Characterization of a Novel Thermostable Oligopeptidase from Geobacillus thermoleovorans DSM 15325

Abstract

A gene (GT-SM3B) encoding a thermostable secreted oligoendopeptidase (GT-SM3B) was cloned from the thermophile Geobacillus thermoleovorans DSM 15325. GT-SM3B is 1,857 bp in length and encodes a single-domain protein of 618 amino acids with a 23-residue signal peptide having a calculated mass of 67.7 kDa after signal cleavage. The deduced amino acid sequence of GT-SM3B contains a conservative zinc metallopeptidase motif (His⁴⁰⁰-Glu⁴⁰¹-X-XHis⁴⁰⁴). The described oligopeptidase belongs to the M3B subfamily of metallopeptidases and displays the highest amino acid sequence identity (40.3%) to the oligopeptidase PepF_Ba from mesophilic Bacillus amyloliquefaciens 23-7A among the characterized oligopeptidases. Secretory production of GT-SM3B was used, exploiting successful oligopeptidase signal peptide recognition by Escherichia coli BL21 (DE3). The recombinant enzyme was purified from the culture fluid. Homodimerization of GT-SM3B was determined by SDS-PAGE. Both the homodimer and monomer were catalytically active within a pH range of 5.0–8.0, at pH 7.3 and 40℃, showing the K_m, V_max, and k_cat values for carbobenzoxy-Gly-Pro-Gly-Gly-Pro-Ala-OH peptidolysis to be 2.17 ± 0.04 × 10^-6 M, 2.65 ± 0.03 × 10^-3 µM/min, and 5.99 ± 0.07 s^-1, respectively. Peptidase remained stable at a broad pH range of 5.0–8.0. GT-SM3B was thermoactive, demonstrating 84% and 64% of maximum activity at 50℃ and 60℃, respectively. The recombinant oligopeptidase is one of the most thermostable M3B peptidase, retaining 71% residual activity after incubation at 60℃ for 1 h. GT-SM3B was shown to hydrolyze a collagenous peptide mixture derived from various types of collagen, but less preferentially than synthetic hexapeptide. This study is the first report on an extracellular thermostable metallo-oligopeptidase.

Introduction

Oligopeptidases are endo-acting peptide bond hydrolases active against unstructured oligopeptides, but unable to cleave proteins [7]. Only oligopeptides of certain length, predominantly less than 30 amino acid residues, are able to reach the active center of oligopeptidases, due to the distinct structural architecture of these enzymes, tailored to block interaction with large macromolecular substrates [30]. Oligopeptidases are found in organisms from all domains of life. The substrate specificity varies within this subgroup of the endopeptidases, which enables oligopeptidases to ensure different biological functions.

According to the catalytic type, oligopeptide hydrolases are serine or metallopeptidases. The MEROPS peptidase database classification attributes serine-oligopeptidases to the family S9 of clan SC, whereas metallo-oligopeptidases are known in the M3 and M13 families of the clan MA (http://merops.sanger.ac.uk [32]). Serine-oligopeptidases (prolyl oligopeptidase and oligopeptidase B) were described as the catalyzers of the bioactive peptide peptidolysis in eukaryotes [8,36]. The homologs of both enzymes were also isolated and described from mesophilic bacteria. The present diversity of the bacterial metallo-oligopeptidases includes oligopeptidase A, oligopeptidase F, oligopeptidase O, and Pz-peptidase A [31]. These peptidases are zinc-dependent and similar in the sense of catalytic mechanism. The active site of these hydrolases demonstrates similar organization, consisting of characteristic motif His-Glu-X-X-His in conjunction with C-terminal glutamate. Both motif’s histidines and C-terminal glutamate act as zinc ligands, while the motif’s glutamate is a catalytic residue [13]. Oligopeptidase F (M03.007) and Pz-peptidase A (M03.010) are grouped together forming the M3 family’s M3B subfamily. Oligopeptidase A (M03.004) was found to have an evolutionarily closer relationship to mammalian thimet oligopeptidase; therefore, it was included in the M3A subfamily. The characterized oligopeptidases O form a distinct group (M13.004, M13.005, M13.009, and M13.010) within the M13 neprilysin family. This branch of metallo-oligopeptidases is relatively distant from M3 peptidases [32].

Almost all oligopeptide metallopeptidases described to date are monomeric and intracellular hydrolases. The biological necessity of metallo-oligopeptidases’ activity in bacteria is not well comprehended if compared with the understanding of homologous peptidases in eukaryotes [15,28]. Generally, metallo-oligopeptidases are described as the participants in intracellular catabolism of oligopeptides. This was extensively demonstrated for oligopeptidase O in lactic acid bacteria analyzing the final stages of casein turnover [33]. In fact, it is also thought that oligopeptidase A is important for the survival of Salmonella enterica ser. Typhimurium during the periods of carbon deficiency [38]. The significance of Pz-peptidases B and A from Geobacillus collagenovorans MO-1 for complete collagen catabolism was assumed, as both metallopeptidases were active against collagenous peptides [24]. Studies on in vivo substrate identification of a particular bacterial metallo-oligopeptidase suggest a role of oligopeptidase F in signal peptides degradation and protein secretion in Lactococcus lactis [22]. Meanwhile, oligopeptidase A is the major soluble enzyme in Escherichia coli, which is able to hydrolyze free lipoprotein signal peptide in vitro [27]. Overproduction of oligopeptidase F inhibits sporulation initiation in Bacillus subtilis as a result of increased peptidolysis of PhrA pentapeptide [15]. In addition to all the presented intracellular oligopeptidases, functions, the activity of intracellular oligopeptidase F in swine pathogen Mycoplasma hyopneumoniae was detected, outlining the possible role of this metallo-hydrolase as a part of the pathogens’ survival strategy [25].

Extracellular homologs of oligopeptidase O are characterized as virulence factors of Porphyromonas gingivalis 381 [3], Mycobacterium tuberculosis [30], and Streptococcus pneumoniae [1]. Secreted oligopeptidase F was reported from soil mesophiles Bacillus licheniformis N22 [4] and Bacillus amyloliquefaciens 23-7A [9]. The functionality of extracellular oligopeptidases from nonpathogenic mesophiles has never been investigated, and is only obscurely hypothesized as peptidolysis of collagenous peptides or extracellular processing of the Phr pheromone family propeptides [9].

