Kinetics of hydrolysis of chitin/chitosan oligomers in concentrated hydrochloric acid

Article abstract of DOI:10.1016/j.carres.2007.02.022

The kinetics of hydrolysis in concentrated hydrochloric acid (12.07 M) of the fully N-acetylated chitin tetramer (GlcNAc₄) and the fully N-deacetylated chitosan tetramer (GlcN₄) were followed by determining the amounts of the lower DP oligomers as a function of time. A theoretical model was developed to simulate the kinetics of hydrolysis of the three different glycosidic linkages in the tetramers. The model uses two different rate constants for the hydrolysis of the glycosidic bonds in the oligomers, assuming that the glycosidic bond next to one of the end residues are hydrolysed faster than the two other glycosidic linkages. The two rate constants were estimated by fitting model data to experimental results. The results show that the hydrolysis of the tetramers is a nonrandom process as the glycosidic bonds next to one of the end residues are hydrolysed 2.5 and 2.0 times faster as compared to the other glycosidic linkages in the fully N-acetylated and fully N-deacetylated tetramer, respectively. From previous results on other oligomers and the reaction mechanism, it is likely that the glycosidic bond that is hydrolysed fastest is the one next to the nonreducing end. The absolute values for the rate constants for the hydrolysis of the glycosidic linkages in GlcNAc₄ were found to be 50 times higher as compared to the glycosidic linkages in GlcN₄, due to the catalytic role of the N-acetyl group and the presence of the positively charged amino-group on the N-deacetylated sugar residue.

Full text of DOI:10.1016/j.carres.2007.02.022

Carbohydrate Research 342 (2007) 1055–1062
Kinetics of hydrolysis of chitin/chitosan oligomers
in concentrated hydrochloric acid
*
˚
Aslak Einbu, Hans Grasdalen and Kjell M. Varum
Norwegian Biopolymer Laboratory (NOBIPOL), Department of Biotechnology,
Norwegian University of Science and Technology, 7491 Trondheim, Norway
Received 21 December 2006; received in revised form 9 February 2007; accepted 19 February 2007
Available online 24 February 2007
Abstract—The kinetics of hydrolysis in concentrated hydrochloric acid (12.07 M) of the fully N-acetylated chitin tetramer (Glc-
NAc4) and the fully N-deacetylated chitosan tetramer (GlcN4) were followed by determining the amounts of the lower DP oligo-
mers as a function of time. A theoretical model was developed to simulate the kinetics of hydrolysis of the three different
glycosidic linkages in the tetramers. The model uses two different rate constants for the hydrolysis of the glycosidic bonds in the
oligomers, assuming that the glycosidic bond next to one of the end residues are hydrolysed faster than the two other glycosidic
linkages. The two rate constants were estimated by fitting model data to experimental results. The results show that the hydrolysis
of the tetramers is a nonrandom process as the glycosidic bonds next to one of the end residues are hydrolysed 2.5 and 2.0 times
faster as compared to the other glycosidic linkages in the fully N-acetylated and fully N-deacetylated tetramer, respectively. From
previous results on other oligomers and the reaction mechanism, it is likely that the glycosidic bond that is hydrolysed fastest is the
one next to the nonreducing end. The absolute values for the rate constants for the hydrolysis of the glycosidic linkages in GlcNAc4
were found to be 50 times higher as compared to the glycosidic linkages in GlcN4, due to the catalytic role of the N-acetyl group and
the presence of the positively charged amino-group on the N-deacetylated sugar residue.
ꢀ 2007 Elsevier Ltd. All rights reserved.
