Chitin and chitosan preparation from shrimp shells using optimized enzymatic deproteinization

Article abstract of DOI:10.1016/j.procbio.2012.07.017

Different crude microbial proteases were applied for chitin extraction from shrimp shells. A Box-Behnken design with three variables and three levels was applied in order to approach the prediction of optimal enzyme/substrate ratio, temperature and incubation time on the deproteinization degree with Bacillus mojavensis A21 crude protease. These optimal conditions were: an enzyme/substrate ratio of 7.75 U/mg, a temperature of 60 °C and an incubation time of 6 h allowing to predict 94 ± 4% deproteinization. Experimentally, in these optimized conditions, a deproteinization degree of 88 ± 5% was obtained in good agreement with the prediction and larger than values generally given in literature. The deproteinized shells were then demineralized to obtain chitin which was converted to chitosan by deacetylation and its antibacterial activity against different bacteria was investigated. Results showed that chitosan dissolved at 50 mg/ml markedly inhibited the growth of most Gram-negative and Gram-positive bacteria tested.

Full text of DOI:10.1016/j.procbio.2012.07.017

Process Biochemistry 47 (2012) 2032–2039
Contents lists available at SciVerse ScienceDirect
Process Biochemistry
journal homepage: www.elsevier.com/locate/procbio
Chitin and chitosan preparation from shrimp shells using optimized enzymatic
deproteinization
Islem Younesa , Olfa Ghorbel-Bellaaj , Rim Nasri , Moncef Chaabouni , Marguerite Rinaudoc,d,
a,∗
a
a
b
Moncef Nasri
a
Laboratory of Enzyme Engineering and Microbiology, National School of Engineering, P.O. Box 1173-3038 Sfax, Tunisia
Laboratory of Industrial Chemistry, National School of Engineering, BP W 3038 Sfax, Tunisia
European Synchrotron Radiation Facility (ESRF), BP 220, 38043 Grenoble Cedex 9, France
Research Centre of Natural Macromolecules (CERMAV-CNRS) affiliated with Joseph Fourier University, BP53, 38041 Grenoble Cedex 9, France
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
Different crude microbial proteases were applied for chitin extraction from shrimp shells. A Box–Behnken
design with three variables and three levels was applied in order to approach the prediction of optimal
enzyme/substrate ratio, temperature and incubation time on the deproteinization degree with Bacillus
mojavensis A21 crude protease. These optimal conditions were: an enzyme/substrate ratio of 7.75 U/mg,
Received 21 January 2012
Received in revised form 11 May 2012
Accepted 14 July 2012
Available online 1 September 2012
◦
a temperature of 60 C and an incubation time of 6 h allowing to predict 94 ± 4% deproteinization. Exper-
imentally, in these optimized conditions, a deproteinization degree of 88 ± 5% was obtained in good
agreement with the prediction and larger than values generally given in literature. The deproteinized
shells were then demineralized to obtain chitin which was converted to chitosan by deacetylation and
its antibacterial activity against different bacteria was investigated. Results showed that chitosan dis-
solved at 50 mg/ml markedly inhibited the growth of most Gram-negative and Gram-positive bacteria
tested.
Keywords:
Shrimp shells
Chitin
Chitosan
Enzymatic deproteinization
Bacillus mojavensis A21
Response surface methodology
© 2012 Elsevier Ltd. All rights reserved.
1
. Introduction
Chitin, the second most abundant biopolymer next to cellu-
level for specific applications. Conventionally, preparation of
chitin from such shellfish wastes involves deproteinization and
demineralization with strong bases and acids. However, the use
of these chemicals may cause a partial deacetylation of the chitin
and hydrolysis of the polymer, resulting in final inconsistent phys-
iological properties [5]. The chemical treatments also create waste
disposal problems, because neutralization and detoxification of the
discharged wastewater are necessary. Furthermore, the interest of
the protein hydrolysate is reduced due to the presence of sodium
hydroxide [6]. To overcome the defects of chemical treatments,
some efforts have been directed toward its substitution by more
eco-friendly processes such as bacterial fermentation and treat-
ment by proteolytic enzymes which have been applied for the
deproteinization of crustacean wastes [7,8].
lose, and its derivatives like chitosan are widely recognized to
have immense applications in many fields [1]. They are widely
used in the food industry, medicinal fields, chemical industries,
textiles, wastewater treatment plants, etc. [2]. The main sources
of raw material for the production of chitin are cuticles of various
crustaceans, principally crabs and shrimps. However, crustacean
shells consist of compact matrices of chitin fibers interlaced with
proteins. These matrices are reinforced through the deposition
of mineral salts, mainly those of calcium [3]. They have to be
quantitatively removed to achieve good accessibility and the
high purity necessary for biological applications. Although many
methods can be found in the literature for the removal of proteins
and minerals, effects on the molecular weight and acetylation
degree cannot be avoided with any of these extraction processes
The aim of this work is to investigate the influence of several
operating parameters such as, enzyme/substrate ratio, temperature
andincubationtimeonthedeproteinizationdegreeof shrimpshells
by non-commercial Bacillus mojavensis A21 crude enzyme.
[
4]. Therefore, a great interest still exists for the optimization of
the extraction to minimize the degradation of chitin, while, at the
same time, bringing the impurity levels down to a satisfactory
Response surface methodology (RSM) is useful for designing
experiments, building models and analysing the effects of sev-
eral independent variables [9,10]. The main advantage of RSM is
the reduced number of experimental trials needed to evaluate
the effect of multiple factors on the response. In order to deter-
mine a suitable polynomial equation that describes the response
∗ Corresponding author. Tel.: +216 74 274 088; fax: +216 74 275 595.
