Secretory Expression, Functional Characterization, and Molecular Genetic Analysis of Novel Halo-Solvent-Tolerant Protease from Bacillus gibsonii

Abstract

A novel protease gene from Bacillus gibsonii, aprBG, was cloned, expressed in B. subtilis, and characterized. High-level expression of aprBG was achieved in the recombinant strain when a junction was present between the promoter and the target gene. The purified recombinant enzyme exhibited similar N-terminal sequences and catalytic properties to the native enzyme, including high affinity and hydrolytic efficiency toward various substrates and a superior performance when exposed to various metal ions, surfactants, oxidants, and commercial detergents. AprBG was remarkably stable in 50% organic solvents and retained 100% activity and stability in 0-4 M NaCl, which is better than the characteristics of previously reported proteases. AprBG was most closely related to the high-alkaline proteases of the subtilisin family with a 57-68% identity. The secretion and maturation mechanism of AprBG was dependent on the enzyme activity, as analyzed by site-directed mutagenesis. Thus, when taken together, the results revealed that the halo-solvent-tolerant protease AprBG displays significant activity and stability under various extreme conditions, indicating its potential for use in many biotechnology applications.

Introduction

As an important group of enzymes, proteases have received considerable attention as biocatalysts in various industrial, agricultural, horticultural, medical, and biotechnological applications [22]. The widespread use of microbial proteases in various processes is attributed to their cost-effective and short-term fermentation, relatively simple purification process, and excellent catalytic properties, including a low substrate specificity and high stability under various reaction conditions [12,15]. For practical applications, enzymes must be active under various extreme conditions. Thus, pursuing proteases with a high stability in the presence of metal ions, surfactants, oxidants, detergents, and organic solvents has attracted much attention [26]. Extensive efforts have recently been made to screen diverse proteases from microorganisms. As a result, several thermostable proteases have been identified in thermophilic microorganisms, such as Bacillus, Pyrococcus, Thermococcus, and Staphylothermus [5,8,10]. In addition, solvent-stable proteases have been found in Saccharopolyspora, Pseudomonas, Geomicrobium, Salimicrobium, and Bacillus [10,14,21]. However, there is still a need for the discovery of halo-tolerant, solvent-stable, and alkali-tolerant proteases, and microorganisms that can grow under high salt, solvent, or high alkaline conditions would seem to be excellent sources [4,13,26].

The enzyme production levels of wild-type producing strains are normally low, which is one of the bottlenecks for widespread application [24]. Although extensive efforts have already been made to screen strains with a high yield and improve strains by mutagenesis and genetic engineering [3,27], there is still no universal method for enzyme production that is suitable for different microorganisms, making such studies laborious and time-consuming. Target gene expression in an exogenous host has been proposed as a simple and effective method to overcome these problems, and B. subtilis is a promising expression system for protease production owing to its high-level protein expression and secretion [28]. Accordingly, in this study, a novel protease gene, aprBG, from an alkaliphilic strain B. gibsonii was cloned and expressed in B. subtilis. Both native and recombinant AprBG proteins were purified and characterized.

 

Materials and Methods

Materials, Bacterial Strains, and Growth Conditions

The B. gibsonii DSM8722 was obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ) and cultured in a TSB medium containing 17.0 g of casein peptone, 3.0 g of soya peptone, 5.0 g of NaCl, 2.5 g of KH2PO4, 2.5 g of glucose, 4.2 g of NaHCO3, 5.3 g of Na2CO3, 1.0 g of MnSO4·H2O, and 1.0 g of MgSO4 per liter. B. subtilis strain WB600 was used as the bacterial host for the secretory expression of the protease gene, and was grown in a 2YT medium (pH7.0) containing 10.0 g of yeast extract, 16.0 g of tryptone, 10.0 g of NaCl, and 0.2 g of MgSO4 per liter. The Bacillus strains were cultured at 30℃.

