Communication of γ Phage Lysin plyG Enzymes Binding toward SrtA for Inhibition of Bacillus Anthracis: Protein–Protein Interaction and Molecular Dynamics Study

Abstract

Bacillus anthracis is a pathogenic, Gram-positive bacterium which chiefly affects the livestock of animals and humans through acute disease anthrax. All around the globe this bio-threat organism damages millions of lives in every year and also most of the drugs were not responding properly in inhibition against this diseased pathogen. In recent development, phage therapy is considered as alternative solution to treat this serious infectious disease. In this study, we elucidated the binding of γ phage lysin plyG enzymes toward the SrtA along with its activator peptide LPXTG. Through protein–protein docking and molecular dynamics simulation studies, we showed the distinguished structure complementarity of SrtA and plyG complex. Especially, MD simulation relates strong and stable interaction occurs between the protein complex structures. These results suggest that additional experimental studies on our approach will lead to availability of better inhibitor against the SrtA.

Keywords::

• Bacillus
• SrtA
• plyG
• LPXTG
• phage
• molecular dynamics

INTRODUCTION

Bacillus anthracis a gram-positive, spore-forming bacterium is the causative agent of anthrax belongs to the Bacillus cereus group which also comprises B. cereus and Bacillus thuringiensis (Rasko et al., 2004, Chung et al., 2006). B. anthracis has long been considered a potential biological warfare agent because they are invisible, silent, odorless, tasteless, easy to disperse, and inexpensive to produce (Hawley et al., 2001). There is an increasing concern within both the scientific and security communities that the ongoing revolution in biology has great potential to be misused in offensive biological weapons programs (Atlas & Dando, 2006). In addition to the bio-threat organisms, the scale of antibiotic resistance now results in annually killing more than 10 million people around the world by the infections not responding to current antibiotics (Brasier, 2008). Most of the drugs are showing deficiency in inhibition against the diseased pathogen and recent studies in phage therapy is considered as another best solution to treat pathogenic bacterial infections (Levin & Bull, 1996).

Bacteriophages (BP) are viruses that attack bacteria, multiply within and cause disruption of bacterial cells (Dabrowska et al., 2001). Phage proteins (lysins) inhibit microbial pathogens via the damage to the surface proteins in the bacterial cell wall (Low et al., 2011). Phage virus must enter inside the host bacterium for replication and also it needs to exit the bacterium to disseminate its progeny phage (Sulakvedlidze et al., 2001). For this mechanism, double-stranded DNA phages have evolved a lytic system to weaken the bacterial cell wall resulting in bacterial lysis (McDonnell & Russell, 1999). Phage lytic enzymes, or lysins, are highly efficient molecules that have been refined over millions of years for this exact purpose (Hermoso et al., 2007). These lysins target the integrity of the cell wall and are designed to attack one of the four major bonds in the peptidoglycan (Fischetti, 2005). Lysins have now been used successfully in animal models to control pathogenic antibiotic-resistant bacteria found on mucosal surfaces and in blood (Fischetti, 2010).

In the year of 2002, Schuch et al., have reported “A bacteriolytic agent that detects and kills Bacillus anthracis” and exploited the inherent binding specificity and lytic action of bacteriophage enzymes called lysins for the rapid detection and killing of B. anthracis, which proved that γ phage-derived plyG enzymes are having the ability to target the B. anthracis and kill it (Parisien et al., 2008). Most B. anthracis strains are sensitive to γ phage, but most B. cereus and B. thuringiensis strains are resistant to the lytic action of γ phage (Schuch et al., 2002). In the year of 2005 Davison et al., have reported the “Identification of the Bacillus anthracis γ Phage Receptor” and from that plyG enzyme binding with surface protein namely SrtA is confirmed (Davison et al., 2005). Previously, the mechanism of plyG inhibiting SrtA is validated by using protein-protein interactions and additional screening of the peptide competitors from γ phage lysin are done, that is screened peptides are having the ability as LPXTG (L = Leucine, P = Proline, X = represent any amino acid, T = Threonine, G = Glycine) competitive inhibitors. In comparison to the activator peptide LPXTG binding motif, γ phage lysin based inhibitor peptides are analyzed towards the binding pocket of SrtA enzymes. Therefore in this study we investigated activator peptide LPXTG interaction with SrtA and complementarity between SrtA and plyG structure.

