Identication and characterization of a peptidoglycan hydrolase from Rhodobacter johrii

The bacterial whole genome sequences are available in the database therefore explored for the varieties of known and unknown proteins. Bacteria harbor various peptidoglycan hydrolases that cleave peptidoglycan and play an important role in the cell division, growth, spore differentiation and development. In the present study, we report a peptidoglycan hydrolase in an endospore producing phototrophic proteobacterium Rhodobacter johrii. The Peptidoglycan Hydrolase of Rba. johrii (PgHR) can actively hydrolyze the intact spore cortex peptidoglycan (sacculi). The protein contains a pre-peptide precursor which has a Hydrolase-2 (PF07486) family conserved domain. PgHR protein has SleB like properties which are spore cortex-lytic enzymes involved in the depolymerization of cortex peptidoglycan present and characterized in Bacillus spp. The expression pattern of PgHR through qRT-PCR suggests its role in stationary phase of Rba. johrii. This is a new type of peptidoglycan hydrolase reported from a proteobacterium.


Introduction
In bacteria there are diverse groups of peptidoglycan hydrolase enzymes which cleaves the bonds of the peptidoglycan polymers and its fragments (Shockman et al. 1996). These enzymes play an important role in cell motility, cell division and competence, cell-wall turnover, peptidoglycan remodeling, cell enlargement, differentiation into spores and germination, digestion of asymmetric septum for prespore engulfment, maturation of cortex of spore, lysis of mother cell and spore cortex lysis (Vollmer et al. 2008;Holtje 1995). These hydrolase cleaves the bond between the cell walls creating pores essential for incorporation of more PG material and expansion of cell wall during cell growth and development of the bacteria (Lee and Huang 2013).
Germination speci c lytic enzymes (GSLE) actively cleave and depolymerizes the peptidoglycan of spore cortex (Ishikawa et al. 1998), these types of enzyme are unable to cleave the intact cell wall of a new germinating bacterial cell which is present just beneath the cortex of spore (Popham et al. 1996). GSLE are two types (i) Spore cortex lytic enzymes (SCLEs) which initially hydrolyze the PG of the spore cortex (ii) Cortical fragment lytic enzymes (CFLEs) further degrade partially hydrolyzed cortex. The B. subtilis has three GSLE-CwlJ, SleB and YkvT of these SleB and CwlJ are well characterized. In B. cereus cortex lytic enzyme is SleL which shows glucosaminidase activity (Chen et al. 2000). There are still peptidoglycan hydrolase that harbor in the genome of bacteria which is yet to be explored, also the function of hydrolase depends on the type of hydrolase.
Rhodobacter johrii is a photosynthetic proteobacterium that was isolated from a pasteurized soil and produces refractile structures which are similar to spores since these structures were stained with malachite green (Girija et al. 2010). In present study, we identi ed a SleB like protein from Rba. johrii, and its comparative characterization involving bioinformatic analysis, enzyme assay and differential expression study was undertaken to understand the property of the new hydrolase.

Material And Methods
Bacterial strains, growth conditions and plasmids Rhodobacter johrii strain JA192 T (= DMS 18678 T = JCM 14543 T = MTCC 8172 T ) was grown in fully lled screw-capped tubes (10 × 100 mm) phototrophically (30°C, 2400 lx) in a mineral medium described previously (Lakshmi et al. 2011) with ammonium chloride (0.1%, w/v) and pyruvate (0.1%, w/v) as nitrogen and carbon sources/edonor, respectively. Bacillus subtilis strain FW2 grown in nutrient broth (Himedia M002) at 37°C overnight in shaking incubator was used for sporulation. The purity of cultures was checked at each step by streaking on nutrient agar (Himedia M001). Escherichia coli strains DH5α and Rosetta (DE3) were grown in LB medium (Himedia M1245) at 37°C and the antibiotic used was kanamycin (34 µg.ml -1 ). The plasmid used for this study was pET SUMO vector. All cultures were maintained in aseptic conditions with sterility precautions to ensure the absence of contamination.

