Comparison of the potential activities of viral and bacterial chitinases

Background: Chitin, a long-chain polymer of N-acetylglucosamine, is a major structural component of the insect exoskeleton and the peritrophic membrane (PM). Chitinases are able to effectively break down glycosidic bonds of chitin polymer thus can be used in agriculture to control plant pathogen insects. These enzymes can be synthesized by higher plants, animals, protista, bacteria, and viruses. Results: In this study, viral and bacterial chitinases were compared for their potential activity on a laboratory test insect. The genes encoding chitinases of Autographa californica nucleopolyhedrovirus (AcNPV) and Cydia pomonella granulovirus (CpGV) were amplified from genomic DNAs by PCR and cloned into the pET-28a (+) expression vector. The chitinase proteins of these 2 viruses (AcNPV-Chi, CpGV-Chi) and Serratia marcescens chitinase C (ChiC) protein which was previously cloned were overexpressed in Escherichia coli. Expressed proteins were purified and confirmed by western blot analysis as 50, 63, and 68 kDa for AcNPV, CpGV, and S. marcescens chitinases, respectively. Enzyme activities of the chitinases were confirmed. Chitinases were also compared to each other in silico. The insecticidal effects of these proteins were evaluated on Galleria mellonella L. larvae. Bioassays were performed on the 3rd instar larvae for each chitinase protein in triplicate. The results showed that although there were differences in enzymatic activities and domain organizations, all 3 microbial chitinases produced almost the same level of insecticidal activity on the test insect. LC50 and LT50 values were compatible with the mortality results. These results were a preanalysis for comparing the effects of microbial chitinases. Conclusion: Potential activity experiments should be carried out on more insects to provide detailed information on the insecticidal effects of bacterial and viral chitinases.

chitinases will reduce nutrition and protection against microbial attacks.
Bacteria and viruses are other chitinase producers. Microbial chitinases weaken and disrupt the chitin structures in insects, thus increasing the risk of microbial infection of the host. Bacteria produce chitinases to obtain nitrogen and carbon as a source of nutrients or precursors and parasitism (Rathore and Gupta 2015). Several genera of bacteria, including Serratia (Bahar et al. 2012), Bacillus (Sasi et al. 2020), Streptomyces (Tran et al. 2019), Enterobacter (Sivaramakrishna et al. 2020), and Vibrio (Jahromi and Barzkar 2018), produce chitinase. A few viral chitinases, mainly species from Baculoviridae, are currently known (Ishimwe et al. 2015). Viral chitinases weaken the host barrier structure to facilitate virus infection and release progeny viruses from infected cells. The most widely studied chitinase from baculoviruses belongs to Autographa californica nucleopolyhedrovirus (AcNPV) species (Hawtin et al. 1997). Viral chitinases and most bacterial chitinases belong to family 18 glycohydrolases that belong to a family of multimodular proteins (Berini et al. 2018). AcNPV chitinase is closely related to Serratia marcescens ChiA thus suggesting that baculovirus chitinase may have been acquired from a bacterium found in insect bodies.
The negative effects of chemicals on the ecosystem and on human health force scientists to find alternative solutions against insect pests. Insect pathogenic bacteria and viruses or their bioactive agents are good options to be used as biological pest control materials.
This study aimed to investigate whether there is a difference between viral and bacterial chitinase enzymes in terms of their potential activity.

In silico analysis of chitinases
AcNPV (NC_001623.1), CpGV (NC_002816.1), and S. marcescens (KF823632) chitinase proteins were analyzed in terms of their domain content at the NCBI conserved domain search database. Amino acid sequences of chitinases were compared to each other using NCBI protein blast and multiple alignment tools.

Construction of chitinase expressing bacterial vectors
Viral genomic DNAs were isolated from virus suspensions, using DNeasy Blood & Tissue Kit (Qiagen, 69506). Specific primers were designed for both AcNPV chitinase (Fw: 5′-GGGGATCCATTCCCGGCACGC-3′; Rv: 5′-GGCTCGAGTTACAGTTCATCTTTAGGT-3′) and CpGV chitinase (Fw: 5′-GGGGATCCAAACCCGGCA CACC-3′; Rv: 5′-GGCTCGAGTCATACTGAATTGC ACAC-3′). BamHI and XhoI restriction endonucleases were added to the 5′ ends of forward and reverse primers, respectively. Both chitinase genes (AcNPV-chi, CpGV-chi) were amplified by PCR from genomic DNAs. The reactions contain 10 ng of genomic DNA, 2.5 μl of both primers of 10 μM, 1 μl of 10 mM dNTPs, 10 μl 5X Phusion HF reaction buffer, and 1 U Phusion HF DNA polymerase. Lastly, the volume was adjusted to 50 μl with sterile dH 2 O. The reaction program was as follows: an initial denaturation step at 98°C for 3 min followed by 30 cycles of 98°C for 50 s, 55°C for 50 s, 72°C for 1 min, and final extension step at 72°C for 10 min. The PCR products were ligated into the pJET1.2/blunt vector according to the manufacturer's instructions. Ligation reactions were transformed into E. coli DH10β cells, and positive clones were confirmed. The generated recombinant plasmids carrying AcNPV-chi and CpGV-chi genes were named as pAcNPV and pCpGV, respectively. The chitinase genes were sequenced by Macrogen Inc. (Amsterdam, the Netherlands). Following the sequence analysis, both chitinase genes containing BamHI and XhoI restriction sites at their 5′ and 3′ ends were introduced into the pET-28a(+) expression vector using the same sites. These plasmids were transformed into E. coli BL21 (DE3) competent cells, and recombinant plasmids were selected in the presence of Kanamycin (50 μg/ml).

