Introduction

Protease S (PrtS) is a secretory zinc metalloprotease of the M4 family found in Photorhabdus bacteria [1]. This enzyme belongs to the group of protealysin-like proteases (PLPs) widespread in bacteria but also occurring in fungi and archaea [2, 3]. The biological functions of PLPs are currently being discussed. However, the available data indicates the involvement of PLPs in the bacterial interaction with animals and plants and possibly in the pathogenesis [1, 3,4,5,6,7,8,9,10,11,12,13,14]. Moreover, PLPs are thought to play a role in interbacterial competition [15, 16]. The study of PLP biological functions can clarify interactions between bacteria and other organisms.

Photorhabdus laumondii are both symbionts of Heterorhabditis nematodes and insect pathogens, providing an opportunity to investigate PLP functions via PrtS in various ecological contexts. Photorhabdus models are being actively used to study the mechanisms of bacteria-host interactions during infection. A wide range of Photorhabdus protein toxins with potent insecticidal activity, as well as a number of low molecular weight metabolites that affect host insect immunity, are known [17]. Yet, there is limited data regarding the PrtS’ role. The injection of substantial quantities of PrtS induces insect death [13]; nevertheless, neither the dose-effect relationship nor the mechanism of action has been defined. PrtS is recognized for inducing melanization in insects, which is attributed to the overstimulation and disturbance of the insect’s innate immune response [13]. The hemolymph of infected insects displays PrtS activity and in vitro exposure to PrtS reduces the antibacterial activity of hemolymph by half, likely resulting from the hydrolysis of antimicrobial peptides [1]. The transcription of the prtS gene was observed during infection [18] but decreased in a phenotypic variant of P. laumondii that exhibits delayed pathogenicity [19]. The evidence suggests that PrtS plays a role in bacterial pathogenesis in insects.

In the present study, to come closer to understanding this role, the recombinant PrtS was produced by expressing the prtS gene in Escherichia coli and purified. Using the purified enzyme, we investigated the interaction between PrtS and photorin, its endogenous inhibitor that was recently discovered [20], studied how the PrtS amount affects insect toxicity and melanization, and identified potential PrtS substrates in the larval hemolymph of the greater wax moth Galleria mellonella.

Materials and methods

Chemical and reagents

In this work we used tris-(hydroxymethyl)aminomethane (Tris) (Amresco, USA), ammonium bicarbonate, 1,10-phenanthroline, piperazine-N, N′-bis(2-ethanesulfonic acid) (PIPES), N-(2-hydroxyethyl) piperazine-N′-(2-ethanesulfonic acid) (HEPES), N-(2-hydroxyethyl)piperazine-N′-(4-butanesulfonic acid) (HEPBS), dithiothreitol (DTT), 2-iodoacetamide, 2,5-dihydroxybenzoic acid, trypsin, phenylthiourea, 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate(CHAPS), NP-40 detergent (Sigma-Aldrich, Germany), dimethyl sulfoxide (DMSO) (MP Biomedicals, France), ampicillin, imidazole (PanReac, Germany), isopropyl-β-D-thiogalactopyranoside (IPTG) (Thermo Fisher Scientific, USA), sodium dodecyl sulfate (SDS) (Helicon, Russia), acrylamide (Dia-M, Russia), N,N’-methylenebisacrylamide (Serva, Germany), and phosphate-buffered saline (PBS, Eco-Service, Russia).

The following reagents were used for genetic construction: Tersus DNA polymerase (Evrogen, Russia), Klenow fragment, T4 DNA ligase (SibEnzyme, Russia), and restriction endonucleases XhoI, XbaI (New England Biolabs, USA), FauNDI, and BglII (SibEnzyme, Russia).

Lennox LB broth (10 g/l peptone (Difco, USA), 5 g/l yeast extract (Difco, USA), and 5 g/l NaCl (PanReac, Germany)) was used to culture bacteria.

Protealysin (Pln), emfourin (M4in), and photorin (Phin) were overproduced in E. coli and purified as described previously [15, 20, 21] Thermolysin (Tln) was obtained from Serva (Germany).