Oligopeptidases are irreplaceable in the final stages of enzymatic preparation of precise length and amino acid sequence composition peptides from native proteins, especially hard-to-degrade animal proteins (collagen, keratin, and elastin) [16]. Metallopeptidases of oligopeptidolytic specificity from mesophiles are introduced for the production of high-value bioactive oligopeptides for medicine as well as cosmetics and are also used in the food industry [39]. Low thermostability and thermoactivity of mesophile oligopeptidases available for biotechnology are the limiting factors that have to be bypassed in order to decrease the cost and increase the efficiency of relevant oligopeptide bioproduction.

Putative oligopeptidases with signal sequences are constantly annotated in the sequenced genomes of thermophiles, but have never been described before. This increasing variety of in silico predicted extracellular metallo-oligopeptidases from thermophilic bacteria is a target of great potential for in-depth characterization, as it is expected that oligopeptidases produced by this type of extremophiles would be thermostable and thermoactive. The only studied oligopeptidases from thermophiles, G. collagenovorans MO-1 cytoplasmic Pz-peptidases B and A, are thermostable and thermoactive metallo-hydrolases [24]. Geobacilli remain an attractive source of biotechnologically beneficial enzymes [42]. Overall, three putative metallo-oligopeptidases were identified in all sequenced genomes of geobacilli. Two of these genes, with and without signal sequence, are homologous to oligopeptidase F, whereas the third gene encodes well-studied Pz-peptidase A [32].

In this study, characterization of a novel extracellular M3B subfamily metallo-oligopeptidase, designated as GT-SM3B, from the casein and collagen-degrading obligate thermophile Geobacillus thermoleovorans DSM 15325 (previously Geobacillus lituanicus DSM 15325T [10]) is presented. The oligopeptidase gene was directly cloned from genomic DNA of G. thermoleovorans DSM 15325. The target hydrolase was overexpressed and purified and is described in detail. The research focused on determination of high temperature impact on GT-SM3B characteristics. The hydrolytic activity of GT-SM3B against collagenous peptides was also explored. This study is the first report on an extracellular metallooligopeptidase from thermophilic bacteria and therefore provides previously unavailable opportunity to compare metallo-oligopeptidases secreted by mesophiles with the oligopeptidase from a thermophile.

 

Materials and Methods

Gene Cloning and Construction of the Recombinant Protein

The full-length GT-SM3B gene was amplified by PCR from G. thermoleovorans DSM 15325 genomic DNA using the following primer pair: M03007kaust1F (5’- CAT ATG GCT AGC ATG ARG CGC TGG C -3’) and M03007kaust1RRS (5’- TCG GAT CCG TCA TGT AGA CCG CTG ACG -3’). The oligopeptidase F gene sequences extracted from the genomic sequences of Geobacillus sp. C56-T3, Geobacillus sp. Y412MC52, Geobacillus sp. Y412MC61, Geobacillus kaustophilus HTA426, and Geobacillus thermoleovorans CCB_US3_UF5 were used for the construction of these primers. The primers were designed by using the PRIMERSELECT component of LASERGENE 6 (DNASTAR). A NheI site (underlined) was incorporated into the forward primer and a BamHI site (underlined) was incorporated into the reverse primer for cloning into pET-28c(+) (Novagen). The GT-SM3B gene was amplified in 50 µl of reaction mixture containing DreamTaq Green PCR Master Mix (2×) (Thermo Fisher Scientific), 0.25 µM of each primer, and 10 ng of genomic DNA. PCR was conducted under the following conditions: initial denaturation at 95℃ for 2 min followed by 29 cycles each consisting of 95℃ for 1 min, 60℃ for 2 min, and 72℃ for 3min, with a final extension step at 72℃ for 7 min. M03007kaust1F and M03007kaust1RRS primed PCR products were digested with NheI and BamHI (Thermo Fisher Scientific) according to the manufacturer’s instructions. The oligopeptidase gene was ligated by T4 DNA ligase (Thermo Fisher Scientific) into the pET-28c(+) vector introducing the N-terminal hexahistidine affinity tag. The ligation products were electroporated into competent Escherichia coli DH5α (Novagen) and the transformants were cultured on a solid Luria-Bertani (LB) medium (1% peptone, 0.5% yeast extract, 1% NaCl, and 1.5% agar) supplemented with kanamycin (30 µg/ml) at 37℃ overnight. The recombinant plasmid, validated by sequencing, was then transformed into Escherichia coli BL21 (DE3) (Novagen).

Sequence Analysis

The signal sequence and the cleavage site of the signal peptide were predicted by using the SignalP 4 server (http://www.cbs.dtu.dk/services/SignalP/ [29]). The search of sequences homologous to GT-SM3B was performed by Standard Protein BLAST, using the National Center for Biotechnology Information server (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The MEGA 6 program [37] was used for GT-SM3B sequence analysis and alignment. The identity percentage between GT-SM3B and homologous oligopeptidases was determined by the Jotun Hein method with the PAM250 residue weight table in the MEGALIGN component of LASERGENE 6 (DNASTAR). The multiple alignments of amino acid sequences of oligopeptidases were generated with the ClustalX 2 program [20] using default settings.

Optimization of Recombinant Oligopeptidase Production

E. coli BL21 (DE3) transformants harboring the expression construct were cultured in LB medium supplemented with kanamycin (30 µg/ml) at 37℃, with constant agitation (180 rpm) for 12 h. Overnight culture was used to inoculate up to 1% (v/v) fresh LB medium supplemented with kanamycin aiming to evaluate the initial conditions for GT-SM3B production. The inoculated culture was incubated under the same temperature and agitation until the OD600 reached 0.5. Then GT-SM3B expression was induced by 1 mM of isopropyl-β- D -thiogalactopyranoside (IPTG) for 4 h.

Different values of the principal overexpression parameters were considered for optimization in order to establish optimal conditions for heterologous soluble GT-SM3B production: E. coli BL21 (DE3) density at the point of induction OD600 0.5 and 0.8; expression temperatures, 25℃ and 30℃; final IPTG concentrations, 0.1, 0.5, and 1 mM; and duration of induction, 2, 4, and 12 h. LB medium and M9 medium with casamino acids 1% (w/v) were compared regarding the positive impact on soluble GT-SM3B production during the optimization. That was carried out using small-scale cultivation of 20 ml of expression cultures. Every expression trial was performed identically as the initial GT-SM3B production evaluation, differing only in a screened combination of target protein production conditions.

