Nutrient broth and chemicals: Coomassie brilliant blue (CBB) R-250, 2-amino-2-(hydroxymethyl)propane-1,3-diol (Tris), Bradford reagent, bovine serum albumin (BSA) and acrylamide/bis-acrylamide solution were purchased from HiMedia (Mumbai, India). Low-molecular-mass marker was purchased from GeneDireX (Taoyuan, Taiwan). N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide (Suc-AAPF-pNA), azocasein, gelatin, 4-aminobenzamidine dihydrochloride, dithiothreitol, sodium dodecyl sulfate (SDS), glycerol, bromophenol blue, trichloroacetic acid (TCA), ethylenediaminetetraacetic acid (EDTA), Triton X-100, ammonium sulfate, sodium acetate, sodium phosphate, glycine, NaCl, NaOH, HCl, KCl, CaCl₂, CuCl₂, ZnCl₂ and MgCl₂ were purchased from Sigma-Aldrich, Merck (St. Louis, MO, USA).

The protease-producing bacterium, termed KB111, used in this study was previously isolated (13) from mangrove forest sediments in Thailand. It exhibited a clear zone surrounding the colonies on skimmed milk agar and exhibited the highest protease activity in liquid medium. KB111 was inoculated in 5 mL of nutrient broth containing 0.5 M NaCl and cultured in a rotary shaker (WDS28; PolyScience, Niles, IL, USA) at 180 rpm and 37 °C. After that, cells were harvested for 8 to 64 h. The absorption of the cultures was checked at 600 nm by a UV spectrophotometer (UV-2401PC; Shimadzu, Kyoto, Japan). The cell cultures were harvested at 14 000×g (U-320R; Boeco, Hamburg, Germany) for 5 min at 4 °C and the crude enzyme supernatant was collected for the protease assay using azocasein as protein substrate. The reaction mixture containing 30 µL of the crude enzyme, 270 µL of 50 mM Tris-HCl (pH=7) and 20 mg/mL azocasein was incubated at room temperature (25 °C) for 30 min. Aliquots were withdrawn to determine the activity, as described in the enzymatic activity section.

The genomic DNA of KB111 was extracted by Purelink^TM HQ mini plasmid purification (Invitrogen, Carlsbad, CA, USA) and the whole genome was sequenced by the Omics Sciences and Bioinformatics Centre, Bangkok, Thailand. The genomic DNA was processed according to Qiagen FX DNA library kit preparation (Hilden, Germany) and a library was sequenced by an Illumina MiSeq sequencer (San Diego, CA, USA) (paired-end 2×250 bp read length).

The crude protein supernatant of KB111 was collected for purification. The protease was precipitated by salting out using ammonium sulfate ranging from 40–50% (m/V) saturation. Then, the precipitate was recovered after centrifugation (U-320R; Boeco) at 14 000×g for 30 min, resuspended in 20 mM Tris-HCl (pH=8) and dialysed overnight against the same buffer. The protein solution was applied to a 5-mL diethylaminoethyl cellulose (DEAE) anion-exchange column (GE Healthcare, Marlborough, MA, USA) pre-equilibrated with 20 mM Tris-HCl (pH=8). The bound protein was subsequently eluted using a gradient of NaCl (0–1 M) in the same buffer. The protein fractions were collected, dialysed overnight against 20 mM Tris-HCl (pH=7) and then concentrated appropriately using a Centricon ultrafiltration unit (Millipore, Bedford, MA, USA) with a 10 kDa molecular cut-off. All purification stages were performed at 4 °C.

The protein sample was prepared by precipitation with 10% (by volume) TCA and dissolving the precipitate in SDS sample buffer, containing 50 mM Tris-HCl (pH=6.8), 2% (m/V) SDS, 0.1 M dithiothreitol, 10% (by volume) glycerol and 0.005% (m/V) bromophenol blue. After that, the solution was neutralized with 2 M NaOH and boiled for 5 min. The protein purity and molecular mass were determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with 12% (m/V) polyacrylamide gel and the protein band was visualized by incubating the gel with staining solution (CBB R-250). The concentration of the purified protein was measured by the method of Bradford (14).

The protein molecular mass was determined by gel filtration column chromatography using a Sephacryl S-200HR column (GE Healthcare). Aldolase (158 kDa), conalbumin (75 kDa), ovalbumin (44 kDa) and ribonuclease (13.7 kDa) were used as protein markers.

