Chitosan Depolymerization: A Simple and Promising Chemical Process for the Preparation of Bioactive Varying Molecular Weight Derivatives

In order to improve functional and biological properties of shrimp chitosan and to facilitate its utilization, chitosan derivatives were prepared by chemical depolymerization using hydrochloric acid (6 M) during 2 h. The obtained chitosan depolymerization products (CDP) were characterized in terms of molecular weight (Mw), acetylation (AD) and polymerization (DP) degrees, solubility, viscosity and crystallinity, and analyzed by FTIR spectroscopy. High Mw-CDP varying from 396 to 94 kDa and low Mw-CDP (< 4.4 kDa), with a DP up to 6 determined by MALDI-TOF, were withdrawn at different hydrolysis time and then desalted on chromatography Sephadex G-25 column. It is clearly demonstrated that the viscosity, the AD and the crystallinity decreased upon depolymerization, especially in low Mw-CDP. While, solubility in water and acetic acid were highly improved. FTIR analysis showed similar spectra of chitosan and CDP. The antibacterial, antifungal and antioxidant properties of CDP were investigated and demonstrated that they were related to its Mw. Indeed, as compared to chitosan, high Mw derivatives, especially C120, possess higher antibacterial and antifungal potentials. While, low Mw-CDP, especially H120, exhibited the highest antioxidant properties. Interestingly, the used chemical depolymerization process seems to be an ecient, simple and easy method to produce bioactive chitosan derivatives with attractive characteristics to be applied in an industrial scale, especially as functional-food components. derivatives (CDP) with varying characteristics were prepared by chemical hydrolysis of shrimp chitosan using HCl (6 M). The antimicrobial effect of the obtained CDP, compared to chitosan, was investigated against seven pathogenic bacteria and two fungi, which were mostly putrefactive microorganisms in food and aquatic preservation. Moreover, their antioxidant potential through different mechanisms was also studied.

In order to remove salts, the insoluble parts, considered as low Mw-CDP, were desalted on a size exclusion chromatography Sephadex G-25 column (100 x 13 mm). The column was eluted with ultrapure water with a ow of 0.5 ml/min. Elution was monitored spectrophotometrically by measuring the absorbance under UV light (OD 260 nm) and the reducing sugar content (SR mg/g chitosane), by the modi ed dinitrosalycilic acid (DNS) method using glucosamine as standard [24], in each collected tube (5 ml). The elution pro les of OD 260 nm and SR were graphed as function of the volume (each 5 ml). After G25, the collected fractions obtained after 30, 60, 90 and 120 min of hydrolysis were freeze-dried and referred as H30, H60, H90 and H120, respectively.

Molecular weight determination by steric exclusion chromatography (SEC-HPLC)
SEC-HPLC was performed with Waters 625 LC System Pump using an Ultrahydrogel column (I.D = 7.8 mm, l = 300 mm) thermostated at 35°C and connected to a Waters 2414 differential refractometer. Samples were dissolved in 0.15 M ammonium acetate/0.2 M acetic acid buffer (pH 4.5), at a concentration of 0.2% (w/v), ltered through a 0.45 µm pore size membrane (Millipore Corporation, Beverly, MA, USA) before injection of aliquots of 20 µl. The ow rate was 0.6 ml/min. The molecular weights (Mw) of the different samples were obtained from the SEC pro les by extrapolation in a calibration curve using different known Mw dextrans as standards.

Mass spectrometric analysis (MALDI-TOF MS) of low Mw-CDP
MALDI-TOF MS was performed using an Auto ex Speed MALDI-TOF mass spectrometer (Bruker Daltonics, Germany) equipped with a Smart Beam TM ND YAG-laser (355 nm) in the positive ion mode. For ionization, 2,5-dihydroxybenzoic acid was used as matrix. All spectra were obtained in re ection mode with an acceleration voltage of 25 kV, a re ector voltage of 26 kV and pulsed ion extraction of 40 ns in positive ion mode. The acquisition range was m/z 50-4000.
The data were obtained by taking the average value of 500 laser shots, with the lowest laser energy necessary to obtain su cient signal-to-noise ratios [2].