Over-expression, purification, and western blot analysis of the recombinant proteins
The AcNPV-Chi, CpGV-Chi, and S. marcescens-ChiC proteins were overexpressed as fusion proteins with the 6×His-tag at their N-terminal in E. coli BL21 (DE3) cells. Chitinase proteins were purified by using the MagneHis™ Protein Purification System Kit (Promega) and dialyzed for 24 h through 1 l of 1X PBS buffer, pH 7.5. The identification and purity of the samples were confirmed by 10% sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) and subsequently Coomassie staining. Western blot analysis was also performed to demonstrate heterologous gene expression at the immunological level. Electrophoresed proteins were transferred to the nitrocellulose membrane. Immunedetection was performed using 1:1000 diluted polyclonal rabbit anti-His-taq antibodies (Abcam) and subsequently 1:1000 diluted polyclonal alkaline phosphataseconjugated goat-anti-rabbit IgG (Millipore). The binding of the antibodies was then visualized with the NBT-BCIP substrate system (Roche).

Chitinase activity assay
Chitinase activities were assayed by using the colorimetric 3,5-dinitrosalicylic acid (DNS) method (Monreal and Reese 1969) with some modifications. The purified bacterial and viral chitinases (0.5 μg) were reacted with 150 μl of colloidal chitin (12.5 mg/ml chitin) as a substrate and incubated at 30°C for 3 h. The reactions were terminated by boiling the mixture in the presence of 300 μl DNS acid reagent for 5 min. In the control reaction, only substrate and DNS were used. The hydrolysis of chitin was measured at 540 nm and run with a glucose standard. One unit of the chitinase activity was defined as the amount of enzyme that liberates 1 μM of reducing sugar per 1 min at 30°C.

Potential activity test
Proteins were tested on G. mellonella at 5 different concentrations (1.5, 1.8, 2, 2.5, and 3 μg protein) per larvae. Bioassays were performed by 30 larvae (3rd instar) for each concentration in triplicate. The larvae were starved for 12 h and then fed with an artificial diet inoculated with desired concentrations of protein. After consumption of the inoculated diets, non-inoculated ones were added and incubated at 26 ± 2°C. Control group larvae were fed on only 1X PBS treated diets. Mortalities were recorded daily for 14 days. An experimental design, showing the insecticidal activity tests, is given in Supplementary Fig. 1.

Statistical analysis
Mortality data were corrected using Abbott's formula (Abbott 1925); LC 50 and LT 50 values were calculated by Probit analysis using MS Excel (Finney 1952).

In silico analysis of chitinases
Chitinase proteins were analyzed for their domain contents. The results showed that viral chitinases had similar domain structures to each other but different than that of bacterial chitinase ( Supplementary Fig. 2). Viral chitinases contain a ChitinaseA_N domain, whose function is binding to chitin substrat, followed by a glyco_18 domain that catalyses the hydrolysis of 1,4-beta-linkages of N-acetylglucosamine in chitin and chitodextrins. S.marcescens-ChiC protein contains a Chi1 domain, whose function is hydrolyzing the chitin, followed by a COG3979 domain that cleaves soluble oligomers to the di-and trisaccharides. Multiple alignments of amino acid sequences of 3 chitinases together did not produce a significant similarity. Likewise, pairwise alignments of amino acid sequences of bacterial chitinase with either AcNPV or CpGV chitinases did not produce similarity.
Although viral chitinases had similar domain organizations, their pairwise identity at amino acid level was only 60%.

Expression, purification, and enzyme activities of chitinases
The sizes of the DNA fragments amplified by PCR for the AcNPV-chi and CpGV-chi genes corresponded with the expected sizes of 1655 and 1784 bp, respectively (Fig. 1a, b). These products were first cloned into the pJET1.2/blunt vectors and subsequently to the pET-28a (+) vectors (Novagen) individually. AcNPV-Chi, CpGV-Chi, and S. marcescens-ChiC proteins were expressed in E. coli with a N-terminal 6xHis-tag. Recombinant Histagged proteins were purified, using the MagneHis™ Protein Purification kit. Purified proteins were analyzed by 10% SDS-PAGE. The electrophoresis process was performed using 2 gels on which samples were loaded identically. In one of the 2 gels stained with Coomassie Brillant Blue, S. marcescens-ChiC, AcNPV-Chi, and CpGV-Chi proteins were observed at sizes of 50, 63, and 68 kDa, respectively, as expected (Fig. 2). The other gel was used for western blot analysis using histidine antibody, which facilitates seeing these proteins by binding to the histidine tag attached to the N-terminal of the proteins. As a result of this analysis, expected chitinase protein sizes were observed on the membrane (Fig. 3). Chitinase activities of these purified proteins, tested by the DNS method, were determined as colorimetric in a spectrometer (Fig. 4).