2-Aminobenzoyl-L-arginyl-L-seryl-L-valyl-L-isoleucyl-L-(ε-2,4-dinitrophenyl)lysine (Abz-RSVIK(dnp)) from Peptide 2.0 (Chantilly, USA) was used as a substrate for protease activity determination.

All other reagents were of reagent grade and were purchased from Helicon (Russia).

Protein analysis

Protein concentration was assayed after Bradford with modifications [22] using IgG as a standard.

The concentrations of purified proteins were determined spectrophotometrically. Published extinction coefficients ε280nm = 52,370 M− 1 cm− 1, ε280nm = 58,200 M− 1 cm− 1, ε280nm = 10,810 M− 1 cm− 1 and 11,460 M− 1 cm− 1 were used for Pln, Tln, M4in, and Phin, respectively [20]. The PrtSt extinction coefficient (ε280nm = 48,360 M− 1 cm− 1) was calculated using the ProtParam tool (http://www.expasy.org/tools/protparam.html).

Proteins were analysed by SDS-PAGE in 12.5% polyacrylamide gel according to Laemmli [23]. Proteins were visualized by Coomassie R-250 Brilliant Blue (Sigma-Aldrich, Germany) staining. Pierce Unstained Protein MW Marker (ThermoFisher Scientific, USA) was used as molecular weight standards.

Transformation of E. coli cells with plasmid DNA was carried out as described by Sambrook et al. [24].

Cloning and construction of expression vectors

Plasmid pETDuet-PrtS was constructed to express a gene encoding PrtS from P. laumondii subsp. laumondii TT01 with a His6-tag at the C-terminus. The first step of pETDuet-PrtS construction included PCR amplification of the protease S gene from P. laumondii subsp. laumondii TT01 using pTT01 [20] as a template and PrtS_F (TGCAAATACAAAACAATAACTACAGAG) and PrtS_XhoI (GAATCTCGAGTTAATGGTGATGGTGATGGTGCTCTTAGTTTTATCTTTATTTTTGTCCT) as primers. The resulting fragment was digested with XhoI. The pET23c vector (Novagen, USA) was treated with FauNDI, Klenow fragment, and XhoI. After ligation, the resulting construct pET23-PrtS was used as the PCR template with primers T7-uni (GTAATACGACTCACTATAGGG) and T7-rev (TATGCTAGTTATTGCTCAGCGG). The amplification product was ligated into the XbaI and XhoI sites of pETDuet-1 (Novagen, USA).

pETDuet-PrtSt plasmid encodes a PrtS with a non-conserved C-terminal region replaced by the AHHHHHH sequence. The resulting protein variant was designated as PrtSt (PrtS truncated). This construction included PCR amplification of a protease S gene fragment from pETDuet-PrtS using primers PrtSt_H6 (CAGACTCGAGTTAATGGTGATGGTGATGGTGCGCAACGCCCACTTCTGTCC) and PrtSt_BglII (ATGCCTGGGAAAGAGCAG). The region corresponding to Ala and His6-tag is boldfaced and the XhoI site is underlined. The resulting fragment was ligated into the XhoI and BglII sites of pETDuet-PrtS.

All cloned fragments were confirmed by sequencing. The obtained genetic constructs were sequenced and all oligonucleotides used were synthesized at Evrogen (Russia).

Protease S purification

In an attempt to produce PrtS in E. coli using the pETDuet-PrtS plasmid, the target protein could not be purified by immobilized metal affinity chromatography (IMAC), possibly due to the proteolytic degradation of the C-terminal region with the His6-tag. It should be noted that such C-terminal regions are generally non-conserved in protealysin-like proteases (PLPs) but they are present with some variations in most PLPs of the Photorhabdus bacteria. The role of this particular region remains unknown in Photorhabdus. However, we speculate that its insignificance for the catalytic function stems from its localization outside of the catalytic domain, its distance from the active site, and its likely unstructured nature (Fig. S1). Furthermore, our collected data suggests that this region is removed from PrtS when it is produced in E. coli. Accordingly, we engineered PrtSt (PrtS truncated), a PrtS with a C-terminal region similar to recombinant protealysin (Pln) from Serratia proteamaculans [2].

