*Article* **Characterization of an Unknown Region Linked to the Glycoside Hydrolase Family 17** β**-1,3-Glucanase of** *Vibrio vulnificus* **Reveals a Novel Glucan-Binding Domain**

**Yuya Kumagai 1,\*, Hideki Kishimura 1, Weeranuch Lang 2, Takayoshi Tagami 2, Masayuki Okuyama <sup>2</sup> and Atsuo Kimura 2,\***


**Abstract:** The glycoside hydrolase family 17 β-1,3-glucanase of *Vibrio vulnificus* (VvGH17) has two unknown regions in the N- and C-termini. Here, we characterized these domains by preparing mutant enzymes. VvGH17 demonstrated hydrolytic activity of β-(1→3)-glucan, mainly producing laminaribiose, but not of β-(1→3)/β-(1→4)-glucan. The C-terminal-truncated mutants (ΔC466 and ΔC441) showed decreased activity, approximately one-third of that of the WT, and ΔC415 lost almost all activity. An analysis using affinity gel containing laminarin or barley β-glucan revealed a shift in the mobility of the ΔC466, ΔC441, and ΔC415 mutants compared to the WT. Tryptophan residues showed a strong affinity for carbohydrates. Three of four point-mutations of the tryptophan in the C-terminus (W472A, W499A, and W542A) showed a reduction in binding ability to laminarin and barley β-glucan. The C-terminus was predicted to have a β-sandwich structure, and three tryptophan residues (Trp472, Trp499, and Trp542) constituted a putative substrate-binding cave. Linker and substrate-binding functions were assigned to the C-terminus. The N-terminal-truncated mutants also showed decreased activity. The WT formed a trimer, while the N-terminal truncations formed monomers, indicating that the N-terminus contributed to the multimeric form of VvGH17. The results of this study are useful for understanding the structure and the function of GH17 β-1,3-glucanases.

**Keywords:** glucanase; *Vibrio*; carbohydrate-binding domain; glycoside hydrolase family 17

#### **1. Introduction**

Marine algae convert marine carbon into algal polysaccharides by photosynthesis. Algal polysaccharides are made up of a variety of glycans. The recycling of algal polysaccharides into carbon dioxide gives us a better understanding of the global marine carbon cycle [1]. Recently, the involvement of marine bacteria in this cycle has been gradually revealed [2]. Laminarin is a major glucose polymer found in marine environments [3]. Therefore, an understanding of the mechanisms underlying the degradation of large algal polysaccharides by enzymes and their modules is useful in order to produce sustainable and renewable raw materials for use in valuable compounds, feeds, and fuels [4].

Endo-β-1,3-glucanases catalyze the hydrolysis of internal β-(1→3)-glucosidic linkages. Endo-β-1,3-glucanases mainly belong to the enzyme families GH16, GH17, and GH3 [5]. The GH16 family is mainly composed of bacterial enzymes that catalyze β-(1→3)-glucan and β-(1→3)/β-(1→4)-glucan [6]. Laminarin is a natural β-(1→3)-glucan with occasional β-(1→6)-glucosyl branches found in marine micro- and macroalgae. Bacteria degrade and metabolize laminarin as a source of glucose [7–12]. The successive hydrolysis of laminarin by GH16 enzymes and GH3 enzymes, a family containing various kinds of glycosidases, has been reported [13–18]. On the other hand, the GH17 family is mainly composed

**Citation:** Kumagai, Y.; Kishimura, H.; Lang, W.; Tagami, T.; Okuyama, M.; Kimura, A. Characterization of an Unknown Region Linked to the Glycoside Hydrolase Family 17 β-1,3-Glucanase of *Vibrio vulnificus* Reveals a Novel Glucan-Binding Domain. *Mar. Drugs* **2022**, *20*, 250. https://doi.org/10.3390/ md20040250

Academic Editor: Hitoshi Sashiwa

Received: 25 February 2022 Accepted: 29 March 2022 Published: 31 March 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

of plant and fungal enzymes. They are classified as pathogen-related proteins [19–21] that contribute to the degradation and biosynthesis of the cell wall. Recently, many GH17 bacterial enzymes have been discovered due to the progress made in sequencing technology; however, the biological functions of bacterial enzymes are still unclear. Studies have shown that proteobacterial species produce an antibiotic biofilm via GH17 glucosyltransferase activity [22,23]. Another study reported that glucosyltransferase activity was modulated to a glucanase activity by a single mutation [24].

The CAZy database provides the taxonomic distribution of the GH17 family (cazy. org/IMG/krona/GH17\_krona.html, accessed on 23 February 2022), revealing that a large number of bacterial enzymes are found within the phylum Proteobacteria (recently renamed Pseudomonadota). Within Proteobacteria, GH17 enzymes are commonly found within the genus *Pseudomonas*, with its diverse members and metabolism. While many *Vibrio* species, which belong to the class Gammaproteobacteria, have GH16 enzymes, a limited number possess enzymes of the GH17 family. The genome of *V. vulnificus* has been sequenced and annotated, and sequence analysis has revealed that one GH16 enzyme (VvGH16) and one GH17 enzyme (VvGH17) exist adjacently in the genome. One GH3 enzyme is located close to the other two enzymes. On the other hand, three GH16 (VbGH16A, VbGH16B, and VbGH16C) and one GH17 (VbGH17A) enzyme of *Vibrio breoganii* 1C10 have been characterized [25].

Several CAZymes have various domains in addition to the catalytic domain, including carbohydrate-binding modules (CBMs) [26–28]. These domains are involved in carbohydrate binding. Tryptophan is an important amino acid residue in carbohydrate binding [26]. The VvGH17 C-terminus has several tryptophan residues. Therefore, we predicted that this region may have functions, such as carbohydrate binding, that can increase the catalytic efficiency or specificity. In this study, we characterized GH17 β-1,3-glucanase of *V. vulnificus* to clarify the unknown region of the protein and found that the N- and C- terminal regions were affiliated with the assembly of monomeric subunits into the multimeric form and the affinity for the substrate, respectively.

#### **2. Results**

#### *2.1. Bioinformatic Analysis of VvGH17*

VvGH17 is composed of 615 amino acids (AAs) comprising 1–22 AAs as signal peptides, 23–86 AAs as an unknown N-terminal region (Uk-N), 87–415 AAs as the GH17 domain, and 416–615 AAs as an unknown C-terminal region (Uk-C) (Figure 1a). Secondary structure prediction showed that Uk-N had a random coil structure, while Uk-C was composed of a β-sheet structure. The structure of VvGH17 was predicted using AlphaFold2 [29] (Figure 1b). The GH17 domain and C-terminus of Uk-C were predicted to have a (β/α)8 barrel structure and a β-sandwich structure, respectively. It was expected that the Uk-C structure possessed some function. Therefore, we attempted to characterize the impact of Uk-N and Uk-C on the catabolic properties of the enzymes.

**Figure 1.** Sequence and predicted structure of VvGH17. (**a**) The scheme of VvGH17. SP—signal peptide; Uk-N—unknown N-terminal region; GH17—catalytic domain of GH17; Uk-C—unknown C-terminal region. Characters highlighted in yellow and those highlighted in black in the sequences are predicted α-fold and β-sheet structures, respectively. (**b**) Predicted three-dimensional structure. Colors are related to the truncated mutation region in this study.

#### *2.2. Biochemical Properties of VvGH17*

VvGH17 was produced in an *Escherichia coli* expression system and isolated using a TALON affinity resin (Takara Bio, Otsu, Japan). The purified enzyme (20 mg) was obtained from one liter of medium and showed a single band of approximately 63 kDa on SDS-PAGE. The optimal temperature and pH were 50 ◦C and 5.0–6.5, respectively (Figure 2a,b). The temperature required for the half inactivation of the hydrolysis activity of VvGH17 at 30 min was 47 ◦C (Figure 2c). The activity of VvGH17 decreased by approximately 80% in the presence of 0.5 M NaCl and retained the activity up to 4.0 M (Figure 2d). The specific activity of VvGH17 in optimal conditions was 65.5 U/mg. VvGH17 hydrolyzed curdlan (insoluble β-(1→3)-glucan) and laminarin, mainly producing laminaribiose, glucose, and laminaritriose; it did not hydrolyze barley β-glucan, which as β-(1→3)- and β-(1→4) linkages (Figure 2e). The products of laminarin hydrolysis by VvGH17 were monitored from 0 to 60 min using gel filtration; the results revealed that VvGH17 hydrolyzed laminarin via an endolytic mechanism (Figure 2f).

**Figure 2.** Characterization of VvGH17. (**a**) The effect of temperature on VvGH17 activity. The enzyme reaction was conducted in a mixture containing 50 mM MES buffer (pH 6.0), 1% (*w/v*) laminarin, and 0.02 mg/mL VvGH17 at 20–70 ◦C for 10 min. (**b**) The effect of pH on VvGH17 activity. The enzyme reaction was conducted in a mixture containing 1% (*w/v*) laminarin, 0.02 mg/mL VvGH17, and 100 mM Britton–Robinson buffer (pH 4.0–10.0) at 45 ◦C for 10 min. (**c**) The effect of temperature on VvGH17 stability. A mixture containing 50 mM MES buffer (pH 6.0) and 0.2 mg/mL VvGH17 was incubated at the indicated temperature for 30 min and placed on ice for 10 min. Then, the enzyme activity was assayed in a mixture containing 50 mM MES buffer (pH 6.0), 1% (*w/v*) laminarin, and 0.02 mg/mL VvGH17 at 45 ◦C for 10 min. (**d**) The effect of NaCl on VvGH17 activity. The enzyme reaction was conducted in a mixture containing 50 mM MES buffer (pH 6.0), 1% (*w/v*) laminarin, 0.02 mg/mL VvGH17, and 0–4.0 M NaCl at 45 ◦C for 10 min. (**e**) Thin layer chromatography (TLC) analyses of the hydrolysis products obtained using VvGH17. One microliter of each reaction mixture was applied for TLC analysis. Mk—marker of glucose and laminaripentaose; '−'—without VvGH17; '+'—with VvGH17. -<sup>1</sup> —curdlan; -<sup>2</sup> —laminarin; -3 —β-glucan. (**f**) Gel filtration chromatography analysis for the hydrolysis of laminarin by VvGH17.

