Hydrolysis rates of 1-glucosyl-2-benzoylhydrazines in aqueous solution

Article abstract of DOI:10.1016/j.carres.2006.12.015

A series of substituted 1-glucosyl-2-benzoylhydrazines were synthesized and their pseudo-first-order rate constants for hydrolysis were determined at pH 4, 5, 6 at 50 °C. All the compounds hydrolyzed quickly (t_1/2 < 3 h) at pH 4.0, but were increasingly stable as the pH approached neutrality.

Full text of DOI:10.1016/j.carres.2006.12.015

Carbohydrate Research 342 (2007) 749–752
Note
Hydrolysis rates of 1-glucosyl-2-benzoylhydrazines
in aqueous solution
Anna V. Gudmundsdottir and Mark Nitz*
Department of Chemistry, 80 St. George Street, Toronto, Ont., Canada
Received 14 October 2006; received in revised form 13 December 2006; accepted 17 December 2006
Available online 22 December 2006
Abstract—A series of substituted 1-glucosyl-2-benzoylhydrazines were synthesized and their pseudo-first-order rate constants for
hydrolysis were determined at pH 4, 5, 6 at 50 ꢀC. All the compounds hydrolyzed quickly (t1/2 < 3 h) at pH 4.0, but were increasingly
stable as the pH approached neutrality.
ꢁ 2007 Elsevier Ltd. All rights reserved.
Keywords: Glycoconjugate; Glycosylhydrazides; Oligosaccharide labelling; Hydrolysis rates
Glycoconjugates are the primary constituents of extra-
cellular matrices and cellular surfaces, where their oligo-
saccharide moieties have been shown to play important
roles in recognition-driven events including cell growth,
development, migration and differentiation.1 The oligo-
saccharide composition at cell surfaces differs between
cell types, and alterations in glycosylation patterns have
been associated with specific disease states. Thus, the
potential exists to design therapeutic agents that interfere
with carbohydrate–protein interactions or exploit the
carbohydrates as biomarkers for vaccine development.2
Despite the biological relevance of glycoconjugates,
their preparation by chemical synthesis is still a challeng-
ing task due to the requirements for extensive use of
orthogonal protection–deprotection strategies to generate
natural O- or N-glycosidic linkages. This problem
becomes acute when only small amounts of isolated oligo-
saccharides are available and multistep synthetic ap-
proaches are not possible. Efforts to circumvent the
synthetic challenges of naturally occurring glycosidic link-
ages have employed alternative conjugation chemistries.
The most commonly used method for glycoconjugate
formation is reductive amination, where the reducing
end of the polysaccharide is condensed with an amine
on the protein or peptide of interest to form a Schiff
base, which is then reduced to form a secondary amine.3
However, reductive amination creates an acyclic struc-
ture at the reducing terminus of the oligosaccharide,
which may have unintended consequences when the bio-
logical activity of the conjugate is evaluated.
Recent developments in chemoselective methods for
forming glycoconjugates have provided new routes to
multifunctional glycoconjugates of high complexity.
These methods use highly nucleophilic functional
groups such as N-alkylhydroxylamines, hydrazides and
semithiocarbazides to specifically condense with the
reducing terminus of the oligosaccharide (Fig. 1).4,5
These methods are very useful for derivatizing the small
amounts of oligosaccharides available from biological
samples as they can be carried out under mild condi-
tions, in the presence of other functional groups found
in biological systems and give the b-pyranosides in a
high yield.
