The Druggable Pocketome of Corynebacterium diphtheriae: A New Approach for in silico Putative Druggable Targets

Syed S. Hassan, Syed B. Jamal, Leandro G. Radusky, Sandeep Tiwari, Asad Ullah, Javed Ali, Behramand, Paulo V. S. D. de Carvalho, Rida Shams, Sabir Khan, Henrique C. P. Figueiredo, Debmalya Barh, Preetam Ghosh, Artur Silva, Jan Baumbach, Richard Röttger, Adrián G. Turjanski, Vasco A. C. Azevedo

Diphtheria is an acute and highly infectious disease, previously regarded as endemic in nature but vaccine-preventable, is caused byCorynebacterium diphtheriae(Cd). In this work, we used anin silicoapproach along the 13 complete genome sequences ofC. diphtheriaefollowed by a computational assessment of structural information of the binding sites to characterize the "pocketome druggability." To this end, we first computed the "modelome" (3D structures of a complete genome) of a randomly selected reference strain Cd NCTC13129; that had 13,763 open reading frames (ORFs) and resulted in 1,253 (∼9%) structure models. The amino acid sequences of these modeled structures were compared with the remaining 12 genomes and consequently, 438 conserved protein sequences were obtained. The RCSB-PDB database was consulted to check the template structures for these conserved proteins and as a result, 401 adequate 3D models were obtained. We subsequently predicted the protein pockets for the obtained set of models and kept only the conserved pockets that had highly druggable (HD) values (137 across all strains). Later, an off-target host homology analyses was performed considering the human proteome using NCBI database. Furthermore, the gene essentiality analysis was carried out that gave a final set of 10-conserved targets possessing highly druggable protein pockets. To check the target identification robustness of the pipeline used in this work, we crosschecked the final target list with another in-house target identification approach forC. diphtheriaethereby obtaining three common targets, these were; hisE-phosphoribosyl-ATP pyrophosphatase, glpX-fructose 1,6-bisphosphatase II, and rpsH-30S ribosomal protein S8. Our predicted results suggest that thein silicoapproach used could potentially aid in experimental polypharmacological target determination inC. diphtheriaeand other pathogens, thereby, might complement the existing and new drug-discovery pipelines.