the latter and the protein amino acids being too
large (Fig. 1). These results strongly suggest that sub-
strate 2 is not bound into the enzymatic active site due
to the lackof hydrogen bonding, as reflected by the elec-
trostatic energy contribution (in this case the dominat-
ing interaction). This is in accordance with
experimental data (i.e., no cleavage).
result could be explained by molecular modelling: sub-
strates 1 and 2 were docked into the active site of the
b-N-acetylhexosaminidase from Aspergillus oryzae CCF
1066 (a typical representative of eukaryotic b-N-acetyl-
hexosaminidases) and the enzyme–substrate complex
interaction energies were compared (Table 2).
Both resonance structures of the azido group
(–N@N+@NÀ and –NÀ–N+„N) were considered in
the calculations. The real structure is a hybrid of the
two. In the case of the gluco-substrate 1, the docking
Table 2. Docking of 4-nitrophenyl- and azido-hexosaminides into the
active site of b-N-acetylhexosaminidase from Aspergillus oryzae CCF
1066
Substratea
Steric
contribution
(kJ molÀ1
Electrostatic
contribution
(kJ molÀ1
)
Total interaction
energyb (kJ molÀ1
)
)
The hydrolysis of substrate 1 by b-N-acetylhexosaminid-
ases from Aspergillus oryzae CCF 1066 (cleavage rate
7% related to 3) and from Talaromyces flavus CCF
2686 (the best one from the screening, cleavage rate
16% related to 3) was characterized by basic kinetic con-
stants (Km, Vmax; Table 3).
1-A
1-B
2-A
2-B
3
À29
À57
À72
À62
À84
À66
À237
À200
À95
À266
À257
À167
À164
À300
À221
À102
À216
À155
4
a A: Azide calculated as –N@N+@NÀ; B: azide calculated as –NÀ–
N+„N.
Results from the Michaelis–Menten plot that both
tested enzymes were inhibited at higher concentrations
of the standard substrate 3 were evident (Ki, Table 3),
b The drop in the enzyme–substrate complex interaction energy reflects
improved binding of the substrate into the enzymatic active site.
Table 3. Hydrolysis of substrates 1 and 3 by selected enzymes
Enzyme source
Azido substrate 1
Vmax (lM minÀ1
4-Nitrophenyl substrate 3
Km (mM)
)
Ki (mM)
Km (mM)
Vmax (lM minÀ1
)
Ki (mM)
A. oryzae CCF 1066
T. flavus CCF 2686
3.1 0.4
0.75 0.09
3.4 0.2
4.4 0.1
0
0
0.75 0.05
0.36 0.04
56
67
2
4
7.0 0.6
1.3 0.1

P. Fialova´ et al. / Tetrahedron Letters 46 (2005) 8715–8718
8717
OH
AcHN
O
OH
OH
O
OH
OH
OH
β-N-Acetylhexosaminidase
O
O
O
O
(T. flavus CCF 2686)
HO
HO
HO
HO
HO
HO
O
N3
N3
N3 +
HO
pH 5.0, 35 ˚C
NHAc
NHAc
NHAc
NHAc
5
6
1
yield 32%
yield 16%
Scheme 1. Enzymatic synthesis of compounds 5 and 6.
OH
OH
AcHN
O
OH
HO
OH
O
OH
OH
β-N-Acetylhexosaminidase
O
O
O
OH
(T. flavus CCF 2686)
HO
HO
R
R
R
+
HO
pH 5.0, 35 ˚C
NHAc
NHAc
NHAc
1
2
7
yield 22%
Scheme 2. Enzymatic synthesis of compound 7, R = N3.
whereas no inhibition was observed with substrate 1.
This fact, together with the substantially improved
water solubility (saturated water solution is 1.1 M for
1 and 0.02 M for 3), enables very efficient transglycosyl-
ation reactions with a low enzyme consumption using
substrate 1 as the glycosyl donor.
Transfer of the b-D-GlcpNAc moiety to acceptor 2 (not
hydrolyzed) catalyzed by b-N-acetylhexosaminidase
from T. flavus CCF 2686 afforded disaccharide 7 in
22% yield (Scheme 2). For experimental data on all
the compounds prepared, see the Supplementary data.
In summary, glycosyl azide 1 proved to be an efficient
and easily preparable glycosyl donor for b-N-acetyl-
hexosaminidase-catalyzed transglycosylation, as demon-
strated by the preparation of three novel disaccharides
5, 6 and 7. Glycosyl azides can serve as interesting alter-
natives to traditional nitrophenyl glycosides.
Glycosidases, which are typically retaining enzymes with
O-glycosides,1a are inverting towards some unusual sub-
strates like glycosyl fluorides (C–F bond).7 Therefore,
the hydrolysis of substrate 1 (C–N bond) was monitored
1
by H NMR to determine the anomeric configuration
of the originating D-GlcpNAc. Since the signal of H-
1b in D-GlcpNAc is hidden under the water resonance,
D-GlcpNAc signals for H-1a (d = 5.01, d, J = 3.5 Hz)
and H-2b (d = 3.48, dd, J = 10.2, 8.4 Hz) were com-
pared. The concentration of b-D-GlcpNAc increases fas-
ter than that of its a-anomer, which is formed from the
b-anomer in water due to mutarotation (Fig. 1 in the
Supplementary data).13 Substrate 1 remained stable
under the conditions. As a result, the b-N-acetylhexos-
aminidase from Aspergillus oryzae CCF 1066 hydro-
lyzed substrate 1 as a retaining enzyme, analogously to
O-glycosides.