Specifically, the chemoselective condensation of
hydrazides with reducing saccharides has found wide
use for the formation of glycoconjugates for applica-
tions including biotin labelling,6 formation of glyco-
arrays7 and the generation of glycopeptide analogues.8
Detailed studies of 1-glycosyl-2-acetylhydrazine have
determined the saccharide to exist as the b-pyranoside,
as a cis/trans equilibrium mixture about the amide bond
*
Corresponding author. Tel.: +1 416 946 0640; e-mail: mnitz@
chem.utoronto.ca
0008-6215/$ - see front matter ꢁ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2006.12.015

750
A. V. Gudmundsdottir, M. Nitz / Carbohydrate Research 342 (2007) 749–752
OH
O
HN O
R
HO
N
R
O
HO
O
OH
O
OH
OH
R
O
H2N NH
Glucose
H
N
HO
HO
N
H
R
OH
S
NH2
H2N NH
O
S
H
N
HO
HO
N
H
NH2
OH
Figure 1. Chemoselective ligation methods based on the reactivity of a
free hemiacetal with a potent nucleophile.
and to be rapidly cleaved in the presence of a strong acid
(0.1 M H2SO4).9 Contradictory reports regarding the
stability of the glycosyl hydrazide linkages under
physiologically relevant conditions have appeared in
the literature.10 But, despite the wide utility of this
chemoselective method, no quantitative studies have
been reported that determine the hydrolytic stability of
the glycosylhydrazides. Here we communicate the rates
of hydrolysis of a series of 1-glucosyl-2-benzoylhydr-
azines at the physiologically relevant pH values of 4.0,
5.0 and 6.0 (Fig. 2).
Figure 3. Hydrolysis of 1-glucosyl-2-phenylhydrazide (50 mM
NaOAc, 5% MeOH, pH 4.0, 50 ꢀC) was monitored using reverse
phase HPLC (C18, MeOH/H2O gradient 7.2–8.0%, 20 min). Phenyl-
hydrazide (13 min) 1-glucosyl-2-phenylhydrazide appears as two peaks
(cis/trans) at retention times 14 min and 16.5 min.
age of total hydrazide observed (Fig. 4). As can be seen
in Table 1, the rate of hydrolysis is strongly pH-depen-
dent, slowing significantly as the pH approaches neu-
trality. This evidently suggests that the rate-limiting
step for hydrolysis is acid catalyzed over the pH range
investigated. The hydrolysis rates of the conjugates fol-
lowed the pKa values for the parent hydrazides with
the most stable conjugate being formed with p-nitro-
benzylhydrazide (pKa 11.28) followed p-chloro-
benzylhydrazide (pKa 12.09), benzylhydrazide (pKa
12.52) and finally the least stable conjugate was formed
with p-methoxyhydrazide (pKa 12.83).12 This suggests
that more electron withdrawn hydrazides may form
even more stable conjugates.
Four 1-glucosyl-2-phenylhydrazides were produced
using acid-catalyzed condensation of glucose with the
respective hydrazide in ethanol (Fig. 2).11 At the limits
1
of detection by H NMR, only the b-pyranosides could
be observed in all cases.
The hydrolysis of the conjugates was followed by
reverse-phase high performance liquid chromatography
(HPLC). A representative time course of an elution pro-
file is presented in Figure 3. As has been previously
observed, the 1-glycosyl-2-acetylhydrazines elute as two
peaks corresponding to the cis–trans isomers about the
amide bond.9 These peaks were shown to reequilibrate
after collection, lyophilization and reinjection. Surpris-
ingly, the cis/trans isomers of the compounds could
1
not be observed by H or 13C NMR. In calculation of
the overall hydrolysis rate, the areas of the cis and trans
isomers were combined. No reverse reaction (glucosyl-
hydrazide formation) was observed under the hydrolysis
conditions.
The observed pseudo-first-order rate constant for
hydrolysis was determined by directly fitting the areas
of 1-gluco-2-benzoylhydrazide remaining as a percent-
OH
O
OH
HO
HO
O
O
H
N
HO
HO
OH
OH
H2O
N
H
+
OH
R
O
Figure 4. Hydrolysis of 1-glucosyl-2-phenylhydrazide in (j) 50 mM
NaOAc (5% methanol) pH 4.0; (d) 50 mM NaOAc (5% methanol)
pH 5.0; (m) 50 mM Na2HPO4 (5% methanol) pH 6.0, at 50 ꢀC. Each
value represents the average of two experiments.
H2N
N
H
R = 1) H, 2) OCH3, 3) Cl, 4) NO2
R
Figure 2. Glucosylhydrazide hydrolysis reactions studied.