The expression of recombinant oligopeptidase was analyzed by SDS-PAGE and quantitative enzyme activity assay in the cell lysate soluble and insoluble fractions. Detection of secreted GT-SM3B in concentrated E. coli BL21 (DE3) secretome was also performed.

Purification of Secreted Enzyme

All purification steps were performed at room temperature, unless otherwise stated. Cells from culture fluid containing secreted GT-SM3B were clarified by centrifugation at 10.000 ×g for 30 min at 4℃. Culture supernatant proteins were salted out using (NH4)2SO4 at 4℃. The fraction precipitated from 30% to 80% of (NH4)2SO4 saturation was recovered by centrifugation, dissolved in buffer A (20 mM Tris–HCl, pH 8.4 at 20℃) and dialyzed against the same buffer at 4℃ for 12 h. The resulting dialysate was loaded onto an anion-exchange column UNO Q6 (12 × 53mm; Bio-Rad) previously equilibrated with buffer A. Proteins bound to the anion-exchange resin were eluted with one-step isocratic flow of buffer A supplemented with NaCl up to 1 M final concentration. Partially purified target oligopeptidase was collected in a flow-through fraction, which was dialyzed for 12 h at 4℃ against buffer B (20 mM MES–NaOH, pH 5.6 at 20℃). After dialysis, the protein sample was subjected to a cation-exchange chromatography step in a UNO S1 column (7 × 35 mm; Bio-Rad) pre-equilibrated with buffer B. The recombinant oligopeptidase was eluted with a linear 30 ml gradient of NaCl (0–1 M) in buffer B at a flow rate of 1 ml/min. Finally, fractions containing GT-SM3B were pooled and dialyzed for 12 h at 4℃ against oligopeptidase sample buffer (50 mM HEPES–NaOH, pH 7.3 at 60℃).

Quantitative Assay of Oligopeptidase Activity

The activity of GT-SM3B was quantitatively assayed by hexapeptide carbobenzoxy-Gly-Pro-Gly-Gly-Pro-Ala-OH (Fluka 27673) peptidolysis. The assay mixture (60 µl) in 50 mM HEPES–NaOH buffer, pH 7.3 at 60℃, contained 2 µM of substrate and 0.5 µg of recombinant oligopeptidase. After the incubation at 60℃ for 10 min with shaking (350 rpm), the hydrolysis was terminated by performing the ninhydrin reaction [34] with slight modifications according to Zhang et al. [43]. Then, the amount of liberated Gly-Pro-Ala was determined spectrophotometrically by measuring the absorbance at 570 nm. Oligopeptidase activity unit was defined as the amount of enzyme that liberated 1 µM of product per minute under optimal conditions. The concentration of liberated product, resulting from peptidolysis by GT-SM3B, was calculated according to a standard curve prepared using known concentrations of Gly-Pro-Ala (Fluka 27674) tripeptide. The assay mixture, immediately applied to ninhydrin reaction without incubation, served as an activity assay control.

Kinetic Analysis

The GT-SM3B kinetic parameters Vmax and Km with carbobenzoxyGly-Pro-Gly-Gly-Pro-Ala-OH as a substrate were determined from initial rate measurement at ten different substrate concentrations between 0.05–5 Km. The liberation of peptidolysis product was monitored at optimal pH and temperature as previously specified for quantitative activity assay. The data were analyzed and fitted to the Michaelis-Menten equation using the GraFit 7 program (Erithacus Software).

Assay of Hydrolytic Activity Against Collagenous Peptides

Mixtures of collagenous peptides were derived from gelatin (from porcine skin; MP Biomedicals), type I collagen (from calf skin; Sigma-Aldrich), and raw collagen extracts (Proteina). Native proteins were dissolved in 100 mM acetic acid at 4℃ for 12 h and dialyzed overnight at the same temperature against 50 mM HEPES–NaOH buffer, pH 7.3 at 60℃ prior to usage. GT-SM3B hydrolytic activity against collagenous peptides was measured analogically as previously specified for quantitative peptidolysis assay.

Determination of Protein Concentration

The protein concentration was estimated with a Coomassie (Bradford) Protein Assay Kit (Thermo Fisher Scientific) following the manufacturer’s recommendations, with bovine serum albumin as the standard. The purified GT-SM3B concentration was determined at 280 nm using the molar extinction coefficient 92.600 M-1 cm-1.

Characterization of Temperature and pH Impact on the Activity and Stability of GT-SM3B

Peptidolytic activity was measured using the synthetic hexapeptide as a substrate in 50 mM HEPES–NaOH buffer, pH 7.3 at temperatures ranging from 10℃ to 95℃ in order to determine the effect of temperature on GT-SM3B activity. The thermostability of GT-SM3B was investigated after incubation of the enzyme solutions in the absence of substrate at temperatures 60℃, 70℃, and 80℃ for 1 h in 50 mM HEPES–NaOH buffer, pH 7.3. Residual peptidolytic activity was measured after the incubation at 60℃. Consequently, the results of these measurements were compared with the initial GT-SM3B activity at 60℃. The influence of different pH values on the activity of GT-SM3B was tested at 60℃ by measuring peptidolytic cleavage in 50 mM sodium acetate (pH 4.0–5.0), 50 mM MES–NaOH (pH 5.0–7.0), 50 mM HEPES–NaOH (pH 7.0–8.0), and 50 mM glycine–NaOH (pH 8.0–10.0) buffers. The pH stability of GT-SM3B was described by determining residual activity after enzyme pre-incubation in the absence of substrate for 1 h at 40℃ in buffers of tested pH values in a range of pH 4.0–10.0 with the specified hexapeptide peptidolysis assay at 60℃.