Keywords: Oligomer; Acid-catalysed hydrolysis; Theoretical model; Reaction rate constants
1. Introduction
(glucosamine) that can be N-acetylated to varying
extents. The glycosidic linkage following an N-acety-
lated unit has been found to be hydrolysed at conditions
favouring acid-catalysed hydrolysis three orders of mag-
nitude faster as compared to the glycosidic linkage fol-
lowing a N-deacetylated unit.2 This can be explained
by two main factors, that is, a catalytic role of the N-
acetyl group at carbon 2 which stabilises the oxocarbo-
nium ion transition state resulting in an enhanced rate of
cleavage of the glycosidic linkage following the N-acet-
ylated residue,3 and the presence of a positively charged
amino-group at carbon 2 in a glucosamine residue that
would shield the glycosidic oxygen from protonation
and thereby result in a decreased rate of cleavage of
the glycosidic linkage following a glucosamine residue.4
A complicating factor when studying acid-catalysed
hydrolysis of chitin oligomers is the concomitant
acid-catalysed N-deacetylation reaction. However, in
Acid-catalysed hydrolysis of glycosidic linkages can
occur with widely varying reaction rates, depending
on, for example, ring size, configuration, conformation
and structure of the glycon. Differences in reaction rates
can be utilised for characterisation of polysaccharides
after purification and subsequent structural character-
isation of the oligosaccharides from acid-catalysed
hydrolysis.1 One example of a very large difference in
the relative rates of acid-catalysed hydrolysis of
glycosidic linkages in the same polysaccharide is from
chitosans, a family of polysaccharides composed of
(1!4)-linked 2-amino-2-deoxy-b-D-glucopyranose units
*
Corresponding author. Tel.: +47 73593324; fax: +47 73591283;
e-mail: kvaarum@nt.ntnu.no
0008-6215/$ - see front matter ꢀ 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.carres.2007.02.022

1056
A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
concentrated hydrochloric acid (12.07 M), the rate of
acid-catalysed cleavage of the glycosidic linkage was
found to be much higher than the N-deacetylation reac-
tion.2,5,6 Thus, at this high acid concentration a study of
the acid-catalysed cleavage of the glycosidic linkages
without concomitant N-deacetylation should be possi-
ble. The acid-catalysed hydrolysis of starch, cellulose
and chitin has been reported to be a nonrandom depoly-
merisation reaction.2,7,8 When the calculated amounts
of D-glucose and oligosaccharides liberated by the
hydrolysis of a given weight of starch or cellulose are
compared with the actual amounts formed, it was found
that there are more products of low degree of polymer-
isation and fewer intermediate-sized products than
would be expected from a completely random hydro-
lysis, for any degree or time of hydrolysis.8,9 Thus, it
appears that acid-catalysed hydrolysis does not occur
by random scission. However, a complicating factor
when studying acid-catalysed depolymerisation of poly-
saccharides such as cellulose and chitin is their insolubil-
ity in aqueous solvents.5
found for the introduction of various sized groups at
C-5.1
Acid-catalysed hydrolysis of model oligosaccharides
of defined composition and chain length has certain
advantages compared to polysaccharides, such as full
water-solubility and fully defined composition and
chain length, and such studies have been mainly
focused on comparing the rates of terminal glycosidic
linkages in relation to internal linkages. Jones et al.10
concluded that the bond of one of the end residues
of isomaltotriose is hydrolysed 1.7 times faster as com-
pared to the reducing end. Weintraub and French11
showed experimentally that the glycosidic linkage next
to the nonreducing end of starch chains was hydrolysed
the fastest by acid. Furthermore, the rate constants
for acid-catalysed hydrolysis of the nonreducing end
glycosidic linkages of maltotriose and maltohexose
were found to be 1.8 times higher as compared to the
other glycosidic linkages.12 Feather and Harris studied
the kinetics of the hydrolysis of cellotriose, and com-
pared their experimental data with model data and
found that the rate of hydrolysis of the nonreducing
end glycosidic linkage was 1.5 times higher compared
to the reducing end glycosidic linkage.13 However, to
the best of our knowledge, modelling of the hydrolysis
of the glycosidic linkages of a tetramer has not been
published. We now report a study on the kinetics of
hydrolysis of the three glycosidic linkages of the fully
N-acetylated chitin tetramer and the fully N-deacetyl-
ated chitosan tetramer in concentrated HCl. The for-
mation of the different oligomers as a function of
time were followed quantitatively, and compared and
optimised to model data on the formation of oligomers
obtained by assuming a different rate for the hydrolysis
of the bond of the nonreducing end residue as com-
pared to the other two glycosidic linkages, and very
good correlations were obtained when rates of hydro-
lysis of the bond of one of the end residues were 2.5
and 2.0 times higher as compared to the other two
glycosidic linkages for the fully N-acetylated and fully
N-deacetylated tetramers, respectively.
The most accepted reaction mechanism for the acid-
catalysed hydrolysis of a glycosidic linkage is illustrated
in Chart 1.
