Synthesis and NMR studies of methyl 3-O-[(R)- and (S)-1-carboxyethyl]-α-D-gluco-, galacto- and manno- pyranosides

Article abstract of DOI:10.1016/S0008-6215(98)00272-9

The title compounds were prepared by condensation of a suitably protected monosaccharide and (R)- or (S)-2-chloropropionic acid followed by removal of the protecting groups. Characterisation by ¹H and ¹³C NMR spectroscopy and NOE difference experiments revealed both chemical shift and enhancement differences between the diastereomers which could be used in determination of the absolute configuration of the 1-carboxyethyl group. Copyright (C) 1998 Elsevier Science Ltd.

Full text of DOI:10.1016/S0008-6215(98)00272-9

Carbohydrate Research 313 (1998) 157–164
Synthesis and NMR studies of methyl 3-O-[(R)- and
(S)-1-carboxyethyl]-a-
D
-gluco-, galacto- and manno-
pyranosides
Lars Andersson, Lennart Kenne *
Department of Chemistry, Swedish Uni6ersity of Agricultural Sciences, P.O. Box 7015, SE-750 07 Uppsala, Sweden
Received 27 July 1998; accepted 7 October 1998
Abstract
The title compounds were prepared by condensation of a suitably protected monosaccharide and (R)- or
1
(S)-2-chloropropionic acid followed by removal of the protecting groups. Characterisation by H and 13C NMR
spectroscopy and NOE difference experiments revealed both chemical shift and enhancement differences between the
diastereomers which could be used in determination of the absolute configuration of the 1-carboxyethyl group.
© 1998 Elsevier Science Ltd. All rights reserved.
Keywords: Diastereomers; Methyl glucosides; Monosaccharide derivatives
as the 13C chemical shift of the signal for the
1. Introduction
carbon substituted with the 1-carboxyethyl
group will appear 8–9 ppm downfield, relative
to the corresponding signal from a non-substi-
tuted sugar. The (R)- or (S)-configuration of
the 1-carboxyethyl group was usually deter-
mined, after reduction of the carboxylic acid
group, by gas chromatographic comparison of
the derived alditol acetates with those of syn-
thetic samples. To avoid synthesis of reference
substances, the need for more direct methods
for determination of the configuration of the
1-carboxyethyl group is obvious. So far two
different alternative methods have been de-
scribed. These are based on (i) NOE measure-
ments on the conformationally rigid lactones
formed between the carboxyl group and an
adjacent hydroxyl group [5] and (ii) circular
dichroism (CD) spectroscopy on methyl gly-
cosides of 1-carboxyethyl substituted sugars
[6]. These methods are not entirely general so
there is still the need for reference substances
Sugar residues etherified with lactic acid are
important components in several bacterial
polysaccharides [1–3]. The biosynthetic path-
way to the lactic acid derivatives involves the
conversion of a phosphoenol pyruvate into an
enol of pyruvic acid which is then reduced to
the (R)- or (S)-1-carboxyethyl group [1]. Both
the (R)- and (S)-forms of the 1-carboxyethyl
groups substituting the 3-, 4- or 6-positions of
different monosaccharide units have been
found [1–4]. The substitution position of the
1-carboxyethyl group has mainly been deter-
mined, after reduction of the carboxylic acid
group, by mass spectrometry of the derived
alditol acetates or the partly methylated aldi-
tol acetates formed. The substitution position
can also be determined by NMR spectroscopy
* Corresponding author. Tel.: +46-18-671573; fax: +46-
18-673477; e-mail: lennart.kenne@kemi.slu.se
0008-6215/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0008-6215(98)00272-9

158
L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
Scheme 1.
and for further improvements of alternative
methods.
condensation minor amounts, 1–6% as deter-
mined by H NMR spectroscopy, of the other
1
At our department the capsular polysaccha-
rides from several strains of Butyri6ibrio fibri-
sol6ens are under investigation and 1-carboxy-
ethyl substituted sugars are common compo-
nents of these polysaccharides [3]. Thus, it is
an advantage for us to have different 1-car-
boxyethyl substituted sugars available in order
to facilitate the structural studies, and for
further development of more convenient
methods for determination of the configura-
tion of the 1-carboxyethyl group. We report
herein the synthesis of the (R)- and (S)- forms
of the methyl 3-O-(1-carboxyethyl)-a-D-gluco-
(1 and 2), galacto- (3 and 4) and manno- (5
and 6) pyranosides.
diastereomer was formed during the reaction
conditions used. These amounts varied with
respect to the absolute configuration of the
2-chloropropionic acid and the stereochem-
istry of the adjacent hydroxyl groups of the
sugar. The protected 3-substituted methyl gly-
copyranosides of glucose and galactose were
purified from the contaminating diastereomers
by crystallisation before removal of the benzyl
ether and the 4,6-O-benzylidene acetal pro-
tecting groups. The latter groups were re-
moved by hydrogenolysis to give the
3-substituted methyl glycopyranosides 1, 2, 3
and 4. We were not able to purify the pro-
tected 3-substituted mannose derivatives from
the contaminating diastereomers before de-
protection. Therefore, isomeric mixtures (98/2
and 94/6, respectively) were obtained from the
deprotected mannosides 5 and 6.