E. coli T7 Express lysY/Iq (New England Biolabs, USA) cells transformed with pETDuet-PrtSt were grown with agitation in 250 ml of Lennox LB containing 100 μg/ml ampicillin at 37 °C for 2 h; 1 mM IPTG was added to the culture and incubation continued at 16 °C for 30 h. Cells were collected by centrifugation (3000 g, 4 °C, 10 min), resuspended in 15 ml of 50 mM Tris-HCl with 2 mM 1,10-phenanthroline, pH 7.4, and ultrasonicated at 4 °C 3 times for 4 min with 1 s impulse / 2 s pause. The cell lysate was centrifuged (9,000 g, 4 °C, 10 min) and the supernatant was loaded onto a 5 ml Nickel XPure Agarose Resin (UBPBio, USA) column pre-equilibrated with 50 mM Tris-HCl, pH 7.4. Elution was performed with 0-300 mM imidazole gradient (15 ml) in the same buffer. After SDS-PAGE analysis, the protease-containing fractions were pooled and concentrated by ultrafiltration on an Amicon Ultra-15 30,000 NMWL Centrifugal Filter Unit (Millipore, USA). The resulting sample was loaded onto a Superdex 75 Increase 10/300 GL column (GE Healthcare, Sweden) equilibrated with 50 mM Tris-HCl, 150 mM NaCl, 2 mM 1,10-phenanthroline, pH 7.4, and eluted with the same buffer at a flow rate of 0.4 ml/min. The fractions with the highest content of the protease were combined and dialyzed against 100 volumes of 50 mM ammonium bicarbonate using Slide-A-Lyzer G2 dialysis Cassette 3,500 MWCO (Thermo Fisher Scientific, USA) for 3 h at 4 °C. After buffer replacement, dialysis was continued overnight at 4 °C. The dialyzed solution was centrifuged (9,000 g, 15 min) and the supernatant was lyophilized.

N-terminal sequencing

For N-terminal sequencing by Edman degradation, proteins were separated by SDS-PAGE and transferred onto a PVDF membrane (Bio-Rad, USA) using the Trans-Blot Turbo system (Bio-Rad, USA) in 50 mM Borate buffer with 10% (v/v) ethanol, pH 9.0. The membrane was sent to the Protein Core Facility of Innsbruck Medical University for sequencing.

Protease activity and inhibition assays

The previously described substrate Abz-RSVIK(Dnp) [25] and analysis method [15, 20] were used for protease activity and inhibition assays. Mass spectrometry data showed that PrtSt, similar to Pln, hydrolyses the substrate at the Ser-Val bond (Fig. S2). Fluorescence was recorded on a CLARIOstar Plus microplate reader (BMG, Germany) at the excitation and emission wavelengths of 320 and 420 nm, respectively, as described earlier by Berdyshev et al. [26]. PrtSt and Tln concentrations were 50 pM in all reactions. The relationship between the rate of Abz-RSVIK(Dnp) hydrolysis by PrtSt and substrate concentration was studied for the substrate concentration range from 3.75 to 120 μM. In the inhibition assay, Abz-RSVIK(Dnp) concentrations were 30 or 90 μM. M4in concentrations were 0.1–3.5 nM for PrtSt (Fig. S3, PrtS + M4in) and 1–20 μM for Tln (Fig. S3, Tln + M4in). Phin concentrations were 0.1–3.5 nM for PrtSt (Fig. S3, PrtS + Phin).

Determination of slow-binding inhibition kinetic parameters (v0, vs, ka), kinetic constants (k+ 3, k-3), and equilibrium constant (Ki) was performed as described previously [15, 20]. The date on the rate of Abz-RSVIK(Dnp) hydrolysis by PrtSt at different substrate concentrations was fitted to the substrate inhibition Eq. (1) to calculate KM and KiS [27]:

$$v = \frac{{{V_{max}} \cdot \left[ S \right]}}{{{K_M} + \left[ S \right](1 + [S]/K_i^S)}}$$
(1)

where v is the reaction velocity, [S] is the substrate concentration, Vmax is the maximum reaction velocity, KM is the Michaelis-Menten constant, KiS is the substrate inhibition constant.