E-mail address: issslem@hotmail.com (I. Younes).
1
359-5113/$ – see front matter © 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.procbio.2012.07.017

I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
2033
Table 1
surface, RSM can be employed to optimize the process for gather-
ing research results better than classical one-variable-at-a-time or
full-factorial experimentation.
Design of experiment-levels of various process parameters of the Box–Behnken
design.
Parameter
Level
In this work, a Box–Behnken design [11] was employed to
establish the relationship between the reaction variables and the
deproteinization degree. Furthermore, the ridge analysis has been
employed to optimize the experimental conditions permitting the
higher deproteinization degree. The purity level of chitin was fol-
lowed by the evaluation of the mineral and protein contents. Chitin
was then converted to chitosan by chemical deacetylation. The
antibacterial activity of the acid-soluble chitosan of shrimp waste
was investigated.
−
1.0
0.0
1.0
X1: enzyme/substrate ratio (U/mg)
X2: temperature ( C)
X3: incubation time (h)
0
40
1
5
50
3.5
10
60
6
◦
2
.5. Experimental design and statistical analysis
In order to describe the nature of the response surface in the experimental
region, a Box–Behnken design was applied. As presented in Table 1, the experi-
mental design involved three parameters (U1: enzyme/substrate, U2: temperature
and U3: incubation time), each at three levels for low, middle and high concentra-
tions. Table 2 represents the design matrix of a 17 trials experiment. For predicting
the optimal point, a second order polynomial function was fitted to correlate the
relationship between independent variables and response. For the three factors this
equation is
2
. Materials and methods
2.1. Raw material
The shrimp (Metapenaeus monoceros) shells were obtained in fresh condition
from a shrimp processing plant located in Sfax, Tunisia. Shell waste were washed
thoroughly with tap water, mixed with distilled water at a ratio of 1:2 (w/v) and
then cooked for 20 min at 90 C. The cooked sample was drained and homogenized
in a Moulinex blender for about 2 min then used for moisture determination and
kept at −20 C until further use.
◦
yˆ = b0 + b1 X1 + b2 X2 + b3 X3 + b12 X1 X2 + b13 X1 X3
®
◦
2
2
2
+ b23 X2 X3 + b11 X1 + b22 X2 + b33 X3
Commercial chitosan [39280-86-9] was provided by MP Biomedicals LLC France;
13
where yˆ is the predicted response, b0 model constant; X1, X2 and X3 are independent
variables; b1, b2 and b3 are linear coefficients; b12, b13 and b23 are cross product
coefficients and b11, b22 and b33 are the quadratic coefficients.
its degree of acetylation, determined by C NMR, was 0.22.
.2. Chemical analysis of shrimp waste homogenate
The moisture and ash content were determined at 105 C and 550 C, respec-
2
Xj: coded variables related to the natural variables Uj by the following equation:
◦
◦
tively, according to the AOAC [12] standard methods 930.15 and 942.05. Total
nitrogen content of shrimp waste was determined by using the Kjeldahl method.
Crude protein was estimated by multiplying total nitrogen content by the factor
of 6.25. Lipids were determined gravimetrically after soxhlet extraction of dried
samples with hexane.
Xj = (Uj − U0j)/Step of variation
where: U0j = (Uj,high – Uj,low)/2
Step of variation of j = (Uj,high + Uj,low)/2
Uj,high and Uj,low: two extreme levels (high and low) given for each natural vari-
able Uj.
2.3. Microbial strains and enzymes preparation
The coded variables Xj are equal to −1 and +1 when the levels of natural variable
Uj are Uj,low and Uj,high, respectively.
B. mojavensis A21 and Bacillus subtilis A26 were isolated from marine water in
The model coefficients were estimated by a least squares fitting of the model
to the experimental results obtained in the design points (runs no. 1–12). The five
replicates at the center point were carried out in order to estimate the pure error
variance.
The software NEMROD W [22] was used for experimental design data analysis
and quadratic model exploitation. The optimal conditions for deproteinization were
obtained by solving the regression equation and also by analyzing the isoresponse
and response surface contour plots using the same software.
Sfax city by Haddar et al. [13] and Agrebi et al. [14], respectively. Bacillus licheniformis
NH1 was isolated by El Hadj Ali et al. [15] from an activated sludge reactor treat-
ing fishery wastewater. Bacillus licheniformis MP1 [16] was isolated from polluted
seawater from Sfax port. Vibrio metschnikovii J1 [17] was isolated from an alkaline
wastewater of the soap industry. Aspergillus clavatus ES1 [18] was isolated from
wastewater. All strains were identified on the basis of the 16S rRNA gene sequenc-
ing and biochemical properties. The medium used for the isolation of A21, A26,
NH1, MP1 and J1 strains was Luria–Bertani broth medium [19] composed of (g/l):
peptone, 10; yeast extract, 5; NaCl, 5 (pH 7.0). The medium used for the isolation of
ES1 strain was consisted of (g/l): peptone, 5.0; yeast extract, 3.0; skimmed milk 25%
2
.6. Chemical demineralization
(
v/v) and bacteriological agar, 12.0 (pH 9.0). Production of proteases was carried out
◦
Demineralization was carried out in a dilute HCl solution. Solid fractions
in optimized medium of each microbial strain. Media were autoclaved at 120 C for
2
each microbial strain, in 250 ml Erlenmeyer flasks with a working volume of 25 ml.