2. Results and discussion
The 1-carboxyethyl substituted methyl gly-
cosides were investigated by NMR spec-
The 1-carboxyethyl substituted methyl gly-
cosides 1, 2, 3, 4, 5 and 6 were synthesised by
the condensation of (S)- or (R)-2-chloropropi-
onic acid [7,8] with protected monosaccharide
derivatives in which all hydroxyl groups but
that of the position for substitution were
blocked. This method was used previously for
the synthesis of other 1-carboxyethyl substi-
tuted sugars [6,9] (Scheme 1).
Inversion of the configuration at C-2% takes
place when the hydroxy group of the sugar
reacts with the 2-chloropropionic acid forming
an ether. The monosaccharide derivatives cho-
sen were the readily available methyl 2-O-ben-
zyl-4,6-O-benzylidene-a-D-gluco- and manno-
pyranosides [10,11]. The galactose derivative
chosen was methyl 2-O-benzyl-4,6-O-benzyli-
dene-a-D-galactopyranoside which was pre-
pared from methyl 4,6-O-benzylidene-a-D-
galactopyranoside [12] by partial benzylation
of the 2-position as described by Garegg et al.
[10]. In addition to the expected product in the
1
troscopy and the H and 13C NMR chemical
shifts are given in Table 1. When the 13C
chemical shifts are compared with those of the
unsubstituted methyl glycosides significant
shifts are only observed for the C-2, C-3 and
C-4 signals with a large downfield shift for the
signal of the substituted carbon and upfield
shifts for the signals of the two adjacent
carbons. The magnitude of these shifts is de-
pendent on the configuration of the 1-car-
boxyethyl group and differences between the
(R)- and (S)-forms are thus observed. For the
methyl glucosides 1 and 2 an upfield shift
(−0.6/−0.9 ppm) for the C-2 signal of 1 and
the C-4 signal of 2 was observed and almost
no shift for the C-4 signal of 1 and the C-2
signal of 2. In addition to the shift differences
for the (R)- and (S)-forms the galactosides
and mannosides show a shift pattern for the
stereochemically similar compounds 3 and 6,

L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
159
Table 1
1H and 13C chemical shifts of 3-O-[(R)- and (S)-1-carboxyethyl]-substituted methyl a-D-glycosides 1, 2, 3, 4, 5 and 6 for solutions
in D2O
Compound
H-1
C-1
H-2
C-2
H-3
C-3
H-4
C-4
H-5
C-5
H-6a/H-6b
C-6
OCH3
OCH3
H-2%
C-2%
H-3%
C-3%
C-1%
(R)-Glcp (1)a
(S)-Glcp (2)
Me Glcpb
4.83
99.62
4.79
100.29
4.80
100.10
4.87
99.79
4.84
100.18
4.84
100.25
4.79
101.21
4.78
3.63
71.46
3.64
72.31
3.55
72.05
3.92
67.96
3.91
68.12
3.81
69.03
4.01
68.50
4.05
3.52
82.79
3.52
82.66
3.66
73.92
3.60
78.06
3.63
78.95
3.81
70.32
3.58
79.97
3.54
3.48
70.56
3.46
69.54
3.40
70.40
4.12
67.30
4.04
67.62
3.97
70.06
3.71
66.70
3.73
3.62
72.54
3.64
72.65
3.64
72.41
3.85
71.47
3.89
71.09
3.89
71.56
3.63
73.26
3.61
3.86/3.75
61.24
3.85/3.73
61.42
3.87/3.75
61.40
3.75/3.74
62.09
3.74/3.74
62.14
3.74/3.74
62.06
3.90/3.76
61.83
3.88/3.74
61.80
3.41
55.64
3.41
55.85
3.42
55.84
3.40
55.66
3.41
55.79
3.41
55.86
3.40
55.51
3.39
55.49
3.41
4.26
79.62
4.23
1.40
19.58
1.39
182.70
183.00
79.48
19.55
(R)-Galp (3)
(S)-Galp (4)
Me Galp
4.11
76.42
4.06
1.37
19.70
1.37
182.41
182.79
77.45
19.40
(R)-Manp (5)
(S)-Manp (6)
Me Manp
4.08
77.61
4.10
1.38
19.35
1.38
182.52
182.22
101.57
4.76
101.69
67.97
3.93
70.74
79.40
3.76
71.38
66.34
3.64
67.60
73.61
3.62
73.38
76.39
19.64
3.90/3.76
61.78
55.53
a (R)-Glcp, methyl 3-O-[(R)-1-carboxyethyl]-a-D-glucopyranoside, etc.
b Me Glcp, methyl a-D-glucopyranoside, etc.
and 4 and 5 in which the adjacent carbons
have the hydroxyl groups in axial and equato-
rial positions. If only the stereochemistry
nearest to the 1-carboxyethyl substituent is
considered, the pairs C-2, C-3 and C-4 are
mirror images (Fig. 1). This means that the
(R)-form of the galactoside 3 has the same
stereochemistry around C-3 as the (S)-form of
the mannoside 6 and this is reflected in similar
shifts for the signals of the axially substituted
carbons and the equatorially substituted car-
bons compared with those of the unsubsti-
tuted methyl glycosides. The shifts for the
signals of 3 C-2/6 C-4, 3 C-3/6 C-3 and 3
C-4/6 C-2 are shifted by approximately the
same amount (ꢀ −1.2, ꢀ7.9 and −2.8
ppm) and this is also observed for the signals
of 4 C-2/5 C-4, 4 C-3/5 C-3 and 4 C-4/5 C-2
(−0.9, 8.6 and ꢀ −2.3 ppm).
