A bacterial strain was selected from others, isolated from infective juveniles (IJs) of Heterorhabditis brevicaudis YS for further work, because it had the highest protease activity based on the clear zone formation around the colonies grown on skim milk agar and on gelatin agar plates. The result of Biolog identification system only supported the identification of the bacterial strain as a Photorhabdus sp. (data not shown). The 16S rDNA sequence of the strain 0805-P5G was deposited in the GeneBank database under accession number EU301784. After analyzing phylogenetic trees, the strain 0805-P5G was most related to P. luminescens subsp. akhurstii strain LN2, it was proposed be within the P. luminescens subsp. akhurstii (Figure 1). We found that the P. luminescens strain 0805-P5G secreted protease from the early logarithmic phase of growth in a variety of media, such as NB, LB and PP3T. The maximum proteolytic activity was obtained from the NB culture medium at eight days post-inoculation with time-course data (data not shown).

The homogeneity of the protease from P. luminescens 0805-P5G was greatly increased after the purification process by FPLC using DEAE column and Q Sepharose column. The final specific activity, fold of purification and recovery is 85.7 U/mg, 10.6 and 8.2%, respectively. The isoelectric point of the protease is 4.2 (Figure 2B).

The authenticity of the protease was confirmed by SDS-PAGE and zymograms. The purified protease showed a single band of 51 kDa in SDS-PAGE (Figure 2A) and exhibited proteolytic activity at the corresponding bands in the zymogram with gelatin and casein, respectively (Figure 3).

The N-terminal amino acid sequencing of the purified protease showed the first 10 amino acid residues, DKDVSGSEKA, which is homologous to alkaline metalloprotease from P. luminescens subsp. akhurstii W14 and P. luminescens subsp. laumondii TT01 [9]. Furthermore, the partial peptide sequence of the protease was identified by LC MS/MS (data not shown). The molecular mass of the purified protease was determined by MALDI-TOF mass spectrometry to be 51.8 kDa, consistent with the results of SDS-PAGE (Figure 2A) and zymograms (Figure 3).

The protease of P. luminescens strain 0805-P5G exhibited highest activity at pH 8 (Figure 4). It remained stable in a wide pH range (5–10) and retained more than 80% of its original activity after incubation for 30 min.

Results showed that the protease had an optimum of 60 °C, with 35% activity retained at 90 °C (Figure 5). In addition, thermal stability results also showed that this enzyme had nearly 100% residual activity in the temperature range of 14–60 °C for 30 min (Figure 5). The long thermal stability showed that the protease is stable, with less than 50% loss in enzyme activity observed upon incubation, even after 10 days at 60 °C (data not shown).

Little inhibition was observed with PMSF and PCMB (Figure 6), which are inhibitors of serine and thiol proteases, respectively. However, significant levels of inhibition were observed following incubation with either EDTA (a general metalloprotease inhibitor) or 1,10-phenanthroline (a zinc metalloprotease inhibitor, but not specific for Zn). There was also a dose-dependent response to these two inhibitors at concentrations between 0.2 and 10 mM (Figure 6). These data are consistent with the supposition that the enzyme might be a Zn metalloprotease.

The bioassay of P. luminescens protease against G. mellonella by injection showed high insecticidal activity. It had the significant insecticidal rate of 100% at the concentration of 11.6 ng/μL. The insecticidal activity decreased clearly when the concentration was decreased (Table 1). The LC₅₀ value against G. mellonella after injection is 3.8 ng/μL (2.1–4.7 ng/μL). The numbers in parentheses are the 95% fiducial limits. In another experiment to test the oral toxicity, both the field-collected and the American strains of P. xylostella were assayed, only the field-collected strain from Taiwan showed high oral toxicity (LC₅₀ is 43 ng/μL). In contrast, low oral toxicity (LC₅₀ is much more than 120 ng/μL) was shown against the American strain of P. xylostella. A general authentic protease (purified from Bacillus polymyxa) purchased from Sigma, with the same concentrations were used both orally and in injection to serve control to both experiments. The results showed that control treatments were non-toxic to insects. Another result of the verification experiment of the purified metalloprotease as the cause of insect toxicity (injected) showed the two controls, both at 3.8 ng/μL, one with 1,10-phenantroline treated then dialyzed enzyme significantly decreased mortality rate (3%), the other with only dialyzed enzyme preparation retained mortality rate (47%).