MATERIALS AND METHODS

System configuration

All the research works were executed on a High Performance Cluster (HPC) operated with Cent OS V5.5 Linux operating platform. Hardware specifications of HPC cluster-Super micro SC826TQ-R1200 LIB series, running with 2 atom processor of 32 Core and 32 GB RAM speed. Software specifications included in screening and MD simulations are commercial version of Schrodinger software package, LLC, New York, NY 2012(Maestro 2012), and academic version of Desmond molecular dynamics package.

Molecular modeling environmental setup

A typical structure file from the PDB was not suitable for immediate use in molecular modeling calculations, so that the crystal structure of SrtA from B. anthracis (PDB ID: 2KW8) and crystal structure of plyG (PDB ID: 2L47) was prepared through Protein Preparation Wizard tool implemented in Maestro 9.2 (Selvaraj et al., 2014). While the missing residues were added using the prime loop modeling which is embedded in Protein Preparation Wizard and the missing loop is filled from the SEQRES records in the PDB file using Prime (Reddy et al., 2013). In order to optimize the protein structures the bond orders were assigned, hydrogen atoms were added, and all the crystallographic waters molecules were removed (Selvaraj & Singh, 2013). The protassign script optimize the hydrogen-bonding network, rotating hydroxyl and thiol hydrogens, which generates appropriate protonation and tautomerization state of HIS, and it also performing chi flips on the ASN, GLN, and HIS residues. Optimized structure of SrtA was minimized until the average root mean square deviation (RMSD) of the non-hydrogen atoms reached 0.3 Å (Sivakamavalli et al., 2014). The activator peptide is constructed using chimera and optimized through Lig prep 2.8 using the OPLS-2005 force field (Selvaraj et al., 2014).

Protein–peptide docking interactions

Protein–peptide docking calculation was performed through Induced Fit Docking (IFD). Here grid was generated based on binding site information of SrtA structure and LPXTG signaling peptide was docked (Chan et al., 2013). However the IFD protocol used in this study was performed in a three consecutive steps (Wang et al., 2008). First, the peptide was docked into a rigid receptor model with scaled-down van der Waals (vdW) radii in a range of 0.5 for both the protein and peptide nonpolar atoms (Selvaraj et al., 2014). A constrained energy minimization was carried out on the SrtA structure, keeping it close to the original crystal structure while removing bad steric contacts (Cherfils et al., 1991). Energy minimization was carried out using the OPLS-2005 force field with implicit solvation model until default criteria were met (Tang et al., 2010). The standard precision mode was used for the initial docking and 20 peptide poses were retained for protein structural refinements (Friesner et al., 2004). In the second step, Prime was used to generate the induced-fit protein–ligand complexes (combination of protein modeling simulation). Each docked conformers in previous step was subjected to side-chain and backbone refinements (Wang et al., 2009). The refined complexes were ranked by Prime energy, and the receptor structures within − 30 kcal/mol of the minimum energy structure were passed for a final round of Glide docking and scoring. In the final step, the peptide is redocked into every refined low-energy receptor structure produced in the second step using Glide XP at default settings. The side-chain orientations have been performed automatically with inclusion of prime in IFD (Selvaraj et al., 2012). An IFD score that accounts for both the protein–ligand interaction energy and the total energy of the system was calculated and used to rank the IFD poses (Sherman et al., 2006).

Protein–protein docking of SrtA and plyG

A geometry-based molecular docking algorithm called Patch Dock (http://bioinfo3d.cs.tau.ac.il/PatchDock) was used to dock the three-dimensional structures of SrtA and plyG. The Patch Dock server predicts the docked transformations that produce good molecular shape complementarity. The algorithm divides the Connolly dot surface representation of the molecules into concave, convex, and flat patches (Cosgrove et al., 2000). The patches are matched according to their complementarities and in order to generate different transformations (Schneidman-Duhovny et al., 2005). A default value of 4Å was used for clustering and redundant solutions were discarded by RMSD clustering (Subramaniam et al., 2009). The Patch Dock output generates the geometric score, desolvation energy, interface area size, and the actual rigid transformation of the solutions (Schneidman-Duhovny et al., 2004). Twenty solutions, out of about 60 predicted SrtA–plyG complexes, were sorted according to their geometric shape. The complementarity scores were analyzed for identifying the residues involved in the protein–protein interface and the interactions are shown by PDB sum–chain interactions.

