Identification and Characterization of a New Alkaline Thermolysin-Like Protease, BtsTLP1, from Bacillus thuringiensis Serovar Sichuansis Strain MC28

Abstract

Thermolysin and its homologs are a group of metalloproteases that have been widely used in both therapeutic and biotechnological applications. We here report the identification and characterization of a novel thermolysin-like protease, BtsTLP1, from insect pathogen Bacillus thuringiensis serovar Sichuansis strain MC28. BtsTLP1 is extracellularly produced in Bacillus subtilis, and the active protein was purified via successive chromatographic steps. The mature form of BtsTLP1 has a molecule mass of 35.6 kDa as determined by mass spectrometry analyses. The biochemical characterization indicates that BtsTLP1 has an apparent K_m value of 1.57 mg/ml for azocasein and is active between 20℃ and 80℃. Unlike other reported neutral gram-positive thermolysin homologs with optimal pH around 7, BtsTLP1 exhibits an alkaline pH optimum around 10. The activity of BtsTLP1 is strongly inhibited by EDTA and a group of specific divalent ions, with Zn²⁺ and Cu²⁺ showing particular effects in promoting the enzyme autolysis. Furthermore, our data also indicate that BtsTLP1 has potential in cleaning applications.

Introduction

Thermolysin-like proteases (TLPs) comprise a group of bacterial and fungal extracellular zinc metalloproteases [1, 11, 25, 34] that share high level of sequence and structural homology. Named after the prototypic thermolysin (E.C. 3.4.24.27) from Bacillus thermoproteolyticus [26, 32], these enzymes are classified as members of the M4 family of metallopeptidases [28]. Bacterial TLPs are initially synthesized as inactive precursors consisting of a N-terminal signal sequence that is cleaved off during secretion, a propeptide that assists the protein folding and maturation, and a mature peptidase sequence [8, 18]. TLP homologs also contain the typical HEXXH motif at the C-terminal, and require one zinc ion for their catalytic activity and multiple calcium ions for stability [4, 8, 16,17].

A large number of TLPs have been identified as virulence factors that are associated with the pathogenicity of various pathogenic microorganisms, indicating their roles as potential drug targets. For example, Npr599 from B. anthracis has been proposed to directly degrade the host tissues and modulate its defense [3], and LasB from pathogenic Pseudomonas aeruginosa can induce vascular cell anoikis through simultaneous proteolysis of specific cell receptors [2]. Meanwhile, TLPs are also widely used in the food industry and one example is the application of thermolysin for the production of the artificial sweetener aspartame [11, 34]. Therefore, all the above indicate that TLPs have large potential in both therapeutic and biotechnological applications.

Biochemical characterizations of both gram-positive and gram-negative bacterial TLPs have uncovered a broad spectrum of different enzymatic properties [3, 5, 13, 14, 18, 19, 23, 24, 36]. For example, de Kreij et al. [5] reported that different Bacillus TLPs displayed specific activities toward both casein and tripeptide substrates. There is also considerable variation in temperature optimum among bacterial TLPs, which can range from 30℃ to 70℃ [5, 35, 36]. On the other hand, except for a few gram-negative alkaline TLPs, such as MprI from Alteromonas sp. strain O-7 [23], a predominant majority of characterized bacterial TLPs show neutral optimal pH between 6 and 8 [1, 3, 13, 18, 36]. Here, we report the identification and biochemical characterization of a novel alkaline TLP, BtsTLP1, from a recently sequenced Bacillus strain, Bacillus thuringiensis serovar Sichuansis strain MC28 [9]. This, to the best of our knowledge, is the first characterized alkaline TLP in grampositive bacterial species. In addition, the potential application of BtsTLP1 in liquid laundry cleaning is also addressed in this report.