Bioinformatic analysis
The whole genome sequence of Rba. johrii with Accession: PRJNA323784 was used for analysis from NCBI database. For full-length protein alignment the sequences used were B. cereus strains, ATCC 14579 (NCBI Accession No. WP001249053), M13 (ASK14870) and MLY1 (WP100910207); B. anthracis strains, ATCC 14578 (WP001249045), Sterne (YP028821) and Ames (AAP26584) using Clustal Omega (Madeira et al., 2019) and rendered sequence similarity by ESPript 3.0 (Robert and Gouet, 2014). The SWISS-MODEL server (Waterhouse et al. 2018) was used to generate a model of the PgHR protein. The model obtained from SWISS-MODEL was veri ed using PROCHECK (Laskowski et al. 1993). Protein domains were searched by a Simple Modular Architecture Research Tool (SMART) (Letunic and Bork 2018).
Protein models were superimposed based on Cα position using PyMol (The PyMOL Molecular Graphics System, Version 1.5.0.4 Schrödinger, LLC). The genetic features including promoter were predicted through the SoftBerry server (Salamov and Solovyevand 2011). The theoretical Mw (molecular weight) and pI were computed through ExPASy online tool (Gasteiger et al. 2005).

Cloning, protein overexpression, and puri cation
The pghr gene ORF lacking starting 23 codons, including codons 24-240 was ampli ed from Rba. johrii genomic DNA by using forward primer (BSCLFSUMOE: 5'-ATATGGATCC GATGTGACGGTGAGCCAGTC-3') and reverse primer (HSCLRSUMOE: 5'-TATAAAGCT TTCAGGAGCCGCCGGGCTG-3'). DNA was extracted using the necessary precautions to prevent contamination. PCR ampli ed product of 674 bp as well as pET SUMO expression vector was digested with BamHI and HindIII restriction enzymes, puri ed and ligated using kits (QIAquick PCR puri cation kit; Qiagen, Inc., USA). The pghr fused with pET SUMO plasmid produced the plasmid pET-SUMO-PGHR, which encodes an N-terminal His 6 -SUMO tagged with PGHR truncated fusion protein. This plasmid was transformed into E. coli DH5α cells and was con rmed by DNA sequencing.
The his 6 -SUMO-PGHR 24-240 fusion protein was overexpressed into E. coli Rosetta (DE3) which was grown in 1 Liter LB medium (Himedia M1245) with 34 µg.ml -1 kanamycin at 37 °C till the OD reached 0.6. At this point, 0.2 mM isopropyl-β-D-thiogalactopyranoside was added and culture was incubated at 37 °C for an additional 6 hours. Cells were harvested by centrifugation (6,500 rpm for 10 mins at 4°C) and resuspended into 25 ml ice-chilled buffer A (50 mM Tris base pH 8.0,300 mM NaCl, 10 mM Imidazole, 0.5 mM phenylmethylsulfonyl uoride [PMSF]). Cell disruption was done by sonication (35% amplitude, 10 min processing time, 20 s intervals). The cell lysate was pelleted by centrifuging (7,000 rpm for 30 min at 4°C) and the supernatant containing the soluble homogenized fusion protein with His-tag was puri ed by metal a nity chromatography with the Ni-NTA resins (Takara Bio.). Proteins were eluted at 100 to 350 mM imidazole gradient in buffer A. Fractions with eluted pure proteins with no other protein contaminants were pooled together, run on SDS-PAGE gels and the appropriate band was excised and sent for MALDI-TOF MS-MS sequencing, outsourced to Sandor Life Sciences, Hyderabad, India. The pooled fractions were dialyzed in buffer with SUMO protease without imidazole and passed again through Ni-NTA resins to yield pure PGHR 24-240 protein. This was concentrated by using an Amicon Ultra 15 centrifugal lter device (Merck Millipore).