Acknowledgements
ˇ
ˇ
This workwas supported by grants FRVS MSMT
No. 1862/04, Czech National Science Foundation No.
203/05/0172, Research Concepts Nos. AV0Z50200510,
MSM 0021620808 and MSM 6007665808, the Ministe-
rio de Educacion y Ciencia (Spain) (Grant, CTQ2004-
00649BQU) and by COST Chemistry D25/0001/01
´
ˇ
(MSMT OC D25.002).
The three best b-N-acetylhexosaminidases for hydrolysis
(Table 1) were tested for their ability to transfer the 2-
acetamido-2-deoxy-b-D-glucopyranosyl moiety from
substrate 1 to different acceptors. The same transglyco-
sylation products were detected (HPLC) in all three
cases. The enzyme from T. flavus CCF 2686 exhibited
the best yield of transglycosylation product and the
highest specific activity. Therefore, it was used for
semi-preparative transglycosylation reactions. Autocon-
densation reactions catalyzed by this enzyme yielded
two isomeric disaccharides 5 and 6 (Scheme 1), which
were separated by preparative HPLC. Compound 5 is
a p-nitrophenyl b-chitobioside14 analogue, which can
be easily conjugated to complex biological structures,
such as immunoactive glycoconjugates,15 after reduction
of the C-1 azido group to an amine.
Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.tetlet.
2005.10.040.
References and notes
1. (a) Krˇen, V.; Thiem, J. Chem. Soc. Rev. 1997, 26, 463–473;
(b) Vocˇadlo, D. J.; Withers, S. G. In Carbohydrates in
Chemistry and Biology; Ernst, B., Hart, G. W., Sinay, P.,
¨
Eds.; Wiley-VCH: Weinheim, 2000; Vol. 2, pp 723–844.
2. (a) Giordano, A.; Andreotti, G.; Mollo, E.; Trincone, A.
J. Mol. Catal. B: Enzym. 2004, 30, 51–59; (b) Monsan, P.;
Paul, F. FEMS Microbiol. Rev. 1995, 16, 187–192.

8718
P. Fialova´ et al. / Tetrahedron Letters 46 (2005) 8715–8718
´
´
3. (a) Husˇakova, L.; Riva, S.; Casali, M.; Nicotra, S.; Kuzma,
9. Day, A. G.; Withers, S. G. Biochem. Cell Biol. 1986, 64,
´
ˇ
M.; Hunˇkova, Z.; Kren, V. Carbohydr. Res. 2001, 331,
914–922.
ˇ
´
´
´
ˇ
´
ˇ
ˇ
´
´
143–148; (b) Fialova, P.; Weignerova, L.; Rauvolfova, J.;
10. Hunkova, Z.; Kren, V.; Scigelova, M.; Weignerova, L.;
Scheel, O.; Thiem, J. Biotechnol. Lett. 1996, 18, 725–730.
11. Mega, T.; Ikenaka, T.; Matsushima, Y. J. Biochem. 1970,
68, 109–117.
´
ˇ
´
Prˇikrylova, V.; Pisvejcova, A.; Ettrich, R.; Kuzma, M.;
Sedmera, P.; Krˇen, V. Tetrahedron 2004, 60, 693–701.
4. (a) Watt, D. K.; Clinch, K.; Slim, G. C. Carbohydr. Res.
12. Weignerova, L.; Vavruskova, P.; Pisvejcova, A.; Thiem, J.;
ˇ ˇ
´ ´ ´
´
2002, 337, 1235–1238; (b) Krist, P.; Kuzma, M.; Pelyvas,
´
ˇ
ˇ
I. F.; Simerska, P.; Kren, V. Collect. Czech. Chem.
Kren, V. Carbohydr. Res. 2003, 338, 1003–1008.
13. Horton, D.; Jewell, J. S.; Philips, K. D. J. Org. Chem.
1966, 31, 4022–4025.
Commun. 2003, 68, 801–811.
5. Yasukochi, T.; Fukase, K.; Kusumoto, S. Tetrahedron
Lett. 1999, 40, 6591–6593.
´
14. Kubisch, J.; Weignerova, L.; Ko¨tter, S.; Lindhorst, T. K.;
6. Chiffoleau-Giraud, V.; Spangenberg, P.; Rabiller, C.
Tetrahedron: Asymmetry 1997, 8, 2017–2023.
7. Williams, S. J.; Withers, S. G. Carbohydr. Res. 2000, 327,
27–46.
8. (a) Shulman, M. L.; Lakhtina, O. E.; Khorlin, A. Y.
Biochem. Biophys. Res. Commun. 1977, 74, 455–459;
Sedmera, P.; Krˇen, V. J. Carbohydr. Chem. 1999, 18, 975–
984.
´
´
15. (a) Krist, P.; Herkommerova-Rajnochova, E.; Rau-
´
ˇ
ˇ
´
´ˇ
volfova, J.; Semenuk, T.; Vavruskova, P.; Pavlıcek, J.;
Bezousˇka, K.; Petrusˇ, L.; Krˇen, V. Biochem. Biophys. Res.
Commun. 2001, 287, 11–20; (b) Semenˇuk, T.; Krist, P.;
´
´
ˇ
ˇ
(b) Szilagyi, L.; Gyo¨rgydeak , Z.Carbohydr. Res. 1985,
Bezouska, K.; Kuzma, M.; Halada, P.; Kren, V. Glyco-
143, 21–41.
conjugate J. 2001, 18, 817–826.