Linocin M18 protein from the insect pathogenic bacterium Brevibacillus laterosporus isolates

Abstract Brevibacillus laterosporus (Bl) is a Gram-positive and spore-forming bacterium. Insect pathogenic strains have been characterised in New Zealand, and two isolates, Bl 1821L and Bl 1951, are under development for use in biopesticides. However, growth in culture is sometimes disrupted, affecting mass production. Based on previous work, it was hypothesised that Tectiviridae phages might be implicated. While investigating the cause of the disrupted growth, electron micrographs of crude lysates showed structural components of putative phages including capsid and tail-like structures. Sucrose density gradient purification yielded a putative self-killing protein of ~30 kDa. N-terminal sequencing of the ~30 kDa protein identified matches to a predicted 25 kDa hypothetical and a 31.4 kDa putative encapsulating protein homologs, with the genes encoding each protein adjacent in the genomes. BLASTp analysis of the homologs of 31.4 kDa amino acid sequences shared 98.6% amino acid identity to the Linocin M18 bacteriocin family protein of Brevibacterium sp. JNUCC-42. Bioinformatic tools including AMPA and CellPPD defined that the bactericidal potential originated from a putative encapsulating protein. Antagonistic activity of the ~30 kDa encapsulating protein of Bl 1821L and Bl 1951during growth in broth exhibited bacterial autolytic activity. LIVE/DEAD staining of Bl 1821L cells after treatment with the ~30 kDa encapsulating protein of Bl 1821L substantiated the findings by showing 58.8% cells with the compromised cell membranes as compared to 37.5% cells in the control. Furthermore, antibacterial activity of the identified proteins of Bl 1821L was validated through gene expression in a Gram-positive bacterium Bacillus subtilis WB800N. Key Points • Gene encoding the 31.4 kDa antibacterial Linocin M18 protein was identified • It defined the autocidal activity of Linocin M18 (encapsulating) protein • Identified the possible killing mechanism of the encapsulins Supplementary Information The online version contains supplementary material available at 10.1007/s00253-023-12563-8.


Introduction
Brevibacillus laterosporus (Bl) is a Gram-positive and spore-forming bacterium belonging to the Brevibacillus brevis phylogenetic cluster (Shida et al. 1996). Numerous strains exhibit pathogenicity against insect pests of different orders including Coleoptera (Kouadio et al. 2021;Salama et al. 2004), Diptera (Bedini et al. 2020;Carramaschi et al. 2015), Lepidoptera (van Zijll de Jong et al. 2016), and also against nematodes (Hamze and Ruiu 2022) and molluscs (De Oliveira et al. 2004) have been documented. Due to the entomocidal action of Bl against multifarious insect pests, there has been a surge of interest in commercialisation for future microbial pesticide development (Baum et al. 2020;Boets et al. 2011;Floris et al. 2008;Glare et al. 2014;Glare et al. 2018;Sampson et al. 2016). In New Zealand, three insect pathogenic strains (Bl 1821L, Bl 1951 demonstrating activity against diamondback moth and mosquitoes (van Zijll de Jong et al. 2016) have been characterised and found to be genetically distinct from the strains of other parts of the world (Glare et al. 2020). Strains Bl 1821L and Bl 1951 are under development as biopesticides, but due to difficulties in mass production, it has been often problematic to harness their insecticidal potential. Therefore, a study was undertaken to find the cause of stunted growth of these two isolates.
Prokaryotes including bacteria often harbour nanocompartments within the cytoplasm (Wiryaman and Toor 2022). These protein-based compartments play key roles in various metabolic and ecological processes ranging from iron homeostasis to carbon fixation (Giessen et al. 2019;McDowell and Hoiczyk 2022). One such class is the encapsulins whose shell proteins seize a fold previously found only in viral capsids (Suhanovsky and Teschke 2015;Wikoff et al. 2000). Encapsulins were first identified in 1994 as a high molecular-weight (HMW) complex in the culture supernatant of a Gram-positive bacterium Brevibacterium linens M18 which exhibited bacteriostatic activity against various strains of Arthrobacter, Bacillus, Brevibacterium, Corynebacterium, and Listeria while N-terminal sequencing of the antagonistic protein identified the encoding gene, lin, in the chromosomal DNA (Eppert et al. 1997;Valdes-Stauber and Scherer 1996). Since then, encapsulins have been described in various bacteria including Mycobacterium leprae (Winter et al. 1995), M. tuberculosis (Rosenkrands et al. 1998), Thermotoga maritima (Hicks et al. 1998), Streptomyces coelicolor (Kawamoto et al. 2001), and Quasibacillus thermotolerans (Giessen et al. 2019). Furthermore, several authors have highlighted the nature, structural features, and the role of these bacterial nanocompartments (Almeida et al. 2021;Andreas and Giessen 2021;Chmelyuk et al. 2023;Giessen 2022;Giessen and Silver 2016;Giessen and Silver 2017;Jones and Giessen 2021;McDowell and Hoiczyk 2022;Wiryaman and Toor 2022).
Herein, we report on the isolation, purification, and characterisation of Linocin M18, a putative antibacterial (encapsulating) protein of the insect pathogenic strains Bl 1821L and Bl 1951.

Bacterial strains and growth conditions
The insect pathogenic strains Bl 1821L and Bl 1951, originally isolated from Brassica seeds (van Zijll de Jong et al. 2016) and held at −80 °C in the Bio-Protection Research Centre Culture Collection, Lincoln University, New Zealand, were used in this study. The strains were routinely grown in Luria-Bertani medium broth (LB Miller, Sigma, St. Louis, MI, USA) on an orbital shaker (Conco,TU 4540,Taibei,Taiwan) at 250 rpm and 30 °C overnight for further usage.