Molecular dynamics simulations of protein–protein interactions

Molecular dynamics simulations were carried out using Desmond, which is known to perform high-speed MD simulations of biological systems on conventional commodity clusters. Here the purpose of the simulation is to obtain the stable conformation of plyG and SrtA structures, binding influence of LPXTG motif with SrtA and also to understand the plyG and SrtA interactions in dynamic state (Klepies et al., 2009). The Desmond program package adopting the OPLS-AA force field parameters were used for EM and MD simulations (Shivakumar et al., 2010). For the MD simulation studies, the structures were solvated using the TIP3P water model and the solvated structures were energy minimized using the steepest descent method, terminating when maximum force is found smaller than 100 kJ/mol^{− 1}/nm^{− 1} (Shafreen et al., 2013). The apo protein, protein–peptide complex, and protein–protein complex are subject to energy minimized using Desmond with OPLS-AA force field parameters (Liu et al., 2013). All systems were subjected to up to 100 steepest descent energy minimization steps before thermalization, for a maximum force of more than 100 KJ/mol^{− 1}/nm^{− 1} (Soares & Caliri, 2013). After thermalization, MD simulation was run at the isothermal-isobaric (NPT) ensemble at a constant temperature (300 K) and pressure (1 bar) with a time step of 2 fs and the relaxation time was applied between 0.1 and 0.4 fs. NVT simulation was carried out for 1ns and the simulated conformers were equilibrated for 10 ns of the time scale (Mori & Okumura, 2013).

RESULT AND DISCUSSION

Structural details of SrtA and plyG

The architecture of SrtA structure is more complicated and it posses the flexible binding region due to the presence of loop structures in the binding region. SrtA unique flexible active site loop, mediates the recognition of lipid-II, the second substrate to which proteins are attached during the anchoring reaction (Weiner et al., 2010). SrtA from B. anthracis adopts an eight-stranded beta barrel structure along with four helices structural elements (Figure 1a and b). The secondary structural elements in the protein have the following topology: H1-β1-β2-H2-β3-β4-H3-β5-β6-H4-β7β8, where H and β refer to helices and strands, respectively. Beginning at the N-terminus, an extended segment (Ala58-Gln63) is positioned in the active site cleft and is then followed by a α-helix (H1, Leu66-Asn71). The barrel structure then begins with strand β1 (Gly81-Ile85), which is connected by a short hairpin that lies anti-parallel to strand β2 (Leu90-Leu95). A 3₁₀ –helix (H2-Glu100-Ser105) then joins strand β2 to β3 (Ala107-Thr109), which lay in parallel. Helix H2 also forms a wall of the active site cleft. In reverse direction of the chain position, strand β4 (Asn120-His126) in an anti-parallel orientation next to strand β3. An extended polypeptide segment containing a 3₁₀-helix (H3, Ile138, and Ser140) wraps around the enzyme to initiate strand β5 (Lys146-Asp151) on the opposite face of the protein. This strand pairs with β1 in an anti-parallel fashion and is separated by a short hairpin from residues in strand β6 (Asn154-Glu165), whose chain also is positioned in an opposite orientation. A long loop containing a 3₁₀-helix (H4, Trp171-Val173) then connects strands β6 and β7 (Glu181-Thr186). The β7 strand runs parallel with respect to the β4 strand and is followed by a structurally disordered loop that reverses the direction of the chain thereby enabling residues in strand β8 (Typ197-Ala208) to hydrogen-bond with residues in strands β6 and β7 in an anti-parallel manner.

Figure 1. (a) 3D Structural representations of SrtA from Bacillus anthracis. (b) Secondary structural alignment of SrtA from B. anthracis.

Two adjacent β-bulges present within strands β6 (Thr109) and β8 (Val204) introduce a kink that enables extensive interactions between the chains and allows them to form opposing faces of the β-barrel structure. Residues His126, Cys187, and Arg196 are completely conserved in sortase enzymes and form the active site. They are located near the end of the sheet formed by strands β4, β7, and β8. Cys187 is situated at the C-terminal end of β7 strand and is bracketed by the side chains of His126 and Arg196 located on strands β4 and β8, respectively.