#### *2.3. C-terminal-Truncated Mutant*

C-terminal-truncated mutants of VvGH17 were constructed for the characterization of Uk-C. The position from 87–415 AA was demonstrated as the conserved domain of GH17. Therefore, Uk-C was defined as the position from 416–615 AA in VvGH17, and the three C-terminal-truncated mutants (ΔC466, ΔC441, and ΔC415) were constructed (Figure 3a). Recombinant proteins of the three mutants were successfully expressed, and we evaluated the enzyme kinetics (Figure 3b, Table 1). The enzyme kinetics *k*cat/*K*m of the WT toward laminarin was 93.0 mM−<sup>1</sup> s−1, and 32.8 and 30.5 mM−<sup>1</sup> s−<sup>1</sup> for ΔC466 and ΔC441, respectively, which are approximately one-third of the *k*cat/*K*<sup>m</sup> values in the WT. The *k*cat/*K*<sup>m</sup> values of ΔC415 were less than 2% of those in the WT. To confirm whether the loss of activity in ΔC415 was derived from folding, the secondary structure was compared using circular dichroism (CD) spectroscopy (Figure 3c). The difference in the CD spectrum (deg cm2 dmol<sup>−</sup>1) between 210 and 230 nm may be a result of the deletion of the C-terminus in VvGH17. From these results, the truncation of the C-terminus in Uk-C (AA 442–615) resulted in a decrease in the catalytic efficiency of VvGH17, and the truncation of the whole Uk-C (416–615 AA) caused the loss of the majority of its activity, suggesting that this is an essential region.

**Figure 3.** C-terminal-truncated mutants of VvGH17. (**a**) Scheme of C-terminal-truncated mutants. (**b**) Enzyme kinetics using laminarin as a substrate. The *k*cat/*K*<sup>m</sup> value of the WT (93.0 mM−<sup>1</sup> s−1) was set at 100%, and the relative values of the other mutants are indicated in the figure. (**c**) CD spectra of the WT and mutants. WT—black line; ΔC466—blue line; ΔC441—red line; ΔC415—pink line.


**Table 1.** Enzyme kinetics of VvGH17 and mutants using laminarin as a substrate.

#### *2.4. Affinity Gel Analysis of C-Terminal-Truncated Mutants of VvGH17*

The truncation of the VvGH17 C-terminus revealed that this region affects the enzyme kinetics (*k*cat and *K*m). This indicates that Uk-C has the potential for carbohydrate binding. To investigate the Uk-C function further, affinity gel analysis was performed (Figure 4). The WT and three C-terminal-truncated mutants showed two bands with and without substrates. The two bands were confirmed as monomers and oligomers (trimer) of the enzyme, as discussed in Section 2.7. Bovine serum albumin (BSA) was used as a marker protein for the mobility shift assay. No affinity toward curdlan was found in the tested enzymes, compared to the gel without substrate. The mobility of the WT monomer was clearly shifted from below the BSA band (without substrate) to upper the BSA band in the gels, confirming its affinity toward laminarin and β-glucan.

**Figure 4.** Affinity gel analysis of the WT and C-terminal-truncated mutants of VvGH17. (**a**) Affinity gel without substrate; (**b**) gel containing curdlan; (**c**) gel containing laminarin; (**d**) gel containing barley β-glucan. Asterisks show the WT monomer bands. The dashed lines show the mobility of BSA.

#### *2.5. Affinity Gel Analysis of Uk-C and Point Mutants of VvGH17*

The affinity of Uk-C and the C-terminus of VvGH17 toward laminarin and β-glucan was revealed. To confirm the important amino acids for substrate binding, point mutants of Uk-C and VvGH17 were constructed. Affinity toward substrates was also evaluated by the mobility as compared with BSA (Figure 5). The mobility of Uk-C in the gel containing laminarin and β-glucan was decreased compared to the gel without substrate. This indicated that Uk-C had a binding ability for laminarin and β-glucan. Tryptophan is an important amino acid for carbohydrate binding [26]. Therefore, we mutated four tryptophans in Uk-C to alanines (W472A, W499A, W542A, and W567A), and the affinity was evaluated by mobility shift assays. The WT and four mutants showed the same mobilities without a substrate. The mobilities of W472A, W499A, and W542A in the gel containing laminarin and β-glucan differed from those of the WT and W567A. The decreased mobility of W472A, W499A, and W542A indicated reduced binding ability, suggesting that the three tryptophans are essential amino acids for substrate binding.

**Figure 5.** Affinity gel analysis of the WT, point mutants of VvGH17, and Uk-C. (**a**) Affinity gel without substrate; (**b**) gel containing laminarin; (**c**) gel containing barley β-glucan. Asterisks indicate the monomer bands showing different mobility. The dashed lines show the mobility of BSA.

#### *2.6. Prediction of Uk-C Structure and Function*

We attempted to clarify the relationship between the predicted three-dimensional structure of Uk-C (AA 441–615 of VvGH17) and the binding ability of the mutants. The predicted Uk-C had two domains: the N-terminus of Uk-C (AA 416–462 of VvGH17) was predicted to be a linker between GH17 and the binding region, and the C-terminus of Uk-C (AA 463–615 of VvGH17) was predicted to be a β-sandwich structure with a possible carbohydrate-binding ability (Figure 1b). Three tryptophan residues (Trp472, Trp499, and Trp542) were located in the putative substrate binding region in the β-sandwich structure (Figure 6). On the other hand, the predicted structure suggested that Trp567 was located outside of the putative substrate binding region. The results of the mutation experiments agreed with the predicted structure.

**Figure 6.** Structural prediction of the C-terminus of VvGH17. AA 441-615 of VvGH17 as predicted by AlphaFold2. Colors from blue to red show the sequence from AA 441 to 615. The locations of the four tryptophans (Trp472, Trp499, Trp542, and Trp567) are indicated.

#### *2.7. N-Terminal-Truncated Mutants of VvGH17*

The catalytic domain of VvGH17 was mapped from AA 87 to 415, and AA 1–22 of VvGH17 were predicted as a signal peptide. The function of the N-terminus in VvGH17 (AA 23–86) was unclear. Therefore, we constructed two N-terminal-truncated mutants (ΔN50 and ΔN65) and evaluated their activity (Figure 7). The *k*cat/*K*<sup>m</sup> values of ΔN50 and ΔN65 toward laminarin were 67.5 and 24.7 mM−<sup>1</sup> s−1, respectively. The predicted three-dimensional structure of VvGH17 showed that the region of AA 51–65 was composed of the bottom of the (β/α)8 barrel structure (Figure 1b). Consequently, the loss of this region in the ΔN65 mutant led to structural instability, resulting in decreased catalytic efficiency (*K*cat/*k*m) (Table 1).

**Figure 7.** N-terminal-truncated mutants of VvGH17. (**a**) Scheme of N-terminal-truncated mutants. (**b**) Enzyme kinetics using laminarin as substrate. The *k*cat/*K*<sup>m</sup> value of the WT (93.0 mM−<sup>1</sup> s−1) was set at 100% and the relative values of the other mutants are indicated in the figure. (**c**) Blue native PAGE of the WT and N- and C-terminal-truncated mutants. \*—WT monomer; \*\*—putative WT trimer from molecular weight.

Two bands from the WT and C-terminal-truncated mutants were observed in native PAGE, as shown in Section 2.4. To confirm the assembly of monomeric subunits into the multimeric form, blue native PAGE was performed for the WT and N- and C-terminaltruncated mutants (Figure 7c). The WT and ΔC466 mutant clearly showed two bands (monomers and putative trimers from the molecular mass), while ΔN50 and ΔN65 showed a single band corresponding to the monomer.

#### **3. Discussion**

In this study, we characterized the GH17 enzymes of *V. vulnificus* using an *E. coli* expression system. Functionally, GH17 enzymes have been reported to be β-(1→3)-glucan hydrolases and transglycosylases. In particular, endo-type β-(1→3)-glucanases are classified into the enzyme commission (EC) number EC 3.2.1.6 endo-1,3(4)-β-glucanase; EC 3.2.1.39 represents glucan endo-1,3-β-D-glucosidase; and EC 3.2.1.73 indicates licheninase. Our results showed that VvGH17 was classified into EC 3.2.1.39 due to its endolytic mechanism and specificity for β-(1→3)-glucan.

VvGH17 produced laminaribiose as the main product regardless of soluble and insoluble β-(1→3)-glucans, showing its potential for oligosaccharide production. The amino acid identity between VvGH17 and VbGH17A from *V. breoganii* 1C10 was low (42% identity). VbGH17A contains a signal peptide, the GH17 catalytic domain, and an unknown region of the C-terminus from amino acid (AA) 411 to 634 [18]. The catalytic domains of VvGH17 and VbGH17A shared 56% identity; however, the identity between the C-termini was low (21%). The hydrolysis products of VbGH17A were oligosaccharides, which were larger than a degree of polymerization (DP) of 4. *V. breoganii* 1C10 has four endo-type β-(1→3)-glucanases, which presumably show synergic activity during the hydrolysis of β-(1→3)-glucans. On the other hand, *V. vulnificus* has two endo-type β-(1→3)-glucanases. We expected several enzymes; however, there was only GH16, which had 57% identity with VbGH16A and produced mainly DP 3 and 4. The GH3 enzymes of *Vibrio* sp. have been shown to have activity toward laminaribiose [30]. Therefore, *V. vulnificus* may metabolize β-(1→3)-glucan by producing small DP oligosaccharides using two endo-type enzymes and then hydrolyzing them using the GH3 enzyme.