Mitomycin C induction of putative antibacterial proteins (ABPs)
To induce the putative ABPs residing in the genome of Bl 1821L and Bl 1951, a modified protocol of Rybakova et al. (2014) was used. Single colonies of Bl 1821L/Bl 1951 were used to inoculate 5 mL of LB broth. The inoculated culture was incubated overnight at 30 °C with shaking on rotatory shaker (Conco, TU 4540, Taiwan) at 250 rpm. Aliquots (500 μL) of the overnight culture of Bl 1821L/Bl 1951 were used to inoculate 25 mL of LB broth in 250 mL conical flasks. This culture was allowed to grow at 30 °C with shaking at 250 rpm until the culture attained turbidity (10-12 h). Mitomycin C (Sigma, Sydney, NSW, Australia) was added into the flasks, which were left shaking overnight at 40 rpm and ambient temperature (24 °C). The flasks were monitored for signs of lysis (clearing of the culture or accumulation of bacterial debris). For induction of putative ABPs of Bl 1821L, mitomycin C at 1 μg/mL (Babar et al. 2022a) and for Bl 1951 at 3 μg/mL (Babar 2021) was used. After induction, the culture was centrifuged at 16,000 × g for 10 min, and the supernatant was passed through a 0.22 μm filter (Merck Millipore, UK) to obtain the cell free supernatants (CFS). Antibacterial activity of the CFS against the Bl 1821L and Bl 1951 as the host bacterium was determined using the Kirby-Bauer disc diffusion test (Kirby et al. 1956).

Purification of putative ABPs
The crude lysates of Bl 1821L and Bl 1951 harbouring the putative ABPs were purified using the sucrose density gradient centrifugation method (Babar et al. 2022b). For the purification, 7.5 mL CFS was transferred into polypropylene centrifuge tubes (Konical, Beckman, Brea, CA, USA) and ultracentrifuged at 151,263 × g in a swing bucket rotor (41Ti, Beckman, Brea, CA, USA) for 70 min. The concentrated pellet was resuspended in 100-150 μL of tris buffer saline (10 × TBS) (40 g NaCl, 1 g KCl, 2.42 g TRIS Base, 16.5 g TRIS-HCl, dH 2 O 490 mL, pH 7.4). Sucrose density gradients (10%, 20%, 30%, 40%, and 50%) were made by applying layers of 1.5 mL of freshly prepared sucrose solution. The highest concentration of sucrose was applied first followed by the lower concentrations. The Bl 1821L andBl 1951 ultracentrifuged preparations (200 μL) were applied on top of the sucrose density gradients and centrifuged at 151,263 × g in a swing bucket rotor (41Ti, Beckman) for 70 min. After ultracentrifugation, each gradient was carefully taken out and tested against Bl 1821L and Bl 1951 as the host bacterium following the Kirby-Bauer disc diffusion assay test. Next, each of the gradients (1.5 mL) was transferred into polypropylene centrifuge tubes to make the final volume of 7.5 mL with the addition of TBS and ultracentrifuged at 151,263 × g for 70 min to concentrate the ABPs. Based on our previous work (Babar et al. 2022b), the purified and concentrated solution of 20% sucrose density gradient of Bl 1821L and 50% of Bl 1951 were cleaned from sucrose residues using an Amicon Ultra-0.5 (10 kDa) centrifugal filter (Merck Millipore, Cork, Ireland) for sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and transmission electron microscopy (TEM).

TEM analysis of crude and purified lysates
For TEM analysis of crude or purified lysates, 5 μL sample of the concentrated pellet was applied to a freshly glowdischarged plastic-coated hydrophilic 200 mesh EM grid (ProSciTech; Thuringowa, Australia) and stained with 3 μL of 0.7% uranyl acetate (UA, pH 5). The sample was subjected to TEM analysis at 18,000 to 25,000 magnifications using a Morgagni 268D (FEI, Hillsboro, OR, USA) TEM operated at 80 KeV. The images were photographed using the TENGRA camera.

SDS-PAGE analysis of purified putative ABPs
The highly purified proteins of Bl 1821L and Bl 1951 were subjected to SDS-PAGE (10%) analysis using the protocol of Laemmli (1970). Ten microlitre of protein ladder (BIO-RAD, Precision Plus Protein™ Standards, Hercules, CA, USA) was loaded. Gels were run for 50 min at 200 volts and then washed four times with H 2 O before staining. For N-terminal sequencing, the protocol of Yasumitsu et al. (2010) involving the RAMA stain (29 mL MQW (milli que water), 12.5 mL CBB (coomassie brilliant blue) stock (1 gm CBB, 300 MeOH, 200 mL MQW), 3.75 mL ammonium sulphate (200 gm, 500 mL MQW), 5 mL of glacial acetic acid per gel) was followed. Prior to overnight destaining in water, the gel was rinsed in water. In addition, the purified and concentrated proteins were also assessed with the silver staining method of Blum et al. (1987).

