Regiospecific synthesis of lactose analog Gal-(β 1,4)-Xyl by transgalactosylation

Article abstract of DOI:10.1139/v02-106

A short enzymatic synthesis of disaccharide 4-O-β-D-galactopyranosyl-D-xylose (1) has been developed, which is of interest as a lactose analog for a non-invasive medicinal determination of lactose intolerance. The starting material, benzyl α-D-xyloside, was obtained by a Fischer-type glycosidation of D-xylose with benzyl alcohol, followed by anomeric differentiation of mixed glycosides using a glycosidase from Aspergillus oryzae. From several commercial β-galactosidases, which were screened for their transgalactosylation capacity, the enzyme from Escherichia coli was found to catalyze a virtually regio- and stereospecific galactosyl transfer from donor compounds o-nitrophenyl β-D-galactoside or lactose to the α-D-xyloside. Subsequent hydrogenolytic deprotection furnished desired disaccharide 1.

Full text of DOI:10.1139/v02-106

7
39
Re gios pe cific s ynthe s is of lactos e analog
Gal-( 1,4)-Xyl by trans galactos ylation
Wolf-Die te r Fe s s ne r a nd J ua n Ma nue l J uá re z Ruiz
Abstract: A short enzymatic synthesis of disaccharide 4-O-ꢀ-D-galactopyranosyl-D-xylose (1) has been developed,
which is of interest as a lactose analog for a non-invasive medicinal determination of lactose intolerance. The starting
material, benzyl ꢁ-D-xyloside, was obtained by a Fischer-type glycosidation of D-xylose with benzyl alcohol, followed
by anomeric differentiation of mixed glycosides using a glycosidase from Aspergillus oryzae. From several commercial
ꢀ
-galactosidases, which were screened for their transgalactosylation capacity, the enzyme from Escherichia coli was
found to catalyze a virtually regio- and stereospecific galactosyl transfer from donor compounds o-nitrophenyl ꢀ-D-
galactoside or lactose to the ꢁ-D-xyloside. Subsequent hydrogenolytic deprotection furnished desired disaccharide 1.
Key words: oligosaccharide synthesis, ꢀ-galactosidase, lactose intolerance.
Résumé : On a mis au point une courte synthèse du disaccharide 4-O-ꢀ-D-galactopyranosyl-D-xylose qui présente de
l’intérêt comme analogue de lactose pour procéder à une évaluation médicinale non invasive de l’intolérance au lactose.
On a obtenu le produit de départ, l’ꢁ-D-xyloside de benzyle, par une glycosylation de type Fischer du D-xylose par
l’alcool benzylique suivie d’une différentiation anomérique des glycosides mixtes à l’aide d’une glycosidase provenant
d’Aspergillus oryzae. Diverses ꢀ-galactosidases commerciales ont été évaluées pour leur capacité à provoquer une trans-
galactosylation; on a trouvé que l’enzyme provenant d’Escherichia coli catalyse le transfert du galactosyle d’une façon
virtuellement régio- et stéréospécique à partir de ꢀ-D-galactoside d’o-nitrophényle ou de lactose agissant comme compo-
sés donneurs vers l’ꢁ-D-xyloside. Une déprotection hydrogénolytique subséquente conduit au disaccharide 1 recherché.
Mots clés : oligosaccharide, synthèse, ꢀ-galactosidase, intolérance au lactose.
[
Traduit par la Rédaction] Fessner and Juárez Ruiz 742
Introduc tion
and the polysaccharide chain (glycosaminoglycan) in pro-
teoglycans of connective tissue (6).
Lactose is a milk disaccharide produced by nearly all
mammalian species. For digestion, it must be hydrolyzed to
D-galactose and D-glucose by lactase, an enzyme which is
present on the brush border of small intestinal enterocytes.