In this study, we identified the Uk-C as a carbohydrate binding-domain. VvGH17 formed a trimer, and the complex structure also showed carbohydrate binding activity. It can be concluded that the N-terminal region is affiliated with the trimerization of VvGH17. A more detailed structure-function analysis is needed. Uk-C showed binding activity with β-(1→3)-glucan and β-(1→3)/β-(1→4)-glucan and no binding activity with curdlan, an insoluble triple helix β-(1→3)-glucan. The TLC results indicated that VvGH17 hydrolyzed curdlan as well as laminarin without the assistance of Uk-C (Figure 2e). This study investigated the affinity of Uk-C for insoluble curdlan, but not the affinity for curdlan gel. Therefore, the binding specificity of Uk-C for a linear β-(1→3)-glucan should be further investigated. Curdlan is produced by the soil bacterium *Agrobacterium* sp., and *V. vulnificus* is a marine bacterium. Therefore, Uk-C might be specific for the marine polysaccharide laminarin. Laminarin is a soluble β-(1→3)-glucan with a β-(1→6)-glycosyl side chain. The difference in polysaccharide structure could affect the binding specificity. CBMs are classified into three types by their ligand binding sites. A-type CBMs recognize crystalline polysaccharide-like cellulose and chitin. B-type CBMs recognize a single glycan by binding a cleft or groove. C-type CBMs recognize the glycan terminus by binding pockets [26]. The Uk-C structure was predicted to be a B-type groove with three tryptophan residues. We confirmed that the mutations constituted the groove. Therefore, the mutation of tryptophan residues decreased the glucan binding ability. A BLAST search of Uk-C (416–615 AA) showed identity with CBM domains linked to other GH17 enzymes. A high AA identity of more than 95% was shared with the CBM domains of the GH17 enzymes from *Vibrio* sp., including *V. fluvialis*, *V. cholerae*, *V. metoecus,* and *V. metschnikovii*. A total of 30–70% identities were shared with enzymes from bacterial species, such as *Enterovibrio* sp., *Porticoccaceae* sp., *Bacteroidetes* sp., and *Grimontia* sp. (Table S1). The CBMs

were classified into 89 families in the CAZy database (accessed on 16 February 2022). Among them, an affinity for β-(1→3)-glucan or β-(1→3)/β-(1→4)-glucan was demonstrated by 18 families: CBM4, 6, 11, 22, 28, 39, 43, 52, 54, 56, 65, 72, 76, 78, 79, 80, 81, and 85. CBM43, linked to eukaryotic GH17 enzymes, generally consists of 90-100 AAs [31]. The C-terminus of VvGH17 consisted of 150 AAs. Therefore, the sequences of CBMs belonging to nine families with around 150 AA residues were selected and aligned using ClustalW (https://www.genome.jp/tools-bin/clustalw, accessed on 23 February 2022) after removing the His-tag sequence, and we visualized the tree using iTOL [32] (Figure 8). The C-terminus of VvGH17 showed the closest relationship to the CBM79 cluster. The AA identity between Uk-C and CBM79 was 15.6-17.2% (Figure S1, Supplementary Materials). The complete genome sequence of *V. vulnificus* was determined in 2011 [33], and CBM79 family was recorded in 2016 [34]; however, Uk-C has not been included. In addition, the Uk-C sequence was not hit by a BLASTP search using CBM79 as query sequence. The distance of VvGH17 and CBM79 was similar to other CBM clusters (Figure 8), indicating that Uk-C is a novel soluble β-(1→3)-glucan and β-(1→3)/β-(1→4)-glucan-binding protein.

**Figure 8.** Phylogenetic tree of VvGH17 C-terminal domain and relatives. The tree was constructed based on ClustalW pairwise sequence alignment using the iTOL visualizing software. The following amino acid sequences were used: CBM4, laminarinase 16A from *Thermotoga maritima*, accession number (AN)—AAD35118, protein data bank (PDB) —1GUI [35]; CBM6, β-1,3-glucanase from *Alkalihalobacillus halodurans* C-125, AN—BAB03955, PDB—1W9T [36]; CBM6, endo-β-1,3-glucanase from *Zobellia galactanivorans*, AN—CAZ95067, PDB—5FUI [36]; CBM22, xylanase Xyn10B from *Acetivibrio thermocellus* YS, AN—CAA58242, PDB—1DYO [37]; CBM65, endoglucanase (EcCel5A) from *Eubacterium cellulosolvens* 5, AN—BAE46390, PDB—2YPJ [38]; CBM72, endoglucanase from uncultured microorganism, AN—EU449484 [39]; CBM76, GH44 from *Ruminococcus flavefaciens*, AN— AAA95959 [34]; CBM78, GH5 from *R. flavefaciens*, AN—WP\_009983134, PDB—4V17 [34]; CBM79, GH9 from *R. flavefaciens*, AN—WP\_009984389 [34]; CBM85, GH10 xylanase from metagenomic data, AN—MH727997 [27]; and VvGH17 from 451–615 AA. His-tag sequences from CBM6\_5FUI, CBM65\_2YPJ, and CBM78\_4V17 were removed. The alignment of the tree and VvGH17 and CBM79 are shown in Figure S1 and Figure S2, respectively.

#### **4. Materials and Methods**

#### *4.1. Materials*

Curdlan was purchased from Fujifilm Wako Pure Chemicals Industries Ltd. (Osaka, Japan); laminarin (*Laminaria digitata*) was from Sigma-Aldrich Corp. (St. Louis, MO, USA); and β-glucan (barley; medium viscosity) was from Megazyme International Ireland Ltd. (Bray, Ireland). Laminaripentaose was prepared by the hydrolysis of curdlan with KfGH64 [40]. All the other reagents were purchased from Wako Pure Chemical Industries (Osaka, Japan).

#### *4.2. Bioinformatic Analysis of VvGH17*

The GH17 gene from *Vibrio vulnificus* (hypothetical protein AOT11\_01225) was obtained from GenBank (accession no. ASM98089.1). The putative conserved domain was

searched using BLASTP [41]. The signal peptide was predicted using the SignalP 4.1 server [42]. Secondary structure prediction was performed using the PSIPRED server [43] and the structure was predicted by AlphaFold2 [29]. Homolog proteins of the C-terminus of VvGH17 (416–615 AA) were searched by BLASTP with the standard algorithm, excluding *Vibrio vulnificus* (taxid: 672); we also removed query covers of less than 40%. A phylogenetic tree was constructed by pairwise sequence alignment.

#### *4.3. Construction, Expression, and Purification of VvGH17*

The expression plasmid of the gene putatively encoding β-(1→3)-glucanase (VvGH17) was constructed as follows: a codon-optimized mature *Vvgh17* gene was synthesized (Eurofins Genomics) for expression in *E. coli* harboring *Nde*I and *Hin*dIII sites at 5 and 3 , respectively. Then, the *Vvgh17* gene was cloned into the *Nde*I-*Hin*dIII site of pET28a to construct an expression vector of pET28a(VvGH17). The recombinant protein was produced in *E. coli* BL21-RIL (DE3) cells (Agilent Technologies, Palo Alto, CA, USA) harboring pET28a(VvGH17) and was purified as previously described [44]. The protein concentrations were determined by absorbance at 280 nm using the molar extinction coefficients for VvGH17 [45].

#### *4.4. Construction of VvGH17 Mutants*

The C-terminal-truncated mutants (ΔC466, ΔC441, and ΔC415), N-terminal-truncated mutants (ΔN50, ΔN65, and UK-C), and point mutants (W472A, W499A, W542A, and W567A) were constructed by polymerase chain reaction using PrimeSTAR MAX DNA polymerase (Takara Bio, Otsu, Japan), primers (Table 2), and pET28a(VvGH17) as a template.


**Table 2.** Sequences of the primers used in this study.

Small characters show amino acid mutations from tryptophan to alanine.

#### *4.5. VvGH17 Standard Activity Assay*

VvGH17 activity was determined at 45 ◦C for 10 min with an appropriate amount of enzyme, 1% (*w/v*) laminarin, and 50 mM 2-morpholinoethanesulfonic acid (MES; pH 6.0). The amount of reducing sugars was determined using the dinitrosalicylic acid method (DNS) [46]. One unit of activity was defined as the amount of enzyme that liberated reducing sugars equivalent to 1.0 μmol glucose per minute. The optimal temperature of VvGH17 was measured as follows: a reaction mixture containing 1% (*w/v*) laminarin and 50 mM MES (pH 6.0) was incubated at 22–70 ◦C for 10 min. The optimal pH of VvGH17 was measured as follows: a reaction mixture containing 1% (*w/v*) laminarin and 100 mM Britton– Robinson buffer (a mixture containing sodium acetate buffer, sodium phosphate buffer, and

glycine–NaOH buffer; pH 4.0–10.0) was incubated at 45 ◦C for 10 min. Temperature stability was determined by measuring the residual activity after incubation in 50 mM MES (pH 6.0) at 30–57 ◦C for 30 min. The effect of NaCl was determined using a mixture containing 1% (*w/v*) laminarin, 50 mM MES (pH 6.0), and 0–4.0 M NaCl at 45 ◦C for 10 min. The *V*max and *K*<sup>m</sup> with laminarin (0.5–40 mg/mL) were determined by the standard Michaelis–Menten equation using nonlinear regression (Origin Software, Lightstone Corp., Tokyo, Japan). All the activity assays were performed in triplicate.