SDS-PAGE was performed using 12% (m/V) polyacrylamide gel containing 1 mg/mL gelatin. The sample was precipitated with 10% (by volume) TCA and boiled in SDS sample buffer for 5 min prior to loading onto the gel. Following electrophoresis, the gel was placed in 1% (by volume) Triton X-100 in phosphate-buffered saline for 2 h at room temperature (25 °C), then transferred into 50 mM Tris-HCl (pH=7) with 0.5 M NaCl for 10 h at room temperature (25 °C) and stained with CBB R-250. A clear zone at the location of the protease band was visualized because of the hydrolysis of gelatin.

Enzymatic activity was measured using a protein substrate (azocasein). The enzymatic reaction mixture containing 30 µL of an appropriate dilution of the enzyme, 270 µL of 50 mM Tris-HCl (pH=7), 0.5 M NaCl, and 20 mg/mL azocasein was incubated for 20 min. After that, 200 µL of 15% (by volume) TCA were added to stop the enzymatic reaction. The mixture was kept on ice for 10 min and centrifuged (U-320R; Boeco) at 14 000×g for 15 min. The supernatant (160 µL) was neutralized by mixing with 40 µL of 2 M NaOH. The absorbance of the resulting solution was measured (UV-2401PC; Shimadzu) at 440 nm. The enzymatic activity of the sample was determined in units (U), which was the amount of enzyme in the assay reaction mixture that increased the absorbance at 440 nm by 0.1 in 1 min.

Specific enzymatic activity was measured using a peptide substrate (Suc-AAPF-pNA). The enzymatic reaction mixture containing 2 µL of an appropriate dilution of the enzyme, 98 µL of 50 mM Tris-HCl (pH=7), 0.5 M NaCl and 2 mM Suc-AAPF-pNA was incubated for 20 min. The absorbance of p-nitroaniline released by enzymatic hydrolysis was measured at 410 nm with a molar absorption coefficient of 8900 M⁻¹ cm⁻¹. The enzymatic activity of the sample was determined in U, which was the amount of enzyme in the assay reaction mixture that produced 1 µmol of p-nitroaniline in 1 min. The specific activity of the enzyme was defined in U of the enzyme per milligram of protein.

The optimum pH was examined at various pH ranging from 4.5 to 10 at 30 °C. Different buffers (50 mM) with 0.5 M NaCl were used in the assay reaction mixtures: sodium acetate (pH=4.5–5.5), sodium phosphate (pH=5.5–7), Tris-HCl (pH=7–9) and glycine-NaOH (pH=9–10). The optimum temperature was examined by incubating the enzyme in 50 mM Tris-HCl (pH=7) with 0.5 M NaCl at various temperatures ranging from 20 to 80 °C. For the effect of NaCl on protease activity, the enzyme was dissolved in 50 mM Tris-HCl (pH=7) and different concentrations of NaCl ranging from 0 to 4 M. To determine the effect of metal ions, KB-SP was pre-incubated for 3 h at 30 °C with 50 mM Tris-HCl (pH=7) with different metal ions: KCl, CaCl₂, CuCl₂, ZnCl₂ and MgCl₂ (final concentration of 10 mM). Enzymatic activity was measured using Suc-AAPF-pNA and expressed in percentage of relative activity by taking the highest activity obtained at optimal pH, temperature and NaCl as 100% and by taking the control (without metal ions) as 100% for metal ions.

The kinetic parameters of the enzyme were assayed using Suc-AAPF-pNA of 0.5–4 mM as a substrate in 50 mM Tris-HCl (pH=7) with 2 M NaCl. The reaction followed the Michaelis–Menten kinetics, the v_max and K_m values were obtained using Lineweaver–Burk plots. The k_cat value was obtained by dividing the v_max value with the concentration of enzyme.

The pH stability of KB-SP was determined by pre-incubating the enzyme for 24 h at various pH ranging from 1 to 12.5 at 30 °C. Different buffers (50 mM) with 0.5 M NaCl were used in the reaction mixture for pre-incubation: KCl-HCl (pH=1–1.5), glycine-HCl (pH=2-3), sodium acetate (pH=4–5), sodium phosphate (pH=6), Tris-HCl (pH=7–9), and glycine-NaOH (pH=10–12.5). The residual activity of the enzyme was analysed at 30 °C using Suc-AAPF-pNA in 50 mM Tris-HCl (pH=7), with 2 M NaCl. To analyse the thermal stability of KB-SP, the enzyme was incubated in 50 mM Tris-HCl (pH=7), with 0.5 M NaCl at 40, 50 and 60 °C. The residual activity of KB-SP was analysed at 30 °C using Suc-AAPF-pNA in 50 mM Tris-HCl (pH=7) with 2 M NaCl. To determine the effect of EDTA treatment on its stability, KB-SP was incubated at 40 and 50 °C in 50 mM Tris-HCl (pH=7) with 10 mM EDTA and the residual activity was analysed at 30 °C using Suc-AAPF-pNA.

