Enzymatic synthesis of alkyl arabinofuranosides using a thermostable α-L-arabinofuranosidase

Article abstract of DOI:10.1016/S0040-4039(02)02381-X

A thermostable α-L-arabinofuranosidase was tested for its ability to perform transglycosylation with different alcohol acceptors. Reactions were characterized by high rates with optimal synthesis being obtained within 10 min. Both primary and secondary al

Full text of DOI:10.1016/S0040-4039(02)02381-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 43 (2002) 9653–9655
Enzymatic synthesis of alkyl arabinofuranosides using a
thermostable a- -arabinofuranosidase
L
Caroline Re´mond,a Mounir Ferchichi,a Nathalie Aubry,a Richard Plantier-Royon,b Charles Portellab
and Michael J. O’Donohuea,*
aInstitut National de la Recherche Agronomique, UMR FARE, 8, rue Gabriel Voisin, BP 316, 51688 Reims Cedex 2, France
bLaboratoire Re´actions Se´lectives et Applications, Associe´ au CNRS (UMR 6519), Universite´ de Reims, Faculte´ des Sciences,
BP 1039, 51687 Reims Cedex 2, France
Received 11 October 2002; revised 15 October 2002; accepted 21 October 2002
Abstract—A thermostable a-L-arabinofuranosidase was tested for its ability to perform transglycosylation with different alcohol
acceptors. Reactions were characterized by high rates with optimal synthesis being obtained within 10 min. Both primary and
secondary alcohols could act as acceptors in transarabinosylation but yields of alkyl arabinosides decreased with increasing alkyl
chain length. © 2002 Elsevier Science Ltd. All rights reserved.
Enzymatic methods for the organic syntheses of sugar-
based compounds have been increasingly developed
and adopted over the last decade. This is because
enzymes present several advantages when compared to
traditional catalysts: high anomeric specificity, mild
reaction conditions, no protection/deprotection steps of
the reactive hydroxyl groups.1
been restricted due to the limited availability of these
enzymes and the need for costly substrates.1 In con-
trast, glycosyl hydrolases are abundant and extensively
used. These enzymes, which normally cleave glycosidic
bonds, can be employed as catalysts under appropriate
conditions to synthesize oligosaccharides2–6 and a vari-
ety of glycoconjugates, including alkylated glycosides.1
The formation of glycosidic bonds is either catalysed by
glycosyl transferases (EC 2.4) or glycosyl hydrolases
(EC 3.2.1). Until recently, the use of the former has
Although a large number of glycosyl hydrolases with
different specificities are known to perform transglyco-
sylation,1,7–9 the majority of reports have described the
Scheme 1.
Keywords: glycosidation; enzyme reactions; carbohydrates; alcohols.
* Corresponding author. Tel.: +33 3 26 35 53 65; fax: +33 3 26 35 53 69; e-mail: michael.odonohue@univ-reims.fr
0040-4039/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(02)02381-X

