Synthesis of β-D-glucopyranuronosylamine in aqueous solution: Kinetic study and synthetic potential

Article abstract of DOI:10.1016/j.carres.2011.08.018

A systematic study of the synthesis of β-D-glucopyranuronosylamine in water is reported. When sodium d-glucuronate was reacted with ammonia and/or volatile ammonium salts in water a mixture of β-D-glucopyranuronosylamine and ammonium N-β-D-glucopyranuronosyl carbamate was obtained at a rate that strongly depended on the experimental conditions. In general higher ammonia and/or ammonium salt concentrations led to a faster conversion of the starting sugar into intermediate species and of the latter into the final products. Yet, some interesting trends and exceptions were observed. The use of saturated ammonium carbamate led to the fastest rates and the highest final yields of β-D-glucopyranuronosylamine/carbamate. With the exception of 1 M ammonia and 0.6 M ammonium salt, after 24 h of reaction all tested protocols led to higher yields of β-glycosylamine/carbamate than concentrated commercial ammonia alone. The mole fraction of α-D-glucopyranuronosylamine/carbamate at equilibrium was found to be 7-8% in water at 30 °C. Concerning bis(β-D-glucopyranuronosyl)amine, less than 3% of it is formed in all cases, with a minimum value of 0.5% in the case of saturated ammonium carbamate. Surprisingly, the reaction was consistently faster in the case of sodium d-glucuronate than in the case of D-glucose (4-8 times faster). Finally, the synthetic usefulness of our approach was demonstrated by the synthesis of three N-acyl-β-D-glucopyranuronosylamines and one N-alkylcarbamoyl-β-D- glucopyranuronosylamine directly in aqueous-organic solution without resorting to protective group chemistry.

Full text of DOI:10.1016/j.carres.2011.08.018

Carbohydrate Research 346 (2011) 2384–2393
Contents lists available at SciVerse ScienceDirect
Carbohydrate Research
journal homepage: www.elsevier.com/locate/carres
Synthesis of b-D-glucopyranuronosylamine in aqueous solution: kinetic
study and synthetic potential
⇑
Ali Ghadban, Luca Albertin , Rédéo W. Moussavou Mounguengui, Alexandre Peruchon, Alain Heyraud
Centre de Recherches sur les Macromolécules Végétales (CERMAV-CNRS), BP53, 38041 Grenoble cedex 9, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 22 March 2011
Received in revised form 16 August 2011
Accepted 18 August 2011
Available online 25 August 2011
A systematic study of the synthesis of b-D-glucopyranuronosylamine in water is reported. When sodium
D
-glucuronate was reacted with ammonia and/or volatile ammonium salts in water a mixture of b-
D-
glucopyranuronosylamine and ammonium N-b- -glucopyranuronosyl carbamate was obtained at a rate
D
that strongly depended on the experimental conditions. In general higher ammonia and/or ammonium
salt concentrations led to a faster conversion of the starting sugar into intermediate species and of the
latter into the final products. Yet, some interesting trends and exceptions were observed. The use of sat-
urated ammonium carbamate led to the fastest rates and the highest final yields of b-D-glucopyranuron-
osylamine/carbamate. With the exception of 1 M ammonia and 0.6 M ammonium salt, after 24 h of
Keywords:
Uronic acid
D-Glucose
reaction all tested protocols led to higher yields of b-glycosylamine/carbamate than concentrated com-
Glycosylamine
Glucuronidation
Glycomonomer
mercial ammonia alone. The mole fraction of
was found to be 7–8% in water at 30 °C. Concerning bis(b-
it is formed in all cases, with a minimum value of 0.5% in the case of saturated ammonium carbamate.
Surprisingly, the reaction was consistently faster in the case of sodium -glucuronate than in the case
of -glucose (4–8 times faster). Finally, the synthetic usefulness of our approach was demonstrated by
the synthesis of three N-acyl-b- -glucopyranuronosylamines and one N-alkylcarbamoyl-b- -glucopyra-
a
-
D-glucopyranuronosylamine/carbamate at equilibrium
D
-glucopyranuronosyl)amine, less than 3% of
N-Acyl-glycosylamine
D
D
D
D
nuronosylamine directly in aqueous–organic solution without resorting to protective group chemistry.
Ó 2011 Elsevier Ltd. All rights reserved.
1
. Introduction
probes.20,21 Beginning in 1986 with the pioneering work of Kochet-
kov and collaborators, four original protocols have been de-
1
4
Uronic acids are monocarboxylic acids formally derived from
scribed for the synthesis of b-glycopyranosylamines in aqueous
1
7,22–25
aldoses by replacement of the hydroxymethyl group CH
2
OH with
or aqueous methanol solutions (Table 1).
They are all based
1
a carboxy group. In nature, they are found in polysaccharides ful-
filling diverse biological and structural functions such as glycos-
aminoglycans (e.g., heparin, hyaluronan, and chondroitin), and
on the use of ammonia and/or volatile ammonium salts, and have
found widespread application in the derivatization of hexoses,
6-deoxyhexoses, and oligosaccharides of different chain
2
13,18,19,21
homoglycuronans (e.g., alginates and pectins). In order to incorpo-
length.
rate uronic acids into glycoconjugates, it would be advantageous to
selectively functionalize their reducing end without resorting to
protective group chemistry, which tends to be rather cumbersome
The main advantage of these aqueous-based methods resides in
their applicability to unprotected and/or charged carbohydrates.
Nevertheless, the considerable amount of salt used, the labor-con-
suming procedures needed to remove it, and the formation of
diglycosylamine restricts their scope in preparative synthesis. Sur-
prisingly, a detailed study on the formation of glycosylamines in
3
in the case of monosaccharides and exceedingly time consuming
4
in the case of oligoglycuronans. A possible solution could be the
transformation of unprotected uronic acids into the corresponding
glycosylamines directly in water.
2
6
aqueous solution is lacking and only two papers claim the prep-
2
2
Glycosylamines have already been used as intermediates in the
aration of glycuronosylamines in aqueous or aqueous methanolic
solution while providing precious little details (only the salt with
carbamic acid was isolated).
5
–7
synthesis of a number of glycoconjugates,
such as glycopep-
8
–14
11,15–17
15,18,19
tides,
surfactants,
glycopolymers,
and N-glycan
In order to palliate to this dearth of information, we have car-
ried out a systematic study of the synthesis of glycuronosylamines
in aqueous solution. In particular, we tried to verify whether such
transformations could be conveniently performed, and to identify
the experimental conditions leading to the maximum yield in the
shortest reaction time, and with the smallest amount of reagents.
⇑
Corresponding author. Tel.: +33 04 76 03 76 25; fax: +33 04 76 54 72 03.
E-mail addresses: luca.albertin@cermav.cnrs.fr, luca_albertin@yahoo.com
(
L. Albertin).
Affiliated with Université Joseph Fourier, and member of the Institut de Chimie
Moléculaire de Grenoble.
0
008-6215/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2011.08.018

A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
2385
Table 1
Experimental protocols reported in the literature for the synthesis of glycosylamines in aqueous solution
a
Method
[carb]
0
(M)
[NH
3
]
0
(M)
Salt
[salt]
0
(M)
T (°C)
React. time
Substrate
GlcNAc
Yield (%)
Kochetkov14
60.2
0.2
0
NH
NH
HCO
HCO
Satd (ꢀ3.6)
b
b
30
42
6 d
36 h
80
100
4
3
Lubineau1
7,26
ꢀ16 M
0
0.2
4
3
D
D
D
-Glc
Gallop22
Likhosherstov24,25
60.06
0.8
(NH
Satd (ꢀ3.3)
3.2
ꢀ25
c
5 d
60
81e
4 2
) CO
3
-GlcA
-GlcA
d
ꢀ7.5 M
NH
2
CO
2
NH
4
20
48 h
Whenever possible, the exact conditions used for uronic acids are listed.
a
Note: [s]
0
indicates the initial concentration of species ‘s’.