Evaluation of an enzyme deficiency (lactase insufficiency,
lactose intolerance) (1, 2) is important in pediatrics and
gastroenterology due to the high frequency of a genetic dis-
position. Because many of the standard diagnostic proce-
dures are invasive, quite drastic, and unacceptable or
dangerous for the testing of infants and young children, a
non-invasive method for lactase determination has been pro-
posed on the basis of 4-O-ꢀ-D-galactopyranosyl-D-xylose (1)
Chemical syntheses of 1 require a judicious installation of
protective groups and suffer from long, tedious reaction se-
quences with little overall productivity (7–9). For a choice of
alternative enzymatic strategies, the biosynthetic route in-
volving specific UDP-Gal dependant galactosyltransferase
enzymes (e.g., ꢀ4GalT7) (10) is burdened by technical diffi-
culties, high costs, and the ambident reactivity of xylose (11,
12). On the other hand, glycosidases have been shown by
numerous examples to allow the facile formation of
glycosides by transglycosylation at low cost (13, 14). Disad-
vantages of this approach, however, include yield limitations
due to competing hydrolysis reactions, and separation prob-
lems because of the formation of regioisomers and other by-
products with similar physical properties, which is due to the
low acceptor selectivity of glycosidases. For example, at-
tempts to form Gal-(ꢀ 1,4)-linkages by transglycosylation to
D-xylose (15, 16) or to different aryl- or alkyl-ꢀ-D-xylosides
(17–20) as acceptors were generally met by concomitant
(
3), which can be administered as a substrate analog to lac-
tose (4, 5). The xylose released from small quantities of this
disaccharide by lactase activity is eliminated in the urine
where it can be quantified colorimetrically. The ꢀ-Gal-(1,4)-
Xyl disaccharide (1) is also well known as a fragment of the
highly conserved linkage region between the core protein
Received 15 March 2002. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 5 June 2002.
Dedicated to J. Bryan Jones on the occasion of his 65th birthday in recognition of his pioneering contributions to preparative
biocatalysis.
1
W.-D. Fessner and J.M. Juárez Ruiz. Darmstadt University of Technology, Department of Organic Chemistry, Petersenstr. 22,
D-64287 Darmstadt, Germany.
1Corresponding author (e-mail: fessner@tu-darmstadt.de).
Can. J. Chem. 80: 739–742 (2002)
DOI: 10.1139/V02-106
© 2002 NRC Canada

7
40
Can. J. Chem. Vol. 80, 2002
formation of the corresponding (ꢀ 1,3)-disaccharides in sig-
nificant proportions.
Here we report a new enzymatic synthesis of 1 based on a
successful regiospecific transgalactosylation to ꢁ-configurated
benzyl D-xyloside.
Scheme 1. Chemoenzymatic synthesis of disaccharide 4. Re-
agents and conditions: (a) crude ꢀ-galactosidase from A. oryzae,
H2O, >60%; (b) ꢀ-galactosidase from E. coli, 23%; (c) Ac2O,
pyridine, 95%; (d) H2–Pd/C, aq. EtOH, 80%.
Re s ults a nd dis c us s ion
The synthetic strategy for constructing free-reducing
disaccharide 1 (for possible use in clinical diagnosis) called
for an anomeric protection group that ideally should confer
several important properties to the synthetic intermediates:
(
i) it needs to be easily introduced in the starting material at
low cost; (ii) it needs to be efficiently removed in the last
step with avoidance of toxic reagents; (iii) it should facilitate
monitoring of conversion and characterization at intermedi-
ate stages; (iv) it should simplify purification of the interme-
diates by a sufficiently high tendency to crystallize
(
particularly in view of a larger scale synthesis); and (v) it
should exert pronounced stereoelectronic effects for sub-
strate recognition during the transglycosylation process for
an enhanced regioselectivity. Thus, from a choice of allyl, 4-
pentenyl, and benzyl protection, the latter was given prefer-
ence because it was expected to best meet these criteria.
Recently, we have developed a simple and efficient proce-
dure for the anomeric differentiation of mixed ꢁ,ꢀ-glycosides
using the hydrolytic function of glycosidases (21). Such
anomeric mixtures are readily obtainable by acid-catalyzed
Fischer-type glycosidation. Because many glycosidases with
suitable substrate specificity and complementary anomeric
preference are readily available from various sources (14),
pure unprotected glycosides are efficiently accessible. Spe-
cifically, the ꢁ-configurated anomeric form of benzyl
xylopyranoside 1 (Scheme 1) can be isolated in high yield at
low cost using crude commercial enzyme preparations from
Aspergillus oryzae (21).