Effects of Metal Ions, Chelators, and Other Chemical Compounds on Oligopeptidase Activity

The pure enzyme samples were pre-incubated for 30 min at 40℃ with metal chlorides Li+, Na+, Mg2+, K+, Ca2+, Mn2+, Fe3+, Co2+, Ni2+, Cu2+, Zn2+, Sn2+, and Ba2+, at 1 mM. The influence of K+, Ca2+, and Na+ on GT-SM3B activity was also tested, applying a concentration of 10 mM, whereas 350 mM concentration for Na+ was also tested. Pre-incubation with purified recombinant oligopeptidase under the same conditions was used to determine the effect of various chemical compounds on GT-SM3B activity. These chemicals were from different compound groups: chelators (ethylenediaminetetraacetic acid (EDTA), ethylene glycol tetraacetic acid (EGTA) at 1 mM), reducing agents (dithiothreitol (DTT), 2-mercaptoethanol (2-ME), Tris(2-carboxyethyl)phosphine (TCEP) at 1 mM), detergents (Tween 20, Tween 40, Tween 60, Tween 80, Triton X-100, Triton X-305 at 0.1% (v/v) and 1% (v/v); Brij 35, Brij 58 at 1 mM; sodium dodecyl sulfate (SDS) at 1 mM and 10 mM), and chaotropic agents (guanidine-HCl at 1 mM; urea at 1 mM and 10 mM). The impact of glycerol (at 1% (v/v) and 5% (v/v)), ethylene glycol (at 0.1% (v/v) and 1% (v/v)) as well as phosphoramidon (at 1 mM) on enzyme activity was also tested. The GT-SM3B activity, applying the described peptidolytic assay at 60℃, was determined after enzyme pre-incubations. Reaction mixtures after pre-incubation with certain chemical compounds at tested concentration, but without incubation at 60℃ after supplementation with hexapeptide substrate, served as a control.

Sodium Dodecylsulfate-Polyacrylamide Gel Electrophoresis (SDSPAGE)

The oligopeptidase’s purity and molecular weight were assessed by 12% glycine SDS-PAGE according to Laemmli [19] under nonreducing conditions at room temperature. The cleavage patterns of collagenous peptides from gelatin, type I collagen, and raw collagen extracts produced by GT-SM3B hydrolysis were also investigated via the indicated electrophoresis technique. SDS-polyacrylamide gels were stained with PageBlue Protein Staining Solution (Thermo Fisher Scientific).

Gelatin Zymography

Zymography was performed in 12% SDS-polyacrylamide gels with copolymerized gelatin (2 mg/ml) under nonreducing conditions without cooling [17]. After the electrophoresis, SDS was washed out using 50 mM HEPES–NaOH, pH 7.3 at 20℃ with 0.1% (v/v) Triton X-100. The detergent was removed by extensive washing with 50 mM HEPES–NaOH, pH 7.3 at 60℃. Zymograms were incubated at 60℃ for 12 h in the incubation buffer (50 mM HEPES–NaOH, pH 7.3 at 60℃). The active bands were visualized by staining with PageBlue Protein Staining Solution (Thermo Fisher Scientific).

Statistical Analysis

All measurements were repeated three times, and average means are presented ± standard deviation. Data were considered as statistically significant for P values of 0.05 or less.

Sequence Accession Number

The nucleotide sequence of gene GT-SM3B from G. thermoleovorans DSM 15325 has been deposited in GenBank under Accession No. KF779146.

 

Results

GT-SM3B Sequence Analysis

The amplification using primers M03007kaust1F and M03007kaust1RRS resulted in a PCR product of about 1,900 bp in length. The sequence of the GT-SM3B gene was 1,857 bp in length, coding a protein of 618 amino acids in length. The molecular mass was calculated to be 70.2 kDa, while the molecular mass of the enzyme without the 23-residue signal sequence was calculated to be 67.7 kDa.

GT-SM3B was the most identical to putative oligopeptidase F with predicted signal peptides of other geobacilli. A comparison of the amino acid sequences revealed 99.2% identity of GT-SM3B to oligoendopeptidase F (protein ID AEV19421.1) of G. thermoleovorans CCB_US3_UF5 and 98.7% identity to oligoendopeptidase F (protein ID ADI26721.1) of Geobacillus sp. C56-T3. The GT-SM3B sequence differed from sequences of G. thermoleovorans CCB_US3_UF5 and Geobacillus sp. C56-T3 oligopeptidases F in five and eight amino acids, respectively. Both proteins indicating such high scores in identity with GT-SM3B are derived from sequenced genomes of geobacilli and only in silico annotated as putative oligoendopeptidases with signal peptides homologous to oligopeptidase F attributed to the M3B subfamily of proteolytic enzymes. It should be noted that, besides oligopeptidase F with signal sequence, the second oligopeptidase F is found in all sequenced genomes of geobacilli. The latter proteins lack signal peptides and actually demonstrate higher identity to holotype oligopeptidase F from L. lactis [32]. The holotype oligopeptidase F is an intracellular hydrolase. To date, Pz-peptidase B from G. collagenovorans MO-1 is the only biochemically described intracellular oligopeptidase F from geobacilli [24], while others remain putative with predicted activity according to sequence homology. GT-SM3B showed the highest 40.6% identity to putative intracellular oligopeptidase F (protein ID ADI27671.1) of Geobacillus sp. C56-T3 and was the least identical, with 38.9% identity, to Geobacillus sp. G11MC16 analogously defined intracellular oligopeptidase F (protein ID EDY07555.1).

GT-SM3B demonstrated equivalent identity (Fig. 1) to both biochemically characterized secreted oligopeptidases F from B. amyloliquefaciens 23-7A [9] and B. licheniformis N22 [2]. It shared 40.3% amino acid sequence identity with oligopeptidase PepFBa (protein ID AAQ08885.2) secreted by B. amyloliquefaciens 23-7A and 38.5% identity with Pz-peptidase (protein ID BAA13561.1) secreted by B. licheniformis N22. A comparison of the GT-SM3B sequence with those of the only studied metallo-oligopeptidases from thermophiles revealed 39.5% sequence identity with G. collagenovorans MO-1 intracellular Pz-peptidase B (protein ID BAD99434.1) and a notably lower identity of 26.2% with G. collagenovorans MO-1 intracellular Pz-peptidase A (protein ID BAD99433.1). Identity of 34.2% was observed between sequences of GT-SM3B and M3B subfamily’s well-described holotype peptidase Lactococcus lactis subsp. lactis IL1403 oligopeptidase F (protein ID AAK05825.1) (Fig. 1) [26]. In addition, GT-SM3B displayed a relatively low 25.3% identity with putative cytoplasmic oligopeptidase F (protein ID BAD75248.1) of Geobacillus stearothermophilus, the crystal structure of which was solved [11].