Solubility in water: NH HCO , 284 g/kg at 30 °C; (NH CO
Room temperature in the original paper.
b
c
4
3
)
4 2
3
, 320 g/L at 20 °C.
d
e
NH
3 3
15 M/CH OH 1:1.
Only the salt with carbamic acid was isolated.
In a previous communication27 we reported that the reaction of so-
dium -glucuronate with ammonia and/or volatile ammonium
salts in water yields the expected b- -glucopyranuronosylamine
and ammonium N-b- -glucopyranuronosyl carbamate for long
2. Results and discussion
2.1. Kinetic study
D
D
D
reaction times (ꢀ24 h), whereas intermediate samples contain a
considerable amount of transient species (Scheme 1). From NMR
and MS experiments we could establish that two such species (6
Sodium
D-glucuronate (1) was reacted with ammonia and/or
volatile ammonium salts (NH HCO or NH CO NH ) in water
according to 18 different protocols at 30 and 40 °C (Scheme 1).
For comparison, some protocols were also applied to -glucose.
4
3
2
2
4
and 9) are the
are precursors to the formation of
and ammonium N- ,b- -glucopyranuronosyl carbamate. To date,
a
anomer of the main products, whereas the others
D
a
/b- -glucopyranuronosylamine
D
The exact experimental conditions used and the final composition
of the gross products are summarized in Table 2 (see also the
Experimental part for the meaning of protocol numbering). Sam-
ples were drawn at preset reaction times, frozen in liquid nitrogen
and freeze-dried overnight to eliminate water and most of the
salts. No further purification was performed, and all reported anal-
yses refer to the gross products obtained this way. In order to mon-
itor the time course of the reaction, individual samples were
a
D
the exact structures of species 5, 7, 8, 10, and 11 are unknown,
other than the fact that 5 and 7 are N-glycosyl carbamates. Based
1
on the H NMR assignments made in that study, we developed a
protocol for quantifying of the different compounds taking part
to the reaction and used it to carry out a kinetic study of the same
transformation under different experimental conditions. For com-
parison, some of experimental conditions tested were also applied
redissolved in cold D O to afford clear solutions of pD ꢀ9 that were
2
1
to
D-glucose. We now report the results of this quantitative study
immediately analyzed by H NMR spectroscopy. Spectra were ac-
together with a few examples of the synthetic potential of our
approach.
quired at 278 K in order to inhibit hydrolysis of the product and
to prevent the peak of residual HDO from interfering with
integration.
(
N-glycosyl carbamates)
5
, 7
CO , NH
2
3
+
-
+
-
Na OOC
Na OOC
O
O
2
CO , NH3
HO
HO
HO
HO
8
, 10, 11
OH
OH
NH
2
NHCO2- NH4+
9
2
6
NH
4
X
2
H O
NH X
4
H O
2
+
-
Na OOC
+
-
+
-
O
Na OOC
Na OOC
NH X
CO , NH3
2
4
O
O
HO
HO
HO
HO
HO
-
+
HO
NH2
NHCO NH
2 4
OH
OH
H O
2
OH
OH
1
3
-
-
-
2
X = OH , HCO , NH CO
2
3
2
NH3
NH3
O
+
-
Na OOC
COO-Na+
O
H
N
HO
HO
OH
OH
OH
HO
4
Scheme 1. Reactions taking place during the synthesis of b-D
-glucopyranuronosylamine 2 in aqueous solution.27

2
386
A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
Table 2
Summary of the ammination experiments performed in this study
a
Entry
Protocolb
Substrate
T (°C)
[NH
3
]
0
(M)
Salt
[salt]
0
(M)
[salt]
0
/[subst.]
0
Highest yield (%)
Time (h)
24
b-carbNHXc
By-productsd
e
e
1
2
3
4
5
6
7
8
9
A.0.S
GlcA
Glc
30
30
40
30
30
40
30
30
30
30
—
—
—
1
5
5
5
8.8
14.5
14.5
—
—
—
—
—
—
—
—
1
5
5
5
8.8
14.5
14.5
14.5
14.5
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
—
4
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
HCO
3
3
3
3
3
3
3
3
3
3
Satd (ꢀ3.6)
Satd (ꢀ3.6)
ꢀ18
ꢀ18
>18
3
1
1
3
3
1
1
75 (19)
64 (14)
78 (22)
67 (21)
80 (34)
75 (22)
84 (17)
86 (17)
84 (28)
82 (22)
69 (47)
84 (51)
87 (45)
89 (47)
88 (48)
89 (23)
87 (6)
87 (29)
79 (22)
80 (32)
75 (19)
84 (12)
86 (17)
86 (31)
81 (25)
82 (78)
33 (32)
11
10
9
f
4
4
4
4
4
4
4
4
4
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
168
e
GlcA
GlcA
GlcA
GlcA
GlcA
GlcA
GlcA
Glc
GlcA
GlcA
GlcA
GlcA
GlcA
GlcA
Glc
GlcA
GlcA
GlcA
GlcA
GlcA
GlcA
GlcA
Glc
Satd (>3.6)
0.60
7.5
f
A.1.06
A.5.02
33
21
15
16
15
13
13
13
8
8
8
8
8
f
0.20
0.20
0.60
0.60
0.20
0.20
1.0
2.0
3.0
4.0
5.0
Satd (>5.4)
Satd (>5.4)
Satd (>5.4)
0.60
0.20
0.20
32.5
9.5
f
A.5.06
A.9.06
A.14.02
33
33
33
f
f
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
168
g
B.0.10
B.0.20
B.0.30
B.0.40
B.0.50
B.0.S
ꢀ30
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
CO
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
2
NH
4
4
4
4
4
4
4
4
4
4
4
4
4
4
4
5
24
24
24
24
24
5.5
96
1.5
g
ꢀ30
10
15
20
25
>27
>27
>27
3
1
1
3
3
g
ꢀ30
g
ꢀ30
g
ꢀ30
e
30
30
40
30
30
40
30
30
30
30
30
30
9
e
e
13
11
19
15
13
14
12
12
14
16
33
f
B.1.06
B.5.02
33
f
32.5
24
33
B.5.06
B.9.06
B.14.02
0.60
0.60
0.20
0.20
—
—
f
33
f
1
1
—
—
33
f
168
f
X.14.00
GlcA
Glc
48
f
—
48
a
Note: [s]
0
indicates the initial concentration of species ‘s’. Unless otherwise specified, [carb]
The numbering of each protocol reveals the experimental conditions used. Hence A.1.06 indicates that the salt used was NH
CO NH , ‘X’ indicates that no salt was added and ‘S’ stands for ‘saturated’.
0
= 0.20 M.
b
4
3
HCO (‘A’), and that the initial concentration
of ammonia and salt was 1 and 0.6 M, respectively. In this context ‘B’ indicates NH
2
2
4
c
Total glycosylamine plus N-glycosyl carbamate; the parenthetical value indicates the yield of free glycosylamine.
Total by-products 4–11.
Solubility in water/NH
End of reaction.
d
e
f
4 3 2 2 4
HCO , 284 g/kg at 30 °C; NH CO NH , 423 g/L at 20 °C.
g
Ambient temperature.
Figure 1 shows the evolution of the proton spectrum with time
during the reaction of sodium -glucuronate (1) with 5 M ammonia
and 0.6 M ammonium carbamate, as well as the assignment of the
peaks used for integration. From these spectra it is clear that, with
the exception of 8 and 11, at least one diagnostic peak per each
species can be accurately integrated.
b
subtracting the integral of anomeric proton of 1 at 4.63 ppm. Also,
D
the contribution of species 8 and 11 was combined in the last factor
since neither can be quantified individually. The mole fraction x of
each species (with the exception of 8 and 11) is then calculated
from:
i
d
i
d
A =n
S
The mole fraction of each compound was then calculated from
the proton spectra on the basis of the assignment previously re-
i
x ¼
ð2Þ
2
7
1
ported. To this end, we assumed that each signal in the H NMR
spectrum is due to a CH group, that is, that there is no rearrangement
of C–H bonds during the reaction. This idea is supported by the
Ammonium N-b-
D
-glucopyranuronosyl carbamate (3) forms
from the reaction of 2 with the CO
of bicarbonate or carbamate anions.