Cloning of Protease-Encoding aprBG Gene

A 467-bp-length fragment of the protease gene was amplified and sequenced using the degenerate primers WY168 and WY169 (Supplemental Table S1). A set of inverse primers WB744 and WB745 (Table S1) was designed according to the determined DNA sequence, and used to obtain the complete ORF sequence of the aprBG gene. The genomic DNA of strain DSM8722 was digested with HindIII or XbaI, and the DNA self-ligated at 4℃ for 18 h after column purification. The ligation mixtures were used as the template DNA for the inverse PCR (IPCR) using the primers WB744 and WB745 (Supplemental Fig. S2). Two amplified fragments with lengths of 803 bp and 1,436 bp, respectively, were purified from the gel and ligated to a pMD19-T vector (TaKaRa, Dalian, China). The complete nucleotide sequence of the aprBG gene was submitted to the GenBank database with the accession number of KC954649.

Plasmid Construction for Recombinant Protein Expression

To construct an E. coli-B. subtilis shuttle vector able to secrete the recombinant protein, a kanamycin resistance gene, the constitutive promoter P43, the engineered levansucrase signal sequence (SPSacB), and the replication origin in B. subtilis were all amplified by a PCR using pWB980 and pDG148 as the templates [29]. The PCR products were then inserted into a pUC19 vector to generate pWYB029 and pWYB055, which carried a different spacer sequence between the P43 and the ATG start codon of SPSacB. WB771 and WB772 (Table S1) designed with HindIII and SalI restriction sites were used to amplify the pro-peptide sequence of the aprBG gene. The point mutations were all introduced by overlapping the PCR using primers WB1072, WB1073, WB1074, WB1075, WB1076, and WB1077 (Table S1).The resulting PCR fragments were cloned into pWYB029 or pWYB055. All the recombinant plasmids harboring the wild-type and mutated aprBG genes were verified by sequencing before being transformed into B. subtilis WB600 cells by electroporation [30]. The positive clones designated as WT, D/A, H/A, and S/A carrying the wild-type and mutated aprBG genes were verified using hydrolysis halos on LB plates containing 1% casein [1].

Production and Purification of Proteases

To produce the native and recombinant proteases (designated AprBGW and AprBGR, respectively), the wild-type producing strain DSM8722 and recombinant strains were cultured in TSB and LB media, respectively, for 12 h as seed cultures, and then transferred to 50 ml of media and incubated with shaking at 30℃.

After cultivation for 36-42 h, the cell-free supernatants were collected by centrifugation and the proteins precipitated using 30-60% ammonium sulfate. The precipitate was collected by centrifugation after incubation overnight at 4℃, dissolved in buffer A (containing 0.01 M NaH2PO4 and 0.01 M Na2HPO4, pH 6.0), and dialyzed overnight against the same buffer to remove any residual ammonium sulfate. The crude extract was then run through a Resource S high-performance column (6.4 × 30 mm diameter) that had been pre-equilibrated with buffer A at a flow rate of 1 ml/min, and eluted using a linear gradient (0-1 M) of NaCl in the same buffer. The active fractions were concentrated using ultrafiltration (10 kDa MW cut-off membrane; Millipore, Glostrup, Denmark) and loaded onto a Superdex 75 10/300 GL high-performance column (10 × 300 mm diameter). The column was then equilibrated with buffer A and the fractions were eluted at a flow rate of 0.8 ml/min. The active fractions were pooled and stored at 4℃ for further analysis. The protein content of each chromatographic fraction was measured using the absorbance at 280 nm with an ÄKTA purifier system (Amersham Biosciences, Piscataway, NJ, USA). In addition, the protein concentration was measured with a BCA protein assay kit (Pierce, Rockford) using bovine serum albumin (BSA) as the standard.

Protein N-Terminal Sequencing

The purified proteases on a 12% acrylamide gel were blotted electrically onto a polyvinylidenedifluoride (PVDF) membrane. The N-terminal sequences were then determined using an automated protein sequencer (ABI Procise 492cLc Protein Sequencing System, Applied Biosystems, USA).