 

Materials and Methods

Phylogeny Analysis of BtsTLP Homologs

The BLAST program was used to identify proteins in the NCBI non-redundant database that share over 85% protein sequence identity to the specific full-length BtsTLP. The CD-HIT program was used to select proteins, where any two would have sequence identity lower than 98%. The resulting proteins, along with nine reported bacterial TLPs, were used for the multiple sequence alignment and the subsequent phylogenetic analyses. With the full-length protein sequences, the multiple sequence alignment was built using Clustal X (ver. 2.0) [21]; and the neighbor-joining tree was constructed using MEGA 6 [29].

Protein Expression and Purification

The nucleotide sequence encoding the propeptide and the mature sequence of BtsTLP1 (BtsTLP1Δ1-36), codon-optimized for B. subtilis expression (Fig. S1), was synthesized and inserted into the expression cassette of the integration vector p2JM103 [33] by Generay (Shanghai, China) to afford plasmid construct p2JM103-BtsTLP1Δ1-36 (Fig. S1). Both p2JM103-BtsTLP1Δ1-36 and p2JM103 (the empty control) were then transformed into B. subtilis. The transformation and protein expression were performed according to a protocol previously described by Vogtentanz et al. [33]. Following the expression, the resulting culture supernatant was collected by centrifugation at 24,000 ×g for 1 h at 4℃.

BtsTLP1 was purified by two successive anion-exchange chromatography steps. The collected supernatant was concentrated and buffer exchanged into 20 mM Tris-HCl (pH 7.5), 1 mM CaCl2, and 10% (v/v) propylene glycol (buffer A), using a VivaFlow 200 ultrafiltration device (Sartorius Stedim, Goettingen, Germany). The concentrated broth was loaded onto a HiPrep Q XL 16/10 column (GE Healthcare, Pittsburgh, USA) pre-equilibrated with buffer A. After the sample loading, the column was washed with buffer A until the UV (A280) baseline became stable, and the bound proteins were then eluted using a linear 32 min salt gradient from 0 to 0.5 M NaCl in buffer A, at a flow rate of 5 ml/min. Fractions containing the target protein, as indicated by the SDS-PAGE analysis, were pooled, concentrated, and buffer exchanged into 20 mM Tris (pH 8.5), 1 mM CaCl2, and 10% (v/v) propylene glycol (buffer B). The resulting concentrate was then loaded onto a HiPrep DEAE FF 16/10 column (GE Healthcare) equilibrated with buffer B and purified using a similar procedure as above. Following the elution, BtsTLP1-containing fractions were pooled, concentrated via the 10K Amicon Ultra device (Merck Millipore Ltd., Darmstadt, Germany), and stored in 40% glycerol at -20℃. The protein concentration was estimated by the BCA assay.

Mass Spectrometry (MS) Analysis

Purified BtsTLP1 (30 μl, 10 mg/ml) was transferred to an Eppendorf tube and chilled Milli-Q water was then added to a total volume of 450 μl. The protein sample was acidified with 50 μl of 1 M HCl and was then treated with 75 μl of 50% (w/v) trichloroacetic acid for protein precipitation. The protein pellet was resuspended in 90% (v/v) ice-cold acetone to remove the acid and salt. After centrifugation at 14,000 rpm for 3 min, the supernatant was removed and the dried pellet was re-dissolved in 20 μl of 8 M urea. The sample (4 μl) was diluted with 26 μl of 50 mM ammonium bicarbonate and then digested with 10 μl of 30 ng/μl trypsin (T-6567; Sigma) or Asp-N (P-3303; Sigma), respectively, with incubation at 37°C overnight. The reaction was quenched by addition of 10 μl of 5 % (v/v) f ormic acid. The digested peptide mixture was diluted 10-fold in 0.1% (v/v) formic acid for LC-MS/MS analysis.

