Date of Award
Doctor of Philosophy (PhD)
Richard J. Kuhn
Committee Member 1
Committee Member 2
Committee Member 3
Flaviviruses are positive single-stranded RNA arthropod-borne viruses within the family Flaviviridae which includes Dengue virus (DENV), West Nile virus (WNV), Yellow fever virus (YFV), Japanese encephalitis virus (JEV) and Zika virus (ZIKV). From DENV alone approximately half of the world lives under the threat of contracting a flavivirus infection, with about 390 million infections globally per year. As arthropod borne viruses, these are easily transmissible, and spread is often di cult to contain. The non-structural protein 1 (NS1) has been shown to be integral in flavivirus pathogenesis. With recently defined roles in increasing infection of mosquitos as well as contributing to vascular leakage, a key phenotype of severe Dengue disease, studies in understanding how this protein controls its various functions are urgently needed.
Chapter 2 of this thesis involves a structure guided mutagenesis approach to determine if DENV and WNV NS1 proteins have similar functions in their corresponding virus lifecycles and if there are central regions of these proteins which are responsible for specific functions. Using a structure and sequence alignment residues were chosen for mutagenesis on the basis of being conserved across all 4 DENV serotypes, but di↵ering from WNV and being surface exposed, to avoid disrupting the protein structure. From a total of 37 single residue changes 3 amino acid substitutions resulted in interesting phenotypes. A P101K and an L270E change in DENV resulted in a blockage of NS1 or virus particle secretion. After further characterization the P101K residue change resulted in a block of both NS1 protein and virus particles to be released, but when transiently expressed was found to be secreted to near wild type (WT) levels. The opposite was observed for L270E in which NS1 was not secreted in either full length virus or transiently expressing systems. Interestingly L270E, however, was able to form intracellular particles, but these were trapped inside cells. This suggested that residue changes to 101 results in a block of secretion potentially due to either an enhanced or reduced interaction with the structural proteins, while L270E could be disrupting host interactions required for secretion. The third residue substituted was at position 343 in NS1 and resulted in opposing changes to DENV or WNV specific infectivity. An E343K change resulted in a decreased infectivity, while in WNV resulted in an increased infectivity, demonstrating an importance of this residue in both DENV and WNV infection. Taken together, our data suggests that there are distinct sites on NS1 that contribute to protein secretion as well as virus infectivity, and that these residues appear to have a conserved function across both Dengue and West Nile viruses.
In Chapter 3 we explored the role of NS1 in regulating a flavivirus infection. Evidence from the previously discovered E343K (DENV) and K343E (WNV) mutant, led us to hypothesize that NS1 can contribute to a flavivirus infection. To this end purified virus and purified NS1 were obtained and used to examine if addition of NS1 to purified virus enhanced virus infectivity. We demonstrated that at several concentrations NS1 increases a flavivirus infection in both a homotypic and heterotypic manner. Furthermore, this increase in infectivity was occurring from an increase in virus attachment. We found that NS1 is able to increase the amount of virus bound to cells, through acting similar to an attachment factor on host cells. DENV attachment is poor with only approximately 0.25 to 1% of virus attaching to cell surfaces. Without NS1 we observed a virus particle attachment within that range and with the addition of NS1, attachment increased to nearly 3% representing an 8-fold increase in virus bound. This resulted in downstream increases in virus entry which correlated to an increased virus specific infectivity. Taken together this data suggests, for the first time, a role of extracellular NS1 species in flavivirus infection, and allows us to establish a role of a non-structural protein to function as an attachment factor.
In Chapter 4, we further investigated this NS1-driven increase in virus infectivity and attempt to recapitulate the specific infectivity defect of E343K using purified E343K protein and purified WT virus. We had previously demonstrated that the addition of NS1 to purified virus has the potential to increase an infection, yet we didn’t relate this back to a less artificial system. Here we were successful in demonstrating that if you deplete E343K NS1 protein from mutant virus containing supernatant then add in purified WT protein, virus infectivity increases to near WT levels. Furthermore, we determined that you can detect NS1 directly on virus particles taken from infected cell supernatants, demonstrating that this NS1-virus interaction occurs during a flavivirus infection. Interestingly a reduced amount of NS1 was found bound onto E343K mutant virus, and when E343K protein was added to purified WT virus, we noticed a reduction in virus specific infectivity, compared to adding equal amounts of WT NS1 protein. Further examination of residues nearby to 343, resulted in finding of additional charge reversal mutations that led to similar phenotypes and defined regions within NS1 that are important for forming interactions with not only virus particles, but also cell surfaces. Together, this data adds to a model in which NS1 interacts with virus particles via charged interactions and helps in attaching particles to cell surfaces leading to increased interactions with cellular entry factors.
The last study, Chapter 5, focuses on virus formation, rather than virus infectivity. Here mutations within DENV NS3 were found to alter virus assembly, potentially at the site of nucleocapsid formation. This project was initially designed to see if the tryptophan W344A mutation in DENV, would result in a similar phenotype as it did in YFV (W349A), suggesting a commonality in the role of NS1 for both DENV and YFV particle assembly. However, this did not occur, despite being both sequence and structurally conserved. Instead residues nearby to the YFV W349 residue were found to disrupt particle assembly, without disrupting particle secretion or infectivity. Trans-rescue assays demonstrated that this defect could be rescued in trans with WT NS3 as well as a replicative dead NS3 mutant, but only partially rescued with a helicase dead NS3 molecule. Further passaging of this mutant resulted in the rescue of virus assembly by addition of an H32Y residue change in the cytoplasmic N-terminus of NS4A. Interestingly this H32Y change rescued an NS3 mutant E338A defect completely, but only partially rescued a D334A NS3 mutant. The data in this chapter adds light to the current lack of knowledge in the flavivirus particle assembly process. With these findings, it appears that NS3-4A act in concert for nucleocapsid formation and that NS3’s helicase activity, might in some way play a role in the packaging of viral RNA into newly forming viral particles.
Dibiaso White, Michael J., "Structure-Function Analysis of a Flavivirus Non-Structural Protein" (2018). Open Access Dissertations. 2031.