Thirteen local Photorhabdus luminescens strains were maintained on NBTA plates [(2.3% nutrient agar (Difco, Franklin Lakes, NJ, USA), 0.0025% bromothymol blue (Merck, Whitehouse Station, NJ, USA), 0.004% 2,3,5-triphenyltetrazolium (Sigma-Aldrich, St. Louis, MO, USA)] at 30 °C and subcultured weekly. 250-mL Erlenmeyer flasks with 50 mL of three different liquid culture media, namely, NB (nutrient broth, Difco, Franklin Lakes, NJ, USA), LB broth (Luria-Bertani Miller, Difco, Franklin Lakes, NJ, USA) and PP3T [(2% Proteose Peptone No. 3 (Difco, Franklin Lakes, NJ, USA) plus 0.5% Tween 60] were inoculated, respectively. The proteolytic activity during bacterial growth was monitored daily by a caseinolytic macroassay. The protease activity in the fractions obtained in the purification steps was tested by a microassay. The macroassay and the microassay of protease activity were both performed as described by Cabral et al.[1]. Aliquots (50 μL) of culture supernatant were obtained at different time intervals, combined with 80 μL of 2% azocasein (Sigma) in 50 mM Tris-HCl (pH 8.5) and incubated at 37 °C for 3 h. The undigested substrate was precipitated by adding 400 μL of 10% trichloroacetic acid to the reaction mixture, followed by incubation for 10 min at 4 °C, followed in turn by centrifugation at 10,000× g for 10 min. The supernatant was neutralized by the addition of 600 μL of 1 N NaOH, and the absorbance at 440 nm (A₄₄₀) was measured. The protease activity in the fractions obtained in the purification steps was tested by a microassay for which 20-μL aliquots from each fraction were added to 40 μL of 3% azocasein in Tris, plus an additional 40 μL of 50 mM Tris-HCl (pH 8.5), followed by incubation at 37 °C for 30 min. The undigested substrate was precipitated by adding 75 μL of 10% trichloroacetic acid to the reaction mixture, followed by incubation for 10 min at 4 °C and centrifugation at 10,000× g for 10 min. Then, 100 μL of supernatant was transferred to a 96-well microtitration plate and neutralized by the addition of an equal volume of 1 N NaOH. The A₄₄₀ was measured by using a microplate reader (Thermo Scientific Inc., Waltham, MA, USA). One unit of enzyme activity was defined as the amount of enzyme that yielded an absorbance change of 0.01.

Proteases production by 13 local P. luminescens was detected on skim milk and gelatin agar plates, which were incubated at 30 °C for five days. Skim milk agar contained 1.5% skim milk powder (Difco, Franklin Lakes, NJ, USA), 1.3% nutrient broth (Scharlau) and 1.5% agar (Oxoid), while skim milk powder was replaced by gelatin in gelatin agar. Microorganisms with proteolytic activity were detected by the formation of clear zones around colonies.

Symbiotic bacteria of nematode Heterorhabditis brevicaudis YS were observed by culturing on MacConkey’s agar and NBTA plates. Of them, a P. luminescens strain 0805-P5G was isolated. The 16S rDNA sequences and phylogenetic analyses were conducted as described in our previous study [10]. The 96-well Biolog GN2 microplate (Biolog Inc., Hayward, CA, USA) was also used to assay the oxidation of 95 carbon substrates as described in our previous study [6].

