Abstract
Human norovirus (HuNV) is the most common cause of acute viral gastroenteritis globally. Most human infections are caused by genogroup II genotype 4 (GII.4) noroviruses, which includes the Sydney 2012 strain of norovirus, the dominant variant which has circulated Australia, Aotearoa New Zealand and Oceania over the last decade. Despite the importance of GII.4 noroviruses, they have been relatively neglected in the structural biology field in favour of the less clinically relevant GI.1 genogroup. The lack of atomic protein structures from GII.4 viruses hampers our understanding of this genogroup and creates a barrier to the development of GII.4-specific antiviral therapeutics. In addition, the protease-polymerase (ProPol) precursor protein has not been solved from either genogroup, despite the essential role it plays in viral replication. This research aimed to solve the atomic structures of three proteins from GII.4 Sydney 2012 norovirus; the mature 3C-like protease and RNA-dependent RNA polymerase (RdRp), and the ProPol precursor. In addition, this project aimed to investigate cross-inhibitory activity of known antivirals against the ProPol precursor protein.
To produce recombinant ProPol protein for structural research, several high-yield expression systems were developed in a Trichoplusia ni (T. ni) insect cell-baculovirus expression system. The development of these expression systems led to the identification of three previously unknown protease cleavage sites targeted by the ProPol enzyme. One target sequence (FQ/GW) shares some sequence homology with a native cleavage target recognized by the ProPol precursor, whilst the other two (FE/KG, FE/KA) share no obvious homology with known target sequences. These findings may have implications for potential host targets cleaved during viral replication. Ultimately, a His6-tag system was developed for the high-yield production of recombinant ProPol protein. Native PAGE gel evidence revealed H6.ProPol exists in both monomer and dimer forms, and can form a binary complex with VPg, the target of nucleotidyl- transfer during viral replication. Thermal shift analysis indicated that this H6.ProPol-VPg complex is stable. RdRp inhibitors were tested by polymerase assay against the H6.ProPol precursor and an effective compound, PPNDS, was identified (IC50: 2.0 μM).
To solve the 3D atomic structure of H6.ProPol, X-ray crystallography was pursued through systematic high-throughput crystallography screening. In more than 3500+ crystallisation conditions, no confirmed H6.ProPol crystals were obtained in this study. However, successful negative stain electron microscopic analysis of H6.ProPol produced a low-resolution envelope map of the dimer form. Cryo-electron microscopy (cryo-EM) was then pursued, which yielded a map of the polymerase domain of H6.ProPol to atomic resolution of 2.61 Å, although the protease domain of H6.ProPol was not resolved because of inherent flexibility at the Pro/Pol linker. The structure revealed that the precursor apopolymerase is highly structurally conserved with related GII.4 mature apopolymerases.
The mature polymerase and protease proteins were subjected to crystallography screening and protein crystals were obtained for both enzymes in a variety of buffers, salts and precipitant conditions. Mature polymerase crystals did not yield a data set from which a structure could be solved in this study. However, mature protease crystals yielded two apoprotease crystal structures at resolutions 2.79 Å and 3.09 Å, and two inhibitor-bound structures at resolutions 1.83 Å and 3.33 Å. The apoprotease structures reveal conformational changes in an arginine residue (R112) near the catalytic triad which regulates the state of the active site, a finding consistent with other protease studies. The inhibitor-bound structures provide insight into GII.4 proteases when covalently bound to potent antiviral compounds, including one structure which was revealed to have a second small molecule binding site within the S4 substrate-binding pocket.
This thesis contributes six new protein structures to the field, the first structures to be solved from the Sydney 2012 variant of norovirus. These include both apo- and inhibitor-bound protease structures. This work also reports the first partial structure of a precursor protein to be solved in the Caliciviridae family of viruses. This work contributes to the structural understanding of GII.4 proteases and may provide a good basis from which structure-aided drug design can be pursued.