A. V. Gudmundsdottir, M. Nitz / Carbohydrate Research 342 (2007) 749–752
751
Table 1. Hydrolysis rates of 1–4
Hydrazide R=
pH 4.0a
kobs (sÀ1) · 10À4
2.5
2.2
1.9
1.5
pH 5.0a
kobs (sÀ1) · 10À5
pH 6.0a
kobs (sÀ1) · 10À6
t1/2 (h)
t1/2 (h)
t1/2 (h)
OMe (2)
H (1)
Cl (3)
0.78
0.89
0.99
1.3
4.3
3.2
2.7
2.2
4.5
6.0
7.2
8.9
5.0
3.7
3.7
3.4
39
51
52
57
NO2 (4)
a 5 mM sample (50 mM NaOAc, 5% MeOH pH 4.0, 50 mM NaOAc 5% MeOH pH 5.0 or 50 mM NaOAc 5% MeOH pH 6.0) incubated at 50 ꢀC
200 lL samples were taken out at different time intervals and quenched with the addition of 400 lL of 4 ꢀC 200 mM phosphate buffer at pH 7.0 and
immediately analyzed by HPLC. Standard deviation between replicate runs was between 3% and 8%.
To determine whether the hydrolysis proceeds via
general or specific acid catalysis, the rates of hydrolysis
were measured at a constant pH but with increasing
buffer concentrations from 50 to 500 mM NaOAc
(Fig. 5). The minimal acceleration in rate observed at
the highest buffer concentration suggests that the hydro-
lysis is predominantly specific acid catalyzed. Thus, the
observed rates can be extrapolated to other buffer sys-
tems where glycoconjugates might be used.
hydrazide being the most stable. The derivatives are sta-
ble at a neutral pH for extended periods of time and
serve as excellent candidates for chemoselective ligation
reactions to prepare neo-glycoconjugates. It has been
observed that with other carbohydrates, mixtures of gly-
cosides be formed with hydrazides and it is likely that
these will hydrolyze at different rates.9 However, given
the hydrolysis rates presented here, caution should be
employed when the glycoconjugates are used under
mildly acidic conditions, such as those found in the lyso-
some, as at these pH values the glycoconjugates would
rapidly hydrolyze.
Under neutral conditions (pH 7, 50 mM Na2HPO4
1
(5% MeOH)) the samples showed no hydrolysis by H
NMR after being kept at room temperature for 37 days,
consistent with the determined hydrolysis rates. An
extrapolation of the observed rates of hydrolysis to
physiological conditions (pH 7.4, 37 ꢀC) yields the half
lives of hydrolysis of approximately 50 days.
1. Experimental
In conclusion, we have established that the reactions
of phenyl hydrazides with glucose occur stereospecifi-
cally to give the b-glycopyranosides. The rate of the
hydrolysis of these glycosides is strongly influenced by
the pH of the solution. Their hydrolysis shows a minor
dependence on the electronic properties of the hydra-
zide, with the conjugate formed with the lowest pKa
1.1. General methods
NMR spectra were recorded on Varian NMR-400 and
Mercury-400 MHz spectrometers. Chemical shifts were
reported in ppm using the solvent residue signals as
reference. HPLC analyses were performed on Dionex
BioLC (PDA-100 Photodiode Array Detector, GS50
Gradient pump and A550 Autosampler) using a Waters
Symmetryꢂ C18 reverse phase analytical column. The
buffers used in hydrolysis experiments were (A):
50 mM NaOAc (5% MeOH) at pH 4.0; (B): 50 mM
NaOAc (5% MeOH) at pH 5.0 and (C): 50 mM Na2H-
PO4 (5% MeOH) at pH 6.0.