The mechanism involves protonation of the glycosidic
oxygen atom to form the conjugate acid, followed by
heterolysis of the exocyclic O-5 to C-1 bond to give a
cyclic carbonium–oxonium ion which most probably
exists in the half-chair conformation having C-2, C-1,
O-5 and C-5 in a plane. Reaction with water then gives
the protonated reducing sugar, and from it, the reducing
sugar is formed. The nonrandom acid-catalysed hydro-
lysis of polysaccharides can be explained by this reaction
mechanism, as the glycosidic linkage at the nonreducing
end is hydrolysed more rapidly than the other glycosidic
linkages in the chain. This conclusion was reached from
the fact that hydrolysis of an internal linkage would
involve, in the formation of the carbonium–oxonium
ion, reorientation of an entire chain; that is, hydrolysis
would be diminished by the introduction of a large
bulky group, in a manner similar to what has been
Chart 1. Schematic illustration of the proposed reaction mechanism for the acid-catalysed hydrolysis of the glycosidic linkage.

A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
1057
2. Experimental
Our model for hydrolysis of a tetramer is based on the
hypothesis that the glycosidic bond next to one of the
end residues of an oligosaccharide undergoes hydrolysis
more rapidly than the other bonds. Based on this
assumption, the model applies two different rate con-
stants: rate constant k1 refers to the hydrolysis of the
glycosidic linkage next to one of the end residues of
the oligomer while the rate constant k2 refers to the
hydrolysis of the other glycosidic linkages.
Chart 2 shows a model for the cleavage and formation
of the different oligomers starting from the tetramer,
and forms the basis for the four differential equations
describing the formation and cleavage of the oligomers
during hydrolysis with rate constants k1 and k2. In Chart
2 it has been assumed that the rate constant for the
hydrolysis of the glycosidic linkage next to the nonreduc-
ing end (k1) is hydrolysed faster than the other glycosidic
linkages.
2.1. Hydrolysis of chitin/chitosan tetramer in
concentrated HCl
To samples of GlcNAc4 and GlcN4 (2 mg) (Seikagaku,
Japan) were added 100 lL concentrated hydrochloric
acid (12.07 M) and the samples were incubated at
40 ꢁC. The reaction was stopped by cooling the solution
on dry ice and then adjusting the pH to 4.5 with NaOH
to prevent heating of the sample upon dilution. Prior to
the reaction, the HCl was saturated with nitrogen gas to
remove oxygen in order to prevent oxidation reactions.
The reaction mixture was kept in reaction tubes which
were sealed during reaction.
2.2. Size-exclusion chromatography (SEC) of reaction
products
The change in the molar concentrations of the differ-
ent oligomers in the reaction mixture during hydrolysis
can be expressed by differential equations derived from
the cleavage and formation of the different oligomers
as seen schematically in Chart 2.
The tetramer can either be split into two dimers or one
trimer and one monomer (Chart 2). Eq. 1 expresses the
decay in the molar concentration of tetramer, [A4], as a
function of [A4], and the rate constants k1 and k2
The oligomers from the neutralised reaction mixtures
were separated on three SuperdexTM30 columns con-
nected in series as previously described.14 (Amersham
Pharmacia Biotech, overall dimensions 2.6 · 180 cm).
The columns were eluted with 0.15 M ammonium ace-
tate, pH 4.5 at a flow rate of 0.8 mL/min. The effluent
was monitored with an online refractive index (RI)
detector (Shimadzu RID 6A, Shimadzu Schweiz GmbH,
Reinach, Switzerland), coupled to a data logger. Sørbot-
ten et al.14 investigated the relationship between detector
response and the mass of injected oligomer when using
fully acetylated and fully N-deacetylated dimer, tetra-
mer and hexamer. Within the accuracy of the area deter-
mination, there was a linear relationship between peak
areas and the amount (mass) of injected oligomer, irre-
spective of DP and degree of acetylation. On the basis
of this, the molar fractions of the different oligomers
can be calculated from the integrals of the chromato-
grams from SEC.
d
½A4ꢀ ¼ ꢁ½A4ꢀ ꢂ ðk1 þ 2k2Þ
ð1Þ
dt
Solving Eq. 1 with initial value of [A4(0)] = 1, gives Eq.