The LC separation system was a Thermo Scientific Accela HPLC system (Fisher Scientific, Pittsburgh, PA, USA) using a Phenomenex Jupiter 4 μ Proteo 90Å C12 column (150 mm × 1.0 mm) (Phenomenex, Torrance, CA, USA) at 30℃. The analysis was started with a gradient from 95% mobile phase A, 0.1% formic acid in a 3:97 (v/v) CH3CN/H2O mixture, to 40% mobile phase B, 0.1% formic acid in a 95:5 (v/v) CH3CN/H2O mixture, for 35 min, followed by a 5-min gradient from 40% to 95% of mobile phase B and held at 95% of mobile phase B for 10 min, at a flow rate of 50 μl/min. The sample (20 μl, in 0.1% formic acid) was injected using a Thermo Scientific Accela Autosampler system (Fisher Scientific). MS detections were performed on a Thermo Scientific LTQ Velos Pro ion trap mass spectrometer equipped with an electrospray ionization source. Positive-ion mode was used, with a full scan range of m/z from 250 to 2,000. MS/MS analysis was done using a data-dependent collision-induced dissociation fragmentation of the top nine most intense ions, with normalized collision energy of 35%, 2 m/z isolation width, activation Q of 0.25, and activation time of 10 ms. Dynamic exclusion was enabled, allowing a repeat count of 2 within 30 sec. LC-MS raw data were acquired using the Thermo Scientific Xcalibur 2.2 SP1 software package. Peptide matching and protein search against the internal database were performed using Thermo Scientific Proteome Discoverer 1.3 with Sequest as a search engine.

The N-terminal sequencing and MALDI-TOF analyses of purified BtsTLP1 were carried out by Shanghai Applied Protein Technology Co. Ltd. (Shanghai, China)

Enzyme Assays

The standard proteolytic assay (100 μl final volume) consisted of 25 mM Tris-HCl (pH 8), 7.5 mg/ml azocasein (Megazyme, Co. Wicklow, Ireland), and 1.6 μg purified BtsTLP1 (or Milli-Q water alone as the blank control). The reactions were carried out at 40℃ for 10 min, and then quenched with 100 μl of 10% (w/v) trichloroacetic acid. Then 120 μl of the resulting supernatant was transferred to a non-binding 96-well microtiter plate (Corning Life Sciences, Tewksbury, USA), and its absorbance at 440 nm (A440) was measured using a SpectraMax 190 reader (Bio-Rad, Hercules, USA). The caseinolytic activity of BtsTLP1, defined as net A440, was calculated by subtracting the A440 of the blank control from that of the enzyme; one unit of protease activity for BtsTLP1 and thermolysin (P1512, Sigma) corresponds to the 0.01 increment of net A440 per minute, per microgram of enzyme, under the assay condition (25 mM Tris-HCl, pH 8, 40℃).

Characterization and Kinetic Analyses of BtsTLP1

BtsTLP1 activity was assayed under different pH, different temperatures, or with different additives (e.g., metal ions and inhibitors). For optimal pH measurement, 1.6 μg of purified BtsTLP1 (or Milli-Q water alone as the blank control) was first mixed with 25 mM sodium acetate/HEPES/glycine buffer at different pH. Following the addition of azocasein (7.5 mg/ml), reactions were incubated at 40℃ for 10 min. The optimal temperature studies were performed using standard assay condition at different temperatures (30-70℃). The effects of specific additives were tested by first mixing the enzyme (or Milli-Q water alone as the blank control) and buffer with each additive on ice for 10 min. Azocasein (7.5 mg/ml) was then added, and the reactions were incubated at 40℃ for 10 min.

The catalytic activity of BtsTLP1 (0.2 μg) was determined at 40℃ for 10 min in 25 mM Tris-HCl (pH 8.0) with variable concentrations (0.18–11.25 mg/ml) of azocasein. The reaction velocity was indicated as the increase of A440 per minute, and the Km values were calculated from a plot of the reciprocal of the reaction velocity against the reciprocal of azocasein concentration.

