Molecular Characterization of Four Alkaline Chitinases from Three Chitinolytic Bacteria Isolated from a Mudflat

Four chitinases were cloned and characterized from three strains isolated from a mudflat: Aeromonas sp. SK10, Aeromonas sp. SK15, and Chitinibacter sp. SK16. In SK10, three genes, Chi18A, Pro2K, and Chi19B, were found as a cluster. Chi18A and Chi19B were chitinases, and Pro2K was a metalloprotease. With combinatorial amplification of the genes and analysis of the hydrolysis patterns of substrates, Chi18A and Chi19B were found to be an endochitinase and exochitinase, respectively. Chi18A and Chi19B belonged to the glycosyl hydrolase family 18 (GH18) and GH19, with 869 and 659 amino acids, respectively. Chi18C from SK15 belonged to GH18 with 864 amino acids, and Chi18D from SK16 belonged to GH18 with 664 amino acids. These four chitinases had signal peptides and high molecular masses with one or two chitin-binding domains and, interestingly, preferred alkaline conditions. In the activity staining, their sizes were determined to be 96, 74, 95, and 73 kDa, respectively, corresponding to their expected sizes. Purified Chi18C and Chi18D after pET expression produced N,N′-diacetylchitobiose as the main product in hydrolyzing chitooligosaccharides and colloidal chitin. These results suggest that Chi18A, Chi18C, and Chi18D are endochitinases, that Chi19B is an exochitinase, and that these chitinases can be effectively used for hydrolyzing natural chitinous sources.

Chitinases belong to the glycosyl hydrolase (GH) families 18,19, and 20 [3]. GH18 and GH19 are mainly endochitinases, and GH20 includes N-acetyl-β-hexosaminidase. GH18 chitinases have been found in various organisms from bacteria to humans, including plants, molds, and vertebrates, while GH19 chitinases are found mainly in plants and are not common in bacteria, having recently been found in only a few strains, such as Chitinophaga sp. [4], Pseudoalteromonas rubra, [5], and Streptomyces alfalfa [6].

Isolation of Chitinolytic Bacteria
From the enriched mudflat sample, 20 colonies were primarily selected on LBCC agar plates. Based on the color and shape of the colony and the size of the clear zone, six colonies were further selected. After measuring the extracellular chitinase activity of the six colonies, three colonies were selected for further experiments: SK10, SK15, and SK16 ( Figure 1). Chitinase activities of the smallest colonies with the largest zone were low in culture broth assays. The 16S rRNA sequence of the isolate SK10 showed 100% similarity with Aeromonas hydrophila (GenBank accession number MG428960), and that of SK15 showed 99.9% similarity with Aeromonas punctata (AM184292) in the NCBI server. That of SK16 showed 98.2% similarity with C. tainanensis (AY264287) and represented a novel species of the genus Chitinibacter, i.e., C. suncheonensis sp. nov. [18]. The isolates SK10 and SK15 were named Aeromonas sp. SK10 and Aeromonas sp. SK15, respectively, and their phylogenetic tree is shown in Figure 2.

Growth and Chitinase Production from the Isolates SK10, SK15, and SK16
When the isolates were cultured with shaking in LB, SK10 and SK15 grew faster than SK16, and all three isolates reached a plateau for growth after 12 h ( Figure S1). When the isolates were cultured with media, such as LB, LBCC, and 0.5 × LB/colloidal chitin, chitinase activities were maximal at 27 or 30 h of culture in an LBCC medium ( Figure S1).
Chi18A had a putative signal peptide comprising 23 amino acid residues and a conserved motif of 307FDGVDIDWE315, which was highly conserved in the GH18 chitinases ( Figure S2). Chi19B had a signal peptide comprising 42 amino acid residues and conserved motifs that were highly maintained in the GH19 chitinases ( Figure S3).

Cloning of Chitinase from SK15
In the library, three active clones, SK15-27, SK15-36, and SK15-65, were screened on the plate. Insert DNA of SK15-36 was 3293 bp in length and contained an ORF, matching a chitinase belonging to GH18, named Chi18C ( Figure 7). Chi18C comprised 864 amino acid residues with a signal peptide of 23 amino acid residues ( Figure S4) and two chitin-binding sites at the C-terminus (Figure 7). Chi18C showed 99.8% similarity with a chitinase from A. punctata and was located with other chitinases from Aeromonas sp. in the phylogenetic tree ( Figure 8). Using a primer set amplifying the conserved motif, it was observed that the same size of DNA fragment appeared in the other two active clones, suggesting that the three clones produced the same kind of chitinase.