2 expressing [A4] as a function of k1, k2 and time, t
½A4ꢀ ¼ eꢁðk þ2k Þt
ð2Þ
1
2
The trimer can only be formed from cleavage of the tet-
ramer, while it can be cleaved to form a dimer and a
A4
k2
A4
2.3. Fitting of the mathematical model for hydrolysis to
experimental data
A3
k1
k2
k1
k1
k2
k2
The differential equations in Section 3 were solved by
programming symbolic algebra in Maple 10. Fitting of
the mathematical model to the experimental data was
done in EXCEL, using an iteration procedure to minimise
the sum of the root mean square deviations for experi-
mental and theoretical molar fractions of all oligomers.
A
A
A A
+
3 1
A2+A1
A1+A2
+
A2+A2
1
3
A2
A3
A4
k2
k2
k1
k1
A2
A1
k2
k1
k1
A3
A4
k1
k2
3. Theory
A1+A1
3.1. Theoretical model for depolymerisation of a tetramer
Chart 2. Cleavage and formation of the different oligomers starting
from the tetramer, A4 (top left), trimer, A3 (top right), dimer, A2
(down left) and monomer, A1 (down right). The hydrolysis of the
glycosidic bonds are described by the rate constants k1 and k2.
A model was developed in order to evaluate the hydro-
lysis of a tetramer compound.

1058
A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
monomer by two different rate constants (Chart 2). The
change in the molar amount of trimer, [A3], in the reac-
tion mixture is expressed in Eq. 3 as a function of [A4]
(Eq. 2), [A3], and the rate constants k1 and k2.
Eqs. 2, 4, 6 and 7 are used to determine [A4], [A3], [A2]
and [A1], respectively as a function of time with different
values of k1 and k2. Using the equations for the molar
concentrations of the four oligomers, the theoretical
model was compared with experimental molar fractions
and estimates of the rate constants k1 and k2 were deter-
mined as described in Section 2.3.
d
½A3ꢀ ¼ ½A4ꢀ ꢂ ðk1 þ k2Þ ꢁ ½A3ꢀ ꢂ ðk1 þ k2Þ
ð3Þ
dt
The first term of Eq. 3 represents the formation of trimer
from tetramer, and the second term represents the deg-
radation of trimer. Solving Eq. 3 with initial conditions
[A4(0)] = 1 and [A3(0)] = 0 gives Eq. 4 which describes
[A3] as a function of time and the rate constants k1
and k2.
4. Results and discussion
4.1. Hydrolysis of the fully N-acetylated chitin tetramer
GlcNAc4 in concentrated HCl
k1 þ k2
1
2
1
2
½A3ꢀ ¼
ꢂ ðeꢁðk þk Þt ꢁ eꢁðk þ2k Þt
Þ
ð4Þ
The size-distribution of the oligomers resulting from
acid-catalysed hydrolysis of (GlcNAc)4 in concentrated
HCl at 40 ꢁC was followed by gel filtration chromato-
graphy as a function of time (Fig. 1). The chromato-
grams show well-resolved peaks of each oligomer
which could be accurately quantified into molar frac-
tions (see Section 2). The tetramer was almost fully
hydrolysed to oligomers of a lower DP within 28 h in
concentrated HCl at 40 ꢁC. No detectable concentration
of oligomers with DP higher than 4 could be seen in the
chromatograms, indicating that transglycosylation
reactions were insignificant. When the hydrolysis was
performed in nitrogen atmosphere (see Section 2), no
development of coloured products could be seen even
after 28 h, and proton NMR-spectra showed essentially
the resonances of the acetylated monomer (data not
shown), indicating that any by-products were only
present in very low concentrations.