Recently, the lysin of γ-phage, PlyG, was identified and reported to specifically lyse B. anthracis (Kikkawa et al., 2007). The C-terminal region of PlyG (amino acid residues 156–233) is sufficient for specific binding to B. anthracis and that Leu190 and Glu199 play critical roles in binding. However, little is known about the catalytic mechanism of action of PlyG. Lysins are generally endo-b-N-acetylglucosaminidases that act on sugar moieties, N-acetylmuramidases (lysozymes) that cleave peptide cross-bridges, or N-acetylmuramoyl L-alanine amidases that hydrolyze amide bonds connecting sugars and peptide constituents (Schmitz et al., 2010). PlyG is predicted to be an N-acetylmuramoyl L-alanine amidase. N-acetylmuramoyl L-alanine amidases typically consist of an N-terminal domain possessing catalytic activity and a C-terminal domain with substrate-binding activity. Their catalytic domains are highly conserved among various amidases, and the catalytic site of PlyG resembles that of T7 lysozyme (Cheng et al., 1994). Cheng et al. reported that the three amino acid residues of T7 lysozyme, His17, Tyr46, and Lys128, are critical for its catalytic activity. The secondary structural elements in the protein have the following topology: α1-H1-α2-H2- α3-α4-H3-α5-H4- H5-α6-H6-H7-H8-α7-H9-α8-α9-H10-α10-, where H, β, and α refer to helices, strands, and sheets, respectively (Figure 2a and b).

Figure 2. (a) 3D Structural representations of plyG from B. anthracis infecting γ phages. (b) Secondary structures of plyG from B. anthracis infecting γ phages.

Protein–peptide interactions studies

Docking with LPXTG was carried out on the binding site of SrtA using IFD; filtering poses based on glide score. The residues Leu66, Val69, Ala70, Asn71, Ala72, Leu74,Lys76, Arg 82, Glu91, Leu92, Pro 93, Ser105, Lys131, Gly132, and Val133 were taken for interaction studies with LPXTG peptide based on active site prediction through sitemap. The activator peptide LPXTG, binds with Ala72, Lys76, Arg82, Glu91, and Ser105 which is considered as LPXTG binding site. The G-score values was used for analysis of binding score for a given protein–peptide complex structure. The docking for the docked complex is showing XP score of − 6.851 kcal/mol, IFD score of − 317.18 and docking energy shows –75.37 kcal/mol from the glide docking. Schrodinger 2D interaction plot showed hydrogen bonding residues (Ala72, Lys76, Arg82, Glu91, and Ser105) with the LPXTG peptide in the binding site of SrtA protein (Figure 3). The peptide binds with barrel structure that begins with strand β1 and connected by a short hairpin that lies anti-parallel to strand β2. The docked poses were minimized using the local optimization feature in Prime, and the energies of complex were calculated using the OPLS-AA force field and GBSA continuum solvent model. Both the docking and binding energies are differing from each other and these energies are showing much support to the docking results.

Figure 3. Interactions of LPXTG peptide with SrtA peptide bound region represented in 2D interactions.

Protein–protein interactions analysis

The structures of SrtA and plyG after optimization were docked to understand the mode of interaction between these two proteins using Patch dock program (http://bioinfo3d.cs.tau.ac.il/PatchDock). The complexes were further optimized by another docking program Fire Dock (http://bioinfo3d.cs.tau.ac.il/FireDock) (Feher & Williams, 2009). The surface representation of SrtA and plyG complex showed excellent shape complementarily (Figure 4a and b).

Figure 4. (a) Interaction of SrtA (Meshed surface) with plyG (Cartoon-ribbon). (b) Interaction of SrtA (Meshed surface-pink) with plyG (Meshed surface-red).

The scores representing, predicted binding free energy for the top-ranked solutions, number of structurally aligned residues appearing within the distance cut-off of 2.25–3.6 Å, and the number of hotspot residues in the protein–protein interface were analyzed. Numbers of hydrogen bonds, hydrophobic, and non-bonded contacts were calculated using PDBSUM program for each of SrtA and plyG protein complex. Initially the SrtA+ LPXTG complex binds with the plyG near the region of LPXTG binding location. This region has fissure binding pocket, which is adoptable for the macromolecular structure plyG to bind with it.

The top protein complex solution was obtained through geometry-based molecular docking algorithm and their output values are represented in Table 1. Whereas the lowest-energy refined protein complex structure was generated using Fire dock and their respective values of global energy, Vdw, atomic contact energy (ACE) and the contribution of hydrogen bonds to the global energy is given in Table 2.

Table 1. Patch Dock output values of top protein complex solution.

Solution No	Geometric score	Interface area size	Atomic contact energy (ACE)	Transformation
1	13770	1853.90	453.38	−2.99 −0.96 −2.47 193.93 −143.58 23.52

Table 2. Fire Dock output values for top refined complex structure.

