Insights Into The Mechanism Of Polyprotein Processing Of Sesbania Mosaic Virus And Characterization Of The Polyprotein Domains

Abstract

1. Viruses are obligate parasites that hijack the host cell machinery to synthesize their own gene products and for their propagation. In order to establish a successful viral infection, viruses have evolved different strategies to evade host check points. Further more, their success also relies in employing varied strategies to express maximum number of functional proteins from their small constrained genome. Polyprotein processing is a widely used strategy of expression by many plant viruses. With limited information available on this aspect for sobemoviruses, the present study was undertaken. 2. The present thesis deals with the mechanism of Sesbania mosaic virus (SeMV) polyprotein processing and functional characterization of the polyprotein domains. SeMV infects Sesbania grandiflora that belongs to the Fabaceae family. It is a positive sense ssRNA virus with a genome length of 4149 nucleotides. The genome encodes four potential overlapping open reading frames (ORFs). ORF1 codes for an 18 kDa protein that is proposed to be involved in the movement of the virus. ORF 3 codes for the coat protein (CP) that encapsidates the viral genomic RNA to form the viral particles. The central ORF codes for polyprotein that has a serine protease domain at its Nterminus that cleaves the polyprotein at specific E-T/S sites to release the functional domains. So far only in SeMV, the E. coli expressed polyprotein, Protease-VPg-RdRp was shown to undergo processing at E325-T326, E402-T403 and E498-S499 releasing protease, VPg, P10 and RdRp domains respectively. 3. Based on the arrangement of the central ORF, the genome organization of SeMV was earlier shown to be like that of SCPMV type. However, recent sequencing data from the laboratory showed that the organization of SeMV gRNA was like that of CfMV type. This would imply that in SeMV the central two ORFs will be translated to give two polyproteins, 2a (Protease-VPg-C-terminal domain) and 2ab (Protease-VPg-RdRp) the C-terminus of 2a and N-terminus of RdRp being different from what was reported previously. Therefore, in the light of the new genome organization for SeMV, the mechanism of processing of polyprotein 2a and 2ab needs to be revisited. 4. SeMV protease domain was shown to require natively unfolded VPg at its Cterminus for its activity. Aromatic stacking interactions between protease and VPg (via W43 residue) were shown to confer the activity to the protease. However, the residues in the protease domain involved in these interactions have not been identified. 5. The objectives of the present studies are • To elucidate the mechanism of processing of polyproteins 2a and 2ab in E. coli and in planta. • To identify residues in the protease domain involved in mediating aromatic stacking interactions with VPg. • To functionally characterize the C-terminal domain of polyprotein 2a. 6. Polyprotein 2a when expressed in E. coli, from the new cDNA clone, got cleaved at the earlier identified sites E325-T326, E402-T403 and E498-S499 to release protease, VPg, P10 and P8 respectively. The specificities of the cleavage sites were established by mutational analysis. 7. Additionally, a novel cleavage was identified within the protease domain at position E132-S133. The polyprotein 2a that was mutated for this site (ΔN70 2a-E132A) showed no release of P8 protein though the polyprotein was intact for E498-S499 site. Unlike other cleavage site mutants, ΔN70 2a-E132A mutant also revealed large accumulation of intact polyprotein, again implying that the mutation not only abolished the proteolytic cleavage at that site but hampered the processing at other sites. The results confirmed that the cleavage at N-terminus of the protease/polyprotein is crucial for an efficient processing in particular for the cleavage between P10-P8. 8. Interestingly, though the sites in polyprotein 2ab are exactly the same as identified in polyprotein 2a, the former got cleaved between Protease-VPg but not between VPg-RdRp. This cleavage site appeared to be rather masked in polyprotein 2ab. Also, the cleavage at E132-S133 site appeared to be rather slow. These results indicate to a differential cleavage pattern, governed probably by the conformation of 2ab. In other words, the local context of the cleavage site and just not the sequence per se could be playing a key role in 2ab polyprotein processing. 9. Products, corresponding to all cleavages identified in E. coli (E132-S133, E325-T326, E402-T403 and E498-S499) were also detected in infected Sesbania leaves. Products corresponding to the sizes of ΔN132 Protease and ΔN132 Protease-VPg were detected suggesting that the removal of the membrane anchoring domain from the protease does occur in planta. Also, detection of band corresponding P8, confirmed that the cleavage between P10-P8 indeed occured in planta too. 10. The trans cleavage experiments suggested that not all of the four cleavages in polyprotein 2a occur in trans (intermolecular). Cleavages at E132-S133 and E498-S499 do not occur in trans impling that cleavages at these sites could only occur in cis (intramolecular) by auto-proteolysis of the polyprotein. 