Bioproduction of Chitooligosaccharides: Present and Perspectives

Chitin and chitosan oligosaccharides (COS) have been traditionally obtained by chemical digestion with strong acids. In light of the difficulties associated with these traditional production processes, environmentally compatible and reproducible production alternatives are desirable. Unlike chemical digestion, biodegradation of chitin and chitosan by enzymes or microorganisms does not require the use of toxic chemicals or excessive amounts of wastewater. Enzyme preparations with chitinase, chitosanase, and lysozymeare primarily used to hydrolyze chitin and chitosan. Commercial preparations of cellulase, protease, lipase, and pepsin provide another opportunity for oligosaccharide production. In addition to their hydrolytic activities, the transglycosylation activity of chitinolytic enzymes might be exploited for the synthesis of desired chitin oligomers and their derivatives. Chitin deacetylase is also potentially useful for the preparation of oligosaccharides. Recently, direct production of oligosaccharides from chitin and crab shells by a combination of mechanochemical grinding and enzymatic hydrolysis has been reported. Together with these, other emerging technologies such as direct degradation of chitin from crustacean shells and microbial cell walls, enzymatic synthesis of COS from small building blocks, and protein engineering technology for chitin-related enzymes have been discussed as the most significant challenge for industrial application.

COS in terms of length (degree of polymerization, DP), degree of de-N-acetylation (DD), and sequence (pattern of acetylation, PA) by enzymatic conversion processes is not straight forward [24]. The expensive cost of chitinases and chitosanases limits their wide application on an industrial scale, even using immobilized enzymes [22,25,26].
We are primarily concerned with the enzymatic hydrolysis of polysaccharides obtained from crustacean shells for the bioproduction of COS. In this review, we introduce the preparation and bioproduction of COS with enzymes from chitinase-and chitosanase-producing bacteria and fungi. Limitations and challenges in the bioproduction of COS are also discussed.

