Rotavirus is the major cause of infantile gastroenteritis. Each year more than 600,000 young children are estimated to die in developing countries throughout the world. Rotavirus infection can be either symptomatic or asymptomatic. But the genetic or molecular basis for rotavirus virulence is not yet clearly understood. NSP4, encoded by genome segment 10, is a multifunctional protein. It is identified as the first viral enterotoxin and is essential for virus morphogenesis and pathogenesis. Analysis of NSP4 from more than 175 strains failed to reveal any sequence motif or amino acid that segregated with the virulence phenotype of the virus. Further, a few studies indicated a lack of consistent correlation between virus virulence and diarrhea inducing ability of the cognate NSP4. To understand the basis for the inconsistency in the enterotoxigenic activity of a few NSP4s reported in a limited number of studies, comparative analysis of the biophysical, biochemical, and biological properties of NSP4ΔN72, which from SA11 and Hg18 was earlier shown to be highly diarrheagenic, from 17 different symptomatic and asymptomatic strains was carried out. To study structure-function relationship we used Thioflavin T fluorescence assay, gel filtration, CD spectroscopy, trypsin susceptibility and enterotoxin assay in newborn mice for all the proteins. Detailed comparative analysis of biochemical and biophysical properties and diarrheagenic activity of the recombinant ΔN72 peptides under identical conditions revealed wide differences among themselves in their resistance to trypsin cleavage, thoflavin T binding, multimerization and conformation without any correlation with their diarrhea inducing abilities. Since earlier studies showed that a secreted peptide (ΔN112) of SA11-NSP4 also induced diarrhea in newborn mice pups, we have generated NSP4ΔN112 deletions from six different strains and tested for their diarrhea inducing ability. The patterns of DD50 values of the ΔN112 peptides was similar to that for ΔN72 peptides, but were 1000-1200-fold less efficient than that of SA11ΔN72. NSP4 exists in multiple forms in the infected cells- as oligomers, higher molecular weight complexes and ER- and cytoplasmic membrane anchored forms. Previous studies suggest that the N-terminal boundary of the oligomerization domain could lie downstream to residue 94 from the N-terminus. A peptide from residue 112-175, secreted from rotavirus infected cells, was reported to induce dose-dependent diarrhea in suckling mice, suggesting that the N-terminal boundary of the enterotoxin activity could lie around residue 112. However, the precise N-terminal boundaries in NSP4 for oligomerization and diarrhea induction have not been identified. To address this question, a large number of deletion mutants C-terminal to residue 94 were generated and tested for their ability to induce diarrhea in newborn mouse pups. Our data suggest that while the deletions ∆N121 to ∆N131 failed to induce diarrhea, ΔN118 was diarrheagenic suggesting that the N-terminal boundary of the minimal diarrhea inducing domain lies between aa 118 and 121. Size exclusion chromatography revealed that residues 95 to 98 are critical and sufficient for oligomerization. Studies on oligomerization further revealed that NSP4ΔN94 exists in pentamers, tetramers and dimers, while deletion mutants C-terminal to aa 94 exist only as dimers. Our studies demonstrate for the first time that not only tetramers but pentamers as well as dimers possess enterotoxigenic properties. Most human rotavirus infections are caused by group A rotaviruses. Within this group, rotaviruses are further classified into subgroups based on the antigenic specificity associated with the protein product of the sixth gene, VP6. Previous studies have mapped SG I specificity to aa position 305 and the region between 296 and 299, and SG II specificity to residue 315 on VP6. However, the subgroup specific determinants on NSP4 have not been identified till date. In this study, we generated several amino acid substitution mutants in the SG I-specific SA11 NSP4∆N72 protein as in previous studies ∆N72 was found to efficiently bind DLPs. Using an enzyme linked immunosorbent assay method, the effect of the mutations in the C-terminal and N-terminal regions in ∆N72 on their binding ability to SG I and SG II DLPs was assayed. Residues at positions 85, 169, 174 and 175 and in the ISVD appear to collectively determine the specificity of binding to DLPs. While the conserved proline and glycines at positions 165, 168 and 162, respectively, are important for maintaining the required conformation for general recognition of DLP. The present study provides important insights towards understanding the determinants in NSP4 for SG-specific DLP interaction.