1
In comparison with the H chemical shifts
from the pertinent methyl glycosides only the
signals from H-2, H-3 and H-4 are affected. A
minor downfield shift (0.06–0.15 ppm) for the
H-2 and H-4 signals and an upfield shift (be-
tween −0.14 and −0.22 ppm) for the H-3
signal are observed. No significant differences
were observed for the (R)- and (S)-forms
of the methyl glycosides but a similar shift
pattern to that obtained for the 13C signals
could be observed for the stereochemically
similar compounds of the methyl galactosides
and mannosides, 3 and 6, and 4 and 5, re-
spectively.
1
Thus the chemical shifts of both the H and
13C signals are dependent on the stereochem-
istry around the substitution position as well
as on the configuration of the 1-carboxyethyl
group. The configurational dependence of the
1H signals is minor but the 13C signals are
Fig. 1. The stereochemistry around C-3 in the 1-carboxyethyl
substituted methyl D-galactosides and D-mannosides. The
stereochemistry for 3 (A) and 6 (B) is shown.

160
L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
Table 2
group if the absolute configuration of the
sugar is known. However, there could be
many other contributions influencing the
NMR data when the sugar substituted with a
1-carboxyethyl group is a component in a
polysaccharide, making the identification
based on these data difficult.
An additional difference between the (R)-
and (S)-1-carboxyethyl substituted methyl a-
D-glycosides is the values for the optical rota-
tion, [h]D. All pairs show a 60° difference with
the (R)-form having the most positive value.
Similar differences have been observed for
other 1-carboxyethyl substituted methyl gly-
cosides [6].
Observed NOEa in % of [(R)- and (S)-1-carboxyethyl]-substi-
tuted methyl a-D-glycosides 1, 2, 3, 4, 5 and 6 when H-3% of
the 1-carboxyethyl group was presaturated (86–98%)
Compound
H-2
H-3
H-4
H-2%
(R)-Glcp (1)b
(S)-Glcp (2)
(R)-Galp (3)
(S)-Galp (4)
(R)-Manp (5)
(S)-Manp (6)
0.033
0.073
0.063
0.077
0.077
0.120
0.087
0.073
0.090
0.210
0.223
0.077
0.067
0.060
0.113c
0.046c
0.067
0.053
1.30
1.30
1.17
1.14
1.03
1.00
a Data were obtained from NOE difference experiments at
400 or 600 MHz, setting the intensity of the H-3% signal to
−300%.
b (R)-Glcp, methyl 3-O-[(R)-1-carboxyethyl]-a-D-glucopyr-
anoside, etc.
c Overlap of signals resolved by addition of pyridine-d5.
3. Experimental
significantly changed and thus they could be
used for the determination of the relative
configuration of the 1-carboxyethyl group.
The methyl glycosides 1, 2, 3, 4, 5 and 6
were subjected to NOE difference experiments
at both 400 and 600 MHz with presaturation
of H-3% of the 1-carboxyethyl group. Only
minor differences in magnitude of the en-
hancements were observed for the experiments
at the two magnetic fields and the relative
enhancements between the signals were the
same. The observed enhancements for the sig-
nals of the ring protons H-2, H-3 and H-4 and
the H-2% signal are given in Table 2. No
significant difference was observed for the
(R)- and (S)-forms of the methyl glucosides 1
and 2 with the adjacent hydroxyl groups in
equatorial positions. However, for the differ-
ent forms of the methyl galactosides, 3 and 4,
and mannosides, 5 and 6, which have the
adjacent hydroxy groups in axial and equato-
rial positions, there are similar enhancements
of the signals of the ring protons for the
compounds with similar stereochemistry (Fig.
1), i.e., compounds 3 and 6, and 4 and 5,
respectively.
General methods.—NaH (55–60% suspen-
sion in oil) was washed with petroleum ether
prior to use. TLC was performed on pre-
coated plates (E. Merck Silica Gel 60 F254),
detection being afforded with 5% H2SO4 and
heating. Column chromatography of synthetic
intermediates was performed on Matrex silica
˚
gel 60 A (35–70 mm, Amicon). Gel perme-
ation chromatography of deprotected prod-
ucts was carried out on a column (2.6×80
cm) of Bio-Gel P-2, using H2O as eluent.
Solvents were used as purchased except for
˚
1,4-dioxane which was dried with 3 A molecu-
lar sieves prior to use. The petroleum ether
fraction used had a bp of 60–70 °C. Solutions
were concentrated under reduced pressure at
temperatures not exceeding 50 °C. Optical ro-
tations were recorded at 20–25 °C using
Perkin–Elmer 141 or Perkin–Elmer 341 po-
larimeters. The [h]D-values for the mannose
derivatives 5, 13, 6 and 14 have been corrected
since the analytical samples contained small
but controlled amounts of the contaminating
diastereomer. High-resolution mass spec-
trometry (HRMS) was applied in the FAB
positive mode on a Jeol JMS-SX/SX-102A
four-sector tandem instrument with either
glycerol or polyethyleneglycol as the matrix.