Bioproduction of Chitin Oligosaccharides and Its Monomer
Chitin oligosaccharides are produced from chitin by using chitinolytic enzymes. These enzymes hydrolyze the glycoside bonds between the sugars and are thus glycoside hydrolases (GH). GHs are classified in the Carbohydrate-Active enZYmes database (CAZy) [27][28][29][30]. The CAZy classification is purely based on the amino acid sequence similarity, which gives very useful information since sequence and structure, and hence mechanism, are related. In CAZy system, enzyme properties such as substrate and product activities, exo-versus endo-binding, processivity, and the presence of additional modules are not taken into account [27][28][29][30]. Chitinases occur in families GH18 and GH19 [31,32].
Chitinases have the unique ability to hydrolyze GlcNAc-GlcNAc bonds and this property discriminates these enzymes from chitosanases.
In COS production, it is practical to use crude enzyme preparations, such as microbial culture supernatants and their solid powders. Preparations of cellulase, protease, lipase, pepsin, chitinase, chitosanase, and lysozyme are mainly applied as commercial enzyme sources for the hydrolytic cleavage of chitin and chitosan (Table 1).
Pretreatment is necessary for enzymatic hydrolysis, because the strong crystallinity and insolubility of chitin in an aquatic environment. For example, chitin is normally treated with strong acids, such as hydrochloric acid (for colloidal chitin) [34] or phosphoric acid (for swollen chitin) [35] to break down crystal structure and increase the accessibility of the substrate to the enzyme. In contrast to chitin, chitosan is soluble in dilute mineral acids, thus forming salts with acids.
Enzymatic or chemical hydrolysis of chitin and chitosan results in a mixture of oligomers, not any specific oligomer, even with prolonged reaction time. Several techniques for separation and purification of COS have been reported, like gel filtration [36], ultrafiltration [37], and ion exchange [38] and metal affinity [39] chromatography. Preparative separation of COS is most commonly based on size, through size exclusion chromatography SEC). A SEC system using SuperdexTM 30 (GE Healthcare) columns, coupled in series, allowed separation of COS with similar DP values ranging from DP 2 to DP 20, independently of DD and PA [36]. Separation of COS can be achieved using cation-exchange chromatography, because protonated amino groups on the deacetylated sugars interact with the ion-exchange resin. With this method, COS of identical DP was separated based on the number of deacetylated units [38].
Non-chitinase enzymes can be applied to the hydrolysis of chitin. GlcNAc was produced from α-chitin (from crab shell) and β-chitin (from squid pen) using crude cellulase preparations from Trichoderma viride (T) and Acremonium cellulolyticus (A) [41]. The yield of GlcNAc was enhanced by mixing cellulases T and A. Crystalline chitin (from bee) and α-chitin (from crab shell) were hydrolyzed with the enzymatic preparation Celloviridin G20x from strain Trichoderma reesei, which includes cellulases and β-glucanases [42]. After 10 days of incubation in these conditions, the yield of GlcNAc reached 86%.
The mixed enzymes (cellulase: lipase = 9:1) from Aspergillus niger, was used in various enzyme concentrations, to investigate the hydrolytic behavior on β-chitin [43]. Increasing substrate concentration while keeping the enzyme concentration constant resulted in higher GlcNAc yield. After 4 days of incubation, the yield of GlcNAc reached 61% from β-chitin (10 mg/mL).
Flake type of chitin together with swollen chitin, colloidal chitin, and powder chitin is also the substrate for the oligosaccharides production. GlcNAc was produced from crab shell α-chitin flake using crude enzyme extract from Aeromonas hydrophila H-2330, with 66%-77% yield after 10 days incubation at 17 °C [44].
Exo-type chitinases can use chitin oligosaccharides as a substrate. (GlcNAc)2 was gradually and completely degraded to GlcNAc by exo-type chitinase (ChiA71) from Bacillus thuringiensis subsp. pakistani with time [45] and by N-acetyl-β-hexosaminidase (StmHex) from Stenotrophomonas maltophilia [46]. GlcNAc was produced from (GlcNAc)4 by enzyme purified from Aeromonas hydrophila SUWA-9 [47].This kind of trial is not intended for the production of monomers, but for the elucidation of enzyme properties, as oligomers are much more expensive and valuable in application.
There are typical reports on the selective preparation of GlcNAc and (GlcNAc)2 [48]. It was found that α-chitin was effectively hydrolyzed to (GlcNAc) by Aeromonas sp. GJ-18 crude enzyme preparation below 45 °C. The enzyme preparation from Aeromonas sp. GJ-18 contained GlcNAcase and N,N′-diacetylchitobiohydrolase [49]. GlcNAcase was inactive above 50 °C, but N,N′-diacetylchitobiohydrolase was stable at this temperature. Therefore, GlcNAc and (GlcNAc)2 were selectively produced from α-chitin at two temperatures. At 45 °C, GlcNAc was the major hydrolytic product with a yield of 74% in 5 days' incubation, while at 55 °C (GlcNAc)2 was the major product with a yield of 35% after 5 days' incubation.

N,N′-Diacetylchitobiose and Chitin Oligosaccharides
As mentioned above, (GlcNAc)2 can be obtained as a major hydrolytic product from enzymes by controlling the ratio of GlcNAcase to N,N′-diacetylchitobiohydrolase activities in the crude enzyme of Aeromonas sp. GJ-18 [50]. After 7 days of incubation, 78.9% and 56.6% of (GlcNAc)2 yields were obtained from swollen α-chitin and powdered β-chitin, respectively, using enzyme preparations that had been pretreated at 50 °C so as to inactivate GlcNAcase.