Rotaviruses, members of the family Reoviridae, are the major etiologic agents of severe, acute dehydrating diarrhea in the young of many mammalian species, including humans, calves and foals. Recent estimates indicate an annual death toll of approximately 600,000 infants due to rotavirus, besides inflicting staggering economic burden worldwide. Most of these deaths occur in the developing countries and India is estimated to account for about a quarter of these deaths. Extensive molecular epidemiology studies carried out by our laboratory have revealed many interesting aspects about rotavirus diversity in this country. Molecular epidemiology of rotaviruses causing severe diarrhea in foals in two organized farms in northern India was carried out. These foal rotaviruses exhibited 5 different electropherotypes (E), E1-E5. Strains belonging to E1, E2 and E5 exhibited G10, P6[1]; G3 and G1 type specificities. Though the E1 strains possessed genes encoding G10 and P6[1] type outer capsid proteins, unlike the G10, P8[11] type strain I321, they exhibited high reactivity with the G6-specific MAb suggesting that the uncommon combination altered the specificity of the conformation-dependent antigenic epitopes on the surface proteins. Strains belonging to electropherotypes E3 and E4 were untypeable. Sequence analysis of the VP7 gene from E4 strains (Erv92 and Erv99), revealed that they represent a new VP7 genotype, G16. Nonstructural protein 4 (NSP4) of rotavirus is a multidomainal, multifunctional protein and is the first viral enterotoxin identified. We have recently reported that the diarrhea-inducing and double-layered particle (DLP)–binding properties of NSP4 are dependent on a structurally and functionally overlapping conformational domain that is conferred by cooperation between the N- and C-terminal regions of the cytoplasmic tail (Jagannath et al., J. Virol, pp 412-425, 2006). Further, a stretch of 40 amino acids (aa) from the C-terminus is predicted to be unstructured and highly susceptible to trypsin cleavage. We examined the role of this unstructured C-terminus of Hg18 NSP4 and SA11 NSP4 on the biological properties of NSP4 using a series of deletion and substitution mutants of the conserved proline and tyrosine residues in this region. Gel filtration, CD spectroscopy and Thioflavin T binding studies showed that these mutants have altered secondary structural contents and either failed to multimerize efficiently or multimerized with altered conformation. The C-terminal ten residues appear to play a regulatory role on multimerization. Proline 168, tyrosine 166 and methionine 175 appear to be critical determinants of DLP binding activity whereas, proline 165 and tyrosine 85 and 131 appears to determine the affinity of binding to DLP in the context of NSP4 ∆N72. Deletion and substitution mutants exhibited severely reduced diarrhea inducing ability and DLP binding property. Of great biological significance is the drastic decrease in the diarrhea inducing ability of the N- and C- terminal deletion mutant ∆N94 ∆C29 that exhibited about 11,000-fold increase in DD50 than the wild type (WT) ∆N72. These studies revealed that the predicted unstructured C-terminus is an important determinant of biological properties of NSP4. Extensive efforts to crystallize the complete cytoplasmic tail (CT) of NSP4 were unsuccessful and to date, the structure of only a synthetic peptide corresponding to aa 95-135 has been reported. Our recent studies indicate that the interspecies variable regions from aa 135-141 as well as the extreme C-terminus are critical determinants of virus virulence and diarrhea-inducing ability of the protein. Here, we examined the crystallization properties of several deletion mutants and report the structure of a mutant recombinant NSP4 from symptomatic (SA11) and asymptomatic (I321) strains that lacked the N-terminal 94 and C-terminal 29 aa (NSP4: 95-146) at 1.67 Å and 2.7Å, respectively. In spite of the high-resolution data, electron density for the stretch of 9 residues from the C-terminus could not be seen suggesting its highly flexible nature. The crystal packing showed a clear empty space for this region. Extension of the unstructured C-terminus beyond aa 146 hindered crystallization under the experimental conditions. The present structure revealed significant differences from that of the synthetic peptide in the conformation of amino acids at the end of the helix as well as crystal packing owing to the additional space required to accommodate the unstructured virulence-determining region. Conformational differences in this critical region effected by the presence or absence of proline or glycine at specific positions in the unstructured C-terminus, could form the basis for the wide range of variation seen in the diarrhea-inducing ability of NSP4 from different strains in newborn mouse pups. Although symptomatic and asymptomatic strains do not generally differ in the presence or absence of the conserved prolines or glycines, they contain a few additional changes that could alter the unique conformation required for optimal biological activity. In conclusion, we demonstrate that the predicted unstructured C-terminal region is indeed highly flexible and is an important determinant of biological functions of the NSP4, mutations in which probably correlates with the virulence properties of the virus.