Isolation of peptidoglycan substrates
Vegetative cell wall (peptidoglycan) from B. subtilis (Sigma Aldrich, Inc., USA) was procured for enzyme assay. Spore sacculi (peptidoglycan) from B. subtilis FW2 was isolated from the puri ed spore isolated as previously described (Bosak et al. 2008). Spore sacculi was isolated from the prepared spores as previously described (Dowd et al. 2008) with few modi cations. Brie y, spore pellet was harvested from 300 ml of B. subtilis FW2 was treated with 35 ml solution containing 50 mM Tris HCl (pH 7.5), 1% SDS, 50mM Dithiothreitol (DTT). This was boiled for 20 min and washed with warm autoclaved Milli-Q water till SDS was removed and centrifuged for 10,000 rpm for 10 min. The supernatant was discarded and the pellet was resuspended in 1 ml of a solution containing 100 mM tris HCl (pH 7.4), 20 mM MgSO 4 , 10 µg DNase I, 50 µg RNase A and incubated at 37˚C for 2 hr. To this 100 µg trypsin (TRT PCK, Worthington) in 10 mM, CaCl 2 was added and incubated at 37˚C overnight. The solution was boiled with 1% SDS and washed in autoclaved Milli-Q water till SDS was removed and dissolved in autoclaved Milli-Q water for analysis (or stored at 4˚C).

Enzyme activity assay
Spore sacculi was dissolved in 1 ml reaction mixture containing 30 mM NaPO 4 (pH 7), 1mM EDTA, 1mM DTT in a cuvette. The enzyme assay was done by observing the % OD 600 loss per minute (1 cm path length cuvette) of 0.2 OD 600 unit.ml -1 spore sacculi in which protein PgHR 24-240 from Rba. johrii was added at 20 nM concentrations and incubated at 25°C. The absorbance at 600 nm was recorded at each 5 min time interval after agitation for 40 min. Similarly, the cell wall was used as a substrate in the same buffer and the same conditions used above in which protein PgHR 24-240 was added at 20 nM concentrations and incubated at 25°C. The enzyme assay was done in triplicate and repeated at least three times.
RNA extraction, cDNA synthesis, and Quantitative RT-PCR for real-time expression Total RNA was extracted from 10 ml of culture samples at different time points using the miRNeasy mini kit (Qiagen Inc., USA) according to the manufacturer's instructions and treated with RNAse-free DNase I (Qiagen Inc., USA) to remove DNA contamination in RNA before reverse transcription. 1 µg of RNA was used to synthesize cDNA in a 10 µl reaction mixture by using Easy Script TM cDNA Synthesis Kit (Applied biological materials, Inc., Canada). The obtained rst-strand cDNA was ampli ed using SYBR green FAST qPCR Master Mix (Kappa Biosystems, Inc., USA) in a 10 µl reaction mixture with target RNA speci c primers on StepOne Real-time PCR detection system (Applied Biosystems). Primer design and speci city were done through Geneious Primer 3 for qRT-PCR (Table S1). Cycling conditions were as follows: an initial denaturation at 95°C for 10 min, followed by 42 cycles of 95°C for 15s, 62°C for the 30s, and 72°C for 30s. qRT-PCR reaction product speci city was tested by melting curve examination and each sample was tested in triplicates. The expression levels of the putative pghr gene were calculated using a relative standard curve method, Ct values of pghr gene were normalized to the Ct values obtained from the ampli cation of the housekeeping gene 16S rRNA gene and relative quanti cation was done by comparative ∆∆Ct method (2^(-∆∆Ct)).

