An alternative bacterial expression system using Bacillus pumilus SG2 chitinase promoter

1. Background
    Chitin (C8H13O5N)n is the second most abundant natural polysaccharide composed of N-acetyl-D-glucosamine (GlcNAc) monomers. Chitin is found in many species including outer skeleton of crustaceans, insects, and as a component of fungi and algae cell walls (1-4). Chitinases (EC.3.2.1.14) are hydrolases able to degrade chitin into GlcNAc. There are endo- and exo-chitinases that degrade chitin as a substrate (2). Different organisms including bacteria, protozoa, fungi, and plants produce chitinases (1, 5, 6). These enzymes have several biological roles in different organisms. For instance, these enzymes are produced to provide carbon and nitrogen sources in bacteria (2, 3, 5, 6).
    Generally, heterotrophic bacteria utilize glucose as a preferred carbon source. Meanwhile, in the presence of glucose, transcription of many genes that use other carbon sources are repressed in a mechanism called carbon catabolite repression (CCR) (7, 8). Most of reported chitinase encoding genes are also regulated by CCR. In Streptomyces plicatus, characterization of the promoter region of two chitinase genes, namely chi63 and chi35, showed their induction by chitin and repression in the presence of glucose (9, 10). Several other chitinase genes including chiA and chiC of Streptomyces lividans and chiA, chiB, chiC, chiD and chiF of Streptomyces coelicolor A3 (5) are similarly regulated by chitin and glucose.  In Bacillus subtilis, most of the genes or operons that are under the control of CCR have a cis-acting catabolite responsive element (cre) (7, 11-14). This cre site (WTGNAANCGNWNNCW) located within, upstream or downstream of a promoter region and interact with a trans-acting complex, carbon catabolite protein A (CcpA) and HPr (S46~P) complex. CcpA is a DNA binding protein that belongs to LacI/GalR transcriptional regulators, which is able to repress or activate the transcription of genes depending on the location of cre site at the target promoter (11, 14-16). HPr is a histidine containing protein that forms a complex with CcpA, when it is phosphorylated at its serine residue number 46 (17).
      B. pumilus strain SG2 is a halophilic and chitinolytic bacterium that produces two chitinases, ChiS and ChiL (18, 19). The expression of chiSL genes is induced by chitin as a sole carbon source, while it is repressed in the presence of glucose. In other words, B. pumilus SG2 chitinase encoding genes is controlled by carbon catabolite repression (CCR) (20). In the current study, activity of PchiS from B. pumilus SG2 was evaluated on a low copy plasmid in B. subtilis in order to investigate the possibility of construction of an expression system.

2. Materials and Methods

2.1. Strains, Media and Growth Condition
    All strains are listed in (Table 1). B. subtilis 3NA (the original genes and promoter is from B. pumilus SG2 strain which was isolated from high salt ecosystem, but cloning and experimental works were carried out in a standard B. subtilis called 3NA as explained in (Table 1) was applied to study the activity of chitinase promoter. Escherichia coli JM109 was used for plasmid propagation. LB-agar supplemented with ampicillin (100 mg.mL-1), chloramphenicol (5 mg.mL-1), or spectinomycin (100 mg.mL-1) were used for selection of E. coli and B. subtilis 3NA transformants. Minimal medium (MI) for transformation of B. subtilis was prepared by mixing 97.5 mL of Spizizen’s minimal salts containing 2 g.L-1 (NH4)2SO4, 14 g.L-1 K2HPO4, 6 g.L-1 KH2PO4, 1 g.L-1 Na3C6H5O7.2H2O, 0.2 g.L-1 MgSO4.7H2O, 5 g.L-1 glucose, supplemented with 0.02% (w/v) casamino acids, and 5 mM MgSO4  (21).
Induction of strain 3NA containing pChi1, pChi2 or pChi3 was carried out using LB medium. LB medium (85 mL) in a 500 mL Erlenmeyer flask was inoculated by 1.7 mL of the overnight culture and incubated at 37°C with 200 rpm shake. Cells were grown to an optical density of 0.4 at 600 nm. Aliquots of 8 mL were divided in 100 mL Erlenmeyer flasks and induced with 0.2% (w/v) inducers, i.e. chitin, chitin+glucose, N-acetylglucosamine (GlcNAc), GlcNAc+glucose or without any inducer as the negative control. Colloidal chitin was prepared as described (18). Samples were collected 1 h after the addition of carbohydrates.
      Minimal medium (96.5 mL of MI medium (without glucose) + 0.02% v/v glycerol) was used to study the b-galactosidase activity of prepared strains, namely 3NA pChi1, 3NA pChi2, 3NA pChi3, Chi7, Chi8 and Chi9. The inducers were 0.2% (w/v) chitin + 0.2% v/v glycerol, 0.2% (w/v) chitin + 0.2% (w/v) glucose, 0.2% (w/v) glucose, and 0.2% (v/v) glycerol. Sample collection was performed at 2 h intervals until 11 h after the induction. All experiments were performed three times and mean values were used for comparison.