Bioproduction of Chitosan Oligosaccharides and Its Monomer
The most important tool in the biodegradation of chitosan to its oligosaccharides is chitosanolytic enzymes. The chitosanases have been prepared from various bacteria and fungi, as shown in Table 2. Chitosanases in glycoside hydrolase (GH) families 5, 7, 8, 46, 75, and 80 [27][28][29][30], have been classified into subclasses I, II, and III based on their substrate specificities toward chitosan [19,[60][61][62][63]. Class I chitosanases can hydrolyze both GlcNAc-GlcN and GlcN-GlcN linkages, and class II chitosanases can split only GlcN-GlcN linkages, whereas class III chitosanases can degrade both GlcN-GlcNAc and GlcN-GlcN linkages. Chitosanases can also be classified into two major categories (endochitosanases and exochitosanases), according to their cleavage sites. Endochitosanases (EC 3.2.1.132; created 1990, modified 2004) cleave a partly acetylated chitosan at random and produces COS. Exochitosanases are usually called exo-1,4-β-D-glucosaminidase (GlcNase, EC 3.2.1.165; created 2008), which cleaves β-D-glucosamine (GlcN) residues continuously from the non-reducing end of the substrate ( [33]. It is easier and more efficient to cleave the β-1,4-glycosidic linkages in chitosan than in chitin, because of its solubility in weakly acidic solutions. Chitosan, being a base, can be solubilized in both inorganic and organic acids, forming salts with acids. Acetic, lactic, and citric acids have all been used to facilitate enzyme hydrolysis. The solubility of chitosan depends on the molecular weight and degree of N-acetylation of the chitosan, concentration of the acid, and temperature. Homogeneous deacetylation of chitin to chitosan (approximately 50% DD) gives a water-soluble polymer.
The solubilized chitosan salt is a preferred substrate for chitosanases in homogeneous aqueous acids. Thus, chitosan is more rapidly hydrolyzed and gives a higher yield than chitin, which is not soluble and heterogeneous in aqueous environment. The products of the hydrolysis reaction of chitosan salt are accordingly the salt of its oligosaccharides. This means that a desalting process is necessary to obtain a pure oligosaccharide. Desalting is a rather complicated and costly process. In commercial products, the content has been expressed as being the salt of acetic, lactic, and citric acid of the oligosaccharides. Glucosamine is supplied in the marketplace as a dietary supplement in the form of a hydrochloric or sulfuric salt.

β-D-Glucosamine
The hydrolytic products of chitosan are mixtures of COS (DP 1-7), dependent on the reaction conditions, even though the monomer becomes a major end product after prolonged reaction time. Thus, chromatographic separation is necessary to purify each monomer, same as in chitin oligosaccharides. The enzymatic preparation of chitosan oligosaccharides is summarized in Table 2.
Chitosan oligosaccharides can be applied for chitosanase as a substrate, not for mass production of glucosamine but for elucidation of the reaction mechanism of chitosanase. GlcN resulted from the hydrolysis of chitobiose, chitopentose, and chitosan from the 79-kDa GlcNase of Nocardia orientalis IFO 12806 [73]. GlcN was also the final product from (GlcN)6 by the 93-kDa chitosanase purified from Trichoderma reesei PC-3-7 [78]. (GlcN)6 appeared to be hydrolyzed to GlcN5 and GlcN at the initial stage of the reaction.
These enzyme preparations contained endochitosanase activity. The endo-type activity is responsible for the production of chitosan oligosaccharides. Endo-type chitosanase is a major contributor to decreasing the viscosity of the reaction solution and fouling of the membrane reactor system [69]. The viscosity of high molecular weight material such as chitosan is involved in producing shearing forces in reactors, changes in enzyme-substrate affinity, and the fouling problem. When mixed with endochitosanase, the viscosity of chitosan decreased rapidly in accordance with production of the oligomers.
Thus  -3GlcNOH were not hydrolyzed. The monomer GlcNor GlcNOH was never detected from the enzyme reaction. These results suggest that D-11 chitosanase recognizes three glucosamine residues in minus position and simultaneously two residues in plus position from the cleavage point [85].
Pantaleone et al. [86] reported the hydrolytic susceptibility of chitosan to a wide range of enzymes, including glycanases, proteases, and lipases derived from bacterial, fungal, mammalian, and plant sources. These nonspecific enzymes have proven hydrolytic ability on chitosan to produce various chitosan oligomers in immobilized papain [79], lysozyme [87], and pronase [88]. Xie et al. [89] reported that chitosan (80% DD) was depolymerized by the cellulase of Aspergillus niger to give COS with DP 3-11. Mixture of cellulase, α-amylase and proteinasewas effective in the production of COS (DP 5~17) [81]. This finding presents another opportunity in the industrial scale production of COS, because these enzymes are active enough, commercially available, cheaper than chitosanase, and easy to handle. A commercial lipase was also applied for enzymatic preparation of COS, where COS with DP 1-6 was produced with 93.8% yield for 24 h hydrolysis at 37 °C [80]. Recently, the commercial α-amylase was used to hydrolyze chitosan from Clanis bilineata larvae skin under the optimal pH and temperature, thereby affording COS with a DP 2-8 [90].
Most chitosanase-producing bacteria require chitosan as a carbon/nitrogen source and chitosanase-inducer. Thus, one can get COS from the culture supernatant of the microorganism.
Reducing COS sugars from shellfish waste were produced in the culture supernatant of Acinetobacter calcoaceticus TKU024, including chitosanases (CHSA1 and CHSA2) after 5 days of incubation [72]. Even though the yield of this method is very low, this kind of trial could be considered as a stepping stone to the direct extraction of COS from bio resources.
Intact cells of Rhizopus oligosporus NRRL2710were digested with a GH-46 chitosanase from Streptomyces sp. N174 [82]. Valuable hetero-and homo-oligosaccharides GlcN-GlcNAc, (GlcN)2-GlcNAc, and (GlcN)2 were produced, functionally, by the enzymatic digestion of the intact cells. The chitosanase digestion of intact fungal cells should be an excellent system for bioconversion of abundant microbial biomass without any environmental impact.

