Dengue virus (DENV) belongs to the Flavivirus genus. Transmitted to its human host with the bite of an infected Aedes aegypti mosquito, DENV has re-emerged as the most globally significant arthropod-borne viral pathogen in annual incidence and morbidity. Antiviral and vaccine development is therefore a major priority for the global medical research community. DENV presents as four distinct serotypes (DENV-1 to DENV-4). Reinfection with a heterologous serotype increases the risk of developing severe disease, hindering effective vaccine development. Despite many years of research into the molecular mechanisms that this important pathogen employs in infecting the host cell, many questions remain unanswered.
In this work we have deployed a range of molecular biological techniques to better understand two key aspects in the viral lifecycle: host cell receptor binding and the fusion of the viral and cellular membrane. The DENV envelope (E) glycoprotein is the major surface protein of the infectious virion. The protein mediates the two critical functions required for virus infectivity, host cell binding and fusion of the viral and cellular membranes.
The E ectodomain consists of three domains, DI, DII and DIII. The surface loops of DIII project furthest from the virion surface, suggesting a role for DIII in host cell receptor binding. We anticipated that DIII would perform well as a candidate for prokaryotic expression, which would allow in vitro analysis of DIII functions independent of the whole virus. We complemented our strategy with cell based infection assays to permit extrapolation of our DIII findings to the whole virus. Accordingly, we designed and expressed a recombinant form of DIII as an MBP fusion protein.
To assess the biological function of our recombinants, we developed a novel medium throughput on-cell far-western based binding assay. We demonstrated the cell binding capacity of our recombinants, and went on to clone DIII from all four DENV serotypes. Using our novel binding assay in a competition based format we revealed significant differences in binding of the four DENV serotypes on mosquito cells. Using a specific antibody and site directed mutagenesis we demonstrated that this specificity is imparted by the FG loop within DIII.
To gain further insight into the receptor binding function of DIII, we undertook a site directed mutagenesis approach of surface lysine residues to alanine within our MBPii DIII(DENV-2) construct. In the early stages of cell binding, DENV interacts with cellsurface negatively-charged polymers called glycosaminoglycans (GAGs). We established a novel GAG based binding ELISA, and identified two key residues, K291 and K295, which demonstrated significantly less binding affinity. We confirmed the importance of these residues using our cell binding assay on human cells. Interestingly, none of the mutations we introduced affected binding to mosquito cells. This result was mirrored when we assessed the ability of our recombinants to inhibit virus infection of human and mosquito cells. Finally, we mutated K291 and K295 to arginine and were able to show a wild-type like phenotype in all three assays, indicating that the GAG interaction was electrostatic in nature.
In addition to its host cell binding function, DIII also plays a key role in the in the large inter-domain reconfigurations involved in the dimer-trimer transition that drives the fusion process. We hypothesised that mutations to conserved residues within DIII that mediate the inter-domain contacts could prevent the binding affinity of exogenous DIII to E and therefore reduce the anti-viral activity of our mutated DIII constructs. Using alanine site directed mutagenesis we examined two conserved residues, Q316 and H317 and showed that in an antiviral assay these recombinants demonstrated significantly reduced activity compared to WT.
To directly assess the anti-fusion activity of our recombinants, we developed a novel live cell imaging assay, supported by computer aided image analysis software. Using this assay we demonstrated the anti-fusion properties of our recombinants, but were unable to distinguish significant differences between the wild type MBP-DIII and our mutated constructs. Further optimization of this experimental approach in a future study could tease out the role of single amino acids in this complex macromolecular process.
This thesis provides new functional insight into the dual role of DIII in cell surface receptor interactions and membrane fusion. We analysed these two roles by introducing site directed mutations to regions thought to be specific to both receptor binding and the fusion process. We developed three novel assays, a GAG-ELISA, cell binding assay and live cell microscopy based DENV fusion assay, and deployed these new systems to investigate the functional differences between our recombinants due to introduced mutations. This work therefore provides both experimental groundwork and functional data that can be used as to further understand the molecular nature of DENV entry.