Fig. 1.Alignment of the amino acid sequence of GT-SM3B (protein ID AHG94995.1) with M3B subfamily holotype oligopeptidase PepF (protein ID AAK05825.1) from L. lactis subsp. lactis IL1403 and both M3B subfamily oligopeptidases with in silico predicted signal peptides– PepFBa (protein ID AAQ08885.2) and Pz-peptidase (protein ID BAA13561.1) from B. amyloliquefaciens 23-7A and B. licheniformis N22, respectively. The residues identical at all aligned positions are underlined with an asterisk (*); the symbol (:) relates to conserved constitutions. Putative signal peptides are highlighted in grey. Residues essential for metal-binding and catalysis are boxed in sequences.

Two elements essential for catalysis were identified (Fig. 1) in the GT-SM3B sequence. The main element was conservative zinc metallopeptidase consensus motif HisGlu-X-X-His (His400-Glu401-Leu-Gly-His404). No less important was the identification of Glu428, positioned after 23 amino acids downstream of the consensus motif functioning as a third zinc ligand. In theory, the minimal distance from the consensus motif to the glutamate residue, if it acts as the third zinc ligand, is 18 residues [31]. The first glutamate in the GT-SM3B sequence C-terminal to the His-Glu-X-X-His motif is positioned a little further, after 23 amino acids. The position of glutamate as the third zinc ligand was shown to be of the same distance as suggested for GT-SM3B in a crystal structure of Pz-peptidase A [16], which, according to sequence homology, is the closest to GT-SM3B of geobacilli oligopeptidases with a solved crystal structure. Conservative sequence elements as well as overall sequence homology to oligopeptidases F from different bacteria enable GT-SM3B attribution to the M3B subfamily of metallo-peptidases.

Recombinant Oligopeptidase Secretion by E. coli

GT-SM3B production under initial conditions had no inhibitory effect on expression strain culture growth and ensured relatively high-level overexpression of GT-SM3B. Unfortunately, almost the whole enzyme produced accumulated in inclusion bodies and was catalytically inactive. Intracellular production of GT-SM3B lacking signal peptide in E. coli BL21 (DE3) was also achieved, but the recombinant protein was also insoluble and inactive (data not shown). However, E. coli, as predicted, recognized native signal peptide of GT-SM3B in silico, despite the introduced N-terminal hexahistidine thrombin-cleavable tag. The total amount of GT-SM3B secreted under initial production conditions was insufficient to be visualized by SDS-PAGE, but was detected by activity assay.

Initial evaluation of GT-SM3B production revealed necessity to select overexpression conditions increasing expressed GT-SM3B solubility in E. coli BL21 (DE3) cytoplasm and/or maximizing oligopeptidase secretion by the expression strain. No influence of GT-SM3B overexpression on expression culture growth was observed during optimization under a variety of conditions.

The attempt to enhance the solubility of GT-SM3B accumulated in the cytoplasm of E. coli BL21 (DE3) cells failed. The lowering of expression temperature from 37℃ to 30℃ or 25℃ equally increased the level of oligopeptidase expression, but not the solubility. Likewise, a lower cell density at the point of induction of OD600 0.5 (not 0.8) slightly increased the expression without any impact on GT-SM3B solubility. IPTG concentrations, applied for induction, positively correlated with the amount of expressed GT-SM3B, that again accumulated intracellularly in the form of the inclusion bodies. Extension of expression time from 2 h to 4 and 12 h did not enhance the intracellular solubility of the recombinant oligopeptidase, but resulted in decline of its expression level. Fortunately, an appreciable amount of secreted GT-SM3B was detected when 0.5 mM IPTG induced GT-SM3B expression for 4 h at 30℃, starting the induction after recombinant E. coli BL21 (DE3) reached OD600 0.5. These conditions, optimal for the recombinant enzyme secretion, were further used for full-scale GT-SM3B secretory overproduction. In GT-SM3B production, LB medium was replaced by M9 medium with casamino acids 1% (w/v), since the latter medium substantially intensified oligopeptidase secretion. If GT-SM3B secretion into selected medium was continued for no more than 4 h, the background proteolytic activity of E. coli secretome proteases was not detected.

Enzyme Purification

Active recombinant peptidase was purified from 500 ml of expression culture fluid to 36-fold by ammonium sulfate fractionation combined with two consecutive ion-exchange chromatography steps. Most of E. coli secretome proteins were separated by anion-exchange chromatography. The final cation-exchange chromatography enabled to concentrate GT-SM3B and remove the last contaminating proteins. Assessment by SDS-PAGE and gelatin zymography confirmed GT-SM3B homogenous purity, revealing homodimerization (Fig. 2). In total, 0.34 ± 0.02 mg of pure recombinant oligopeptidase with a specific activity of 5.3 ± 0.03 U/mg was obtained. GT-SM3B was stable and non-aggregating throughout the purification procedure, including the oligopeptidase sample buffer.

Fig. 2.Results of GT-SM3B purification and hydrolysis of collagenous peptides. 1: GT-SM3B gelatin zymography; 2: Purified GT-SM3B; M: PageRuler Unstained Protein Ladder (Thermo Fisher Scientific); 3: Gelatin control; 4: Gelatin-derived oligopeptide hydrolysis with GT-SM3B; 5: type I collagen control; 6: type I collagen-derived oligopeptide hydrolysis with GT-SM3B; 7: Raw collagen extract control; 8: Raw collagen extract-derived oligopeptide hydrolysis with GT-SM3B. Collagenous oligopeptide hydrolysis reactions with GT-SM3B were performed in 50 mM HEPES–NaOH, pH 7.3 at 60℃. The hydrolysis reactions were performed for 12 h. Controls were treated under identical conditions, but without GT-SM3B.

Effects of Temperature and pH on the Activity and Stability of GT-SM3B

The optimum temperature for maximum GT-SM3B peptidolytic activity was 40℃, among the ten different temperatures tested (Fig. 3A). Oligopeptidase showed a typical bell-shaped curve with the hydrolytic activity in the temperature range of 10–80℃. The thermoactivity of the GT-SM3B at temperatures 50℃ and 60℃ was 84% and 64% relative to the optimum enzyme thermoactivity. The peptidase activity of 18% was detected at 10℃, while GT-SM3B activity decreased at the temperature over 60℃. The characterized metallo-oligopeptidase was inactive at 95℃ (Fig. 3A), but still demonstrated 34% of its activity at 70℃ and 2% at 90℃.