Since prior attempts for D-xylopyranoside transgalactosyl-
ation were mainly performed on the ꢀ-configurated benzyl
glycoside with unsatisfactory regioselectivity (17–20), ꢁ-
anomer 1 was chosen for our study. Several commercial ꢀ-
galactosidases (EC 3.2.1.23, e.g., from A. oryzae, E. coli,
and CloneZyme® library enzymes of undisclosed origin)
were screened for transgalactosylation activity in phosphate
buffer at pH 7.0 using 1 as the acceptor component and o-
nitrophenyl ꢀ-D-galactoside (o-NPG) as the donor compo-
nent. Because individual regioisomers were not revealed
upon TLC analyses using various solvents, crude product
mixtures were analyzed by NMR spectroscopy for composi-
tion. Practically all enzymes tested were found to catalyze
the desired reaction to a varying extent, producing disa-
ccharide mixtures composed of 2 and considerable fractions
of the undesired (ꢀ 1,3)-regioisomer. No formation of the (
umn chromatography. Residual acceptor 1 could, however,
be easily removed by selective extraction with ethyl acetate,
while the more hydrophilic disaccharide derivative 2 re-
mained in the aqueous phase. Further purification was per-
formed by column filtration through a plug of reversed-
phase silica gel. Free galactose was readily removed upon
washing with water, while ethanol elution furnished pure
benzyl-protected disaccharide 2 in 23% yield as a crystalline
solid. Unambiguous structural proof was obtained by full
characterization of corresponding peracetate 3.
It should be emphasized that the judicious choice of
benzyl protection did indeed strongly facilitate product re-
covery and purification by easily scalable extractive manipu-
lations and thus obviated the need for an exacting high-
performance chromatographic separation. Although reaction
conditions have not been further optimized, the yield is
within the range commonly obtainable by preparative-scale
transglycosylations (13, 14). An improvement may be ex-
pected from applications of glycosynthases that have been
engineered to prevent product hydrolysis (22). Concerning
the regioselectivity of the galactosyl transfer, it should be
compared with the unfavorable 5:2 ratio of 3-/4-
regioisomers obtained for the corresponding benzyl ꢀ-
xyloside with the same enzyme. The unusually high
regioselectivity in the case of ꢁ-xyloside may be due to a
pronounced bend in the molecular shape of 1 induced by the
steric bulk of the ꢁ-anomeric benzyl group. Rather than the
linearly extended geometry of the ꢀ-xyloside, this seems to
render the relative setting of the 3- and 4-hydroxyl groups
sufficiently different from each other to facilitate discrimina-
tion by the galactosidase in favor of the desired compound.
The transgalactosylation was also tested using lactose as
an alternative, inexpensive galactosyl donor. As expected,
the reactions proceeded more slowly compared with those
with the highly reactive o-NPG donor. Following the same
work-up procedure, pure disaccharide 2 was obtained in
15% yield. The lower yield, as compared with the o-NPG
donor, is in line with the relatively lower conversion rate be-
cause of the increasing likelihood of product hydrolysis at
longer reaction times. Attempts were also made for an ex-
tended usage of the biocatalyst. Because a continuous operation
ꢀ
1,2)-disaccharide was observed, which is probably due to
the steric inaccessibility to the biocatalysts. The enzyme
from E. coli was unique, however, because it yielded the de-
sired 1,4-regioisomer 2 as the only disaccharide product de-
tectable by high-resolution NMR spectroscopy (de O 95%).
Although the transgalactosylation proceeded with excel-
lent regioselectivity, the presence of unreacted acceptor
glycoside and free galactose from competing hydrolysis of
the donor complicated product purification by silica gel col-
©
2002 NRC Canada

Fessner and Juárez Ruiz
741
of the transglycosylation would be difficult to achieve given
the propensity of the product to be hydrolyzed under the re-
action conditions, batchwise recycling of the enzyme offered
a better solution. Indeed, when the protein was recovered
from the reaction mixture by ultrafiltration, it could be re-
used for at least four identical consecutive runs without no-
ticeable detrimental effects on the reaction velocity or prod-
uct yield.