Enzyme Activity Assay

The protease activity was assayed in 50 mM glycine–NaOH (pH 9.5) using 1% azocasein as the substrate, and the activity quantified by measuring the absorbance at 366 nm. One unit of enzyme activity was defined as the amount of enzyme that catalyzed the formation of 1 μmol of azodye per minute. The detailed procedure was performed as previously reported [3].

Determination of Proteolytic Properties

The catalytic properties of AprBGW and AprBGR were determined in a reaction mixture containing 1% azocasein as the substrate under the various experimental conditions described below.

Effects of temperature and pH. The optimal temperature was determined by analyzing the enzyme activity in 50 mM glycine–NaOH (pH 9.5) in the temperature range of 30-90℃. The optimal pH was determined at 75℃ using buffers with different pHs [3]. The relative activity was calculated as a percentage of the maximal activity. The thermal and pH stabilities of the enzyme were assayed by determining the residual activity after incubating 10 μl of the purified enzyme at 30-60℃ for 0-120 min or at a pH of 5.0-12.0 for 12 h. The initial activity was taken as 100%.

Effects of metal ions and inhibitors. The effects of various metal ions on the enzyme activity were determined by incubating the reaction in mixtures containing 10 mM CaCl2, MgCl2, BaCl2, MnCl2, NiSO4, CoCl2, FeSO4, or CuSO4. To evaluate the effects of inhibitors, a soybean trypsin inhibitor (SBTI), β-mercaptoethanol (ME), urea, dithiothreitol (DTT), guanidine hydrochloride (GnHCl), ethylene glycol bis(2-aminoethyl) tetraacetic acid (EGTA), ethylenediaminetetraacetic acid (EDTA), and PMSF were used. The purified enzyme was pre-incubated with each inhibitor for 30 min at room temperature before the residual activity was measured. The enzyme activity incubated under identical conditions, but in the absence of a metal ion or inhibitor, was defined as 100%.

Effects of NaCl and organic solvents. To determine the influence of NaCl on the protease stability, AprBG was incubated in 0-4 M NaCl solutions for 24 h at room temperature before the residual activity was assayed. To determine the effects of various organic solvents, the purified enzyme was incubated with 10% or 50% (v/v) n-butanol, toluene, n-hexane, formamide, DMSO, ethanol, isopropyl alcohol, xylene, benzene, acetone, or ethyl acetate for 24 h. The enzyme activity assayed under identical conditions without NaCl or a solvent was defined as 100%.

Effects of surfactants, oxidants, and detergents. The effects of surfactants and oxidants on the AprBG stability were evaluated by incubating 10 μl of the purified enzyme in solutions of Triton X-100, Tween-20, -40, -60, and -80, sarkosyl, SDS, or H2O2 at room temperature before the residual activity was measured. Similarly, the effects of commercial detergents (Tide, Ariel, Keon, and Cnice) were also investigated to assess the potential application of AprBG in the laundry industry. The endogenous proteases of the detergents were inactivated by incubation in boiling water for 1 h. Next, 10 μl of the purified enzyme was mixed with each detergent at a final concentration of 1% (w/v) and the residual activity determined after incubation at room temperature for 1 h. The relative enzyme activity against the data obtained under the same conditions without a surfactant, oxidant, or detergent was then calculated.

Kinetic measurements. The kinetic parameters of the purified protease were determined with Lineweaver-Burk plots using the following natural, modified, and synthetic substrates: 0.07-1.20 mM casein, 0.4-40.00 mM azocasein, 0.075-1.500 mM AAPF, 0.2-1.0 mM AAPL, and 0.2-1.0 mM AAVA. The reaction was performed in 50 mM glycine–NaOH (pH 9.5) at 75℃ for 5 min after adding 0.002 mg of the purified protein into the reaction mixture. The specific activity was defined as the amount of enzyme that catalyzed the formation of 1 μM of product per minute. Thereafter, Kcat was calculated by dividing the value of Vmax with the quantity of the enzyme [19].

