Five monophosphates of methyl L-glycero-α-D-manno-heptopyranoside: Synthesis, hydrolysis and migrations

Article abstract of DOI:10.1016/S0008-6215(98)00037-8

From suitably protected methyl α-D-mannopyranosides five methyl L- glycero-α-D-manno-heptopyranosides were synthesized by the one-carbon-atom chain elongation at C-6 and converted to five monophosphates (1-5) having the PO(OH)₂ group at O-2, -3, -4, -6 and -7. Compounds 1-5 were exposed to acidic and basic hydrolytic conditions used in lipopolysaccharide analysis and the products and their proportion were determined. Under acidic conditions, besides hydrolysis of the glycoside, migrations and hydrolytic cleavage of the phosphate residue were found. Under basic conditions the phosphates were stable.

Full text of DOI:10.1016/S0008-6215(98)00037-8

Carbohydrate Research 307 (1998) 55±67
Five monophosphates of methyl l-glycero-ꢀ-d-
manno-heptopyranoside: synthesis, hydrolysis
and migrations
Barbara Grzeszczyka, Otto Holstb, Sven Muller-Loenniesb,
Aleksander Zamojskia,*
a Institute of Organic Chemistry, Polish Academy of Sciences, 01-224 Warsaw, Poland
b Forschungszentrum Borstel, D-23845 Borstel, Germany
Received 26 May 1997; accepted 20 December 1997
Abstract
From suitably protected methyl ꢀ-d-mannopyranosides ®ve methyl l-glycero-ꢀ-d-manno-hep-
topyranosides were synthesized by the one-carbon-atom chain elongation at C-6 and converted to
®ve monophosphates (1±5) having the PO(OH)2 group at O-2, -3, -4, -6 and -7. Compounds 1±5
were exposed to acidic and basic hydrolytic conditions used in lipopolysaccharide analysis and the
products and their proportion were determined. Under acidic conditions, besides hydrolysis of the
glycoside, migrations and hydrolytic cleavage of the phosphate residue were found. Under basic
conditions the phosphates were stable. # 1998 Elsevier Science Ltd. All rights reserved
Keywords: Bacterial sugars; Methyl l-glycero-ꢀ-d-manno-heptopyranoside monophosphate; Heptose synthesis
1. Introduction
group under, at least, basic or weakly acidic
hydrolytic conditions.
In continuation of the work on l-glycero-d-
manno-heptose (ld-manHepp) monophosphates
having the PO(OH)2 residue unambiguously loca-
ted at O-2, -3, -4, -6, and -7 [1] we extended the
investigation to a similar series of monophosphates
bound to methyl l-glycero-ꢀ-d-manno-heptopyr-
anoside. These compounds enabled the study of
phosphoryl group migration under typical condi-
tions used in lipopolysaccharide degradation. The
ꢀ-glycosidic methyl group eliminated the problems
connected with the presence of ꢀ,ꢁ-anomers and
eliminated also side reactions due to the free HO-1
Synthetic phosphates of l-glycero-d-manno-
heptose are scarcely known; only preparations
of ld-manHepp 1-phosphate [2], dl-manHepp
* Corresponding author.
0008-6215/98/$19.00 # 1998 Elsevier Science Ltd. All rights reserved
PII S0008-6215(98)00037-8

56
B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
7-phosphate (enantiomer of ld-manHepp) [3], 70-
(2-aminoethyl)phosphate [4] and 4- and/or 40-
phosphates of ꢀ-(1!3)-linked di-ld-manHepp [5]
have been described.
same reagent yielded 6b (Scheme 1), having a free
HO-6 group available for the Swern oxidation.
After oxidation, the aldehyde 7b was reacted with
benzyloxymethylmagnesium chloride (8, R5=Bn)
to yield methyl 2-O-allyl-3,4,7-tri-O-benzyl-l-gly-
cero-ꢀ-d-manno-heptopyranoside (10), methyl 2-O-
allyl-3,4,7-tri-O-benzyl-d-glycero-ꢀ-d-manno-hep-
topyranoside (14), and methyl 2-O-allyl-3,4,7-tri-
O-benzyl-d(l)-glycero-ꢁ-l-gulo-heptopyranoside
(15) (Table 1).
Compound 6c (Scheme 1) having O-3 blocked
with an allyl group and being a precursor of the
heptoside substrate for phosphorylation at O-3,
was obtained from methyl 6-O-trityl-ꢀ-d-manno-
pyranoside by stannylation with Bu2O followed by
allylation with allyl bromide, then benzylation with
benzyl bromide at O-2 and O-4, and detritylation.
Further steps consisted of a Swern oxidation to
aldehyde 7c and chain elongation with 8 (R5=Bn),
thus leading to methyl 3-O-allyl-2,4,7-tri-O-ben-
zyl-l-glycero-ꢀ-d-manno-heptopyranoside (16),
methyl 3-O-allyl-2,4,7-tri-O-benzyl-d-glycero-ꢀ-d-
manno-heptopyranoside (20), and methyl 3-O-allyl-
2,4,7-tri-O-benzyl-d(l)-glycero-ꢁ-l-gulo-heptopy-
ranoside (21) (Table 1).
2. Results
Synthesis.ÐThe synthesis of ®ve monophos-
phates of methyl l-glycero-ꢀ-d-manno-heptopyr-
anoside (1±5) was carried out essentially according
to the protocol employed previously [1], the main
di€erence being the replacement of benzyl ꢀ-d-
mannopyranoside as substrate by commercially
available methyl ꢀ-d-mannopyranoside (6a).
As a precursor of the heptoside substrate for
phosphorylation at O-4 methyl 2,3-di-O-benzyl-4-
O-(p-methoxybenzyl)-ꢀ-d-mannopyranoside (6d)
was obtained from methyl 4,6-O-(p-methoxy-
benzylidene)-ꢀ-d-mannopyranoside (22) by benzy-
lation and subsequent selective ring opening with
Me3SiCl±NaBH3CN. Chain elongation with 8
(R5=Bn) gave methyl 2,3,7-tri-O-benzyl-4-O-(p-
methoxybenzyl)-l-glycero-ꢀ-d-manno-heptopyrano-
side (23) and methyl 2,3,7-tri-O-benzyl-4-O-(p-
methoxybenzyl)-d(l)-glycero-ꢁ-l-gulo-heptopy-
ranoside (27) (Table 1). In this reaction formation
of the dd-manno stereoisomer was not detected.
Scheme 1.
Below are shortly described the syntheses of
speci®cally blocked methyl l-glycero-ꢀ-d-manno-
heptopyranosides permitting a ready access to
substrates for phosphorylation having OH groups
free at C-2, -3, -4, -6, and -7. In all chain elongation
reactions, besides the main product A of the ld-
manno con®guration, substantial amounts of the
dd-manno stereoisomer B as well as methyl hepto-
side C derived from C-5 ``inverted'' aldehyde [1]
were obtained. The general course of the syntheses
is presented in Scheme 1.
For the synthesis of the heptoside substrate for
phosphorylation at O-2, methyl exo-2,3:4,6-di-O-
benzylidene-ꢀ-d-mannopyranoside was used (9, cf.
Scheme 1 in ref. [1]). Selective opening of the 2,3-
O-benzylidene ring by reduction with LiAlH4±
AlCl3, followed by allylation and further selective
opening of the 4,6-O-benzylidene ring with the