9654
C. Re´mond et al. / Tetrahedron Letters 43 (2002) 9653–9655
Scheme 2.
use of gluco- and galactosidases for transglucosylation
or transgalactosylation, respectively, in the presence of
various nucleophile acceptors. This tendency reflects the
high availability and the wide-ranging applications of
glucosides and galactosides. In contrast, fewer reports
have described enzymatic catalysts for C-5 sugar trans-
formations, despite the recent development of medi-
alcohol used, the maximum yield was obtained after 10
min of incubation at 60°C. Longer incubation times led
to the gradual AbfD3-catalyzed hydrolysis of the reac-
tion products. This reaction rate is very fast compared
to many of the other glycosyl hydrolase-catalyzed reac-
tions described in the literature, although other exam-
ples of fast reactions have been reported.18,19 Once
maximum product yield was achieved, the enzyme was
inactivated to avoid hydrolysis (100°C, 10 min) and
preparative TLC (AcOEt/AcOH/H2O: 7/2/2) was per-
formed to facilitate reaction product purification. Accu-
rate quantification of product yield was achieved using
reverse phase HPLC (equipped with a dynamic light
scattering detector and a C-18 column from which
products were eluted using a gradually increasing gradi-
ent of acetonitrile in water: 2–60% v/v during 10 min,
flow rate 1 mL/min). Table 1 shows the maximum
yields obtained with several different alcohols. Propan-
2-ol was the least reactive acceptor (24%), while
methanol and ethanol gave the highest yields (54%). Up
to a chain length of four carbons (with the exception of
propan-2-ol), a very gradual decrease in yield was
observed. However, in the presence of n-pentanol no
product was detected, even when the reaction was
performed in the presence of a variety of co-solvents.
cally-relevant
letter, we report the synthesis of alkyl
nosides using a purified thermostable transglycosylating
a- -arabinofuranosidase (EC 3.2.1.55).
L
-arabinose-based compounds.10 In this
-arabinofura-
L
L
The reactions between p-nitrophenyl-a-L-arabinofura-
noside (pNP-ara) and various primary and secondary
alcohols are illustrated in Scheme 1.
In order to perform enzyme-catalyzed reactions, recom-
binant arabinofuranosidase (AbfD3) was first expressed
in Escherichia coli and purified as previously
described.11,12 Although the synthesis of pNP-ara was
previously described,13,14 we carried out a multi-gram
scale synthesis according to methods reported for other
monosaccharides. Methyl
first benzoylated and then transformed in 1-O-acetyl-
L
-arabinofuranoside15 was
2,3,5-tri-O-benzoyl-L
-arabinofuranose.16 Glycosylation
with p-nitrophenol was achieved17 and the benzoyl
groups were removed with dilute sodium hydroxide to
yield pNP-ara in a 40% overall yield (Scheme 2).
Table 1. Alkyl arabinosides obtained with AbfD3 in the
presence of pNP-ara and alcohols
Transglycosylations were performed by incubating
pNP-ara (5 mM in water) and alcohol acceptors (23%,
v/v) in the presence of AbfD3 (59 IU) at 60°C. The
progress of reactions was monitored using TLC in
AcOEt/AcOH/H2O (7/2/2). In the absence of alcohol
acceptors, pNP-arabinobiose was formed indicating
that pNP-ara can act both as the donor and the accep-
tor (data not shown). In contrast, in the presence of
alcohols, only alkylated products were formed, suggest-
ing that, in the role of nucleophile acceptor, alcohols
outcompete both water and pNP-ara. In order to max-
imise product formation it was necessary to identify the
optimal reaction time. By varying incubation time it
was possible to demonstrate that, regardless of the
Acceptors
Productsa
Yieldsb (%)
Methanol
Ethanol
Methyl-a-
Ethyl-a-L-arabinofuranoside
L
-arabinofuranoside
54
54
n-Propanol
Propan-2-ol
n-Butanol
Butan-2-ol
n-Propyl-a-
i-Propyl-a-
n-Butyl-a-
i-Butyl-a-
L-arabinofuranoside 45
L
-arabinofuranoside
24
42
41
33
L
-arabinofuranoside
-arabinofuranoside
-arabinofuranoside
L
Allyl alcoholc Allyl-a-
L
n-Pentanol
None detected
Not applicable
a Structures resolved by 1H NMR analysis.
b Yields determined by HPLC and based on pNP-ara initially added.
c CH3CN (24%, v/v) was used as cosolvent.

