Immunoinformatics-based design of a multi-epitope vaccine for Marburg virus.

Marburg viruses (MARV) are negative-stranded RNA viruses belonging to filoviridae family. This virus causes severe haemorrhagic fever in humans and non-human primates, with a high fatality rate. Currently, there are no vaccines or drugs have been approved to induce productive immunity or control viral infection. As a result, vaccination against this virus is essential for reducing mortality rates. The current study aimed to identify CTL (cytotoxic T lymphocytes) and B-cell epitopes of MARV using in silico tools. A total of 3697 CTL epitopes and 4577 B-cell epitopes were predicted in the viral proteome of MARV. A reverse vaccinology approach was used to reduce the predicted CTL epitopes by adjusting MHC class I processing, immunogenicity, and other parameters. Finally, epitopes that are non-toxic, antigenic, non-allergenic, and non-homologous to the human proteome were chosen. Among these, 29 novel immunodominant MARV CTL epitopes were docked to their respective HLA alleles and the stability of the interaction was assessed using molecular dynamics simulation. All HLA-epitope complexes were found to be stable indicating that the predicted epitopes are binding with good affinity. Finally, a multi-epitope vaccine with three B-cell epitopes and ten CTL epitopes was designed. The secondary and tertiary structures of the vaccine construct were predicted, refined, and validated. The vaccine construct's codons were then optimized for maximum protein expression. In silico cloning was used to insert the gene construct into the pcDNA 3.1 (+) vector. This final construct can be used to develop an effective epitope-based MARV vaccine.