N-terminal sequencing and bioinformatic analysis
Purified bands of the protein of interest of Bl 1821L and Bl 1951 were excised for N-terminal sequence analysis. Peptides generated by trypsin (Promega, Madison, WI, USA) cleavage were analysed using liquid chromatography mass spectrometry (LC-MS). The obtained peptide masses were compared with the National Centre for Biotechnology Information (NCBI) library using the genomes of Bl 1821L (GenBank accession NZ_CP033464.1) and Bl 1951 (GenBank accession RHPK01000003, contig 1). Amino acid sequences from the genomes were assessed and characterised using UniProt database (https:// www. unipr ot. org; accessed 25 September 2019), BLASTp (Basic Local Alignment Search Tool) (https:// blast. ncbi. nlm. nih. gov; accessed 23 April 2020), ExPasy (https:// www. expasy. org; accessed 23 April 2020), and the CLUSTALO (Clustal Omega) (https:// www. unipr ot. org; accessed 23 April 2020). The identified ~30 kDa putative antibacterial protein of Bl 1821L and Bl 1951 was analysed using Geneious basic (Kearse et al. 2012). Amino acid sequence alignments were performed using CLUSTALO (https://www. unipr ot. org; accessed 23 April 2020). The genomes of both the insect pathogenic strains were searched for the presence of secondary metabolism biosynthesis gene clusters using antiSMASH (antibiotics and secondary metabolite analysis shell) (https:// antis mash. secon darym etabo lites. org; accessed 23 April 2020) and BAGEL4 (identifying the prokaryotic genes involved in the biosynthesis of ribosomally synthesised and post translationally modified peptides (RiPPs) and (unmodified) bacteriocins) (http:// bagel4. molge nrug. nl; accessed 23 April 2020) (Medema et al. 2011;van Heel et al. 2018). The antiSMASH-predicted gene clusters were compared with the similar known gene clusters using the programme MIBiG 3.0 (minimum information about a biosynthetic gene cluster) (https:// mibig. secon darym etabo lites. org; accessed 10 April 2023) (Terlouw et al. 2023). Bioinformatic tools AMPA (http:// tcoff ee. crg. cat/ apps/ ampa; accessed 23 April 2020) (Torrent et al. 2012) and CellPPD (Designing of Cell Penetrating Peptides) (http:// crdd. osdd. net/ ragha va/ cellp pd; accessed 23 April 2020) (Gautam et al. 2013) were used to identify the motifs related to bactericidal activity and potency of cell penetrating peptides (CPPs) of 31.4 kDa putative encapsulating protein (EP) of Bl 1821L and Bl 1951. AMPA is an automated web server used to predict the antimicrobial regions in a protein. It is based on an antimicrobial propensity scale that takes into account the physical and chemical properties of each amino acid like hydrophobicity and amphipathicity and the relevance of amino acid position for antimicrobial activity (Torrent et al. 2012). In this study, an antimicrobial index (AI) set at a default value of 0.225 was used, and an AI of ˂ 0.225 was considered a positive hit for an antimicrobial peptide (AMP). CellPPD is a support vector machine (SVM) that scores each amino acid residue with an SVM score, and an SVM >0 is considered a positive CPP (cell-penetrating peptide) hit (Gautam et al. 2013).

Bactericidal activity assay
To determine the bactericidal activity of the putative ABPs of Bl 1821L and Bl 1951 in crude and purified forms, growth 1 3 assay was performed. Five hundred microlitre of crude lysate harbouring the putative ABPs was pipetted into 5 mL of the overnight grown culture of the host bacteria. For the control treatment, a similar volume of LB broth/TBS buffer without putative ABPs was used. All the flasks with/without putative ABPs were maintained at 30 °C on a shaking incubator (Conco, TU 4540, Taiwan). Samples were drawn from each treatment after 1, 3, 6, 12, 18, and 24 h of incubation to determine the number of viable cells (CFU/mL) and OD 600nm . Cell biomass (OD 600nm ) was measured using the Ultrospec-10 spectrophotometer (Amersham Biosciences, Amersham, UK). To determine the number of viable cells (CFU/mL), tenfold serial dilutions (10 −1 to 10 −6 ) of each time interval were prepared, and 100 μL from each dilution was spread in duplicate on two independent LB agar plates. After incubation at 30 °C for 2-3 days, the colonies were counted using a colony counter (Cole-Palmer™ Stuart™, Thermo Fisher Scientific, UK) and converted into log 10 CFU/mL. To define potential ABP antagonistic activity, the percentage change in number of viable cells compared to the control treatment (without ABPs) after treatment with the putative ABPs was calculated. Four independent sets of experiments with biological replications were performed, and the pooled data were subjected to statistical analysis using the ANOVA (analysis of variance) test through the Genstat 20th edition programme. Similar to this, an experiment with the purified ~30 kDa putative EP of Bl 1821L and Bl 1951 was performed.

LIVE/DEAD staining
Ten milliliters of overnight culture of the host bacterium (Bl 1821L or Bl 1951) was aliquoted into two parts. One part was treated with the ~30 kDa-purified putative EP, and the other part, serving as the control, was treated with the TBS buffer. Two volumes (100 μL and 200 μL) of purified putative EPs (~30 kDa) were used. All the treatments with/without TBS or putative EPs were incubated at 30 °C, and the samples were drawn from each treatment (1, 3, 6, 12, 18, 24 h) for LIVE/DEAD staining. For fluorescent microscopy, 5 μL of SYTO9/propidium iodide (PI) stain (LIVE/DEAD® BacLight™ bacterial viability assay kit; Invitrogen, Carlsbad, CA, USA) was mixed with an equal volume of host bacteria with/without ~30 kDa-purified putative EP for 5 min in a UV-safe tube. Next, 5 μL of this mixture was pipetted onto a slide, and 3 μL of molten agarose (0.1%) was added to reduce the cell movement. Cells were examined under a BX51 fluorescent microscope (Olympus, Sydney, NSW, Australia) at × 100 magnification. An Olympus DP74 camera (Olympus, Sydney, NSW, Australia) attached to the microscope was used to take the images with the help of CellSens 2.1 software (Olympus, Sydney, NSW, Australia). SYTO9 (green) stain was visualised using a FITC filter (Olympus, Sydney, NSW, Australia) with excision of 550 nm wavelength, and for PI stain (red), a wavelength of 650 nm was used. Immediately, both the images were overlapped using the CellSens 2.1 software to visualise the status of cells (alive or dead).