Turnover number kcat was calculated from Eq. (2):

$${k}_{cat}= {V}_{max}/{E}_{t}$$
(2)

where kcat is the turnover number, Et is the concentration of enzyme catalytic sites (for PrtSt, it equals the enzyme concentration).

The cleavage site of Abz-RSVIK(Dnp) by PrtSt was identified in 10 mM NH4HCO3 as described by Karaseva et al. [25]. The volume of the reaction mixture was 100 μl with 30 μM Abz-RSVIK(Dnp) and 0.5 nM PrtSt. The mixture was incubated for 0 and 90 min at 37 °C and immediately frozen at − 20 °C and lyophilized. Control samples were produced in a similar way but without the enzyme or the substrate. The dried samples (≈ 3 nmole of peptide) were dissolved in 300 μl of mixed water-acetonitrile (1:1) with 1 mM 1,10-phenanthroline and then diluted 1:10 in water-acetonitrile (1:1) with 0.1% formic acid. Mass spectrometric analysis of the obtained peptide solutions was carried out on a Q-exactive HF-X mass spectrometer (Thermo Scientific, USA) in the positive ionization mode using an electrospray ion source (ESI). The sample solution was injected into ESI at a rate of 5 μl/min. The following settings were used for mass spectrometric analysis: spray voltage, 4 kV; capillary temperature, 290 °C; and sheath gas flow rate, 8 l/min. Panoramic scanning was performed in the mass range from 200 to 2000 m/z at a resolution of 70,000. The resulting mass spectrum of each sample was obtained by averaging 100 scans.

The proteolytic activity was determined by Azocasein hydrolysis as described previously [1]. Three independent measurements were carried out for each experimental point in all reactions.

Effect of chelating agents and zinc ions on PrtSt activity

Two ml of ZnSO4 (0.5 or 1 mM), 1,10-phenanthroline (0.5 or 1 mM) or EDTA (10 or 50 mM) in 50 mM Tris-HCl, pH 7.4, were supplemented with 2 μl of 1 μM PrtSt in the same buffer and incubated for 10 min. Ten μl of the resulting solution was added to 110 μl of Abz-RSVIK(Dnp) in the same buffer containing the tested compound in the same concentration, and fluorescence changes were measured as described above. The Abz-RSVIK(Dnp) concentration in the reaction mixture was 30 μM. The obtained activity values were checked against the enzyme activity measured similarly but without tested compounds.

Temperature stability of PrtSt

Temperature stability was determined by incubation of 1 nM PrtSt in 50 mM Tris-HCl, pH 7.4, at 40–70°С for 2–150 min, after which samples were placed on ice. Protease activity was determined by hydrolysis of Abz-RSVIK(Dnp) as described above. Its concentration in the reaction mixture was 30 μM.

pH optimum of PrtSt activity

The pH optimum was identified by measuring PrtSt activity in buffers containing 50 mM PIPES, 50 mM HEPES, and 50 mM HEPBS; pH was adjusted to 5.5–9.5 by 1 M HCl or 1М NaOH. Protease activity was determined by Abz-RSVIK(Dnp) hydrolysis as described above. The Abz-RSVIK(Dnp) concentration in the reaction mixture was 30 μM.

Insect toxicity

Insect toxicity of Pln, Tln, and PrtSt was tested on last instar larvae of G. mellonella weighing 250–350 mg (Eco Baits, Russia). Protease solutions with concentrations of 4, 6, 8, 10, 14, 16, 18, 20, and 24 μM were made in sterile PBS. A volume of 10 μl was injected into the left one of the first pair of proleg using a BD Micro-Fine Plus 1 ml syringe (USA) with a 30G×8 needle and an infusion pump LSP04-1 A (Longer Pump, China). The control larvae were injected with 10 μl of PBS. Postinjection larvae were placed in Petri dishes and incubated at 30°С. The living and melanized animals were counted 24 h later. Three independent experiments with 15 larvae in each experimental and control group were conducted.