The cultures were centrifuged 5 min at 10,000 rpm, and the cell-free supernatants
were recovered and concentrated by the addition of solid ammonium sulfate to 80%
saturation.
obtained after hydrolysis by A21 crude protease were treated with 1.5 M HCl in 1:10
0 min. Cultivations were performed on a rotatory shaker in optimal conditions for
◦
(
w/v) ratio for 6 h at 50 C under constant stirring (150 rpm). The chitin product was
filtered through four layers of gauze with the aid of a vacuum pump and washed
◦
to neutrality with deionized water and then dried for 1 h at 60 C. Demineralization
was expressed as percentage and computed by the following equation [21]:
Protease activity was measured by the method described by Kembhavi et al. [20]
using casein as a substrate.
DDM = [
(MO × O) − (MR × R)] × 100
(2)
%
MO × O
2
.4. Deproteinization of shrimp waste by proteases
where MO and MR are ash contents (%) before and after demineralization; while, O
and R represent the mass (g) of deproteinized shell and demineralized residue in
dry weight basis, respectively.
Microbial crude enzyme preparations were tested for their deproteinization effi-
ciency. Two commercial enzymes, bromelain (Smart city) and alcalase (Novozyme)
were chosen as control for deproteinization experiments. Deproteinization tests
were carried out in a thermostated stirred Pyrex reactor (300 ml). Shrimp waste
homogenate (15 g) were mixed with 45 ml distilled water. The pH and temperature
of the mixture were adjusted to the optimum conditions for each enzyme: pH 10.0,
2.7. Deacetylation of chitin
◦
The purified chitin was treated with 12.5 M NaOH in 1:10 (w/v) ratio at 140
C
for 4 h until it was deacetylated to chitosan. After filtration, the residue was washed
with deionized water, and the crude chitosan was obtained by drying in a dry heat
◦ ◦ ◦
0 C for A21, NH1 and MP1 enzymes; pH 8.0, 40 C for A26; pH 11.0, 40 C for J1, pH
5
8
◦
◦
◦
.5, 40 C for ES1, pH 8.0, 50 C for alcalase and bromelain. Then, the shrimp waste
incubator at 50 C overnight.
proteins were digested with crude enzymes. The reaction was then stopped by heat-
ing the solution at 90 C during 20 min to inactivate enzymes. The solid phase was
◦
2.8. 13C CP/MAS-NMR spectroscopic analysis
washed and then pressed manually through four layers of gauze. Deproteinization
was expressed as percentage and computed by the following equation [21]:
Chitosan structural analysis was carried out by 13C NMR (nuclear magnetic res-
onance) with CP/MAS technique (cross-polarization, magic-angle-spinning) using
DDP = [
(PO × O) − (PR × R)] × 100
(1)
%
a BRUKER-ASX300 instrument. NMR spectra were recorded at a 13C frequency of
PO × O
7
5.5 MHz (field of 7.04 T). CP/MAS sequence was used with the following parame-
where PO and PR are the protein concentrations (%) before and after hydrolysis;
while, O and R represent the mass (g) of original sample and hydrolyzed residue in
dry weight basis, respectively.
ters: the 13C spin lattice relaxation time was 5 s; powdered samples were placed in
an alumina rotor used for the double airbearing-type MAS system and spun as fast
as 8 kHz; contact time was 8 ms.

2
034
I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
Table 2
The actual design of experiments and response of deproteinization.
Design pointa
Enzyme/substrate ratio (X1)
Temperature (X2)
Incubation time (X3)
Deproteinization (%)
Experimentalb
Predicted
1
2
3
4
5
6
7
8
9
0
10
0
10
0
10
0
10
5
5
5
5
5
40
40
60
60
50
50
50
50
40
60
40
60
50
50
50
50
50
3.5
3.5
3.5
3.5
1
1
6
6
1
1
6
6
3.5
3.5
3.5
3.5
3.5
26
54
38
79
30
75
36
82
68
75
76
86
69
71
68
76
70
26
59
33
79
32
72
39
80
66
76
73
87
71
71
71
71
71
10
11
12
13
14
15
16
17
5
5
5
5
Bold values represent the five replicates at the center point used to estimate the pure error.
a
Experiments were conducted in a random order.
The values given in the table are the average of three independent experiments.
b
The degree of acetylation (DA) of the samples was determined by dividing the
intensity of the resonance of the methyl group carbon by the average intensity of
the resonances of the glycosyl ring carbon atoms. The DA was calculated using the
in Fig. 1, high deproteinization degrees were obtained with the
crude enzymes of A21, A26, J1 and MP1 (at about 76 ± 4%), while the
two others NH1 and ES1 were exerting significantly lower values
65 ± 3% and 59 ± 3% respectively). B. mojavensis A21 strain, rarely
described in the literature, which produces at least six different
proteases, was retained. Otherwise, commercial enzymes, brome-
lain and alcalase, were used as control group. These enzymes were
chosen as control for deproteinization experiments. Deproteiniza-
tion degrees obtained with bromelain and alcalase were lower than
those obtained with crude enzymes; it reached 67 ± 3% and 54 ± 3%
for bromelain and alcalase, respectively.
following relationship [23]:
(
I[CH3]
%
DA =
× 100
(3)
(
I[C1] + I[C2] + I[C3] + I[C4] + I[C5] + I[C6])/6
I is the intensity of the particular resonance peak.