The purified protein, termed KB-SP, was excised as a single band from the SDS-PAGE gel and dehydrated. After drying, the presence of the amino acids on the gel was analysed using liquid chromatography–tandem mass spectrometry (LC-MS/MS) by First BASE Laboratories, Selangor, Malaysia. The sample was digested with trypsin and extracted into peptides following standard methods (15). Peptides were analysed using a nano HPLC-MS/MS system (Shimadzu), containing a triple time-of-flight (TOF) mass spectrometer model TripleTOF® 5600+ (Sciex, Framingham, MA, USA) via electrospray ionization (ESI) interface. The tryptically digested peptides were injected into an Agilent Zorbax 300SB-C18 column, 3.5 µm particle size (Agilent Technologies, Santa Clara, CA, USA) to separate peptides by linearly increasing the gradient of φ(formic acid, water)=0.1%/acetonitrile. Protein spectra were analysed and identified using Mascot sequence matching software (16) and the UniProt database (17) with the taxonomy of the genus Bacillus.

The amino acid sequence of KB-SP was used to generate the homology or comparative 3D modelling structure using the Expasy SWISS-MODEL server (18). For model building, the template was selected based on the highest quality of the QMEAN score. After modelling, the structural validation of the model was checked and analysed by PROCHECK using the PDBsum server (19), and the secondary structure of the protein was predicted using the ProFunc server (20). The 3D images of structures to analyse the location of the amino acid residues were generated using PyMOL software (v. 2.4) (21).

All experiments in this study were carried out in three replicates and the data represent the mean value±standard error. The SPSS software v. 22 (22) was used for statistical analysis by one-way ANOVA and differences in the mean values were subjected to Duncan’s multiple range test at 0.05 level (p≤0.05).

Strain KB111 was isolated from mangrove forest sediments in Chanthaburi, Thailand. The isolate was examined and further identified as a Gram-positive, rod-shaped bacterium, and its protease production was optimized on skimmed milk agar containing 30 mg/mL NaCl by a point inoculation technique and in liquid medium using azocasein as substrate. Based on 16S rRNA gene sequence analysis, the isolate KB111 was determined as being closely related to Bacillus sp. (13). Furthermore, the whole genomic DNA of KB111 was identified, and it displayed average nucleotide identity (ANI) of 99.68% with Bacillus licheniformis LMG 17339. Strain LMG 17339 was isolated from silage in Denmark by the Danish Veterinary Laboratory and was grown aerobically on nutrient agar or trypticase soy agar at 37 °C (23). Its genome contains 21 genes encoding proteases/peptidases (24). However, no research on purification and characterization of the proteases/peptidases in this genome has been reported. The B. licheniformis protease termed subtilisin Carlsberg was the first bacterial protease used in detergent, in the 1960s. More recently, a large group of commercial alkaline proteases have been derived from Bacillus strains, such as B. amyloliquefaciens, B. licheniformis and B. clausii. For example, Alcalase and Savinase from Novozymes are industrial alkaline proteases from B. licheniformis and B. clausii, respectively (25). Furthermore, carbohydrase and protease enzymes of B. licheniformis strains are listed in the fourth edition of the Food Chemicals Codex for application in food processing (26).

The relationship of KB111 protease activity with bacterium growth was examined during incubation at 37 °C for 8 to 64 h using azocasein as a substrate. Azocasein is a non-specific protease substrate and is normally used to determine the proteolytic activity of proteases. The result shows that KB111 exhibited the highest protease activity ((38.0±1.3) U/mL) after 56 h of incubation, at the end of the stationary phase, and it gradually decreased during the stationary phase (Fig. 1). This result corresponds to those for the haloalkaline protease from Bacillus sp. SM2014 (27) and the alkaliphilic strain B. pumilus MCAS8 (28). In addition, most Bacillus spp. exhibit their highest protease activity during the post-exponential and stationary phases (29, 30). These results indicate the role of the extracellular protease of this organism in survival during the stationary phase in the environment ecology (29).