Due to the previously reported much faster hydrolysis
of the glycosidic bond compared to the hydrolysis of the
N-acetyl linkage of the chitin oligomers in concentrated
hydrochloric acid, insignificant N-deacetylation of the
oligomers is expected to take place in the time span of
the complete hydrolysis of tetramer into monomer.5,6
The chromatograms in Figure 1 show no detectable
amounts of any partially N-deacetylated oligomers,
which would appear as separate peaks eluting just after
the fully N-acetylated oligomer of the same DP.14
Hence, no significant N-deacetylation can be detected
even after the longest reaction time displayed in Figure
1 (28 h at 40 ꢁC in concentrated HCl), when the molar
fraction of monomer is about 0.90.
k2
Eq. 5 expresses the change in the molar amount of the
dimer, [A2], with time.
d
½A2ꢀ ¼ ½A4ꢀ ꢂ 2k2 þ ½A3ꢀ ꢂ ðk1 þ k2Þ ꢁ ½A2ꢀ ꢂ k1
ð5Þ
dt
The first term represents the formation of dimer from
the tetramer, the second term the formation of dimer
from the trimer and the last term the degradation of
the dimer (Chart 2). Solving Eq. 5 with initial conditions
[A4(0)] = 1, [A3(0)] = 0 and [A2(0)] = 0 gives Eq. 6
which expresses the [A2] as a function of time, k1 and k2
!
!
2
2
ðk1 þ k2Þ
ðk1 þ k2Þ
1
½A2ꢀ ¼ 1 þ
ꢂ eꢁk t
ꢁ
2k22
k22
!
2
ðk1 þ k2Þ
1 ꢁ
2k22
ꢂ eꢁðk þk Þt
ꢁ
ꢂ eꢁðk þ2k Þt ð6Þ
1
2
1
2
The increase in the molar amount of monomer, [A1],
with time is similarily given by Eq. 7 (see Chart 2)
d
½A1ꢀ ¼ ½A4ꢀ ꢂ ðk1 þ k2Þ þ ½A3ꢀ ꢂ ðk1 þ k2Þ þ ½A2ꢀ ꢂ 2k1
dt
ð7Þ
The three terms of Eq. 7 represents the formation of
monomer from tetramer, trimer and dimer, respectively.
Solving Eq. 7 with initial conditions [A4(0)] = 1,
[A3(0)] = 0, [A2(0)] = 0 and [A1(0)] = 0 gives an expres-
sion for [A1] as a function of time, k1 and k2.
!
2
ðk1 þ k2Þ
½A1ꢀ ¼ 2 þ
ꢂ ð1 ꢁ eꢁk t
Þ
1
k22
4.2. Comparison of experimental data and model data
ꢀ
ꢁ
k1 þ k2
k1
k2
1
2
þ
ꢂ
1 ꢁ 2
ꢂ ð1 ꢁ eꢁðk þk Þt
Þ
The mathematical model for the hydrolysis of a tetramer
(see Section 3) was fitted to the molar fractions deter-
mined from the chromatograms in Figure 1 by using
an iteration procedure to minimise the sum of all root
mean square deviations (RMSD) for experimental and
theoretical molar fractions of all oligomers. Molar frac-
tions of the four different oligomers together with the
k2
2
k1 þ k2
ðk1 þ k2Þ
2k1
þ
ꢁ
ꢁ
k1 þ 2k2 k2ðk1 þ 2k2Þ k1 þ 2k2
!!
2
ðk1 þ k2Þ
ꢂ ð1 ꢁ eꢁðk þ2k Þt
Þ
ð8Þ
1
2
ꢂ 1 ꢁ
2k22

A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
1059
optimised theoretical curves for molar fractions are plot-
ted as a function of time in Figure 2.
A4
150
100
50
The excellent fit between experimental and theoretical
data with k1 2.5 times larger than k2 indicates that the
hydrolysis of the tetramer is a nonrandom process,
and that one of the end residues is hydrolysed signifi-
cantly faster as compared to the other two glycosidic
linkages in the tetramer. From previous results on starch
and cellulose oligomers,10–12 and most accepted reaction
mechanism (Chart 1), it is likely that the glycosidic bond
that is hydrolysed fastest is the one next to the non-
reducing end. The difference between the two rate con-
stants is somewhat larger as compared to differences
found in, for example, cellotriose.13
0.5 h
3.5 h
A3
A2
A1
0
1020
770
820
870
920
970
Time (min)
150
100
50
A4
A3
A2
A1
0
1020
770
820
870
920
970
The robustness of the model with respect to predicting
the optimised values of k1 and k2 from the experimental
data was tested. This is visualised as a contour plot of
the sum of all root mean square deviations for experi-
mental and theoretical molar fractions of all oligomers
as a function of k1 and k2 in Figure 3. From Figure 3,
it can be seen that the sum of all root mean square devi-
ations has a global minimum when k1 = 2.5 · 10ꢁ5 sꢁ1
and k2 = 1.0 · 10ꢁ5 sꢁ1. Thus, these rate constants rep-
resent the optimal fitting of the theoretical data to our
experimental results and are the best estimates for the
two rate constants in the experiment. The data in Figure
3 clearly visualises that the hydrolysis is a nonrandom
process. The diagonal line indicates a random hydrolysis
were k1 = k2, whereas the global minimum is located
away from the diagonal. The density of the contour
intervals around the global minimum also reflect the size
of the error estimates for the determined rate constants
k1 and k2. From the plot, it can be seen that small
changes in the rate constants give significantly higher
values for the sum of all root mean square deviations,
indicating that the optimised rate constants are robust
with respect to predicting rate constants with good
accuracy.