NMR spectra were recorded at 400 MHz (1H)
and 100 MHz (13C) with a Bruker DRX 400
instrument except for some of the NOE-differ-
ence spectra, which were recorded on a Bruker
When adding the information obtained
1
13
from all the H and C NMR chemical shifts
and the NOE experiments then there are sig-
nificant differences between the (R)- and (S)-
forms of the methyl 3-O-(1-carboxyethyl)-
a-glycosides and it should be possible to iden-
tify the configuration of the 1-carboxyethyl

L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
161
4.40 (q, 1 H, J 7 Hz, carboxyethyl-CH), 4.26
(dd, 1 H, J6a,6b 9.5, J5,6a 4.5 Hz, H-6a), 3.92 (t,
1 H, J2,3=J3,4 =9.5 Hz, H-3), 3.78 (ddd, 1 H,
J4,5 10 Hz, H-5), 3.72 (dd, 1 H, H-6b), 3.59
(m, 2 H, H-2, 4), 3.33 (s, 3 H, OCH3), 1.47 (d,
3 H, J 7 Hz, carboxyethyl–CH3); 13C NMR: l
174.7 (COOH), 137.0–125.8 (aromatic), 101.3
(PhCH), 98.1 (C-1), 81.6 (C-4), 78.4 (C-2),
78.1 (C-3), 77.3 (carboxyethyl–CH), 73.1
(PhCH2), 68.9 (C-6), 62.2 (C-5), 55.3 (OCH3),
18.9 (carboxyethyl–CH3).
Hydrogenolysis of compound 7 (112 mg,
252 mmol) dissolved in 10 mL 1,4-dioxane was
catalysed by Pd–C (50 mg) and carried out
for 4 h at 20–25 °C. The catalyst was removed
by filtration and one drop of aq conc NH3
was added to the filtrate to avoid accidental
hydrolysis and possible lactone formation un-
der the following concentration to dryness.
This gave the crude ammonium salt of 1
which was purified on a column of Bio-Gel
P-2 followed by passage through a column
(1×5 cm) of cation exchanger (Dowex 50,
Na+). After lyophilisation, the Na salt of 1
(66 mg, 91%) was afforded as a white amor-
phous powder; [h]D +144° (c 1.0, H2O);
HRMS: Calcd for C10H18O8+H: 267.1079,
found 267.1102 [M+H]+.
DRX 600 spectrometer. Chemical shifts are
given relative to Me4Si using internal acetone
(lH 2.225, lC 31.07) for D2O solutions at
30 °C, whereas NMR spectra recorded for
samples in CDCl3 are referenced to internal
Me4Si (l 0.00). NMR spectra on all fully
protected 1-carboxyethyl substituted sugars
were recorded in CDCl3 containing 1.5%
pyridine-d5 due to poor stability in CDCl3.
NMR spectra were recorded at 30 °C where
nothing else is stated. Assignments of signals
were done from standard H,H-COSY and
1
HMQC experiments. H NMR chemical shifts
and coupling constants of overlapping signals
were obtained from the cross-peaks in the 2D
spectra. Melting points were determined on a
METTLER FP62 instrument and are
uncorrected.
Methyl 3-O-[(R)-1-carboxyethyl]-h-D-glu-
copyranoside (1).—A solution of methyl 2-O-
benzyl-4,6-O-benzylidene-a-D-glucopyranoside
[10] (1.00 g, 2.69 mmol) and (S)-2-chloropro-
pionic acid [7,8] (1.26 g, 11.6 mmol) in 1,4-
dioxane (50 mL) was stirred with NaH (1.8 g,
75 mmol) for 18 h at 50°C, whereupon the
mixture was cooled and water (15 mL) added
to destroy excess NaH. A two-phase mixture
was obtained. The upper phase, containing the
product, was a 1,4-dioxane–water mixture
and the lower viscous phase contained mostly
salt. The upper phase was collected, washed
with petroleum ether (3×15 mL), and aci-
dified (pH 2–3, HCl), and the product was
extracted with CH2Cl2 (3×30 mL). The
CH2Cl2 solution was washed with H2O (15