Acetylation degree measurement
The potentiometric determination of the acetylation degree (AD) was carried out following the rst derivative UV spectrophotometric method given by Muzzarelli et al. [27]. UV measurements were carried out in a spectrophotometer in the wavelength range 190-240 nm. A calibration curve of N-acetylglucosamine (10-40 ppm) and samples, at a concentration of 0.1% (w/v), were prepared in 0.01 M acetic acid.

Functional properties evaluation
The intrinsic viscosity [η] of chitosan and its derivatives was studied according to Rinaudo  Where ηsp/C is the reduced viscosity (ml/g); k H is the Huggins' coe cient and C is the sample concentration (g/ml).
The solubility in water and acetic acid (1%) was performed according to the method of Fernandez-Kim [29]. The solubility of the sample was determined and calculated using the following equation: Where W 1 and W 2 are the initial and nal tube weights containing sample, respectively, and W 0 is the initial tube mass.

X-ray diffraction measurements (XRD)
XRD patterns were performed using an X-ray diffractometer (Philips X'Pert SW) equipped with a copper anode. Samples were continuously scanned from 0 to 50° (2θ) at 45 kV and 40 mA. The crystallinity index (CrI) was calculated as described by Focher et al. [30]: Where I 110 is the crystallinity diffraction intensity at 2θ = 20° and I am is the amorphous diffraction intensity at 2θ = 16°.
2.4.6. Fourier-transform infrared spectroscopy analysis (FTIR) FTIR spectra were recorded at 25°C, in the spectral range frequencies of 500-4000 cm − 1 and at a resolution of 4 cm − 1 , on a Performer Spectra Tech spectrometer (Agilent Technologies, Carry 630 series) linked to an attenuated total re ectance accessory with a diamond crystal. Calibration was done using background spectrum recorded from the clean and empty cell. The agar well diffusion method was rstly performed, at 50 mg/ml of sample concentration, as described by Vanden Berghe & Vlietinck [31] on a Mueller-Hinton and malt-extract agar medium for antibacterial and antifungal activities, respectively. Acetic acid (0.1%) was used as negative control and gentamicin and cycloheximide were used as positive controls for antibacterial and antifungal activity, respectively. The diameter of the growth inhibition zone around each well (including well diameter of 6 mm) was measured and reported as the antimicrobial activity.
Further, the minimum inhibitory concentration (MIC) of chitosan and CDP, de ned as the lowest sample concentration that inhibited the visible growth of tested bacteria after 24 h, was determined by liquid growth inhibition assay in a sterile 96-well microplates assay system according to Farag et al. [32]. The minimum bactericide concentration (MBC), which determines the su cient sample concentration to destroy 99.99% of the microorganism population, was also studied. MIC and MBC values are expressed in % (or in g/100 ml).

Reducing power activity
The ability of BHA, chitosan and its derivatives to reduce iron (III) was evaluated according to the method described by Yildirim et al. [35]. Higher absorbance of the reaction mixture showed higher reducing power. The reducing power activity was reported as the effective concentration which allows to obtain an absorbance of 0.5.

Total antioxidant assay
The total antioxidant ability of chitosan and CDP to reduce Mo (VI) to Mo (V) and to form a phosphate/Mo (V) complex at acidic pH was tested as reported by Prieto et al. [36]. The total antioxidant activity was reported as α-tocopherol equivalents and BHA was used as a positive control.

Statistical analysis
All experiments were carried out in triplicate, and average values with standard deviation errors are reported. Mean separation and signi cance were analyzed using the SPSS software package ver. 17.0 professional edition (SPSS, Inc., Chicago, IL, USA) using ANOVA analysis with Duncan post hoc testes. Differences were considered signi cant at p < 0.05.

Preparation of CDP by acid depolymerization of chitosan
Among the different methods of chitosan depolymerization, acid hydrolysis with HCl is preferred, especially on an industrial scale [23,8]. Varying Mw-CDP were prepared from shrimp chitosan by using HCl (6 M). The hydrolysis reaction was carried out for different time intervals in order to determine the reaction time needed to obtain CDP with certain characteristics. After hydrolysis, the insoluble compounds were named C30, C60, C90 and C120, obtained at 30, 60, 90, 120 and 240 min, respectively.
While, the soluble parts were desalted on an exclusion chromatography G25. To collect the desalted fractions, the absorbance under UV light at OD 260 nm and the reducing sugar amount in each obtained fraction (of 5 ml), were measured and results are depicted in Fig. S1. The collected fractions were freeze-dried and named H30 (tubes from 20 to 110), H60 (tubes from 20 to 103), H90 (tubes from 20 to 107) and H120 (tubes from 20 to 75), obtained at 30, 60, 90, 120 and 240 min, respectively.

