Molecular and bioinformatic strategies in reducing the burden of influenza infections in slum areas of Dhaka

Abstract

Influenza epidemics are caused by rapid evolution of the viral genotypes and play a significant role in the annual mortality and morbidity due to respiratory tract infections in tropical countries like Bangladesh. In low-income countries, influenza associated hospitalization is more prevalent in impoverished population such as slums. In Bangladesh, data on influenza infections in slums of densely populated Dhaka city is limited. Aims of this study were to detect circulatory influenza strains in slums, genetic characterization by molecular and serological methods, assessment of antiviral drug susceptibility of the virus and prediction of putative candidate epitope by in silico approach for alternate vaccine design. In this study course, 993 nasal and throat swabs were collected from patients presented influenza-like illness (ILI) from Rayerbazar, Hazaraibagh, Mohammadpur slums in Southeast Dhaka, Bangladesh between June 2012 and August 2013. Influenza viruses were detected using real time RT-PCR. One hundred one (10%) samples were identified as influenza positive, 47 influenza A (19 A/H1N1pdm09 and 26 A/H3N2) and 54 influenza B viruses including both B/Yamagata and B/Victoria lineages. Influenza positive samples were antigenically characterized by Hemagglutination Inhibition Assay (HAI) which revealed that Bangladeshi strains were antigenically similar to the WHO recommended vaccine strain for the Northern Hemisphere. Complete genome sequencing of two representative influenza A strains were conducted by sanger method and analyzed by BLAST, BioEdit, and Mega tools. Based on complete genome sequence, A/H1N1pdm and A/H3N2 strains were almost identical to other contemporary Bangladeshi strains as well as the globally circulating strains. In the phylogenetic analysis, HA gene of A/H1N1pdm and A/H3N2 were clustered in clade 6B and 3C.3, respectively. Compared to the vaccine strain, A/H1N1pdm showed at least three mutations, K163Q, S185T, and S203T, respectively in the HA antigenic sites (Sa, Sb, and Ca) while receptor binding sites remained conserved. Mutations were also observed at residue E374K of HA which is essential for membrane fusion. N-linked glycosylation sites were conserved in HA gene but one alteration at position 42 in NA gene was identified. The A/H3N2 showed two mutations at the antigenic site A with substitutions T144A and B, R158G in HA while receptor binding sites remained conserved. No strains had mutations that have been reported as responsible for enhanced virulence. Antiviral assay was performed in confluent monolayer of MDCK cells in 96-well plate in duplicate by MTT assay. Both strains were susceptible to the antiviral drugs routinely Abstract xvi | P a g e prescribed. Candidate epitope prediction was performed through in silico approach and putative peptide vaccine was determined. This study demonstrates several intriguing findings towards the understanding of genetic diversity of influenza viruses in Bangladesh and prediction of alternate vaccine approaches to reduce the burden of the disease. First, at least four types of influenza viruses were circulating during the study period which justifies that a quadrivalent influenza vaccine formulation that includes influenza A, influenza B, and both lineages of influenza B viruses could be more effective to reduce influenza disease burden in the country. Second, although Bangladeshi strains and WHO recommended vaccine strains are antigenically similar, several mutations were identified. It is interesting to see whether these mutations have any impact on vaccine efficacy. Third, all slum strains were found to be sensitive to the drugs routinely used for influenza treatment. The strains were analyzed based on known molecular markers for the drug resistance which could be useful for the clinical management of the patients particularly during pandemic situation. Fourth, prediction of HA and NA based candidate epitopes on the study strains could facilitate alternate vaccine approach. Notably, in Bangladesh the influenza vaccination has not been implemented in the national vaccination schedule and the vaccine effectiveness among Bangladeshi population is remained unknown. Therefore, assessment of current vaccine effectiveness as well as efforts to assess alternate vaccine approaches are required. In summary, the findings of this study contribute to understanding the characterization of slum influenza viruses that will be useful for routine surveillance, potential drug recommendation. Moreover, candidate epitope prediction will guide for alternate and improved vaccine for the control of influenza in future.