Benzyl ꢁ-D-xylopyranoside (1)
The glycosidation mixture (21) from acidic equilibration
of D-xylose with benzyl alcohol (0.1 M) was hydrolyzed in
phosphate buffer (0.2 M, pH 4.5) at 37°C using the crude ꢀ-
galactosidase preparation from A. oryzae with TLC monitor-
ing. The unreacted benzyl ꢁ-D-xyloside (1) was isolated ac-
cording to the published procedure (21) as a colorless
crystalline solid in >60% absolute yield.
Hydrogenolytic cleavage of the benzyl protection group in
2
was performed on 10% Pd/C catalyst. Complete conver-
sion was only achieved by applying quite forcing conditions
60 bar H , 50°C), which may be related to the lower reac-
Benzyl ꢀ-D-galactopyranosyl-(1,4)-ꢁ-D-xylopyranoside (2)
(
o-NPG donor
2
tivity of an ꢁ-glycoside. The resultant product was purified
by filtration through reversed-phase silica gel column using
water as the eluent to furnish the free disaccharide 4 in 80%
yield. The spectroscopic data of this compound was identical
to values published for the disaccharide from chemical syn-
thesis (7).
Compound 1 (100 mg, 0.4 mmol) and o-nitrophenyl ꢀ-D-
galactoside (250 mg, 0.8 mmol) were dissolved in phosphate
buffer (8.0 mL, 50 mM; 1.0 mM MgCl , 5.0 mM mer-
2
captoethanol, pH 7.0), and the mixture was incubated with ꢀ-
galactosidase from E. coli (130 ꢃL, 215 U) on a rotary
shaker (50 rpm) at room temperature. Reaction was moni-
tored for product formation by TLC (MeOH–toluene, 2:1,
Rf (2) = 0.50). After 4 h, an additional equivalent of o-NPG
was added to the solution, and agitation was continued for
Conc lus ions
1
h. The enzyme was removed by heating the solution at
A short, versatile enzymatic synthesis has been developed
for the lactose analog Gal-(ꢀ 1,4)-Xyl 4 by a highly regio-
and stereoselective transgalactosylation of benzyl ꢁ-D-xylo-
side 1 as the acceptor. Unlike the situation for free xylose or
the corresponding ꢀ-D-xyloside, the ꢀ-galactosidase from
E. coli efficiently directs galactosyl transfer to 4-OH of the
acceptor, which is likely supported by the steric influence of
the bulky ꢁ-anomeric substituent that is presumed to restrict
a competing access to 3-OH. The benzyl group is of further
advantage for simple and effective isolation of the target
disaccharide 2 from complex product mixtures by allowing
group-specific extraction procedures. Thus, this synthesis is
promising for the preparation of larger disaccharide quanti-
ties.
80°C for 3 min, followed by centrifugation. Free nitrophenol
and residual acceptor were removed by extraction with ethyl
acetate (3 × 10 mL), and the aqueous phase was filtered
through a plug of RP-silica gel (5.0 g). After washing with
water (20 mL), the product was eluted with ethanol to fur-
nish 2 as a colorless solid (37 mg, 23%).
Lactose donor
Compound 1 (23 mg, 0.1 mmol) and D-lactose (65 mg,
0.2 mmol) were dissolved in phosphate buffer (2.0 mL,
–
1
50 mM; 1.0 mmol L MgCl , 5.0 mM mercaptoethanol,
2
pH 7.0), and the mixture was incubated with ꢀ-galactosidase
from E. coli (30 ꢃL, 50 U) with shaking. After 3 h, a further
equivalent of lactose was added to the solution, and agitation
was continued for 2 h. The product was isolated as above to
give pure 2 (4 mg, 15%).
Expe rim e nta l
Repetitive-batch operation
To a solution containing D-lactose (3.7 g, 10.8 mmol) in
General
phosphate buffer (20 mL, 50 mM; 1.0 mmol L–1 MgCl2,
5
(
NMR spectra were recorded on Bruker ARX-300 and
Bruker AVANCE 500 spectrometers; chemical shifts are ref-
erenced to internal TMS (CDCl ) or TSP (D O; 0.00 ppm).