Sequence Analysis and Molecular Modeling

The amino acid sequence of AprBG was compared with other known proteins obtained from the EMBL/GenBank/DDBJ databases using available BLAST methods (http://www.ncbi.nlm.nih.gov/blast/). A conserved sequence alignment of AprBG and other related proteases was created using CLC Main Workbench software (ver. 6.8.1). Moreover, a phylogenetic tree was constructed with MEGA ver. 5.05 using the neighbor-joining method. Based on the crystal structures of the proteases Pron-Tk-Sp (PDB code, 3afgB) from Thermococcus kodakaraensis and BLS (PDB code, 1gciA) from B. lentus, the structures of the pro-AprBG and mature AprBG proteins were modeled using Swiss-Model (http://swissmodel.expasy.org/) and the quality of the final models was analyzed using QMEANclust (http://swissmodel.expasy.org/qmean/cgi/index.cgi).

 

Results

Sequence Analysis of aprBG Gene

The entire aprBG gene (1,128 bp) was amplified from B. gibsonii DSM8722 using a degenerate and inverse PCR. The positions of the primers used in this study are depicted in Supplemental Fig. S1. The complete gene sequence began with an ATG codon and ended with a TAA codon, which was deduced to encode 375 amino acids. A 19-residue signal peptide directing secretion was also identified using the SignalP 4.0 server [18]. When compared with other proteases in the EMBL/GenBank/DDBJ databases (Fig. 1A), the amino acid sequence of AprBG exhibited the highest identity of about 68% and 67% to AprS from Bacillus sp. G-825-6 and M-protease from B. clausii KSM-K16, respectively [22]. Additionally, AprBG showed a less than 57% similarity to AprB from Bacillus sp. B001 and other related proteases [2]. Therefore, AprBG is a novel protease that is most closely related to the high-alkaline proteases belonging to the Subtilisin family (Fig. 1B).

Fig. 1.Conserved amino acid sequence alignment (A) and phylogenetic tree (B) of AprBG and its closest related proteases. M-protease (Q99405) from B. clausii KSM-K16; YaB (P20724) from Bacillus sp. strain YaB; AprN (BAA25184) from Bacillus sp. B21-2; AprS (D29688) from Bacillus sp. G-825-6; AprB (GU136486) from Bacillus sp. B001; AH101 (D13158) from Bacillus sp. strain AH-101; ALP1 (Q45523) from Bacillus sp. strain NKS-21; Subtilisin Carlsberg (P00780) from B. licheniformis; rSAPB (AM748727) from B. pumilus CBS; Subtilisin BPN’ (Q44684) from B. amyloliquefaciens; KP-43 (AB051423) from Bacillus sp. strain KSM-KP43. Conserved active sites are indicated with “▲”. The scale bar represents 0-60% sequence divergence.

Heterologous Secretory Expression of aprBG

To increase the levels of enzyme production and simplify the purification process, the aprBG gene was expressed in the recombinant strain B. subtilis WB600. Two vectors, pWYB029 and pWYB055, which carried different spacer sequences between the strong promoter P43 and the ATG start codon of SPSacB, were constructed for the heterologous secretory expression of aprBG (Fig. 2A). Thus, while pWYB029 and pWYB055 both carried the same promoter P43, vector pWYB029 harbored a long spacer sequence of 2,754 bps, whereas pWYB055 carried a normal spacer sequence of 8 bps. The strain harboring pWYB029 produced a larger proteolytic halo on the agar plate, corresponding to a 3-fold increase of protein production when compared with the strain carrying pWYB055 (Figs. 2B and 2C). Therefore, the secretion of protease AprBG was more efficiently expressed in the recombinant strain with the newly constructed vector pWYB029.

Fig. 2.Genetic maps of two E. coli-B. subtilis shuttle expression vectors and their production levels for expressing protease genes. (A) Plasmids pWYB029 and pWYB055. (B) Expression level of aprBG gene cloned into pWYB055 (aprBG-055) and pWYB029 (aprBG-029). (C) Proteolytic halos formed by recombinant strains harboring aprBG-055 and aprBG-029. B. subtilis WB600 cells harboring pWYB055 were used as controls.