2
liberated by the decomposition
19,21
Since it decomposes fairly
2 2
absence of signals attributable to R C@O or RCH –OH carbons in
easily upon standing in aqueous solution and/or during repeated
1
3
the C and DEPT NMR spectra. A normalizing constant ‘S’ was then
calculated according to the formula:
21
freeze drying cycles, its proportion was included in the calcula-
tion of product yields in Table 2. Also, throughout this report com-
1
2
i
d
i
d
pound 4 and intermediate species 5–11 will be all considered as
by-products of the synthesis of b-D-glucopyranuronosylamine (2),
unless otherwise specified.
As mentioned above, all methods reported in the literature for
the synthesis of glycosylamines in aqueous or aqueous methanol
solution (Table 1) involve the use of volatile ammonium salts that
X
A
n
S ¼
i¼1
ꢀ
ꢁ
1
4
b
:63
A
3:27 ꢁ A
1
a
:23
1b
4:63
2
3:18
3
4:70
5
5:63
6
5:37
¼
A5 þ A
þ A
þ A
þ
þ A
þ A
2
7
5
9
10
þ A
þ A
þ þA
release CO upon decomposition (carbonate, bicarbonate or carba-
:13
4:84
4:58
2
ꢀ
ꢁ
4
5
5
7
5:13
10
mate) and trap the product as the more stable N-glycosyl carba-
A
4:20—4:40 ꢁ A ꢁ 2A
ꢁ 3A
ꢁ 3A
:63
4:58
28,29
mate.
The rationale behind this choice is that studies
þ
ð1Þ
5
published in the period of 1950–1970 demonstrate that, upon
standing in concentrated aqueous ammonia, carbohydrates de-
grade and epimerize to give a large number of compounds30 of
which the corresponding glycosylamines are sometimes a minor
component. Since none of these reports focused on uronic acids
i
d
where A indicates the area of the methine signal of compound i
having a chemical shift d (ppm), and n is the number such methine
groups in species i (e.g., there are two equivalent H2 protons in
diglycosylamine 4). It should be noted that since the H2 signal of
b
superimposes with that of 1 , the former was corrected by
i
d
though, we decided to react D-glucose and sodium D-glucuronate
4

A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
2387
H-5
3
H-2
3
H-1
3
H-1
2
H-2
2
1
1
+
others
H-1
5
H-5
6
H-1
6
H-1
7
H-1
1
0
H-1
β
H-2
β
H-1
α
H-5
α
5
.6
5.4
5.2
4.8
4.6
4.4
4.2
4.0
3.8
3.6
3.4
3.2
1
H / ppm
Figure 1. Evolution of the 1D 1H spectrum (400.13 MHz) with time during the transformation of sodium
D
-glucuronate (1) into a mixture of 2 and 3 according to protocol
1
B.5.06. The assignment of some peaks is also shown (D
2
O, 278 K, d ( H)TSP = ꢁ0.017 ppm).
with commercial aqueous ammonia (ꢀ14.5 M) at 30 °C and to ana-
lyze the gross products by NMR spectroscopy (protocol X.14.00 in
Table 2). As expected, in the case of glucose equal amounts of by-
products and of glycosylamine/N-glycosyl carbamate formed: 21%
and 33% after 25 and 48 h, respectively. By contrast, for the same
0
2
4
6
8
10 24 48 72 96 120 144 168
by-products
0
.3
0.2
.1
0.0
0.3
0.2
0.1
0.0
x
0
reaction times sodium
D-glucuronate afforded 70% and 82% of 2 +
3, with only 20% and 16% of by-products, respectively. The latter
result is rather positive, especially taking into account the simplic-
ity of the procedure, and it will be used as benchmark for all other
protocols tested.
0
0
0
.8
.6
.4
0.8
0.6
0.4
0.2
0.0
products
2
.1.1. Saturated salt and concentrated ammonia protocols
x
We began our study by comparing the kinetics obtained with a
species
substrate protocol
þ
ꢁ
2+3
4
5-11
saturated solution of NH HCO , that is, the original protocol of
4
3
GlcA
Glc
A.0.S
A.0.S
B.0.S
B.0.S
1
4
Kochetkov and collaborators (A.0.S; entries 1 and 2 in Table 2),
0.2
0.0
GlcA
Glc
with those obtained with a saturated solution of ammonium carba-
mate (B.0.S; entries 16 and 17 in Table 2), which is less stable and
thus easier to eliminate by freeze-drying.23 Surprisingly, when
0
2
4
6
8
10 24 48 72 96 120 144 168
t / h
judged from the time required to obtain a 67% of 2 + 3, the carba-
mate protocol was about nine times faster than the bicarbonate
one (Fig. 2). Moreover, the equilibrium (plateau) fraction of 2 + 3
was 75% in the case of A.0.S and 89% in the case of B.0.S. Part of this
difference can be explained by the higher solubility of
Figure 2. Evolution of the composition with time for the reaction of sodium
glucuronate and -glucose with saturated ammonium bicarbonate or ammonium
carbamate solutions at 30 °C (see Table 2, entries 1, 2, 16, and 17). The mole
fractions (x) of different species are shown.
D-
D
þ
ꢁ
þ
ꢁ
NH NH
2
CO2 (>5.4 M vs ꢀ3.6 M for NH HCO ), and by the fact
4
4
3
that each formula unit of carbamate decomposes to give 2 mole-
cules of ammonia, but other factors must be at work. Similar
Concerning the amount of by-products 5–11, in case of GlcA
protocol B.0.S led to an initial increase that was followed by a rapid
stabilization around 8%, a value similar to that observed for A.0.S
(11%). The salt effect was less pronounced in the case of Glc, with
ꢀ10% of by-products in the presence of ammonium carbamate
against ꢀ6% for ammonium bicarbonate. Finally, in all cases the
amount of diglycosylamine remained close to zero during the first
33 h of reaction, and it only started to increase later on (the latter
data is only available for Glc).
behavior was observed with D-glucose (entries 2 and 17): to attain
a 52% yield in glycosylamine/N-glycosyl carbamate it took protocol
B.0.S one eighth of the time required using protocol A.0.S. What is
more, the rate of formation of glycosylamine/N-glycosyl carbamate
is ꢀ4–5 times slower in the case of
D-glucose than in the case of
GlcA (1), although the same fraction of product is present at
equilibrium.

2
388
A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
Although the high yield and the short reaction time obtained
a) 0.5 0
0.4
2
4
6
8
10
24
28
32
(
with protocol B.0.S are absolutely remarkable, the large amount
of salt used could restrict its scope in synthesis. For instance, oligo-
glycuronans are not soluble at high ionic strength. Thus, we tested
the effect of a decreasing amount of ammonium carbamate on the
reaction outcome (entries 11–15 in Table 2). As it turned out, by
0.5
0.4
0.3
0.2
0.1
0.0
by-products
[
NH3]
0
0
0
.3
.2
.1
x
þ
ꢁ
0.0
using a 2 M initial concentration of NH NH
2
CO2 , 84% of 2 + 3 is
4
still obtained after 24 h of reaction.