B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
57
Table 1
Methyl heptosides via chain elongation: yields and HR MS data
b
Series
Methyl heptoside
No.
Yielda %
[ꢀ]d (c)
HR MS (LSIMS)
Calcd
Found
2P
3P
ld-manno
dd-manno
d(l)l-gulo
ld-manno
dd-manno
d(l)l-gulo
ld-manno
d(l)l-gulo
ld-manno
dd-manno
d(l)l-gulo
ld-manno
dd-manno
d(l)l-gulo
10
14
15
16
20d
21
23
27
28
30
31
32
36
37
32.0
15.8
12.4
36.3
11.5
9.0
28.6
20.2
40.7
16.5
15.2
34.5
20.8
13.8
+24 (1.2)
+55 (1.2)
+11 (1.3)
+25 (0.9)
n.d.e
+2.2 (0.9)
+13 (1.2)
+6 (1.1)
+18 (1.0)
+34 (1.0)
0 (1.3)
557.25152
557.25093
557.25093
557.25093
n.d.
557.25145
637.27921
637.27676
607.26715
649.27739
607.26943
557.25093
4P
6P
637.27773
607.26717
649.27773f
607.26717
557.25152
7P
+21 (1.0)
+3 (1.3)
a Isolated yield after repeated chromatography.
b In CHCl3 at 20‹3 ꢀC.
c [M+Na]+.
1
d Compound 20 was not analytically pure. Its identity was con®rmed by H NMR analysis.
e Not determined.
f Exact mass determination was made for the acetylated product.
the substrates 12, 18, 25, and 34, respectively, for
the selective phosphorylation. Obviously, hepto-
side 28 could be phosphorylated directly. The 13C
NMR data of the various products of the l-gly-
cero-ꢀ-d-manno series including the O-acetylated
derivatives of 12, 18, 25, 28, and 34 (13, 19, 26, 29,
and 35) are collected in Table 2.
For the phosphorylation of 12, 18, 25, 28, and 34
2-dimethylamino-5,6-benzo-1,3,2-dioxaphosphepane
(DMABDP, 38) was employed [1], and the pro-
ducts obtained were in situ oxidized with m-chloro-
peroxybenzoic acid to the corresponding 5,6-
With respect to the heptoside substrate for
phosphorylation at O-6 methyl 2,3,4,7-tetra-O-
benzyl-l-glycero-ꢀ-d-manno-heptopyranoside (28)
was obtained by chain elongation of known methyl
2,3,4-tri-O-benzyl-ꢀ-d-manno-hexodialdo-1,5-pyr-
anoside (7e), prepared from compound 6e, with 8
(R5=Bn). Heptoside 28 was accompanied by the
dd-stereoisomer 30 and the ``inverted'' product 31
(Table 1).
For the preparation of the substrate for phos-
phorylation at O-7, aldehyde 7e was reacted with
allyloxymethylmagnesium chloride (8, R5= All, cf.
Scheme 5 in ref. [1]). Besides the main product,
methyl 7-O-allyl-2,3,4-tri-O-benzyl-l-glycero-ꢀ-
d-manno-heptopyranoside (32) two stereoisomeric
products were also formed: methyl 7-O-allyl-2,3,4-
tri-O-benzyl-d-glycero-ꢀ-d-manno-heptopyrano-
side (36) and methyl 7-O-allyl-2,3,4-tri-O-benzyl-
d(l)-glycero-ꢁ-l-gulo-heptopyranoside(37)(Table1).
In order to obtain the various ®nal heptoside
substrates, ®rst the hydroxyl group at C-6 in 10,
16, 23, and 32 was benzylated to form 11, 17, 24,
and 33, respectively. Then, in a next step, the allyl
(or p-methoxybenzyl) group was removed to form
benzo-2-oxo-1,3,2-phosphepan-2-yl
derivatives
39, 40, 41, 42, and 43 (for 13C NMR data, see
Table 3). Exhaustive hydrogenation (H2, Pd-C)
yielded the ®ve methyl l-glycero-ꢀ-d-manno-hep-
topyranosides 1±5 having phosphoryl groups at O-
2, -3, -4, -6, and -7, respectively. Spectral data of
these products are collected in Tables 4 and 5.
Hydrolysis and migrations.ÐMigration of the
phosphoryl group in polyol monophosphate sys-
tems under acidic conditions is known since a long
time [6]. Investigations involved glycerol [7±9],
inositol [10], and nucleoside [11,12] phosphates.
The mechanism of the PO(OH)2 group migration
was carefully studied [13,14].
Although numerous syntheses of sugar phos-
phates have been described in the literature (e.g.
[15±22]), so far the migration of phosphate residues
was studied to a very limited extent. Apparently,

58
B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
Table 2
13C NMR data of protected methyl l-glycero-ꢀ-d-manno-heptopyranosidesa
Substituents
13C chemical shifts
No.
R1
R2
R3
R4
R5
C-1
C-2
C-3
C-4
C-5
C-6
C-7
10
11
12
13
16
17
18
19
23
24
25
26
28
29
32
33
34
35
All
All
H
Bn
Bn
Bn
Bn
All
All
H
Ac
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
pMB
pMB
H
H
Bn
Bn
Bn
H
Bn
Bn
Bn
H
Bn
Bn
Bn
H
Ac
H
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
All
All
H
99.45
99.04
100.39
98.71
99.39
98.97
97.87
98.99
99.33
99.23
99.15
99.27
99.34
98.91
99.33
98.95
99.12
99.05
74.47
74.94
67.94
68.37
74.46
74.97
78.30
75.73
74.59
75.58
75.00
74.99
74.51
73.90
74.56
74.27
74.30
73.98
80.03
80.40
80.59
78.57
79.90
80.30
72.07
73.89
80.26
81.19
79.94
77.74
80.23
80.48
80.24
80.63
80.34
80.54
74.18
74.13
74.80
74.82
74.14
74.16
76.05
73.24
74.01
75.42
66.17
67.81
74.23
73.77
74.21
74.19
74.22
73.84
70.70
71.24
70.57
70.93
70.70
71.27
70.42
71.08
70.80
71.83
71.35
69.37
70.76
69.43
70.70
71.20
73.09
70.99
67.89
74.10
73.57
73.69
67.92
74.12
75.05
74.89
67.94
74.41
73.75
73.72
67.93
68.79
67.85
74.92
75.83
74.15
71.44
70.15
69.98
69.98
71.06
70.15
70.12
70.03
71.49
69.95
69.65
69.33
71.44
67.83
71.32
69.88
62.52
62.87
Ac
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Ac
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Bn
Ac
a
Signals of acetyl, allyl, benzyl and methyl groups occurred at their normal positions and are omitted.
Table 3
13C NMR data of benzyl-protected, phosphorylated methyl l-glycero-ꢀ-d-manno-heptopyranosidesa
C-1
C-2
C-3
C-4
C-5
C-6
C-7
39
40
41
42
43
99.04
(2.8)b
99.16
73.01
(5.4)
76.82
78.34
(4.6)
78.86
73.63
71.16
75.23
70.05
69.88
69.02
73.89
(6.4)
74.27
(6.3)
74.91
71.12
75.01
73.84
98.94
99.93
99.11
74.35
73.81
73.90c
78.36
80.68
80.48
70.30
(5.5)
69.90
(5.6)
70.74
74.60
(5.6)
74.85
(7.6)
68.74
(1.7)
66.40
(5.7)
74.15c
a
Signals of carbon atoms belonging to benzyl groups and 5,6-benzo-2-oxo-1,3,2-dioxaphosphepane are omitted.
b 3
c
J
C,P
or 2JC,P
.
Can be interchanged.
the earliest report was by Levene and Raymond
[15], who, on phosphorylation of 5-blocked 1,2-O-
isopropylidene-ꢀ-d-xylofuranose with POCl3 and
pyridine followed by hydrolysis of the product with
acetic and sulfuric acids, obtained only d-xylose 5-
phosphate instead of the expected 3-phosphate.
Baddiley, Buchanan and Carss [18] found that
heating of d-ribitol 5-phosphate in 1 M HCl
(15 min, 100 ꢀC) caused migration of the PO(OH)2
group to positions 2 and 3. Pyridine-catalyzed
migration of the thiophosphoryl group in sugar ꢁ-
hydroxyphosphorodithioates was studied by
Michalska et al. [23]. More often no migration of
the phosphoryl group was noted [2,3,16,19,20,22].
Cawley and Letters [21] synthesized methyl ꢀ-d-
mannopyranoside 4- and 6-phosphates, and
found that under acidic (2 M HCl, 100 ꢀC, 2 h) or
alkaline (0.5 M NaOH, 100 ꢀC, 4 h) conditions
both compounds were stable and no migration of
the PO(OH)2 group was detected.