Expression of putative ABPs in the Gram-positive bacterium B. subtilis WB800N
Genomic DNA of Bl 1821L was extracted using the DNeasy Blood and Tissue Kit (250) Figure S1) were cloned (pHT01) into a Grampositive bacterium Bs WB800N expression system. Three constructs were made, pHT01-hypo (25 kDa), pHT01encap (31.4 kDa), and pHT01-hypo.encap containing both the genes. To form these, vector primers (Table 1.) were used to amplify each region and their combinations using CloneAmp (TakaRa Bio, San Jose, CA, USA) master mix according to the manufacturer's instructions. The bands of appropriate size were visualised and excised on a DarkReader blue light table (Clare Chemical Research, Dolores, CO, USA) before purification using a NucleoSpin gel and PCR purification kit (Macherey-Nagel, Allentown, PA, USA) according to the manufacturer's instructions.

Cloning into the plasmid pHT01
PCR-amplified sequences were cloned into plasmid pHT01 (MoBiTec GmbH, Göttingen, Germany) ( Figure S2) using an Infusion HD cloning kit (TakaRa Bio, San Jose, CA, USA). Briefly, plasmid pHT01 was digested with BamHI and XmaI (The New England BioLabs, Ipswich, MA, USA) then gel purified as above. Purified linear plasmid and respective PCR products were used in the infusion cloning reaction, according to the manufacturer's instructions. Cloning reactions were then used to transform Escherichia coli stellar cells (Clontech Laboratories, TakaRa Bio, San Jose, CA, USA) according to manufacturer's instructions with selection on LB agar containing 100 μg/mL carbenicillin. Plasmid DNA was extracted using a NucleoSpin Plasmid Mini kit (Macherey-Nagel, USA). Plasmids were sequenced by Macrogen (Seoul, Korea), and sequencing data were aligned to those of the respective constructs using Geneious Prime-2020 (Biomatters, New Zealand) and then manually curated. Plasmid DNA from one correct construct was used to transform Bs WB800N. The Gram-positive bacterium, Bs WB800N (nprE aprE epr bpr mpr::ble nprB::bsr Δvpr wprA::hyg cm::neo; NeoR) was used for transformation (Jeong et al. 2018;Wu et al. 2002). Electrocompetent cells of Bs WB800N were prepared following the protocol of Lu et al. (2012). Validated Bs WB800N transformants of pHT01-hypo (A1, A2), pHT01encap (B1, B2), and pHT01-hypo.encap were independently cultivated overnight in LB broth until the culture attained OD 600nm 0.7-0.8 and aliquoted into two parts. An aliquot (7.5 mL) was induced with 1 mM IPTG (Isopropyl β-D-1thiogalactopyranoside), and the other was uninduced. Both induced and uninduced samples were harvested after 3.5 and 24 h of treatment. The resultant cultures were centrifuged at 6000 × g for 10 min at 4 °C to spin down the cells. Cell pellets were stored at −20 °C, and the supernatant filter was sterilised by passing twice through a 0.22 μm filter.

Assay test of expressed proteins
CFS of Bs WB800N containing the (uninduced and induced) proteins corresponding to the pHT01-hypo (25 kDa), pHT01-encap (31.4 kDa), and pHT01hypo.encap encoding both the proteins was evaluated using the Kirby-Bauer disc diffusion test (Kirby et al. 1956) against Bl 1821L and Bl 1951 as the host bacteria. CFS of Bs WB800N without transformation (pHT01-hypo, pHT01-encap, pHT01-hypo. encap) was tested for bioactivity, and none was detected.

Bioactivity assay, electron microscopy, and SDS-PAGE analysis of putative ABPs
Mitomycin C-induced CFS of Bl 1821L and Bl 1951in the bioactivity assay against the producer and the vice versa strain produced a prominent zone of inhibition, and no such activity was noted in the negative control ( Figure S3). TEM analysis of the crude lysates of Bl 1821L (Fig. 1a) and Bl 1951 ( Fig. 1b) showed similar structures including globular or capsid, hollow sheath, contractile phage tail sheath, and polysheath-like structures. Purified lysates of Bl 1821L and Bl 1951 after 10 kDa molecular weight cutoff (MWCO) membrane concentration yielded a single protein band of ~30 kDa (Fig. 2a, c). Assessments of the concentrated protein of Bl 1821L under electron microscope predominantly displayed the presence of globular or hexagonal structures that resembled phage capsids (Fig. 2b) while for Bl 1951 globular or phage capsid-like structures and long nanotubes similar to polysheaths were seen (Fig. 2d).

Identification of ~30 kDa putative antibacterial protein
The amino acid sequences derived from N-terminal sequencing of the purified ~30 kDa protein band were used to predict the corresponding gene in the Bl 1821L genome (GenBank accession NZ_CP033464.1). Analysis identified several matches to a hypothetical protein (A0A502I846, covering 57% of the amino acid sequence; Figure S8) and a bacteriocin family protein (A0A502IA18, covering 78% of the amino acid sequence; Figure S9). The genes encoding the hypothetical protein (GenBank accession WP_113757161) and the putative bacteriocin family protein (GenBank accession WP_113757162) residing adjacently were identified (Fig. 3, Figure S10). Amino acid sequence analysis through ExPasy computed the molecular weight of the identified hypothetical and bacteriocin family proteins of Bl 1821L to be 25 kDa and 31.4 kDa. The identification of two proteins suggested co-migration of proteins. The molecular mass of the predicted 31.4 kDa protein was within the expected range estimated by SDS-PAGE analysis, i.e., ~30 kDa (Fig. 2a). The same genes and predicted proteins were also identified in the Bl 1951 genome (GenBank accession RHPK01000003.1, contig 1). For Bl 1951, the genes encoding these proteins were annotated as putative encapsulating protein for a DyP-type peroxidase or ferritin-like protein (31.4 kDa) and hypothetical protein (25 kDa) in the GenBank annotated deposits (Fig. 3, Figure  S11).