2.2. Construction of Expression Plasmids and Strains
      Plasmids and oligonucleotide primers are listed in (Table 1). Construction of pUP-Chi2Δcre was carried out using overlapping PCR to fuse the chitinases promoter to the coding region of β-galactosidase gene. PCR1 was performed using ChiSLF10 and UP-CRE2 primer pairs and plasmid pUPChi2 as a template. PCR2 was performed using primers (ChiSR1J and UP-CRE1) and plasmid pUPChi2 as a template. PCR3 was performed using primers (ChiSLF10 and ChiSR1J) and 20 ng of each PCR1 and PCR2 amplicons as templates. The final PCR product was inserted into pDHAFB via Cfr9I and SphI restriction sites. The final cassette was integrated into the amyE locus of B. subtilis chromosome to have a constitutive expression of chitinase. Using this method, the a-amylase encoding gene was inactivated and the transformants were selected based on their chloramphenicol resistance (22). In order to create pChi1, pChi2 and pChi3 constructions, promoter of chitinase gene of B. pumilus SG2 was amplified in a PCR using s5767 and s5768 oligonucleotides and pUP-Chi2, pUP-Chi2Dcre, and pUP-Chi2Dcresig as the templates. The amplified fragment was inserted into a derivative of the theta replicating pBS72 plasmid (pSUN279.2), which is a low copy number plasmid in B. subtilis, upstream of the lacZ as the reporter gene using NheI (NEB) and AflII (NEB) restriction enzymes (Figure 1). The pSUN279.2 has also a pUC18 origin of replication for plasmid propagation in E. coli.
    To integrate the PchiSDcresig-chiS cassette into the chromosome of B. subtilis 3NA, plasmid pUP-Chi2Dcresig was used containing an amyE integration cassette. In this way, strain 3NA containing pChi1, pChi2 or pChi3 was transformed to create Chi7, Chi8, and Chi9, respectively.
2.3. Enzyme Activity
Production of α-amylase was detected by the addition of iodine solution (0.5% iodine in 1% potassium iodide solution)  in nutrient agar containing 1% (w/v) starch (23), while production of chitinase was detected using MI agar plates containing 0.5% (w/v) chitin and 0.001% (v/v) Congo Red (24). b-galactosidase activity was measured y using o-nitrophenyl-D-galactopyranoside (ONPG) as substrate according to Miller assay (25) and the result was given as Miller unit (MU). T-test statistical analysis was used to demonstrate the significance and reliability of the quantitative results.