Chitin Deacetylase and Chitooligosaccharides
The chemical conversion from chitin to chitosan, the most difficult and cost-demanding step, has been done with 50% NaOH and high temperature in industrial applications. In nature, this step occurs at comparatively lower temperatures and neutral pH, by chitin deacetylase (CDA, EC 3.5.1.41; created 1976) [91]. The enzyme hydrolyzes the linkage between the acetyl group and the amine group in N-acetyl-D-glucosamine residues of chitin. Thus, bioconversion of chitin oligosaccharides to chitosan oligosaccharides can be achieved by CDA and vice versa.
A solid natural substrate shrimp chitin could be deacetylated with an 11% deacetylation by CDA from Saccharomyces cerevisia [98]. Pre-hydrolysis of crystalline shrimp chitin by grape chitinases increased the deacetylation triggered by CDA and produced COS with a degree of deacetylation of 67% [98]. It is well known that the high crystallinity of chitin microfibril, by the hydrogen bond-stabilized packaging of chitin polymer, greatly impedes the access of the enzyme to the deacetylation reaction site in the chitin molecule. Enzymatic deacetylation is profoundly affected by the physical properties of the substrate, such as crystallinity, degree of deacetylation, particle size, and origin [99]. Pretreatment to destroy the crystalline structure prior to addition of the enzyme seems to be desirable, in order to improve the deacetylation rate and produce novel chitosan polymers and oligomers. The preparation of COS including N-deacetylation and transglycosylation are summarized in Table 3.

Transglycosylation and Chitooligosaccharides
In addition to hydrolytic activity, some chitinolytic enzymes possess certain level of transglycosylation ability, that is, the ability to transfer the released oligosaccharide moiety to a suitable acceptor to form a new glycosidic bond. The transglycosylation activity of these chitinolytic enzymes suggests great potential for the synthesis of size-and stereo-specific chitin/chitosan oligomers, or even polymers and their derivatives.
Another example of enzymatic transglycosylation shows its diversity and potential in commercial applications. Fujimoto et al. [106] have reported the synthesis of gentiooligosaccharides (DP 3-9) from gentiobiose using a crude enzyme preparation from P. multicolor. The transglycosylation was shown to take place in two stages by a combination of β-glucosidase and β-(1-6)-glucanase. In the beginning, β-glucosidase produced gentiotriose from gentiobiose, and then β-(1-6)-glucanase acted on the resulting gentiotriose to produce a series of gentiooligosaccharides (DP 3-9) by transglycosylation. The transglycosylation reaction has high potential in the small-scale preparation of high value glycoside products applicable in medical and industrial fields.
The chitinase-catalyzed glycosylation using the sugar oxazoline substrate was applied to the stepwise elongation of GlcNAc unit by utilizing two enzymes, chitinase for ring-opening polyaddition of N-acetyllactosamine(Gal-β-(1→4)-GlcNAc) oxazoline (Gal = galactose) and β-galactosidase for the removal of the galactose unit from the produced oligosaccharide. [110][111][112][113][114]. The hydrolytic removal affords a new glycosyl acceptor with the GlcNAc unit at the non-reducing end. Repetition of these procedures using the two enzymes enabled the synthesis of chitooligosaccharides with desired chain lengths.
Interestingly, N-acetyllactosamine oxazoline was found to bepolymerized by chitinase A1 from Bacillus circulans WL-12 catalysis under basic conditions, giving rise to a novel oligosaccharide having the β-(1→4)-β-(1→6)-linked repeating unit in the main chain [115]. The DP of the resulting oligosaccharides was up to 5 based on the disaccharide. This was the first example of enzymatic glycosylation forming β-(1→6)-glycosidic linkage by chitinase catalysis.
The main disadvantage using the chemo-enzymatic approach is poor yield, because the product becomes necessarily a substrate for the enzyme ( Table 4). Suppression of the chitinase-catalyzed hydrolysis of the product during the enzymatic polymerization is a challenge for chemoenzymatic glycosylation of COS. Wild-type chitinaseA1 from Bacillus circulans WL-12 has D202 and E204 residues as a DXE (D = Asp, X = any amino acid, E = Glu), a general sequence at the catalytic domain of chitinase [116,117]. A mutant chitinase E204Q (Q = Gln) exhibited less hydrolysis activity of the produced oligosaccharides by the enzymatic glycosylation, probably less protonation ability toward the oxygen of the glycosidic linkage by replacement of COOH in Glu with CONH2 in Gln [118]. Table 4. Chemoenzymatic preparation of chitin and chitosan oligosaccharides (COS) and its derivatives.