.0 mM mercaptoethanol, pH 7.0) was added a solution of 1
1.00 g, 4.2 mmol) in acetonitrile (1.0 mL). The mixture was
3
2
placed in an ultrafiltration cell and was incubated with ꢀ-
galactosidase from E. coli (1.3 mL, 2150 U) at room temper-
ature for 3.5 h with slow stirring. The product-containing so-
lution was collected by pressure application, and the enzyme
that was retained by the membrane was re-used for a subse-
quent, identical conversion.
Mass spectra were recorded on a Bruker Esquire-LC system
ESI), and elemental analyses were performed on a Heraeus
CHN-O-Rapid system. Optical rotations were measured with
(
a PerkinElmer 241 apparatus in a 10-cm cuvette at ꢂ =
5
89 nm. For ultrafiltration, an Amicon 8400 cell equipped
with a YM 10 membrane was used. Column chromatography
was performed on Merck 60 silica gel (0.063–0.200 mesh),
and analytical thin-layer chromatography was performed on
Merck silica gel plates 60 GF254 using anisaldehyde stain for
detection. C -silica gel 100 (Fluka, 0.040–0.063 mm) was
The combined reaction volumes from consecutive batches
were processed by extractive removal of 1 using ethyl ace-
tate (3 × 30 mL), followed by solid-phase extraction of 2 on
RP-silica gel essentially as described above. Isolated yield
1
8
2
0
3
(
34 mg (10%); [ꢁ]D +63.7 (c 1.0, water); R = 0.42
used for reversed-phase (RP) adsorption. ꢀ-Galactosidases
from A. oryzae (Sigma, G-5160 or G-7138) and from E.
coli (Roche Molecular Biochemicals, 150797 or Sigma, G-
f
1
EtOAc–isopropanol–H O, 9:4:2). H NMR (500 MHz,
2
D O) ꢄ: 7.49–7.41 (m, 5H, H ), 5.00 (d, 1H, 1>-H), 4.77 (d,
2
ar
1
H, 1-H ), 4.62 (d, 1H, 1-H ), 4.46 (d, 1H, 1>>-H), 3.92 (d,
6
008) were used as commercially supplied. The
a b
®
1H, 4>>-H), 3.87–3.78 (m, 4H, 3>-H, 5>-H , 5>>-H, 6>>-H ),
CloneZyme glycosidase kit from Recombinant Biocatalysts
Inc. (now Diversa) was used for transglycosylation screen-
ing.
a a
3
3
.75 (dd, 1H, 5>-H ), 3.70 (dd, 1H, 4>-H), 3.66–3.63 (m, 2H,
b
>>-, 6>>-H ), 3.60 (dd, 1H, 2>-H), 3.52 (dd, 1H, 2>>-H) (J
=
b
1a,1b
©
2002 NRC Canada

7
42
Can. J. Chem. Vol. 80, 2002
1
1
1.7, J1 >, 2 > = 3.7, J2 >, 3 > = 9.3, J3 >, 4> = 8.1, J4 >, 5b> = 3.9, J
1.8, J1 >, 2 > = 7.8, J2 >, 3 > = 9.8, J3 >, 4 > = 3.3 Hz). C NMR
2 i o m
31.16 (C ), 104.6 (C-1>>), 100.3 (C-1>), 79.6 (C-3>), 78.1
p
=
Ac knowle dgm e nts
5a >, 5b>
1
3
This work was partially supported by the Fonds der
Chemischen Industrie. J.M.J.R. thanks the Consejo Nacional
de Ciencia y Tecnologia (CONACYT) México, for a re-
search fellowship. We are grateful to Dr. P. Rasor from
Roche Molecular Biochemicals for a generous gift of the ꢀ-
galactosidase from E. coli.