A two-column strategy was developed for protease purification. All the purification steps were performed at 4 °C with AKTKprime (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA). Following growth on NB media, 150 mL of eight-day culture was centrifuged at 30,000× g for 30 min at 4 °C. The supernatant was then applied to a HiTrap DEAE Sepharose Fast Flow column equilibrated with 50 mM K₂PO₄ (pH 8.6). The bound proteins were eluted in a linear gradient of 8 mL from 0 to 0.2 M KCl in 50 mM K₂PO₄ (pH 9.3), followed by a linear gradient of 10 mL from 0.2 to 0.25 M KCl and, finally, a linear gradient of 5 mL from 0.25 to 1.0 M KCl. The fractions with protease activity were then pooled, dialyzed against 50 mM Tris-HCl (pH 8.5) and loaded onto a HiTrap Q Sepharose XL column equilibrated with the same buffer. The bound proteins were eluted with 40 mL of NaCl in a linear gradient of 0 to 1.0 M in 50 mM Tris-HCl (pH 8.5). The protease activity in the fractions obtained in the purification steps was assayed by a microassay as described by Cabral et al.[1].

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed in 12% polyacrylamide gel. P. luminescens protease has high affinity for substrates, such as casein and gelatin [5]; therefore, zymograms were performed as described by Schmidt et al.[11] and Cabral et al.[1], with minor modifications. Polyacrylamide gels (12%) were copolymerized with 0.05% gelatin or casein. The samples were applied in nonreducing Laemmli buffer without boiling and run at 100 V. The gels were washed twice in 2.5% (w/v) Triton X-100 solution and then incubated in 50 mM Tris-HCl (pH 7.5) with 5 mM CaCl₂ and 200 mM NaCl for 2 h at 30 °C [5]. The gels were stained with 0.1% amido black in methanol-acetic acid-water (40:10:50, v/v/v) and later washed until the zones of substrate hydrolysis were visible. Protease requires bound calcium to stabilize its structure [5], and we therefore added 5 mM CaCl₂ to the development buffer of zymography.

Polyacrylamide isoelectric focusing (IEF) was performed in a Mini-PROTEAN 3 gel system (Bio-Rad) using ampholytes of the pH range 3 to 10. Approximately 3–5 μg of the purified protein along with standards (pI 4.45–9.6) were applied to the ready gel (Bio-Rad) according to the manufacturer’s instructions. The IEF gel was stained with Coomassie brilliant blue.

The N-terminal amino acids of purified protease were sequenced by automated Edman degradation. The exact molecular mass of the purified protease was determined by matrix-assisted laser desorption ionization-time-of-flight (MALDI-TOF) mass spectrometry. Both techniques were carried out by the Mission Biotech Company (Taipei, Taiwan).

Twenty μl of the purified proteases (116 ng/μL) activity was determined by varying the pH of the assay reaction mixture using the following buffers: 200 mM glycine (pH 3), 200 mM sodium acetate (pH 4–5), 200 mM sodium phosphate (pH 6), 200 mM Tris-HCl (pH 7–8) and 200 mM Tris-glycine (pH 9–10). To determine the stability of the protease, it was preincubated in different buffers (pH 3–10) for 30 min. The maximum activity was defined as 100%. The relative activity (% of maximum) was calculated. The averages from three experiments are shown. The proteolytic activity was accessed by using the azocasein microassay described as Section 3.1. Twenty microliter of the purified proteases (116 ng/μL) was added to 40 μL of 3% azocasein in Tris plus an additional 40 μL of 50 mM Tris-HCl (pH 8.5), followed by incubation at 37 °C for 30 min. The undigested substrate was precipitated by adding 75 μL of 10% trichloroacetic acid to the reaction mixture, followed by incubation for 10 min at 4 °C and centrifugation at 10,000× g for 10 min. Then, 100 μL of supernatant was transferred to a 96-well microtitration plate and neutralized by the addition of an equal volume of 1 N NaOH. The A₄₄₀ was measured by using a microplate reader (Thermo Scientific Inc., Waltham, MA, USA).