The extracellular protease KB-SP was purified by the addition of ammonium sulfate and HiTrap DEAE (anion-exchange chromatography). The enzymatic activity in each step was determined using 2 mM Suc-AAPF-pNA at 30 °C in 50 mM Tris-HCl (pH=7). Suc-AAPF-pNA as a synthetic peptide peptidyl-pNA was used to determine the enzymatic activity because the extracellular serine protease and alkaline protease showed the highest specificity towards Suc-AAPF-pNA (31, 32). KB-SP was purified 8.6-fold with a (20.1±0.2) % yield and a specific activity of (15.0±0.2) U/mg in the final purification step (Table 1). The total mass of KB-SP was (2.8±0.1) mg purified from 1 L of culture. In SDS-PAGE analysis, the purified KB-SP showed a major activity band at 70 kDa by both CBB staining and zymography and some weak bands at 38 and 28 kDa in CBB staining (Fig. 2). KB-SP was denatured by TCA treatment and boiling in the presence of SDS prior to SDS-PAGE. However, KB-SP exhibited activity in the gel (zymography) probably because it is partially refolded after dialysis against 1% (by volume) Triton X-100 in phosphate-buffered saline. Protease activity in the gel was also observed in a previous study (33). By gel filtration column chromatography, the molecular mass of purified KB-SP was calculated to be 68 kDa, which corresponds to a major band on SDS-PAGE, suggesting that KB-SP is a monomer.

The purified KB-SP exhibited more than 50% of the highest activity at pH=6.5–10, the optimum being at pH=7 in 50 mM Tris-HCl. However, at pH=7 and 8 the activity was at 95% (Fig. 3a). Considering temperature dependence, KB-SP exhibited the highest activity at 50 °C, but it did not differ significantly from that at 40 °C. The protease exhibited approx. 60 and 70% of the highest activity at 30 and 60 °C, respectively (Fig. 3b). This result corresponds with a previous study in which the crude enzyme of isolate KB111 was determined, with its optimal pH=7.0 and temperature of 40 °C with azocasein as substrate (13). This result is similar to that for Neutrase 0.8L (Novozymes) with optimal activity at pH=7 and 40–50 °C (34).

Considering the effect of NaCl on protease activity, KB-SP exhibited high activity in the broad salinity range of 2–4 M NaCl and lost 80% of its activity in the absence of NaCl (Fig. 3c). This result suggests that KB-SP is a halotolerant protease. It has been reported in a previous study that isolate KB111 can grow at NaCl concentrations ranging from 0.5 to 1.5 M. It exhibits activity in the presence of 0–2 M NaCl (13), a range similar to those observed for the halotolerant B. licheniformis BA17 (0–3.4 M NaCl) (12) and B. licheniformis TD4 (0–2 M NaCl) (35). However, B. licheniformis BA17 and TD4 have not yet been studied to determine the effect of NaCl on the protease activity of the purified enzyme. The high activity of purified KB-SP in a broad salinity range was similar to that of the alkaline and halophilic proteases produced by B. luteus H11 at 1–5 M NaCl (3). Nevertheless, the effect of NaCl concentration on KB-SP activity was superior to that on the activity of other Bacillus spp. haloalkaline proteases because approx. 80% of the activity of KB-SP was retained at 1–4 M NaCl, while more than 80% of the activity of Bacillus sp. SM2014, Bacillus sp. APCMST-RS7 and B. iranensis X5B was retained at 0–2, 1–1.5 and 0.5–1.5 M NaCl, respectively (27, 36, 37). In a previous study, halophilic proteases were obtained from Bacillus spp., Halobacillus spp., Salinivibrio sp., Pseudoalteromonas sp., Filobacillus sp., Virgibacillus sp. and Chromohalobacter sp. (9). Halophiles can live in an environment with high salt concentration due to their adaptation mechanism which prevents the diffusion of NaCl into the cells and unique structural features of proteins to maintain their function under high-salt conditions (8, 10).