The LD50 and ED50 values were calculated as described by Agarwal et al. [28]. Data analysis was performed using the GraphPad Prism version 9.5 (GraphPad Software, USA).

Proteomic analysis

The purified PrtSt components were analyzed by SDS-PAGE. The protein bands of interest were excised from the gel and treated with DTT, 2-iodoacetamide, and then trypsin using the standard method [29]. Tryptic peptides were extracted from the gel as described previously [30]. An LC-MS/MS analysis was performed using a nano-flow ultra-high pressure system nanoElute (Bruker Daltonics, Germany) coupled with a mass spectrometer timsTOF (Bruker Daltonics) using a CaptiveSpray (Bruker Daltonics) ion source. The obtained mass spectra were processed by PEAKS Studio X Pro [31].

The effect of PrtSt on hemolymph proteins was studied after incubation of G. mellonella larvae on ice for 15 min. The hemolymph of 50 larvae was collected into 1.5 ml pre-cooled tubes and 2 μg/ml of phenylthiourea was added to prevent melanization. The hemocytes were removed by centrifugation at 200 g, 4 °C for 5 min and the supernatant was centrifuged at 18,600 g, 4 °C for 10 min. The cell-free hemolymph was diluted with sterile PBS to 60 mg/ml protein. The obtained solution (150 μl) was supplemented with an equal volume of purified PrtSt in PBS (0.6 mg/ml). The mixture was incubated at 30°С for 15 min and the reaction was stopped by 0.3 μl 1 M 1,10-phenanthroline, frozen, and stored at -70°С until analysis. In the control, PrtSt and hemolymph solutions were incubated separately, 1,10-phenanthroline was added to 2 mM to protease, and solutions were combined. The PrtSt solution incubated under the same conditions and supplemented with 1,10-phenanthroline was used as an additional control.

Before 2D electrophoresis, the hemolymph of G. mellonella was reprecipitated with a mixture of chloroform/methanol/water and dissolved in a mixture of sample buffer for isoelectric focusing (40 mM Tris-НСl, рН 9.5, containing 8 M urea, 2 M thiourea, and 4% of mixture of CHAPS and NP40), incubated 30 min at the room temperature, and centrifuged at 14,000 g for 15 min. The proteins were labelled using 3Dye 2D DIGE Kit (Lumiprobe, Russia) in accordance with the manufacturer’s instructions.

Differential 2D gel-electrophoresis and tryptic digestion of the proteins were performed as described earlier [32]. Mass spectra were recorded on an Ultraflex II MALDI-ToF-ToF mass spectrometer (Bruker Daltonics) equipped with an Nd laser. The [MH]+ molecular ions were measured in reflector mode; the accuracy of mass peak measurement was 70 ppm.

Data analysis and presentation

Data analysis and construction of charts were performed using the GraphPad Prism version 9.5 (GraphPad Software, USA). SDS-PAGE analysis was performed on a Gel Doc XR + system (Bio-Rad) using Image Lab software. Differential 2D gels were scanned on a Typhoon Trio scanner (Amersham). Identification of protein hemolymph was carried out by peptide fingerprint search by Mascot software (Matrix Science, USA) through the G. mellonella protein database. Protein scores greater than 45 are supposed to be significant (p < 0.05).

Results

Preparing of protease S

PrtSt (PrtS truncated) accumulated in cells mostly in a soluble activity form and was purified by immobilized metal affinity chromatography (IMAC) and gel permeation chromatography (GPC). Details on protein purification are presented in Fig. 1; Table 1. The purified protease was not electrophoretically homogeneous, which we attribute to autoproteolysis. Along with the major band corresponding to the 35.7 kDa protein (band B36 in Fig. 1A), we also detected a minor constituent (∼ 10%) of approximately 31.6 kDa in the gel (band B32 in Fig. 1A). However, mass spectrometry demonstrated that both the major and minor bands match the protease S (Fig. S4); indicating that the minor constituent is a degradation product.