(
125.7 MHz, D O) ꢄ: 139.8 (C ), 131.6 (C ), 131.3 (C ),
1
(
C-4>), 75.5 (C-3>>), 74.3 (C-5>>), 74.0 (C-2>), 73.5 (C-2>>),
2.6 (C-1), 71.5 (C-4>>), 63.9 (C-5>), 62.0 (C-6>>). MS m/z
7
+
(
%): 425.3 ([M + Na] 100).
Benzyl 2,3,4,6-tetra-O-acetyl-ꢀ-D-galactopyranosyl-(1,4)-
2
,3-di-O-acetyl-ꢁ-D-xylopyranoside (3)
Disaccharide 2 (40 mg, 0.1 mmol) was taken up in dry
Re fe re nc e s
pyridine (3.0 mL), acetic anhydride (2.0 mL) was added, and
the solution was stirred at room temperature for 24 h. Water
was added (20 mL), and the mixture was extracted with
EtOAc (3 × 10 mL). The combined organic phases were
washed consecutively with water, 1.0 N HCl, sat. NaHCO3,
and brine, and then dried over Na SO . The solvent was re-
1. E.H. Rings, R.J. Grand, and H.A. Buller. Curr. Opin. Pediatr.
6, 562 (1994).
2
3
. H.Y. Naim. Histol. Histopathol. 16, 553 (2001).
. J.J. Aragon, A. Fernandez-Mayoralas, J. Jimenez-Barbero, M.
Marin-Lomas, A. Rivera-Sagredo, and D. Villanueva. Clin.
Chim. Acta, 210, 221 (1992).
2
4
4
5
. A. Rivera-Sagredo, F.J. Canada, O. Nieto, J. Jimenez-Barbero,
and M. Martin-Lomas. Eur. J. Biochem. 209, 415 (1992).
. P. Fernandez, F.J. Canada, J. Jimenez-Barbero, and M. Martin-
Lomas. Carbohydr. Res. 271, 31 (1995).
moved in vacuo, and the residue recrystallized from Et O–
2
petrol ether mixtures to give peracetate 3 as colorless crys-
tals (95 mg, 95%); mp 102°C; R = 0.47 (chloroform–
f
1
MeOH, 40:1). H NMR (300 MHz, CDCl ) ꢄ: 7.35–7.23 (m,
3
6
7
. S.B. Selleck. Trends Genet. 16, 206 (2000).
. A. Rivera-Sagredo, A. Fernandez-Mayoralas, J. Jimenez-Barbero,
M. Martin-Lomas, D. Villanueva, and J.J. Aragon. Carbohydr.
Res. 228, 129 (1992).
H, H ), 5.42 (t, 1H, 3>-H), 5.33 (dd, 1H, 4>>-H), 5.06 (dd,
H, 2>>-H), 4.96 (dd, 1H, 3>>-H), 4.93 (d, 1H, 1>-H), 4.73 (dd,
ar
H, 2>-H), 4.71 (d, 1H, 1-H ), 4.47 (d, 1H, 1-H ), 4.46 (d,
a
b
H, 1>>-H), 4.08 (m, 2H, 6>>-H) 3.86 (dd, 1H, 5>>-H), 3.78
8
. B. Lindberg, L. Roden, and B.G. Silvander. Carbohydr. Res. 2,
413 (1966).
ab
1
1
1
.99, 1.95 (6 P s, 18H, Ac) (J1a,1b = 12.2, J1 >, 2> = 3.3, J2 >, 3>
0.2, J3 >, 4> 9.4, J4 >, 5a> = 7.6, J5a >, 5b> = 8.0, J1 >, 2 > = 7.6, J2 >, 3 >
=
=
9. B. Erbing, B. Lindberg, and T. Norberg. Acta Chem. Scand.
Ser. B, 32, 308 (1978).
10. R. Almeida, S.B. Levery, U. Mandel, H. Kresse, T.
Schwientek, E.P. Bennett, and H. Clausen. J. Biol. Chem. 274,
26165 (1999).
11. T. Wiemann, Y. Nishida, V. Sinnwell, and J. Thiem. J. Org.
Chem. 59, 6744 (1994).