Fig. 3.Effects of temperature on GT-SM3B. (A) The effect of temperature on activity of GT-SM3B. The purified oligopeptidase was assayed at temperatures ranging from 10℃ to 95℃ in 50 mM HEPES–NaOH, pH 7.3 with carbobenzoxy-Gly-Pro-Gly-Gly-Pro-Ala-OH peptidolysis assay to determine the effect of temperature on activity. Peptidolysis at optimal temperature (5.3 ± 0.03 U/mg) was taken as 100% of oligopeptidase activity. (B) The effect of temperature on stability of GT-SM3B. The thermostability of GT-SM3B was investigated after incubation of the enzyme solutions in the absence of substrate at indicated temperatures for 1 h in 50 mM HEPES–NaOH, pH 7.3. Residual peptidolytic activity was measured after the incubation. Peptidolysis at 60℃ (3.6 ± 0.06 U/mg) was taken as 100% of oligopeptidase activity.

GT-SM3B retained 71% of the initial activity after 1 h exposure to 60℃ (Fig. 3B), giving extrapolated half-life (t1/2) estimates in excess of 1.5 h. At 70℃, the t1/2 value of the oligopeptidase was approximately 40 min. The thermal inactivation of free GT-SM3B at 80℃ was almost complete, with an observed residual activity below 2% (Fig. 3B).

The GT-SM3B displayed pH optimum at pH 7.3 and had an effective activity level above 50% within a wide pH range of 5.0–8.0 (Fig. 4). Below neutral pH, the GT-SM3B activity had tendency towards gradual decline to 44% of hydrolytic activity at pH 4. The slight shift of the pH profile of GT-SM3 to more alkaline pH was noticed, because activity at pH 8.0 was by approximately 7% higher than measured at pH 7.0. GT-SM3B peptidolysis declined sharply above pH 8.0, and at pH 10.0 it was not more than 5% (Fig. 4). The purified oligopeptidase was stable at pH from 5.0 to 8.0 (Fig. 4). The treatment of GT-SM3B at strongly acidic or alkaline conditions led to the reduction of enzyme activity. Residual activity of GT-SM3B at pH 4.0 and 9.0 was 60% (Fig. 4), and at pH 10.0 was 40%.

Fig. 4.The effect of pH on the activity and stability of GT-SM3B. The influence of different pH values on the GT-SM3B activity was tested at 60℃ with the carbobenzoxy-Gly-Pro-Gly-Gly-Pro-Ala-OH peptidolysis assay. GT-SM3B pH stability was described by determining residual activity after enzyme pre-incubation in the absence of substrate for 1 h at 40℃ in buffer of pH 4–10 range with the specified hexapeptide peptidolysis assay at 60℃. Peptidolysis at 60℃ (3.6 ± 0.06 U/mg) was taken as 100% of oligopeptidase activity.

Influence of Metal Ions and Various Compounds on the GT-SM3B Activity

The effects of metal chlorides and other reagents on GT-SM3B activity are shown in Tables 1 and 2, respectively.

Table 1.aFinal concentration, 1 mM or as indicated. bHexapeptide cleavage at 60℃ was taken as 100%, corresponding to 3.4 ± 0.06 U/mg.

Table 2.aHexapeptide cleavage at 60℃ was taken as 100%, corresponding to 3.4 ± 0.06 U/mg. ND, not detectable.

The presence of Li+, K+, Sn2+, and Ba2+ at concentration of 1 mM as well as K+ at 10 mM concentration had no effect on the activity of recombinant oligopeptidase. The addition of Ca2+ up to 1 or 10 mM had negligible stimulation on GT-SM3B catalytic activity. The Mg2+ and Na+ ions at 1 mM concentration were stimulators, enhancing GT-SM3B enzymatic activity by about 20%. Interestingly, at 10 and 350 mM, Na+ cation had the same stimulatory activity on GT-SM3B as at 1 mM concentration. The Co2+ ion at 1 mM was a weak inhibitor of GT-SM3B, because it had decreased oligopeptidase activity only by about 6%. Other divalent cations (Mn2+, Ni2+, Cu2+, and Zn2+) that acted as GT-SM3B inhibitors at 1 mM concentration, reduced the enzymes activity by approximately 50%. Fe3+ was the strongest oligopeptidase metal ion inhibitor, with inhibition activity of over 75%.

Chelation by EDTA caused complete GT-SM3B inhibition, but the same 1 mM concentration of EGTA had no influence on the hydrolase catalysis. At 1 mM, reducing agents modulated GT-SM3B differently. Enzyme activity was suppressed by DTT and 2-ME by more than 65% and 9%, respectively. TCEP stimulated GT-SM3B activity by about 10%. All tested detergents inactivated recombinant oligopeptidase at a different extent; from approximately 14% inhibition by 0.1% Triton X-100 up to the complete elimination by 10 mM of SDS and 1 mM of Brij 35 or Brij 58. Urea at 1 mM and 10 mM concentrations had no impact on GT-SM3B activity. However, the other chaotropic agent guanidine-HCl at 1 mM inhibited the characterized enzyme by nearly 20%. The presence of glycerol at 1% concentration ensured the enhancement of GT-SM3B activity by more than 11%. The glycerol solution of 5% concentration and both applied concentrations of ethylene glycol produced weak inhibitory effects on GT-SM3B activity, ranging from 7% to 24%.

Evaluation of Oligopeptidase Oligomerization and Substrate Specificity

Electrophoresis revealed oligomerization of the purified enzyme (Fig. 2). The molecular mass of the GT-SM3B oligomer was close to 150 kDa. The observed molecular mass indicates that the GT-SM3B oligomer is formed during homodimerization. The homodimer of the characterized oligopeptidase was visualized in SDS-PAGE at lower concentration than the GT-SM3B monomer.

Hydrolytic activity of the GT-SM3B monomer and dimer was detected via gelatin zymography (Fig. 2). The localization of clear unstained zones in zymograms agreed well with the sizes of the GT-SM3B monomer and homodimer. The intensity of activity zones corresponded to the concentrations of monomer and dimer as well.