Physicochemical characterization of CDP
The Mw reduction of samples withdrawn during chitosan hydrolysis was followed by SEC-HPLC analysis, and the obtained average Mw values are given in Table 1. The chromatograms of SEC ( Fig. 1) showed that the peak of the elution curve of the original chitosan (Mw = 1244.70 kDa) shifted towards longer retention time as the Mw of CDP decreased evidencing depolymerization. Results revealed that insoluble parts obtained after chitosan hydrolysis and recovered at different incubation times (30, 60, 90 and 120 min), exhibited an average Mw of 396.46, 170.65, 109.63 and 94.10 kDa for C30, C60, C90 and C120, respectively. These compounds were referred as high Mw-CDP. While, SEC-HPLC analysis of the desalted and freeze-dried soluble parts showed that in all cases, the retention times were out of the dextran calibration curve showing a Mw lower than 4.4 kDa. These samples were considered as low Mw-CDP. Table 1 Average molecular weight (Mw), intrinsic viscosity, acetylation degree (AD), acetic acid (AS) and water (WS) solubilities and crystallinity index (CrI) of chitosan and its derivatives obtained by chemical hydrolysis of shrimp chitosan. ND indicates data were not determined.
Means with different letters (a-c) and within a column indicate signi cant difference (p < 0.05).
In order to evaluate the oligomer composition of low Mw-CDP, samples were analyzed by MALDI-TOF MS and identi ed as sodium adducts [M + Na] + . The assigned structure of each sample is given in Table 2. It appeared that during hydrolysis, the O-glycosidic and the N-acetyl linkages between chitosan residues are chemically-hydrolyzed, leading to a mixture of acetylated (GlcNac-oligomers) (A) and deacetylated chito-oligomers (GlcN-oligomers) (D) with DP up to 7. After 2 h of hydrolysis, the yielded products, generated by chemical hydrolysis of chitosan were a mixture of hetero-chito-oligomers with DP 2, 3, 4 and 5 4 (GlcNac) 1 (D4A1), (GlcN) 6 (D6) as the major components.  The acetylation degree (AD) of chitosan and its derivatives was further determined and results are depicted in Table 1. It appeared that there is no signi cant difference in AD between chitosan and high Mw-CDP (p > 0.05), except of C120 in which the AD slightly decreased (p < 0.05). However, all low Mw-CDP showed a signi cant decrease in the AD values as compared to chitosan and high Mw-CDP. The decrease of the AD in chitosan derivatives obtained by acid hydrolysis was previously reported [23].
Moreover, the hydrolysis of chitosan was highlighted by the signi cant decrease in the intrinsic viscosity of chitosan (7.81 ± 0.21 dl/g) as compared to CDP, which is correlated with the decrease in Mw. Results, illustrated in Table 1, showed that the decrease of the viscosity of high Mw-CDP was more pronounced during the rst 30 min (2.07 dl/g), followed by a slow decrease during the rest of the reaction (0.66 ± 0.01 dl/g for C120). While, low Mw-CDP are not viscous.
In addition, solubility in acetic acid (1%) and water of chitosan and CDP were studied. Results illustrated in Table 1 showed that, as expected, the samples solubility increases with the Mw decrease. Contrarily to chitosan, high Mw-CDP are soluble in acetic acid (1%), while, they exhibited a poor water solubility (< 22%) (Table 1). However, low Mw-CDP were totally soluble in acetic acid 1% and water (100%). The reason for the increased solubility of chitosan derivatives was the destruction of intra macromolecular and inter chain hydrogen bonds, which alters the secondary structure of chitosan, decreasing its crystallinity and unfolding its molecular chains [1].
Besides, the acid degradation of chitosan was accompanied by a decrease in the crystallinity. The x-ray diffraction patterns of chitosan and high Mw-CDP (C30 and C120) are illustrated in Fig. 2. Chitosan exhibited two crystalline peaks at 2 θ = 10 and 20°, which corresponded to its hydrated polymorphe structure, with a crystallinity index (CrI) of 74.4% (Table 1). However, after chemical hydrolysis, the crystallinity decreased in C30 and C120, showing a less intense peak at 2 θ = 20°, while the peak at 2 θ = 10° disappeared, suggesting their amorphous structure. The CrI of C30 and C120 were 45.25 and 25.76%, respectively (Table 1). Low Mw-CDP are not crystalline (Data not shown).
The physicochemical properties of chitosan and CDP (C30, C120, H30 and H120) were also studied by FTIR analysis, showing similar spectra and con rming the conservation of the basic structure of chitosan during hydrolysis. FTIR spectra (Fig. 3) showed characteristic absorption bands at 3370 and 3296 cm − 1 attributed to the stretching vibration of -OH and NH groups, respectively. The small peak at around 2875 cm − 1 was ascribed as CH 2 and CH 3 groups. The signal observed at 1650 cm − 1 was characteristic of the amide I band (C = O in the NH-COCH 3 group). At around 1562 cm − 1 , appeared the band of amine (NH).
The peak at around 1420 cm − 1 was characteristic to the C-H bending vibration of CH 2 groups. Similar FTIR spectra of chitosan and its derivatives were obtained by Affes et al. [25] and Li et al. [37].