2.9. Antimicrobial activity of chitosan
The microorganisms used for antimicrobial activity were Micrococcus luteus
ATCC 4698), Escherichia coli (ATCC 25922), Pseudomonas aeruginosa (ATCC 27853),
(
Klebsiella pneumoniae (ATCC 13883), Bacillus cereus (ATCC 11778), Staphylococ-
cus aureus (ATCC 25923) Salmonella typhi and Enterococcus faecalis (ATCC 29212).
Antibacterial activity assays were performed according to the method described by
Berghe and Vlietinck [24]. Chitosan was dissolved at 50 mg/ml in 0.1% acetic acid (pH
Secondly, the enzymatic deproteinization of shrimp shells by B.
mojavensis A21 crude enzyme was carried out under several exper-
imental conditions, including various enzyme/substrate ratios,
different temperatures, pH and incubation times. Reaction carried
4
1
.68). The inoculum suspension (200 ␮l) of the tested microorganisms, containing
6
◦
0
colony forming units (CFU/ml) of bacteria cells were spread on Muller–Hinton
out at 50 C without enzymes resulted in a deproteinization degree
agar. The inoculums were allowed to dry for 5 min. Then, bores (3-mm depth, 6-mm
diameter) were made using a sterile borer and were loaded with 50 ␮l of each sam-
ple. Well with only acetic acid (without chitosan) was used as a negative control (pH
of about 38 ± 2%. Indeed, some proteins associated to chitin by elec-
trostatic forces or hydrogen bonds, could be dissociated by thermal
treatment. However, other proteins are linked to chitin by covalent
bonds, their removal requires severe chemical or enzymatic treat-
ments [25]. The deproteinization rate with an E/S ratio of 1 was high
3
.25). Gentamycin was used as positive reference. The Petri dishes were kept, firstly,
◦
◦
for 1 h at 4 C, and then were incubated for 24 h at 37 C. Antibacterial activity was
evaluated by measuring the diameter of the growth inhibition zones in millime-
ters (including well diameter of 6 mm) for the test organisms and comparing to the
controls. The measurements of inhibition zones were carried out for three sample
replications, and values are the average of three replicates.
(
71 ± 2%), which shows the effectiveness of the enzyme preparation
of B. mojavensis A21, and further increase in enzyme concentration
increased slightly (from 71 ± 2% to 77 ± 2%) the deproteinization
rate. Beyond 20 U/mg the deproteinization rate remained constant.
B. mojavensis A21 crude alkaline protease characterization [13]
2.10. Statistical analysis
All experiments were carried out in triplicate, and average values with standard
◦
showed an optimal activity at pH 8.0–11.0 and at 60 C, using casein
deviation errors are reported. Mean separation and significance were analyzed using
the SPSS software package (SPSS, Chicago, IL). Correlation and regression analysis
was carried out using the EXCEL program.
3
. Results and discussion
3.1. Enzymatic deproteinization of shrimp waste by microbial
proteases
Firstly, we tried to point out the main parameters playing a
role on enzymatic deproteinization. In this view, we tried to select
the most effective enzymes. Microbial proteases from B. mojaven-
sis A21, B. subtilis A26, A. clavatus ES1, B. licheniformis MP1, B.
licheniformis NH1 and V. metschnikovii J1 were tested for their
deproteinization efficiency. Deproteinization tests were conducted
for 3 h under conditions of optimal enzyme activity and stability
with enzyme/substrate ratios equal to 20 U/mg of protein. As shown
Fig. 1. Deproteinization of shrimp waste (%) by protease preparations. Alcalase;
Bromelain; A21: B. mojavensis A21; J1: V. metschnikovii J1; MP1: B. licheniformis
MP1; A26: B. subtilis A26; NH1: B. licheniformis NH1 and ES1: A. clavatus ES1.

I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
2035
Table 3
Analysis of Variance (ANOVA) for the fit of experimental data to response surface model.
Source of variation
Sum of squares (SS)
Degrees of freedom (DF)
Mean square
Fexp
Significance
Regression
Residual
Lack of fit
Pur error
5706.44
153.80
115.00
38.80
9
7
3
4
634.04
21.97
383.33
9.70
28.85
0.0102***
10.9 NS
3.95
Total
5860.24
16
***: indicates significant at the level 99.9%. N.S.: indicates non-significant at the level 95.0%
as a substrate. It was extremely stable in the pH range of 7.0–11.0. In
this work, the preliminary study showed that pH has a little impact
on the deproteinization degree in the range of activity and stability
of enzymes (pH 8.0–11.0) and only temperature and incubation
time could have a considerable impact on the deproteinization
degree.
deproteinization degree is shown in Fig. 2. The right parts and the
left parts of both plots refer to the maximization and the mini-
mization of the response, respectively. The distance (r) from the
center of the design is indicated in abscissas. Fig. 2a shows that the
optimum response reached as a function of the distance r. Fig. 2b
displays the coordinates for each factor, in codified variables, of the
points of plot 2a. As it can be seen in Fig. 2a, the deproteinization
degree significantly increases from the center of the domain to the
boundary (r = 1) where it reaches 88.87%. In addition, Fig. 2b shows
that, close to the maximum, the deproteinization degree is more
With regards to these results, for the selected enzyme B.
mojavensis A21 crude alkaline protease, E/S ratio (U ), temperature
1
(
U ) and incubation time (U ) were selected as effective operating
2 3
variables on the deproteinization degree. In this view, we applied
a response surface methodology allowing to predict the optimum
conditions for enzymatic deproteinization of shrimp shells.
sensitive to the variations of the incubation time (X ) and the tem-
3
perature (X ) than that of E/S ratio (X ). To reach the maximum, all
2
1
factors must tend toward relatively high values.