mL), dried (Na2SO4), filtered, and concen-
trated to give crude methyl 2-O-benzyl-4,6-O-
benzylidene-3-O-[(R)-1-carboxyethyl]-a-D-
glucopyranoside (7), which was contaminated
by 6% of methyl 2-O-benzyl-4,6-O-benzyli-
dene-3-O-[(S)-1-carboxyethyl]-a-D-glucopy-
ranoside (8). Recrystallisation from toluene–
petroleum ether gave pure 7 (814 mg, 68%) as
white needles; mp 171–172 °C, [h]D +26° (c
0.5, CHCl3); HRMS: Calcd for C24H28O81+
Na: 467.1682, found 467.1663 [M+Na]+; H
NMR: l 10.25 (s, 1 H, COOH), 7.48–7.28 (m,
10 H, Ar–H), 5.54 (s, 1 H, PhCH), 4.81 (d, 1
H, Jgem 12 Hz, PhCH2), 4.70 (d, 1 H, Jgem 12
Hz, PhCH2), 4.64 (d, 1 H, J1,2 3.5 Hz, H-1),
Methyl 3-O-[(S)-1-carboxyethyl]-h-D-glu-
copyranoside (2).—Starting from methyl 2-O-
benzyl-4,6-O-benzylidene-a-D-glucopyranoside
(1.03 g, 2.76 mmol), (R)-2-chloropropionic
acid (1.00 g, 9.21 mmol) and NaH (1.65 g, 69
mmol), using the same experimental and
work-up conditions as in the synthesis of 7,
methyl
2-O-benzyl-4,6-O-benzylidene-3-O-
[(S)-1-carboxyethyl]-a-D-glucopyranoside (8)
contaminated by less than 1% of 7 was ob-
tained. Recrystallisation from toluene–
petroleum ether yielded pure 8 (890 mg, 73%)
as colourless crystals; mp 167.5–168.5 °C, [h]D
+18° (c 1.0, CHCl3); HRMS: Calcd for
C24H28O8+H: 445.1862, found 445.1902
1
[M+H]+. H NMR: l 8.4 (s, 1 H, COOH),
7.48–7.27 (m, 10 H, Ar–H), 5.54 (s, 1 H,
PhCH), 4.81 (d, 1 H, Jgem 12 Hz, PhCH2), 4.67
(d, 1 H, Jgem 12 Hz, PhCH2), 4.59 (d, 1 H, J1,2
3.5 Hz, H-1), 4.43 (q, 1 H, J 7 Hz, car-
boxyethyl–CH), 4.26 (dd, 1 H, J6a,6b 10, J5,6a
4.5 Hz, H-6a), 3.95 (t, 1 H, J2,3=J3,4 =9.5

162
L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
EtOAc to give pure 9 (2.41 g, 46% over all
yield). Crystallisation of 9 from toluene–
petroleum ether gave white crystals: mp
100.6–101.5 °C; [h]D +59° (c 1.1, CHCl3);
HRMS: Calcd for C21H24O6+H: 373.1651,
Hz, H-3), 3.82 (ddd, 1 H, H-5), 3.71 (dd, 1 H,
J5,6b 10 Hz, H-6b), 3.64 (dd, 1 H, J4,5 9 Hz,
H-4), 3.59 (dd, 1 H, H-2), 3.38 (s, 3 H,
OCH3), 1.48 (d, 3 H, J 7 Hz, carboxyethyl–
CH3); 13C NMR: l 174.8 (COOH), 137.8–
126.1 (aromatic), 101.8 (PhCH), 98.8 (C-1),
80.8 (C-4), 79.5 (C-2), 78.9 (C-3), 77.4 (car-
boxyethyl–CH), 73.7 (PhCH2), 69.0 (C-6),
62.1 (C-5), 55.5 (OCH3), 19.1 (carboxyethyl–
CH3).
Hydrogenolysis of 8 (121 mg, 272 mmol),
followed by purification and exchange of
cations, was done in the same way as for
compound 7. This gave the Na salt of 2 (75
mg, 95%) as a white amorphous powder; [h]D
+81° (c 1.0, H2O); HRMS: Calcd for
C10H18O8+Na: 289.0899, found 289.0895
[M+Na]+.
1
found 373.1642 [M+H]+; H NMR (CDCl3,
25 °C): l 7.55–7.25 (m, 10 H, Ar–H), 5.54 (s,
1 H, PhCH), 4.80 (d, 1 H, Jgem 12 Hz,
PhCH2), 4.78 (d, 1 H, J1,2 3.5 Hz, H-1), 4.65
(d, 1 H, Jgem 12 Hz, PhCH2), 4.27 (dd, 1 H,
J3,4 3.5, J4,5 1.0 Hz, H-4), 4.24 (dd, 1 H, J6a,6b
12.5, J5,6a 1.5 Hz, H-6a), 4.13 (dd, 1 H, J2,3 10
Hz, H-3), 4.05 (dd, 1 H, J5,6b 1.5 Hz, H-6b),
3.82 (dd, 1 H, H-2), 3.66 (m, 1 H, H-5), 3.36
(s, 3 H, OCH3), 2.46 (s, 1 H, HO-3); 13C
NMR (CDCl3, 25 °C): l 137.6–126.3 (aro-
matic), 101.2 (PhCH), 98.9 (C-1), 76.8 (C-2),
76.1 (C-4), 73.3 (PhCH2), 69.3 (C-6), 68.5
(C-3), 62.3 (C-5), 55.6 (OCH3).
