Studies on the hemagglutinin spike protein VP4 of human rotaviruses

Abstract

Rotaviruses belong to the family Reoviridae and are the major cause of severe, acute infantile gastroenteritis in humans and animals worldwide. Rotavirus diarrhoea accounts for about a million deaths per year, primarily in developing countries such as India. Rotaviruses are nonenveloped, tripleshelled, icosahedral viruses consisting of an outer shell composed of two proteins, VP4 and VP7; an intermediate shell composed of VP6; and an inner shell composed of VP2 that encloses the core containing VP1 and VP3 complexed with the genomic doublestranded RNAs. The outer shell is primarily made up of the glycoprotein VP7, which forms the smooth surface on which 60 spikelike projections-dimers of VP4-are located. Rotaviruses are classified into seven serogroups (A-G) on the basis of groupspecific epitopes on VP6. Group A rotaviruses are the primary cause of infantile gastroenteritis in humans. Viruses within a serogroup are further classified into “subgroups” and “serotypes.” Subgroup specificity is determined by epitopes on VP6, while the outer capsid proteins VP4 and VP7 define P serotype and G serotype, respectively. Using neutralizing serotypespecific monoclonal antibodies, about 14 G serotypes have been identified. Due to the lack of similar reagents against VP4, rotaviruses have been classified into P types primarily based on relative amino acid homology among VP4 proteins; at least 20 P types have been identified in humans and animals. The rotavirus genome comprises 11 segments of doublestranded RNA (dsRNA), numbered 1-11 according to electrophoretic mobility. The average genome size of group A rotavirus is 18,557 bp. Except for segment 11, each segment potentially encodes a single polypeptide. All rotavirus genes are A+Urich (58-67%), and neither genomic nor plussense RNAs are polyadenylated. The plusstrand carries a capped 5 end. Gene segment 4 encodes the outer capsid spike protein VP4, which plays key roles in hemagglutination, virulence, neutralization, protease susceptibility, infectivity, cell attachment, growth restriction in vitro, and plaque morphology. VP4 is cleaved by trypsin into VP5* and VP8*, and this cleavage enhances infectivity both in vivo and in vitro. VP4 induces protective immunity in animals and is immunogenic in children, making it a potential candidate for a recombinant rotavirus vaccine. Immunological and sequence analyses of VP4 and VP7 from global isolates have shown considerable antigenic variation. Little was known about the antigenic nature of Indian human rotaviruses. This study contributes to understanding the genetic and antigenic variation among Indian isolates. Human rotavirus strain IS2 (subgroup I, G2 serotype) was isolated from a child with diarrhoea in Bangalore. Using dsRNA segments 1, 2, 3, and 4 purified directly from a clinical sample, a cDNA library was constructed using the Okayama-Berg method. Screening with a segment4specific probe yielded five positive clones (10, 11, 30, 43, 47). Terminal sequence analysis showed none were fulllength, but two overlapping clones, pBS11 and pBS47, spanning nt 829-2359 and 1-1655 respectively, allowed reconstruction of fulllength gene 4 via a unique PflMI site at nt 1375. Complete nucleotide sequencing of gene 4 revealed a 2359nt ORF (nt 10-2337) encoding VP4 of 775 amino acids. Comparative analysis showed IS2 VP4 is closely related (95% identity) to other P4 alleles from DS1, RV5, L26, and L27. IS2 differed by 33 amino acids from RV5 and DS1, with notable substitutions of neutral residues by acidic ones at positions 49, 392, 662, and 678. These changes converted the typically basic P4 VP4 protein into an acidic protein (pI 6.26), making IS2 unique among known P4 strains. These findings and other studies indicate that Indian rotaviruses exhibit substantial sequence variation. Because human rotaviruses grow poorly in culture, VP4 yields are very low. Therefore, expression in E. coli was attempted for structural and functional studies. Previous attempts by others to express fulllength VP4 in E. coli yielded only insoluble fragments; VP5* and VP8* were successfully expressed only as insoluble fusion proteins, while baculovirus systems yielded VP4 that could not be purified economically. In this study, three vector systems were evaluated for soluble expression of VP4 in E. coli. Vectors pKK2333 (tac promoter) and pET20(b) (T7 promoter) produced only degraded polypeptides. To overcome instability, VP4 was expressed as a soluble maltosebinding protein (MBP) fusion using pMALc/p. VP4, VP5*, VP8*, and VP8* deletion mutants (E7, N, C) were expressed and purified by singlestep amylose affinity chromatography. For comparison, VP8* and its deletion mutants (S, H) from bovine rotavirus strain 1321 (G10P11) were also expressed as MBP fusions. Hemagglutination assays on recombinant proteins showed that MBPVP4 and MBPVP8* of IS2 hemagglutinated 1dayold chick RBCs, while VP5*, E7, N, C and MBP alone did not. Similar results were obtained for strain 1321. Thus, the hemagglutination domain resides within VP8* and is compromised by large N or Cterminal deletions. AntiMBPVP4 antibodies inhibited hemagglutination, confirming the conformational authenticity of expressed VP4 and VP8*. Hemagglutinationinhibition assays demonstrated that antiIS2 VP4 antibodies inhibited hemagglutination in IS2 as well as SA11 (simian, G3), UK (bovine, G6), and OSU (porcine, G5) strains, indicating the presence of conserved conformational epitopes across P types. AntiMBP antibodies had no effect. These findings show that antibodies raised against purified VP4 can detect diverse rotavirus strains and are suitable for ELISAbased diagnosis in clinical samples.