products
For comparison, a number of protocols based on the combina-
tion of concentrated aqueous ammonia (14.5 M) and ammonium
salts were investigated as well. Figure 3 shows the results obtained
0
.8
0.8
0.6
0.4
0.2
0.0
0.6
x
1
7
[
NH3]
with Lubineau’s protocol (A.14.02; entries 9 and 10 in Table 2) as
well as with a modification in which ammonium bicarbonate was
replaced with the less stable ammonium carbamate (B.14.02; en-
tries 24 and 25 in Table 2). It is evident that the same kinetic pro-
files were obtained with the two salts, that the conversion in
glycosylamine is ꢀ7–8 times faster for GlcA (1) than for Glc, and
that the amount of diglycosylamine remains close to zero during
the first 33 h of reaction and starts to increase later on (the latter
data are only available for Glc). Also, when compared with the cor-
responding saturated salt protocols (Fig. 2), it is clear that a much
larger fraction of by-products is present with concentrated ammo-
nia during the first ꢀ10 h of reaction, and that it takes ꢀ24 h to at-
tain a plateau value (ꢀ10% in all cases). This observation is
consistent with intermediate species taking longer to transform
into the final products 2 and 3. Concerning the absolute rate of
reaction, for both sugars it is comparable to that observed with
protocol A.0.S (saturated ammonium bicarbonate) although some-
what higher yields of 2 + 3 are obtained for long reaction times. It is
interesting to observe that, with respect to concentrated ammonia
alone (X.14.00; entries 26 and 27 in Table 2), the addition of 0.2 M
ammonium salt results in a moderate increase in the proportion of
glycosylamine/N-glycosylcarbamate in the case of GlcA (82% after
0
0
0
.4
.2
.0
protocol
species
2+3
4
5-11
A.1.06
A.5.06
A.9.06
0
0
2
2
4
4
6
6
8
t / h
10
24
24
28
28
32
32
8
10
(b)
by-products
0
.6
0.6
0.4
0.2
0.0
0.4
[NH3]
x
0
0
.2
.0
products
0.8
0.8
0.6
0.4
0.2
0.0
0
0
0
0
.6
.4
.2
.0
[
NH3]
x
protocol
species
4
2+3
5-11
B.1.06
B.5.06
B.9.06
2
4–25 h of reaction), and in a two-fold increase in the case of Glc
(
43%), both measured after 24–25 h of reaction.
0
2
4
6
8
10
24
28
32
t / h
2
.1.2. Effect of a lower concentration of reagents
In order to optimize the synthesis of b- -glucopyranuronosyl-
D
Figure 4. Evolution of the composition with time for the reaction of sodium
D-
amine, we looked into the possibility of reducing the amount of re-
agents used. At first, we examined the effect of a decreased
concentration of ammonia for a given concentration of ammonium
glucuronate with different concentrations of ammonia and (a) 0.6 M ammonium
bicarbonate (Table 2, entries 4, 7, and 8), or (b) 0.6 M ammonium carbamate
(entries 19, 22, and 23). The mole fractions (x) of the different species are shown.
þ
ꢁ
salt. Figure 4a shows that, in the presence of 0.6 M NH HCO
4
3
(
entries 4, 7, and 8 in Table 2), the use of 9 M or 5 M ammonia re-
0
2
4
6
8
10 24 48 72 96 120 144 168
by-products
sults in almost identical kinetic profiles, whereas a further diminu-
tion to 1 M ammonia slows the reaction down significantly and
results in ꢀ20% less glycosylamine formed at equilibrium. The
opposite can be said about the amount of by-products, which goes
from 10% to 20% at equilibrium when the ammonia concentration
decreases from 9 to 1 M. A qualitatively similar effect is observed
0
0
0
0
0
0
.5
.4
.3
.2
.1
.0
0.5
0.4
0.3
0.2
0.1
0.0
x
þ
ꢁ
with NH NH
2
CO2 (entries 19, 22, and 23 in Table 2), but the same
4
is much less pronounced (Fig. 4b). The fraction of diglycosylamine
4 remained close to zero in all cases. Here it should be noted that
after 33 h, the yield in 2 + 3 is virtually identical for protocols
A.14.02 and A.5.06, meaning that a small increase in the amount
of salt effectively compensates for a three-fold decrease in ammo-
nia concentration. The same can be said for protocols B.14.02 and
B.5.06.
Encouraged by the good results of protocols A/B.5.06, and fol-
lowing the same rationale of minimizing the amount of reagents
used, we also tested the effect of a lesser amount of ammonium
salt (entries 5 and 20 in Table 2). As shown in Figure 5, in the pres-
ence of 5 M ammonia a reduction from 0.6 to 0.2 M of the initial
concentration of ammonium bicarbonate/carbamate only results
in a slight decrease in the rate of formation of 2 + 3, and in a
ꢀ5% decrease of their proportion at equilibrium. Concerning the
0
0
0
0
0
.8
.6
.4
.2
.0
0.8
0.6
0.4
0.2
0.0
products
x
substrate protocol
species
4
2
+3
5-11
GlcA
Glc
GlcA
Glc
A.14.02
A.14.02
B.14.02
B.14.02
0
2
4
6
8
10 24 48 72 96 120 144 168
t / h
Figure 3. Evolution of the composition with time for the reaction of sodium
glucuronate and -glucose with concentrated ammonia (14.5 M) and one equiva-
lent ammonium salt (0.2 M; entries 9, 10, 24, and 25 in Table 2). The mole fractions
x) of the different species are shown.
D-
D
(

A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
2389
0
2
products
2
4
6
8
10
24
28
32
amount of by-products 5–11, there was no effect in the case of
ammonium bicarbonate, while a higher concentration of ammo-
nium carbamate led to a higher fraction of 5–11 being present
during the first hours of reaction. This difference disappeared after
0
0
0
0
.6
.4
.2
.0
0.6
0.4
0.2
0.0
by-products
x
ꢀ
7 h, though. The fraction of diglycosylamine 4 remained close to
zero in all cases, and attained a maximum of ꢀ3% after 33 h.
2
.1.3. Effect of temperature
The last parameter to be tested was the reaction temperature,
0
0
0
0
0
.8
.6
.4
.2
.0
0.8
0.6
0.4
0.2
0.0
and in particular whether an increase in temperature would accel-
erate the rate of reaction without increasing the amount of by-
products (Fig. 6). As it turned out, raising the temperature from
30 to 40 °C resulted in a four-fold increase in the rate of production
of glycosylamine/N-glycosylcarbamate in the case of protocol A.0.S
(entries 1 and 3 in Table 2), and in a three-fold increase in the case
of B.0.S (entries 16 and 18), both measured at 70% conversion. All
the same, the highest yield attained remains almost identical for
the two temperatures, and for long reaction times, a decrease in
the proportion of 2 + 3 is actually observed in the case of B.0.S: This
certainly results from the poorer thermal stability of ammonium
x
protocol
species
4
2+3
5-11
A.5.02
A.5.06
B.5.02
B.5.06
0
4
6
8
t / h
10
24
28
32
Figure 5. Effect of ammonium salt concentration (0.2 M or 0.6 M) on the evolution
of the composition with time for the reaction of sodium -glucuronate with 5 M
ammonia (Table 2, entries 5, 7, 20, and 22). The mole fractions (x) of different
D
2
3
carbamate, which decomposes at T >30 °C. As for the amount of
by-products 5–11, raising the temperature resulted in a marginal
increase, whereas the fraction of diglycosylamine 4 remained close
to zero in all cases.
species are shown.
0
2
4
6
8
10
24
28
32
(
a) 0
.3
.2
.1
.0
0.3
0.2
0.1
0.0
by-products
0
1
2
3
4
5
6
7
8
9
24
32
(
a)
0
1
0
.0
.9
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
x
0
conditions
substrate
protocol
species
T
GlcA
B.5.02
1
0
1
0
0.8
2+3
6+9
4
o
temperature 30
C
0
0
0
0
.7
.6
.5
.4
.0
.8
.6
.4
.2
.0
NH CO NH
products
1.0
0.8
0.6
0.4
0.2
0.0
2
2
4
sat.
other intermediates
x
0
x
protol temp
species
4
2+3
5-11
0
0
0
o
A.0.S 30
C
C
C
0.3
0.2
NH HCO
o
4
3
sat.