Structuraj investigations on the Ddi1 (DNA-damage inducible protein 1) of Leishmania major and on the rotavirus nonstructural protein NSP4 were carried out. Ddi1 belongs to the ubiquitin receptor family of proteins. One of its domains is similar to the retroviral aspartic proteinases. It has been shown that this domain is the target of HIV-protease inhibitors that were being used in the treatment of AIDS and it was observed that these drugs effectively controlled opportunistic diseases caused by many parasitic protozoa such as Leishmania and Plasmodium species. The retroviral protease-like domains present in Ddi1 proteins of these organisms were identified as the targets of these drugs. Structural studies on Ddi1 from L. major have been carried out, in an attempt to provide a platform for the design of anti-protozoal compounds. Rotavirus NSP4, the first viral enterotoxin to be identified, is a multifunctional glycoprotein that plays critical roles in viral pathogenesis and morphogenesis. As part of an ongoing project on the structural characterization of NSP4, we determined the structure of the diarrhea-inducing region of this protein from the rotavirus strain MF66. Chapter 1 presents an overview of Ddi1 and NSP4 of the rotavirus with an emphasis on their structural features. The methods employed during the course of the present work are described in Chapter 2. Structural studies on the retroviral protease-like domain of Ddi1 (Ddi1-RVP) of L. major is presented in Chapter 3. Apart from this domain, Ddi1 of L. major also has a ubiquitin-associated and ubiquitin-like domains whereas P. falciparum has only the ubiquitin-associated domain. Activity of the full length Ddi1 of L. major and the retroviral protease domain of P. falciparum using an HIV protease substrate was shown to be inhibited by an HIV protease inhibitor, saquinavir. Binding of saquinavir to the proteins was also confirmed by Biolayer Interferometry studies. The crystal structure of the retroviral protease domain of L. major Ddi1 has been determined. It forms a homodimeric structure similar to that of HIV protease and the reported structure of the same domain from Saccharomyces cerevisiae. The loops in Ddi1-RVP are similar to the 'flap' regions of the HIV protease which close-in upon substrate/inhibitor binding; they are visible in the electron density maps, unlike the case of the S. cerevisiae protein. Though the native form of the domain shows an open dimeric structure, normal mode analysis reveals that it can take up a closed conformation resulting from relative movements of the subunits. The present structure of Ddi1-RVP of L. major with the defined 'flap'-like loops will be helpful in the design of effective drugs against protozoal diseases, starting with HIV protease inhibitors as the lead compounds. Chapter 4 describes the structural investigations carried out on the diarrhea-inducing region of the nonstructural protein NSP4 of the rotavirus strain MF66 which forms an α-helical coiled-coil structure. Crystal structures of a synthetic peptide and of two recombinant proteins spanning this region showed parallel tetrameric organization of this domain with a bound Ca2+ ion at the core. Subsequently, we determined the structure of NSP4 from a different strain as a pentamer without the bound Ca2+ ion. This new structure provides more insights into understanding some of the functions of NSP4 such as the release of ions into the cytoplasm and binding to the double-layered particle (DLP). We also established conditions responsible for these structural transitions. The crystal structure of the coiled-coil domain of NSP4 presented in this chapter shows an entirely different structure which is an antiparallel tetramer. This explains our failure to determine the structure by the molecular replacement method using known oligomers. The structure was solved by the Sulphur-SAD method using diffraction data collected with Cr Ka radiation. The study reveals that the structural diversity of NSP4 is not limited. We could relate sequence variations and pH conditions to the differences in oligomeric assemblies. Surface properties of the domain suggest that the new form is likely to interact with different sets of proteins compared to those that interact with the parallel tetramers or pentamers. Further investigations are needed to establish this property.