Bactericidal activity of putative encapsulating protein
AMPA analysis of the 31.4 kDa putative EP sequence of Bl 1821L revealed an index value below the threshold level, which suggested its bactericidal potency ( Figure S12). The region of amino acid similarity covered amino acids −2 to 279 with a propensity value of 0.001 (0%) and a mean value of 0.001 ( Figure S12). CPP specific amino acids and motifs identified in the Bl 1821L (identical to the Bl 1951) EP (31.4 kDa) sequence are presented in Table S4 and Figure S13. The identified genes and amino acid content of N-terminal sequenced and bioinformatically analysed ~30 kDa protein band of Bl 1951 were 100% identical to the identified 31.4 kDa Bl 1821L putative EP. Therefore, the resultant bactericidal activity predicted through bioinformatic tools AMPA and CellPPD is likely to be similar. The crude lysate of Bl 1951 harbouring the putative ABPs in the liquid assay exhibited the autocidal activity by showing a non-significant decrease of 48.4% in the number of viable cells after 6 h of incubation at 30 °C (Table S5, Figure S14). However, for the same time intervals, the addition of ~30 kDa-purified putative EP of Bl 1821L and Bl 1951, antibacterial activities against both the isolates, were noted (Table S7, S9, Fig. 4, Figure S17). Spectrophotometric readings (OD 600nm ) of ~30 kDa-purified putative EP-treated and untreated cultures similar to the crude lysates revealed very small changes across the evaluated time intervals (Table S6, S8, S10, Figure S15, S16, S18).

LIVE/DEAD staining
LIVE/DEAD staining of Bl 1821L-treated culture with the ~30 kDa-purified putative EP of Bl 1821L did not show a higher population of red cells (cells with compromised membranes) after 1 h of incubation than in the control treatment (without EP) (Fig. 5). After 3 h, more red cells were seen with the ~30 kDa putative EP-treated culture as compared to the control where yellow/orange colour cells were predominant (Fig. 5). Bl 1821L cells with compromised cell membranes (red) were prominent after 6 and 12 h of incubation with the purified putative EP, which was similar to the control (Fig. 5). Notably, the appearance of red cells coincided with the decrease in the number of viable cells (48.9%) 6 h after treatment with the purified putative EP but in contrast to the 12-h treatment (Table S7, Figs. 5 and 6). On a percentage basis, 6 h after treatment, 58.8% of the putative EP-treated and 37.5% of the untreated cells were red (Fig. 6). LIVE/DEAD staining of Bl 1821L-treated culture with the ~30 kDa-purified putative EP of Bl 1821L at a Fig. 4 Number of viable cells (log 10 CFU/mL) of Bl 1821L and Bl 1951 with/without treatment of purified Bl 1821L putative encapsulating protein (~30 kDa) after incubation at 30 °C for various time intervals higher volume (200 μL) resulted in more cells with compromised membranes (red) from the initial 1 h of incubation up to 6 h when compared to the control (without EP). However, yellow/orange cells in the Bl 1821L culture with the purified ~30 kDa EP were dominant from 12 to 24 h of incubation at 30 °C ( Figure S19).

Expression of B. laterosporus 25 kDa and 31.4 kDa proteins in B. subtilis WB800N
The CFS of the expressed proteins pHT01-hypo (25 kDa), pHT01-encap (31.4 kDa), and the both (pHT01-hyo.encap) were used for further analysis. Bs WB800N supernatant (pHT01-hypo, A1) expressing the 25 kDa hypothetical protein demonstrated antagonistic activity against Bl 1951 by developing a zone of inhibition of 11.7 mm and 12 mm with induced and uninduced cultures respectively after 3.5 h of induction (Table S11, Figure S20b). For Bl 1821L, the same supernatant demonstrated no significant activity (Table S12, Figure S20a). Similarly, 25 kDa hypothetical protein (pHT01-hypo, A2) after 3.5 h of induction exhibited similar activity against both strains (Bl 1821L andBl 1951) by developing a zone of inhibition of 11.7 mm and 11 mm respectively. However, after 24 h of induction, antagonistic activity was almost absent (Table S11, S12, Figure S21a, S21b). Purification of the corresponding protein was only visualised on SDS-PAGE with the 25 kDa hypothetical protein sample (pHT01-hypo, A1) ( Figure S20c), and no protein bands were seen with the other concentrated protein (pHT01-hypo, A2). Unexpectedly, a defined band of ~48 kDa in the 40% gradient, faint bands of the same level in other gradients, and prominent protein bands of >48 kDa in the 20% gradient were visualised on SDS-PAGE with the purified and concentrated 25 kDa hypothetical protein (pHT01-hypo, A1) ( Figure S20c). Assay test of the CFS of Bs WB800N without any transformation (pHT01-hypo, pHT01-encap, pHT01-hypo.encap) showed no zones of inhibition against Bl 1821L and Bl 1951 (Table S11, S12, Figure  S22). In addition, no bands were observed on silver-stained SDS-PAGE of these concentrated supernatants ( Figure S22). Antibacterial activity was found in CFS (pHT01-encap, B1) containing 31.4 kDa putative EP against Bl 1821L as the host bacterium, developing a lysis zone of 12.3 mm after 24 h of induction and 11.7 mm without induction (Table S12, Figure S23a). For Bl 1951, a halo zone of 11 mm was noted 24 h after induction (Table S11, Figure S23b). SDS-PAGE analysis of the 31.4 kDa-concentrated putative EP (pHT01-encap, B1) after 3.5 h of induction displayed a very faint band of ~30 kDa ( Figure S23c), but after 24 hours of induction, in addition to the ~30 kDa band, other proteins were also present ( Figure S23d). Putative 31.4 kDa EP (pHT01-encap, B2) exhibited inhibitory activity solely against Bl 1821L as the host bacterium by producing a zone of inhibition of 11 mm after 24 h of incubation (Table S12, Fig. 7a), and a purified band of ~30 kDa in 40% gradient was visualised on SDS-PAGE (Fig. 7c).