Solution No	Global energy	Attractive vdW	Replusive vdW	Atomic contact energy (ACE)	Hydrogen bond
1	− 87.13	− 41.96	7.52	− 14.04	− 5.71

Molecular dynamics simulations

Computational analysis of protein–protein interactions is having much correlation with X-ray crystallographic studies, though more protein–protein complex is found in protein data bank. Proteomics approach through molecular dynamics simulations are discussed here, potential functional and physical interactions are predicted using in silico methods. Explicit solvent MD simulations of SrtA+ LPXTG, plyG, and SrtA+ LPXTG+ plyG complex protein showed good stability in the simulation point. Interaction of SrtA with plyG before simulation shows initial interactions were formed through hydrogen bond interactions and non-bonded interactions. The protein complex of plyG and SrtA shows strong interactions from the initial to final level of 10ns of simulations. The potential energy of all the dynamics shows stable state and it informs that the simulation event with these three proteins and protein–protein complex are well equilibrated. The stability of the protein was obtained using the Desmond-RMSD analysis. RMSD of SrtA lies between 2.3Å and 2.9Å and RMSD of plyG lies between 2.1Å and 3.8Å for 10 nanoseconds of MDS run (Figure 5). When comparing both apo proteins of SrtA and plyG, the SrtA–plyG complex shows more variations from the protein structures and bind efficiently with each other but the architecture of the protein remained relatively similar after simulation.

Figure 5. RMSD graph and Mean variation plot represents the SrtA and plyG in MD simulation event of 10ns timescale.

The final stable protein structures after simulation event are represented in ribbon models in (Figures 1a and 2a). SrtA and plyG complex best in terms of global energy and ACE of the complex had 7 hydrogen bonds and 112 non-bonded contacts (Figure 6). The change in initial and final interactions between the PPI complex is due to the negative region of donor protein interacts with positive region of acceptor protein. Even though docking algorithm fails to allocate the interactions between the SrtA and plyG complexes, so we had concluded the possible interactions through molecular dynamics simulation approach. The hydrogen bonds and non-bonded contacts are shown in cartoon representation (Figures 7 and 8) and here the Chain A represents the SrtA and the Chain B represents the plyG. When comparing the SrtA, the plyG is less stable because the size of the protein is high. The loop structures present in the SrtA are fluctuating high and in complex structures, those loops are helpful in binding with plyG. Final complex structure shows that, plyG structure overall covers 60% of the SrtA within it and their surface representations are shown in the Figure 4A and B. The region of SrtA+ LPXTG motif bound complex shows high affinity binding toward the plyG, due to the activation role played by the LPXTG binding motif. The initial level of the SrtA and plyG complexes lacks in binding efficiency due to donor/acceptor manifestations and dynamic pose of plyG and SrtA complex shows high-affinity plyG toward binding with SrtA.

Figure 6. Bonded interactions between SrtA and plyG protein–protein complex in the MD simulation event of 10ns timescale.

Figure 7. Cartoon representations of bonded interactions between SrtA and plyG protein–protein complex in the MD simulation event of 10ns timescale in each ns difference.

Figure 8. Interaction of SrtA with plyG before and after simulation shows initial interactions were formed only through bonding interactions in simulation of 10ns protein Chain A represents the SrtA and Chain B represents the plyG.

CONCLUSION

In this study we described the unavailability of potential drugs toward the inhibition of B. anthracis SrtA and application of using phage-based proteins against them. Hence we explored the interactions of activator peptide LPXTG with SrtA and additionally plyG enzyme. The Patch molecular docking algorithm produces definite molecular shape complementarity between SrtA and plyG structure. While molecular dynamics simulation analysis informs the strong bonding interaction occurs between the plyG and SrtA complex structures throughout the timescale. Further work must therefore concentrate on the phage protein to obtain valid peptidomimetic drug for SrtA inhibition.

Acknowledgment

Chandrabose Selvaraj gratefully acknowledges CSIR for the Senior Research Fellowship (SRF). R. Bharathipriya Sincerely thank the facillities offered by the UGC Innovative program in Department of Bioinformatics, Alagappa University.

Declaration of interest: The author report no declaration of interest. The author alone is responsible for the content and writing of the paper.

This Research work was partially supported by a Council of Scientific & Industrial Research, India (CSIR sanction no: 37(1491)/11/EMR-II), and Dr. Sanjeev Kumar Singh acknowledging CSIR for financial support.