Direct Degradation and Separation of Chitin from Crustacean Shells or Squid Pens
The direct degradation and separation of α-chitin from crab and shrimp shells, and microbial cell walls poses a significant challenge. In light of the difficulties associated with traditional chitin oligosaccharides' production processes, environmentally compatible and reproducible degradation alternatives are desirable. However, crustacean shells are not soluble in standard aqueous media and their crystallinity is potentially too high to be degraded enzymatically. To solve these problems, pretreatment to break the chitin crystal structure is widely considered.
In this regard, chitin substrates were pretreated with steam explosion prior to enzymatic reaction [122]. A 11.28% reduction of the crystallinity index was observed with steam explosion and a 23.6% yield of chitin oligosaccharides with DP up to 5 was achieved. Interestingly, Nakagawa et al. [123] reported a breakthrough in the direct production of oligosaccharides from chitin and crab shells. They introduced a combination protocol of mechanochemical grinding and enzymatic hydrolysis, and produced GlcNAc and (GlcNAc)2 directly from crab shells and chitin. The direct degradation ratio of the chitin in crab shell was close to 100%. For this purpose, they developed a novel "converge" mill [124], a derivative of the medium ball mill. Mechanochemical grinding, with the converge mill, was found to be extremely effective for pretreating chitin and crab shell before enzymatic digestion [125].
Breaking down the chitin crystal structure of biomaterials improves enzymatic degradation, allowing the enzymes to easily access and exert catalytic action. Wu and Miao [126] showed that mechanochemical treatment markedly increases the glucose yield from enzymatic corn flour hydrolysis. Similar results were observed by Fujimoto et al. [127] for lignocellulosic biomass. Van Craeyveld et al. [128] also showed improvement of extractability and molecular properties of Psyllium seed husk arabinoxylan by ball milling.
Mechanochemical grinding also provides an additional advantage. In the case of chitosan oligosaccharide production, salts are inevitably formed from aquatic reaction mixtures. Sometimes the bound salts limit application of COS. Salt production can be circumvented by mechanochemical grinding of the substrate, followed by enzymatic hydrolysis.