Cloning of Chitinase from SK16
In the library, three active clones, SK16-4366, SK16-4369, and SK16-9122, were screened on the plate. The insert DNA of SK16-4366 was 3215 bp in length and contained an ORF, matching a chitinase belonging to GH18, named Chi18D ( Figure 9). Chi18D comprised 664 amino acid residues with a signal peptide of 31 amino acid residues and the conserved GH18 motif ( Figure S5). Chi18D showed 54.6% similarity (the highest level of similarity) with a chitinase from Doohwaniella chitinasigens and was located close to chitinase in the phylogenetic tree ( Figure 10). Using a primer set amplifying the conserved motif, it was observed that the same size of DNA fragment appeared in the other two active clones, suggesting that the three clones produced the same kind of chitinase.   The optimum temperature for Chi18A was 50 • C ( Figure 11A), and activity was decreased at 70 • C to a half of the maximum. The optimum pH for Chi18A was 8.0 ( Figure 11B). Regarding heat stabilities, Chi18A was stable at 50 • C for up to 1 h preincubation but lost >60% of its original activity after 15 min preincubation at 60 • C ( Figure 11C). The half-lives of Chi18A at 60 and 70 • C were 25.0 and 13.8 min, respectively. Cations Mg 2+ and K + increased Chi18A activity to 120% and 117%, respectively, and Cu 2+ , Fe 2+ , Mn 2+ , and Zn 2+ decreased its activity to 42.6%, 44.1%, 68.5%, and 88.1%, respectively ( Figure 11D). In the experiments on concentration dependency, Mg 2+ and K + showed maximal activity at 5.0 and 7.5 mM, respectively ( Figure S6).

Chi18D
The optimum temperature for Chi18D was 50 • C ( Figure 14A), and its activity was decreased at 70 • C to 80% of the maximum. The optimum pH for Chi18D was 9.5 ( Figure 14B). Regarding heat stabilities, Chi18D was stable at 50 • C for up to 1 h preincubation; however, it rapidly lost >60% of its activity after 15 min preincubation at 60-80 • C ( Figure 14C). The half-life of Chi18D at 70 • C was 9.3 min.

Site-Directed Mutagenesis of Chitinases
When Glu315, Glu315, and Glu408 of Chi18A, Chi18C, and Chi18D, respectively, were mutated to Ala, the activities of the mutants were found to be lost completely on an LBCC plate and assay of the crude extract ( Figure 15), verifying that those residues were catalytic sites of the enzymes.

Analysis of the Molecular Masses of Chitinases Using Activity Staining
By activity staining with MUCh 2 after SDS-PAGE and the renaturation method, the molecular masses of Chi18A, Chi19B, Chi18C, and Chi18D were estimated to be 96, 74, 95, and 73 kDa, respectively ( Figure 16). The results closely corresponded to the expected values. For Chi18A, two active bands, corresponding to 75 and 65 kDa, were detected due to the internal cleavage of the enzyme.

Purification of Chi18C and Chi18D
Each enzyme was purified using the Ni-NTA method after Chi18C and Chi18D were separately expressed with a pET28a(+) vector. Chi18C was greatly induced by IPTG, purified to homogeneity, and exhibited an active band at the corresponding position ( Figure 17). Chi18D was purified with a minor, smaller protein band and showed an active band ( Figure 17). The results revealed that Chi18C and Chi18D were successfully expressed and purified in active forms.