Purification and N-Terminal Amino Acid Sequence of AprBG

The AprBGW and AprBGR enzymes were both purified using ammonium sulfate precipitation, followed by cation exchange and gel filtration chromatographies. The purity and molecular mass of the enzymes were visualized by SDS-PAGE as a single band of 27 kDa (Supplemental Fig. S2). The production and recovery yield of the recombinant AprBG (Lanes 6 and 9 in Fig. S2) in B. subtilis WB600 were both higher than those in the original host strain (Lanes 1 and 4 in Fig. S2). The first 10 amino acid residues of AprBGW and AprBGR were determined to be QTVPWGITRV for both proteins, confirming the purity and consistency of the two enzymes (Supplemental Fig. S3). According to the N-terminal sequence of the mature AprBGW and AprBGR, the AprBG protein contained a 19-residue signal peptide, an 88-residue pro-sequence, and a 268-residue mature protein. The molecular mass of AprBG was calculated to be 27.3 kDa, which matched the sizes of the purified AprBGW and AprBGR enzymes (Fig. S2).

Kinetic Parameters of AprBG Toward Different Substrates

As shown in Table 1, AprBGW and AprBGR both exhibited a high affinity and hydrolytic efficiency toward all the substrates examined in this study. According to the Km values, the substrate preference was casein > AAPF > AAVA > AAPL > azocasein. When casein was used as the substrate, the Kcat/Km values of AprBGW and AprBGR were 2.6 × 105 min-1 mM-1 and 2.3 × 105 min-1 mM-1, respectively. The catalytic efficiency exhibited by both enzymes toward the natural substrate was clearly higher than that exhibited toward the modified and synthetic substrates. Proteases from Bacillus sp. B001 and B. pumilus CBS have been reported to exhibit a higher affinity and catalytic efficiency toward modified and synthetic substrates, respectively [2,12]. Therefore, the higher affinity and catalytic efficiency exhibited by AprBG toward the natural substrate suggest potential applications in various biotechnological processes [5].

Table 1.aAprBGW and AprBGR are AprBG proteases from wild-type producing strain DSM8722 and recombinant strain, respectively.

Effects of Temperature and pH on AprBG Activity and Stability

AprBGW and AprBGR were both determined to be active over a broad range of temperatures from 30℃ to 90℃ and over a range of pHs from 5 to 11, with the optimal activity at 75℃ and pH 9.5 (Figs. 3A and 3C). In addition, AprBGW and AprBGR were both stable up to 60℃, retaining above 90% and 50% activity after incubation for 30 and 120 min, respectively (Fig. 3B). Both enzymes also showed relatively high stability in a broad range of pHs from 5 to 12, and retained more than 80% activity after 12 h of incubation (Fig. 3D). Therefore, these results suggest that AprBG is a thermostable enzyme and could be used to improve the reactions performed in industrial applications related to food, paper, detergents, and drugs [8].

Effects of Various Inhibitors and Metal Ions on Enzyme Activity

The effects of various inhibitors and metal ions on AprBGW and AprBGR are shown in Table 2. Fe2+ did not significantly affect the activity of either enzyme, whereas Cu2+ had a strong inhibitory effect. Ca2+ increased the activity of both enzymes by 7-10-fold. Mg2+, Ba2+, Mn2+, Ni2+, and Co2+ also increased the enzyme activities by 203%, 152%, 136%, 125%, and 118%, respectively. Neither enzyme was inhibited or only slightly inhibited by SBT1, ME, DTT, GnHCl, urea, and EGTA. Meanwhile, EDTA was identified to inhibit both enzymes, suggesting that metal ion(s) are necessary for the protease activity of AprBG [26]. Furthermore, AprBG may be a serine protease with a serine residue in its active site, as the protease was strongly inhibited by PMSF [9].

Table 2.aAprBGW and AprBGR are AprBG proteases from wild-type producing strain DSM8722 and recombinant strain, respectively.