Considering the effect of metal ions, the enzymatic activity of KB-SP in the presence of the divalent cations Ca²⁺ and Mg²⁺ and monovalent cation K⁺ at a final concentration of 10 mM increased by 22, 25 and 30%, respectively, when compared to the control (Table 2). The activity of control was determined after dialysis against 10 mM EDTA. These results are similar to those reported for B. licheniformis MP1 and BA17 for each metal ion at 5.0 mM (32, 38). In addition, the enzymatic activity of B. licheniformis RP1 and P003 and halophilic proteases from Bacillus spp. such as B. luteus H11 and B. iranensis X5B had also been found to increase with the addition of Ca²⁺ and Mg²⁺ ions at a final concentration of 1.0–5.0 mM (3, 30, 37, 39). Nevertheless, ZnCl₂ and CuCl₂ had a negative effect on KB-SP activity. It was inhibited by 80 and 97% in the presence of Cu²⁺ and Zn²⁺ ions, respectively. These results are similar to those reported for B. luteus H11, B. licheniformis RP1, and Bacillus sp. APCMST-RS7 for each metal ion at 1.0 and 5.0 mM (3, 30, 36). The results of metal ion influence indicated that KB-SP could be used in shrimp waste powder because it contained 13.45, 0.58, 0.12 and 0.07% of Ca²⁺, Na⁺, Mg²⁺ and K⁺ ion, respectively (30), which markedly affected the increased KB-SP activity.

The kinetic parameters of KB-SP were determined at 30 and 50 °C using Suc-AAPF-pNA as a substrate. The v_max value of KB-SP at 30 and 50 °C was (23.5±1.1) and (47.7±2.9) U/mg, respectively (data not shown). The K_m value of KB-SP was rather similar at 30 °C (1.0±0.2 mM) and 50 °C (1.0±0.1 mM), indicating the same substrate binding affinity at both temperatures. However, the k_cat/K_m values of KB-SP at 30 °C ((25.6±2.4) s⁻¹ mM⁻¹) were considerably lower than at 50 °C ((54.1±2.3) s⁻¹ mM⁻¹) (data not shown). The results show that KB-SP was characterized by a low turnover number and catalytic efficiency at low temperatures.

The purified KB-SP was fully stable at pH=7–11, unlike at pH≤6 and ≥12. It lost >75% of its activity at pH≤3 and ≥12 (Fig. 4a). The effect of pH on the stability and activity of KB-SP was similar to the findings of previous studies on proteases from other bacteria having a pH=7–12 optimal for haloalkaline proteases (3, 27, 32, 40) and the alkaline protease belonging to B. licheniformis (38, 39, 41, 42). However, KB-SP was more stable at alkaline pH than the commercial enzymes subtilisin Carlsberg and Alcalase from B. licheniformis, which show pH stability at 8–10 and 6–10 (34), respectively.

The results for KB-SP against irreversible heat inactivation showed that it was stable for at least 3 h at 40 °C and it lost activity with a half-life of 90 and 10 min at 50 and 60 °C, respectively (Fig. 4b). This result confirms that KB-SP is active and stable in the mesophilic temperature range. The optimum temperature and thermal stability of KB-SP were found to be similar to those of the halotolerant B. licheniformis BA17 with a half-life of 90 and 12 min at 50 and 60 °C, respectively (32). In general, alkaline proteases from B. licheniformis and halophilic proteases from other bacteria remain active and stable at temperatures ranging between 40 and 70 °C (38, 41, 43, 44).

Considering the effects of EDTA treatment on stability, KB-SP lost approx. 40% of activity by incubation at 40 °C for 10 min and rapidly lost activity with a half-life of 5 min at 50 °C. These results suggest that the stability of KB-SP can be lost in the presence of EDTA (Fig. 4c). EDTA is a chelating agent that can bind to metal ions to form a complex in the molecule. Therefore, metal ions are required for the maximum stability of KB-SP; these ions tightly bind to the enzyme and are difficult to remove in the absence of EDTA.

The results of biochemical characterization indicate that KB-SP may be useful in various applications in industry at relatively high alkaline pH and salt concentrations. KB-SP can be applied in the food industry as a milk-clotting enzyme to produce milk curd (45) and protein hydrolysate using fish, meat and soy as raw material (46). Furthermore, it can be used in the feed industry as an additive that increase the protein digestibility of high-protein meal such as soybean (47), and for deproteinization of waste in the seafood industry (38).