The substrate specificity of the oligopeptidase towards the hexapeptide carbobenzoxy-Gly-Pro-Gly-Gly-Pro-AlaOH and mixtures of short-chain collagen-derived oligopeptides was determined at 60℃ (Table 3). GT-SM3B showed the highest preferentiality towards the synthetic hexapeptide. A kinetic analysis of the reaction with this substrate revealed Km, Vmax, and kcat values of 2.17 ± 0.04 × 10-6 M, 2.65 ± 0.03 × 10-3 µM/min, and 5.99 ± 0.07 s-1, respectively. The specific peptidolytic activity at 60℃ was by almost 36% lower comparing with the GT-SM3B peptidolysis maximum of 5.3 ± 0.03 U/mg at optimal temperature and pH. The enzyme’s specific hydrolytic activity against oligopeptide mixtures from the acetic acid solubilized gelatin and type I collagen was similar. The GT-SM3B relative hydrolysis activity on these peptide mixtures was reduced by about one third. When evaluating GT-SM3B specificity, the raw collagen extract oligopeptide mixture was most poorly utilized of all tested substrates. Recombinant metallo-hydrolase was inactive against azocoll even after 12 h of incubation at 30–60℃ (data not shown). GT-SM3B hydrolytic activities were visualized (Fig. 2) by SDS-PAGE. Positive cleavage patterns of the collagenous substrates hydrolysis reactions performed at 60℃ for 12 h were observed in stained SDS-PAGE gels.

Table 3.aHexapeptide cleavage at 60℃ was taken as 100%.

 

Discussion

The supposition that secreted zinc-dependent oligopeptidase is encoded in the G. thermoleovorans DSM 15325 genome was proven by GT-SM3B sequence amplification from chromosome of this obligately thermophilic bacteria. Out of all metallo-oligopeptidases with the known biochemical characteristics, the GT-SM3B amino acid sequence was the most identical to PepFBa and Pz-peptidase secreted by mesophilic bacteria B. amyloliquefaciens 23-7A [9] and B. licheniformis N22 [4], respectively. The last characterized metallo-hydrolase with comparably high identity to GT-SM3B was intracellular Pz-peptidase B from G. collagenovorans MO-1 [24]. Results of GT-SM3B sequence similarity analysis indicate enzyme novelty for characterization, as the enzyme amino acid sequence identity with the so-far studied metallo-oligopeptidases does not exceed 41%. The GT-SM3B sequence in silico analysis revealed elements governing metal-binding and catalytic mechanisms, as typical for gluzincin metallopeptidases. All to date known metallo-oligopeptidases are gluzincins [32], so GT-SM3B is no exception. Sequence homology assigns GT-SM3B to the M3B subfamily of the only bacterial metallopeptidases limited to oligopeptidolytic specificity.

To the best of our knowledge, GT-SM3B production exploiting secretion by an expression strain in order to achieve soluble and active enzyme is the first time when secretory production was used for metallo-oligopeptidase. M3B peptidases that were produced by heterologous overexpression remained soluble in the E. coli cytoplasm [22,35]. The only bacterial metallo-oligopeptidase with signal peptide produced in E. coli (PgPepO from P. gingivalis 381) was also smoothly expressed at 25–28℃ for 1 h, remaining intracellularly soluble, but was not secreted by E. coli BL21 (DE3) pLysS [5]. The tendency, recorded in literature [28], to maintain soluble metallo-oligopeptidase in cytoplasm overexpressing target hydrolases at reduced temperature and IPTG concentration did not suit in the case of GT-SM3B. The decrease of temperature and the fall of expressed GT-SM3B concentration in E. coli BL21 (DE3) cytoplasm were not the changes that could increase the amount of intracellularly soluble and/or active oligopeptidase. Overexpression of GT-SM3B had no influence on expression strain growth dynamics. This fact also supports the conclusion that aggregated GT-SM3B was inactive inside E. coli. However, the beneficial medium and GT-SM3B production parameters for enzyme secretion were determined during optimization. The way GT-SM3B was produced granted the active and not aggregation prone enzyme’s final mature form to be not affected by additional amino acids.

The GT-SM3B temperature optimum at 40℃ is similar to PepFBa optimum at 45℃ [9] and comparable with B. licheniformis Pz-peptidase and PgPepO activities at 37℃ [4,5]. Only G. collagenovorans MO-1 thermoactive intracellular Pz-peptidases B and A significantly exceed the GT-SM3B optimum, 30℃ and 25℃, respectively [24]. The range of GT-SM3B effective activity is one of the widest ever reported for metallo-oligopeptidases. The high thermoactivity of GT-SM3B, for 20℃ above optimum, is characteristic for only this oligopeptidase. The temperature of GT-SM3B activity maximum is low, as for an enzyme from the obligate thermophile G. thermoleovorans DSM 15325 that has growth maximum between 55℃ and 60℃ with growth minimum at 55℃ [18]. Previously characterized bacterial metallo-oligopeptidases secreted by mesophiles, namely PepFBa, B. licheniformis Pz-peptidase, and PgPepO, were the most active at or near the temperature of microorganism growth optimum.

The comparison (Table 4) of GT-SM3B thermostability with the previously reported for other metallo-oligopeptidases shows that GT-SM3B oligopeptidase is as thermostable as Pz-peptidase A and slightly less stable than Pz-peptidase B of G. collagenovorans MO-1. Unfortunately, the most complete data on the thermostability of secreted bacterial metallo-oligopeptidases are available only for B. licheniformis Pz-peptidase that was thermostable at 50℃, but was rapidly inactivated at temperatures above 55℃ [4]. Additionally, 1 h stability at 40℃ of PepFBa [9] was also reported. GT-SM3B is the first thermostable extracellular metallo-oligopeptidase, because it is more thermostable than B. licheniformis Pz-peptidase and PepFBa. The thermostability of GT-SM3B is also the highest among the non-thermoactive metallo-oligopeptidases.

Table 4.Comparison of thermostability between different bacterial metallo-oligopeptidases.

The characterized oligopeptidase has broad coinciding pH activity and stability ranges of pH 5.0–8.0 at 60℃. The profiles of GT-SM3B pH activity and stability have more resemblance to the profiles known for intracellular metallo-oligopeptidases from lactococcci [26,41] and G. collagenovorans MO-1 [24] than to the extracellular metallo-oligopeptidase from mesophiles, regarding the profiles’ widths and characteristic shift to alkaline pH. The GT-SM3B pH optimum displayed at pH 7.3 is closer to optimums of PgPepO at pH 6.8–7.0 [5] and PepFBa at pH 7.0 [9], as B. licheniformis Pz-peptidase requires pH 7.8 for maximal activity [4]. If compared with the pH optimums of intracellular metallo-oligopeptidases, the GT-SM3B pH optimum is almost identical to PepB, a streptococcal oligopeptidase F with pH optimum at 7.4 [21].