B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
59
Table 4
NMR data for methyl l-glycero-ꢀ-d-manno-heptopyranoside monophosphates 1±5 and their Na+ saltsa,b
Chemical shifts (ꢂ)
Phosphate
H-1/C-1 H-2/C-2/2-P
H-3/C-3/3-P
H-4/C-4/4-P H-5/C-5 H-6/C-6/6-P
H-7a/C-7/7-P H-7b
2P (1)
H+
4.984
100.48
100.7
4.370
74.40
74.10
0.81
3.854
71.20
71.39
3.895
67.10
67.26
3.598
72.00
72.06
4.063
69.53
69.56
3.745
63.73
63.76
3.745
c
Na+
H+
4.952
100.12
4.339
75.17
1.99
4.143
69.75
3.811
70.82
3.920
66.84
3.595
71.97
4.065
69.47
3.750
63.66
3.750
3.995
3.774
3.754
3.766
3.896
3.786
4.059
3.989
3P (2)
4P (3)
6P (4)
7P (5)
4.799
101.01
4.308
73.63
0.66
4.219
77.36
4.20
3.985
66.18
3.641
71.31
4.074
69.72
3.995
63.21
Na+
H+
4.840
100.92
4.102
69.82
4.019
65.83
3.641
71.82
4.094
69.61
3.774
63.65
4.782
101.49
3.968
70.52
3.943
71.18
4.386
71.53
1.41
4.317
72.39
3.65
3.704
71.00
4.082
69.25
3.754
63.29
Na+
H+
4.830
101.56
3.968
70.72
3.972
71.05
3.673
70.82
4.169
69.23
3.766
63.28
4.773
102.05
3.935
70.79
3.783
71.04
3.870
66.64
3.731
71.94
4.581
72.91
1.44
4.524
74.49
4.03
3.808
62.31
Na+
H+
4.826
101.99
3.946
70.64
3.852
71.30
3.989
66.71
3.712
71.17
3.911
61.55
4.780
101.80
3.928
70.79
3.771
71.67
3.868
66.85
3.628
71.51
4.216
68.49
4.059
66.57
1.04
3.989
67.17
2.50
Na+
4.797
101.80
3.945
70.75
3.791
71.61
3.889
66.73
3.647
71.28
4.220
68.00
a
Resonance signals of OCH3 groups occurred within ꢂ 3.38±3.42 (1H) and 54.98±55.69 (13C).
b The samples possessed the following pH: 2P (1): 1.3; Na+ salt: 5.3. 3P (2): 1.5; Na+ salt: 6.6. 4P (3): 1.4; Na+ salt: 6.0. 6P (4): 1.2;
Na+ salt: 5.9. 7P (5): 1.3; Na+ salt: 6.0.
c After KOH treatment of the sample.
sphere. The results of the hydrolyses were studied
by high-performance anion-exchange chromato-
graphy at high pH (HPAEC); note that the heptose
phosphates were stable under alkaline conditions.
The elution times of 1±5 on CarboPac PA-10 using
two concentration gradient systems and of ®ve l-
glycero-ꢀ-d-manno-heptopyranose phosphates (44±
48) obtained earlier [1] are collected in Table 6. It
must be noted that phosphates 1, 2, and 5 displayed
very similar elution times in both gradients. Also
In the present study the methyl l-glycero-ꢀ-d-
manno-heptopyranoside monophosphates 1±5 have
been exposed to di€erent hydrolytic conditions
often employed in the degradation of lipopoly-
saccharides (LPS): (D1) 1 M hydrochloric acid,
100 ꢀC, 2 h; (D2) 4 M tri¯uoroacetic acid, 100 ꢀC,
4 h; (D3) 4 M potassium hydroxide, 120 ꢀC, 16 h.
All reactions were performed in a nitrogen atmo-
phosphates 3 and 4 gave a common peak. A better
resolution was found for the free heptose phos-
phates: 44, 47, and 48 which gave nicely separated
peaks, whereas 45 and 46 formed a joint signal.
Monophosphates 1±5 were exposed to the
hydrolytic conditions D1±D3 using their free acids
and disodium salts. The results of all reactions and
proportions of products identi®ed are tabulated in

60
B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
Table 5
Table 7
Transformations of methyl l-glycero-ꢀ-d-manno-heptopyrano-
side 2-phosphate (1)
3
2JP,C and JP,C coupling constants of methyl l-glycero-ꢀ-d-
manno-heptopyranoside phosphates 1±5
13C-atom
Reaction conditions
Form
Products
C-2 C-3 C-4 C-5 C-6 C-7
1
44 45 48 49
2P (1) H+
5.1 5.4
5.1 5.4
3.7 4.4
5.5 5.8
2.2 3.3 1.9
5.6 6.4
D1
D2
D3
H+
Na+
H+
10.1 7.2 11.8 Ð 70.9
16.1 8.9 16.5 Ð 58.5
14.6 2.6 7.8 2.0 73.5
15.4 3.2 9.5 2.0 71.9
2P (1) H+ after KOH
2P (1) Na+
3P (2) H+
Na+
3P (2) Na+
4P (3) H+
H+ 100
Na+ 100
4P (3) Na+
6P (4) H+
4.3
Ð
6.1
5.7 5.3
4.2 4.0
6P (4) Na+
7P (5) H+
that the equatorial phosphate group was more
stable than the axial one. Only 10±15% of the
unchanged substrate remained in the mixtures. In
contrast, 100% of 1 remained unchanged when the
compound was exposed to alkaline conditions D3.
Detection of a minute amount of the 7-phos-
phate 48 in the D2 experiment may indicate that
the PO(OH)2 group is capable to migrate through
the whole carbon skeleton to form the most stable
primary heptose phosphate. Another explanation
might also be o€ered, namely, a direct esteri®ca-
tion of HO-7 by the released phosphoric acid. This
explanation could be also considered for the acidic
hydrolyses of 2 and 3 (vide infra). Therefore, a
blank experiment was performed consisting of
heating methyl l-glycero-ꢀ-d-manno-heptopyrano-
side with phosphoric acid in tri¯uoroacetic acid
under the D2 conditions. Besides the free heptose
49 resulting from hydrolysis of the methyl glyco-
side, not a trace of any heptose phosphate could be
detected with HPAEC.
The results of the hydrolysis of the 3-phosphate
(2, Table 8) were similar to those obtained for the
2-phosphate: 12±16% of the unchanged substrate,
two phosphorylated heptoses 44 and 45(46), in a
similar proportion as in the case of 1, large amounts
of demethylated and dephosphorylated heptose
49, and a minute amount of 7-phosphate 48. The
results from Tables 7 and 8 may indicate that 1 and
2 under acidic conditions undergo glycoside
8.5 4.7
7.6 4.1
7P (5) Na+
Tables 7±11. Under acidic conditions (D1 and D2)
for all phosphates three reactions have been noted:
(i) hydrolysis of the glycoside, (ii) migration of the
phosphate group, and (iii) hydrolysis of the phos-
phate. All phosphates were stable under alkaline
conditions, which was proven for phosphates 1±4
by 13C NMR analysis before and after the alkaline
treatment.
In the case of the 2-phosphate (1, Table 7)
hydrolysis of the glycoside and of the phosphate
were the main reactions under acidic conditions
with formation of the free heptose 49 as the dom-
inating product. The presence of ld-Hepp 2- and
3(4)-phosphates (44 and 45(46)) indicated that the
hydrolysis of the glycoside was faster than of the
phosphate monoester. Migration of the phosphate
moiety from O-2 to O-3(O-4), to form 45(46) in ca.
2±3:1 predominance over 44, demonstrated clearly
Table 6
Elution times of methyl l-glycero-ꢀ-d-manno-heptopyranoside
monophosphates 1±5 and of l-glycero-d-manno-heptopyra-
nose monophosphates 44±48 and l-glycero-d-manno-hepto-
pyranose 49 on CarboPac PA-10 (HPAEC)
Heptose phosphate
No.
aa
bb No.
b
2P
3P
4P
6P
7P
1
2
3
4
5
19.60c 12.53 44 20.67
19.67c 12.73 45 21.47e
20.40d 13.20 46 21.47e
20.40d 13.07 47 20.20
19.53c 12.53 48 18.87
Table 8
Transformations of methyl l-glycero-ꢀ-d-manno-heptopyrano-
side 3-phosphate (2)
49
4.67
Reaction conditions
Form
Products
a
Gradient a, 1±20% B (cf. Experimental).
Gradient b, 1±50% B.
2
44 45 48 49
b
c Mixture of 1, 2 and 5 could not be separated using gradients
a and b.
D1
D2
H+ 16.2 9.5 19
Ð 55.3
Na+ 13.3 7.4 16.5 Ð 62.7
H+ 14.9 3.5 8.8 2.6 70.1
Na+ 12.7 2.7 7.2 2.6 74.7
d
Mixture of 3 and 4 could not be separated using gradients a
and b.
e
ld-manHepp 3P (45) and 4P (46) could not be separated.