3. Results

3.1. Activity of the Chitinase Promoter on a pBS72 Derivative in B. subtilis
     Regulation of PchiS and its derivative PchiSDcresig has been already investigated in B. subtilis using an integrated PchiS-lacZ cassette (20). To develop a new gene expression system for B. subtilis, PchiS was inserted upstream of lacZ, as a reporter gene, on pSUN279. Plasmid pSUN279 is an E. coli (oripUC18)-B. subtilis (oripBS72) stable low copy number shuttle vector in B. subtilis. Transformation of B. subtilis 3NA with pChi1 (PchiS) was followed by induction of the transformed mutants in LB in the presence of chitin or GlcNAc with (out) glucose. Measurement of the b-galactosidase activity of the 3NA pChi1 strain showed low amount of the b-galactosidase activity up to 18 MU after 1 h of induction in LB. Since there was not a significant difference in the b-galactosidase activity with chitin or GlcNAc as an inducer compared to the uninduced culture, it was concluded that the expression from the PchiS promoter was constitutive (Figure 2). In contrast, the addition of glucose reduced b-galactosidase activity of 3NA pChi1 by 2.6- and 2.5-fold in the presence of chitin and GlcNAc, respectively (Figure 2). In order to investigate the effect of the cre site on the activity of PchiS, two shortened versions of PchiS, namely PchiSDcre and PchiSDcresig, were inserted upstream of lacZ. The PchiSDcre lacked only the cre site, whereas PchiSDcresig had an additional deletion upstream of the cre site (for the exact promoter sequence see Heravi, Shali (20)). Like 3NA pChi1, the b-galactosidase activities of strains 3NA pChi2 (PchiSDcre) and 3NA pChi3 (PchiSDcresig) were measured. The results revealed that the deletion of the cre site increased the activity in all culture conditions (Figure 2). Unlike 3NA pChi1, in both 3NA pChi2 and 3NA pChi3, PchiS was constitutive. As expected, glucose had no effect on the production of b-galactosidase and the level of enzyme was significantly increased up to 67 MU in 3NA pChi2 and 3NA pChi3. Likewise, no drastic difference was observed between the b-galactosidase activity of 3NA pChi2 and 3NA pChi3 (Figure 2). Altogether, the results indicated that PchiSDcre and PchiSDcresig had higher activities than PchiS. In between, only PchiS was regulated by glucose.

3.2. Strain and Media Optimization for the Activity of PchiS and its Derivatives
      Since the b-galactosidase activities of strains 3NA pChi1, 3NA pChi2 and 3NA pChi3 were too low in LB, MI medium containing glycerol was used with the addition of different carbohydrates (glycerol, glucose and chitin). The β-galactosidase activity of each strain was measured with 2 h intervals up to 11 h (Figure 3A-C). Comparing to LB medium, the results indicated an increase in b-galactosidase activity. Moreover, the b-galactosidase activity of each construct was similar in different conditions, showing that the PchiS and its derivatives were constitutive in minimal medium with the defined composition. The higher b-galactosidase activities were obtained for 3NA pChi2 and 3NA pChi3 after 11 h of induction with about 800-1,100 MU.
      Plasmid pUP-Chi2Dcresig with the PchiSDcresig-chiS cassette was integrated into the genome of strain 3NA pChi1, 3NA pChi2 and 3NA pChi3 at amyE locus. The newly constructed strains Chi7, Chi8 and Chi9 were able to degrade chitin and use it as carbon source. The integration of PchiSDcresig-chiS into the amyE gene was confirmed using starch agar (Figure 4A-4F). As a result, strains 3NA pChi1, 3NA pChi2 and 3NA pChi3 showed a halo around their colonies resulted by starch degradation (negative controls (Figure 4A-4C), whereas strains Chi7, Chi8 and Chi9 showed no amylase activity and therefore, they showed no halo around the colonies (Figure 4D-4F). Likewise, expression of chitinase was confirmed using chitinase-Congo Red agar. The production of clear halos around the colonies of Chi7, Chi8 and Chi9 showed the chitinase activity compared to strains 3NA pChi1, 3NA pChi2 and 3NA pChi3 (Figure 4A¢-4F¢).
   Given that chitinase breaks the chitin branches increasing the amount of GlcNAc as carbon source, the activity of PchiS in strain Chi7 with the addition of different carbohydrates was investigated with 2 h interval up to 11 h (Figure 3D). Strain Chi7 showed identical activity with different carbon sources similar to 3NA pChi1. Overall, the b-galactosidase activity of Chi7 was 2- to 3-fold higher than 3NA pChi1.                  Additionally, the activities of PchiSΔcre and PchiSΔcresig from strains Chi8 and Chi9 in the same growth condition were measured. The b-galactosidase activities of both Chi8 and Chi9 in different carbohydrate sources were similar and greater than Chi7 (Figure 3D-4F). Comparison of Chi8 and Chi9 with their parental strains 3NA pChi2 and 3NA indicated that the integration of chiS enhanced the b-galactosidase activity in Chi9 (Figure 3B-F). Furthermore, the b-galactosidase activity of Chi9 was slightly increased in the presence of glycerol and chitin glycerol compared to chitin with (out) glucose reaching up to 1,800 MU (Figure 3F). Overall, the b-galactosidase activity was significantly increased by changing the medium to a defined medium. Moreover, the PchiS and its derivatives showed constitutive activity in the minimal medium.