Fig. 1
figure 1

Purification of protease S (PrtSt) and analysis of the purified protein. A SDS-PAGE analysis. M, molecular weight standards; Lysate, cell lysate of E. coli T7 Express lysY/Iq (pETDuet PrtSt); IMAC, PrtSt sample after immobilized metal affinity chromatography; GPC, PrtSt sample after gel permeation chromatography. The arrows indicate the major (B36) and minor (B32) bands of PrtSt with a mass of ∼ 35.7 kDa and ∼ 31.6 kDa. B Analysis of purified PrtSt by gel permeation chromatography on a Superdex 75 10/300 GL column. The molecular weight of PrtSt was calculated from the calibration curve

Full size image
Table 1 Purification of the recombinant protease S (PrtSt)
Full size table

The analysis of the obtained PrtSt preparation by GPC demonstrated that the enzyme is in the solution as a monomer (Fig. 1B). The molecular weight of the purified protease determined by SDS-PAGE and GPC corresponds to the mature enzyme (an N-terminal propeptide of about 50 amino acids is removed during PLP maturation). The processing point was identified through N-terminal sequencing of the major and minor components (B36 and B32, respectively, in Fig. 1) of the purified PrtSt. The major N-terminal sequence of B36 is VINES, which corresponds to the previous data for mature PrtS [13]. No such N-terminal sequence was found for B32. Additionally, both samples were found to contain PDNGD and PQPGH (Fig. S4), which likely corresponds to degradation products of the enzyme.

Thus, a preparation of recombinant protease S was acquired with around 95% of the desired enzyme, with the full-length mature protease form comprising approximately 85% of the overall protein.

Enzymatic properties of protease S

The enzymatic activity of PrtSt was evaluated using Abz-RSVIK(Dnp) as a substrate. The relationship between the rate of hydrolysis and the substrate concentration indicated the enzyme inhibition by the substrate (Fig. 2), while no such inhibition was observed for Pln [25, 26]. Abz-RSVIK(Dnp) hydrolysis by PrtSt can be described by the following parameters: KM = 13 ± 3 μM, kcat = 38 ± 4 s-1, and substrate inhibition constant KiS = 310 ± 100 μM (presented as mean ± standard error (SE)).

Fig. 2
figure 2

The dependence of the initial rates of Abz-RSVIK(Dnp) hydrolysis by protease S (PrtSt) on the substrate concentration. PrtSt concentration is 50 pM. Three independent measurements were carried out for each experimental point. Values are represented as mean ± SD

Full size image

The optimum activity of PrtSt occurs at neutral pH (Fig. 3), which agrees with the published data on protease S from Photorhabdus sp. Az29 (amino acid sequence identity of the enzymes is 81%) and is typical of most metalloproteases of the M4 family [33, 34].

Fig. 3
figure 3

Effect of pH on protease S (PrtSt) activity for Abz-RSVIK(Dnp) hydrolysis. Three independent measurements were carried out for each experimental point. Values are represented as mean ± SD

Full size image

PrtSt shows no high thermal stability (Fig. 4), which agrees with the published data on protease S from Photorhabdus sp. Az29 [1]. After exposure at 40 and 50°С for 2 h, PrtSt activity is reduced by about a quarter and a half, respectively; while the exposure at 60°С for 30 min reduces it by over 80%.

Fig. 4
figure 4

Protease S (PrtSt) activity after incubation at 40 (●), 50 (▼), 60 (▲), and 70 °C (■). A0 and A, Abz-RSVIK(Dnp) hydrolysis activity by PrtSt before and after incubation, respectively. Three independent measurements were carried out for each experimental point. Values are represented as mean. SD is below 0.1 for all points

Full size image

The activity of PrtSt is effectively inhibited by 1,10-phenanthroline, whereas EDTA does not inhibit the enzyme completely even at 50 mM (Table 2). The same inhibition pattern was previously demonstrated for protease S from Photorhabdus sp. Az29 [1] as well as some other PLPs [33]. PrtSt activity, as well as other zinc metalloproteases [34], is inhibited by high (more than 0.05 mM) concentrations of Zn2+ (Table 2). Such an effect for PLPs was demonstrated for the first time.