150
Time (min)
A4
100
50
0
A3
A2
A1
5 h
7 h
1020
770
820
870
920
970
Time (min)
150
100
50
A4
A3
A1
A2
0
770
1020
820
870
Time (min)
920
970
150
A1
100
50
A2
12 h
16 h
28 h
A3
A4
0
1020
770
820
870
Time (min)
920
970
150
100
50
0
A1
A2
A3
1.0
0.9
0.8
A4
0.7
0.6
0.5
0.4
1020
770
820
870
920
970
A4
A3
A2
A1
Time (min)
150
100
50
A1
0.3
0.2
0.1
0.0
A2
A3
A4
0
0
5
10
15
20
25
30
770
1020
820
870
920
970
Time (hours)
Time (min)
Figure 2. Molar fractions of reaction products as a function of time
during acid-catalysed hydrolysis of the fully N-acetylated chitin
tetramer (A4) in concentrated hydrochloric acid at 40 ꢁC. Theoretical
values for the molar fractions are indicated by continuous lines, with
Figure 1. Chromatograms showing the size-distribution of oligomers
after acid-catalysed hydrolysis of tetra-N-acetylchitotetraose with
increasing reaction times in concentrated HCl at 40 ꢁC. The gel-
filtration chromatograms (SuperdexTM30) display peaks of the tetramer
(A4), trimer (A3), dimer (A2) and monomer (A1).
k1 = 2.5 Æ 10ꢁ5 sꢁ1 and k2 = 1.0 Æ 10ꢁ5 sꢁ1
.

1060
A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
200
Total sum of RMSD
D4
5
4
3
2
1
150
100
50
5 h
D3
0
800
900
1000
1100
-50
200
150
Time (min)
D4
100
50
0
21 h
D3
D2
800
900
1000
1100
-50
Time (min)
200
150
100
50
D4
1
2
3
4
5
166 h
260 h
336 h
k1 ( x 10-5 s-1)
D3
D2
0
Figure 3. Contour plot of the sum of all Root Mean Square Deviations
for experimental and theoretical molar fractions of all oligomers as a
function of k1 and k2. Contour interval = 0.05.
800
900
1000
1100
-50
Time (min)
200
150
100
50
D4
4.3. Hydrolysis of the fully N-deacetylated chitosan
tetramer GlcN4 in concentrated HCl
D3
D2
0
800
-50
900
1000
1100
The size-distribution of the oligomers resulting from
acid-catalysed hydrolysis of the fully N-deacetylated
chitosan tetramer in concentrated HCl at 40 ꢁC was also
followed by gel filtration chromatography as a function
of time (Fig. 4). The chromatograms show well-resolved
peaks of oligomers with DP 2–4 which could be quanti-
fied into molar fractions (see Section 2).
Time (min)
200
150
100
50
D4
D3 D2
0
800
900
1000
1100
No evidence of significant transglycosylation reac-
tions (oligomers with DP > 4) or byproducts (coloured
products or unidentified resonances in the proton
NMR-spectra) when the hydrolysis reaction was per-
formed in nitrogen gas atmosphere. The only exception
was the most extensively hydrolysed sample (460 h),
which developed a slight yellow-brown colour.
-50
Time (min)
200
150
100
50
D2
460 h
D3
D4
0
As expected, due to the previously found very much
slower rate constant for the acid-catalysed cleavage of
the glycosidic linkages followed by a GlcN-unit, the
fully N-deacetylated oligomer was hydrolysed much
slowly than the acetylated oligomer in concentrated
HCl. This can be clearly seen from the very much
extended reaction times in order to obtain detectable
amounts of the lower oligomers in Figure 4 as compared
to Figure 1.