1.1.1. 1-Glucosyl-2-benzoylhydrazine (1). To a suspen-
sion of glucose (477 mg, 2.7 mmol) in 2 mL of ethanol
and two drops of glacial acetic acid was added phen-
ylhydrazide (540 mg, 4.0 mmol). The reaction mixture
was stirred and heated at reflux for 3 h. The product pre-
cipitated and was purified by hot ethanol filtration and
1
gave 654 mg (83%) of 1. H NMR (400 MHz, D2O): d
7.75 (m, 2H, Ar), 7.63 (m, 1H, Ar), 7.53 (m, 2H, Ar),
4.22 (d, 1H, J1,2 = 9.0 Hz, H-1), 3.92 (dd, 1H,
J6a,6b = 12.2 Hz, J5–6a = 2.1 Hz, H-6a), 3.73 (dd, 1H,
J5–6b = 5.7 Hz, H-6b), 3.57 (t, 1H, J3,4 = 9.0 Hz, H-3),
3.46 (ddd, 1H, J4,5 = 9.8 Hz, H-5), 3.40 (t, 2H,
J2,3 = 9.3 Hz, H-4, H-2); 13C NMR (100 MHz, D2O):
d 170.8, 132.5, 132.0, 128.9 (2), 127.4 (2), 90.0, 77.0,
Figure 5. Determination of specific acid hydrolysis of 1-glucosyl-2-
phenylhydrazide in (j) 50 mM NaOAc (5% methanol) pH 4.0; (d)
250 mM NaOAc (5% methanol) pH 4.0; (m) 500 mM NaOAc (5%
methanol) pH 4.0, at 50 ꢀC. Each value represents the average of two
experiments.

752
A. V. Gudmundsdottir, M. Nitz / Carbohydrate Research 342 (2007) 749–752
76.4, 70.8, 69.7, 61.0; ESI-MS m/z calcd for
ESI-MS m/z calcd for [C13H17N3O8+Na]+: 366.0925.
Found: 366.0907.
[C13H18N2O6+Na]+: 321.1064. Found: 321.1057.
1.1.2. 1-Glucosyl-2-p-methoxybenzoylhydrazine (2).
Using the same procedure as described for compound
1, glucose (300 mg, 1.7 mmol) and p-meth-
oxyphenylhydrazide (415 mg, 2.5 mmol) gave 363 mg
1.2. General procedure for hydrolysis experiments
A 5 mM solution of the hydrazide was prepared by dis-
solving compound 1 (3.3 mg, 10 lmol) in 1.9 mL buffer
A and 100 lL MeOH. The solution was incubated at
50 ꢀC in a water bath, and 200 lL samples were taken
out at defined time intervals and quenched by the addi-
tion of 400 lL of 4 ꢀC 200 mM phosphate buffer at
pH 7.0. The sample was kept on ice until the analysis
was carried out by HPLC. The data was fit to an expo-
nential decay using the software Origin, v = k[glucos-
ylhydrazide], where k equals the pseudo-first-order
kinetic constant for the appearance of the product.
1
(66%) of 2. H NMR (100 MHz, D2O): d 7.79 (d, 2H,
J = 8.9 Hz, pOMePh), 7.11 (d, 2H, J = 8.9 Hz, pO-
MePh), 4.25 (d, 1H, J1,2 = 9.0 Hz, H-1), 3.94 (dd, 1H,
J6a,6b = 12.4 Hz, J5–6a = 2.3 Hz, H-6a), 3.91 (s, 3H,
–OCH3), 3.75 (dd, 1H, J5–6b = 5.8 Hz, H-6b), 3.58 (t,
1H, J3,4 = 9.0 Hz, H-3), 3.47 (ddd, 1H, J4,5 = 9.8 Hz,
H-5), 3.41 (t, 1H, H-4), 3.40 (t, 1H, J2,3 = 9.1 Hz, H-
2); 13C NMR (100 MHz, D2O): d 170.2, 162.4, 129.5
(2), 124.5, 114.2 (2), 90.2, 77.1, 76.4, 70.9, 69.8, 61.0,
55.6; ESI-MS m/z calcd for [C14H20N2O7+Na]+:
351.1166. Found: 351.1162.
Supplementary data
1.1.3.
1-Glucosyl-2-p-chlorobenzoylhydrazine
(3).
Using the same procedure as described for compound
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.carres.
2006.12.015.