Table 2 Effect of chelating agents and Zn2+ on protease S (PrtSt) activity
Full size table

We studied the sensitivity of PrtSt to two protein inhibitors of M4 peptidases, photorin (Phin) from P. laumondii [20] and emfourin (M4in) from S. proteamaculans [15], both of which belong to the recently discovered family I104 (https://www.ebi.ac.uk/merops/cgi-bin/pepsum?id=I104.001;type=I). Phin and M4in act as potent slow-binding competitive inhibitors of PrtSt (Fig. S3). Their inhibition constants (Ki) were determined to be 210 ± 10 and 190 ± 40 pM, respectively. Comparing these values, as well as our determined Ki for the M4in-Tln pair (660 ± 210 pM) (Fig. S3), with the previously determined Ki values for Phin-Tln, Phin-Pln [20], and M4in-Pln [15] interactions makes it clear that inhibitors from the same species most efficiently inhibit PLPs (Fig. 5).

Fig. 5
figure 5

Comparison of inhibition constants (Ki) for photorin (Phin) and emfourin (M4in) effect on protease S (PrtS), protealysin (Pln), and thermolysin (Tln). Constants are represented as mean + SE. Published Ki values for Pln and Tln [16] were used

Full size image

The specific proteolytic activity of the obtained PrtSt determined by Azocasein hydrolysis was 28,240 ± 1,577 U/mg, which is consistent with the published value (26,518 U/mg) for protease S from Photorhabdus sp. Az29 [1]. At the same time, the activity of PrtSt was significantly higher compared to the Tln (18,940 ± 1,322 U/mg) and Pln (7,170 ± 1,330 U/mg) samples employed in this study.

Insect toxicity and melanization

The study utilized G. mellonella, the greater wax moth, as a model to investigate the insect toxicity of PrtS and its potential as a melanization factor. Similar tests were also conducted on Pln and Tln. Figure 6 presents the dependence of insect melanization and toxicity on the concentration of enzymes, while Table 3 shows the median lethal (LD50) and effective (ED50) doses inducing melanization in half of the animals.

Fig. 6
figure 6

Mortality (red) and melanization (blue) of Galleria mellonella larvae 24 h after injection of protease S (PrtSt), protealysin (Pln), or thermolysin (Tln)

Full size image
Table 3 LD50 and ED50 (melanization) under the influence of protease S (PrtSt), protealysin (Pln), and thermolysin (Tln) on Galleria mellonella
Full size table

Effect of protease S on hemolymph proteins in G. mellonella

To identify potential protease S substrates, hemolymph samples from G. mellonella larva were treated with the purified PrtS The protease-treated and control specimens were analyzed by two-dimensional difference gel electrophoresis (Fig. 7). Differentially represented protein spots were identified using a MALDI mass-spectrometry after trypsinolysis (Table 4).

Fig. 7
figure 7

Two-dimensional difference gel electrophoresis of hemolymph G. mellonella. Hemolymph proteins after treatment with PrtSt were labelled with Cyanine3 (green). Hemolymph proteins without PrtSt treatment were labelled with Cyanine5 (red). PrtSt incubated without hemolymph was labelled with Cyanine2 (blue). The arrows indicate the identified protein spots, the numbers correspond to Table 4

Full size image
Table 4 Potential protein targets of protease S in G. mellonella hemolymph
Full size table

Discussion

The initial outcome of this study is the production of recombinant protease S. Due to the enzyme’s propensity for degradation, which complicated manipulations with it, we opted to eliminate the protein’s C-terminal segment and conduct purification in the presence of an inhibitor. After analyzing the resultant sample, we found that the protease’s specific activity, thermostability, pH optimum, and inhibition pattern were not impacted by the treatment, thus enabling the enzyme’s utilization for further research.