Methyl 2-O-benzyl-4,6-O-benzylidene-h-D-
galactopyranoside (9).—To a solution of
Acetylation of a sample of pure 9 gave,
after purification, an analytical sample of 10
as a syrup; [h]D +128° (c 1.0, CHCl3);
HRMS: Calcd for C23H26O7+H: 415.1756,
methyl
4,6-O-benzylidene-a-D-galactopyra-
noside (4.00 g, 14.2 mmol) prepared according
to [12], tetrabutylammoniumhydrogen sulfate
(1.00 g, 2.95 mmol), and benzyl bromide (4.17
g, 24.4 mmol) in CH2Cl2 (100 mL) were added
in aq 5% NaOH (10 mL) as described by
Garegg et al. [10]. The mixture was boiled
under reflux and vigorous stirring for 4 days,
allowed to cool, and the aqueous phase was
removed. The organic phase was dried
(Na2SO4), filtered, and concentrated. The
crude product was a mixture of mainly the
3-O- and the 2-O-benzyl derivatives, but mi-
nor amounts of the 2,3-di-O-benzyl derivative
and starting material were also present. To be
able to separate the 3-O-benzyl- from the
2-O-benzyl derivative on a silica gel column,
the mixture was acetylated (Ac2O in pyridine;
25 °C, overnight). After concentration, the
products were separated using 3:2 petroleum
ether–EtOAc as the eluent. The pooled frac-
tions containing the 3-O-acetyl derivative
were concentrated to dryness to give methyl
3-O-acetyl-2-O-benzyl-4,6-O-benzylidene-a-D-
galactopyranoside (10) contaminated with 12–
15% of the 2,3-di-O-benzyl derivative
according to 1H NMR. The mixture was
deacetylated with NaOMe in MeOH, neu-
tralised (Dowex 50 H+), and after evapora-
tion the residue was purified on a silica gel
column eluted with 1:2 petroleum ether–
1
found 415.1850 [M+H]+; H NMR (CDCl3):
l 7.55–7.24 (m, 10 H, Ar–H), 5.49 (s, 1 H,
PhCH), 5.28 (dd, 1 H, J2,310.5, J3,4 3.5 Hz,
H-3), 4.80 (d, 1 H, J1,2 3.5 Hz, H-1), 4.74 (d,
1 H, Jgem 12 Hz, PhCH2), 4.62 (d, 1 H, Jgem 12
Hz, PhCH2), 4.45 (dd, 1 H, J4,5 ꢀ1 Hz, H-4),
4.21 (dd, 1 H, J6a,6b 12.5, J5,6a 1.5 Hz, H-6a),
4.10 (dd, 1 H, H-2), 4.05 (dd, 1 H, J5,6b 1.5
Hz, H-6b), 3.71 (m, 1 H, H-5), 3.40 (s, 3 H,
OCH3), 2.08 (s, 3 H, OAc); 13C NMR
(CDCl3): l 170.7 (OAc), 138.3–126.2 (aro-
matic), 100.8 (PhCH), 99.2 (C-1), 74.3 (C-4),
73.6 (C-2), 73.5 (PhCH2), 70.9 (C-3), 69.3
(C-6), 62.1 (C-5), 55.6 (OCH3), 21.1 (OAc).
Methyl 3-O-[(R)-1-carboxyethyl]-h-D-gal-
actopyranoside (3).—Starting from 9 (0.80 g,
2.15 mmol), (S)-2-chloropropionic acid (0.49
g, 4.5 mmol) and NaH (1.54 g, 64 mmol)
using the same experimental and work-up
conditions as in the synthesis of 7, methyl
2-O-benzyl-4,6-O-benzylidene-3-O-[(R)-1-car-
boxyethyl]-a-D-galactopyranoside (11), con-
taminated with ca. 5% of methyl 2-O-benzyl-
4,6-O-benzylidene-3-O-[(S)-1-carboxyethyl]-
a-D-galactopyranoside (12), was obtained. Re-
crystallisation from toluene–petroleum ether
yielded 11 (0.63 g, 66%) as white crystals: mp
167.5–168.5 °C; [h]D +104° (c 1.0, CHCl3);

L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
163
boxyethyl–CH), 75.6 (C-3), 75.3 (C-4), 73.7
(PhCH2), 69.3 (C-6), 62.5 (C-5), 55.5 (OCH3),
18.8 (carboxyethyl–CH3).
Hydrogenolysis of 12 (103 mg, 232 mmol)
followed by purification and cation exchange,
in the same way as for compound 7, gave the
Na salt of 4 (65 mg, 98%) as a white amor-
phous powder; [h]D +118° (c 1.2, H2O);
HRMS: Calcd for C10H18O8+Na: 289.0899,
found 289.0844 [M+Na]+.
HRMS: Calcd for C24H28O8+H: 445.1862,
found 445.1857 [M+H]+; 1H NMR: l 9.79 (s,
1 H, COOH), 7.54-7.25 (m, 10 H, Ar-H), 5.57
(s, 1 H, PhCH), 4.84 (d, 1 H, Jgem 12 Hz,
PhCH2), 4.77 (d, 1 H, J1,2 3.5 Hz, H-1), 4.65 (d,
1 H, Jgem 12 Hz, PhCH2), 4.34 (dd, 1 H, J4,5
1
Hz, H-4), 4.26 (q, 1 H, J 7 Hz, carboxyethyl–
CH), 4.26 (dd, 1 H, J6a,6b 12.5, J5,6a 1.5 Hz,
H-6a), 4.07 (dd, 1 H, H-2), 4.07 (dd, 1 H, J5,6b
2 Hz, H-6b), 3.98 (dd, 1 H, J2,3 10, J3,4 3.5 Hz,
H-3), 3.62 (m, 1 H, H-5), 3.32 (s, 3 H, OCH3),
Methyl 3-O-[(R)-1-carboxyethyl]-h-D-man-
nopyranoside (5).—Starting from methyl 2-O-
benzyl-4,6-O-benzylidene-a-D-mannopyrano-
side [11] (1.04 g, 2.78 mmol), (S)-2-chloropro-
pionic acid (0.75 g, 6.9 mmol) and NaH (1.93
g, 80 mmol), using the same experimental and
work-up conditions as in the synthesis of 7,
crude methyl 2-O-benzyl-4,6-O-benzylidene-3-
O-[(R)-1-carboxyethyl]-a-D-mannopyranoside
(13) contaminated with 2% of methyl 2-O-ben-
zyl-4,6-O-benzylidene-3-O-[(S)-1-carboxy-
ethyl]-a-D-mannopyranoside (14) and some
aliphatic contaminants, was obtained as a
13
1.51 (d, 3 H, J 7 Hz, carboxyethyl–CH3); C
NMR: l 174.3 (COOH), 137.6-126.1 (aro-
matic), 100.8 (PhCH), 98.4 (C-1), 75.9 (C-3),
75.0 (C-2), 74.2 (carboxyethyl–CH), 73.8 (C-
4), 73.2 (PhCH2), 69.4 (C-6), 62.3 (C-5), 55.5
(OCH3), 19.2 (carboxyethyl–CH3).