To determine what type of enzyme it is, the purified KB-SP was incubated with the serine protease inhibitor, 4-aminobenzamidine dihydrochloride (final concentration of 10 mM) and its protease activity was determined by zymography. No activity band was detected by zymography, indicating that KB-SP belongs to the serine protease family. In addition, KB-SP retained more than 60% of its activity when incubated with 10 mM EDTA, a metalloprotease inhibitor (Fig. 4c), indicating that KB-SP is not a metalloprotease. After that, the amino acid sequence of purified KB-SP was identified by LC-MS/MS as RVVVPANQTGKA, RVTSVTVEPGAKQ, KGVAPEATLLAYRV, RGVIPFVDKAENAKN, i.e. with a high score matching with that of peptidase S8 (accession no. KUL11341) of B. licheniformis LMG 17339. This 2421 bp gene encodes 806 amino acid residues with a molecular mass of 86 kDa. The amino acid sequence of this gene contains a signal sequence, a propeptide and a mature domain, termed Prepro-KB-SP, and it was compared with those of the serine protease Vpr from B. subtilis (accession no. AAA22881), termed Prepro-BaVpr, in this study (Fig. 5 (48-50)). Based on amino acid sequence identities, Prepro-KB-SP displayed 72.21% similarity to Prepro-BaVpr. The signal sequences for protein secretion from the cell membrane were estimated for Prepro-KB-SP and Prepro-BaVpr using SignalP 5.0 (50) or experimentally determined for BssE (48) and SubC (49) and are underlined in Fig. 5. In a previous study, the propeptide BaVpr was determined as being composed of 132 amino acids and processed from a propeptide form upon cleavage of the peptide bonds between Gln132 and Met133 (51). These amino acids are also conserved in KB-SP, probably because KB-SP comprises a propeptide form (Pro-KB-SP, Thr1-Gln132) and a mature domain (Met133-Glu778). The mature domain was calculated from the amino acid sequence as having a mass of 68 kDa, which corresponds to the activity staining of the gel (zymography) of KB-SP with a molecular mass of ~70 kDa. Therefore, the purified enzyme may represent the mature domain of KB-SP. Furthermore, weak bands of 38 and 28 kDa were observed on SDS-PAGE (Fig. 2), corresponding to the previous result for BaVpr (51). These weak bands may represent the proteolytic fragment of the mature domain. The three residues of a serine protease that form a catalytic triad are completely conserved in the KB-SP sequence as Asp161, His205 and Ser516, and the asparagine residue that forms an oxyanion hole is also conserved as Asn295. Between the histidine and serine residues of the active site, KB-SP and BaVpr have an unusual 132 amino acid residue, longer than the other serine protease of B. subtilis (subtilisin E; BssE) and B. licheniformis (subtilisin Carlsberg; SubC) (Fig. 5). Vpr was identified as a minor extracellular protease with an Asp/His/Ser catalytic triad, which is similar to that found in the active site of subtilisin (51, 52). Collagen, fibrin, casein, gelatin and albumin can be used as protein substrates for Vpr (53). Vpr from B. subtilis and B. licheniformis has been identified as a fibrinolytic enzyme and a milk-clotting protease, respectively (45, 54). Vpr contains an N-terminal signal sequence, a propeptide, a catalytic domain (peptidase S8) and a C-terminal protease-associated (PA) domain (52). The PA domain is also conserved in KB-SP with 73.75% amino acid sequence similarity to BaVpr. According to a previous study, the long C-terminal PA domain has a high molecular substrate-binding specificity (53). The Vpr from Bacillus sp. has been found to have an optimal pH=7–8.5 and temperature of 37–50 °C (45, 52, 53), close to those of KB-SP at pH=7 and 50 °C. However, no information is available about the effect of NaCl on its activity. Therefore, this is the first report indicating that Vpr exhibits maximum activity in the presence of NaCl. However, further functional and structural studies are necessary to understand the autoprocessing, folding and substrate binding of KB-SP.

Due to the lack of information about the structure of Vpr in the Protein Data Bank (PDB) (19), the amino acid sequence of the mature domain (KB-SP) was used to generate a homology model using the SWISS-MODEL server (18). KB-SP displayed maximum sequence similarity of 26.68% to C5a peptidase (ScpA, a multidomain cell-envelope subtilase) from Streptococcus pyogenes [3eif.1.A]. Thus, ScpA was used as a template to predict the 3D structure of KB-SP. The model of the KB-SP structure was visualized and analysed using PyMOL software (21), as shown in Fig. 6a. The KB-SP structure consists of 13 β-sheets and 12 α-helices in the catalytic domain, 4 β-sheets and 3 α-helices in the PA domain, and 10 β-sheets and 1 α-helix in the C-terminal domain. Similar to all members in the serine protease family, KB-SP contains a catalytic triad consisting of three active site residues, Asp161, His205 and Ser516 (Fig. 6b). Folding of the main chain of the catalytic domain of KB-SP is related to that of the subtilisin domain with the subtilisin-like α/β domain, but the PA domain is not available in bacterial subtilisins. However, ScpA contains a PA domain with 8 β-sheets and 2 α-helices and has 37.66% amino acid sequence similarity to the PA domain of KB-SP.