Investigação in silico de epítopos oriundos de linhagens de Mycobacterium avium subsp. hominissuis como candidatos vacinais

Abstract

Non-tuberculous mycobacteria are environmental mycobacteria responsible for a growing number of systemic and respiratory infections affecting mostly children, elders and immunocompromised individuals. The Mycobacterium avium Complex comprises Mycobacterium avium as well as M. intracellulare and the major responsible for the reported cases to this day. M. avium has been recently classified as containing four subspecies with different infectivities as well as different hosts. One of those subspecies, Mycobacterium avium subsp. hominissuis has been isolated from humans and swines, whereas other varieties are found in cattle, birds and wild animals. To this moment, MAC infections are controlled with the use of multiple antibiotics through long, expensive and sometimes inefficient treatment regimens. The identification of effective targets for controlling such organisms is an essential and challenging task as surface proteins, which are key target molecules in several successful immunotherapies, are difficult to isolate. In addition, the design of immunotherapies and vaccine formulations depends on the identification of peptides of immunological interest which are usually found through repetitive and expensive experimental protocols. In this study applied computational tools to investigate surface proteins with exposed and ubiquitous immunogenic portions to strains of Mycobacterium avium subsp. hominissuis. To achieve that, 32648 amino acid sequences obtained from the NCBI database for Mycobacterium avium subsp. hominissuis were submitted to TMHMM for detection of alpha-helix transmembane domain, which were present in 3426 of those sequences. These proteins were clustered in 577 groups by CMG Biotools according to their homology as to identify membrane proteins common to all the organisms of interest. Those sequences were then submitted to available methods obtained at IEDB to classify their affinity to a list of 27 MHC alleles frequent in human populations. Peptides with the highest predicted immunogenicities were selected and 112 clusters with core proteins and high MHC affinities were selected. Crossing information between IEDB and TMHMM allowed for the selection of the 58 clusters in which at least one peptide was predicted to be placed on the outer portion of membrane. We also calculated peptide A. conservation (their presence in different strains), where 60% of clusters are formed by ubiquous peptides and B. promiscuity (the number of distinct MHCs to which they bind), where only a single cluster has a peptide that binds to four distinct MHCs with high affinities. As for vaccine epitope candidates, a minimum set with nine peptides of high binding affinity to the highest possible number of distinct MHCs were selected, interacting with 15 molecules. None of those nine sequences showed potential to cross-react with human or swine proteins. The protocol executed for this work can be applied to other organisms as means to identify possible vaccine application candidates.