Effects of Various Surfactants, Oxidants, and Detergents on Enzyme Stability

AprBGW and AprBGR were found to be remarkably stable under various conditions (Table 3). Both enzymes remained extremely stable in the presence of different non-ionic surfactants. After incubating with 5% of Triton X-100 and Tween-20, -40, -60, and -80 for 72 h, the enzymes exhibited an enhanced residual activity of 114-207%. Moreover, AprBGW and AprBGR retained 130% and 96% activity in 1% sarkosyl, respectively. After incubation with an oxidant for 0.5 h, AprBGW and AprBGR retained 82% and 88% activity, respectively. To determine the compatibility of these enzymes with detergents, 1% commercial laundry detergents with different compositions were used. AprBGW and AprBGR remained highly stable in Ariel, retaining 100% and 111% of their initial activity, respectively. Furthermore, the two enzymes retained about 64-85%, 51-78%, and 78-79% activity in the presence of Cnice, Tide, and Keon, respectively.

Table 3.aw/v; bv/v; cAprBGW and AprBGR are AprBG proteases from wild-type producing strain DSM8722 and recombinant strain, respectively.

Effects of NaCl on AprBG Activity and Stability

The AprBGW and AprBGR proteases showed a high activity and stability with various salt concentrations. Both proteases displayed approximately 110-130% activity when the reaction buffer contained 1-4 M NaCl, with the optimum activity at 2-3 M (Fig. 3E). AprBGW and AprBGR also retained about 105-138% activity after incubation in 1-4 M salinity for 24 h and displayed the highest stability at 3 M (Fig. 3F). Moreover, the results confirmed that the AprBGW and AprBGR proteases consistently exhibited excellent functional activity and stability at salt concentrations as high as 4 M NaCl (equal to 23.4%).

Fig. 3.Effects of temperature, pH, and NaCl on AprBGW and AprBGR activity and stability. (A) Effects of temperature on enzyme activity. (B) Stability of enzymes after incubation at 30℃, 40℃, 50℃, and 60℃ for 0-120 min. (C) Effects of pH on enzyme activity. (D) pH stability of enzymes after incubation at pH 5-12 for 12 h. (E) Effects of NaCl on enzyme activity. (F) Stability of enzymes after incubation in 0-4 M NaCl solution for 24 h. AprBGW and AprBGR are AprBG proteases from wild-type producing strain DSM8722 and recombinant strain, respectively.

Effects of Various Organic Solvents on AprBG Stability

Organic solvents with log P values between -1.51 and 3.5 were selected to evaluate their effects on AprBG. As shown in Table 4, AprBGW and AprBGR were both remarkably tolerant to organic solvents at concentrations of 10-50% (v/v) after incubation for 24 h. Both enzymes retained more than 100% of their original activity in the presence of xylene, benzene, ethyl acetate, and isopropyl alcohol. In fact, their residual activities in the presence of the above solvents increased to 103-125%, 113-136%, 108-139%, and 110-141%, respectively. Furthermore, AprBGW and AprBGR displayed 93-123%, 93-111%, 95-112%, 98-125%, and 95-112% residual activity in the presence of n-hexane, toluene, n-butanol, acetone, and ethanol, respectively. When exposed to polar solvents, AprBGW and AprBGR retained about 85-99% and 87-117% activity in the presence of DMSO (log P = -1.49) and formamide (log P = -1.51), respectively. This noticeable tolerance of AprBG to various organic solvents makes the enzymes very attractive for peptide synthesis to achieve a high yield of peptides [17].

Table 4.aAprBGW and AprBGR are AprBG proteases from wild-type producing strain DSM8722 and recombinant strain, respectively.

Structural Mechanism for AprBG Secretion and Maturation

The overall molecular structure of AprBG was compared with the proteases Pron-Tk-Sp (PDB code, 3afgB) from Thermococcus kodakaraensis and BLS (PDB code, 1gciA) from B. lentus. The potential catalytic triads were deduced to be Asp119, His168, and Ser321 and located closely within the 3D structure of the pro-AprBG and mature protein (Fig. 4). As shown in Fig. 4, the residues Gln107 and Gln108, as the end residue of the pro-sequence and start residue of the mature protein, respectively, were close to the catalytic center.