Discussion
In this study, we determined the identity of an antibacterial protein produced by two insect pathogenic isolates Bl 1821L and Bl 1951. Initially, the genome sequence analysis of the Bl 1821L and Bl 1951through antiSMASH and BAGEL4 revealed the presence of gene clusters associated with the production of non-ribosomally synthesised peptides (petrobactin, bogorol A, basiliskamide A, zwittermicin A, and laterocidine) and ribosomally synthesised peptides (laterosporulin, sactipeptides, UviB, and LAPs). Importantly, among the latter, group matches were found to the previously defined low molecular-weight (LMW) bacteriocins such as laterosporulin (5.75 kDa) of Bl GI-9 (Singh et al. 2012) and UviB (8.78 kDa) of Bt serovar israelensis (Anderson et al. 2005). However, subsequent experimental work involving the isolation and sucrose density gradient purification yielded a homogeneous protein of ~30 kDa on SDS-PAGE. N-terminal sequencing of the ~30 kDa-purified putative antibacterial protein bands of Bl 1821L and Bl 1951 located the putative genes encoding the HMW proteins (bacteriocins) in their genomes. A gene encoding the 25 kD hypothetical protein at the 3′ end of the identified putative EP (31.4 kDa) encoding gene was also found in both the strains. Bioinformatic analyses of the 31.4 kDa EP sequence using the tools AMPA and CellPPD identified the potential bactericidal motifs. The killing potency of the putative ABPs present in the crude and purified lysates of Bl 1821L and Bl 1951 against the cells of both the strains was determined growing in broth and corroborated with the LIVE and DEAD staining. Furthermore, the genes encoding the 25 kDa hypothetical protein, 31.4 kDa putative EP, and both the genes (25 kDa and 31.4 kDa) of Bl 1821L were transformed in a Gram-positive bacterium Bs WB800N for heterologous expression, with activity confirmed for expression of both proteins (Jeong et al. 2018).
Genes encoding the identified putative antibacterial protein of Bl 1821L and Bl 1951 shared >97% amino acid identity to the family 1 encapsulin nanocompartment shell protein belonging to the multispecies of the genus Brevibacillus. Nanocompartment proteins are generally divided into Valdes-Stauber and Scherer 1996), and so far, 31 out of 35 prokaryotic phyla are known to encode encapsulin-like operons (Giessen and Silver 2017). In addition, the nanocompartments also contain ferritin-like compounds or peroxidase enzymes (McHugh et al. 2014). Ferritin family proteins include three sub-families: classical ferritin (Ftn), bacterioferritin (Bfr), and the DNA-binding protein from starved cells (Dps) (Andrews 2010;Theil et al. 2013). The first two categories (Ftn and Bfr) perform diverse functions including ribonucleotide reductase (Åberg et al. 1993), protecting DNA from oxidative damage (Grant et al. 1998), and iron storage (Bradley et al. 2014). However, Dps proteins are unique on account of their role in detoxification of iron as opposed to its storage (Karas et al. 2015;Orban and Finkel 2022), and a failure to detoxify iron may prove fatal to the cells (Bradley et al. 2020;He et al. 2016). Mycobacterium tuberculosis, a pathogenic bacterium, accommodates two putative iron storage proteins, BfrA (Rv1876) a bacterioferritin and BfrB (Rv3841) a ferritin-like protein. The expression of both bfrA and bfrB is regulated by the binding of iron-activated ideR (iron-dependent regulator) to the tandem operator sites present at the 5′ end of the iron storage gene (Cole et al. 1998). Reddy et al. (2012) generated mutants lacking these functional genes. Interestingly, the authors found that the mutant bacterium under iron-deficient conditions not only showed a significant decrease in growth but was also vulnerable to oxidative stress (Reddy et al. 2012). In this context, it is possible that the 31.4 kDa putative EP of Bl 1821L and Bl 1951 might be undergoing iron deficiency, or the ferritin proteins failed to detoxify the excessive iron which contributed towards the stunted growth of both the strains.
Encapsulin, originating from the supernatants of Br. linens M18, was initially designated "bacteriocins" due to their inhibitory effect on the growth of several Gram-positive bacteria, but no lethal effect was noticed against the Gram-negative bacteria (Eppert et al. 1997;Valdes-Stauber and Scherer 1996). The newly found 31 kDa biomolecule (Linocin M18) lost its antagonistic activity upon purification (Valdés-Stauber and Scherer 1994). In our study, the sucrose density gradients containing the purified ~30 kDa putative EP (Linocin M18) of Bl 1951 and Bl 1821L sustained bioactivity in the paper disc assay. Furthermore, the ~30 kDa-purified putative EP of Bl 1951 and Bl 1821L in the liquid assay exhibited an autolytic activity which was evident from the decrease in Typically, the putative ABPs (bacteriocins) inhibit growth and kill a broad array of microbial species or just closely related strains of the same species, and the producer strains are immune to their toxic effects (Ghequire and De Mot 2015;Riley and Wertz 2002). However, some members of a genetically identical population can kill their siblings (autocidal) (Popp and Mascher 2019) like bacteriocin hyicin 3682 (Fagundes et al. 2011). Miljkovic et al. (2019, while assessing the antimicrobial potential of different Bl strains (BGSP7, BGSP9, BGSP11), also found their pesticidal potential against the potato beetle, Leptinotarsa decemlineata, and some crop pathogenic fungi. The authors in their preliminary work noticed the autocidal (selfkilling) activity of Bl BGSP7, Bl BGSP9, and Bl BGSP11 strains which they cited due to the synthesis of numerous antimicrobial