#### *4.6. Analysis of Hydrolysis Products by TLC and Gel Filtration Chromatography*

The products of VvGH17 hydrolysis were analyzed by TLC using a silica gel 60 plate (Merck). The substrates (curdlan, laminarin, and β-glucan: 10 mg/mL) were hydrolyzed with 0.01 U/mL of VvGH17 for 24 h, and the reaction was terminated by heating at 100 ◦C for 10 min. The hydrolysis products (1 μL) were developed in ethyl acetate, acetic acid, and water (2:2:1, *v/v/v*); sugars were detected by spraying a solution of 10% (*v/v*) sulfuric acid in ethanol and then heating at 100 ◦C for 10 min.

The distribution of the hydrolysis products of laminarin was analyzed using highperformance liquid chromatography (HPLC) with a Superdex Peptide 10/300 GL column (GE Healthcare UK Ltd., Little Chalfont, UK) and a Corona Charged Aerosol Detector (Thermo Scientific Inc., Chelmsford, MA, USA). Laminarin (10 mg/mL) was hydrolyzed with 0.01 U/mL of VvGH17 from 0 to 60 min, and the reaction was terminated by heating at 100 ◦C for 10 min. The samples were eluted using water with a flow rate of 0.3 mL/min.

#### *4.7. CD Spectroscopy*

The secondary structures of VvGH17 were determined by CD spectroscopy using a J-720WI spectrometer (Jasco Corp. Tokyo, Japan). The proteins were dissolved at a final concentration of 0.1 mg/mL in 50 mM MES buffer (pH 6.0). The spectra were acquired at 37 ◦C using a 0.2 cm cuvette. The molar ellipticities (per residue) were calculated using the equation [*θ*] = 100(*θ*)/(*lcN*), where [*θ*] is the molar ellipticity per residue, (*θ*) is the observed ellipticity in degrees, *l* is the optical path length in centimeters, *c* is the molar concentration of the protein, and *N* is the number of residues in the protein.

#### *4.8. Polyacrylamide Gel Electrophoresis (PAGE) Analysis*

The assays for the binding activity of the proteins were performed by affinity gel electrophoresis, according to the procedure described by Zhang et al. [47]. A stacking gel containing 3 wt% polyacrylamide in 1.5 M Tris-HCl buffer (pH 8.3), a native gel with 12 wt% polyacrylamide containing 0.1 wt% polysaccharides (curdlan, laminarin, and barley β-glucan), and a control gel without polysaccharides were prepared. Each protein (1 μg) was loaded onto the gel, and the gels were electrophoresed at 4 ◦C and 100 V for 3 h. The gels were then stained with Coomassie brilliant blue G-250 for protein visualization.

Blue native PAGE was performed using a 5–10% gradient gel at 4 ◦C and 150 V held constant for 3.5 h using an anode buffer (50 mM tricine, 15 mM bis-Tris/HCl, pH 7) and cathode buffer (50 mM tricine, 15 mM bis-Tris/HCl, pH 7, 0.02% (*w/v*) Coomassie blue G250).

#### **5. Conclusions**

In this study, we characterized the unknown domains of the GH17 β-(1→3)-glucanase of *V. vulnificus*. The WT formed a trimer, but the N-terminal truncations formed monomers. Therefore, the N-terminus contributes to the assembly of monomeric subunits into the multimeric form of VvGH17. The C-terminal region showed an affinity for β-(1→3)-glucan and β-(1→3)/β-(1→4)-glucan. The C-terminus was predicted to have a β-sandwich structure, and three tryptophan residues (Trp472, Trp499, and Trp542) were located at the substrate binding site using mutational analysis. A BLAST search revealed that the C-terminal region of GH17 was conserved among Gammaproteobacteria. The results of this study are useful for understanding bacterial GH17 enzymes and oligosaccharide preparation.

**Supplementary Materials:** The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/md20040250/s1, Table S1: BLAST search result for Uk-C; Figure S1: AA alignment of Uk-C and CBM79; Figure S2: AA alignment of Figure 8.

**Author Contributions:** Conceptualization, Y.K.; methodology, Y.K., H.K., W.L., T.T., M.O. and A.K.; writing—original draft preparation, Y.K.; writing—review and editing, W.L., T.T. and M.O.; supervision, H.K. and A.K.; funding acquisition, Y.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Japan Society for the Promotion of Science KAKENHI Grant No. 16K18748.

**Conflicts of Interest:** The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

#### **References**


#### *Article* **Identification and Characterization of Three Chitinases with Potential in Direct Conversion of Crystalline Chitin into** *N***,***N-* **-diacetylchitobiose**

**Xue-Bing Ren 1, Yan-Ru Dang 1, Sha-Sha Liu 1, Ke-Xuan Huang 1, Qi-Long Qin 1,2, Xiu-Lan Chen 1,3, Yu-Zhong Zhang 1,2,3, Yan-Jun Wang 1,4,\* and Ping-Yi Li 1,\***


**Abstract:** Chitooligosaccharides (COSs) have been widely used in agriculture, medicine, cosmetics, and foods, which are commonly prepared from chitin with chitinases. So far, while most COSs are prepared from colloidal chitin, chitinases used in preparing COSs directly from natural crystalline chitin are less reported. Here, we characterize three chitinases, which were identified from the marine bacterium *Pseudoalteromonas flavipulchra* DSM 14401T, with an ability to degrade crystalline chitin into (GlcNAc)2 (*N,N*'-diacetylchitobiose). Strain DSM 14401 can degrade the crystalline α-chitin in the medium to provide nutrients for growth. Genome and secretome analyses indicate that this strain secretes six chitinolytic enzymes, among which chitinases Chia4287, Chib0431, and Chib0434 have higher abundance than the others, suggesting their importance in crystalline α-chitin degradation. These three chitinases were heterologously expressed, purified, and characterized. They are all active on crystalline α-chitin, with temperature optima of 45–50 ◦C and pH optima of 7.0–7.5. They are all stable at 40 ◦C and in the pH range of 5.0–11.0. Moreover, they all have excellent salt tolerance, retaining more than 92% activity after incubation in 5 M NaCl for 10 h at 4 ◦C. When acting on crystalline α-chitin, the main products of the three chitinases are all (GlcNAc)2, which suggests that chitinases Chia4287, Chib0431, and Chib0434 likely have potential in direct conversion of crystalline chitin into (GlcNAc)2.

**Keywords:** chitinases; crystalline chitin; chitooligosaccharides; *N*,*N* -diacetylchitobiose; *Pseudoalteromonas*

#### **1. Introduction**

Chitin is a polymer of *N*-acetyl-D-glucosamine (GlcNAc) and is the second most abundant polysaccharide after cellulose in nature. Chitin is mainly present in arthropod exoskeletons, fungal cell walls, and insect cuticles in a crystalline form, which is intractable, highly hydrophobic, and insoluble in water [1]. Chitin has three polymorphic isomers, including α-chitin, β-chitin, and γ-chitin. Among them, α-chitin is the most common form found in fungi, insect exoskeletons, and shells of crustaceans. α-chitin is harder to degrade than β-chitin and γ-chitin as it has a higher degree of recalcitrance, which decreases the accessibility of the individual polymer chains [2]. Colloidal chitin is normally prepared by treating natural chitin with strong acids to break the crystal structure and increase the

**Citation:** Ren, X.-B.; Dang, Y.-R.; Liu, S.-S.; Huang, K.-X.; Qin, Q.-L.; Chen, X.-L.; Zhang, Y.-Z.; Wang, Y.-J.; Li, P.-Y. Identification and Characterization of Three Chitinases with Potential in Direct Conversion of Crystalline Chitin into *N*,*N* -diacetylchitobiose. *Mar. Drugs* **2022**, *20*, 165. https://doi.org/ 10.3390/md20030165

Academic Editors: Yuya Kumagai, Hideki Kishimura and Benwei Zhu

Received: 12 January 2022 Accepted: 23 February 2022 Published: 24 February 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

accessibility of the substrate to enzymes. Therefore, colloidal chitin is usually used as the substrate for chitinase characterization.

The annual production of chitin in the ocean exceeds billions of tons [3,4], which is a good source for the production of chitooligosaccharides (COSs) and GlcNAc. Due to their various bioactive activities, COSs and GlcNAc have been widely applied in agriculture, medicine, cosmetics, and foods. For example, COSs have protective effects against infections and enhanced antitumor properties [5,6]. GlcNAc and (GlcNAc)2 (*N*,*N* diacetylchitobiose) can serve as cosmetic ingredients, dietary supplements, and osteoarthritis therapeutics [7–9].

The chitinolytic enzymes contain chitinases (EC 3.2.1.14), mainly from the GH18 and GH19 families, and β-*N*-acetylglucosaminidases (EC 3.2.1.52), mainly from the GH20 and GH3 families. While several β-*N*-acetylglucosaminidases have been reported to be active on chitin [10–12], the hydrolysis of chitin into COSs and/or GlcNAc is predominantly catalyzed by chitinases [13]. Chitinases include endochitinases and exochitinases, which are widely produced by bacteria [14], fungi [15], and plants [16], playing key roles in natural chitin degradation and recycling. Many bacteria-derived chitinases have been characterized, predominantly with colloidal chitin or chitooligosaccharides (or synthetic chitooligosaccharide analogs) as the substrate. Reported bacterial chitinases are mostly mesophilic enzymes with optimal temperatures at 40–60 ◦C [17–25]; only a few have been found to be cold-active enzymes, such as CHI II of *Glaciozyma antarctica* PI12 (15 ◦C) [26] and ChiA of *Pseudoalteromonas* sp. DL-6 (20 ◦C) [27]. The pH optima of bacterial chitinases are over a wide range. For example, chitinases from *Streptomyces chilikensis* RC1830 [24], *Pseudoalteromonas tunicata* CCUG 44952T [25], and *Bacillus* sp. R2 [21] showed their highest activity at neutral pHs (7.0–7.5), those from *Micrococcus* sp. AG84 [22], *Pseudoalteromonas* sp. DC14 [23], and *Citrobacter freundii* haritD11 [28] at basic pHs (8.0–9.0), and those from *Moritella marina* ATCC 15381 [29] and *Paenicibacillus barengoltzii* CAU904 [17] at acidic pHs (3.5 and 5.0, respectively). Chitinases are good tools to prepare COSs and GlcNAc from chitin. Because the natural source of chitin is crystalline chitin, chitinases that can efficiently hydrolyze crystalline chitin have better application potential in preparing COSs and Glc-NAc from natural chitin sources than those only active on colloidal chitin. However, so far, only a few crude enzymes produced by wild strains and recombinant chitinases have been reported to be used in preparing COSs and GlcNAc from crystalline chitin [27,30–33]. Thus, it is necessary to identify and characterize more chitinases that can efficiently hydrolyze crystalline chitin for preparing COSs and GlcNAc from natural chitin sources.