Results of GT-SM3B sequence in silico analysis were confirmed by the effect of chelators, divalent metal ions and phosphoramidon, indicating this enzyme to be a zinc-containing metallopeptidase. The crucial importance of the divalent ion was demonstrated by complete GT-SM3B activity inhibition by EDTA and phosphoramidon, but the characterized enzyme was not Ca2+ dependent. Milimollar excess of Zn2+ strongly inhibited GT-SM3B, and this is typical for zinc metallopeptidases [12]. Other secreted or thermostable oligopeptidases of the M3B subfamily were similarly susceptible to Fe3+, Co2+, Ni2+, and Cu2+, like GT-SM3B. Pz-peptidases from G. collagenovorans MO-1 and B. licheniformis N22 were more tolerant to Mn2+ than GT-SM3B [4,24]. The characterized metallo-oligopeptidase was more intensively stimulated by Mg2+ than Pz-peptidase B from G. collagenovorans MO-1 [24]. The amount of Na+ up to at least 350 mM was a positive activity modulator for GT-SM3B. However, 1 M of NaCl decreases the activity of lactoccocal oligopeptidase F by 25% [6]. GT-SM3B does not contain cysteine residues; therefore, DTT inhibition of GT-SM3B can be based on DTT chelating properties [26]. The analysis of the impact of compounds often used in biotechnology on GT-SM3B activity is the most detailed ever reported for metallo-oligopeptidase. GT-SM3B tolerated urea and glycerol and was able to remain effectively active in the presence of most tested reagents.

The estimation of GT-SM3B homodimerization was expected. The homologous secreted hydrolases from bacilli PepFBa [9] and B. licheniformis Pz-peptidase [4] were purified as homodimers. It is likely that the in vivo homodimer is a native state of GT-SM3B, but the in vitro monomer is also catalytically active at the same extent as the oligomer. Secreted M13 family oligopeptidase PgPepO was found to be a monomeric [5]. Intracellular bacterial metallo-oligopeptidases are mainly known as monomers.

The dissolvent of gelatin, type I collagen, and raw collagen extract in acetic acid as well as exposure to 60℃ during inspection of GT-SM3B hydrolytic activity were the manipulations that produced short oligopeptides accessible for the GT-SM3B active center. M3B peptidases cleave peptides ranging from 5 to 23 amino acids in length [9,26]. Having a broad oligopeptidolytic specificity to collagen-derived oligopeptides, GT-SM3B did not hydrolyze azocoll, an intact dye-impregnated Type I collagen. GT-SM3B was relatively more specific to hydrolyze oligopeptides derived from gelatin and Type I collagen, substrates identical in polypeptide chain sequences. Type I collagen oligopeptides comprise only a fraction among other peptides of collagen types in raw collagen extract, which explains the weaker utilization of the extract’s oligopeptide mixture by GT-SM3B. Carbobenzoxy-Gly-Pro-Gly-Gly-Pro-Ala-OH, the optional substrate for oligopeptidase activity evaluation [4,38], was actively hydrolyzed by GT-SM3B between Gly-Gly and/or Pro-Gly. The Vmax of this reaction was 2.65 ± 0.03 × 10-3 µM/min, with Km and kcat values of 2.17 ± 0.04 × 10-6 M and 5.99 ± 0.07 s-1, respectively. Unfortunately, these kinetic parameters of GT-SM3B activity cannot be compared with kinetic characteristics of other oligopeptidases as carbobenzoxy-Gly-Pro-Gly-Gly-Pro-Ala-OH has been never used for kinetic analysis before. M3B metallo-oligopeptidases lack aminopeptidolytic activity and require a minimum of three amino acids from the peptide carboxyl terminus [26]; this enables to predict the hexapeptide cleavage sites with high degree of confidence.

Taken together, the GT-SM3B substrate specificity and accordance of enzyme’s characteristics with G. thermoleovorans DSM 15325 optimal growth conditions suggest a role of the characterized oligopeptidase in bacteria nutrition. The GT-SM3B-encoding thermophile is a proteolytic bacterium [18]. The described secreted oligopeptidase functioning extracellularly would increase the efficiency of G. thermoleovorans DSM 15325 collagen catabolism cascade. According to this model, GT-SM3B would take intermediate place as a cascade member, acting in partnership with extracellular G. thermoleovorans DSM 15325 collagenolytic proteases and intracellular metallo-oligopeptidases. It is noteworthy that G. thermoleovorans DSM 15325 encodes and produces intracellular homologs of Pz-peptidases B and A (unpublished results), the oligopeptidases that ought to support collagen degradation by collagen-derived peptides hydrolysis after their uptake inside a cell of G. collagenovorans MO-1 [16,24]. Possible partial overlapping substrate specificity of GT-SM3B and intracellular G. thermoleovorans DSM 15325 metallo-oligopeptidases, as usual for lactococci and lactobacilli metallo-oligopeptidases, would be fully reasonable. The G. collagenovorans MO-1 genome is not sequenced and there are no data of whether this thermophile also encodes secreted metallo-oligopeptidases.

Housekeeping functions, not necessary protein turnover, are attributed for a part of characterized bacterial M3B peptidases that are constitutively expressed. GT-SM3B is not a constitutively secreted enzyme. The hydrolase was not identified by MALDI-TOF mass spectrometry analysis in the G. thermoleovorans DSM 15325 secretome (unpublished results), when bacteria were cultivated in minimal medium with glucose as the sole carbon and energy source [14]. Secreted metallo-oligopeptidases from bacilli PepFBa [9] and B. licheniformis N22 Pz-peptidase [4] are supposed to participate in the degradation of peptides for cell nutrition, whereas P. gingivalis 381-secreted PgPepO ensures invasion of this pathogen [3].

The described putative hydrolase is the first secreted thermostable metallo-oligopeptidase with possible function in G. thermoleovorans DSM 15325 to respond to environmental changes. The determined set of GT-SM3B characteristics unites the qualities of metallo-peptidases earlier separately attributed to the thermoactive metallo-oligopeptidases or to the peptidases of this catalytic type from mesophiles.