Hydrolysis of Chitooligosaccharides by the Purified Chi18C and Chi18D
When the purified Chi18C was reacted with the chitooligosaccharides (Ch 2 to Ch 6 ) as substrates, the major product of Ch 3 to Ch 6 was Ch 2 , whereas Ch 2 was not hydrolyzed ( Figure 19). The amount of Ch 2 as product was proportional to the degree of polymerization of chitooligosaccharides; the hydrolysis of Ch 6 was the highest among those tested. With colloidal chitin as a substrate, Ch 2 was the main product, and the amount increased with the reaction time ( Figure 19). When the purified Chi18D was reacted with the chitooligosaccharides and colloidal chitin as substrates, the hydrolysis results were the same as those for Chi18C (Figure 19), suggesting that Chi18C and Chi18D are endochitinases. The monosaccharide Ch 1 spots were observed in the reactions of E 3 to E 6 , and the spots in E 3 and E 5 were more intensive than those in E 2 , E 4 , and E 6 . This pattern suggests that Chi18C and Chi18D hydrolyze chitooligosaccharides consisting of more than two units of NAG by degradation of Ch 3 and Ch 5 into Ch 2 and Ch 1 .

Discussion
Marine bacterial chitinases are considered sources of energy, eco-friendly agents, and industrial biocatalysts [11]. In this study, we identified two strains, SK10 and SK15, as Aeromonas sp., among three isolates from a marine source, i.e., a mudflat. In a previous report, SK16 was deposited as a new species of C. suncheonensis sp. nov. [18]. Furthermore, four chitinase genes were cloned from them: chi18A and chi19B from Aeromonas sp. SK10, chi18C from Aeromonas sp. SK15, and chi18D from C. suncheonensis sp. nov. Three GH 18 family members, i.e., Chi18A, Chi18C, and Chi18D, were endochitinases, and GH 19 family member Chi19B was found to be an exochitinase, based on an inability to hydrolyze colloidal chitin but an ability to hydrolyze MUCh 2 .
No information about the chitinase gene from Chitinibacter sp. has been reported in the literature; however, recently, a chitinase gene was isolated and cloned from C. tainanensis, and an encoded enzyme was characterized [19].
Chi18A and Chi18C have a high identity, with a score of 80.46%; however, there are some differences in properties; in optimum temperatures, i.e., 50 • C and 60 • C, respectively; and in ion effects, i.e., Chi18A was activated by Mg 2+ and K + with relative activities of 120% and 117%, respectively, but Chi18C was not activated. In addition, Chi18A was inhibited by Fe 2+ , but Chi18C was inhibited by Ca 2+ .
In general, endochitinase cleaves substrates of a large size by attacking the inside randomly, i.e., endo-acting, to produce oligomers and finally disaccharide N,N"-diacetylchitobi ose as a main product. Exochitinase cleaves the substrates by hydrolyzing non-reducing ends to produce the disaccharide as a product but does not attack inside the substrates. The patterns can be distinguished experimentally by primarily observing halo formation abilities for colloidal chitins (large-sized substrates) and additionally disaccharide production for oligosaccharides (small-sized substrates) as a main product. If both are observed, the enzyme is an endochitinase. However, if a halo is not formed for colloidal chitins, the enzyme is an exochitinase. In this study, Chil18C and Chi18D could degrade colloidal chitin to form halos, as well as Chi18A, and the major product was dimer. Therefore, Chi18C and Chi18D, as well as Chi18A, showed endo-type reactions, i.e., endochitinase, like most of the family 18 chitinases [15,30,38], except ChiC [32] and BthChi74 from Bacillus thuringiensis [35], which showed an exo-type reaction, i.e., exochitinase (Table 2), producing Ch 2 as its main products. Most of the family 19 chitinases were reported to be endochitinase. However, in our experiment, Chi19B exhibited an exo-type reaction based on non-hydrolytic results for colloidal chitin and hydrolytic results for MUCh 2 , which are similar to Ch 3 in their structures. Like Chi19B, it was reported that Chi19 from Vibrio proteolyticus hydrolyzed colloidal chitin to release small soluble oligosaccharides at an early stage, noting that Chi19 was not a strict exo-type reaction and an exo-like chitinase because its products contained a small amount of Ch 3 and Ch 4 [29] (Table 2). Further study of the substrate specificity of Chi19B will be necessary. In predicting 3D structures, models of Chi18A and Chi18C showed the highest identity to Chitinase A (PDB code: 1 x 6l.1.A) and Chitinase A (PDB code: 1ffr.1.A), respectively. Due to the high molecular weights of Chi18A (92.7 kDa) and Chi18C (91.6 kDa), which are unique cases of chitinases, there were no predicted models that covered full domains entirely. However, two structures were predicted, matching the N-and C-terminal parts of Chi18A or Chi18C with the highest identity ( Figure 20). In the N-terminal part-containing structure of Chi18A (24-565), ChBD and the catalytic domain were predicted to be A24-A131 and L155-D565, respectively, with an identity of 74.42%. For the C-terminal partcontaining structure of Chi18 A (155-815), two Ig-Like domains were predicted at P574-P758, which were previously reported [41] (Figure 20), and another ChBD was predicted at D768-Q809 with an identity of 18.40%. Chi18C showed similar patterns with Chi18A, i.e., N-terminal ChBD at A24-A108, catalytic domain at K155-Y562, Ig-like domains at P570-K752, and C-terminal ChBD at A768-Q809 ( Figure 20). Collectively, Chi18A and Chi18C were predicted to have two ChBD and two Ig-like domains, suggesting that they promote substrate affinity with a large binding surface [41].
Chi19B showed the highest identity (32.95%) with chitinase C (PDB code: 1wvv. 2.B). However, a model could not be predicted to cover the full domains ( Figure 21). In the predicted model, ChBD and the catalytic domain were predicted at A24-G95 and K172-C427 in the N-terminal part-(42-427) containing structure, and the cellulose-binding domain was predicted at Y481-I592 in the C-terminal part-(481-592) containing structure ( Figure 21). According to the 3D modeling result, we checked activity with carboxylmethylcellulose (CMC) using an LB agar plate containing 1% CMC, and then stained with Congo red. However, activity was not detected.   Chi18D showed the highest identity (44.01%) with chitinase B (PDB code: 1kfw.1.A), but it could not cover the full domains. Therefore, we independently performed the Nterminal part-containing (1-225) structure and confirmed that it showed the highest identity (34.52%) with ChBD of Deacetylase DA1 (PDB code: 4ny2.1.A) ( Figure 22). Collectively, the ChBD and catalytic domain were predicted at A31-Q161 and P226-A661, respectively ( Figure 22). These differences in their structures might suggest that they have different substrate specificities depending on the degree of polymerization of chitin, as observed in the halo formation differences for colloidal chitin hydrolysis.