3
.2. Optimization of the shrimp waste deproteinization using A21
The isoresponse curves and the response surface were drawn by
plotting the response variation against two factors, while the third
is held constant at its mean level (Fig. 3). The examination of these
figures shows that, an important effect on the deproteinization effi-
ciency was provided by the E/S ratio (X1). Indeed, Fig. 3a shows that
when the E/S ratio is low, between 0 and 5 U/mg, the temperature
has no significant effect on the deproteinization degree. However,
temperature has a very significant positive effect on the response
when the E/S ratio is beyond 5 U/mg. In this way, a deproteinization
degree of above 85% is obtained with an E/S ratio of 7.5 U/mg and
proteases
The optimization of the experimental conditions of shrimp
waste deproteinization was achieved using a Box–Behnken design
and carried out under several experimental conditions including
various E/S ratio (in the range of 0–10 U/mg), temperature (in the
◦
range of 40–60 C) and incubation time (in the range of 1–6 h). The
pH was maintained at 9.0. Table 2 shows the real experimental
conditions and the measured responses. These data are used to
establish a mathematical relation between the set of parameters
and the degree of deproteinization.
◦
a temperature of 60 C.
The same could be said for the reaction time effect, when com-
pared to that of temperature. Fig. 3b represents the isocontour
plots when the level of the temperature is fixed to its average level
3
.2.1. Model equation and validation
The coefficients of the postulated model were calculated on
◦
(50 C). It showed that yields higher than 85% could be reached
by fixing the E/S ratio and incubation time at levels higher than
the basis of the experimental responses (Table 2) by the least
square regression using the NEMROD W software. The fitted model,
expressed in coded variables, is represented by the equation:
7
.5 U/mg and 6 h, respectively.
Finally, it could be seen from Fig. 3c that, by keeping X1 at its
average level (5 U/mg), high deproteinization degrees are obtained
with high levels of both temperature and incubation time.
yˆ = 70.4 + 20 X + 6.75 X + 4 X + 3.25 X X + 0.25 X X
1
2
3
1
2
1 3
2
1
2
2
+
0.75 X X − 20.825 X − 0.325 X2 + 6.175 X3
2
3
3.2.3. Comparison between model and experimental results
It is necessary to make an analysis of the variance (ANOVA) using
the F-test to attest the good quality of the fitting [26]. Results of
the analysis of variance for the fitted model are summarized in
Table 3. They clearly indicated that the regression sum of squares is
statistically significant at the level 99.9%. Moreover, the coefficient
From this model analysis, the optimal experimental conditions
◦
were fixed at: E/S ratio 7.75 U/mg, temperature 60 C and incu-
bation time 6 h which allow to obtain 94 ± 4% of yield. In order
to confirm this prediction, an additional independent run was
conducted using these selected variable levels. Results obtained
showed that the observed deproteinization degree (88 ± 5%) was
not significantly different from the predicted value. This result con-
firmed that the empirical model derived from RSM can be used to
adequately describe the relationship between the factors and the
shrimp shells deproteinization response using the crude enzyme of
B. mojavensis A21.
The deproteinization activity of B. mojavensis A21 crude pro-
tease was better than many proteases reported in previous studies.
Deproteinization rate obtained using A21 proteases was even
better than values reached using fermentation which gives habit-
ually higher deproteinization rate. Busto and Healy [27] compared
the effects of microbial and enzymatic deproteinization. A max-
imum value of 82% was achieved with Pseudomonas maltophilia
after six days and 64% with purified microbial protease under the
same condition. Fermentation using the culture supernatant from
2
of determination, R , for the conversion yield was 0.974. This means
that 97.4% of the observed variation is attributed to the variable
effects. Thus, it is concluded that the predicted model well fitted
the experimental data.
On the other hand, the validity of the model has been established
by comparing the variance related to the lack of fit to that of pure
error which demonstrated the non-significance of the lack of fit
(
Table 3).
The valid model was used to predict response values in the stud-
ied domain and to draw isoresponse contour plots and response
surfaces.
3.2.2. Graphical interpretation of the response surface model
The representation of the optimum path computed from
the ridge analysis [26] of the response surface fitted for the

2
036
I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
Fig. 2. Ridge analysis: optimal response plot (a) and optimal coordinate plot (b).
Pseudomonas aeruginosa K-187 allows a deproteinization rate of
demineralization conditions used in this study reduce the mineral
content to permissible limits in the chitin. Indeed, the ash content
was reduced to about 1.9%. This was lower than that found by Sini
et al. [32]. This low ash content for chitin indicated the suitabil-
ity of removal of calcium carbonate and other minerals from the
raw material. There were no significant differences in the mois-
ture content and ash among the two chitins (p > 0.05). An important
difference concerns the protein content significantly higher in the
chitin isolated after enzymatic deproteinization (p < 0.05); com-
plete removal of the residual protein associated with the chitin was
not achieved even if the residual yield is lower than usually found
in literature.
◦
7
8% after a seven-day incubation at 37 C [28]. Fermentation con-
ditions by P. aeruginosa A2 were optimized by response surface
methodology and maximal deproteinization of 89% was achieved
after five days incubation [7]. Bacillus cereus SV1 proteases were
also applied for enzymatic deproteinization. Protein removal from
shrimp waste reached the same value of deproteinization degree
obtained by A21 proteases but using higher E/S ratio (20 instead
of 7.75 U/mg) [8]. The fact that deproteinization cannot reach 100%
is explained by the non-accessibility of enzymes to some proteins
protected by chitin and minerals. Indeed, the proteins in the inner
layer of shrimp shell waste are protected by the outer layer chitin,
from the attack by proteases and thus, no further proteolysis could
occur [25]. In addition, some portions of peptides are suggested to
be linked covalently to a small number of the C-2 amino groups of
chitins [25].