molecules. In addition, antiSMASH analysis of the three potential insect pathogenic isolates predicted a 31.4 kDa family 1 encapsulating nanocompartment shell protein (Linocin M18), but the following research revealed that among the several purified antagonistic compounds, one antimicrobial molecule of 1.5 kDa produced by the Bl BGSP7 and Bl BGSP9 and two antimicrobial molecules of 1.5 kDa and 6 kDa produced by Bl BGSP11 were found to be more potent than others. Though, the BLASTp analysis of the 31.4 kDa-identified putative EP protein (Linocin 18) in our study exhibited 97.1%, 97.1%, and 96.4% amino acid identity with the similar protein of Bl BGSP7, Bl BGSP9, and Bl BGSP11 respectively. The Gram-positive bacterium Bacillus subtilis (Bs) exhibits the self-killing activity under nutritional stress environment (Popp and Mascher 2019). The Bs cells at the onset of sporulation secrete extracellular killing factors and lyse the sibling non-sporulating cells that have not developed immunity to these toxins. In addition, the resultant lysis also releases nutrients from the dead cells into the starved medium that the surviving sporulating cells can feed on, this behaviour was called "cannibalism" (González-Pastor 2011;Lamsa et al. 2012). The cannibalistic phenomenon has some biological significance for the bacterial populations as the Bs population could be at risk if most of the members engage in sporulation at once (Höfler et al. 2016).
Electron micrographs of the crude lysates of both the Bl 1821L and Bl 1951 showed similar structures including the globular or phage capsid-like (encapsulating) and contractile phage tail-like structures. Typically, family 1 encapsulin is assembled from a single type of shell protein into compartments of 24 to 42 nm diameter with either T1, T3, or T4  (Akita et al. 2007;Sutter et al. 2008). Therefore, the appearance of the ~30 kDa-purified protein suggests that ~30 kDa assembles as an icosahedral (capsidlike) particle. Recently, we isolated, purified, and characterised the putative phage tail-like protein (bacteriocin) of ~48 kDa from the insect pathogenic isolate Bl 1821L (Babar et al. 2022a). Based on these findings, it can be summarised that the crude lysates of Bl 1821L and Bl 1951 harbour the ~30 kDa and ~48 kDa putative ABPs. Notably, these putative ABPs of both the isolates demonstrated autocidal activity, but the purified ~30 kDa putative EP showed antagonistic activity against Bl 1821L and Bl 1951. Variation in the antagonistic activity of crude lysate (~30 kDa and ~48 kDa) and the purified putative EP (~30 kDa) might be attributed to the prevalence of a toxin-antitoxin mechanism (Hayes and Kędzierska 2014). In this stance, it is likely that in the crude state, either the ~30 kDa or ~48 kDa putative ABP is masking the toxic effect of the other.
Encapsulin proteins seize a fold of capsid protein of lambdoid bacteriophage in their shell protein (formerly referred to as linocin-like protein) (Gabashvili et al. 2020;Suhanovsky and Teschke 2015); therefore, it is likely that encapsulins and the capsid proteins of tailed phages have evolved from a common ancestor (Duda and Teschke 2019; Orlova et al. 2012). However, viral capsids and encapsulin nanocompartments functionally differ; the former transport viral genomes from one cell to another and the latter are involved in metabolic activities (Giessen 2022;Nichols et al. 2017). A search of the identified 31.4 kDa family 1 encapsulin nanocompartment shell protein (Linocin M18) of Bl 1821L and Bl 1951 in the Pfam (protein families) database revealed its association with the clan (CL0373) of phage coat superfamily proteins (Mistry et al. 2021). Virus structural proteins (capsid) often encode various bioactive peptides such as antimicrobial peptides (AMPs) and CPPs (Freire et al. 2015;Järver and Langel 2006). Bioinformatic tool AMPA revealed the bactericidal potential of the 31.4 kDa putative EP, and CellPPD identified the motifs in the bactericidal stretch that have the faculty to penetrate the cells. Typically, CPPs are short cationic peptides, usually 5-30 amino acid residues, which can be found in a wide range of natural sources from microbes to plants and animals (Milletti 2012;Zorko and Langel 2022). Our findings corroborate a recent study of a peptide derived from the dengue virus capsid protein, PepR, which demonstrated rapid bactericidal activity against Staphylococcus aureus (Pinto et al. 2019). Furthermore, several authors deliberated that the viral protein-derived peptide (vAMP 059) and viral protein-derived CPPs (vCPP 0769 and vCPP 2319) are all bactericidal in nature and derived from viral capsids (Dias et al. 2017;Freire et al. 2015).
Intact cell membrane, metabolic activity, and reproducibility are the three accepted general parameters for the viability assessment of microorganisms (Oliver 2005). The conventional colony-forming unit (CFU) counting method takes into account only one of these parameters and reproducibility and provides only information about viable and culturable cells. The method is prone to errors due to the factors involved in bacterial growth on agar plates (Davis 2014). However, an alternative to culture-based detection is the assessment of cell viability using fluorescent dyes (LIVE/DEAD staining), including membrane potential and integrity that is being employed to determine bacterial viability (Berney et al. 2007;Stiefel et al. 2015). LIVE/DEAD staining permits detection of cell states other than live and dead, as those unable to grow, including live injured (where membrane integrity is compromised) cells that are unable to grow on agar plates (Léonard et al. 2016;Stiefel et al. 