40 o
B.0.S 30
o
4
0 C
0
0
.1
.0
0
0
2
2
4
4
6
6
8
t / h
10
24
24
28
28
32
32
0
0
1
2
1
3
4
2
5
6
7
8
4
9
24
24
32
32
t / h
8
T
10
(
b)
3
(b)
0.4
0.2
0.0
0.8
by-products
0.4
0.2
0.0
0.8
1.0
0.9
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
x
0
0
0
0
.8
.7
.6
.5
conditions
substrate
protocol
GlcA
B.0.S
o
temperature 30
C
o
40
C
products
species
species
0
0
0
0
.6
.4
.2
.0
0.6
0.4
0.2
0.0
x
1
x
0.4
.3
2+3
6+9
4
protocol temp
o
2+3
4
5-11
0
3
0
C
A.5.02 30
o
o
o
o
C
C
C
C
other inermediates
40
0.2
0.1
B.5.02 30
40
0
.0
0
2
4
6
8
10
24
28
32
0
1
2
3
4
24
32
t / h
t / h
Figure 6. Effect of temperature on the evolution of the composition with time for
the reaction of sodium -glucuronate with (a) saturated ammonium salt (Table 2,
entries 1, 3, 16, and 18) and (b) 5 M ammonia containing ammonium salt 0.2 M
entries 5, 6, 20, and 21). The mole fractions (x) of the different species are shown.
D
Figure 7. Comparison between the kinetics at 30 °C of (a) a ‘slow’ protocol (B.5.02)
and (b) the fastest protocol tested (B.0.S; entries 16 and 20 in Table 2). The mole
fractions (x) of different species are shown.
(

2
390
A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
+
-
Na OOC
O
H
N
HO
HO
Cl
OH
O
1
9
+
-
Na OOC
O
O
O
O
O
H
N
HO
HO
1
0 eq
88%
a)
O
O
OH
(
O
Cl
Cl
20
1
0 eq
6
9 % (a)
+
-
Na OOC
+
-
Na OOC
O
i) NH CO NH 5 M,
O
HO
HO
4
2
2
R.T., 24 hrs
HO
HO
O
NHX
X: H, CO
OH
OH
ii) Freeze-drying
OH
-
Cl
eq
1
2+3
2
8
2 %
6
+
-
Na OOC
O
8
0 %
HO
HO
H
N
(
b)
O
OH
OCN
O
68 %
(a)
O
2
1
OH
2
eq
O
H
HO
HO
N
OH
+
-
Na OOC
HO
HO
O
O
O
2
3
H
N
H
N
O
OH
O
2
2
Scheme 2. Synthesis of N-acyl-b-
D
-glucopyranuronosylamines 19–21 and of N-alkylcarbamoyl-b-
OH 1:1, 0 °C, 0.5 h.
2 3 2
D-glucopyranuronosylamine (22). Conditions: (a) Na CO , H O/DMSO 8:2,
0
°C to rt, 25 h; (b) Na CO , H O/CH
2
3
2
3
The same type of experiment was carried out with protocols A/
of the reaction, the proportion of residual GlcA 1 was 5% in the case
of B.5.02 and 2% in case of B.0.S, and both systems contained the
B.5.02 (Fig. 6b; entries 5, 6, 20, and 21 in Table 2). In this case, raising
the temperature from 30 to 40 °C resulted in a two-fold increase in
the rate of formation of 2 + 3 (measured at 60% yield), but the evap-
oration of ammonia was rapid enough to lead to ꢀ5% lower yields
after 24 h. Accordingly, at 40 °C the proportion of 5–11 was signifi-
cantly lower during the first ꢀ10 h of reaction and bottomed out
around 13% for longer reaction times, that is, the same value ob-
served at 30 °C. Finally, for both temperatures tested the amount
of diglycosylamine 4 remained close to zero during the first ꢀ6 h
and increased to ꢀ3% by the end of the reaction.
same amount (7%) of the
a anomers 6 + 9. We checked the final
amount of 6 + 9 for three other protocols and found that nearly the
same value was obtained with A.5.06 (8%), A.14.02 (8%), and
B.9.06 (7%). This suggests that 7–8% is about the equilibrium fraction
of a-D-glucopyranuronosylamine/carbamate in water at 30 °C.
2.2. Synthetic potential
Glucuronidation is a major detoxification pathway in all verte-
brates. The reaction is catalyzed by UDP-glucuronosyltransferases
(EC 2.4.1.17) and involves the transfer of a glucuronic acid residue
from UDP-glucuronic acid to countless substances possessing hy-
droxyl, amino, carboxyl, or thiol groups, thus converting them to
2
.1.4. Fate of intermediate species
Finally, it is worth looking more closely at the results obtained at
0 °C for a ‘slow’ protocol (B.5.02; entry 20 in Table 2) and for the
3
3
1
fastest protocol tested (B.0.S; entry 16). Figure 7 shows that after
0 min of reaction between 65% (B.5.02) and 70% (B.0.S) of the initial
sugar had already been consumed. In the case of B.5.02 though, only
half of it was transformed into ,b- -glucopyranuronosylamine and
ammonium N- ,b- -glucopyranuronosyl carbamate, whereas the
other half yielded intermediates 5, 7, 8, 10, and 11. It then took
2.5 h for the concentration of intermediates to reach its lowest va-
water-soluble b-D-glucopyranuronosyl derivatives. What is more,
3
pharmacologists have conjugated a number of drugs to glucuronic
acid to allow for more effective delivery, while the toxicity of some
3
2
a
D
substances is effectively reduced by glucuronidation. In the field
3
3
a
D
of drug delivery, Oku and Namba have demonstrated that lipo-
somes carrying b- -glucopyranuronosyl residues at their surface
D
3
bind to a lower extent to macrophage-like cells, have prolonged
circulation, and passively accumulate in tumor tissue, thus reduc-
ing the toxic side effects of the therapeutic agents.
lue of 6% and for the yield in 2 + 3 to attain a maximum value of 80%.
By contrast, in the case of B.0.S the fraction of intermediates present
after 30 min was only 10%, and their concentration reached a mini-
mum of ꢀ2% in just 2.5 h. Consequently, at the same reaction time
In this context, the possibility to transform
D-glucuronic acid
into b- -glucopyranuronosyl derivatives without resorting to pro-
D
the system already contained ꢀ60% of
amine and ammonium N- ,b-
and it reached a maximum yield of 89% in barely 5.5 h. At the end
a,b-D
-glucopyranuronosyl-
tective group chemistry would be highly advantageous. To probe
the synthetic potential of our approach, four test reactions were
carried out in which the title compound was reacted with various
a
D-glucopyranuronosyl carbamate,

A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
2391
Table 3
b
Summary of the characterization of compounds 1 –3 and 19–23
Compounda d 1H/ppm (m, J/Hz, assignment)
d 13C/ppm (assignment)
m/z (structure, monoisotopic ion mass, molecular
ion)
ꢁ
1
2
3
1
b
4.63 (d, 8.0, H1), 3.27 (m, H2), 3.50 (H3), 3.51 (m, H4),
.72 (H5)
98.57 (C1), 76.70 (C2), 78.25 (C3), 74.53 193 (1
(C4), 78.93 (C5), 178.64 (C6)
H
, 193.0, [MꢁH] ); 217 (1