Bacteria of the genus *Pseudoalteromonas* are widely distributed in the ocean, accounting for 2–3% of total bacterial abundance in upper ocean waters [34,35]. Many strains in this genus contain multiple chitinase-encoding genes [36], and some have been reported to secrete chitinases [18,23,25,27,37,38]. Furthermore, some chitinases from *Pseudoalteromonas* have been characterized. The GH18 chitinase Chi23, from *Pseudoalteromonas aurantia* DSM6057, is a thermostable enzyme with activity towards crystalline chitin in acidic conditions (pH 3.0–6.0) [18]. The GH18 chitinases ChiA and ChiC from *Pseudoalteromonas* sp. DL-6 [27,37] and ChiB from *Pseudoalteromonas* sp. O-7 [38] are cold-active enzymes with temperature optima at 20–30 ◦C. The GH19 chitinase Ptchi19 from *Pseudoalteromonas tunicata* CCUG 44952<sup>T</sup> was active at 20–50 ◦C and pH 6.0–9.5 [25]. The chitinase purified from the fermentation broth of *Pseudoalteromonas* sp. DC14 exhibited halo-alkali and thermo-stable properties [23]. Despite these studies, *Pseudoalteromonas* chitinases with potential in preparing COSs/GlcNAc from natural crystalline chitin have rarely been reported. The aim of this study is to identify and characterize chitinases with activity on crystalline chitin from marine *Pseudoalteromonas* bacteria and to evaluate their potential in preparing COSs/GlcNAc from natural crystalline chitin. In this study, the ability of 26 *Pseudoalteromonas* type strains to use crystalline chitin as a carbon source for growth was investigated, and *Pseudoalteromonas flavipulchra* DSM 14401T (hereafter strain DSM 14401), which was isolated from surface seawater [39], was found to have the highest degradation rate on crystalline α-chitin. The extracellular chitinases secreted by strain DSM 14401

were further identified by genome and secretome analyses. Three chitinases with high abundance in the secretome were heterologously expressed in *Escherichia coli* BL21 (DE3) and biochemically characterized. The hydrolytic products released from crystalline chitin by these chitinases were further investigated. The results suggest that these chitinases likely have potential in the preparation of (GlcNAc)2 from natural crystalline chitin.

#### **2. Results and Discussion**

#### *2.1. The Ability of Strain DSM 14401 to Utilize Crystalline Chitin*

To obtain *Pseudoalteromonas* strains that can secrete chitinases to efficiently degrade crystalline chitin, 26 type *Pseudoalteromonas* strains (Table S1) were cultured in a liquid medium containing chitin flakes (crystalline α-chitin) as carbon source, and their growth and the degree of degradation of chitin flakes were observed. Strain DSM 14401 showed the greatest degradation rate of chitin flakes. This strain was able to degrade most of the chitin flakes in the medium in 5 days (Figure 1A). The growth curve and the extracellular chitinase activity of strain DSM 14401 during cultivation were also investigated (Figure 1B). The strain was cultured in a medium containing 0.05% peptone, 0.01% yeast powder, and 3% chitin flakes; the same medium without chitin flakes was used as a control. Strain DSM 14401 grew rapidly in the first 10 h in both media, with or without chitin flakes. After 10 h, the growth stagnated in both media, likely due to the depletion of the absorbable nutrients, such as peptone and yeast powder. After 40 h, while the cell number in the control medium began to decrease slowly and no extracellular chitinase activity was detected during the cultivation, both the cell number and the extracellular chitinase activity in the medium containing chitin flakes began to continuously increase until 68 h (Figure 1B). Based on this result, it can be speculated that, after absorbable nutrients were depleted, strain DSM 14401 began to secrete chitinases to degrade the chitin flakes in the medium into COSs/GlcNAc, which were absorbed by the strain to support its growth.

**Figure 1.** Growth and extracellular chitinase activity of *P. flavipulchra* DSM 14401<sup>T</sup> cultured on crystalline chitin. (**A**) Cultures of strain DSM 14401 at 25 ◦C for 0 and 5 days. (**B**) The growth curve (black line) and the extracellular chitinase activity (red line) of strain DSM 14401. Strain DSM 14401 was cultured in a minimal medium containing 0.05% peptone, 0.01% yeast powder, and 3% (*w*/*v*) chitin flakes at 25 ◦C and 180 rpm. The extracellular chitinase activity was measured with chitin powder as the substrate at 50 ◦C. Strain DSM 14401 cultured in the same medium without chitin flakes and under the same conditions was used as the control.

Paulsen et al. reported that 27 *Pseudoalteromonas* strains have the ability to degrade crystalline chitin [36]. Strain *Pseudoalteromonas* sp. DC14 was also reported to be able to degrade crystalline chitin [23]. In addition, 5 chitinases from *Pseudoalteromonas* strains have been expressed and characterized, including ChiA and ChiC from *Pseudoalteromonas* sp. DL-6 [27,37], ChiB from *Pseudoalteromonas* sp. O-7 [38], PtChi19p from *P. tunicata* CCUG 44952T [25] and Chi23 from *P. aurantia* DSM6057 [18]. Among them, chitinases ChiA, PtChi19p, and Chi23 have activity on crystalline chitin based on substrate specificity analysis [18,25,27]. These reports indicate that many *Pseudoalteromonas* strains can produce chitinases with activity on crystalline chitin. Consistently, strain DSM 14401 was most likely to secrete chitinases with activity on crystalline chitin due to its high degradation rate on crystalline α-chitin.

#### *2.2. Identification of the Chitinases Secreted by Strain DSM 14401*

To ascertain the chitinolytic enzymes secreted by strain DSM 14401, genomic analysis was carried out to find putative chitinolytic enzyme-encoding genes in strain DSM 14401. There are 11 genes encoding putative chitinolytic enzymes in strain DSM 14401, which were named Chia2822, Chib0431, Chib0434, Chia4287, Chib0889, Chib0721, Chia2290, Chia3704, Chib0633, Chib0719 and Chib0710. Chia2822, Chib0431, Chib0434, and Chia4287 are potential chitinases belonging to the GH18 family (Figure 2). Of these, Chib0431, Chib0434, and Chia4287 belong to the GH18A subfamily that mainly contains processive exochitinases [40–42], and Chia2822 belongs to the GH18B subfamily that mainly contains non-processive endochitinases [43,44]. Multiple sequence alignments suggest that all these GH18 chitinases of strain DSM 14401 contain a DxDxE catalytic motif (Figure S1), which is conserved in the GH18 chitinases [45]. Chitinase Chib0889 belongs to the GH19 family that mainly contains chitinases found in plants [46]. Two GH19 chitinases, LYS177 and LYS188, from *Pseudomonas* Ef1 have been reported to have lysozyme activity and they are clustered with phage/prophage endolysins based on the phylogenetic analysis [47]. However, the GH19 chitinase, Chib0889, of strain DSM 14401 was nested in the cluster of chitinases from Proteobacteria (Figure S2), implying that Chib0889 may function as a chitinase rather than a lysozyme. Chib0721, Chia2290, Chia3704, Chib0633, and Chib0719 from the GH20 family, and Chib0710 from the GH3 family are potential β-*N*-acetylglucosaminidases. The predicted domain architectures of these chitinolytic enzymes are shown in Figure 3. Except for Chib0710, the other chitinolytic enzymes all have a signal peptide predicted by SignalP 5.0, implying that they are likely secreted enzymes. Among these enzymes, Chib0633 and Chib0710 are single-domain enzymes, while the others are all multi-domain enzymes containing one or more carbohydrate-binding domains (Big\_7, CBM\_5\_12, and CHB\_HEX) in addition to their catalytic domains. The CBMs (carbohydrate-binding modules) in chitinases were reported to facilitate enzyme movement along a chitin chain during processive action and to stimulate the substrate to decrystallize [48–51].

Secretome analysis was further performed to identify the chitinolytic enzymes secreted by strain DSM 14401 cultured in the medium containing 3% chitin flakes as the sole carbon source. The extracellular proteins tightly absorbed on the chitin flakes were collected for secretome analysis when approximately half of the chitin flakes in the medium were degraded after 85 h. Finally, 6 of the putative chitinolytic enzymes were detected in the secretome. Of these, the 4 GH18 chitinases accounted for 97.50% of the abundance, and the GH19 and GH20 chitinolytic enzymes each accounted for 1.25% (Table 1), which suggests the importance of the GH18 chitinases in the degradation of crystalline chitin. Of the GH18 chitinases, Chia4287 was the most abundant (48.75%), followed by Chib0431 (25.00%), Chib0434 (15.00%) and Chia2822 (8.75%). The five putative β-*N*-acetylglucosaminidases with a predicted signal peptide were not found in the secretome, which may be secreted to the periplasm.

**Figure 2.** Phylogenetic analysis of chitinases Chib0431, Chib0434, Chia4287, and Chia2822 with other GH18 chitinases. The phylogenetic tree was constructed by the Neighbor-Joining method. Bootstrap analysis of 1000 replicates was conducted.