Identi cation and sequence analysis
From the genome of Rba. johrii a hydrolase protein sequence (Accession. no. WP069331589) was retrieved. Protein has 28% amino acid identity, 67% coverage with B. subtilis, and 26% identity, 65% coverage with B. cereus spore cortex lytic enzyme (SleB). Hydrolase protein of Rba. johrii was further characterized and compared with Bacillus spore cortex lytic enzyme (SleB). The functional domain and characteristics of the protein were analyzed with the SMART search tool (Table S2), which predicted a signal peptide from 1 to 23 a.a and a conserved hydrolase 2 superfamily (pfam07486) domain from 131 to 234 a.a (Fig. 1A). The genetic features were identi ed, its promoter is having -10 box (CGCAAAAAT) from 121 bp to 113 bp and -35 box (TGGACG) from 141 bp to 136 bp upstream of the start codon (ATG) of coding sequence having total length of 723 bp of this hydrolase gene (Fig. 1B). The gene encodes a polypeptide of 240 amino acid residues with an expected molecular mass of ~25,557 Da, which is highly basic having a theoretical pI of 9.82.
We named this protein as Peptidoglycan Hydrolase from Rba. johrii abbreviated as PgHR. Multiple sequence alignment of full-length PgHR protein with the other SleB of Bacillus spp. was highly conserved at the catalytic domain with divergence at other parts (Fig. 2). Further, the sequence motif EXRGE having alternate glutamate (E) residues has been considered as a signature sequence of SleB proteins (Jing et al. 2012) which is strictly conserved in PgHR protein at position Glu128 and Glu132, simultaneously the catalytic residue glutamate (Glu128) was also conserved in PGHR of Rba. johrii (Fig. 2).
Secondary structure prediction and model superimposition PgHR protein of Rba. johrii secondary structure was model through homology modeling using the SWISS-MODEL server, from 109-234 a.a using a template of germination-speci c lytic transglycosylase SleB crystal structure (PDB ID -4f55.1) of Bacillus cereus ATCC 14579 (Li et al. 2012). The structure has 7 αhelices, 3 β-sheets, 11 loops and turns (Fig. 3A). PROCHECK tool was used to validate the model obtained from SWISS-MODEL with Ramachandran plot indicating >99% residues in the allowed region, only one residue amounting to ~1% in the disallowed region (Fig. 3B). PGHR model of Rba. johrii and template model was superimposed based on Cα position using PyMol. The root means square deviation (RMSD) for the superimposition of the model and the template was found to be 1.96 Å (Fig. S1).
The predicted model for PgHR of Rba. johrii and SleB of B. cereus were superimposed which indicated that catalytic residue Glu128 (glutamate) is conserved and present exactly in the same location of the groove between the α and β domains as previously reported in B. cereus (Li et al. 2012). The PgHR groove is lined with aromatic residues Phe167, Phe209, Trp218, His231, Phe233 which are conserved in all SleB type proteins (Fig. 4A) and these groove amino acids are responsible for binding and cleavage of spore cortex peptidoglycan (Li et al. 2012). In the superimposition of PgHR model of Rba. johrii with SleB of B. anthracis ATCC 14578 (Fig. 4B) it was found that the residues Tyr153, Gly164, are replaced by the same functional group amino acid and Trp218 is conserved, shown in red color. These residues are hypothesized for their role in the binding of muramic-δ-lactam (Jing et al. 2012). Whereas the catalytic, signature, aromatic residues Glu132, Arg147, Val148, Pro154, Phe167 were conserved in the exact location around the catalytic groove (Fig. 4B).

Substrate speci city and in vitro enzyme assay
The rst twenty-three amino acids at the N-terminal of PgHR 24-240 protein are predicted as signal sequence and were deleted. The C-terminal hydrolase domain containing 24 to 240 amino acids was overexpressed in the pET-SUMO-PgHR 24-240 vector along with histidine and Small Ubiquitin-like modi er tag (His 6 -SUMO) at N-terminal as a fusion protein in E. coli Rosetta (DE3). Overexpressed protein PGHR 24-240 with N-terminal His 6 -SUMO tag remained in a soluble form which was puri ed with Ni-NTA a nity followed by cation-exchange chromatography was con rmed by western blotting using antibodies against the His 6 tag. The sequence from MALDI-TOF MS-MS completely identical (Fig. S2) with the sequence of PgHR protein (Accession. no. WP069331589) of Rba. johrii and also checked on SDS gel (Fig. 5A). The tag was removed by DeSUMOylases cysteine protease treatment and the SDS-PAGE shows the expected molecular mass of PGHR (~25 KDa) after digestion of Rba. johrii (Fig. 5B). The digested protein was further puri ed with cation-exchange chromatography resulting in a pure protein free from His 6 -SUMO tag. Puri ed PgHR 24-240 protein activity of cleaving the spore cortex was analyzed by enzyme assay. When B. subtilis FW2 spore sacculi were incubated with recombinant PgHR 24-240 protein for 40 min, a loss of 50% of the substrate was observed at OD 600 (Fig. 6A). This con rms that puri ed PGHR shows activity on B. subtilis spore cortex peptidoglycan. When vegetative cell wall (peptidoglycan) from B. subtilis was used as a substrate for PGHR 24-240 protein, OD loss was not observed showing no enzyme activity (Fig. 6B).