B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
61
Table 9
Transformations of methyl l-glycero-ꢀ-d-manno-heptopyrano-
side 4-phosphate (3)
Table 11
Transformations of methyl l-glycero-ꢀ-d-manno-heptopyra-
noside 7-phosphate (5)
Reaction conditions
Form
Products
Reaction conditions
Form
Products
47 48 49
10 84.7 5.3
9.6 87.9 2.5
11.4 81.6 7.0
9.6 80.5 9.9
3
44 45 48 49
5
D1
D2
D3
H+
Na+
H+
Ð
Ð
Ð
6.7 19.3 11.2 62.8
10.8 33.9 12.1 43.2
3.2 15.5 11.1 70.2
D1
D2
D3
D4
H+
Na+
H+
Ð
Ð
Ð
Ð
Na+ 3.3 3.3 16.3 11.6 65.5
Na+
H+
Na+ 100
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
H+ 100
Na+ 100
H+ 100
Na+ 100
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
Ð
hydrolysis in the ®rst step, followed by migration
of the phosphate group and, eventually, by hydro-
lytic cleavage of the PO(OH)2 group. A formal
objection remains thatÐbecause of peak over-
lappingÐsubstrate peaks in Tables 7 and 8 may
represent, in fact, mixtures of 1, 2 and 5. While this
can be a real possibility (1 and 2, cf. Table 6), the
presence of 5 does not appear to be very likely as
this phosphate disappears completely under the
acidic hydrolysis conditions employed (cf.
Table 11).
to induce the migration of the PO(OH)2 group
towards the primary hydroxyl group. Thus, the
location of the phosphoryl group at the primary
alcohol center is especially favored.
On acidic (D1 and D2) treatment of 4, demethy-
lation and migration were the main processes, and
the 6- and 7-phosphates 47 and 48 were formed
with the latter in
a distinct predominance
(Table 10). Under alkaline conditions 6-phosphate
4 remained unchanged.
4-Phosphate 3 remained unchanged under alka-
line conditions and disappeared almost completely
after acidic treatment (Table 9). The main product
was, as above, free heptose 49. Most interesting
was the presence of three (four) heptose phos-
phates: 2- (44), 3- (45) (and/or 4- (46)), and 7-
phosphate (48). This means that the phosphoryl
residue may migrate in both directions: towards O-
2 and O-7, whereby the last direction appears to be
preferred.
Most interestingly was the behaviour of the 7-
phosphate 5 (Table 11). Whereas it was stable
under weakly acidic (D4: 1% acetic acid, 100 ꢀC,
2 h) and alkaline (D3) conditions, in tri¯uoroacetic
acid (D2) it underwent a complete demethylation
with formation of over 80% of heptose 7-phos-
phate 48 and ca. 10% of 6-phosphate 47 and free
heptose 49. Formation of 6-phosphate 47 from 5
demonstrated that back migration from O-7 to O-6
was possible, and an equilibrium 47:48 = ca. 1:8
was established. Location of the phosphate residue
at the primary hydroxyl position is thermo-
dynamically preferred. The reason of this stability
might be of steric nature; MM calculations [25] of
6- and 7-phosphates pointed at lower strain energy
of the latter.
1
Hydrogenation of 28 led to 6-phosphate 4. H
and 13C NMR spectra recorded soon after isola-
tion of 4 pointed to a pure substance (cf. Table 4).
However, 4 on standing in water solution iso-
merized partially to 5; after 2 weeks a mixture was
formed containing ca. 62% of 4 and 38% of 5. The
amount of 5 increased gradually on further
standing. It appears that the acidity of 4 (higher
than phosphoric acid: pH 1.3, cf. [24]) is sucient
3. Conclusions
Table 10
Transformations of methyl l-glycero-ꢀ-d-manno-heptopyrano-
side 6-phosphate (4)
Phosphates 1±5 are stable at room temperature
and their phosphoryl groupsÐwith the exception
of 4Ðdo not undergo spontaneous migration. This
was proven by recording NMR spectra immedi-
ately after the synthesis and after several days.
From the data in Tables 7±11 several conclu-
sions can be drawn. The results of the reactions
performed on free acids and on disodium salts
Reaction conditions
Form
Products
47 48
1.0 20.5 63.5 15.0
4
49
D1
D2
D3
H+
H+
Ð
9.7 80.9 9.4
Ð
H+ 100
Ð
Ð

62
B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
remainÐfor obvious reasonsÐthe same within the
experimental error. Treatment of heptose mono-
phosphates with strong alkali results neither in
degradation nor in migration of the phosphate
group. Under the harsh acidic conditions employed
the methyl glycoside and phosphate ester are
cleaved to a substantial extent. For 1, 2, and 3 the
free heptose is the main product of hydrolysis. The
methyl group at the anomeric center is removed
faster than the phosphate function. This permits to
follow the migrations of the PO(OH)2 group. It is
important to note that not only the ``1,2-cis''
migrations (between O-2 and O-3, or O-6 and O-7)
but also between O-3 and O-4 (``1,2-trans'') and
between O-4 and O-6 (probably via a dioxaphos-
phorinane intermediate) are possible. It is well
known that for the 1,2-migrations two mechanistic
pathways are possible: through a cyclic inter-
mediate or through a direct, H+-intermediated
phosphate transfer [13,14]. We suppose that in case
of heptose phosphates both pathways must be
involved. In neither case a transient cyclic phos-
phate was observed.
CH2Cl2, dried, and concentrated to dryness. The
residue was puri®ed by chromatography with 4:1
hexane±ether to yield the allylated product (9.89 g,
99%); [ꢀ]d +51.3ꢀ (c 2.1, CHCl3); 1H NMR
(C6D6): ꢂ 5.38 (s, 1 H, PhCH), 4.66 (d, 1 H, J1,2
1.6 Hz, H-1), 4.42 (t, 1 H, H-4), 4.19 (dd, 1 H, J6a,5
4.8, J6a,6b 10.2 Hz, H-6a), 4.07 (dd, 1 H, J3,2 3.2, J3,4
9.9 Hz, H-3), 3.86 (ddd, 1 H, J5,4 9.8, J5,6a 5.0 Hz,
H-5), 3.76 (dd, 1 H, H-2), 3.69 (t, 1 H, H-6b).
Methyl 2-O-allyl-3,4-di-O-benzyl-a-d-mannopyr-
anoside (6b).ÐTo a solution of methyl 2-O-allyl-3-
O-benzyl-4,6-O-benzylidene-ꢀ-d-mannopyranoside
(5.8 g) in CH2Cl2 (100 mL) was added LiAlH4
(2.26 g), and a solution of AlCl3 (6.8 g) in ether
(20 mL) was added dropwise. The mixture was
boiled under re¯ux for 45 min, whereupon the
excess of the hydride was decomposed with EtOAc
(10 mL) and water (10 mL). The product was
extracted with ether, and the organic extract was
dried and concentrated to dryness. The residue was
chromatographed with 4:1 hexane±EtOAc to yield
6b (5.6 g, 96%); [ꢀ]d +44.9ꢀ (c 1.2, CHCl3); H
1
NMR (CDCl3): ꢂ 4.71 (d, 1 H, J1,2 1.8 Hz, H-1),
3.75 (dd, 1 H, J2,3 2.6 Hz, H-2), 3.60 (ddd, 1 H,
J5,6a 2.8, J5,6b 4.9, J5,4 9.6 Hz, H-5).
4. Experimental
Methyl 4,6-O-(p-methoxybenzylidene)-a-d-manno-
pyranoside (22).ÐA solution of methyl ꢀ-d-man-
nopyranoside (19.4 g) in Me2SO (120 mL) was stir-
red with 4A molecular sieves for 20 min, then p-
methoxybenzaldehyde (16.3 g), ZnCl2 (15.7 g), and
p-toluenesulfonic acid (2.07 g) were added. After 6
days of stirring at room temperature the mixture
was neutralized with aq 25% ammonia (57 mL),
®ltered through Celite, poured into water, extrac-
ted with CHCl3, and the organic layer was dried
and concentrated. The residue was chromato-
graphed with 9:1 hexane±EtOAc to yield methyl
2,3:4,6-di-O-(p-methoxybenzylidene)-ꢀ-d-manno-
The general methods, unless otherwise indicated,
were the same as in the previous paper [1]. HPAEC
was performed as described [1] with the modi®ca-
tion that a column of CarboPac PA-10 (4Â240 mm)
was used; elutions were carried out with linear
concentration gradients of either 1±20% B (eluent
A containing 1 M NaOAc, gradient a) or 1±50%
B (gradient b) in 0.1 M NaOH (eluent A) at a ¯ow
rate of 1 mL/min. Methyl 2,3,4-tri-O-benzyl-ꢀ-d-
mannopyranoside [26], methyl 2,3,4-tri-O-benzyl-
ꢀ-d-manno-hexodialdo-1,5-pyranoside [26], methyl
3-O-benzyl-4,6-O-benzylidene-ꢀ-d-mannopyrano-
side [27], methyl 3-O-allyl-6-O-trityl-ꢀ-d-manno-
pyranoside [28], methyl exo-2,3:4,6-di-O-iso-
propylidene-ꢀ-d-mannopyranoside (9) [29] were
obtained according to literature methods.
pyranoside (13.4 g, 31%) {[ꢀ]d
8.6ꢀ (c 0.5,
CHCl3), exo:endo mixture 1:1; Anal. Calcd for
C23H26O8: C, 64.17; H, 6.09. Found: C, 64.00; H,
6.19} and 22 (6.8 g, 22%); [ꢀ]d +59.8ꢀ (c 1.6,
1
CHCl3); H NMR (200 MHz) (CDCl3): ꢂ 5.50 (s, 1
Methyl 2-O-allyl-3-O-benzyl-4,6-O-benzylidene-
a-d-mannopyranoside.ÐTo a solution of methyl 3-
O-benzyl-4,6-O-benzylidene-ꢀ-d-mannopyranoside
[27] (9.0 g) in DMF (260 mL) cooled to 0 ꢀC was
added 50% oil suspension of NaH (3.48 g). After
20 min a solution of allyl bromide (6 mL) in DMF
(6 mL) was added dropwise. After 3 h the excess of
the hydride was decomposed with MeOH, the
mixture was poured into ice-water, extracted with
H, MeOC6H4CH), 4.69 (s, 1 H, H-1), 3.8±4.3 (m, 6
H, H-2,3,4,5,6a,6b), 3.79 (s, 3 H, MeOC6H4CH),
3.37 (s, 3 H, OMe). Anal. Calcd for C15H20O7: C,
57.68; H, 6.46. Found: C, 56.98; H, 6.80.
Methyl 2,3-di-O-benzyl-4-O-(p-methoxybenzyl)-
a-d-mannopyranoside (6d).ÐCompound 22 (13.9 g)
was conventionally benzylated (NaH, BnBr, DMF,
0 ꢀC) to yield the 2,3-di-O-benzyl derivative (20.7 g,
95.1%). A solution of this compound (9.8 g) in