800
-50
900
1000
Time (min)
1100
Figure 4. Chromatograms showing the size-distribution of oligomers
after acid-catalysed hydrolysis of fully N-deacetylated chitosan tetra-
mer with increasing reaction times in concentrated HCl at 40 ꢁC. The
gel-filtration chromatograms (SuperdexTM30) display peaks of the
tetramer (D4), trimer (D3) and dimer (D2).
It was not possible to quantify the N-deacetylated
monomer, glucosamine, as it was eluting together with
the large salt peak (resulting from large amounts of
NaCl from neutralisation of the samples of concen-
trated HCl) in our chromatographic system. Quantifi-
cation of oligomer fractions was therefore restricted to
the tetramer, trimer and dimer, which made it impossi-
ble to determine the correct molar fractions of the dif-
ferent oligomers. Instead, the sum of relative molar
fractions of tetramer, trimer and dimer was set equal

A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
1061
Hydrolysis of acetylated tetramer
Hydrolysis of N-deacetylated tetramer
a
b
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
A3
A2
A4
D3
D2
D4
0
100 200 300 400 500 600 700 800 900 1000
0
5
10
15
20
25
30
Time (hours)
Time (hours)
Figure 5. Relative molar fractions of tetramer (A4/D4), trimer (A3/D3) and dimer (A2/D2) (as a fraction of the sum of tetramer, trimer and dimer)
as a function of time during acid-catalysed hydrolysis in concentrated HCl at 40 ꢁC. Hydrolysis of acetylated tetramer (a) and N-deacetylated
tetramer (b). Optimal fitting of the theoretical values for the relative fractions are indicated by continuous lines.
to 1. In order to fit the mathematical model for hydro-
lysis of chitin tetramer with the experimental results
from hydrolysis of the chitosan tetramer, fractions of
tetramer, trimer and dimer as a fraction of the sum
of all three oligomers were plotted as a function of
time in Figure 5b. The results from hydrolysis of
GlcNAc4 were also plotted in the same way in order
to directly compare the data from the hydrolysis of
GlcN4 (Fig. 5a).
The continuous lines in Figure 5 show the optimal
fitting of the theoretical fractions to the experimental
results. The optimal fit was determined when the ratio
k1/k2 is 2.0 and 2.4 for the hydrolysis of N-deacetylated
and acetylated tetramer, respectively. These results indi-
cate that the hydrolysis is a nonrandom process for both
tetramers, and the glycosidic linkages next to one of the
end residues are hydrolysed faster than the two other
glycosidic linkages. As for the results with the chitin
oligomers, it is likely to assume that the glycosidic bond
that is hydrolysed fastest is the one next to the non-
reducing end.
4.4. Summary of estimated reaction rate constants
Table 1 summarises values of the reaction rate constants
which give the optimal fittings to experimental values in
figure as well as the reaction rates determined for the
fully N-acetylated chitin tetramer.
The estimated rate constants derived from fitting of
our model in Figure 5 are based on data from only three
of the four oligomers present, while the estimated rate
constants from Figure 2 are based on all four oligomers
present during reaction. From the data for hydrolysis of
the fully N-acetylated chitin tetramer, it can be seen that
the accuracy is slightly better when using all four oligo-
mers (Table 1, last column, row 1 and 2). However, the
ratio k1/k2 for the hydrolysis of the two different oligo-
mers are the same within the accuracy of the ratios.
The reaction rate constants of the determined for
GlcNAc4 herein are comparable to those determined
for GlcNAc2 previously at the same conditions, taken
into account that the previous results were obtained
in deuterated hydrochloric acid.6 The reaction rate
constants for the hydrolysis of the glycosidic bonds
following an acetylated unit is about 50 times higher
as compared to the glycosidic bond following a
N-deacetylated unit (Table 1), and can be explained
by the catalytic role of the N-acetyl group and the
presence of the positively charged amino group in the
N-deacetylated unit (see Section 1). This difference
between the two constants is at least an order of
From the continuous lines of the optimal fit in Figure
5b, it seems likely that the model slightly overestimates
the amount of tetramer and underestimates the amount
of dimer during the initial reaction times. One possible
explanation can be a difference in all three reaction rate
constants for the cleaving of the three bonds in the
tetramer, increasing when going from the reducing end
towards the nonreducing end, or vice versa.
