Structural and Functional Studies of Morbillivirus Proteins

Abstract

Synopsis of Work Done by Abdur Rahaman (S.R. No 111495443) For the Award of Ph.D. Degree in Faculty of Science, Indian Institute of Science, Bangalore, India

STRUCTURAL AND FUNCTIONAL STUDIES OF MORBILLIVIRUS PROTEINS

Rinderpest and Peste des petits ruminants viruses (RPV & PPRV) are two important members of the genus Morbillivirus under the family Paramyxoviridae, having negative-sense single-stranded RNA as their genome. The encapsidated viral RNA genome is embedded inside a host-derived envelope and codes for six structural genes, namely nucleocapsid protein (N), phosphoprotein (P), matrix protein (M), fusion protein (F), hemagglutinin protein (H), and large protein (L).

N protein forms the nucleocapsid that encapsidates viral RNA to form N-RNA, which acts as a template for viral transcription and replication by the viral RNA polymerase complex comprising L and P proteins, where L has the actual polymerization activity and P acts as a transcription transactivator. H and F are surface glycoproteins, where H recognizes and binds to specific host receptors, and F brings about fusion between virus and cells, as well as infected cells and uninfected cells. The present study aims at structural and functional characterization of RPV P and PPRV F proteins that play a pivotal role in the viral life cycle.

The phosphoprotein is so named because of its property of being phosphorylated by a cellular kinase. P protein, in addition to being an integral part of the polymerase complex, has other biological activities that involve the formation of the N0-P complex, a precursor for nucleocapsid formation by keeping N protein in soluble form. The P protein is known to form multimers, but the multimerization state as well as its dependence on phosphorylation for multimerization varies among different viruses in this family.

In the present study, the multimerization status of RPV P protein has been examined, and the domain involved in multimerization has been identified. For this purpose, full-length P protein, earlier cloned in pRSET vector, was expressed and purified to near homogeneity by Ni-NTA agarose column as histidine-tagged protein. The purified protein was subjected to Size Exclusion Chromatography (SEC) using Sephacryl S-300 as matrix. The bacterial-expressed P protein was found to be an oligomer, which was further confirmed by Dynamic Light Scattering (DLS).

The oligomerization domain was located in the C-terminus by using two halves of the protein already available in the laboratory. To further narrow down the region responsible for oligomerization and also to determine the actual oligomerization status, two constructs corresponding to two halves of the C-terminal domain were cloned, expressed, and purified to near homogeneity. The purified proteins were subjected to SEC on Sephadex G-75 column. The results showed that the coiled-coil region in the C-terminal domain is responsible for multimerization to form a tetramer, which was substantiated by DLS and chemical crosslinking studies. Multimerization is not dependent on phosphorylation, as bacterially expressed P protein used in this study is not phosphorylated. The multimerization of P protein is essential for its biological function(s), as demonstrated using a minigenome construct, and the results also show that the multimerization domain is highly conserved both structurally and functionally between RPV and PPRV P proteins.

The domain organization of the P protein has been extensively characterized by secondary structure prediction and CD spectra analysis. P protein could be divided into two major domains: the N-terminal domain is highly unstructured, whereas the C-terminal is rich in -helix content. The C-terminal domain can be further divided into two sub-domains: the multimerization domain, which is coiled-coil in nature, and the rest of the C-terminus, which is also -helical in nature. The three-dimensional structure of the multimerization domain was modeled based on homology using Sendai virus multimerization domain (crystal structure known) as the template structure.

As sequence identity between these two sequences was very poor, different structural properties like secondary structure formation, hydrogen bond formation, and solvent accessibility were taken into consideration for aligning the sequence and were found to be very conserved. The modeled structure is a four-stranded coiled-coil, mainly stabilized by hydrophobic residues coming from four different strands. The model was further substantiated by CD and NMR spectra analysis.

To see the effect of phosphorylation on P protein structure, bacterially expressed purified P protein was phosphorylated in vitro using recombinant casein kinase II and monitored by CD and fluorescence spectroscopy. There was very little change in both secondary and tertiary structure of P protein upon phosphorylation. To verify the same for the protein made in a eukaryotic system, the wild-type and phosphorylation-null mutant P proteins were cloned and expressed in baculovirus expression system. Both proteins were purified using Ni-NTA agarose column. The mutant was found to be unstable, giving cleaved products after purification. The purified wild-type P protein was used for CD and fluorescence analysis and compared with that of bacterially expressed P protein. The results showed different conformational states of P protein in the two expression systems. The baculovirus-expressed P protein was found to be more folded, as indicated by higher secondary structure content. The difference in conformation was further confirmed by thermal melting analysis using CD spectroscopy.

To understand the structure-function relationship of RPV P protein, crystallization studies were undertaken. For this purpose, one deletion mutant containing all the functional domains was used for crystallization trials. The bacterially expressed protein was purified to near homogeneity by Ni-NTA agarose column, and biochemical characterization was performed by SEC and sucrose density gradient ultracentrifugation. The protein was found to be a multimer, where the dimer was disulfide-linked. Crystals were obtained in one of the conditions and diffracted up to 2 Å on a rotating anode image plate. Preliminary X-ray crystallographic characterization showed that the crystal falls in monoclinic system with space group P2 or P2, having four monomeric subunits in the unit cell. Repeated attempts to reproduce the crystallization for structural determination failed, likely due to the highly aggregating nature of this protein, as monitored by DLS and SEC. After several attempts, conditions for isolating homogeneous protein were standardized for future crystallization trials. Crystals were also obtained for the multimerization domain of P protein in one of the conditions.

The fusion protein (F) in this group of viruses is involved in the actual fusion of the viral envelope with the host cell membrane, enabling viral entry into the host cell. The N-terminal hydrophobic peptide in the F1 subunit is involved in the fusion reaction, where two highly conserved heptad repeats, namely HR1 and HR2, are also implicated in this process in enveloped viruses. To understand the structural and functional importance of these two conserved heptad repeats in the fusion process, both regions from PPRV F were cloned after PCR amplification and expressed as fusion proteins in a bacterial expression system. The peptides were purified by Ni-NTA agarose column and cleaved from the fusion tag by Factor Xa protease.

Both heptad repeats were subjected to SEC on Sephadex G-75. HR1 eluted from the column corresponding to a trimer, whereas HR2 was a monomer. This result was substantiated by chemical crosslinking using dithio-bis-succinimidylpropionate (DSP) as crosslinker. CD spectra showed an -helical nature for HR1, with HR2 being mainly unstructured. Both HR1 and HR2 peptides were sensitive to proteinase K digestion when incubated individually, whereas the complex of HR1 and HR2 was resistant under similar assay conditions. This indicates that HR2 interacts with HR1 to form a stable complex. This complex is implicated to form the fusion core, where the inner core is made by a triple-stranded coiled-coil structure of HR1, and HR2 forms the outer layer.

Similar observations are made in all enveloped viruses, where the complex is presumed to be formed during or following the fusion process. Although there is low sequence identity among these heptad repeats of enveloped viruses, there is striking structural and functional similarity. Based on these observations, a structural model was built for the fusion core using SV5 as the template structure. The model confirms the triple-stranded coiled-coil nature of the fusion core, forming a hexamer, i.e., a trimer of HR1-HR2 dimers. The modeled structure is stabilized by conserved hydrophobic residues. Both peptides were found to be involved in the viral fusion process, as shown by inhibition of virus-mediated fusion (syncytia formation) by these peptides.