B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
63
acetonitrile (160 mL) was stirred with 3A molecular
sieves for 10 min. The solution was cooled (ice-
water), NaBH3CN (4.81 g) was added, and after a
few minutes a solution of trimethylsilyl chloride
(9.7 mL) in acetonitrile (20 mL) was added drop-
wise. After completion of the reduction (TLC) the
mixture was ®ltered through Celite, poured into dil
aq NaHCO3, extracted with CH2Cl2, dried, and the
organic layer was concentrated to dryness. The
residue was chromatographed with 9:1!4:1 hexane±
ether. The ®rst compound eluted was methyl 2,3-di-
O-benzyl-6-O-(p-methoxybenzyl)-ꢀ-d-mannopyr-
(BnOCH2Cl) and stirring was continued for 2 h.
The temperature was then lowered to 30 ꢀC
(AllOCH2MgCl) or 20 ꢀC (BnOCH2MgCl) and a
solution of aldehyde (4 mmol) in abs THF (10 mL)
was added dropwise. The mixture was stirred at
these temperatures for 2 h and was slowly brought
to room temperature while stirring for another 12 h.
A cold (0 ꢀC) aq solution of NH4Cl (150 mL) was
added and the products were extracted with ether.
The ether extract was dried, concentrated to dry-
ness, and the residue was chromatographed with
hexane±EtOAc (9:1 or 85:15) to give derivatives of
methyl heptosides in the order: ``inverted'' ®rst, fol-
lowed by ld-, and dd-manHepp as the last. The
assignment of con®guration to the stereoisomeric
methyl heptosides was based on regularities
anoside (3.13 g, 20%); [ꢀ]d 6.3ꢀ (c 1.7, CHCl3); H
1
NMR (CDCl3): ꢂ 4.77 (d, 1 H, J1,2 1.7 Hz, H-1),
4.03 (t, 1 H, J4,3&J4,5&9.4 Hz, H-4). Anal. Calcd
.
for C29H34O7 1/2H2O: C, 69.16; H, 7.01. Found:
C, 69.11; H, 7.11. The second one that eluted was
6d (5.2 g, 34%); [ꢀ]d +32ꢀ (c 1.2, CHCl3); 1H
NMR (C6D6): ꢂ 4.70 (d, 1 H, J1,2 1.8 Hz, H-1), 4.27
(t, 1 H, J4,5 = J4,3 = 9.6 Hz, H-4), 4.04 (dd, 1 H,
J3,2 3.1 Hz, H-3). Anal. Calcd for C29H34O7: C,
70.42; H, 6.93. Found: C, 70.25; H, 7.03.
Synthesis of protected derivatives of methyl l-gly-
cero-a-d-manno-heptopyranoside. General method.Ð
A solution of oxalyl chloride (1 mL, 11 mmol) in
CH2Cl2 (25 mL) was cooled to 50 Ä 60 ꢀC and
a solution of Me2SO (1.7 mL, 22 mmol) in CH2Cl2
(5 mL) was added slowly. After 5 min a solution of
suitably protected methyl mannopyranoside with a
free HO-6 group (10 mmol) in CH2Cl2 (20 mL) was
added dropwise. Stirring at 60 ꢀC was continued
for 1 h, whereupon Et3N (7 mL, 50 mmol) was
added. After 5 min of stirring, the mixture was
allowed to attain room temperature. To the solu-
tion was added water (50 mL), and the mixture was
extracted with CH2Cl2. The organic extract was
dried (MgSO4) and concentrated to dryness. The
remaining oil was additionally dried by azeotropic
distillation (3±4 times) with small portions of anh
benzene and was used in the next step without any
additional puri®cation.
1
observed in H NMR coupling constants of H-6,
-7a and -7b. For derivatives of ld-manHepp the
coupling constants were: J6,5 1.0±1.4 Hz, J7a,6 6.3±
7.2 Hz, J7b,6 5.8±6.3 Hz, and J7a,7b 9.3±9.6 Hz. For
dd-manHepp: J6,5 2.0±2.1 Hz, J7a,6 4.5±4.7 Hz, J7b,6
6.7±7.2 Hz, and J7a,7b 10.0±10.7 Hz. It must be
added that the con®guration of benzyl glycosides of
ld-manHepp and dd-manHepp obtained in the
same way was earlier [30,31] con®rmed by prepara-
tion of diethyl dithioacetals displaying identical
physical data as the original compounds. The con-
®guration of methyl 2,3,4-tri-O-benzyl-d-glycero-ꢀ-
d-manno-heptopyranoside, obtained by deacetyla-
tion and deallylation of 36, was independently con-
®rmed by CD [32]. The l-gulo con®guration of the
``inverted'' methyl heptosides follows from 1H
NMR coupling constants within the pyranoside
ring, i.e. J1,2 8.2 Hz.
Methyl 2-O-allyl-3,4,7-tri-O-benzyl-d(l)-gly-
cero-b-l-gulo-heptopyranoside (15).Ð1H NMR
(CDCl3): ꢂ 4.62 (d, 1 H, J1,2 8.2 Hz, H-1), 3.82 (dd,
1 H, J5,6 9.0, J5,4 1.2 Hz, H-5), 3.79 (t, 1 H, J3,2 3.4,
J3,4 3.5 Hz, H-3), 3.75 (dd, 1 H, H-4), 3.70 (dd, 1 H,
J7a,6 3.1, J7a,7b 9.6 Hz, H-7a), 3.62 (dd, 1 H, J7b,6
5.4 Hz, H-7b), 3.50 (dd, 1 H, H-2); 13C NMR
(CDCl3): ꢂ 101.7 (C-1), 76.14, 74.56, 73.98, 72.45,
and 68.01 (C-2,3,4,5,6), 71.12 (C-7). HR MS
(LSIMS): Table 1.
To dry magnesium turnings (0.384 mg, 16 mmol)
was added, under freshly distilled THF (1.5 mL),
sublimed HgCl2 (50 mg), and a few drops of neat
pure alkoxymethyl chloride, freshly distilled before
use, were added while lowering the temperature
Methyl 2-O-allyl-3,4,7-tri-O-benzyl-l-glycero-
a-d-manno-heptopyranoside
(10).Ð1H NMR
to
15 ꢀC (for allyloxymethyl chloride) or to
(CDCl3): ꢂ 4.68 (d, 1 H, J1,2 1.7 Hz, H-1), 4.10 (t, 1
H, J4,5 9.5 Hz, H-4), 3.86 (dd, 1 H, J3,2 3.1, J3,4
9.5 Hz, H-3), 3.72 (dd, 1 H, H-2), 3.66 (dd, 1 H,
J7a,6 7.4, J7a,7b 9.4 Hz, H-7a), 3.66 (dd, 1 H, J5,6
0 Ä 5 ꢀC (for benzyloxymethyl chloride). When
the formation of the Grignard reagent has started,
the remaining amount of alkoxymethyl chloride
(16 mmol) in THF (3 mL) was added at
18 Ä 20 ꢀC (AllOCH2Cl) or at 0 Ä 5 ꢀC
1.0 Hz, H-5), 3.57 (dd, 1 H, J7b,6 5.9 Hz, H-7b); 13
NMR (CDCl3): Table 2. HR MS (LSIMS): Table 1.
C