**Table 1.** The extracellular chitinolytic enzymes secreted by strain DSM 14401 identified by secretome analysis.

<sup>a</sup> Peptide-Spectrum Matches. <sup>b</sup> Abundance was calculated based on the proportion of the PSMs of a chitinolytic enzyme in the sum of PSMs of all chitinolytic enzymes in the secretome.

It has been reported that chitinolytic strains belonging to the genus *Pseudoalteromonas* usually have two GH18 chitinase genes in their chitin degradation clusters [36]. In addition, many *Pseudoalteromonas* species also contain one or more GH19 chitinase genes [36]. However, the removal of the GH19 chitinase gene from strain *Pseudoalteromonas rubra* S4059 had no significant influence on the growth of the strain on crystalline α-chitin [52], suggesting

that the GH19 chitinase is likely unimportant in the utilization of crystalline chitin. In contrast, the removal of the GH18 chitinase gene *chiD* from strain *Cellvibrio japonicus* Ueda107 made it unable to grow on crystalline α-chitin [53], indicating that the GH18 chitinase plays an important role in the crystalline chitin degradation of this strain. Moreover, it has been reported that (GlcNAc)2 and larger chitooligosaccharides can induce the expression of chitinases in *Vibrio furnissii* 7225 and *Vibrio cholerae* O1 [54]. For strain DSM 14401, although its genome contains a GH19 chitinase gene, a GH3 β-*N*-acetylglucosaminidase gene, and 5 GH20 β-*N*-acetylglucosaminidase genes in addition to 4 GH18 chitinase genes, secretome analysis showed that it mainly secreted the GH18 chitinases when crystalline α-chitin was present, which suggests that the GH18 chitinases likely play a main role in the degradation of crystalline α-chitin in this strain.

**Figure 3.** Domain architecture of the 11 chitinolytic enzymes of *P. flavipulchra* DSM 14401T. Protein sequences were analyzed on the HMMER website, and domains were illustrated by different colors based on their functional annotations. The Pfam IDs corresponding to the function annotations are as follows: Big\_7, bacterial Ig domain (PF17957); GH18, glycosyl hydrolases family 18 (PF00704); ChitinaseA\_N, ChitinaseA\_N-terminal domain (PF08329); CBM\_5\_12, carbohydrate-binding module (PF02839), ChiC, Chitinase C (PF06483); GH19, glycoside hydrolase family 19 (PF00182), CHB\_HEX, putative carbohydrate-binding domain (PF03173); GH20, glycosyl hydrolase family 20 (PF00728); GH3, glycosyl hydrolase family 3 (PF00933).

#### *2.3. Characterization of the GH18 Chitinases with Activity on Crystalline Chitin*

The high abundance of the GH18 chitinases in the secretome of strain DSM 14401 implies that they are likely to be the chitinases with activity on crystalline chitin. Thus, 3 GH18 chitinases, Chia4287, Chib0431, and Chib0434, with high abundance in the secretome, were selected to be expressed and characterized. Genes encoding Chia4287, Chib0431, and Chib0434 were heterologously expressed in *E. coli* BL21 (DE3), and the recombinant proteins were purified by NTA-Ni Sepharose affinity chromatography (Figure 4). The purification folds for Chib0431, Chib0434 and Chia4287 were 6.75, 5.33, and 7.30, respectively (Table S2). As shown in Figure 4, the 3 purified recombinant proteins have apparent molecular weights of approximately 88 kDa (Chib0431), 112 kDa (Chib0434), and 51 kDa (Chia4287), consistent with their theoretical molecular weights (Table 1).

**Figure 4.** The SDS-PAGE analysis of recombinant proteins Chib0431, Chib0434, and Chia4287. Lane M, protein molecular mass marker; Lane 1, the cell lysate of *E. coli* containing recombinant protein Chib0431; Lane 2, the purified recombinant protein Chib0431; Lane 3, the cell lysate of *E. coli* containing recombinant protein Chib0434; Lane 4, the purified recombinant protein Chib0434; Lane 5, the cell lysate of *E. coli* containing recombinant protein Chia4287; Lane 6, the purified recombinant protein Chia4287. The enzyme bands are indicated by arrows.

To investigate the substrate specificity of these 3 chitinases, the enzyme activities of Chib0431, Chib0434, and Chia4287 toward colloidal chitin, chitin powder, chitosan, microcrystalline cellulose, 4-Methylumbelliferyl *N*-acetyl-β-D-glucosaminide (MUF-GlcNAc) [55], 4-Methylumbelliferyl-β-D-*N*,*N* -diacetylchitobioside hydrate (MUF-(GlcNAc)2) [56], and 4-Methylumbelliferyl-β-D-*N*,*N* ,*N*-triacetylchitotrioside (MUF-(GlcNAc)3) [57] were determined. As shown in Table 2, all the three chitinases had activity toward colloidal chitin, crystalline chitin, MUF-(GlcNAc)2, and MUF-(GlcNAc)3, but neither had activity toward chitosan, microcrystalline cellulose, or MUF-GlcNAc. Among them, Chia4287 had the highest activity towards chitin powder, followed by Chib0431 and Chib0434, which is consistent with their amount in the secretome. Chitinases Chia4287 and Chib0431 exhibited higher activities toward MUF-(GlcNAc)3 than MUF-(GlcNAc)2, suggesting that both enzymes likely function as endochitinases. In contrast, Chib0434 showed approximately 10-fold higher activity toward MUF-(GlcNAc)2 than MUF-(GlcNAc)3, suggesting that Chib0434 tends to act as an exochitinase.

With chitin powder as the substrate, the three chitinases were biochemically characterized. Both chitinases Chib0431 and Chia4287 showed optimum temperatures at 50 ◦C, and Chib0434 at 45 ◦C (Figure 5A). For their thermal stability, Chib0431 retained approximately 100% activity at 40 ◦C and more than 61% at 50 ◦C after 120 min incubation but lost all its activity at 60 ◦C in 15 min (Figure 5B). Chib0434 retained 100% activity at 40 ◦C after 120 min incubation but lost all its activity at 50 ◦C in 90 min and at 60 ◦C in 30 min (Figure 5C). Chitinase Chia4287 retained high activity (≥89%) when incubated at 40 ◦C for 120 min (Figure 5D). Chitinases Chib0434 and Chia4287 both showed highest activity at pH 7.5 and Chib0431 at pH 7.0 (Figure 6A). For their pH stability, the 3 chitinases all exhibited high stability (retaining ≥80% activity) from pH 5.0 to 11.0 in the Britton–Robinson buffer after 10 h incubation at 4 ◦C (Figure 6B). They all showed highest activity at 0 M NaCl (Figure 6C) but maintained high activity (≥92%) in 1–5 M NaCl after 10 h incubation at 4 ◦C (Figure 6D). Therefore, the 3 chitinases have temperature optima of 45–50 ◦C and pH

optima of 7.0–7.5, indicating that they are all neutral and mesophilic enzymes. They are all stable at 40 ◦C and in the pH range of 5.0–11.0, and all have excellent salt tolerance.

**Table 2.** The substrate specificity of the three chitinases of strain DSM 14401 a.


<sup>a</sup> The data in the table are from three experiment repeats (mean ± SD). <sup>b</sup> ND means that the enzyme activity was not detectable.

**Figure 5.** Effect of temperature on the activities and stabilities of chitinases Chib0431, Chib0434, and Chia4287. (**A**) Effect of temperature on the activities of Chib0431, Chib0434, and Chia4287. The activities of each enzyme were measured at its optimal pH with chitin powder as the substrate. The highest activity of each enzyme was defined as 100%. (**B**) Effect of temperature on the stability of Chib0431. (**C**) Effect of temperature on the stability of Chib0434. (**D**) Effect of temperature on the stability of Chia4287. In B, C, and D, the residual activities of each enzyme were measured at its optimal temperature and pH with chitin powder as the substrate, and the activity of each enzyme without incubation was defined as 100%. The graphs show data from triplicate experiments (mean ± SD).

**Figure 6.** Effects of pH and NaCl on the activities and stabilities of chitinases Chib0431, Chib0434, and Chia4287. (**A**) Effect of pH on the activities of Chib0431, Chib0434, and Chia4287. The activities of each enzyme were measured at its optimal temperature with chitin powder as the substrate. The highest activity of each enzyme was defined as 100%. (**B**) Effect of pH on the stabilities of Chib0431, Chib0434, and Chia4287. The residual activities of each enzyme were measured at its optimal temperature and pH with chitin powder as the substrate. (**C**) Effect of NaCl concentration on the activities of Chib0431, Chib0434, and Chia4287. The activities of each enzyme were measured at its optimal temperature and pH with chitin powder as the substrate. The activity of each enzyme in 0 M NaCl was defined as 100%. (**D**) Effect of NaCl concentration on the stabilities of Chib0431, Chib0434, and Chia4287. The residual activities of each enzyme were measured at its optimal temperature and pH with chitin powder as the substrate. The highest activity of each enzyme was defined as 100%. The graphs show data from triplicate experiments (mean ± SD).

Many chitinases have been heterologously expressed and characterized with colloidal chitin or synthetic chitooligosaccharide analogs. As shown in Table 3, the temperature and pH optima of the reported chitinases and their thermostability are quite diverse. So far, several *Pseudoalteromonas* GH18 chitinases have been characterized (Table 3). The chitinase Chi23 from *P. aurantia* DSM6057 was reported to be thermostable but active toward crystalline chitin only in acidic conditions (pH of 3.0–6.0) [18]. Chitinases ChiA and ChiC from *Pseudoalteromonas* sp. DL-6 [27,37] and ChiB from *Pseudoalteromonas* sp. O-7 [38] are all cold-active enzymes with optimal activities at 20–30 ◦C and low thermostability. The 3 mesophilic chitinases, Chib0431, Chib0434, and Chia4287, characterized in this study are active toward crystalline chitin at neutral pH conditions (pH 7.0–7.5) and have good thermostability and pH- and salt-tolerance, which, therefore, may be good candidates for industrial application.