BtsTLP1 Autolysis Test

The protease autolysis test (10 μl final volume) contained 8 μg of purified BtsTLP1 and 10 mM of each individual divalent ion (or Milli-Q water as control). The reactions were kept at 40℃ for 0 or 15 min, followed by the addition of 10 μl of 1 M HCl. Eight microliters of the resulting solution was subjected to SDS-PAGE analyses. The relative amount of the target protein on the SDSPAGE gel was estimated by Image Lab software (Bio-Rad, Hercules, USA).

Microswatch Analyses of BtsTLP1

The tested fabrics, E-116 and P-S-38, were purchased from Center for Testmaterials (The Netherlands), with each microswatch placed in each well of a 96-well microtiter plate. The commercial detergent (Tide, Clean Breeze; Proctor & Gamble, USA) was preheated at 95℃ for 1 h before being diluted in 5 mM HEPES buffer (pH 8.2) to a final concentration of 0.788 g/l. Assays were initiated by adding 200 μl of diluted detergent in the presence or absence of a specific amount of purified BtsTLP1 (or Milli-Q water alone as the blank control) to each microswatch well. After 20min incubation in an incubator at 32℃, with agitation at 1,150 rpm, 120 μl of the resulting solution from each well was transferred into a fresh plate and its absorbance at 600 nm (for E-116) or 405 nm (for P-S-38) was measured using a SpectraMax 190 reader. The protease subtilisin Carlsberg (P8038, Sigma) was used for comparison purposes.

Sequence Accession Number

The protein sequences for BtsTLP1 and BtsTLP2 have been deposited in the GenBank database under the accession numbers AFU15727 and AFU16748, respectively. The synthesized nucleotide sequence encoding BtsTLP1Δ1-36 has been deposited in the GenBank database under the accession number KP280067.

 

Results

Identification of Thermolysin-Like Proteases (TLPs) from Bacillus thuringiensis Serovar Sichuansis Strain MC28

A BLAST search using the amino acid sequence of fulllength thermolysin (UniProt Accession No. P00800) led to the identification of two TLPs from Bacillus thuringiensis serovar Sichuansis strain MC28 (GenBank accession numbers AFU15727 and AFU16748, herein named BtsTLP1 and BtsTLP2, respectively). The full-length BtsTLP1 or BtsTLP2 shares 42% or 58% sequence identity to thermolysin. Both are classified as members of the peptidase family M4 in the MEROPS database [28]. Similar to other bacterial TLPs, both BtsTLP1 and 2 were predicted by SignalP 4.1 to be extracellular proteases with the first 36 aa and 27 aa respectively, as their putative N-terminal signal sequences [27]. Sequence alignment between the BtsTLPs and thermolysin (Fig. 1) indicated that in addition to the Nterminal signal peptide, each BtsTLP also has a pro-region that shares low identity to the propeptide of thermolysin, followed by a conserved C-terminal peptidase domain containing the HEXXH motif.

Fig. 1.Comparison of the full-length protein sequence between BtsTLP1 (GenBank Accession No. AFU15727), BtsTLP2 (GenBank Accession No. AFU16748), and thermolysin (UniProt Accession No. P00800). Protein sequence alignment was performed using Clustal X (ver. 2.0) [21]. The signal sequence, propeptide, and the mature sequence of thermolysin [32] are shown in italic, lowercase, and capital letters, respectively. The conserved HEXXH motif is underlined. The zinc binding sites [16], the proposed active sites, and thermoylisn-Ca(4) binding sites [16, 17] in the mature sequence of thermolysin are marked by solid triangles, circles, and squares, respectively.

In addition, the phylogeny analysis of BtsTLPs and their corresponding homologs clearly separated these two groups of enzymes into distinct evolutionary clades. As shown in Fig. 2, BtsTLP2 was grouped with thermolysin and other reported gram-positive TLPs, such as Npr599 from B. anthracis [3] and TLP-cer from B. cereus [5], implying that this enzyme may share similar properties with those neutral metalloproteases. However, the function and biochemical characteristics of BtsTLP1 and its homologs are still ambiguous. Therefore, to gain a better understanding about BtsTLP1, the recombinant enzyme was expressed in B. subtilis and its properties were studied.