2015). Furthermore, it allows the loss of membrane integrity to be visualised over time (Nocker et al. 2017). Therefore, LIVE/ DEAD staining was performed in this study to look into the status of host cells after treatment with the ~30 kDa-purified putative EP of Bl 1821L. The staining displayed more Bl 1821L cells with compromised membranes (red) 6 to 12 h after treatment, and a pronounced effect was noted with an increase in the volume of purified putative EP. Interestingly, the appearance of red cells occurred at the same time as a decrease in the number of viable cells at 6 h after treatment with the ~30 kDa-purified putative EP as compared to the control. Typically, cells with compromised membranes (red) are considered to be dead or approaching death (Oliver 2010;Robertson et al. 2019). However, in this study, yellow/ orange cells were also seen after LIVE/DEAD staining of host cells in the overlay images which have been included the category of dead cells in the past (Leuko et al. 2004;Lu et al. 2014;Stiefel et al. 2015). The cause might be insufficient replacement of SYTO9 with PI and the retention of both the dyes within the cells at the same time (Kirchhoff and Cypionka 2017;Lu et al. 2014). A decrease in the colony count and a concurrent increase in the number of cells with compromised cell membranes (red) in LIVE/DEAD staining assay suggests that the CPPs residing in the ~30 kDa-purified putative EP may be involved in the formation of pores to lyse the susceptible cells (Sanderson 2005).
The Gram-positive bacterium, Bs, is an effective expression system for foreign proteins due to the structural differences in the outer cell membrane compared to the Gram-negative bacterium, E. coli (Su et al. 2020;Zweers et al. 2008). The gene lin encoding the Linocin 18 protein (bacteriocin) from the bacterium Br. linens M18 has been successfully expressed in E. coli (Valdes-Stauber and Scherer 1996). However, to enhance the stability of secreted proteins, an extracellular protease deficient mutant Bs WB800N (Jeong et al. 2018;Nguyen et al. 2011) was used in this study. The role of genes encoding the 25 kDa hypothetical and 31.4 kDa putative EP (Linocin M18) in the antagonistic activity was affirmed by expressing the corresponding genes individually (pHT01-hypo, 25 kDa), (pHT01-encap, 31.4 kDa) and together pHT01-hypo.encap in Bs WB800N. Notably, in the bioactivity assays, each expressed hypothetical (pHT01-hypo), encapsulating (pHT01-encap), and pHT01-hypo.encap encoding both the proteins expressed their antagonistic activity. BLASTp analysis of 25 kDa hypothetical protein shared 98.6% amino acid identity to a hypothetical protein of Brevibacterium sp. JNUCC-42. Interestingly, the CFS of heterologously expressed 25 kDa hypothetical protein (pHT01hypo) exhibited more potent activity as compared to 31.4 kDa putative EP (pHT01-encap), but SDS-PAGE analysis showed only purified bands of ≥50 kDa. The absence of the expected protein band might be due to the loss of expressed protein in the multiple purification steps (Walker 2010) or low expression of the proteins of interest (Lu et al. 2012). The putative EP (31.4 kDa) of Bl 1821L was successfully expressed in Bs WB800N at low levels. Bioactivity assay of Bs WB800N (pHT01-encap) supernatant containing 31.4 kDa protein exhibited antibacterial activity against Bl 1821L, and a prominent purified band of ~30 kDa was visualised on SDS-PAGE. However, activity of the 31.4 kDa protein (pHT01-encap, B1) was found against both strains Bl 1821L and Bl 1951, which might be due to the variation in the expression of pHT01-encap, B1 and pHT01-encap, and B2. Notably, the expression of both the encoded 25 kDa hypothetical and 31.4 kDa putative encapsulating proteins in Bs WB800N (pHT01-hypo.encap) displayed an inhibitory action against both the strains (Bl 1821L and Bl 1951), and purification of the corresponding supernatant yielded a protein band of ~30 kDa along with other proteins. All the constructs used in this study expressed the antibacterial activity in Bs WB800N, which is otherwise absent in this strain. Based on the literature, this is not only the first report of the involvement of family 1 encapsulin nanocompartment shell protein (Linocin M18) from the genus Brevibacillus in the putative antibacterial activity, but also among the entomopathogenic bacteria.
Overall, this research identified, purified, and characterised homologous genes encoding the 31.4 kDa family 1 encapsulin nanocompartment shell protein (Linocin M18) in Bl 1821L and Bl 1951 genomes. In addition, a gene encoding the more potent 25 kDa hypothetical protein was also identified at 3′ end of the identified putative EP. Antagonistic activity of the crude and purified lysates differed remarkably. The ~30 kDapurified putative EP of Bl 1951 and Bl 1821L was found to be active against both the isolates, but the autocidal activity was more pronounced with the crude lysates of both the strains. The autocidal activity of the identified 31.4 kDa putative EP (Linocin M18) may pose implications in harnessing the insecticidal potential of these isolates in future. Furthermore, based on bioinformatics analyses, this study proposed different killing mechanisms of 31.4 kDa putative EP of Bl 1821L and Bl 1951. These include the activation of stress relevant transcriptional regulator family proteins (PadR and MarR) or YtxJ protein under some stresses (such as malnutrition of iron), failure of ferritin protein to detoxify iron, and cell-penetrating peptide activity. However, these likely mechanisms require further investigation to validate their mechanistic role in the stunted growth of Bl 1821L and Bl 1951strains.