A COMBINED GENETIC AND CHIMERIC ANALYSIS OF THE FLAVIVIRAL NON-STRUCTURAL PROTEINS

A successful flaviviral life cycle involves several coordinated events between viral proteins and host factors. The polyprotein processing at the surface of the ER membrane results in the formation of several replication proteins that bring about changes in the ER membrane making it permissive for viral genome amplification. Non-structural proteins 4A (NS4A) and non-structural protein 4B (NS4B) are two of the most important integral membrane proteins of DENV that are essential part of the viral replicase complex. The cleavage at NS4A-2K-NS4B is temporally and spatially regulated. The cleavage at the N-terminal of 2K is carried out by viral NS2B/3 protease while host signalase cleaves on the C-terminal side at the ER lumen to give rise to a mature NS4B protein. This thesis primarily focuses on demonstrating the function of 2K as an independent peptide rather than simply a signal sequence, and the role 2K plays, when present as 2K-NS4B vs NS4B. Moreover, this thesis has attempted to explore the function of transmembrane domains (TMDs) in replication separating them from their membrane anchor function. This thesis will also describe the development of a ZIKV replicon and its use in screening small molecule inhibitors in the last chapter.

In Chapter 2 of the thesis, we established 2K as an independent, information carrying peptide rather than just a signal peptide. A strategy involving chimeric virus generation and mutational analysis supported the notion that 2K is rather unique and important for viral replication and infectious particle production. Using an interserotypic 2K chimeric virus, it was established that the 2Ks of DENV are serotype specific, however, they are interchangeable with a huge fitness cost in infectious particle production. We further showed that individual amino acid residues towards then end of h-region and C-terminus of the 2K peptide affect viral replication and infectious particle production. Moreover, it was shown that the 2K peptide consists of a highly conserved ‘DNQL’ region at its N-terminal that plays an important role in viral replication.

Chapter 3 details the mechanistic aspect of the effects observed in interserotypic 2K chimeric viruses. The interserotypic chimeric viruses were comparable to wild type in replication, however, they were deficient in infectious particle production early in the life cycle. The major change to be noted in the chimeric viruses was the absence of signalase cleavage at the 2K-NS4B junction. We demonstrated that in a virus infected system, 2K-NS4B and NS4B populations are always present which led us to look for any specific functions of the cleaved vs uncleaved 2K-NS4B protein. Using a transcomplementation system where NS4B was presented in the absence of 2K, we showed that particle production can be rescued in the interserotypic 2K chimeric viruses. It was further concluded using NS4B truncations that the property of NS4B to rescue particle production was concentrated in the ER luminal loop. Further, alanine scanning mutagenesis of the conserved residues of ER loop resulted in pinpointing T198 and its involvement in the early stages of viral packaging.

Chapter 4 examined the role of TMDs of NS4A and NS4B and attempted to define their roles separately from their membrane anchoring functions. Several interserotypic TMD chimeric viruses were generated to address the function of these domains. We concluded that TMD1 and TMD3 of NS4A could be replaced with partial success across the DENV serotypes, whereas, TMD2 was serotype specific. The specificity of TMD2 of NS4A is not contributed by a single amino acid and should be a function of the secondary structure formed by TMD2 as it sits on the inner leaflet of the ER membrane. We demonstrated the variable roles different TMDs of NS4B play in viral replication using a similar strategy of reverse genetics of chimeric viruses. TMD1 of NS4B was replaceable with no to minimal effect, whereas, the remaining four showed variable effect upon substitution. More importantly, we demonstrated how the reorientation of TMD5 of NS4B post NS2B/3 cleavage might vary in different serotypes of DENV using revertant virus obtained from the TMD5 interserotypic chimera. Analysis of interserotypic cytosolic and ER luminal loop chimeras of NS4B pointed to functional conservation of the cytosolic loop between DENV-2 and DENV-3, whereas, the remaining cytosolic loops and the ER loops showed variable level of defects upon substitution, suggesting their functions in serotype-dependent manner.

Chapter 5 describes the construction and characterization of a ZIKV replicon system and use of it to screen several small molecule inhibitors of the flaviviruses MTase. Several small molecule inhibitors of flavivirus N-7-MTase were designed/synthesized in Dr. Arun K Ghosh’s lab which would target the extra pocket unique to the flavivirus SAM-binding site. We analyzed the docking of a set of these compounds into MTase domain of NS5 of ZIKV, DENV and YFV and screened them for their ability to inhibit replication of ZIKV, DENV and YFV. A huge variation in the activity profile of these compounds were observed against different flaviviruses even though these compounds were targeted against the highly conserved MTase domain of flavivirus NS5. GRL-002- and GRL-004-16-MT specifically inhibited ZIKV replication with low micromolar IC₅₀ value, while these compounds showed little to no effect on DENV and YFV. On the other hand, compounds GRL-007-, GRL-0012- and GRL-0015-16-MT demonstrated a dual inhibitory effect against DENV and YFV albeit the CC₅₀ values of the GRL-012 and GRL-015 were concerning. Compounds GRL-007-16-MT showed broad spectrum activity against ZIKV, DENV and YFV even though it was slightly cytotoxic to Vero cells. Moreover, GRL-002-16 was inhibitory to YFV while ineffective against DENV, whereas, GRL-016-16 had the opposite effect. Our results reveal the differential efficacies of the small molecule inhibitors targeting N-7-MTase. The experimental data suggests these compounds have different cytotoxicities in different cell lines and the compounds act in a virus-specific way. Nonetheless, we were able to shortlist some potent compounds for future modifications.