Isolation of Chitinolytic Bacterial Strains and Culture Conditions
Chitinolytic bacterial strains were previously isolated from an enriched mudflat in Suncheon Bay, Republic of Korea, using Luria-Bertani (LB) medium containing 0.2% colloidal chitin (LBCC) agar plates [18]. Colloidal chitin was prepared using Hsu and Lockwood's [42] method with modifications. After incubation at 30 • C for three days, positive colonies with large zones of hydrolysis were selected. Extracellular chitinase activity in the supernatant was measured after being grown at 30 • C for 18 h in an LBCC medium.

Identification of Bacterial Strains
The genomic DNAs of the selected strains were isolated using a Genomic DNA Extraction Kit (SolGent, Daejeon, Korea), and their 16S rRNA sequences were determined by SolGent, as described previously [43]. Sequence similarities of the 16S rRNA and phylogenetic trees were searched and analyzed using the BLASTN program [44,45] on the NCBI website. The SK16 strain was previously deposited in KCTC and the DSM under the numbers KCTC 23839 and DSM 25421, respectively, as a new species, C. suncheonensis sp. nov., and its 16S rRNA sequence was deposited in GenBank under the accession number JN981166 [18]. Strains SK15 and SK10 were deposited in KCTC under the numbers KCTC 42713 and KCTC 42714, respectively, as Aeromonas sp. SK15 and Aeromonas sp. SK10. The 16S rRNA sequences of SK15 and SK10 were deposited in GenBank under the accession numbers MZ573230 and MZ573228, respectively.

Monitoring the Growth and Chitinolytic Activities of the Isolates SK10, SK15, and SK16
Three strains were cultured for 36 h at 37 • C with shaking at 150 rpm in 200 mL of LB medium 1-L flasks. Cell growth was monitored by sampling every 3 h. Effects of substrate addition on extracellular chitinolytic activity were analyzed for 36 h using culture supernatants sampled every 3 h with LB, LBCC, and 0.5 × LB containing 0.2% colloidal chitin.