Although such deproteinization percentage is lower than that
obtained by chemical treatment, enzymatic deproteinization helps
to avoid many drawbacks of chemical treatment, over-hydrolysis,
breakdown of chitin, etc.
3.3. Chemical demineralization
3.5. Preparation of chitosan
In the recovery of chitin from shrimp waste, associated minerals
The major procedure for obtaining chitosan is based on the alka-
line deacetylation of chitin with strong alkaline solution. In this
study, chitosan was prepared from chitin obtained by a treatment
should be removed as a second stage. As a consequence, shrimp
waste deproteinized by enzymatic treatment was subjected to mild
acid treatment in order to remove minerals. The demineralization
◦
with 12.5 M NaOH in 1:10 (w/v) ratio at 140 C for 4 h. The effi-
◦
ciency of this treatment is evaluated by the acetylation degree of
chitosan determined by Nuclear Magnetic Resonance (NMR). Solid
state 13C RMN spectroscopy was used in order to verify the purity
of the chitosan sample by the chemical shifts and the intensities of
was completely achieved within 6 h at 50 C after treatment with
1
.5 M HCl solution at a ratio of 1:10 (w/v).
One of the factors determining the good quality of chitin is the
low mineral content [29]. Chitin obtained in this work present con-
tent of minerals as low as those reported in other works [30] at
around of 1.9%.
The treatments employed to extract chitin from the shrimp
waste allowed the recovery of 18.5 ± 2.3% of its initial dry mass
as a water insoluble white fibrous material, which indicates that a
good yield for the chitin extraction was attained and no pigments
were present in the chitin. Other studies reported similar yields for
the extraction of chitin [29,30].
13
the C absorption peaks [33].
3.6. 13C CP/MAS-NMR spectroscopic analysis
NMR is one of the most powerful tools in the study of
polysaccharide composition and sequential structure. NMR is a
non-destructive method resulting in retained structure and con-
formation of the polysaccharide, making it possible to monitor
3.4. Chitin characterization
Table 4
Properties of the chitins obtained by deproteinization with B. mojavensis proteases
The same raw material was treated by alkali (NaOH 1.25 M
after 4 h incubation at 80 C) [8]; this treatment gives chitin 2
(1) and by alkali deproteinization (2).
◦
%
Raw material
Chitin (1)
Chitin (2)
(
Table 4). In Table 4, the characteristics of the raw material, chitin
Moisture
Chitin
Ash
Protein
Lipid
Appearance
67.0 ± 1.1
33.6 ± 2.3
33.2 ± 0.2
27.1 ± 1.3
6.0 ± 0.3
–
4.6 ± 1.1
18.5 ± 2.3
1.9 ± 0.1
12.0 ± 1.6
–
3.9 ± 0.5
20.0 ± 2.0
1.4 ± 0.1
6.2 ± 1.3
–
prepared by enzymatic treatment (chitin 1) and that obtained by
alkaline treatment (chitin 2) are compared. The ground shrimp
waste before pre-treatment contained a relatively high contents
of chitin (33.5 ± 2.3%) and ash (33.2 ± 0.2%). These results are
comparable with those reported by previous studies [29,31]. The
White flakes
Yellowish flakes

I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
2037
Fig. 3. Three-dimensional response surface and contour plots for the effect on the deproteinization degree of: E/S ratio and temperature at constant time 3.5 h (a) E/S ratio
◦
and incubation time at constant temperature 50 C. (b) Temperature and incubation time at constant E/S ratio 5.00 U/mg (c).
reactions and other structural and physical properties under dif-
ferent solvent conditions.
to that of commercial chitosan. There are 8 signals for the 8 car-
bon atoms of chitosan. The C1-C6 carbons of N-acetylglucosamine
monomeric unit are observed between 50 and 110 ppm, indicating
high structural homogenity. The carbonyl group is around 173 ppm,
while the methyl group of the acetyl group produced a peak at
around 23 ppm. The less the peaks of carbonyl and acetyl groups
are, the more efficient the deacetylation reaction is. Fig. 4 shows the
1
3
Solid state C CP/MAS-NMR is known to be very sensitive to
changes in the local structure. 13C CP/MAS-NMR spectrum of the
chitosan prepared by enzymatic deproteinization and commercial
chitosan prepared by alkaline treatment, are shown in Fig. 4. NMR
analysis of the shrimp waste chitosan gave similar peak pattern

2
038
I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
Fig. 4. 13C CP/MAS-NMR solid-state spectra of chitosans. (1) Chitosan prepared from chitin obtained by enzymatic deproteinization with B. mojavensis proteases and chemical
demineralization. (2) Commercial chitosan obtained by chemical deproteinization and chemical demineralization.
Table 5
chitosan spectrum, in which the deacetylation of chitin is evident,
The diameters of inhibition zones against Gram-positive and Gram-negative
bacteria.
since there are tight peaks at 23 and 173 ppm, that correspond to
the CH and C O groups, respectively. The other peaks correspond
3
Diameterofinhibitionzones(mm)
to C1 (ı104.63), C2 (ı55.71), C3 (ı73.93), C4 (ı83.63), C5 (ı76.14)
and C6 (ı61.46). Note in the spectrum that the removal of pro-
teins were efficient during the extraction, once there are no other
peaks, suggesting a great purity of the product [34]. The presence
of some background noise could be due to the presence of a possi-
ble by-product or impurity in the sample, especially in commercial
chitosan (Fig. 4).