For the first time, the study examined the impact of recently discovered protein protease inhibitors of the I104 family on PrtS. These inhibitors’ genes in bacteria are colocalized with the genes of protealysin-like proteases (PLPs) [15]. The analysis of the obtained and published data [20] established that I104 inhibitors show significant species-specificity: Phin from P. laumondii most effectively inhibited PrtS activity just as M4in from S. proteamaculans inhibited Pln (Fig. 5). This finding affirms that cognate proteases and inhibitors have undergone evolutionary adaptation, and consequently endogenous proteases represent the natural targets of I104 inhibitors. Therefore, it is crucial to investigate the function of PLPs, especially in the P. laumondii model with regard to the corresponding inhibitors.

For the first time, we established the dependence of the insect toxic effect and melanization on the amount of PrtS (Fig. 6). The insect toxicity of PrtS proved much lower compared to its homologs from the non-entomopathogenic bacteria, Pln and Tln (Table 3). This can be attributed to the low specific activity of PrtS. Nevertheless, the efficiency of hydrolysis of protein (see Results, 3.2. Enzymatic Properties of PrtSt) and Abz-RSVIK(Dnp) peptide (kcat/KM is 2.9 ± 0.4 s-1 μM-1 for PrtSt (this work); 2.0 ± 0.1 s-1 μM-1 for Pln; and 1.5 ± 0.1 s-1 μM-1 for Tln [25]) by PrtS was higher compared to two other enzymes. In contrast to insect toxicity, the melanization ability was indistinguishable for the three enzymes (Table 3). At the same time, it was shown convincingly that both insect toxicity and melanization are defined by the enzyme activity [13]. Thus, it can be inferred that the differences in toxicity observed are attributable to the fine substrate specificity of proteases. Earlier findings on Tln [50] and elastase (LasB) from Pseudomonas aeruginosa [51] suggested a direct relationship between the insect toxicity of M4 proteases and insect melanization. The data obtained here for PrtS argues against this since pronounced melanization is observed in the presence of non-lethal enzyme doses (Fig. 6). Overall, it can be stated that both as insectotoxins and melanization inducers, homologues from non-entomopathogenic bacteria are not inferior to PrtS. Thus, it is not improbable that these PrtS properties are not specific and are not essential for the infection process.

PrtS is a secretory enzyme that can be found in the hemolymph of infected insects [19], suggesting specific targets for this protease in the hemolymph. The proteomic analysis of G. mellonella hemolymph exposed to PrtS was conducted, leading to the identification of potential protein substrates of the enzyme (Table 4). No published experimental findings regarding the function of this proteins in G. mellonella are available, but their orthologs in other organisms are known. Analysis of these data suggests that many of the proteins revealed are associated with the immune response, in particular, in insects. Adapting to the insect immune response is an important factor in the interaction of P. laumondii and other entomopathogenic bacteria with the host. In particular, these bacteria produce a number of low molecular weight compounds and proteins that affect the insect host immunity by acting on host proteins [52,53,54,55]. Thus, effector proteins and toxins such as LopT, TccC3, TccC5, Mcf1 and Photox secreted by Photorhabdus suppress the host insect cellular response through inhibition of phagocytosis by insect haemocytes [56]. In addition, Photorhabdus, like other entomopathogenic bacteria, release secondary metabolites that affect host insect immunity by inhibiting eicosanoid biosynthesis, resulting in suppression of melanization and nodule formation [52, 57]. Apparently, PrtS is in this category and its function is to modulate the insect immune response to bacterial infection.

Conclusion

The data obtained supports the previous proposal that photorin from P. laumondii is a specific inhibitor of PrtS and suggests an evolutionary co-adaptation between I104 inhibitors and protealysin-like proteases. The insect toxicity of PrtS is presumably negligible for the bacterial infection process since its role is to impair the insect’s immune system. Therefore, future investigations concerning the biological function of PLPs on the P. laumondii model should concentrate on the detailed analysis of PrtS interactions (given the involvement of photorin) with the immune system during in vivo infection.