**Table 3.**

Characteristics

 of bacterial chitinases.


**Table 3.** *Cont.*

#### *2.4. Analysis of the Products of the Chitinases on Crystalline Chitin*

In order to investigate the application potential of the three chitinases in preparing COSs/GlcNAc from natural chitin, we analyzed the degradation products of Chia4287, Chib0431, and Chib0434 towards crystalline chitin. The reaction mixtures, containing chitin powder and chitinases, were incubated at their respective optimal temperatures for different time periods (15 min, 30 min, 1 h, and 3 h). The COSs/GlcNAc released from chitin in the supernatants of the mixtures were analyzed by gel filtration chromatography on a Superdex Peptide 10/300 GL column. For Chib0431 and Chib0434, during the 3 h degradation of crystalline chitin, (GlcNAc)2 was always the predominant product, with only a slight amount of GlcNAc (Figure 7A,B). However, in the hydrolytic products of Chia4287 on crystalline chitin, although (GlcNAc)2 was also the main product, the proportion of GlcNAc was much higher compared to that in the hydrolytic products of Chib0431 and Chib0434 (Figure 7C). Together, these results indicate that Chia4287, Chib0431, and Chib0434 can degrade crystalline chitin into (GlcNAc)2 and GlcNAc, with (GlcNAc)2 as the main product. These results imply that they may have potential in the preparation of (GlcNAc)2 from natural crystalline chitin.

**Figure 7.** Analysis of the degradation products of the three chitinases on crystalline chitin. (**A**) The degradation product of Chib0431. (**B**) The degradation product of Chib0434. (**C**) The degradation product of Chia4287. Chitin powder was degraded by the chitinases at their respective optimal temperatures for different times (15 min, 30 min, 1 h, and 3 h). The reaction system with enzyme inactivated at 100 ◦C for 10 min was used as the control. The reaction was terminated by boiling at 100 ◦C for 10 min, and then the reaction mixtures were centrifuged at 17,949× *g* for 10 min. The products in the supernatants were analyzed by gel filtration chromatography on a Superdex Peptide10/300 GL column (GE Healthcare, Sweden), which were monitored at a wavelength of 210 nm. The injected volume was 10 μL. DP1-DP6 are chitooligosaccharide markers. DP1, GlcNAc; DP2, (GlcNAc)2; DP3, (GlcNAc)3; DP4, (GlcNAc)4; DP5, (GlcNAc)5; DP6, (GlcNAc)6.

COSs/GlcNAc have been widely prepared with a variety of crude enzymes from wild strains and purified recombinant chitinases, most of which were prepared with colloidal chitin [17,58–61]. So far, however, there have been only a few chitinases used to prepare COSs/GlcNAc from natural crystalline chitin. The enzyme cocktail of strain *Paenibacillus* sp. LS1 can produce GlcNAc and (GlcNAc)2 with minor (GlcNAc)3 from crystalline αchitin [30]. The crude enzyme of *Aeromonas hydrophila* H-2330 mainly produces GlcNAc from crystalline α-chitin [31]. The chitinase ChiA of strain *Pseudoalteromonas* sp. DL-6 is an endochitinase, and its products on crystalline α-chitin are a mixture of chitin COSs (DP 2–6), with (GlcNAc)2 as the major product [27]. The mixture of purified chitinases SaChiB and SaHEX of strain *Streptomyces alfalfa* ACCC40021 can enhance the conversion of crystalline α-chitin to GlcNAc [62]. The chitinase of strain *Chitinibacter* sp. GC72 can degrade practical-grade chitin into GlcNAc [33]. The three chitinases characterized in this study can degrade crystalline α-chitin into (GlcNAc)2, suggesting their potential in direct conversion of natural crystalline chitin into (GlcNAc)2.

#### **3. Materials and Methods**

#### *3.1. Bacterial Strains and Experimental Materials*

The 26 type strains of genus *Pseudoalteromonas* were purchased from Deutsche Sammlung von Mikroorganismen and Zelkulturen (DSMZ) or Japan Collection of Microorganisms (JCM). Chitin powder (crystalline α-chitin), MUF-GlcNAc, MUF-(GlcNAc)2, and MUF- (GlcNAc)3 were purchased from Sigma-Aldrich (St. Louis, MO, USA). Chitin flakes, purchased from Yuan Cheng Group (Wuhan, China), are crystalline α-chitin. Colloidal chitin was prepared as previously described [18]. GlcNAc, (GlcNAc)2, (GlcNAc)3, (GlcNAc)4, (GlcNAc)5, and (GlcNAc)6 were purchased from BZ Oligo Biotech Co., LTD (Qingdao, China). Chitosan was purchased from Sangon Biotech (Shanghai, China). BCA protein assay kit was purchased from Thermo Scientific (Boston, MA, USA). Other chemicals were of analytical grade and commercially available.

#### *3.2. Screening of Strain DSM 14401*

The 26 type strains of genus *Pseudoalteromonas* (Table S1) were cultivated at 25 ◦C and 180 rpm in the TYS medium composed of 0.5% (*w*/*v*) peptone, 0.1% (*w*/*v*) yeast powder, and artificial seawater (pH 7.8). When the OD600 of the culture was approximately 1.0, 2 mL cell suspension was collected and the cells were washed with the minimal medium (30 g/L NaCl, 0.5 g/L NH4Cl, 3 g/L MgCl2·6H2O, 2 g/L K2SO4, 0.2 g/L K2HPO4, 0.01 g/L CaCl2, 0.006 g/L FeCl3·6H2O, 0.005 g/L Na2MoO4·7H2O, 0.004 g/L CuCl2·2H2O, 6 g/L Tris, pH 7.6) three times. Then, the washed cells were inoculated into the minimal medium supplemented with 0.05% (*w*/*v*) peptone, 0.01% (*w*/*v*) yeast powder, and 3% (*w*/*v*) chitin flakes and cultivated at 25 ◦C and 180 rpm for 5 days. Their growth and the degree of degradation of the chitin flakes were observed every day. Among them, strain DSM 14401 showed the highest degradation rate on crystalline α-chitin, which was then chosen for further study. The OD600 of the culture of this strain in the medium was measured at different time intervals, as indicated in Figure 1, to produce its growth curve. The washed cells were cultured in the same medium without chitin flakes and in the same conditions as the control.

#### *3.3. Extracellular Chitinase Activity Assay of Strain DSM 14401*

During the cultivation of strain DSM 14401 in the above liquid medium with or without chitin flakes, 1 mL of culture was taken out at different intervals, as indicated in Figure 1. The cultures were filtered with a 0.22 μm filter to remove the bacterial cells, and the filtrate was used for the extracellular chitinase activity assay. A 200 μL mixture consisting of 50 mM Tris-HCl (pH 7.0), 3% chitin powder, and 50 μL of filtrate was incubated at 50 ◦C for 2 h. The mixture was then centrifuged at 17,949× *g* for 2 min at 4 ◦C and the supernatant obtained was used for the reducing-sugar assay by the DNS method [63]. The control mixture contained a pre-boiled filtrate instead of the filtrate. Subsequently, the optical density at 550 nm was measured to quantify the released reducing sugar. The amount of reducing sugar generated was calculated using GlcNAc as a standard. One unit of enzyme

activity was defined as the amount of enzyme that liberated 1 μmol of reducing sugar per minute.

#### *3.4. Bioinformatics Analysis*

The genome DNA of strain DSM 14401 was sequenced by our lab [64]. The putative chitinases of this strain were determined according to dbCAN [65] analyses. Signal peptides of the chitinases were predicted by SignalP 5.0 (http://www.cbs.dtu.dk/services/SignalP/ (accessed on 12 January 2022)) [66]. The domain architectures of the chitinases were predicted on the HMMER website (https://www.ebi.ac.uk/Tools/hmmer/search/hmmscan (accessed on 12 January 2022)) [67]. The phylogenetic tree was constructed based on the Neighbor-Joining method and using the Poisson model with MEGA X after multiple alignments of the sequences by MUCLE [68]. Sequences alignment results were visualized using the ESPript 3.0 server [69]. The molecular weights of the chitinases were predicted by the ExPASy Server (https://web.expasy.org/compute\_pi/ (accessed on 12 January 2022)) [70].

#### *3.5. Secretome Analysis*

Strain DSM 14401 was cultured at 25 ◦C and 180 rpm in a medium containing the minimal medium and 3% chitin flakes. When approximately half of the chitin flakes were degraded, the culture was centrifuged at 8228× *g* at 4 ◦C for 6 min. The precipitates were resuspended using 20 mM Tris-HCl (pH 8.0) containing 1 M NaCl, and then centrifuged at 1157× *g* at 4 ◦C for 3 min. This step was repeated three times. The resultant precipitates were resuspended using 50 mM Tris-HCl (pH 8.0) containing 6 M Guanadine-HCl, and then centrifuged at 15,557× *g* at 4 ◦C for 10 min. The supernatant was moved into an ultrafiltration tube (15 mL, 3 kDa). The Guanadine-HCl in the supernatant was removed by adding 50 mM Tris-HCl to the ultrafiltration tube (molecular weight cut-off, 3 kDa) and centrifugation (4629× *g* for 10 min at 4 ◦C) for three times. Then, the proteins in the supernatant were precipitated by 50 mL acetone containing 10% trichloroacetic acid and 0.1% dithiothreitol overnight at −20 ◦C. The precipitates were harvested and washed by 80% acetone and 100% acetone successively, and then lyophilized. The lyophilized sample was successively denatured, reduced, and alkylated by denaturation buffer (0.5 M Tris-HCl, 2.75 mM EDTA, 6 M Guanadine-HCl), dithiothreitol (1 M), and iodoacetamide (1 M), respectively. The sample solution was further replaced with 25 mM NH4HCO3 solution by centrifugation ultrafiltration (15,294× *g* for 15 min at 4 ◦C) in an ultrafiltration tube (1 mL, 3 kDa). The sample was digested using trypsin at 37 ◦C for 12 h, and the resultant peptides were desalted on a C18 column (ZipTip C18, Millipore, Billerica, MA, USA). The desalted peptides were analyzed using the mass spectrometer Orbitrap Elite (Thermo Fisher Scientific, Bremen, Germany) coupled with Easy-nLC 1000 (Thermo Fisher Scientific, Bremen, Germany). Finally, the raw data was analyzed against the genome of strain DSM 14401 using Thermo Scientific Proteome DiscovererTM 1.4. The mass spectrometry proteomics data have been deposited to the ProteomeXchange [71] Consortium via the PRIDE [72] partner repository with the dataset identifier PXD030600. The reviewer account details: Username: reviewer\_pxd030600@ebi.ac.uk; Password: 1QCP2jqI.