Fig. 2.Neighbor-joining tree of BtsTLP homologs. Percentage bootstrap values of 500 replicates, with a value larger than 90, are given at the corresponding branch point. Branch lengths are shown to scale. For each protein, its GI number or GenBank accession number, as well as the related species information are provided. The BtsTLPs described in this study and nine previously reported bacterial TLPs are marked by solid square and triangles, respectively. TLP-cer (GI#1709341), TLP-ste (GI#135736), thermolysin (GI#93141324), Neutral protease B (GI#730171), and TLP-sau (GI#8928563) were reported by de Kreij et al. [5]. Npr599 (GI#30260755), BVMP (GI#45736603), SSMP (GI#89111269), and MprI (GI#18143405) were described by Chung et al. [3], Kim et al. [19], Hatanaka et al. [13], and Miyamoto et al. [23], respectively.

Active Mature BtsTLP1 Is Extracellularly Produced in Bacillus subtilis

A clearly visible protein band (Fig. 3A, lane 3, marked by the arrow head) was detected by SDS-PAGE analysis of the crude broth from B. subtilis cells expressing recombinant BtsTLP1, which was absent in the broth of cells expressing empty vector (Fig. 3A, Lane 2). The expressed protein was subsequently purified by successive anion-exchange chromatography steps (Fig. 3A, Lane 4) and the proteolytic activities were detected in both the purified and the corresponding crude samples but not in the control broth (Fig. 3B).

Fig. 3.Expression and initial characterization of recombinant BtsTLP1. (A) SDS-PAGE of crude broth collected from B. subtilis expressing control empty vector (lane 2) or recombinant BtsTLP1 (lane 3), and the purified BtsTLP1 resulted from successive anion-exchange chromatographs (lane 4, marked by arrow). (B) Proteolytic activities of crude broths and the purified BtsTLP1. Assays were carried out at 40℃ for 10 min, by mixing an aliquot of each protein sample (10 μl of each crude broth or 8 μg of purified BtsTLP1) with 7.5 mg/ml azocasein in 100 μl of 25 mM Tris-HCl (pH 8) buffer.

The identity of the purified protein was determined by LC-MS/MS (Fig. S1) to be the conserved C-terminal peptidase domain of the full-length BtsTLP1, suggesting that its pro-region, like those in other bacterial TLPs, was cleaved off during maturation. N-terminal sequencing further revealed the mature protein’s first four amino acids to be ITGF. Combining data obtained from LC-MS/MS, Nterminal sequencing, and the sequence alignments, the mature sequence of BtsTLP1 was thus confirmed (from ITGF to VGVK, Fig. S1). The resulting theoretical molecular mass was calculated to be 35,650 Da, consistent with the MALDI-TOF measurement (35,650 Da). Therefore, we concluded that the mature BtsTLP1 was extracellularly produced in B. subtilis as an active protease.

Biochemical Characterization of BtsTLP1

With azocasein as the substrate, the specific activity of BtsTLP1 was calculated to be 3.29 ± 0.03 U, which is around 10 times lower compared with thermolysin (34.77 ± 0.95 U). The enzyme was shown to be active between 20℃ and 80℃ (Fig. 4A), with a temperature optimum around 60℃. Interestingly, unlike other well-characterized gram-positive TLPs [3, 13, 18, 36], BtsTLP1 displayed an alkaline pH optimum around 10 and could retain at least 75% of its activity in the pH range of 9.5-11 (Fig. 4B). Kinetic studies suggested that the apparent Km of BtsTLP1 for azocasein was 1.57 ± 0.23 mg/ml. In addition, the finding that BtsTLP1 was strongly inhibited by metal-chelating agents (EDTA), but not by common serine or cysteine protease inhibitors (PMSF or E-64, respectively) (Table 1), strongly supported our hypothesis of BtsTLP1 as a metalloprotease.