C. Re´mond et al. / Tetrahedron Letters 43 (2002) 9653–9655
9655
While the influence of increasing carbon number on
References
transglycosylation has already been reported,20,21 the
sudden loss of activity in the presence of n-pentanol is
rather surprising and suggests that either we had not
succeededinidentifyingtheappropriateco-solventorthat
AbfD3 is rapidly inactivated in the presence of higher
alcohols.
1. van Rantwijk, F.; Woudenberg-van Oosterom, M.; Shel-
don, R. A. J. Mol. Catal. B: Enzym. 1999, 6, 511–532.
2. Nilsson, K. G. I. Carbohydr. Res. 1987, 167, 95–103.
3. Lopez, R.; Fernandez-Mayorolas, A. Tetrahedron Lett.
1992, 33, 5449–5452.
4. Binder, W. H.; Hanspeter, K.; Schmid, W. Tetrahedron
1994, 50, 10407–10418.
5. Toone, E. J.; Simon, E. S.; Bednarski, M. D.; Whitesides,
G. M. Tetrahedron 1989, 45, 5365–5422.
6. Williams, S. J.; Withers, S. G. Carbohydr. Res. 2000, 327,
27–46.
With the exception of ethane-1,2-diol which gave a 20%
mixed product yield, all the alkylated reactions products
were successfully purified and analysed using H NMR
1
spectroscopy. Inallcases, onlythecorresponding a-L-ara-
binofuranosides were detected, witnessed by H-1 and H-2
coupling constants typical of a trans-relationship (J1,2 <2
Hz). This is coherent with the notion that the use of
glycosidases leads to very tight anomeric selectivity.
Moreover, these results confirm the findings of previous
work on the hydrolytic mechanism of F-51 arabinofura-
nosidases,12,22 which suggested that hydrolysis proceeds
via a retention mechanism in which the anomeric configu-
ration is conserved. Additionally, AbfD3 was able to
efficiently transfer the arabinofuranose moiety onto
secondary alcohols. Indeed, in the case of the butyl
alcohols, reaction yields were very similar (42 and 41%
for butan-1 and 2-ol, respectively). In the case of a chiral
secondary alcohol (butan-2-ol), formation of both
diastereomers was obtained in almost identical propor-
tions according to 1H NMR. This latter observation
indicatesthattheactionofAbfD3wasnotenantiospecific.
Although not all the glycosidases tested have been found
to be enantiospecific, various reports have indicated that
glycosidase enantioselectivity is highly variable and can
depend on the exact source of the enzyme. Likewise, in
the case of some b-galactosidases from E. coli preferential
transfer of galactosyl moieties to the (R)-enantiomers of
chiral alcohols was observed,23,24 whereas the b-galactosi-
dase from Sulfolobus solfataricus was highly selective for
the (S)-enantiomer in the case of the secondary hydroxyl
group of propane-1,2-diol.25
7. Vic, G.; Thomas, D. Tetrahedron Lett. 1992, 33, 4567–
4570.
8. Baker, A.; Turner, N. J.; Webberley, M. C. Tetrahedron:
Asymmetry 1994, 5, 2517–2522.
9. Basso, A.; Ducret, A.; Gardossi, L.; Lortie, R. Tetra-
hedron Lett. 2002, 43, 2005–2008.
10. Chu, C. K.; Hong, J. H.; Choi, Y.; Du, J.; Lee, K.; Chun,
B. K.; Boudinot, F. D.; Peek, S. F.; Korba, B. E.;
Tennant, B. C.; Cheng, Y.-C. Drug. Fut. 1998, 23, 821–
826.
11. Debeche, T.; Cummings, N.; Connerton, I.; Debeire, P.;
O’Donohue, M. J. Appl. Environ. Microbiol. 2000, 66,
1734–1736.
12. Debeche, T.; Bliard, C.; Debeire, P.; O’Donohue, M. J.
Prot. Eng. 2002, 15, 21–28.
13. Fielding, A. H.; Hough, L. Carbohydr. Res. 1965, 1,
327–329.
14. Kelly, M. A.; Sinnott, M. L.; Widdows, D. Carbohydr.
Res. 1988, 181, 262–266.
15. Schulze, O.; Voss, J.; Adiwidjaja, G. Synthesis 2001,
229–234.
16. Wang, Z.; Prudhomme, D. R.; Buck, J. R.; Park, M.;
Rizzo, C. J. J. Org. Chem. 2000, 65, 5969–5985.
17. Honma, K.; Nakazima, K.; Uematsu, T.; Hamada, A.
Chem. Pharm. Bull. 1976, 24, 394–399.
18. Ooi, Y.; Hashimoto, T.; Mitsuo, N.; Satoh, T. Chem.
Pharm. Bull. 1985, 33, 1808–1814.
19. Stevenson, D. E.; Stanley, R. A.; Furneaux, R. H. Bio-
technol. Bioeng. 1993, 42, 657–666.
20. Itoh, H.; Kamiyama, Y. J. Ferment. Bioeng. 1995, 80,
510–512.
21. Watt, D. K.; Ono, H.; Hayashi, K. Biochim. Biophys.
Acta 1998, 1385, 78–88.
22. Pitson, S. M.; Voragen, A. G. J.; Beldman, G. FEBS
Lett. 1996, 398, 7–11.
In conclusion, this preliminary survey of AbfD3 trans-
glycosylation activity has revealed that this enzyme may
be a useful tool for chemoenzymatic syntheses involving
L
-arabinose. Further experiments are now underway to
investigate the ability of AbfD3 to perform transglycosyl-
ation in the presence of other nucleophile acceptors,
especially sugars other than L-arabinose.
23. Crout, D. H. G.; MacManus, D. A.; Critchley, P. J.
Chem. Soc., Perkin Trans. 1 1990, 7, 1865–1868.
24. Matsumura, S.; Yamazaki, H.; Toshima, K. Biotechnol.
Lett. 1997, 19, 583–586.
25. Trincone, A.; Improta, R.; Nucci, R.; Rossi, M.; Gam-
bacorta, A. Biocatalysis 1994, 10, 195–210.
Acknowledgements
Financial support for this work was provided by the
‘Contrat de Plan Etat-Re´gion’, through the Glycoval
project. The authors thank F. Cherouvrier for the
synthesis of pNP-ara and H. Baillia for the NMR spectra.