Modelome derived intra-species broad spectrum drug and vaccine targets identification in the animal pathogen Corynebacterium pseudotuberculosis

Abstract

Corynebacterium pseudotuberculosis (Cp) is the etiological agent of several infectious and contagious chronic diseases, including caseous lymphadenitis (CLA), ulcerative lymphangitis, mastitis, and edematous skin disease thus affecting a broad spectrum of animal hosts, including ruminants that cause a huge decrease in wool production and carcass quality, thus threatening economic and dairy industries worldwide. In this work, first, manual annotation was performed using Artemis: the annotation tool, for two C. pseudotuberculosis strains. For Cp162, isolated from a camel in UK and completely sequenced using the SOLiD v3 Plus NGS platform, a genome of 2,293,464 bp in size, having 52.17% GC content, 2,002 coding sequences, 4 rRNA operons, 49 tRNA genes, and 87 pseudogenes, was obtained. For strain P54B96, isolated from an antelope in South Africa and sequenced using the Ion Torrent PGM NGS platform, a genome of 2,337,657 bp in size, having 52.2% GC content, 2,084 coding sequences, 4 rRNA operons, 49 tRNA genes, and 62 pseudogenes, was obtained. They were deposited in the NCBI GenBank database under accession numbers CP003652 and CP003385, respectively. Motivated by an increasing demand for new drug and/or vaccine targets, secondly, a novel strategy using a highthroughput workflow, called MHOLline, was followed to construct the modelome (3D models) of 15 C. pseudotuberculosis strains from various hosts and countries. Only Very High, High, Good and Medium to Good quality sequences were used from MHOLline classified groups(G2). Using a locally installed BALST NCBI tool, a set of 331 conserved proteins with 3D structures was selected showing 95-100% intra-species sequence similarity. The host proteomes considered in this study were human, horse, cow and sheep. Further filtering this core-modelome for essential and non-host homologous proteins, resulted in a final set of 10 proteins. Among these, only 4 proteins were identified as essential and non-host homologous and were considered as putative drug and vaccine targets, subjected to virtual screening and docking analyses. The druggability score, a drug target prioritization parameter, among others, was also calculated for all these proteins, allowing the prediction of druggable protein cavities. We further extrapolated our research to the other 6 essential host homologous proteins. A deep structure analysis at their residue level confirmed some conserved active site residues either exhibiting different conformations or even completely different residues. A multiple sequence alignment proved the residues conservation in these targets. Further, the role of these targets in different bacterial metabolic pathways, pathogenicity and virulence were also determined. It was proposed that the 6 essential host homologous proteins might also provide an extended choice for therapeutic targets. It is expected that our data will facilitate selection of C. pseudotuberculosis proteins for successful designing of new drugs and vaccines for a broad range of hosts.