64
B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
Methyl 2-O-allyl-3,4,7-tri-O-benzyl-d-glycero-
(CDCl3): ꢂ 101.9 (C-1), 76.07, 74.84, 73.69, 72.43,
and 68.03 (C-2,3,4,5,6); HR MS (LSIMS): Table 1.
Methyl 2,3,4,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (28).Ð1H NMR (CDCl3):
ꢂ 4.70 (d, 1 H, H-1), 4.17 (dt, 1 H, H-6), 4.17 (t, 1
H, J4,5 9.6 Hz, H-4), 3.88 (dd, 1 H, J3,2 3.1, J3,4
9.4 Hz, H-3), 3.77 (dd, 1 H, J1,2 1.9 Hz, H-2), 3.56
(dd, 1 H, J7a,6 7.2, J7a,7b 9.5 Hz, H-7a), 3.65 (bd, 1
H, H-5), 3.57 (dd, 1 H, J7b,6 6.0 Hz, H-7b); 13C
NMR (CDCl3): Table 2. HR MS (LSIMS): Table 1.
The d-glycero-d-manno stereoisomer (30) was
characterized as its 6-O-acetyl derivative.
Methyl 6-O-acetyl-2,3,4,7-tetra-O-benzyl-d-
glycero-a-d-manno-heptopyranoside.Ð1H NMR
(CDCl3): ꢂ 5.56 (ddd, 1 H, J6,5 2.0, J6,7a 4.6 Hz, H-
6), 4.71 (bs, 1 H, H-1), 3.98 (t, 1 H, H-4), 3.86 (dd,
1 H, J3,4 9.0, J3,2 3.0 Hz, H-3), 3.80 (dd, 1 H, J5,4
9.9 Hz, H-5), 3.76 (dd, 1 H, J7a,7b 10.7 Hz, H-7a),
3.74 (dd, 1 H, H-2), 3.68 (dd, 1 H, J7b,6 7.2, H-7b);
13C NMR (CDCl3): ꢂ 98.69 (C-1), 80.22 (C-3),
74.74, 74.66, 71.76, and 71.67 (C-2,4,5,6), 68.18 (C-
7). HR-MS (LSIMS): Table 1.
Methyl 7-O-allyl-2,3,4-tri-O-benzyl-d(l)-gly-
cero-b-l-gulo-heptopyranoside (37).Ð1H NMR
(CDCl3): ꢂ 4.70 (d, 1 H, J1,2 8.2 Hz, H-1), 3.80 (dd,
1 H, J5,6 9.0, J5,4 0.9 Hz, H-5), 3.71 (t, 1 H, J3,4
3.8 Hz, H-3), 3.70 (dd, 1 H, H-4), 3.66 (dd, 1 H,
J7a,6 3.1, J7a,7b 9.7 Hz, H-7a), 3.57 (dd, 1 H, J7b,6
5.7 Hz, H-7b), 3.56 (dd, 1 H, J2,3 2.7 Hz, H-2); 13C
NMR (CDCl3): ꢂ 101.9 (C-1), 76.10, 74.88, 73.76,
72.56, and 67.94 (C-2,3,4,5,6), 71.10 (C-7). HR MS
(LSIMS): Table 1.
Methyl 7-O-allyl-2,3,4-tri-O-benzyl-l-glycero-
a-d-manno-heptopyranoside (32).Ð1H NMR
(CDCl3): ꢂ 4.71 (bs, 1 H, H-1), 4.17 (t, 1 H, H-4),
4.14 (bt, 1 H, H-6), 3.88 (dd, 1 H, J3,2 3.1, J3,4
9.5 Hz, H-3), 3.78 (dd, 1 H, J2,1 1.9 Hz, H-2), 3.64
(dd, 1 H, J5,6 1.0, J5,4 9.8 Hz, H-5), 3.60 (dd, 1 H,
J7a,6 7.3, J7a,7b 9.5 Hz, H-7a), 3.52 (dd, 1 H, J7b,6
6.0 Hz, H-7b); 13C NMR: Table 2. HR MS
(LSIMS): Table 1.
a-d-manno-heptopyranoside (14).Ð1H NMR
(CDCl3): ꢂ 4.68 (d, 1 H, J1,2 2.1 Hz, H-1), 3.94 (t, 1
H, J2,3 2.5 Hz, H-2), 3.67 (dd, 1 H, J7a,6 6.7, J7a,7b
10.1 Hz, H-7a), 3.60 (dd, 1 H, J7b,6 3.1, H-7b); 13C
NMR (CDCl3): ꢂ 99.09 (C-1), 80.25 (C-3), 76.07,
74.40, 71.99, and 71.01 (C-2,4,5,6), 70.82 (C-7).
Methyl 3-O-allyl-2,4,7-tri-O-benzyl-d(l)-gly-
cero-b-l-gulo-heptopyranoside (21).Ð1H NMR
(CDCl3): ꢂ 4.62 (d, 1 H, J1,2 8.2 Hz, H-1), 3.78 (dd,
1 H, J5,6 9.0, J5,4 1.2 Hz, H-5), 3.71 (dd, 1 H, J4,5
1.4, J4,3 3.7 Hz, H-4), 3.71 (dd, 1 H, J7a,6 3.1, J7a,7b
9.8 Hz, H-7a), 3.63 (t, 1 H, J3,2 3.3 Hz, H-3), 3.63
(dd, 1 H, J7b,6 5.5 Hz, H-7b), 3.53 (dd, 1 H, H-2);
13C NMR (CDCl3): ꢂ 101.8 (C-1), 75.85, 74.88,
73.79, 72.48, and 68.00 (C-2,3,4,5,6), 71.17 (C-7).
HR MS (LSIMS): Table 1.
Methyl 3-O-allyl-2,4,7-tri-O-benzyl-l-glycero-
a-d-manno-heptopyranoside (16).Ð1H NMR
(CDCl3): ꢂ 4.69 (bs, 1 H, H-1), 4.12 (t, 1 H, H-4),
3.76 (dd, 1 H, J3,2 3.1, J3,4 9.3 Hz, H-3), 3.74 (dd,
J2,1 1.8 Hz, H-2), 3.65 (dd, 1 H, J7a,6 7.2, J7a,7b
9.5 Hz, H-7a), 3.63 (dd, 1 H, J5,6 1.0, J5,4 9.8 Hz,
H-5), 3.57 (dd, 1 H, J7b,6 6.0 Hz, H-7b); 13C NMR
(CDCl3): Table 2. HR MS (LSIMS): Table 1.
Methyl 3-O-allyl-2,4,7-tri-O-benzyl-d-glycero-ꢀ-
d-manno-heptopyranoside (20) was not isolated in
analytically pure state.
Methyl 2,3,7-tri-O-benzyl-4-O-(p-methoxybenzyl)-
d(l)-glycero-b-l-gulo-heptopyranoside (27).Ð1H
NMR (CDCl3): ꢂ 4.67 (d, 1 H, J1,2 8.2 Hz, H-1),
4.02 (ddd, 1 H, J6,5 8.7, J6,7a 3.1, J6,7b 5.4 Hz, H-6),
3.81 (dd, 1 H, J5,4 1.0 Hz, H-5), 3.97 (t, 1 H, J3,4
3.6 Hz, H-3), 3.62 (dd, 1 H, J7b,7a 9.7 Hz, H-7b),
3.53 (dd, 1 H, J2,3 2.7 Hz, H-2); 13C NMR (CDCl3):
ꢂ 101.8 (C-1), 76.10, 74.81, 73.25, 72.41, and 68.06
(C-2,3,4,5,6), 71.16 (C-7). HR MS (LSIMS):
Table 1.
Methyl 2,3,7-tri-O-benzyl-4-O-(p-methoxybenzyl)-
l-glycero-a-d-manno-heptopyranoside (23).Ð1H
NMR (CDCl3): ꢂ 4.62 (d, 1 H, J1,2 1.7 Hz, H-1),
4.09 (dt, 1 H, J6,5 1.1 Hz, H-6), 4.08 (t, 1 H, J4,5
9.6 Hz, H-4), 3.79 (dd, 1 H, J3,2 3.1, J3,4 9.5 Hz, H-
3), 3.70 (dd, 1 H, H-2), 3.58 (dd, 1 H, J7a,6 7.3,
J7a,7b 9.4 Hz, H-7a), 3.55 (dd, 1 H, H-5), 3.49 (dd, 1
H, J7b,6 5.9 Hz, H-7b); 13C NMR (CDCl3): Table 2.
HR MS (LSIMS): Table 1.
The d-glycero-ꢀ-d-manno (36) stereoisomer was
characterized as its 6-O-acetyl derivative
Methyl 6-O-acetyl-7-O-allyl-2,3,4-tri-O-benzyl-
d-glycero-a-d-manno-heptopyranoside.Ð1H NMR
(CDCl3): ꢂ 5.50 (ddd, 1 H, J6,5 2.1, J6,7a 4.8, J6,7b
6.9 Hz, H-6), 4.72 (bs, 1 H, H-1), 3.99 (t, 1 H, H-4),
3.87 (dd, 1 H, J3,2 3.1, J3,4 9.1 Hz, H-3), 3.79 (dd, 1 H,
J5,4 9.9 Hz, H-5), 3.75 (dd, 1 H, J2,1 1.8 Hz, H-2), 3.74
(dd, 1 H, J7a,7b 10.7 Hz, H-7a), 3.64 (dd, 1 H, H-7b);
13C NMR (CDCl3): ꢂ 90.06 (C-1), 80.64 (C-3), 75.14,
75.06, 72.12, and 72.04 (C-2,4,5,6), 68.34 (C-7).
Methyl 2,3,4,7-tetra-O-benzyl-d(l)-glycero-b-l-
gulo-heptopyranoside (31).Ð1H NMR (CDCl3): ꢂ
4.68 (d, 1 H, J1,2 8.2 Hz, H-1), 4.04 (ddd, 1 H, J6,7a
3.1, J6,7b 5.3, J6,5 9.0 Hz, H-6), 3.82 (dd, 1 H, H-5),
3.65 (dd, 1 H, J7a,7b 9.5 Hz, H-7b); 13C NMR