Deprotection of 11 (105 mg, 236 mmol)
followed by purification and cation exchange,
in the same way as for compound 7, gave the
Na salt of 3 (64 mg, 94%) as a white amor-
phous powder; [h]D +183° (c 1.1, H2O);
HRMS: Calcd for C10H18O8+Na: 289.0899,
found 289.0910 [M+Na]+.
Methyl 3-O-[(S)-1-carboxyethyl]-h-D-galac-
topyranoside (4).—Starting from 9 (1.16 g,
3.11 mmol), (R)-2-chloropropionic acid (0.71
g, 6.5 mmol) and NaH (2.28 g, 95 mmol),
using the same experimental and work-up
conditions as in the synthesis of 7, methyl
2 - O - benzyl - 4,6 - O - benzylidene - 3 - O - [(S) -
1-carboxyethyl]-a-D - galactopyranoside (12),
contaminated with ꢀ2% of 11, was obtained.
Recrystallisation from toluene–petroleum
ether yielded 12 (1.15 g, 83%) as white crys-
tals; mp 138.5–139.5 °C; [h]D +71° (c 1.0,
CHCl3); HRMS: Calcd for C24H28O8+H:
445.1862, found 445.1886 [M+H]+; 1H
NMR: l 10.6 (s, 1 H, COOH), 7.55–7.25 (m,
10 H, Ar-H), 5.57 (s, 1 H, PhCH), 4.81 (d, 1
H, Jgem 12 Hz, PhCH2), 4.78 (d, 1 H, J1,2 3.5
Hz, H-1), 4.63 (d, 1 H, Jgem 12 Hz, PhCH2),
4.51 (bs, 1 H, H-4), 4.39 (q, 1 H, J 7 Hz,
carboxyethyl–CH), 4.23 (dd, 1 H, J6a,6b 12.5,
J5,6a 1.5 Hz, H-6a), 4.05 (dd, 1 H, J5,6b 2 Hz,
H-6b), 4.00 (m, 2 H, H-2, 3), 3.65 (bs, 1 H,
H-5), 3.39 (s, 3 H, OCH3), 1.52 (d, 3 H, J 7
Hz, carboxyethyl–CH3); 13C NMR: l 175.6
(COOH), 138.3–126.5 (aromatic), 101.2
(PhCH), 99.2 (C-1), 77.3 (C-2), 76.3 (car-
1
syrup in ꢀ95% yield as estimated from H
NMR. It was not possible to separate com-
pounds 13 and 14 on silica gel, but the aliphatic
contaminants were removed by passage
through a silica column (99:1, CHCl3–MeOH).
Purification of 13 from the isomer 14 by crys-
tallisation with different counter ions was tried
and column chromatography on Sephadex
LH-20 (3×25 cm), using EtOAc as eluent was
also tried, but without success. The analytical
sample of 13 was thus contaminated by 2% of
1
14 as determined by H NMR; [h]D +48° (c
1.1, CDCl3); HRMS: Calcd for C24H28O8+H:
445.1862, found 445.1899 [M+H]+; 1H
NMR: l 9.5 (s, 1 H, COOH), 7.5–7.2 (m, 10
H, Ar–H), 5.62 (s, 1 H, PhCH), 4.98 (d, 1 H,
Jgem 12 Hz, PhCH2), 4.82 (d, 1 H, Jgem 12 Hz,
PhCH2), 4.65 (d, 1 H, J1,2 1.5 Hz, H-1), 4.43 (q,
1 H, J 7 Hz, carboxyethyl–CH), 4.24 (dd, 1 H,
J6a,6b 10, J5,6a 4.5 Hz, H-6a), 4.17 (t, 1 H,
J3,4=J4,5 =10 Hz H-4), 4.11 (dd, H-2), 3.98
(dd, 1 H, J2,3 3.5 Hz, H-3), 3.86 (t, 1 H, J5,6b
10 Hz, H-6b), 3.76 (ddd, 1 H, H-5), 3.30 (s, 3
H, OCH3), 1.52 (d, 3 H, J 7 Hz, carboxyethyl–
CH3); 13C NMR: l 176.3 (COOH), 138.7–
126.0 (aromatic), 101.5 (PhCH), 101.1 (C-1),
79.4 (C-4), 77.5 (C-2), 76.8 (C-3), 76.3 (car-
boxyethyl–CH), 74.0 (PhCH2), 68.9 (C-6), 64.0
(C-5), 54.8 (OCH3), 19.2 (carboxyethyl–CH3).