11. The Thr at P1’ did not make a site trans cleavable. Interestingly, SeMV protease was found to cleave even an E-S site in trans but only when present at positions 324-325 and 402-403, suggesting that trans cleavage in SeMV is governed by the context rather than the Thr at P1’position of the cleavage site. The E498-S499 site was found to be highly stringent not only for the mode of its cleavage (cis cleavage) but also for its sequence (E-S only). A Thr substitution for Ser at this site, made it non cleavable in cis. 12. The results reveal that the polyprotein processing in SeMV is regulated by a number of strategies, viz. a) availability of the cleavage site depending on the conformation of the flanking domains (E132-S133 and E402-T403 cleavages in 2ab). b) Mode of recognition (cis or trans). c) Context/position of the cleavage site. 13. Based on the sequences of all four cleavage sites identified, a consensus has been drawn for SeMV serine protease cleavage site, i.e., N/Q-E-T/S-X (where X is an aliphatic residue) at P2-P1-P1’-P2’ position respectively. 14. With a view to understand the structural reasons for such high specificity, the residues in the S1 and S2 binding pocket, that recognize the substrate P1 and P2 residues respectively, were identified based on the structural comparison of SeMV protease with other Glu/Gln specific proteases. Mutational analysis of these residues clearly demonstrated that H298, T279 and N308 of the S1-binding pocket that would bind the substrate glutamate are crucial for the protease activity. R309 that forms the S2 binding pocket is also crucial for protease activity. 15. Also, the P2 (Asn/Gln) residue recognized by R309 plays an important role in determining the substrate specificity. A positively charged residue Lys was not tolerated at this position. SeMV protease was also shown to efficiently cleave the peptide bond C-terminus to an uncharged Gln in vivo suggesting that it is a Glu/Gln specific protease. 16. An interesting feature of the SeMV protease domain is the presence of a disulphide bond that holds the S1-binding pocket. However, unlike for the cellular counterparts like trypsin, the disulphide was found to be not essential for either the SeMV protease activity or structural stability. 17. Protease and VPg domains were proposed to be involved in aromatic interactions that conferred activity to the protease. The structure of protease revealed a stack of aromatic residues (W271, F269. Y315 and Y319) exposed to the solvent. Mutational analysis was performed to identify their role in mediating the interactions and hence the activity of protease. H275, though not a part of exposed aromatic stack in the protease, was chosen for mutational analysis as it lies close to the W271 in sequence and is conserved in the protease domain across all the known sobemoviruses. The in vivo and trans cleavage assays suggested that residues W271 and H275 but not Y315 or Y319 are crucial for protease activity. 18. The Far-UV CD spectrum of protease-VPg is characterized by a positive peak at 230 nm, signifying the aromatic interactions. Far-UV CD spectral analysis of the aromatic mutants showed that W271 and H275, but not F269 and Y319 are the major contributors of the 230 nm positive peak, confirming the direct involvement of these residues in the stacking interactions with W43 of VPg. Thermal stability studies, fluorescence spectroscopy and 1D-NMR spectroscopy studies also confirmed the histidine aromatic interactions between W271, H275 of protease with W43 of VPg. 19. The loss in aromatic interactions in the mutants caused Protease-VPg to aggregate, suggesting that the aromatic interactions between protease and VPg not only conferred activity to the protease but also the active oligomeric status. 20. In silico analysis of the C-terminal domain showed that it has no significant similarities with any known functional proteins. The region corresponding to P8 was amplified and cloned in pRSET C vector, over-expressed and purified. 21. The purified His-tagged P8 showed mass abnormality on the SDS-PAGE. However, the mass spectrometric analysis of the purified protein showed that it had a molecular mass of 9.766 kDa as is expected for a His-tagged P8. P8 is highly basic, which could possibly explain its anomalous behaviour on the SDS-PAGE. The purified recombinant P8 protein was found to be natively unfolded. In vitro binding studies revealed that P8 had nucleic acid binding property. The protein was also found to be phosphorylated both in vitro and in vivo conditions. 22. Interestingly, P18, (a precursor of P8) but not P8, was found to possess an inherent ATP hydrolyzing property. Optimum conditions for the ATPase assay were found to be Tris HCl pH 8.0, 37 ºC, 5 mM MgCl2. The activity was linear upto 20 mins. P18 could utilize all NTPs and dNTPs. Studies revealed that ATPase activity resided in the P10 domain of P18, though P8 region could enhance the activity. Conclusively, the results demonstrate that the C-terminal domains of polyprotein 2a have ATPase and nucleic acid binding activity and could therefore have possible roles in movement and replication.