B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
65
Benzylation at O-6 in methyl l-glycero-a-d-
manno-heptopyranosides 10, 16, 23, and 32.ÐTo a
cooled (0 ꢀC) solution of methyl heptopyranoside
(1 mmol) in abs DMF (10 mL) was added NaH
(2 mmol, 50% oil suspension). The suspension was
stirred for 10 min, then a solution of benzyl chlor-
ide (2 mmol) in abs DMF (1 mL) was added. Stir-
ring was continued for 8 h, whereupon the excess of
the hydride was decomposed with MeOH and the
mixture was poured into ice-water. The solution
was extracted with ether, and the organic layer was
dried and concentrated to dryness. The residue was
puri®ed by chromatography in 95:5 or 9:1 hexane±
EtOAc.
Methyl 2-O-allyl-3,4,6,7-tetra-O-benzyl-l-glycero-
a-d-manno-heptopyranoside (11).ÐYield: 87.6%;
[ꢀ]d +45ꢀ (c 1.1, CHCl3); 13C NMR (CDCl3):
Table 2. HR MS (LSIMS): C39H44O7 +Na+
[M+Na]+; Calcd: 647.2985. Found: 647.2951.
Methyl 3-O-allyl-2,4,6,7-tetra-O-benzyl-l-glycero-
a-d-manno-heptopyranoside (17).ÐYield: 85.5%;
[ꢀ]d +34ꢀ (c 1.1, CHCl3); 13C NMR (CDCl3):
Table 2. HR MS (LSIMS): C39H44O7+Na+
[M+Na]+; Calcd: 647.2985. Found: 647.2981.
Methyl 2,3,6,7-tetra-O-benzyl-4-O-(p-methoxyben-
zyl)-l-glycero-a-d-manno-heptopyranoside (24).Ð
Yield: 79.2%; [ꢀ]d +23ꢀ (c 0.53, CHCl3); 13C
NMR (CDCl3): Table 2. HR MS (LSIMS): C44H48
O8 +Na+ [M+Na]+; Calcd: 727.3247. Found:
727.3255.
Methyl 7-O-allyl-2,3,4,6-tetra-O-benzyl-l-glycero-
a-d-manno-heptopyranoside (33).ÐYield: 97.2%;
[ꢀ]d +38ꢀ (c 1.4, CHCl3); 13C NMR (CDCl3):
Table 2. HR MS (LSIMS): C39H44O7 +Na+
[M+Na]+; Calcd: 647.2985. Found: 647.2981.
De-allylation of 11, 17 and 33.ÐTo a solution
of heptopyranoside derivative 11, 17 or 33
(0.65 mmol) in a mixture of EtOH (9 mL), benzene
(3 mL), and water (1 mL), was added 1,4-diamino-
bicyclo[2,2,2]octane (15 mg) and the solution was
heated to 80 ꢀC. Wilkinson's catalyst (42 mg) was
added, the mixture was boiled under re¯ux for 3 h,
then left at room temperature overnight. The mix-
ture was ®ltered and the ®ltrate was concentrated
under lowered pressure. The remaining oil was
dissolved in 15:1 acetone±water, and to the solu-
tion were added HgO (149 mg) and HgCl2
(187 mg). The suspension was stirred (0.5 h) at
room temperature, ®ltered, and concentrated, and
the residue was dissolved in ether. The solution was
washed with aq 50% KI, aq 10% NaHSO3, and aq
1% NaHCO3, then dried and concentrated. The
products were chromatographed with 7:3 light
petroleum±ether.
Methyl 3,4,6,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (12).ÐYield: 90.9%; [ꢀ]d
+59ꢀ (c 0.97, CHCl3); 13C NMR (CDCl3): Table 2.
HR MS (LSIMS): C36H40O7+Na+ [M+Na]+;
Calcd: 607.2672. Found: 607.2672. 2-O-Acetyl
derivative (13).Ð[ꢀ]d +30ꢀ (c 1.0, CHCl3); HR MS
(LSIMS): C38H42O8+Na+ [M+Na]+; Calcd:
649.2777. Found: 649.2762. 13C NMR (CDCl3):
Table 2.
Methyl 2,4,6,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (18).ÐYield: 99.9%; [ꢀ]d
+20ꢀ (c 1.3, CHCl3); 13C NMR (CDCl3): Table 2.
HR MS (LSIMS): C36H40O7 +Na+ [M+Na]+;
Calcd: 607.2672. Found: 607.2672. 3-O-Acetyl
derivative (19). [ꢀ]d +20ꢀ (c 1.8, CHCl3). HR MS
(LSIMS): C38H42O8+Na+ [M+Na]+; Calcd:
649.2777. Found: 649.2774. 13C NMR (CDCl3):
Table 2.
Methyl 2,3,4,6-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (34).ÐYield: 76.0%; [ꢀ]d
+38ꢀ (c 1.3, CHCl3), Lit. [33]: [ꢀ]d +37.8ꢀ (c 1.1,
CHCl3); 13C NMR (CDCl3): Table 2. 7-O-Acetyl
derivative (35).Ð[a]d +32ꢀ (c 1.6, CHCl3); 13C
NMR (CDCl3): Table 2.
Methyl 2,3,6,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (25).ÐTo a solution of 24
(466 mg) in 9:1 acetonitrile±water (10 mL) was
added cerium-ammonium nitrate (743 mg), and the
mixture was stirred at room temperature. After 1 h
(TLC in 2:1 hexane±ether) the mixture was diluted
with CH2Cl2 (15 mL) and poured into sat aq
NaHCO3 (30 mL). The organic extract was dried
and concentrated to dryness. The residue was
chromatographed with 3:1 hexane±ether to yield 25
(336 mg, 87%); [ꢀ]d 5.1ꢀ (c 1.84, CHCl3); 13C
NMR (CDCl3): Table 2. HR MS (LSIMS):
C36H40O7+Na+ [M+Na]+; Calcd: 607.2672.
Found: 607.2672. 4-O-Acetyl derivative (26).Ð[ꢀ]d
+16ꢀ (c 1.3, CHCl3). 13C NMR (CDCl3): Table 2.
HR MS (LSIMS): C38H32O8+Na+ [M+Na]+;
Calcd: 649.2777. Found: 649.2780.
Phosphorylation of methyl tetra-O-benzyl-l-gly-
cero-a-d-manno-heptopyranosides 12, 18, 25, 28,
and 34 using 2-dimethylamino-5,6-benzo-1,3,2-
dioxaphosphepane (38). General method.ÐTo a
solution of methyl tetra-O-benzyl-l-glycero-ꢀ-d-
manno-heptopyranoside (585 mg, 1 mmol) in abs
CH2Cl2 (10 mL) was added tetrazole (210 mg,
3 mmol), and the mixture was stirred. After 10 min
a solution of 38 (253 mg, 1.2 mmol) in CH2Cl2