#### *3.6. Expression and Purification of Chitinases Chib0431, Chib0434, Chia4287*

The gene sequences of Chib0431, Chib0434, and Chia4287 without the signal peptide were cloned from the genomic DNA of strain DSM 14401 and inserted into the NdeI and XhoI sites of the expression vector pET-22b(+). The constructed recombinant plasmids were then transformed into *E. coli* BL21(DE3) for protein expression. The constructed recombinant *E. coli* BL21(DE3) strains were cultured at 37 ◦C in liquid LB medium containing 100 μg/mL ampicillin. When the OD600 of the cultures reached 0.6–1.0, 0.45 mM isopropyl thio-β-D-galactoside (IPTG), used as an inducer, was added into the cultures, and the cultures were incubated at 18 ◦C for 16 h. Then, the recombinant *E. coli* cells in the cultures were collected via centrifugation and crushed by sonication in the lysis buffer (100 mM NaCl, 5 mM imidazole, 50 mM Tris-HCl pH 8.0). The recombinant proteins of

Chib0431, Chib0434, and Chia4287 in the cell extracts were further purified by affinity chromatography with Ni-NTA agarose resins (Qiagen, Santa Clarita, CA, USA), followed by desalination on PD-10 Desalting Columns (GE Healthcare, Piscataway, NJ, USA), using 10 mM Tris-HCl containing 100 mM NaCl (pH 8.0) as the running buffer. The purified proteins were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) [73]. The protein concentrations were determined using a BCA protein assay kit with bovine serum albumin (BSA) as the standard.

#### *3.7. Enzyme Assays*

The activities of the three purified chitinases towards chitin powder, colloidal chitin, chitosan, microcrystalline cellulose, MUF-GlcNAc, MUF-(GlcNAc)2 and MUF-(GlcNAc)3 were assayed in 50 mM Tris-HCl at their respective optimal temperatures and pHs (50 ◦C and pH 7.0 for Chia4287, 50 ◦C and pH 7.5 for Chib0431, 45 ◦C and pH 7.5 for Chib0434). When the insoluble chitin powder, chitosan, or microcrystalline cellulose was used as the substrate, the reaction mixture contained 190 μL 50 mM Tris-HCl, 3% (*w*/*v*) substrate and 10 μL enzyme, which was incubated for 1 h for Chia4287 or 2 h for Chib0431 and Chib0434. When colloidal chitin was used as the substrate, the reaction mixture contained 190 μL 0.75% (*w*/*v*) colloidal chitin in 50 mM Tris-HCl and 10 μL enzyme, which was incubated for 40 min. After incubation, the activities of the chitinases towards these substrates were determined using the DNS method [63]. The enzyme activity (U) was defined as the amount of enzyme that required to release 1 μmol GlcNAc equivalent reducing sugar from the substrate per minute. When MUF- GlcNAc, MUF-(GlcNAc)2, or MUF-(GlcNAc)3 was used as the substrate, the enzyme activity was assayed for 15 min with the reaction mixture contained 790 μL 1 mM substrate in 50 mM Tris-HCl and 10 μL enzyme, which was incubated for 15 min and then terminated by an addition of 0.4 M NaCO3. The enzyme activity (U) was defined as the amount of enzyme that required to release 1 μmol MUF from the substrate per minute.

#### *3.8. Characterization of the Chitinases*

The purified Chib0431, Chib0434, and Chia4287 were characterized with chitin powder as substrate. The effect of temperature on the enzyme activity was measured by assaying the enzyme activity at different temperatures (0–80 ◦C for Chia4287; 10–70 ◦C for Chib0431 and 20–60 ◦C for Chib0434) and their respective optimal pHs. The effect of pH on the enzyme activity was measured by assaying the enzyme activity in the Britton-Robinson buffer at different pHs (pH 4.0–9.0 for Chia4287; pH 5.0–10.0 for Chib0431 and Chib0434) and their respective optimal temperatures. Effect of salinity on the enzyme activity was assayed by assaying the enzyme activity in 50 mM Tris-HCl containing different concentrations of NaCl (0–5 M for Chib0431 and Chia4287; 0–2 M for Chib0434) at their respective optimal temperatures and pHs.

For the thermal stability assay, the purified chitinases were incubated at 40 ◦C, 50 ◦C, or 60 ◦C for 0–120 min, and the residual activities towards chitin powder were measured at an interval of 15 min under their respective optimal temperatures and pHs. For the pH stability assay, the purified chitinases were incubated in the Britton-Robinson buffers ranging from pH 3.0 to pH 11.0 at 4 ◦C for 10 h, and the residual activities towards chitin powder were measured at their respective optimal temperatures and pHs. For the halotolerance assay, the purified chitinases were incubated in 50 mM Tris-HCl containing different concentrations of NaCl (0–5 M) at 4 ◦C for 10 h, and the residual enzyme activities towards chitin powder were measured at their respective optimal temperatures and pHs.

#### *3.9. Analysis of the Products Released from Crystalline Chitin by the Chitinases*

The purified Chib0431, Chib0434, and Chia4287 (10 μL) were incubated with 3.0% chitin powder in 190 μL of 50 mM Tris-HCl (pH 7.0) for different times (15 min, 30 min, 1 h, and 3 h) at their respective optimal temperatures. The reaction was terminated by boiling at 100 ◦C for 10 min, and the reaction mixtures were centrifuged at 17,949× *g* for 10 min. Then, the products in the supernatants were analyzed by gel filtration chromatography on a Superdex Peptide 10/300 GL column (GE Healthcare, Uppsala, Sweden), which were monitored at 210 nm using a UV detector. The injected volume was 10 μL. The products were eluted with 0.2 M ammonium hydrogen carbonate for 90 min with a flow rate of 0.3 mL/min. The reaction system containing 10 μL enzyme pre-heated at 100 ◦C for 10 min was used as the control. A mixture of GlcNAc, (GlcNAc)2, (GlcNAc)3, (GlcNAc)4, (GlcNAc)5, and (GlcNAc)6 was used as the marker.

#### **4. Conclusions**

COSs have wide application in agriculture, medicine, cosmetics, and foods. While most COSs are now prepared with colloidal chitin, there are only a few reports of chitinases with potential in the preparation of COSs from natural crystalline chitin. In this study, three chitinases with activity on crystalline chitin were identified from a marine *Pseudoalteromonas* strain and characterized. These chitinases are all neutral mesophilic enzymes, which are most active at 45–50 ◦C and pH 7.0–7.5, and have high stability at 40 ◦C, pH 5.0–11.0, and in 5 M NaCl. The main products of the three chitinases on crystalline chitin are all (GlcNAc)2, suggesting that these chitinases have potential in preparing (GlcNAc)2 via direct degradation of natural crystalline chitin. Further studies such as improving the expression amount of these chitinases and their degradation efficiency on crystalline chitin are underway.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/md20030165/s1, Figure S1: Multiple sequence alignments of Chib0431, Chib0434, Chia4287, and Chia2822 with known GH18 chitinases.; Figure S2: Phylogenetic analysis of Chib0889 with other GH19 chitinases.; Table S1: General information of 26 type strains of Pseudoalteromonas.; Table S2: Purification of the recombinant enzymes Chib0431, Chib0434, and Chia4287. References [39,74–96] are cited in the supplementary materials.

**Author Contributions:** Conceptualization, X.-L.C. and Y.-Z.Z.; Investigation, Y.-J.W.; Methodology, X.-B.R., S.-S.L. and Y.-J.W.; Project administration, X.-L.C. and P.-Y.L.; Resources, X.-L.C. and Y.-Z.Z.; Software, Y.-R.D. and K.-X.H.; Supervision, Q.-L.Q., X.-L.C. and P.-Y.L.; Writing—original draft, X.-B.R.; Writing—review & editing, X.-L.C. and P.-Y.L. All authors have read and agreed to the published version of the manuscript.

**Funding:** This work was supported by the National Science Foundation of China (grants 42176229, 31870101, U2006205, 31870052 and awarded to P.-Y.L., Q.-L.Q., X.-L.C. and X.-L.C., respectively), the Major Scientific and Technological Innovation Project (MSTIP) of Shandong Province (2019JZZY010817 awarded to Y.-Z.Z.), Taishan Scholars Program of Shandong Province (tspd20181203 awarded to Y.-Z.Z.).

**Institutional Review Board Statement:** This article does not contain any studies involving human participants or animals performed by any of the authors.

**Data Availability Statement:** Proteomic data are available via ProteomeXchange with identifier PXD030600.

**Acknowledgments:** We would like to thank Andrew McMinn from the University of Tasmania, Australia, for editing this paper. We would like to thank Cai-Yun Sun and Rui Wang from State Key Laboratory of Microbial Technology of Shandong University for help and guidance in Bioscreen C Microbiology reader.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