The activity was evaluated by measuring 20 μL of the purified proteases (116 ng/μL) activity at 4, 14, 23, 28, 37, 50, 60, 65, 70, 80 and 90 °C in 50 mM Tris-HCl (pH 8) for 30 min. The effect of temperature on protease stability was determined by measuring the residual activity after 30 min of pre-incubation at different temperature. The maximum activity was defined as 100%. The relative activity (% of maximum) was calculated. The averages from three experiments are shown. The long-term stability of the enzyme was determined by measuring the residual activity at one-day intervals for 10 days of pre-incubation at 60 °C. The proteolytic activity was accessed by using the azocasein microassay described as Section 3.1. Twenty microliter of the purified proteases (116 ng/μL) was added to 40 μL of 3% azocasein in Tris, plus an additional 40 μL of 50 mM Tris-HCl (pH 8.5), followed by incubation at 37 °C for 30 min. The undigested substrate was precipitated by adding 75 μL of 10% trichloroacetic acid to the reaction mixture, followed by incubation for 10 min at 4°C and centrifugation at 10,000× g for 10 min. Then, 100 μL of supernatant was transferred to a 96-well microtitration plate and neutralized by the addition of an equal volume of 1 N NaOH. The A₄₄₀ was measured by using a microplate reader (Thermo Scientific Inc., Waltham, MA, USA).

Twenty microliter of the purified proteases (116 ng/μL) were incubated for 30 min at room temperature in the presence of phenylmethanesulfonyl fluoride (PMSF) (1.5, 2, 2.5, 3, and 10 mM), p-chloromercuribenzoate (PCMB) (1.5, 2, 2.5, 3, and 10 mM), ethylenediaminetetraacetic acid (EDTA) (1, 2, 5, 8, and 10 mM) and 1,10-phenantroline (0.2, 1, 3, and 8 mM). The activity assayed in the absence of protease inhibitors was defined as the reference control. The remaining proteolytic activity was accessed by using the azocasein microassay described as Section 3.1. Twenty microliters of the purified proteases (116 ng/μL) was added to 40 μL of 3% azocasein in Tris, plus an additional 40 μL of 50 mM Tris-HCl (pH 8.5), followed by incubation at 37 °C for 30 min. The undigested substrate was precipitated by adding 75 μL of 10% trichloroacetic acid to the reaction mixture, followed by incubation for 10 min at 4 °C and centrifugation at 10,000× g for 10 min. Then, 100 μL of supernatant was transferred to a 96-well microtitration plate and neutralized by the addition of an equal volume of 1 N NaOH. The A₄₄₀ was measured by using a microplate reader (Thermo Scientific Inc., Waltham, MA, USA). The averages from three experiments are shown.

For the Galleria haemolymph injection bioassay, aliquot 10 μL of each sample (11.6 ng/μL, 9.5 ng/μL, 7.3 ng/μL, 6.5 ng/μL, 5.8 ng/μL, 4.8 ng/μL and 3.7 ng/μL; the purified proteases were diluted with 50 mM Tris-HCl containing 250 mM NaCl.), which had been filter sterilized, was injected into seventh-instar greater wax moth (G. mellonella) aseptically using a 10 μL syringe (Hamilton). The oral bioassay was performed also with two strains (a field-collected strain and an American strain) of laboratory-reared diamondback moth (P. xylostella) under constant conditions (temperature 25 °C, relative humidity 60%–70%, under 15/9 h light/dark photoperiod). The field strain was collected locally, and the American strain was kindly provided by Syngenta (Wilmington, DE, USA) Ltd. P. xylostella larvae reared on the surface of the artificial diet that contain more than 10 ingredients [12] and can be effectively assayed using the surface contamination method. Proteases to be assayed orally were mixed with an appropriate artificial diet and fed to third-instar larvae at 25 °C. A general authentic protease (purified from Bacillus polymyxa) purchased from Sigma with the same concentrations were used both orally and by injection to serve control treatments (non-insect toxic) to both experiments to justify insect specificity of the isolated protease. The mortality rate for each concentration for 30 insect larvae was recorded after incubation for three days at 25 °C. The concentration-mortality regressions were then evaluated. At least five concentrations resulting in mortalities >10% and <90% were desirable. The data for 50% lethal concentrations (LC₅₀), which were estimated using a Probit analysis program [13], were the averages from three experiments. A further verification experiment of the purified metalloprotease as the cause of insect toxicity (oral and injected) was performed. Two controls, one with 1,10-phenantroline treated then dialyzed enzyme, the other with only dialyzed enzyme preparation at the same concentration were assayed.