New Enzymes with Better Properties
Solubilization has become the first choice for breaking down the chitin crystal structure in chitin and crab shells to improve enzymatic degradation. Solvents are necessary to solvate the substrate, and these are still being studied along with continuing research seeking enzymes that are stable in organic solvents or heterogeneous solvent systems. Enzymes that function in an extreme environment are important for this purpose and their mode of action differ fundamentally from of glycoside hydrolases.
During researches on enzymes capable of efficiently degrading recalcitrant polysaccharides such as cellulose and chitin for biofuels, it has been speculated about the existence of a substrate-disrupting factor that could make the crystalline substrate more accessible to hydrolytic enzymes [129]. Vaaje-Kolstad et al. [130] showed that CBP21 (CBP for chitin-binding protein), produced by the chitinolytic bacterium Serratia marcescens, is an enzyme that catalyzes cleavage of glycosidic bonds in crystalline chitin, thus opening up the in accessible polysaccharide material for hydrolysis by normal glycoside hydrolases. The products were identified as chitin oligosaccharides with a normal sugar at the non-reducing end and an oxidized sugar, 2-(acetylamino)-2-deoxy-D-gluconic acid (GlcNAcA), at the other end. This is an oxidative enzyme boosting the enzymatic conversion of recalcitrant polysaccharides. This finding helps understand the factors involved in degradation of crystallinechitin and the chitin cycle in nature and provides a stake to more precisely handle the process for COS production from crystalline chitin and even raw materials such as crab and shrimp shells.
Recently, several commercially available hydrolytic enzymes including lysozyme, cellulase, papain, pectinases, and hemicellulase were found to catalyze the cleavage of the glycoside bond in chitin and chitosan [83]. Use of these nonspecific enzymes together with specific chitinase, chotosanase, glycosyltransferase, and CDA certainly opens a possible route to optimize the hydrolysis reactions controlling the production of chitooligosaccharides. Although different nonspecific enzymes have been used to obtain COS from chitin and chitosan, due to the limited capacity of most hydrolytic enzymes there is still interest in finding new enzymes with better properties.
Branchzyme is a relatively inexpensive commercial preparation that contains a branching glycosyltransferase from Rhodothermus obamensis expressed in Bacillus subtilis. This enzyme catalyzes the transfer of a segment of a 1,4-α-D-glucan chain to a primary hydroxyl group in a similar glucan chain to create 1,6-α-linkages, thereby increasing the number of branch points [131]. Interestingly, Branchzyme was active on chitosan and produced COS with DP 2-20, with a higher concentration having COS with DP 3-8 [132]. Recently, a family 46 chitosanase from S. coelicolor A3(2) was employed to degrade both a fully deacetylated chitosan and a 68% deacetylated chitosan for the production of a series of COS and to study in-depth the enzyme's mode of action [133].

Manipulation of the Size-Distribution of Oligomers Produced by Enzymatic Bioconversion
The DD, DP, the molecular weight distribution, and N-acetylation pattern (PA) of the resulting COS mixture depend on the starting material (mostly DD of chitin and chitosan) and the specificity of the enzyme used. Product mixtures can be enriched for certain compounds by optimizing the substrate-enzyme combination. This is illustrated by several studies on enzymatic degradation of chitosans [134][135][136][137][138][139][140]. The degradation of chitosan by the family 18 chitinases, ChiA, ChiB and ChiC, from Serratia marcescens has been studied in detail [36,134,141,142]. The biphasic kinetics of the reaction with ChiB was illustrated, as in the initial rapid phase and the second much slower phase of the reaction [36]. Sorbotten et al. [36] showed that the size-distribution of oligomers could be manipulated by varying the DD of the chitosans (DD 87%, 68%, 50%, and 35%) with ChiB; the products get longer as the DD goes up. Very interestingly, Sikorski et al. [137,142] have produced a model for the degradation of different chitosans with ChiB, which is capable of predicting the length distributions and yield of the products after extended hydrolysis. These profiles allow for selection of optimal reaction and substrate parameters for efficient production of oligomers with desired lengths. Altogether, this suggests that production of oligomers with desired lengths can be achieved by manipulating the choice of the starting materials with specific DD, and PA, the choice of the enzyme, and the choice of the processing time [130].

Genetic Engineering Technology
Genetic engineering technology is applicable and promising for mass production of recombinant enzymes and understanding mode of action, which is useful for bioproduction of COS and synthesis of structure-specific oligosaccharides. Martinez et al. [51] declared that enzymatic synthesis of COS has been a matter of research by exploiting the transglycosylation activity of retaining family 18 chitinases. The mutagenesis of two GH-18 glycoside hydrolase, B. circulans WL-12 chitinase A1 (Bc ChiA1) and Trichoderma harzanium chitinase 42 (Th Chit42) abolished the hydrolytic activity of chitin [51]. The mutants D200A and D202A of Bc ChiA1, together with D170N of Th Chit42 proved to be active for chitinbiose oxazoline polymerization, and also for coupling reaction between Gal (β1→4) chitinbiose oxazoline and chitinpentaose at neutral pH. These mutants have retained the ability to catalyze transglycosylation reaction on natural COS. Such mutants could be considered as chitin transglycosylases.
Chitin synthase, a processive inverting enzyme of glycosyltransferase family 2, transfers GlcNAc from UDP-GlcNAc to preexisting chitin chains in reactions that are typically stimulated by free GlcNAc. This enzyme can be applied for synthesis of chitin oligosaccharides. By using a recombinant Saccharomyces cerevisiae chitin synthase with UDP-GlcNAc, GlcNAc2 was obtained as a major reaction product [143]. Formation of both COS and insoluble chitin was stimulated by GlcNAc2 and by N-propanoyl-, N-butanoyl-, and N-glycolyl-glucosamine.