Antibacterial potential
The antibacterial potential of chitosan and its derivatives was rstly performed using the agar well diffusion method. Results reported in Table 3 showed that chitosan and all CDP inhibited the growth of the seven studied pathogenic bacteria, with varying diameters of inhibition and with higher inhibitory potential against Gram (−) than Gram (+) strains. Moreover, the resistance of the bacterial strains tested towards chitosan and its derivatives decreased in the following order: B. cereus > M. luteus > L. monocytogenes > S. aureus > E. sp > S. enterica > P. aeruginosa. Further, high Mw-CDP, especially C120 (Mw = 94.10 kDa, AD = 5.94%), possess signi cantly higher diameters of growth inhibition comparing to chitosan and low Mw-CDP (p < 0.05). For the strains S. enterica, P. aeruginosa and L. monocytogenes, C120 showed similar antagonistic effect than that of the positive control gentamycin (p > 0.05). Subsequently, the MIC and MBC were determined to compare the antibacterial potential of these compounds. As depicted in Table 4.A and 4.B, the MIC and MBC values uctuate from 0.0012 to 0.039% and from 0.01 to 0.3125%, respectively, against Gram (−) strains and from 0.0048 to 0.3125% and 0.019 to 1.875%, respectively, toward Gram (+) strains, con rming the better inhibitory potential against the Gram (−) bacteria obtained with agar diffusion method. Similarly, among the tested samples, C120 exhibited the lowest MIC and MBC values, leading to the highest antibacterial potential, as compared to chitosan, the other high Mw-CDP (C120 > C90 > C60 > C30) and low Mw-CDP (H120 < H90 < H60 < H30). Further, the higher inhibitory potential of this sample was obtained toward P. aeruginosa, S. enterica, E. sp, followed by S. aureus. Furthermore, MBC values demonstrate a bacteriostatic effect that all the samples against the seven tested bacteria with a ration MBC/MIC ≥ 4.  [2] demonstrated that the antimicrobial activity of chitooligosaccharides (COS) is clearly closely dependent on their Mw and decreased with its declining and that this effect was more pronounced against the Gram (−) bacteria E. coli than the Gram (+) strain L. monocytogenes. Furthermore, Jeon et al. [39] illustrated that Mw of chitosan is an important factor governing the inhibition of Gram (−) and Gram (+) bacteria and that it should be higher than or around 10 kDa for effective inhibition. Uchida et al. [40] reported that chitosan hydrolysate obtained by partial hydrolysis of chitosan exhibited better antibacterial activity than highly hydrolyzed chitosan materials. Ueno et al. [41] reported that COS with an average Mw < 2.2 kDa, was not capable of suppressing the bacterial growth, but COS with Mw around 5.5 kDa suppressed the growth dose dependently.
Besides, owing to its higher Mw, chitosan possesses high viscosity and low solubility, which result in decrease of its inhibitory activity. However, chitosan derivatives with lower Mw and viscosity and higher solubility are advantageous as antimicrobial agents compared to native chitosan [15,17]. Further, the antibacterial activity of chitosan and CDP was demonstrated to be correlated to its positive charge which interact with the negatively charged residues of macromolecules at the surface of bacteria (cell wall, DNA, RNA …), thus inhibiting its growth [42,43]. The greater negative charges on the cell surface of Gram (−) strains could explain their sensitivity to CDP than Gram (+) bacteria.