Fig. 4.Effects of temperature and pH on BtsTLP1 activity. (A) The activity of recombinant BtsTLP1 was analyzed at different temperatures in 25 mM Tris-HCl (pH 8), and was reported as the relative activity where the activity at the optimal temperature was set to be 100%. Each value is the mean of triplicate reactions. (B) The activity of purified BtsTLP1 was analyzed in 25 mM sodium acetate/HEPES/glycine buffer at different pH, and was reported as the relative activity where the activity at the optimal pH was set to be 100%. Each value is the mean of at least triplicate reactions.

Table 1.aValues represent the mean ± SD of three experiments.

The effects of various metal ions on the catalytic activities of BtsTLP1 were further explored. As shown in Table 1, BtsTLP1 was strongly inhibited by Ni2+ and Cu2+. In addition, it lost half of its activity in the presence of 2 mM exogenous Zn2+, Co2+, Mg2+, or Mn2+. In contrast, incubation with Ca2+ resulted in a slight increase of BtsTLP1’s proteolytic activity (Table 1). Next, the purified BtsTLP1 was mixed with each of the metal ions at 40℃ for 15 min before being subjected to SDS-PAGE analysis. Whereas no significant difference was observed following its incubation with Ca2+, Mg2+, or Ni2+ (Fig. 5, Fig. S2), about 25-30% of the target protein was degraded in the presence of exogenous Cu2+ and Zn2+ (Fig. 5B). The Zn2+-dependent BtsTLP1 autolysis was later corroborated in additional time-course experiments (Fig. S2). In the case of Co2+ and Mn2+, however, the electrophoresis was unsuccessful, as both ions interfered with protein migration on the gel (data not shown). Collectively, our data suggest that certain divalent ions could destabilize BtsTLP1 and facilitate its autolysis.

Fig. 5.Effects of selected divalent ions on the stability of purified BtsTLP1. (A) SDS-PAGE of divalent-ion-treated BtsTLP1. The degradation fragments with a size around 21 or 15 kDa are marked with arrow heads. The experiment was repeated more than three times and a representative gel is shown. (B) Percentage of intact BtsTLP1 following the divalent ion treatment. The protein amount of control sample at time zero was set to be 100%. Each value is the mean of at least triplicate experiments.

BtsTLP1 May Have Potential Application in Liquid Laundry Cleaning

Microbial alkaline proteases have been commercially exploited in an extensive array of detergent formulations as an active protein-degrading ingredient [20]. Owing to its alkaline pH optimum, we therefore conducted a performance evaluation of BtsTLP1 in liquid laundry cleaning by using the microswatch assay described in US patent US2010/0192985(A1). A sample of subtilisin Carlsberg, a wellknown protease used in multiple industrial applications including cleaning [10], was also tested here for comparison purpose. Two different types of microswatches, E-116 (cotton soiled with blood/milk/ink) and P-S-38 (aged egg yolk with pigments on polyester), were applied as test materials. As shown in Fig. 6, despite its lower performance compared with subtilisin Carlsberg, the purified BtsTLP1 could clearly enhance the cleaning performance of the detergent on both microswatch stains, suggesting that BtsTLP1, or a variant of it, might serve as a suitable protease for detergent applications.

Fig. 6.Cleaning performance of BtsTLP1 and subtilisin Carlsberg in liquid laundry detergent on microswatches E-116 (A) and P-S-38 (B). Each value is the mean of six replicates. The dashed line represents the cleaning performance of the blank control (Milli-Q water alone).

 

Discussion

Our study described the identification of two TLPs (BtsTLP1 and 2) from insect pathogen Bacillus thuringiensis serovar Sichuansis strain MC28 as well as the functional characterizations of BtsTLP1, a novel alkaline gram-positive TLP.