164
L. Andersson, L. Kenne / Carbohydrate Research 313 (1998) 157–164
Hydrogenolysis and purification of 13 (128
76.3 (C-2), 75.5 (carboxyethyl–CH), 73.8
(PhCH2), 68.8 (C-6), 64.1 (C-5), 55.0 (OCH3),
19.1 (carboxyethyl–CH3).
mg, 288 mmol) was done on the isomeric
mixture (98:2) followed by purification and
cation exchange in the same way as for
compound 7. This gave the Na salt of 5 (75
mg, 90%) as a white amorphous powder; [h]D
+77° (c 1.0, H2O); HRMS: Calcd for
C10H18O8+Na: 289.0899, found 289.0898
[M+Na]+.
Hydrogenolysis of 14 (116 mg, 261 mmol)
was done on the isomeric mixture (94:6) fol-
lowed by purification and cation exchange in
the same way as for compound 7 to give the
Na salt of 6 (63 mg, 84%) as a white amor-
phous powder; [h]D +12° (c 1.0, H2O);
HRMS: Calcd for C10H18O8+H: 267.1079,
found 267.1083 [M+H]+.
Methyl 3-O-[(S)-1-carboxyethyl]-h-D-man-
nopyranoside (6).—Starting from methyl 2-O-
benzyl-4,6-O-benzylidene-a-D-mannopyrano-
side (1.03 g, 2.76 mmol), (R)-2-chloropropi-
onic acid (0.75 g, 6.9 mmol) and NaH (1.65 g,
69 mmol), using the same experimental and
work-up conditions as in the synthesis of 7,
crude 14, contaminated with 6% of 13 and
some unidentified aliphatic contaminants, was
Acknowledgements
This work was supported by grants from
the Swedish Natural Science Research Coun-
cil.
1
obtained in ꢀ80% yield, according to H
NMR. The aliphatic contaminants were re-
moved by chromatography on silica gel (99:1,
CHCl3–MeOH). It was not possible to sepa-
rate 14 from 13, so the analytical sample of 14
References
[1] B. Lindberg, Ad6. Carbohydr. Chem. Biochem., 48 (1990)
279–318.
[2] Y.A. Knirel, N.K. Kochetkov, Biochemistry (Moscow),
59 (1994) 1325–1383.
[3] R.J. Stack, Appl. En6iron. Microbiol., 54 (1988) 878–883.
[4] M. Andersson, S. Ratnayake, L. Kenne, L. Ericsson,
R.J. Stack, Carbohydr. Res., 246 (1993) 291–301.
[5] W.B. Severn, J.C. Richards, J. Am. Chem. Soc., 115
(1993) 1114–1120.
[6] M. Andersson, L. Kenne, R. Stenutz, G. Widmalm,
Carbohydr. Res., 254 (1994) 35–41.
[7] E. Fischer, K. Raske, Ber. Deutsch. Chem. Ges., 39
(1906) 3981–3995.
[8] P. Sinay¨, M.D.A Halford, M.S. Choudhary, P.H. Gross,
R.W. Jeanloz, J. Biol. Chem., 247 (1972) 391–397.
[9] N.K. Kochetkov, B.A. Dmitriev, L.V. Backinowsky,
Carbohydr. Res., 51 (1976) 229–237.
[10] P.J. Garegg, T. Iversen, S. Oscarson, Carbohydr. Res., 50
(1976) C12–C14.
[11] Y. Kondo, K. Noumi, S. Kitagawa, S. Hirano, Carbo-
hydr. Res., 123 (1983) 157–159.
1
contained 6% of 13 according to H NMR;
[h]D +1° (c 1.5, CHCl3); HRMS: Calcd for
C24H28O8+H: 445.1862, found 445.1883
1
[M+H]+; H NMR: l 8.5 (s, 1 H, COOH),
7.5–7.3 (m, 10 H, Ar–H), 5.62 (s, 1 H,
PhCH), 4.81 (d, 1 H, Jgem 12 Hz, PhCH2), 4.74
(d, 1 H, Jgem 12 Hz, PhCH2), 4.73 (d, 1 H, J1,2
1.5 Hz, H-1), 4.25 (dd, 1 H, J6a,6b 10, J5,6a 4,5
Hz, H-6a), 4.22 (t, 1 H, J4,5 10 Hz, H-4), 4.21
(q, 1 H, J 7 Hz, carboxyethyl–CH), 3.93 (dd,
1 H, J2,3 3.5, J3,4 10 Hz, H-3), 3.86 (dd, 1 H,
J5,6b 10 Hz, H-6b), 3.84 (d, 1 H, H-2), 3.79
(ddd, 1 H, H-5), 3.35 (s, 3 H, OCH3), 1.43 (d,
3 H, J 7 Hz, carboxyethyl–CH3); 13C NMR: l
174.1 (COOH), 137.8–126.1 (aromatic), 102.1
(PhCH), 100.3 (C-1), 77.9 (C-4), 77.2 (C-3),
[12] M.E. Evans, Carbohydr. Res., 21 (1972) 473–475.
.