Based on the structural homology model, the peptide bond between the pro-peptide and the mature protein may have been hydrolyzed by the active sites. The 88-residue pro-sequence was cleaved, and AprBG folded into a mature protease composed of 9 α-helices and 9 β-strands (Fig. 4).

Fig. 4.Structural model for the maturation process of AprBG. The 3D structural model image was prepared using the PYMOL program.

To investigate the effect of the predicted active sites on the secretion process of AprBG, D/A, H/A, and S/A variants, in which the catalytic Asp119, His168, and Ser321 had been separately substituted with Ala, were cloned into pWYB029 and expressed in B. subtilis WB600. The precursor and mature proteases were first analyzed by determining the protease activity on casein plates and in a cell-free supernatant. The recombinant WB600 expressing the AprBG wild-type gene exhibited high protease activity in the supernatant (Fig. 5A) and produced a large proteolysis halo on the casein plate (Fig. 5B), indicating that AprBG can efficiently synthesize and secrete in an extracellular milieu. Meanwhile, no evident protease activity was detected in the recombinant WB600 producing the D/A, H/A, and S/A variants, indicating that the formation of the active AprBG was dependent on the three catalytic sites. An SDS-PAGE analysis further confirmed that neither the pro-peptide nor the mature protein of the three mutant proteases was detected in the supernatant. Therefore, these results suggest that the mutant proteins were not sufficiently secreted or may have caused intracellular/extracellular degradation of the overexpressed mutant protease that was incorrectly folded (Fig. 5C). Therefore, the failure to secrete the mutant AprBG suggests that the secretion process of extracellular proteases is dependent on the catalytic center.

Fig. 5.Secretion and activation process of AprBG in vivo. (A) Level of active protease produced by various recombinant strains expressing wild-type (WT) and mutant (D/A, H/A, and S/A) aprBG genes in culture medium. (B) Phenotypes of various recombinant strains grown in casein plates. (C) SDS-PAGE analysis of supernatant fractions from various recombinant strains. B. subtilis WB600 cells harboring pWYB029 were used as controls. Arrows indicate pro-AprBG and mature AprBG.

 

Discussion

Bacterial proteases, especially those from Bacillus, represent the most important enzymes with commercial applications. Hence, extensive studies have been performed to discover novel (or improved) proteases in order to provide alternatives with superior performance under harsh application conditions. In the present study, a novel halo-solventtolerant protease AprBG produced by B. gibsonii DSM8722 was purified, characterized, and efficiently expressed in B. subtilis.

Based on a structural homology model and site-directed mutagenesis analysis, the AprBG precursor protein was synthesized as a pre-pro-peptide, secreted with the propeptide, and subsequently matured and activated by autoprocessing. In keeping with a common phenomenon among proteases, excluding trypsin, chymotrypsin, and elastase, the pro-sequence of AprBG also acted as an intramolecular chaperone and directed the efficient secretion and correct folding of the protein [25].

Using a newly constructed pWYB029 vector harboring a long sequence between the promoter and the ATG start codon of the signal peptide (SP) (Fig. 2A), an efficient secretion and a high-level production of the AprBG protease were achieved by the recombinant strain. The elevated expression of the protease was attributed to the genetic insulator effect of the spacer, which was also observed by Mutalik et al. [16], who developed a standard junction between the promoter and the gene of interest to achieve precise and reliable gene expression in E. coli.

Extensive biochemical studies of the purified native and recombinant enzymes revealed that AprBG exhibited superior performance under various conditions. In contrast with proteases that are inhibited by Ba2+, Mn2+, Ni2+, and Co2+, the activity of AprBG was significantly increased by most of the tested metal ions [6,10,26]. According to existing literature, the optimum temperatures for most known microbial proteases are below 70℃, with normal stability below 60℃ [5]. Only a few proteases from thermophilic and hyperthermophilic bacteria and archaea are stable above 70℃ [8,20]. However, AprBG exhibited its optimal activity at 75℃ and pH 9.5, with a high stability up to 60℃ and pH range of 5-12. Therefore, the thermoalkali-stability of AprBG indicates its potential application as a detergent additive, which requires stability across a wide range of pHs [24].