H
, 217.0,
+
+
3
[M+Na] ); 239 (1Na, 239.0, [M+Na] ); 387 ((1
H
)
2
,
ꢁ
+
3
87.1, [MꢁH] ); 455 ((1Na
2
) , 455.0, [M+Na] )
ꢁ
+
4.09 (d, 8.8, H1), 3.18 (dd, 9.0, H2), 3.48 (H3), 3.48 (d, 9.5, 87.68 (C1), 76.76 (C2), 79.03 (C3),74.71 192 (2
H4), 3.69 (m, H5)
H
, 192.1, [MꢁH] ); 194 (2
H
, 194.1, [M+H] );
+
(C4), 79.92 (C5), 179.14 (C6)
216 (2Na, 216.0, [M+H] ); 238 (2Na, 238.0,
+
ꢁ
[
M+Na] ); 385 ((2
)
H 2
, 385.1, [MꢁH] )
+
4.70 (d, 9.1, H1), 3.34 (t, 9.0, H2), 3.52 (H3), 3.48 (H4),
85.68 (C1), 74.64 (C2), 79.02 (C3), 74.55 238 (3H, H, 238.0, [M+H] ); 260 (3H, H, 260.0,
+
3
.73 (d, 9.4, H5), 6.25 (d, 9.8, NHCO
2
)
(C4), 79.87 (C5), 178.88 (C6), 165.86
NHCO
82.16 (C1), 78.84, 78.60, 73.90, 73.73,
4.96 (CH Cl), 173.10 (amide)
[M+Na] )
(
2
)
ꢁ
9
5.03 (d, 9.1, H1), 4.20 (s, CH Cl)
2
268, 270 (22
H
, 268.0, 270.0, [MꢁH] ); (22Na
,
ꢁ
4
2
290.0, 292.0, [MꢁH] ); (22
H
, 292.0, 294.0,
+
+
[
M+Na] ); 314, 316 (22Na, 314.0, 316.0, [M+Na] )
ꢁ
2
2
2
2
0
5.06 (d, 8.6, H1), 3.80 (d, 9.5, H5)
5.04 (d, 9.0, H1), 3.81 (d, 9.6, H5)
4.82 (d, 9.1, H1), 3.74 (d, 9.5, H5)
82.12 (C1), 80.97, 78.95, 74.44, 175.54
amide)
81.95 (C1), 80.65, 78.89, 74.37, 132.08
CH @CH), 172.04 (amide)
83.69 (C1), 80.62, 138.49 (CH
29.67 (CH @C), 162.33 (urea)
260 (19
H
, 260.1, [MꢁH] ); 262 (19
H
, 262.1,
+
+
(
[M+H] ); 306 (19Na, 306.1, [M+Na] )
ꢁ
1
246 (20
H
, 246.1, [MꢁH] ); 268 (20Na, 268.0,
ꢁ
+
(
2
[MꢁH] ); 292 (20Na, 292.0, [M+Na] )
ꢁ
2
2
@C),
347 (21
[M+Na] ); 717 ((21
H
+
, 347.1, [MꢁH] ); 393 (21Na, 393.1,
ꢁ
1
2
H
)21Na, 717.2, [MꢁH] )
3b
5.05 (d, 9.2, H1); 3.87 (dd, 12.4, 2.1, H6); 3.72 (dd, 12.4, 82.12 (C1), 80.30 (C5), 79.12 (C3), 74.48 232 (23 232.1; [MꢁH] ); 256 (23 256.1,
ꢁ
+
+
5.4, H6); 3.54 (m, H3 and H5); 3.44 (m, H2 and H4); 6.32 (C2), 71.92 (C4), 63.26 (C6), 132.09,
2
[M+Na] ); 489 ((23) , 489.2, [M+Na] )
(
m, CH @CH); 5.87 (dd, 6.6, 4.9, CH @CH) 132.11 (CH @CH), 172.02 (amide)
2
2
2
a
Solvents used for NMR analyses: D
Assignments obtained by HMQC.
2 2 6 2 6
O (19, 21, 23); 8:2 D O–DMSO-d (20) or 5:4 D O–DMSO-d (22).
b
acylating agents and with an isocyanate (Scheme 2). In particular, a
mixture of 2 + 3 was prepared according to protocol B.0.50 and
purified by two freeze-drying cycles. This procedure slightly re-
duces the global yield (82% vs 88% for a single cycle) but effectively
removes all ammonium salt. The gross product was then reacted
with chloroacetic anhydride, methacrylic anhydride, acryloyl chlo-
ride, and 2-isocyanoethyl methacrylate in cold aqueous/organic
into intermediate species 5–11, and of the latter into the final
products 2 and 3. Yet, some interesting trends and exceptions were
observed:
i. B.0.S (saturated ammonium carbamate) was both the fastest
protocol tested and the one leading to the highest final
yields of 2 + 3 (89% in 5.5 h at 30 °C; 87% in 1.5 h at 40 °C).
In particular, at 30 °C it was nine times faster than Kochet-
kov’s and Lubineau’s methods. A qualitatively similar result
solutions. This way, N-acyl-b-D-glucopyranuronosylamines 19–21
and N-alkylcarbamoyl-b- -glucopyranuronosylamine 22 where
D
obtained in 68–88% yield. The compounds were not isolated, but
their formation was confirmed by MS and NMR analysis of the
crude products, as summarized in Table 3. In particular, we found
that for compounds 19–21 the chemical shift and coupling con-
stant of C1 (82 ppm) and H1 (5.05 ppm, doublet, 9 Hz) are almost
was obtained with
be a general finding.
D-glucose, and we postulate that this may
ii. Whenever a lesser amount of salt (or a lower ionic strength)
is needed, protocols B.0.20/30 and B.5.06 should be pre-
ferred to A/B.9.06 and A/B.14.02, since they lead to the same
yield in 2 + 3 (84–86% after 24–33 h) while requiring much
less reagent.
iii. With the sole exception of A/B.1.06, after 24 h of reaction all
tested protocols led to higher yields of 2 + 3 than concen-
trated commercial ammonia alone (X.14.00).
identical to those of N-(prop-2-enoyl)-b-
23), a pure sample of which was prepared in our laboratory. The
same can be said for compound 22, whose diagnostic peaks match
D-glucopyranosylamine
(
those reported in the literature for several N-arylcarbamoyl-b-D-
glucopyranosylamines.34 Finally, it is interesting to note that the
yields of 19 and 21 are comparable to those reported by Manger
iv. The equilibrium fraction of
about 7–8% in water at 30 °C.
v. Concerning bis(b- -glucopyranuronosyl)amine (4), less than
3% formed in all cases, with a minimum value of 0.5% in the
case of B.0.S (after 24 h).
vi. The formation of b-glycopyranosylamine/N-b-glycopyrano-
syl carbamate is consistently faster in the case of sodium
a anomers 6 + 9 was found to be
et al.2 and Kallin et al. for the analogous derivatives of N-acetyl-
glucosamine (95%) and lactose (88%), respectively; whereas the
yield of 22 matches those obtained by Somsák et al.35 for the syn-
1
19
D
thesis of N-arylcarbamoyl-b-
45–76%). Here it should be noted that compound 19 represents
a classic starting point for the synthesis of 1-N-glycyl-b-D-glycosyl
D-glucopyranosylamines in pyridine
(
1
8,21
derivatives,
while compounds 20–22 can be either functional-
D
-glucuronate than in the case of
D-glucose (4–8 times
35
ized by thiol-ene chemistry or used as glycomonomers for radical
polymerization.
faster).
3
6,37
Finally, the synthetic usefulness of our approach was demon-
strated by transforming a mixture of b- -glucopyranuronosyl-
amine (2) and N-(b- -glucopyranosyluronic acid) carbamate (3)
into N-acyl-b- -glucopyranuronosylamines 19–21 (69–88% yield)
and N-alkylcarbamoyl-b- -glucopyranuronosylamine 22 (68%
yield) directly in aqueous/organic solution without resorting to
protective group chemistry. The findings from this study are now
being applied to the regioselective functionalization of oligoglycu-
ronans directly in aqueous solution.