The degree of acetylation in both chitosans was determined
using Eq. (3). The degree of acetylation was 22% and 4% in the
commercial sample and the prepared sample, respectively. A
deacetylated chitin with a rate of 70–90% and low protein content
is considered as a good final product, the best one being that with
the higher degree of deacetylation.
Chitosana
Chitosanb
Gram −
Escherichia coli
12.4 ± 1.5
R
7 ± 1.2
R
Pseudomonas aeruginosa
Klebsiella pneumoniae
Salmonella typhi
12 ± 1.2
7 ± 0.8
10 ± 0.6
9 ± 0.4
Gram +
Staphylococcus aureus
Bacillus cereus
Enterococcus faecalis
Micrococcus luteus
8 ± 0.5
9.5 ± 0.5
R
9 ± 1.2
7 ± 0.6
9 ± 0.5
R
8.4 ± 0.8
Diameter well: 6 mm; R = resistant.
a
Chitosan prepared from chitin obtained by enzymatic deproteinization with B.
mojavensis proteases and chemical demineralization
b
Commercial chitosan obtained by chemical deproteinization and chemical dem-
ineralization.
[
36]. However, No et al. [37] showed stronger bactericidal effects
3.7. Antimicrobial activity
of chitosan toward Gram-positive than Gram-negative bacteria.
Antimicrobial activity of chitosans against several bacterial
4. Conclusion
species has been recognized and is considered as one of the most
important properties linked directly to their possible biological
applications. The antimicrobial activity of chitosan dissolved in
acetic acid 0.1% was investigated against four Gram-positive and
four Gram-negative bacteria. As shown in Table 5, both chitosans
inhibited the growth of all Gram-negativeand Gram-positivebacte-
ria tested excepted P. aeruginosa and E. faecalis. Chitosan obtained
by deproteinization with B. mojavensis proteases categorized as low
degree of acetylation showed higher inhibition activity than the
commercial one with higher degree of acetylation. These results
are in line with the works of Chen et al. [35] who reported that
antibacterial activity increase in the order of chitin deacetylation
degree. This inhibition is more important against Gram-negative
than Gram-positive bacteria tested. Gram-negative bacteria, with
lipopolysaccharide at the outer surface providing negative charges,
seemed to be very sensitive to chitosan while the sensitivity of
Gram-positive bacteria that can have variable amounts of nega-
tively charged teichoic acids at their outer surface varied greatly
A Box–Behnken design with three variables and three levels
was applied for the determination of deproteinization efficiency
of shrimp shells with enzymatic treatment by B. mojavensis A21
proteases. A21 proteases were found to remove up to 88 ± 5% of
the shell proteins in agreement with optimization using response
surface methodology. The optimal conditions for deproteinization
were: an enzyme/substrate ratio of 7.75 U/mg, a temperature of
◦
6
0 C and an incubation time of 6 h. Chitin obtained by enzymatic
deproteinization was then converted to chitosan, with a low degree
of acetylation, which was found to exhibit remarkably antibacterial
activities.
Acknowledgment
This work was supported by grants from Ministry of Higher
Education and Scientific Research, Tunisia.

I. Younes et al. / Process Biochemistry 47 (2012) 2032–2039
2039
References
[19] Miller JH. Experiments in molecular genetics. New York: Cold Spring Harbor
Laboratory Press; 1972.
[
[
[
[
[
20] Kembhavi AA, Kulkarni A, Pant A. Salt-tolerant thermostable alkaline protease
from Bacillus subtilis NCIM no. 64. Appl Biochem Biotechnol 1993;38:83–92.
21] Rao MS, Mu n˜ oz J, Stevens WF. Critical factors in chitin production by fermen-
tation of shrimp biowaste. Appl Microbiol Biotechnol 2000;54:808–13.
[
1] Sini TK, Santhosh S, Mathew PT. Study of the influence of processing parameters
on the production of carboxymethylchitin. Polym J 2005;46:3128–31.
2] Muzzarelli RAA. Advances in chitin science, vol. 1. France: Jacque Andre; 1996.
3] Saito Y, Okano T, Gaill F, Chanzy H, Putaux JL. Structural data on the intra-
crystalline swelling of beta-chitin. Int J Biol Macromol 2000;28:81–8.
4] Kurita K, Tomita K, Tada T, Ishii S, Nishimura S, Shimoda K. Reactivity charac-
teristics of a new form of chitosan. J Polym Sci Polym Chem 1993;31:485–91.
5] Brine CJ, Austin PR. Chitin variability with species and method of preparation.
Comp Biochem Physiol 1981;69B:283–6.
[
[
22] Mathieu D, Nony J, Phan-Than-Luu R. NEMROD W software LPRAI. Marseille;
2000.
[
[
[
[
23] Ottøy MH, Varum KM, Smidsrød O. Compositional heterogeneity of heteroge-
neously deacetylated chitosanes. Carbohydr Polym 1996;29:17–24.
24] Berghe DVA, Vlietinck AJ. Screening methods for antibacterial and antiviral
agents from higher plants. In: Dey PM, Harborne JB, editors. Methods in plant
biochemistry. London: Academic Press; 1991. p. 47–69.
6] Gagne N, Simpson BK. Use of pyrolytic enzymes to facilitate the recovery of
chitin from shrimp wastes. Food Biotechnol 1993;7:253–63.
7] Ghorbel-Bellaaj O, Hmidet N, Jellouli K, Younes I, Maâlej H, Hachicha R, et al.