Direct Fermentation of Raw Materials such as Crab and Shrimp Shells
The direct fermentation of raw materials such as crab and shrimp shells presents another opportunity in the production of chitin and COS. Chitin oligosaccharides were directly obtained from 1.5% shrimp head powder after 2 days of fermentation with protease-and chitosanase-producing Bacillus cereusTKU022 [144]. (GlcNAc)2, (GlcNAc)4, (GlcNAc)5, and (GlcNAc)6 were all identified in the culture supernatant. Even though the concentration of the products is still very low (0.3-201.5 μg/mL), these findings are meaningful in terms of direct production from raw materials.

Reactor Systems
Finally, the reactor system used should be considered to be a critical factor for efficiency and high yield in bioproduction. There are three types of reactors normally used in the preparation of chitosan oligosaccharides, including batch reactors, column reactors and ultrafiltration reactors [145]. Batch reactors represent the simplest method used for the enzymatic production of COS. This method has a few drawbacks, including limited reuse of reacting enzyme and continuous production of COS, difficulty in controlling molecular weight of COS, low yield and high cost. Continuous preparation of COS is possible by packing immobilized enzymes into a column reactor that the substrate passes through. Though this method has several advantages over batch reactors, the poorer affinity of immobilized enzymes to chitosan substrate than that of free enzyme limits the activity of enzyme and the usage of column reactors in the commercial preparation of COS.
The ultrafiltration membrane reactor has been applied for overcoming the problems of the reusability of enzymes in batch reactors and the poor affinity of the substrate toward immobilized enzymes in column reactor systems, but suffers from the problem of membrane fouling [25,69]. A dual ultrafiltration membrane reactor system composed of a column reactor packed with immobilized enzymes and an ultrafiltration membrane reactor, has been invented and applied in COS production [146]. Using this reactor, chitosanase from Bacillus pumilus BN-262 continuously produced chitosan oligomers of DP 3~6 from chitosan (89% DD), free from any fouling problem.

Conclusions
Bioproduction of COS with enzymes and microorganismshas been studied for decades. However, the yield of bioproduction is still lower and the cost is higher than traditional chemical methods. Crude rather than pure enzyme preparations of chitinase, chitosanase, lysozyme, cellulase, protease, lipase, and pepsin were preferred for this purpose for the practical production of the oligosaccharides. The transglycosylation activity of chitinolytic enzymes was successfully exploited for the synthesis of desired chitin oligomers and their derivatives. Chitin deacetylase is also applicable to the preparation of oligosaccharides.
The direct degradation and separation of chitin from crab and shrimp shells and microbial cell walls presents a significant challenge in bioproduction of COS. The direct production of oligosaccharides from chitin and crab shells was achieved by using a combination of mechanochemical grinding and enzymatic hydrolysis. In case of chitosan oligosaccharides, salt forms are inevitably formed from aquatic reaction mixtures, which limit the application. However, the mechanochemical grinding of the substrate followed by enzymatic hydrolysis can circumvent this.
Breaking down the crystal structure of chitin and crab shells is necessary to improve enzymatic degradation of the substrate. Solubilization is the first choice for destruction of chitin crystallinity. Once solubilized in organic or heterogeneous solvent systems, the substrate must receive hydrolytic enzymes with special properties. Thus, screening high-potential enzymes that function in extreme environments is an important factor for the optimization of bioproduction.
The direct fermentation of raw biomaterials like crab and shrimp shells presents another opportunity in the production of chitin and COS. Even though the efficiency of this method is still very low, it is significant in terms of the direct production of COS from the row materials.
To overcome these inefficiencies, the optimization of a reactor system appropriate to enzymatic bioproduction is called for. A dual ultrafiltration membrane reactor system has been described that is capable of overcoming the problems of enzyme reusability in batch reactors as well as the poor affinity of the substrate toward immobilized enzymes in standard column reactor systems.