4. Discussion
    Since chitin is an abundant polysaccharide in nature, it can be used as a cheap carbon source in growth media. Nevertheless, the genome of B. subtilis 3NA which was used to analyse the promoter in this study, contains no chitinase encoding gene to secret into the surrounding environment and sense the chitin and degrade it, the chitinase gene of B. pumilus SG2 plus its upstream promoter was inserted into the genome of this standard B. subtilis. promoter. Therefore, Chitinase encoding genes of B. pumilus SG2 are expressed by two promoters with a cre site in between. The activity of this promoter region is inducible by chitin and repressible by glucose in B. pumilus SG2, while the promoter activity is only repressible by glucose in B. subtilis (20). In the present study, the activity of chitinase promoter was investigated on a low copy plasmid using lacZ (encodes b-galactosidase), as a reporter gene. In addition to the wild type PchiS, two deletions were carried out in the cis-regulatory element of the PchiS to test the possible effect of these deletions on the promoter activity.
     Firstly, LB medium was used as the basal medium, which resulted in a weak activity of the wild type promoter in B. subtilis (max. 18 MU). Deletion of the cre site and its flanking region increased the activity slightly although the level of lacZ expression was low and constitutive (approx. 70 MU). This could be due to the shortened untranslated region of the lacZ mRNA on pChi2 and pChi3. Similarly, shortening of the untranslated region of lacZ mRNA expressed by PmtlA increased the b-galactosidase activity in the B. subtilis host strain (26). Replacing the LB medium with M9 medium led to weak growth of the strains (data not shown). Hence, MI minimal medium containing casamino acids was utilized as the basal media for the chitinases producing strains. In the latter medium, stronger activity of the chitinase promoter was observed. Comparison of the different growth conditions with the negative control showed that the activities of these promoters neither can be regulated nor was strong in B. subtilis. Besides, catabolite repression was not observed in the presence of glucose in minimal medium. Integration of the chitinase had increased the b-galactosidase activity in the presence of chitin and chitin + glucose in strain Chi7 compared with Chi1 as well as Chi3 compared with Chi9. B. subtilis contains no chitin degradation system (27, 28); however, it has only an N-acetylglucosamine utilization system (29). Therefore, the over-produced chitinase could provide further N-acetylglucosamine for the growth of the cells. Nevertheless, the increase of the b-galactosidase activity in the absence of chitin remains unclear.
In current study, B. subtilis selected as a host organism for recombinant protein expression. B. subtilis has significant potential for production of industrial enzyme including proteases, a-amylase and lipases (30). Using B. subtilis as an expression system has several advantages. The genome of this species has been sequenced. In addition, B. subtilis classified as a GRAS organism and without production of any harmful exotoxins or endotoxins. In addition, it could naturally secret recombinant proteins into the extracellular medium, which facilitate purification stage. Despite these benefits, using B. subtilis has several drawbacks including lack of suitable expression vectors (31-33). Until now several expression systems have been developed in B. subtilis. Among these expression systems are the starch-inducible amylase promoter, xylose-inducible xylA promoter, Pglv promoter, prophage-derived heat-inducible gene expression systems and B. subtilis expression system involving the mannose operon (34). In addition E. coli lac repressor-based expression system has been developed for B. subtilis that could produce very high levels of recombinant proteins after induction using IPTG (30, 34, 35).
      The chitinase promoter developed in this study, has potential to be used in an expression vector that could be induced by chitin. In addition, compared to the other inducers like IPTG and lactose, chitin is definitely cheaper and more available as an inducer