The Bacillus thuringiensis MC28 strain was isolated from Mu Chuan virgin forest located in south China [9] and has been reported to be highly toxic to lepidopterous and dipterous insects [9, 30, 31]. Although several novel insecticidal spherical crystal proteins have been identified [30, 31], it is still ambiguous whether other types of proteins, such as TLPs, could be involved in the pathogenicity of strain MC28. Interestingly, BLAST analyses against the NCBI database returned no protein sequence with 60% or greater identity to each full-length BtsTLP in non-pathogenic Bacillus species such as Bacillus subtilis (data not shown), whereas BtsTLP homologs with over 85% identity were widely distributed among Bacillus cereus, Bacillus anthracis, and other Bacillus thuringiensis strains (Fig. 2). Such results led us to hypothesize the role of BtsTLPs as virulence factors of MC28, which is partially supported by the close phylogenetic relationship between BtsTLP2 and Npr599, a reported pathogenic factor from B. anthracis [3] (Fig. 2). Clearly, additional mutagenic studies would be necessary to further elucidate the biological roles of BtsTLPs.

It is widely accepted that an excess amount of Zn2+ or Cu2+ can inhibit the proteolytic activities of thermolysin homologs. Although the exact mechanism was unclear, a few studies suggested that binding to these divalent metal ions could alter the conformation of the wild-type enzyme. For example, a 3D structural analysis of thermolysin saturated with exogenous Zn2+ indicated that in addition to the catalytic Zn2+, a second Zn2+ could interact with the enzyme side chains, thereby excluding the substrate from the active site [15]. Similarly, Fukasawa et al. [7] proposed that Cu2+ substitution of the active-site Zn2+ in thermolysin homologs might increase the rigidity of the catalytic site, limiting its interaction with both the substrate and solvent H2O. It is very likely that such mechanisms could also account for the BtsTLP1 inhibition, as the active sites are conserved within the thermolysin family (Fig. 1). Our data show that both Zn2+ and Cu2+ were capable of inducing BtsTLP1 autolysis, which could provide new insights into their inhibitory role. Taking a closer look at the autolysis profile of BtsTLP1, the presence of protein fragments around 21 and 15 kDa in both assays (Fig. 5A) resembles the degradation pattern of thermolysin following its treatment by Co2+ [12] and EDTA [6], where the cleavage site is proposed to be near the thermolysin-Ca(4) binding sites. Such an observation might raise the speculation that, similarly to the EDTA- or Co2+-promoted degradation of thermolysin, the Zn2+- and Cu2+-induced BtsTLP1 autolysis might also involve Ca2+ depletion, presumably by either metal ion displacement or conformation changes. This would be substantiated by the high sequence homology of thermolysin-Ca(4) binding sites between thermolysin and BtsTLP1 (Fig. 1). Furthermore, it is noteworthy that Cu2+-treated BtsTLP1 seemed to undergo a greater extent of degradation compared with its Zn2+-treated counterpart (Fig. 5A). However, our data (Fig. 5B) in fact found very similar amounts of the intact protease following both treatments. Therefore, the reason that the Zn2+-treated BtsTLP1 appeared to be “cleaner” was very likely due to its higher residual activities (Table 1). Apparently, further mechanistic elucidation would require unambiguous identification of the autolysis products as well as thorough understanding of the conformational dynamics of BtsTLP1.

Proteases are some of the most important industrial enzymes, and the majority of those exploited commercially belong to the serine proteases family [10, 22]. Recently, several alkaline metalloproteases have been reported [23, 34], which arouse our interests in studying the industrial applications of this particular group of enzymes. The microswatch data reported here suggest the potential application of BtsTLP1 in laundry cleaning. Meanwhile, considering its alkaline pH optimum and proteolysis activity, BtsTLP1, similar to the alkaline serine proteases, may also be applied in other industrial applications that require extreme pH conditions, such as leather treatment, specific peptide syntheses, and food or feed ingredient processing [10]. Additional efforts will be needed to further explore applications for this enzyme, realizing that its peak activity at elevated temperature (60℃) and elevated pH (10) could fill a need in industrial settings.