D
3
. Conclusions
D
D
When sodium
and/or volatile ammonium salts in water, the rate of formation of
b- -glucopyranuronosylamine (2) and ammonium N-(b- -gluco-
pyranuronosyl carbamate (3) strongly depended on the experi-
mental conditions. In general higher ammonia and/or ammonium
salt concentrations led to a faster conversion of the starting sugar
D-glucuronate (1) was reacted with ammonia
D
D
D

2
392
A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
4
4
. Experimental
and pulse sequence recycle times of 3 s were used. The probe tem-
perature was calibrated in the range 275–300 K using neat metha-
nol. The NMR probe was pre-equilibrated at 278 K, individual
.1. Materials and methods
samples were redissolved in D
2
O (ꢀ3% w/w), transferred to a
Unless otherwise specified, all chemicals were reagent grade
NMR tube, and immediately lowered into the instrument magnet
1
and used as received. Ammonium bicarbonate (P99.0%), ammo-
nium carbamate (P99.5%), -glucose (P99.0%), and -glucuronic
acid sodium salt monohydrate (99%) were from Fluka Chemical
Co. Ammonia (28% w/w, Carlo Erba), D O (99.9%-D, Euriso-top),
-isocyanatoethyl methacrylate (>98.0%, TCI Europe), methacrylic
for analysis. 1D H spectra were obtained with 32 scans and 32 K
D
D
data points, and were re-processed using MestReNova software
(v5.1). Sodium 3-(trimethylsilyl)propanoate (TSP) or sodium 3-(tri-
methylsilyl) propane-1-sulfonate (DSS) were used as an internal
2
1
13
2
reference. Chemical shifts (in ppm) for H and C nuclei were ref-
1
13
anhydride (94%, Sigma–Aldrich), SiO
2
(15–40
lm, 60 Å, E. Merck),
erenced to dTSP = ꢁ0.017 ppm ( H) and dTSP = ꢁ0.149 ppm ( C), or
1 13
sodium carbonate monohydrate (P99.5%, Sigma–Aldrich), and
to dDSS = 0.000 ppm ( H and C). Mass spectrometry analyses were
performed with a Waters ZQ (Altrincham, GB) single quadrupole
atmospheric pressure ionization mass spectrometer fitted with a
Z electrospray interface (ESI). The instrument was calibrated with
TLC plates (SiO F254, 15 m, 60 Å, E. Merck) were obtained from
2
l
the indicated suppliers. Acryloyl chloride (P96.0%, Fluka) and
methacryloyl chloride (P97.0%, Fluka) were distilled under re-
duced pressure. Distilled or deionized water was used in all
experiments.
ꢁ
1
mass spectra generated by electrospray ionization of a 0.1 mol L
solution of sodium iodide in aqueous acetonitrile (50%, v/v) in the
mass range of 23–1972 amu. Nitrogen was used as the drying and
ꢁ1
4
.2. Protocol numbering
nebulizing gas. Samples (ꢀ1 mg mL ) were dissolved in deionized
water or water/MeOH mixtures and infused to the ESI interface at
ꢁ
1
Sodium
acted at 30 and 40 °C with ammonia and/or volatile ammonium
salts (NH HCO or NH CO NH ) in water according to different
D
-glucuronate (1) (Scheme 1) and
D-glucose were re-
constant flow rate (50
4.5. Synthesis of b- -glucopyranuronosylamine (2)
-Glucuronic acid sodium salt monohydrate (1.00 g, 4.27 mmol)
lL min ).
D
4
3
2
2
4
protocols. The numbering of each protocol reveals the exact exper-
imental conditions used:
D
and ammonium carbamate (8.32 g, 0.106 mol) were weighed in a
50-mL round-bottom flask and dissolved in 21.3 mL of de-ionized
water (experiment AP10-12, protocol B.0.50). A magnetic bar was
added, the flask was sealed with a rubber septum, a disposable
needle (21 G) was passed through the septum to prevent pressure
build-up, and the mixture was stirred at ambient temperature
(ꢀ25 °C) and 300 rpm. After 24 h of reaction, the content of the
flask was frozen in liquid nitrogen and freeze dried overnight.
The resulting powder was redissolved in deionized water
(80 mL) and submitted to a second cycle of freeze-drying. The
obtained white fluffy solid (1.13 g) had the following molar
ꢂ The letter indicates the type of solid salt added (A, NH
NH CO NH , X, no salt).
ꢂ The middle number indicates the initial formal concentration of
4 3
HCO ; B,
2
2
4
ꢁ1
liquid ammonia in mol L
.
ꢂ The last number indicates the initial formal concentration of salt
ꢁ
1
ꢁ1
in 10 mol L
.
4 3
For example, A.1.06 indicates that the salt used was NH HCO ,
and that the initial concentration of ammonia and salt were 1
and 0.6 M, respectively.
1
composition (as determined by H NMR): 1 (10%), 2 (65%), 3
4
.3. Kinetic study
(17%), 4 (1%), 6 + 9 (4%), 5 + 7 + 8 + 10 + 11 (3%). The sample
was then sealed in a round-bottom flask and stored in a freezer
(ꢁ18 °C) until needed. The increase in the mass of the sample is
mostly due to the presence of N-glycosylcarbamates 3, 5, and 6,
whose molar mass is larger than that of the starting sugar 1 and
of glycosylamine 2.
In a typical experiment (AG09-30_P2, protocol B.5.06), D-gluc-
uronic acid sodium salt monohydrate (0.468 g, 2.0 mmol) and
ammonium carbamate (0.468 g, 6.0 mmol) were weighed in a 25-
mL round-bottom flask. A magnetic bar was added together with
1
0 mL of aqueous ammonia (5 M), the flask was sealed with a rubber
septum, and a disposable needle (21 G) was passed through the sep-
tum to prevent pressure build-up. The flask was plunged in an oil
bath pre-heated at 30 °C and stirred at 200 rpm. At pre-set intervals,
4.6. Synthesis of N-acyl-b-D-glucopyranuronosylamines (19–21)
and of N-{[2-((2-methylprop-2-enoyl)oxy)ethyl]carbamoyl}-b-D-
glucopyranuronosylamine (22)
ꢀ
250 lL samples were drawn with a syringe, transferred to a glass
vial, diluted with 2 volumes of water, and frozen in liquid nitrogen.
In a typical experiment, b-
of gross product) was dissolved in (a) 8:2 1 M Na
(3.9 mL) or (b) 1:1 1 M Na CO –DMSO (4.4 mL) at 0 °C under stir-
ring. A calculated volume of acylating agent was added dropwise
either as neat liquid (a) or as 1.5 M solution in THF (b). The result-
ing mixture was then left stirring on ice for 60 min (a) or 30 min
D
-glucopyranuronosylamine (100 mg
All samples were freeze-dried overnight and stored in a freezer
2
CO –DMSO
3
(
ꢁ18 °C) untilneeded. The mole fraction ofeach compound was then
2
3
1
calculated by H NMR spectroscopy according to (Eq 2). For proto-
cols A/B.0.S, saturation was ensured by the constant presence of so-
lid salt at the bottom of the flask.
(
b), and at ambient temperature for another 24 h (a). MeOH was
4
.4. Instrumental analyses
eliminated by rotary evaporation (b) and any unreacted acylating
agent was removed by solvent extraction (EtOAc, 2 ꢃ 5 mL). Fol-
lowing further rotary evaporation at ambient temperature, the
remaining solution was diluted with 2 volumes of water and
freeze-dried overnight (experiments WM_03, WM_06, AP10-11,
AP10-16 and AP10-17). Yields were calculated from the H NMR
spectra of the gross products by integrating to one the anomeric
protons signals.
The composition of samples from the kinetic study was deter-
mined by NMR spectroscopy on a Bruker DPX400 spectrometer
equipped with a Variable Temperature (VT) module (resonance
frequency of 400.13 and 100.62 MHz for H and C nuclei, respec-
tively). Two different 5-mm detection probes were used: QNP
1
13
1
(
direct) and BBIZ (inverse). Unless otherwise specified, 90° pulses

A. Ghadban et al. / Carbohydrate Research 346 (2011) 2384–2393
3. Timoshchuk, V. A. Russ. Chem. Rev. 1995, 64, 675–703.
2393
4
.7. Synthesis of N-(prop-2-enoyl)-b-D-glucopyranosylamine (23)
4
5
6
7
.
.
.
.
Leendert, J. v. d. B.; Jeroen, D. C. C.; Remy, E. J. N. L.; Jasper, D.; Herman, S. O.;
Gijsbert, A. v. d. M. Eur. J. Org. Chem. 2007, 2007, 3963–3976.
Kamoda, S.; Nakano, M.; Ishikawa, R.; Suzuki, S.; Kakehi, K. J. Proteome Res.
2005, 4, 146–152.