Antifungal activity
The antifungal potential of chitosan and CDP was further studied using agar well diffusion method. Results depicted in Table 3 showed that the tested compounds were most effective against Fusarium solani than Rhizoctonia solani. Similarly to the antibacterial activity, the same correlation between antifungal potential and chitosan Mw was observed. Indeed, C120 (Mw = 94.10 kDa) exhibited the highest antagonistic effect towards the two fungi, compared to chitosan, as well as the other high and low Mw-CDP. In the same context, Affes et al. [15] and Rahman et al. [44] demonstrated that medium Mw-CDP were signi cantly more effective than the native chitosan and low Mw-CDP. The exact mechanism of the antifungal activity of chitosan is not well-known and various hypotheses were proposed, depending on the particular type of fungus and the molecular characteristics of chitosans (Mw, AD and DP) [1,16,40,12].

Antioxidant activity
The DPPH and ABTS radical-scavenging activities, reducing power and total antioxidant activity of CDP were studied and compared to the chitosan and BHA.
DPPH and ABTS radical scavenging activities were reported as the effective concentration needed to reduce DPPH and ABTS radicals by 50% (IC50). The reducing power was reported as the concentration needed to obtain an absorbance of 0.5 (DO 700 nm = 0.5) and the total antioxidant potential was measured at a concentration of 5 mg/ml. Table 5 showed that, the antioxidant potential of CDP was signi cantly higher than that of BHA but lower than that of chitosan. Moreover, the Mw reduction of CDP resulted in increased antioxidant activity. Indeed, H120 exhibited the highest antioxidant potential (0.60, 0.20, 0.87 and 240.24 ± 1.75 for DPPH and ABTS radical-scavenging activities, reducing power and total antioxidant activity, respectively) than the other varying Mw-CDP (p < 0.05). Values are means ± standard deviation (n = 3). Means with different letters (A-I) and within a column indicate signi cant difference (p < 0.05).

Results illustrated in
Similar results, in vitro, were reported by Zhou et al. [19], who stated that chitosan oligomers obtained by chitosan hydrolysis using ChiEn3 chitinase exhibited better DPPH scavenging activity than the undigested chitosan. Further, Laokuldilok et al. [20] observed that COS with Mw of 5.1 kDa possess better DPPH radical activity than chitosan and COS with 14.3 and 4.1 kDa. As well, the increase of the antioxidant activity towards ABTS of low Mw-CDP, in comparison with chitosan and high Mw-CDP, was in accordance with those reported by Affes et al. [15] and Chang et al. [45]. Furthermore, Huang et al.
[46] and Laokuldilok et al. [20] demonstrated that lower Mw chitosan derivatives showed greater reducing power capacity.
The lower antioxidant potential of high Mw-CDP was mainly due to their higher inter and intramolecular bonding and compact structure which restrict the chance of exposure of their amine groups [1].

Conclusion
Current attention on chitosan focuses to its derivatives because of their advantages, including its reduced Mw and water-solubility. In this work, we showed that the chitosan chemical depolymerization, during 2 h using HCl, is a simple and practical method to produce high and low Mw-CDP at different hydrolysis times. The obtained chitosan derivatives, with reduced Mw, viscosity and crystallinity and higher solubility have been demonstrated, as compared to the native chitosan, to possess better antibacterial, antifungal and antioxidant effects, which were found to be in uenced by their Mw, among other factors. Interestingly, such depolymerization process may be considered as a promising method for the preparation of effective varying Mw chitosan derivatives to be used as functional food additives and natural food preservatives for food-processing applications. Figure 1 SEC-HPLC chromatograms of chitosan and its derivatives (C30, C60, C90 and C120) obtained by chitosan depolymerization. Figure 2 X-ray diffractograms of chitosan and its derivatives (C30 and C120).