Differential expression pattern analysis of PGHR in growth phases
The expression pattern of pghr was analysed at different growth phases (lag, mid-log, stationary, and decline) of Rba. johrii. The results showed no signi cant difference in the expression pattern until the stationary phase as compared to lag phase (Fig. 7). However, during the late stationary phase, the expression levels of pghr were up-regulated (~4.5-fold) and eventually down-regulated in the late decline phase. We do not nd any differentiation in the cells from lag to stationary phase and also the increased expression of pghr at late stationary ( Fig. 7). At the late decline phase, the pghr expression was decreased.

Discussion
This study provides evidence (Accession. no. WP069331589) of a Peptidoglycan Hydrolase in Rba. johrii that speci cally cleaves the spore cortex peptidoglycan (sacculi) but unable to cleave the cell wall peptidoglycan. This protein has a spore cortex lytic enzyme (SleB) speci c sequence motif EXRGE having alternating glutamate (E) residues (at positions 128 and 132) and catalytic residue (E128) present in PgHR of Rba. johrii. Further, PgHR has conserved hydrolase-2 superfamily (pfam07486) catalytic domain which is present in SleB type proteins (Jing et al. 2012).
In silico predicted secondary structure of PGHR was similar to the SleB of B. cereus with the conserve catalytic groove. Structural comparison of PGHR of Rba. johrii with SleB of Bacillus spp. showed the conserve catalytic glutamate residue and aromatic amino acids present within the catalytic groove responsible for binding and cleavage of spore cortex peptidoglycan (Fig. 4A) as reported in SleB proteins (Li et al. 2012;Jing et al. 2012). SleB enzymes cleave the spore cortex by speci cally identifying the muramic-δ-lactam (MAL) residue present in the cortex peptidoglycan of spore and it is unable to cleave the intact cell wall (Popham et al. 1996). Three amino acids which play an important role in identi cation and binding to muramic-δ-lactam (MAL) was found to be conserved in the Sle protein of Rba. johrii (Fig. 4B). Overall bioinformatics analysis suggested that PgHR protein has spore cortex lytic enzyme like properties. SleB protein has peptidoglycan (PG)-binding domain in the N-terminal region, the role of this domain is insigni cant, as the truncated protein without this domain did not affect the function of the protein (Heffron et al. 2011). Similarly, PgHR protein does not have peptidoglycan (PG)-binding domain but has the spore cortex lytic enzyme like properties.
The enzyme assay was also performed with the cell wall peptidoglycan as a substrate where SleR protein did not cleaved this substrate (Fig. 6B). This assay implicates the association of the enzyme PgHR in cleaving the spore cortex and not the cell wall in vitro. Real-time expression pro ling of pghr in Rba. johrii at different growth phases demonstrated that pghr expression was not signi cant between lag to stationary phase, while its levels spiked during the late stationary phase (Fig. 7). The expression pattern of SleR protein clearly indicates that this protein is not a prerequisite during early phases of development like in the peptidoglycan remodeling for cell division and growth (seen during the log phase). Further, it also indicates the role of PgHR during stationary phase which corroborate with sporogenesis. However, there is a need for further in-depth study to elucidate the excat function of PgHR protein in Rba. johrii.
Likewise, comparably in a gram-negative δ-proteobacteria Myxococcus xanthus forming spore-like fruiting bodies, a CbgA protein was identi ed as a homolog based on sequence similarity (35% identity and 57% similarity) to SpoVR protein of B. subtilis (Beall and Moran 1994). When CbgA protein was characterized in M. xanthus it was found to be a similar property protein (Tengra et al. 2006). Similarly, in this study we identi ed a new peptidoglycan hydrolase protein from Rba. johrii and further con rmed through in silico, in vitro, and expression pro le-based characterization.

Conclusion
The available whole genome sequences are exploited for the discovery of new uncharacterized proteins.
In this study, the bioinformatics analysis with current available tools, enzyme assay and expression study provides insights into a new peptidoglycan hydrolase from Rba. johrii that needs in depth functional study.