1, glucose (150 mg, 0.8 mmol) and p-chlorophenylhydr-
1
azide (284 mg, 1.6 mmol) gave 182 mg (66%) of 3. H
NMR (400 MHz, D2O): d 7.76 (d, 2H, J = 8.5 Hz,
pClPh), 7.58 (d, 2H, J = 8.5 Hz, pClPh), 4.26 (d, 1H,
J1,2 = 9.0 Hz, H-1), 3.94 (dd, 1H, J6a,6b = 12.2 Hz,
J5–6a = 1.9 Hz, H-6a), 3.75 (dd, 1H, J5–6b = 5.8 Hz,
H-6b), 3.58 (t, 1H, J3,4 = 9.0 Hz, H-3), 3.47 (ddd, 1H,
J4,5 = 9.7 Hz, H-5), 3.41 (t, 1H, H-4), 3.40 (t, 1H,
J2,3 = 9.1 Hz, H-2); 13C NMR (100 MHz, D2O): d
169.8, 138.1, 130.7, 129.0 (4), 90.0, 77.1, 76.4, 70.9,
69.7, 61.0; ESI-MS m/z calcd for [C13H17N2O6Cl+Na]+:
355.0667. Found: 355.0667.
References
1. Koeller, K. M.; Wong, C. H. Nat. Biotechnol. 2000, 18,
835–841.
2. Danishefsky, S. J.; Allen, J. R. Angew. Chem., Int. Ed.
2000, 39, 836–863.
3. Gray, G. R. Arch. Bioch. Biophys. 1974, 163, 426–428.
4. Rodriguez, E. C.; Marcaurelle, L. A.; Bertozzi, C. R.
J. Org. Chem. 1998, 63, 7134–7135.
5. Larsen, K.; Thygesen, M. B.; Guillaumie, F.; Willats, W.
G. T.; Jensen, K. J. Carbohydr. Res. 2006, 341, 1209–
1234.
6. Leteux, C.; Childs, R. A.; Chai, W.; Stoll, M. S.;
Kogelberg, H.; Feizi, T. Glycobiology 1998, 8, 227–236.
7. Lee, M.-r.; Shin, I. Org. Lett. 2005, 7, 4269–4272.
8. Peluso, S.; Imperiali, B. Tetrahedron Lett. 2001, 42, 2085–
2087.
1.1.4. 1-Glucosyl-2-p-nitrobenzoylhydrazine (4). Using
the same procedure as described for compound 1, glu-
cose (820 mg, 4.6 mmol) and p-nitrophenylhydrazide
(1.5 g, 8.4 mmol) gave 940 mg (60%) of 4. 1H NMR
(400 MHz, D2O): d 8.38 (d, 2H, J = 8.7 Hz, pNPh),
7.98 (d, 2H, J = 8.7 Hz, pNPh), 4.29 (d, 1H, J1,2
=
8.9 Hz, H-1), 3.93 (dd, 1H, J6a,6b = 12.2 Hz,
J5–6a = 2.1 Hz, H-6a), 3.76 (dd, 1H, J5–6b = 5.8 Hz,
H-6b), 3.59 (t, 1H, J3,4 = 9.0 Hz, H-3), 3.49 (ddd, 1H,
J4,5 = 9.7 Hz, H-5), 3.42 (t, 2H, J2,3 = 9.3 Hz, H-4, H-
2); 13C NMR (100 MHz, D2O): d 168.8, 149.8, 138.3,
128.8 (2), 124.0 (2), 90.0, 77.1, 76.5, 70.9, 69.7, 61.0;
9. Bendiak, B. Carbohydr. Res. 1997, 304, 85–90.
10. Ridley, B. L.; Spiro, M. D.; Glushka, J.; Albersheim, P.;
Darvill, A.; Mohnen, D. Anal. Biochem. 1997, 249, 10–
19.
11. Takeda, Y. Carbohydr. Res. 1979, 77, 9–23.
12. Titov, E. V.; Korzhenevskaya, N. G.; Rybachenko, V. I.
Ukr. Khim. Zh. (Russ. Ed.) 1968, 34, 1253–1256.