Table 1. Estimated reaction rate constants, k1 and k2, and the ratio (k1/k2) for the hydrolysis of GlcNAc4 and GlcN4
k1 (sꢁ1 k2 (sꢁ1
)
)
Ratio (k1/k2)
Hydrolysis of N-acetylated tetramer (estimate from Fig. 2)
Hydrolysis of N-acetylated tetramer (estimate from Fig. 5)
Hydrolysis of N-deacetylated tetramer (estimate from Fig. 5)
2.5 0.1 · 10ꢁ5
2.4 0.1 · 10ꢁ5
4.6 0.2 · 10ꢁ7
1.0 0.1 · 10ꢁ5
1.0 0.1 · 10ꢁ5
2.4 0.2 · 10ꢁ7
2.5 0.3
2.4 0.4
2.0 0.3

1062
A. Einbu et al. / Carbohydrate Research 342 (2007) 1055–1062
magnitude lower as compared to the difference in the
rate constants as determined by us previously.2 How-
ever, these rate constants were determined at a different
acid concentration and temperature (0.4 M HCl,
60 ꢁC). We speculate that the difference may be
explained by a higher relative increase in the rate
constant with increasing acid concentration of the N-
deacetylated tetramer as compared to the N-acetylated
tetramer. However, this could not be verified in a con-
trol experiment (data not shown) where the relative
rates of the two oligomers were compared at 6 and
12 M HCl, where both oligomers showed the same
relative rate increase (ꢃ6 times). It is, however, possible
that the relative increase in rate constant of the two
tetramers could be different at acid concentrations be-
tween 0.4 and 6 M, which may explain the discrepancy.
Acknowledgements
This work was supported by the Norwegian Biopolymer
Laboratory (NOBIPOL) and the Norwegian Research
Council, Grant 143184/140 (‘Calanus’). We thank Olav
Smidsrød for valuable discussions.
References
1. BeMiller, J. N. Adv. Carbohydr. Chem. 1967, 22, 25–108.
2. Varum, K. M.; Ottøy, M. H.; Smidsrød, O. Carbohydr.
Polym. 2001, 46, 89–98.
3. Piszkiewicz, D.; Bruice, T. J. Am. Chem. Soc. 1968, 90,
5844–5848.
4. Moggrigde, R. C. G.; Neuberger, A. J. Am. Chem. Soc.
1938, 745, 265–269.
5. Rupley, J. A. Biochim. Biophys. Acta 1964, 83, 245–251.
˚
˚
6. Einbu, A.; Varum, K. M. Biomacromol 2007, 8, 309–314.
7. Kerr, R. W. In Chemistry and Industry of Starch;
Academic Press: New York, 1950, Chapter 14.
8. BeMiller, J. N. In Starch; Academic Press: New York,
1965; pp 495–499.
9. Purves, C. B. In Cellulose and Cellulose Derivatives;
Interscience: New York, 1954.
10. Jones, R. W.; Dimler, R. J.; Rist, C. E. J. Am. Chem. Soc.
1955, 77, 1659–1663.
11. Weintraub, M. S.; French, D. Carbohydr. Res. 1970, 15,
241–250.
5. Conclusions
Reaction rate constants for the hydrolysis of the glyco-
sidic linkages in GlcNAc4 was found to be 50 times
greater as compared to the glycosidic linkages in GlcN4.
The two tetramers behave similarly in that one of the
end glycosidic linkages is hydrolysed 2–2.5 times faster
than the other two glycosidic linkages, probably the
nonreducing end residue. The model for the depolymer-
isation of the tetramers described herein can be used to
model selectivity in other degradation mechanisms with
tetramers as substrates, such as chitinases, chitosanases,
amylases and cellulases.
12. Weintraub, M. S.; French, D. Carbohydr. Res. 1970, 15,
251–262.
13. Feather, M. S.; Harris, J. F. J. Am. Chem. Soc. 1967, 89,
5661–5664.
˚
14. Sørbotten, A.; Horn, S. J.; Eijsink, V. G. H.; Varum, K.
M. FEBS J. 2005, 272, 538–549.