Shrimp waste fermentation with Psceudomonas aeruginosa A2: optimization
of chitin extraction conditions through Plackett–Burman and response surface
methodology approaches. Int J Biol Macromol 2011;48:596–602.
[
25] Jeuniaux C, Compère P, Goffinet G. Structure, synthèse et dégradation
des chitinoprotéines de la cuticule des crustacés décapodes. Ital
986;53:183–96.
26] Montgomery DC. Design and analysis of experiments. 3rd ed. New York: Wiley;
991.
27] Bustos RO, Healy MG. Microbial deproteinization of waste prawn shell. In: Insti-
tution of Chemical Engineers Symposium Series. Rugby, England: Institution of
Chemical Engineers; 1994. p. 13–5.
J Zool
1
[
[
1
[
8] Manni L, Ghorbel-Bellaaj O, Jellouli K, Younes I, Nasri M. Extraction and charac-
terization of chitin, chitosan, and protein hydrolysates prepared from shrimp
waste by treatment with crude protease from Bacillus cereus SV1. Appl Biochem
Biotechnol 2010;162:345–57.
[
28] Oh YS, Shih IL, Tzeng YM, Wang SL. Protease produced by Pseudomonas
aeruginosa K-187 and its application in the deproteinization of shrimp and
crab shell wastes. Enzyme Microb Technol 2000;27:3–10.
[9] Myers RH, Montgomery DC. Response surface methodology. Process and prod-
uct optimization using designed experiments. New York: John Wiley & Sons;
1
995.
[
29] Tolaimate A, Desbrieres J, Rhazi M, Alagui A. Contribution to the preparation
of chitins and chitosans with controlled physico-chemical properties. Polymer
[
[
[
[
10] Lewis GA, Mathieu GA, Phan-Tan-Luu R. Pharmaceutical experimental design.
New York: Marcel Dekker Inc; 1999.
11] Box GEP, Behnken DW. Some new three-level designs for the study of quanti-
tative variables. Technometrics 1960;2:455–75.
12] AOAC. Official methods of analysis. 12th ed. Washington: DC: Association of
Official Analytical Chemists; 1975.
2003;44:7939–52.
[
[
30] Percot A, Viton C, Domard A. Optimization of chitin extraction from shrimp
shells. Biomacromol 2003;4:12–8.
31] Canizares E, Gonzalez José A, Hau L, Osorno H. Evaluación de dos alternativas de
tratamiento químoco para la obtención de quitina a partir de residuos sólidos
del camarón. Alimentaria 2002;331:31–4.
13] Haddar A, Bougatef A, Agrebi R, Sellami-Kamoun A, Nasri M. A novel surfactant-
stable alkaline serine-protease from a newly isolated Bacillus mojavensis A21:
purification and characterization. Process Biochem 2009;44:29–35.
14] Agrebi R, Haddar A, Hmidet N, Jellouli K, Manni L, Nasri M. BSF1 fibrinolytic
enzyme from a marine bacterium Bacillus subtilis A26: purification, biochemical
and molecular characterization. Process Biochem 2009;44:1252–9.
15] El Hadj Ali N, Agrebi R, Ghorbel-Frikha B, Sellami-Kamoun A, Kanoun S, Nasri
M. Biochemical and molecular characterization of a detergent stable alka-
line serine-protease from a newly isolated Bacillus licheniformis NH1. Enzyme
Microb Technol 2007;40:515–23.
16] Jellouli K, Ghorbel-Bellaaj O, Ben Ayed H, Manni L, Agrebi R, Nasri M. Alkaline-
protease from Bacillus licheniformis MP1: purification, characterization and
potential application as a detergent additive and for shrimp waste depro-
teinization. Process Biochem 2011;46:1248–56.
17] Jellouli K, Bougatef A, Manni L, Agrebi R, Siala R, Younes I, et al. Molecular and
biochemical characterization of an extracellular serine-protease from Vibrio
metschnikovii J1. J Ind Microbiol Biotechnol 2009;36:939–48.
[
[
[
[
[
[
32] Sini TK, Santhosh S, Mathew PT. Study on the production of chitin and chi-
tosan from shrimp shell by using Bacillus subtilis fermentation. Carbohydr Res
[
[
2007;342:2423–9.
33] Raymond L, Morin FG, Marchessault RH. Degree of deacetylation of chi-
tosan using conductometric titration and solid-state NMR. Carbohydr Res
1993;246:331–6.
34] Pauline AT, Simionato JI, Garcia JC, Nozaki J. Characterization of chi-
tosan and chitin produced from silkworm chrysalides. Carbohydr Polym
2006;64:98–103.
[
35] Chen YM, Chung YC, Wang LW, Chen KT, Li SY. Antibacterial properties of chi-
tosan in waterborne pathogen. J Environ Sci Health A Tox Hazard Subst Environ
Eng 2002;37:1379–90.
36] Devlieghere F, Vermeulen A, Debevere J. Chitosan: antimicrobial activity, inter-
actions with food components and applicability as a coating on fruit and
vegetables. Food Microbiol 2004;21:703–14.
[
37] No HK, Park NY, Lee SH, Meyers SP. Antibacterial activity of chitosans and
chitosan oligomers with different molecular weights. Int J Food Microbiol
[
18] Hajji M, Kanoun S, Nasri M, Gharsallah N. Purification and characterization of
an alkaline serine-protease produced by a newly isolated Aspergillus clavatus
ES1. Process Biochem 2007;42:791–7.
2002;74:65–72.