Cloning and Analysis of Chitinase Genes from the Isolates SK10, SK15, and SK16
Chitinase genes were cloned from the isolates SK10, SK15, and SK16 using pUC19 and Escherichia coli DH5α (Yeastern Biotech. Co., Taipei, Taiwan), as described previously [43], with slight modifications. E. coli transformants were primarily grown on LB agar plates supplemented with ampicillin (50 µg ml −1 ), X-Gal, and IPTG for about 24 h at 37 • C. The transformants were tooth-picked to LBCC agar plates, grown for about 24 h at 37 • C, and the colonies with a hydrolysis zone were selected. The nucleotide sequence of the insert DNA was determined by SolGent. The conserved region of the gene was identified with BlastN or BlastP of BLAST of NCBI (http://www.ncbi.nlm.nih.gov, accessed on 18 August 2021), and a phylogenetic tree of the gene was constructed using DNA/MAN (Lynnon Biosoft, version 4.11, Quebec, QC, Canada). The signal peptide was predicted with SignalP 5.0 in CBS (http://www.cbs.dtu.dk/services/SignalP/, accessed on 18 August 2021) [46]. The molecular mass and pI of the encoded protein were predicted, and multiple alignments were constructed using DNA/MAN (Lynnon Biosoft, version 4.11, Quebec, QC, Canada). Identified chitinase gene sequences of chi18A, chi19B, chi18C, and chi18D were deposited in GenBank under the accession numbers MZ673655, MZ673656, MZ673657, and MZ673658, respectively. Three dimensional structures of them were predicted by SWISS-MODEL (https://swissmodel.expasy.org/, accessed on 21 August 2021).

Subcloning of Chitinase Genes from the Active Clone SK10-5
Three open reading frames (ORFs) in the active clone SK10-5 were amplified individually or combinatorially using the primer sets ( Table 1). The insert DNA fragments of the six kinds of subclones were verified, and the activities of the subclones were analyzed on an LBCC agar plate.

Enzyme Assay
Chitinase activity was assayed in a 1.0 mL reaction mixture containing 50 mM sodium acetate buffer (pH 7.0) and 0.5% colloidal chitin. At the end of the reaction, for 30 min at 37 • C, the amounts of reducing sugar released were determined using the DNS method after centrifugation of the mixtures at 12,000 rpm for 5 min [47]. One unit of enzyme activity was defined as the amount of enzyme that liberated 1 µmol of reducing sugar per minute under the conditions. NAG was used as the standard for this method.

Biochemical Characterization of Chitinases
The effects of pH and temperature on enzyme activities were investigated using crude or purified enzymes at pH values with 50 mM of universal buffer (from pH 3.0 to 12.0) at temperatures from 40 to 80 • C and using 50 mM of sodium acetate buffer (pH 7.0). Enzyme thermostability was analyzed by preincubation without substrate for 0, 15, 30, and 60 min at 40-80 • C. The influence of various cations on enzyme activity was determined at concentrations of 5.0 mM for Na + , K + , Mg 2+ , Ca 2+ , Mn 2+ , Fe 2+ , Cu 2+ , or Zn 2+ .

Estimation of the Molecular Mass of the Chitinases by Activity Staining
The molecular masses of the chitinases were estimated using crude extracts employing the activity staining method, as previously described [48], except with MUCh 2 instead of MUG, and with modification. Shortly, after SDS-PAGE, the gels were washed three times with 20% isopropanol for 15 min to remove SDS and with 50 mM of sodium acetate buffer (pH 7.0) three times for 15 min to renature the enzyme, then soaked in a buffer containing 2 mM of MUCh 2 for 30 min at 4 • C. The gels were then transferred onto glass plates and incubated at 50 • C for 5-15 min. The MUCh 2 -hydrolyzing activity was photographed as fluorescent bands under UV light.

Site-Directed Mutagenesis
Three chitinase genes were changed using a site-directed mutagenesis method and a QuikChange II kit (Stratagene, Santa Clara, CA, USA). Primer sets were designed to mutate Glu (a tentative catalytic residue of each enzyme) to Ala ( Table 3). The substitutions were made by PCR amplification using 160 ng DNA, 10 pmol primer, 2.5 mM dNTP, and 2.5 U of Pfu Ultra HF DNA polymerase (Stratagene) with pre-denaturation at 95 • C for 30 sec and 12 cycles (denaturation at 95 • C for 30 sec, annealing at 55 • C for 60 sec, and extension at 68 • C for 6 min). The PCR products were treated with DpnI, and the resulting products were transformed to E. coli XL-1 blue super-competent cells (Stratagene). The substitutions were confirmed by nucleotide sequencing.