Potential activity of chitinase proteins
To analyze the potential activity of AcNPV-Chi and CpGV-Chi and S. marcescens ChiC proteins, G. mellonella larvae were fed on purified and dialyzed proteins. The results showed that the effects on G. mellonella larvae were not much different from each other. However, CpGV-Chi, at 2 and 2.5 μg/larvae concentrations, produced a slight high mortality on insects (Fig. 5). The daily number of dead insects was recorded. Mortality effects of chitinase proteins over time were also examined ( Fig. 6). LC 50 and LT 50 values were correlated with the mortality results (Tables 1 and 2).

Discussion
Potential effects of chitinase proteins which belong to Autographa californica nuleopolyhedrovirus (AcNPV), Cydia pomonella granulovirus (CpGV), and Serratia marcescens were tested and compared to each other. The results showed that the potential effects of all 3 chitinases were comparable, with only a slightly high mortality rate caused by CpGV chitinase. Increased chitinase concentration produced high mortality rates. Amino acid sequence alignments of chitinases showed that S. marcescens-ChiC and viral chitinases did not show significant similarity to each other. Also, domain positions were opposite in S. marcescens-ChiC and viral chitinases, while the glyco_18 and GH_18 domains located at the C terminal of viral chitinases showed similar domains located at the N terminal of S. marcescens-ChiC. S.marcescens-ChiA, structurally analyzed previously by Danişmazoğlu et al. (2015), had a similar domain organization to those of AcNPV-Chi and CpGV-Chi.
Enzyme activities of chitinases were confirmed using the DNS method before activity analysis. The results showed that CpGV-Chi activity was higher than S. marcescens-ChiC activity and much higher than AcNPV-Chi Fig. 6 Mortality rates of purified AcNPV-Chi, CpGV-Chi, and Serratia marcescens-ChiC proteins on Galleria mellonella larvae depending on time.
Mortality rates were carried out using the 2.5 μg per larvae protein concentration activity. Although there was a small difference between bacterial and viral chitinases in terms of potential activity results, this difference was also detected in the in vitro enzyme activity assay. Chitinase protein of AcNPV in conjunction with virusencoded cathepsin caused liquefaction of the Trichoplusia ni (Hubner, 1803) larvae in the latter stages of infection (Hawtin et al. 1997). They showed that infections of T. ni larvae with mutant viruses lacking either chitinase or cathepsin did not cause liquefaction of dead insect bodies. In the present study, dead G. mellonella larvae showed softening, but the body remained intact for several days after death. Failure to observe body liquefaction in dead G. mellonella larvae may be due to not using cathepsin protein with chitinase in infections. In another study, the AcNPV chitinase A (ChiA) gene was expressed in E. coli cells, subsequently purified and tested on Bombyx mori (Linnaeus, 1758) larvae at a dose of 1 μg/g of larval body weight. They obtained 100% mortality and concluded that AcNPV-ChiA may offer interesting new opportunities for pest control (Raw et al. 2004). Daimon et al. (2007) constructed a recombinant Bombyx mori NPV (BmNPV) in which BmNPV ChiA was replaced by CpGV ChiA. This recombinant BmNPV was tested on B. mori larvae. Their results showed that CpGV ChiA were able to complement the absence of BmNPV ChiA in the terminal liquefaction process of infected larvae.
The potential activity of S. marcescens ChiC protein, used in this study, was previously studied by Danişmazoğlu et al. (2015), who tested the mortality not only on ChiC protein but also chiA and chiB proteins of S. marcescens on Malacosoma neustria (Linnaeus, 1758) and Helicoverpa armigera larvae. Among these 3 chitinase proteins, ChiC produced the highest potential activity against used insect larvae. At another study, the same S. marcescens chitinase B and C proteins were transformed to Bacillus thuringiensis strains containing crystal proteins, and the potential activity of the recombinant bacteria was tested on G. mellonella larvae (Ozgen et al. 2013). Mortality rates recorded were between 45 and 55%. In the present study, the chitinase C protein of S. marcescens was tested on G. mellonella larvae and displayed around 50% potential activity at 2 μg protein amount per larvae. Insecticidal proteins, which provide important tools to control insect pests, can be used in conjunction with microorganisms. Virion-free proteins of a GV (GVPs) were used with an NPV on H. armigera larvae (Shigeyuki and Chie 2007). Their results showed that the addition of GVPs successfully enhanced NPV infectivity in H. armigera without any negative influence on NPV pathogenicity.

Conclusion
Two viral and a bacterial isolates were compared to each other in terms of their potential activity. The results showed that all the 3 chitinases produced close insecticidal effects on G. mellonella larvae. Testing these proteins on more insects would be more informative to conclude which microbial chitinases are more effective. Chitinases and other insecticidal proteins alone or in conjunction with other control agents may contribute to the development of new biopesticides.