The potential industrial application of AprBG was further investigated by assessing its stability in the presence of various surfactants and oxidants [10]. In most of the tested agents, AprBGW and AprBGR were both very stable (retaining more than 100% activity) in non-ionic surfactants and the detergent Ariel. Similar results have also been observed for the detergent-stable proteases from B. mojavensis A21 and B. circulans, which retained above 94% activity in Ariel [6,7,20]. Nevertheless, SAPB from B. pumilus CBS was reported to be less compatible with Ariel than AprBGW and AprBGR, retaining only approximately 67% of its initial activity [11]. Therefore, the high resistance of AprBG to detergent formulations indicates its potential for industrial application, especially in commercial detergents [15].

Enzymes with halotolerant properties are usually isolated from halophilic bacteria. For example, a protease from a haloalkaliphilic species, AH-6, is active in a range of 0-4 M salt, yet only displays 75% residual activity with 4 M NaCl after incubation for 4 h [4]. Several other proteases from Bacillus have also recently been reported to be halotolerant and stable with 0-10% NaCl, but their activity and stability decrease with an increasing concentration [13,26]. Moreover, the SAPB protease has been reported to show enhanced residual activity from 58% to 105% in 2 M salt after mutagenesis [12]. However, when compared with these proteases, AprBG is a highly halotolerant enzyme that displays above 100% activity in the presence of (or after incubation in) 0-4 M salinity (Fig. 3). As salt is one of the key components in the granulation of enzymes for detergent additives, this halotolerant property of AprBG makes it especially useful as a detergent additive [10,12].

The log P values of various organic solvents define the polarity of different solvents. That is, the lower the log P value, the greater the polarity and toxicity of the solvent [21]. Thus, when considering the degree of disruption related to the polarity of a solvent, a solvent with a log P < 4.0 is considered to be extremely toxic [10]. Although proteases from B. subtilis DM-04 and Saccharopolyspora sp. A9 have already been found to be tolerant to organic solvents, they only showed residual activities of 68-110% and 35-99%, respectively [19,21]. A protease from B. licheniformis RSP-09-37 has also been reported to be tolerant to acetonitrile, hexane, and isooctane, but lost its activity in the presence of DMSO [23]. Therefore, the stability of AprBG is superior to most other reported proteases, except for BHAP from B. horikoshii [13]. Thus, when including its highly halotolerant property, AprBG is a rarely reported halo-solvent-tolerant enzyme.

When compared with the biochemical properties of other reported proteases, AprBG is stable in the presence of various organic solvents, detergents, oxidants, and detergents and displays superior performance at high temperatures and across a wide range of pHs. Moreover, AprBG is remarkably functional and stable in the presence of 0-23.4% NaCl, which is better than previously reported proteases (Table S2). Therefore, these properties make AprBG potentially useful for industrial bioremediation processes, including the treatment of wastewater contaminated with carbohydrates or the amelioration of salt marshes polluted with organic solvents [14].

In conclusion, a novel halo-solvent-tolerant protease produced by B. gibsonii DSM8722, AprBG, was purified, characterized, and efficiently produced in B. subtilis. The deduced protein exhibits a 57-68% identity with the highalkaline proteases of the subtilisin family. Based on a structural homology model and site-directed mutagenesis, the secretion and maturation mechanisms of AprBG were further addressed. The kinetic parameters toward different substrates revealed that AprBG displays a high affinity and hydrolytic efficiency toward natural substrates. The biochemical characterization of the native and recombinant AprBG enzymes consistently showed superior performance with various metal ions, surfactants, oxidants, and commercial detergents, especially in the presence of high salinity and various solvents. Thus, when taken together, the attractive properties of AprBG are of considerable interest for diverse applications in the future.