12. Y. Hara and K. Suyama. Eur. J. Biochem. 267, 830 (2000).
3. K.G.I. Nielsson. In Modern methods in carbohydrate synthe-
sis. Edited by S.H. Khan and R.A. O’Neill. Harwood Aca-
demic, Amsterdam. 1996. p. 518.
4. M. Scigelova, S. Singh, and D.H.G. Crout. J. Mol. Catal. B:
Enzym. 6, 483 (1999).
5. E. Montero, J. Alonso, F.J. Canada, A. Fernandez-Mayoralas,
and M. Martin-Lomas. Carbohydr. Res. 305, 383 (1998).
6. J.J. Aragon, F.J. Canada, A. Fernandez-Mayoralas, R. Lopez,
M. Martin-Lomas, and D. Villanueva. Carbohydr. Res. 290,
0.4, J3 >, 4 >> 3.4, = J4 >, 5 > = 0.9, J5 >, 6a > = 6.6, J5 >, 6b >, = 7.1 Hz)
C NMR (75.4 MHz, CDCl ) ꢄ: 172.2 (C=O), 138.9 (C ),
1
3
3
i
1
7
7
2
30.4 (C ), 129.9 (C ), 129.7 (C ), 103.1 (C-1>>), 96.9 (C-1>),
o m p
9.1 (C-4>), 73.1 (C-5>>), 72.9 (C-2>), 72.8 (C-3>>), 72.1 (C-3>),
1.6 (C-1), 71.1 (C-2>>), 68.9 (C-4>>), 63.2 (C-6>>), 61.4 (C-5>),
2.5, 22.6, 22.8 (6 P CH ). Anal. calcd. for C H O
3
30 38 16
1
(
654.61): C 55.04, H 5.85; found C 55.12, H 5.88.
ꢀ
-D-Galactopyranosyl-(1,4)-D-xylose (4)
1
1
1
To a solution of benzyl disaccharide 2 (160 mg,
.4 mmol) in 50% aqueous ethanol (10 mL) containing ace-
0
tic acid (0.25 mL) was added Pd/C catalyst (10%; 400 mg),
and the suspension was stirred in an autoclave under hydro-
gen (60 bar) at 50°C. After 24 h (TLC control EtOAc–
isopropanol–water, 3:2:2), the reaction mixture was filtered
through Celite, and the solution was concentrated in vacuo.
The resultant residue was taken up in water (10 mL) and fil-
tered through a plug of RP silica gel (2.0 g). After further
elution with water (20 mL), the combined aqueous phases
were freeze-dried to provide free disaccharide 4 as a color-
less solid, which could be recrystallized from acetone. Yield
2
09 (1996).
7. R. Lopez and A. Fernandez-Mayoralas. J. Org. Chem. 59, 737
1994).
8. R. Lopez and A. Fernandez-Mayoralas. Tetrahedron Lett. 33,
449 (1992).
1
1
(
5
19. T. Yasukochi, C. Inaba, K. Fukase, and S. Kusumoto. Tetrahe-
dron Lett. 40, 6585 (1999).
20. T. Yasukochi, K. Fukase, Y. Suda, K. Takagaki, M. Endo, and
S. Kusumoto. Bull. Chem. Soc. Jpn. 70, 2719 (1997).
21. J.M. Juarez Ruiz, G. Oꢀwald, M. Petersen, and W.-D. Fessner.
J. Mol. Catal. B: Enzym. 11, 189 (2001).
2
0
9
(
4
9 mg (80%); mp 164°C (mp 160–168°C (7)); [ꢁ] +21.5
D
20
1
c 1.0, water; [ꢁ]D +22 (7)). H NMR (300 MHz, D O) ꢄ:
.56 (d, J1 >, 2 > = 7.3 Hz, 1>>-H), 3.97–3.56 (m, 12H).
2
1
3
C
NMR (75.4 MHz, D O) ꢄ: 105.19 (C-1>>), 82.91 (C-1>),
2
2
2. L.F. Mackenzie, Q. Wang, R.A.J. Warren, and S.G. Withers. J.
7
6
7.95, 75.40, 74.71, 73.76, 72.70, 71.45, 65.48, 63.89,
3.44.
Am. Chem. Soc. 120, 5583 (1998).
©
2002 NRC Canada