D
-Glucose (2.00 g, 11.1 mmol) and ammonium bicarbonate
2.60 g, 32.8 mmol) were weighed in a 250-mL round-bottom flask
(
Likhosherstov, L. M.; Novikova, O. S.; Shibaev, V. N. Russ. Chem. Bull. 1998, 47,
and dissolved in 55 mL of 5 M ammonia (experiment WM_04, pro-
tocol A.5.06). A magnetic bar was added, the flask was sealed with a
rubber septum, a disposable needle (21 G) was passed through the
septum to prevent pressure build-up, and the mixture was stirred
at 35 °C and 300 rpm. After 30 h the flask was fitted to a rotary
evaporator, and the reaction mixture was concentrated to ꢀ½ of
the initial volume (p = 35 mbar, Tbath = 25 °C). More water was
added (ꢀ30 mL), and the process was repeated once. The resulting
solution was freeze-dried overnight. A white fluffy solid (2.15 g)
1
214–1217.
Vetter, D.; Tate, E. M.; Gallop, M. A. Bioconjugate Chem. 1995, 6, 319–322.
8. Ishiwata, A.; Takatani, M.; Nakahara, Y.; Ito, Y. Synlett 2002, 634–636.
9.
Monsigny, M.; Quetard, C.; Bourgerie, S.; Delay, D.; Pichon, C.; Midoux, P.;
Mayer, R.; Roche, A. C. Biochimie 1998, 80, 99–108.
1
1
1
1
1
0. Zeng, X.; Murata, T.; Kawagishi, H.; Usui, T.; Kobayashi, K. Biosci., Biotechnol.,
Biochem. 1998, 62, 1171–1178.
1. Plusquellec, D.; Brenner-Henaff, C.; Leon-Ruaud, P.; Duquenoy, S.; Lefeuvre, M.;
Wroblewski, H. J. Carbohydr. Chem. 1994, 13, 737–751.
2. Mellet, C. O.; Blanco, J. L. J.; Fernandez, J. M. G.; Fuentes, J. J. Carbohydr. Chem.
1993, 12, 487–505.
3. Urge, L.; Otvos Jr, L.; Lang, E.; Wroblewski, K.; Laczko, I.; Hollosi, M. Carbohydr.
Res. 1992, 235, 83–93.
4. Likhosherstov, L. M.; Novikova, O. S.; Derevitskaya, V. A.; Kochetkov, N. K.
Carbohydr. Res. 1986, 146, C1–C5.
containing b- -glucopyranosylamime was obtained that was mixed
with Na CO O (6.13 g, 49.5 mmol) and redissolved in 108 mL of
ꢄH
2
:1 CH OH–H O under stirring (WM_06). The resulting solution
D
2
3
2
1
3
was cooled in an ice bath and acryloyl chloride (4.10 g, 45.2 mmol,
diluted in 31 mL of anhyd THF) was added over a period of 5 min
under vigorous stirring. After 30 min the reaction flask was fitted
to a rotary evaporator, and the volatiles were eliminated at ambient
temperature. The reaction mixture was then diluted with water
15. Halila, S.; Manguian, M.; Fort, S.; Cottaz, S.; Hamaide, T.; Fleury, E.; Driguez, H.
Macromol. Chem. Phys. 2008, 209, 1282–1290.
1
6. Retailleau, L.; Laplace, A.; Fensterbank, H.; Larpent, C. J. Org. Chem. 1998, 63,
08–617.
17. Lubineau, A.; Auge, J.; Drouillat, B. Carbohydr. Res. 1995, 266, 211–219.
8. Tuzikov, A. B.; Gambaryan, A. S.; Juneja, L. R.; Bovin, N. V. J. Carbohydr. Chem.
000, 19, 1191–1200.
6
1
2
(
60 mL) and 2-PrOH (50 mL) and 63 g of chromatographic grade
SiO were added. In order to adsorb the gross product on silica,
the flask was re-fitted to the rotary evaporator and water was elim-
inated as a binary azeotrope with 2-PrOH (T = 80 °C, 88% w/w alco-
hol) before evaporating the resulting slurry to dryness (p = 45 mbar,
bath = 40 °C). The dry silica gel was then charged on the top of a pre-
packed column and eluted with 9:1 CH CN–H O. Fractions contain-
ing the product (R 0.34, 8:2 CH CN–H O) were pooled, stabilized
with a few grains of BHT, concentrated at the rotary evaporator,
and freeze-dried overnight. Isolated yield: 1.45 g (52%) of white
fluffy powder. The sample was stored in a freezer (ꢁ18 °C) until
needed. See Table 3 for spectroscopic characterization.
1
9. Kallin, E.; Loenn, H.; Norberg, T.; Elofsson, M. J. Carbohydr. Chem. 1989, 8, 597–
611.
2
20. Rice, K. G. Anal. Biochem. 2000, 283, 10–16.
2
1. Manger, I. D.; Rademacher, T. W.; Dwek, R. A. Biochemistry 1992, 31, 10724–
b
1
0732.
22. Vetter, D.; Gallop, M. A. Bioconjugate Chem. 1995, 6, 316–318.
2
3. Fulks, G.; Fisher, G. B.; Rahmoeller, K.; Wu, M.-C.; D’Herde, E.; Tan, J., A Review
of Solid Materials as Alternative Ammonia Sources for Lean NOx Reduction
with SCR. In Diesel Exhaust Emission Control, 2009, SAE International:
Warrendale, 2009; pp 2009–2001–0907.
T
3
2
f
3
2
2
4. Likhosherstov, L. M.; Novikova, O. S.; Shibaev, V. N. Dokl. Chem. 2003, 389, 73–
7
6.
2
5. Likhosherstov, L. M.; Novikova, O. S.; Shibaev, V. N. Dokl. Chem. 2002, 383, 89–
92.
26. Campa, C.; Donati, I.; Vetere, A.; Gamini, A.; Paoletti, S. J. Carbohydr. Chem.
001, 20, 263–273.
2
2
7. Ghadban, A.; Albertin, L.; Condamine, E.; Moussavou Mounguengui, R. W.;
Acknowledgments
Heyraud, A. Can. J. Chem. 2011, 89, 987–1000.
2
2
3
3
8. Johnson, S. L.; Morrison, D. L. J. Am. Chem. Soc. 1972, 94, 1323–1334.
9. Ewing, S. P.; Lockshon, D.; Jencks, W. P. J. Am. Chem. Soc. 1980, 102, 3072–3084.
0. Kort, M. J. Adv. Carbohydr. Chem. Biochem. 1971, 25, 311–349.
1. Grancharov, K.; Naydenova, Z.; Lozeva, S.; Golovinsky, E. Pharmacol. Ther. 2001,
89, 171–186.
We thank the French Ministère de l’Enseignementsupérieur et de
la Recherche, the Réseau de Recherche Chimie pour le Développe-
ment Durable (CNRS-INRA), the Région Rhône-Alpes (Cluster Chi-
mie) and the Agence Nationale de la Recherche Scientifique (ANR-
32. Kroemer, H. K.; Klotz, U. Clin. Pharmacokinet. 1992, 23, 292–310.
33. Oku, N.; Namba, Y. Methods Enzymol. 2005, 391, 145–162.
34. Somsák, L.; Felföldi, N.; Kónya, B.; Hüse, C.; Telepó, K.; Bokor, É.; Czifrák, K.
Carbohydr. Res. 2008, 343, 2083–2093.
0
9-CP2D-02). We also thank Isabelle Jeacomine, Stéphanie Boullan-
ger, and Dr. Bernard Brasme for their assistance with NMR and MS
analyses, and Professor M. Rinaudo for constructive criticism.
35. Hoyle, C. E.; Bowman, C. N. Angew. Chem., Int. Ed. 2010, 49, 1540–1573.
36. Klein, J.; Begli, A. H.; Engelke, S. Makromol. Chem.-Rapid 1989, 10, 629–636.
37. Kallin, E.; Loenn, H.; Norberg, T.; Sund, T.; Lundqvist, M. J. Carbohydr. Chem.
1991, 10, 377–386.
References
1
2
.
.
McNaught, A. D. Pure Appl. Chem. 1996, 68, 1919–2008.
Vandamme, E. J.; De Baets, S.; Steinbüchel, A. Polysaccharides II: Polysaccharides
from Eukaryotes; Wiley-VCH: Weinheim, 2002.