Expression and Purification of Two Chitinases from SK-15 and SK-16 Clones
Two chitinases of the active clones of SK-15 and SK-16 were highly expressed using a pET28a(+) vector. For pET expressions of Chi18C from SK-15 and Chi18D from SK16, primers were designed and used (Table 4). Table 4. Primers for pET expressions of Chi18C and Chi18D. Chi18C F  5  TTTGAATTCATGTTAAGTCCAAAACTTTCC  3  30  Chi18C R  5  TTTAAGCTTTCAGTTGCAGCTCGCC  3  25  Chi18D F  5  TTTGGATCCATGAGGATGGAGACCCTTATG  3  30  Chi18D R  5  TTTGAGCTCTTGACGAGCTTTACCCATTC  3  29 Expression of the protein was monitored by varying the IPTG concentrations (0.5 and 1.0 mM) and induction times (6 and 12 h). After each clone was cultured for 12 h, suspended with the lysis buffer, incubated for 15 to 30 min on ice, and centrifuged at 12,000× g for 15 min at 4 • C under the native conditions, the supernatant was loaded into the Ni-NTA column (Qiagen, Hilden, Germany), washed, and eluted with the buffer following the manufacturer's protocol. Protein concentrations were determined at 595 nm by the Bradford method using bovine serum albumin (BSA) as standard [49]. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was conducted on 11.5% polyacrylamide gels [50], which were stained with Coomassie Brilliant Blue R-250 or a Plus One Silver Staining Kit (GE Healthcare, Uppsala, Sweden).

Hydrolysis of Chitooligosaccharides and Colloidal Chitin by Purified Chitinases
The end products of the chitooligosaccharides were analyzed by incubating 0.1 U of the enzyme for 6 h and 12 h at 50 • C in the presence of 5 mM chitooligosaccharides, Ch2 to Ch6, or 0.2 U of the enzyme for 24 h with 0.5% colloidal chitin in 50 mM sodium acetate (pH 7.0). Reaction products were separated using thin-layer chromatography on a Silica gel 60 F254 plate (20 × 20 cm) (Merck, Darmstadt, Germany) with 1-butanol: acetic acid: water = 2:1:1 (v/v) as the developing solvent [51]. The products were visualized by spraying with ethanol-sulfuric acid (95:5, v/v) followed by drying for 10 min at 100 • C.

Conclusions
In this study, four chitinases were cloned and characterized from two Aeromonas spp. and C. suncheonensis. In an active clone from SK10, a gene cluster was found containing Chi18A, Pro2K, and Chi19B. It was found that Chi18A was an endochitinase and that Chi19B was an exochitinase based on the combinatorial amplification of the genes and hydrolysis experiments for colloidal chitin and MUCh 2 . Pro2K was a metalloprotease; however, its role was unclear in this study. Chi18C and Chi18D were cloned from SK15 and SK16, respectively. In the activity staining analysis, the molecular masses of Chi18A, Chi19B, Chi18C, and Chi18D were 96, 74, 95, and 73 kDa, respectively, corresponding to the expected sizes of the proteins.
Furthermore, Chi18C and Chi18D were purified after expression using a pET vector. Using the purified enzymes, it was found that both hydrolyzed chitooligosaccharides and colloidal chitin produce N,N -diacetylchitobiose as the major product, indicating that both are endochitinases. In summary, the four chitinases in this study had signal peptides and, interestingly, had high molecular masses and alkaline preferences compared to other chitinases. In the 3D model prediction, Chi18A and Chi18C were predicted to have two ChBDs and two Ig-like domains, such that they would be expected to promote substrate affinity by a large binding surface. Chi18D was predicted to have a ChBD at the N-terminus. Chi19B was predicted to have a ChBD at the N-terminus and a cellulose-binding domain at the C-terminus; however, cellulolytic activity was not detected. In addition, Chi18A, Chi18C, and Chi18D were endochitinases with ChBD, and Chi19B was an exochitinase. Further studies on substrate specificity will be necessary, however. These four chitinases can be useful for breaking chitinous materials and producing small chitooligosaccharides.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.