66
B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
(1 mL) was added and the mixture was again
stirred. After 2 h at room temperature (TLC
in 1:1 hexane±ether) the solution was cooled to
O-phosphate (1).ÐYield: 99.8%; [ꢀ]d +27.5ꢀ (c
0.52, H2O).
Methyl l-glycero-a-d-manno-heptopyranoside 3-
O-phosphate (2).ÐYield: 94.6%; [ꢀ]d +40.4ꢀ (c 0.5,
H2O).
Methyl l-glycero-a-d-manno-heptopyranoside 4-
O-phosphate (3).ÐYield: 98.3%; [ꢀ]d +60.1ꢀ (c
0.22, H2O).
Methyl l-glycero-a-d-manno-heptopyranoside 6-
O-phosphate (4).ÐYield: 97.3%; [ꢀ]d +49.5ꢀ (c 0.5,
H2O).
60 ꢀC, and
a solution of m-ClC6H4CO3H
(1.8 mmol) in CH2Cl2 (5 mL) was added. The
reaction was completed in 15 min (TLC in 1:1
hexane±EtOAc), and the mixture was diluted with
EtOAc (20 mL), washed with aq 1% NaHCO3 and
aq 1% NaHSO3. The organic layer was dried and
concentrated to dryness. The residue was chroma-
tographed with 4:1 hexane±EtOAc to yield pure
phosphates.
Methyl l-glycero-a-d-manno-heptopyranoside 7-
O-phosphate (5).ÐYield: 98.9%; [ꢀ]d +32ꢀ (c 0.8,
H2O).
Methyl 2-O-(5,6-benzo-2-oxo-1,3,2-dioxaphos-
phepan-2-yl)-3,4,6,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (39).ÐYield: 92.1%; oil;
[ꢀ]d 10.6ꢀ (c 6.85, CHCl3); 13C NMR (CDCl3):
Table 3. HR MS (LSIMS): C44H47PO10+Na+
[M+Na]+; Calcd: 789.2805. Found: 789.2780.
Methyl 3-O-(5,6-benzo-2-oxo-1,3,2-dioxaphos-
phepan-2-yl)-2,4,6,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (40).ÐYield: 73.2%; oil;
[ꢀ]d +18ꢀ (c 1.1, CHCl3); 13C NMR (CDCl3):
Table 3. HR MS (LSIMS): C44H47PO10+Na+
[M+Na]+; Calcd: 789.2805. Found: 789.2780.
Methyl 4-O-(5,6-benzo-2-oxo-1,3,2-dioxaphos-
phepan-2-yl)-2,3,6,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (41).ÐYield: 88.1%; oil;
[ꢀ]d +11ꢀ (c 1.0, CHCl3); 13C NMR (CDCl3):
Table 3. HR MS (LSIMS): C44H47PO10+Na+
[M+Na]+; Calcd: 789.2805. Found: 789.2788.
Methyl 6-O-(5,6-benzo-2-oxo-1,3,2-dioxaphos-
phepan-2-yl)-2,3,4,7-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (42).ÐYield: 85.2%; oil;
[ꢀ]d 3.5ꢀ (c 1.1, CHCl3); 13C NMR (CDCl3):
Table 3. HR MS (LSIMS): C44H47PO10+Na+
[M+Na]+; Calcd: 789.2805. Found: 789.2795.
Methyl 7-O-(5,6-benzo-2-oxo-1,3,2-dioxaphos-
phepan-2-yl)-2,3,4,6-tetra-O-benzyl-l-glycero-a-d-
manno-heptopyranoside (43).ÐYield: 88.2%; oil;
[ꢀ]d +16ꢀ (c 0.91, CHCl3); 13C NMR (CDCl3):
Table 3. HR MS (LSIMS): C44H47PO10+Na+
[M+Na]+; Calcd: 789.2805. Found: 789.2788.
Hydrogenation of methyl O-(5,6-benzo-2-oxo-1,3,
2-dioxaphosphepan-2-yl)-tetra-O-benzyl-l-glycero-
a-d-manno-heptopyranosides.ÐTo a solution of
phosphate 39 (40, 41, 42 or 43) (421 mg, 0.5 mmol)
in 9:3 EtOH±EtOAc was added 10% Pd-C (380 mg)
and the suspension was hydrogenated overnight.
Filtration through Celite and concentration of the
®ltrate left a foam. The 1H, 13C and 31P NMR data
of the phosphates are collected in Table 4.
Acknowledgements
This work was ®nanced in part by grant no. 2
P303 124 04 from the State Committee for Scien-
ti®c Research (KBN) and by the Deutsche For-
schungsanstalt fur Luft- und Raumfahrt e. V.
(grant POL 150-96). We thank Mrs. Regina Engel
for expert technical assistance.
References
[1] B. Grzeszczyk, O. Holst, and A. Zamojski, Carbo-
hydr. Res., 290 (1996) 1±15.
[2] M. Teuber, R.D. Bevill, and M.J. Osborn, Bio-
chemistry, 7 (1968) 3303±3308.
[3] P. Szabo, J. Chem. Soc., Perkin Trans. I, (1974)
920±923.
[4] G.J.P.H. Boons, P. Hoogerhout, J.T. Poolman,
G.A. van der Marel, and J.H. van Boom, Bioorg.
Med. Chem. Lett., 1 (1991) 303±308.
[5] K. Ekelof and S. Oscarson, J. Carbohydr. Chem.,
14 (1995) 299±315.
[6] O. Bailly, Ann. Chim. (Paris), 6 (1916) 96±154,
215±278.
[7] M.C. Bailly, Compt. rend., 208 (1939) 443±445;
Bull. Soc. Chim. Fr., (1942) 314±350, 405±438.
[8] P.E. Verkade, J.C. Stoppelenburg, and W.D.
Cohen, Rec. Trav. Chim. Pays-Bas, 59 (1940) 886±
892.
[9] E. Charga€, J. Biol. Chem., 144 (1942) 455±458.
[10] Y. Watanabe in Atta-ur-Rahman (Ed.), Selective
reactions and total synthesis of inositol phos-
phates, Studies in Natural Products Chemistry, Vol.
18, Elsevier, Amsterdam, 1996, pp 391±456.
[11] M. Oivanen, R. Schnell, W. P¯eiderer, and H.
Lonnberg, J. Org. Chem., 56 (1991)3623±3628 and
literature cited therein.
Methyl l-glycero-a-d-manno-heptopyranoside 2-

B. Grzeszczyk et al./Carbohydrate Research 307 (1998) 55±67
67
[12] M. Sekine, H. Tsuruoka, S. Iimura, H. Kusuoku,
and T. Wada, J. Org. Chem., 61 (1996) 4087±4100.
[13] W.D. Fordham and J.H. Wang, J. Am. Chem.
Soc., 89 (1967) 4197±4203.
[23] M. Michalska, E. Brzezinska, and P. Lipka, J. Am.
Chem. Soc., 113 (1991) 7945±7951.
[24] W.D. Kumler and J.J. Eiler, J. Am. Chem. Soc., 65
(1943) 2355±2361.
[14] S.L. Buchwald, D.H. Pliura, and J.R. Knowles, J.
Am. Chem. Soc., 106 (1984) 4916±4922.
[15] P.A. Levene and A.L. Raymond, J. Biol. Chem.,
107 (1934) 75±83.
[16] J. Baddiley, J.G. Buchanan, and L. Szabo, J.
Chem. Soc., (1954) 3826±3832.
[17] W.E. Harvey, J.J. Michalski, and A.R. Todd, J.
Chem. Soc., (1951) 2271±2278.
[18] J. Baddiley, J.G. Buchanan, and B. Carss, J. Chem.
Soc., (1957) 1869±1876.
[19] P. Szabo and L. Szabo, J. Chem. Soc., (1960) 3758±
3762 and 3762±3768.
[20] A. Danan, M. Mondange, S.R. Sarfati, and P.
Szabo, J. Chem. Soc., Perkin Trans. I, (1982) 1275±
1282.
[25] PC Model, Serena Software, Bloomington, IN
47402.
[26] H.B. Boren, K. Eklind, P.J. Garegg, B. Lindberg, and
A. Pilotti, Acta Chem. Scand., 26 (1972) 4143±4146.
[27] M.A. Nasheed, Carbohydr. Res., 60 (1978) 200±205.
[28] A.J. Varma and C. Schuerch, J. Org. Chem., 46
(1981) 799±803.
[29] Z. Szurmai, L. Balatoni, and A. Liptak, Carbo-
hydr. Res., 254 (1994) 301±309.
[30] K. Dziewiszek and A. Zamojski, Carbohydr. Res.,
150 (1986) 163±171.
[31] B. Grzeszczyk and A. Zamojski, Carbohydr. Res.,
262 (1994) 49±57.
[32] I. Akritopoulou-Zanze, K. Nakanishi, H.
Stepowska, B. Grzeszczyk, A. Zamojski, and N.
,
[21] T.N. Cawley and R. Letters, Carbohydr. Res., 19
(1971) 373±382.
[22] Ch. Berlind, S. Oscarson, and G. Widmalm, Car-
bohydr. Res., 263 (1994) 173±180.
Berova, Chirality, 9 (1997) 699±712.
[33] G.J.P.H. Boons, R. Steyger, M. Overhand, G.A.
van der Marel, and J.H. van Boom, J. Carbohydr.
Chem., 10 (1991) 995±1007.