Estudi de la immunogenicitat de Norovirus humans i de la seua interacció amb receptors cel·lulars

Resumen

Noroviruses (NoVs) are the main cause of sporadic cases and outbreaks of acute gastroenteritis (Tam et al., 2012; Gastanaduy et al., 2013) and are globally associated with a large burden of disease (Patel et al., 2008). NoVs are a highly diverse group of viruses, although over the past two decades most reported NoV outbreaks and epidemics have been caused by NoV GII.4 genotype. Phylogenetic analyses of the GII.4 strains circulating in the last 20 years have shown that this genotype can be divided into distinct variants, which peak and wane over time in a similar pattern to that described for influenza viruses (Buesa et al., 2008; Siebenga et al., 2009; Koelle et al., 2006). In recent years different susceptibilities to NoV infection, depending on their HBGA phenotypes have been reported (Hutson et al., 2004). NoVs belong to the Caliciviridae family and are classified in 5 genogroups (GI-GV), although genogroup I (GI) and GII cause most human NoV infections. The genome is organized in three open reading frames (ORFs): ORF1, ORF2 and ORF3. The VP1, encoded by ORF2, is the major capsid protein, which is further organized into the N-terminal (N), the shell (S), and the protruding (P) domains. The P domain can be further divided into two subdomains: P1 and P2 (Prasad et al., 1999). The P1 subdomain forms the anchoring portion of the P dimer connecting it to the S domain, while the P2 subdomain is exposed on the surface of the capsid protein and is the most variable region of the virus. The main epitopes for immunorecognition and the histo-blood group antigen (HBGA) binding domains reside within this P2 subdomain. The emergence and accumulation of mutations along the P2 subdomain is the main driver of evolution for GII.4 strains, and results in new epidemic strains with altered antigenicity and HBGA binding properties (Allen et al., 2008, 2009; Shanker et al., 2011; Debbink et al., 2012). The cellular receptors that lead to norovirus infections have not been completely characterized yet, but it has been proposed that NoVs are likely to attach to either HBGA expressed on the gastroduodenal epithelial cells of secretor-positive individuals (Marionneau et al., 2002). However, a recent study has shown that NoV could bind to enterocytes independently of HBGAs (Murakami et al., 2013). To date no cell culture system has shown evidence of supporting norovirus propagation, Caco-2 cells, originally derived from human colorectal carcinoma expressing carbohydrates of the histo-blood group family on their surface, allow significant attachment of norovirus VLPs but only when these cells are differentiated. The C-terminal of the VP1 (P domain) of NoVs has the ability to auto-assemble into subviral particles, termed P particles (Tan & Jiang, 2005). Previous work showed that P particles retain the capacity to bind to saliva samples and to synthetic HBGA and that are also immunogenic, making them a valuable tool for the study of viral attachment and for the development of vaccines (Tan et al., 2008, 2011). In this study, the baculovirus expression system in insect cells was used to produce and purify the biologically active form of NoV VLPs from GII.4-1999 (GII.4-v0), GII.4-Hunter_2004 (GII.4-v2) and GII.4-Den Haag_2006b (GII.4-2006b) NoV strains. On the other hand, P particles (P-GI.1, P-GII.4-VA387, and P-GII.9) and the P domain (P-GII.4-2007_Apeldoorn) were produced in bacteria (E. coli) as recombinant proteins tagged with histidines (6xHis), and purified by FPLC in Ni-NTA columns. After the immunization of mice with the different purified NoV VLPs, the mAbs obtained by limiting dilution strongly recognized the P domain of the homologous variant, being indicative of the immunodominance of this region and corroborating the importance of the epitopes previously described in this domain (Zakikhany et al., 2012). Moreover, the av0 mAb also recognized the homologous P particle (P-GII.4-VA387 with a 97’88% of identity), indicating that P particles are antigenically comparable with the entire VLPs. Polyclonal antiserum were obtained from VLPs mix (GII.4-v0, GII.4-v2 and GII.3) immunized rabbit and recognized all of the variants used, both VLPs and P particles, with the exception of the P particle derived from a GI.1 strain. Therefore P particles contain variant-specific epitopes within genotypes but not between genogroups. Another objective of this thesis was to determine the epitope recognized by the anti-2006b mAb (3C3G3) which was performed by phage display technique. This mAb recognizes an epitope formed by 11 residues located into the P2 subdomain of NoV VP1 capsid protein, close to the blockade epitopes described. Thus being able to conduct inhibitory studies to identify biological activity. The binding properties to D-Caco-2 cells of virus-like particles (VLPs) of the different variants of NoV GII.4 and of P particles have been assessed by immunofluorescence, as well as inhibition of binding by: a) saliva; b) porcine gastric mucin (PGM); c) the monoclonal antibodies (mAbs) produced against NoV VLPs; and d) anti-Lewis antigens (Lea, Leb, Lex and Ley) or anti-H antigens (H1 and H2). These assays were performed in order to further investigate the interactions between NoVs and the cellular surface. The results were also compared with binding and blocking salivary assays by ELISA. All VLPs, but not P particles in the same conditions, were able to bind to saliva and D-Caco-2 cells. The different mAbs anti-NoV used blocked the VLP binding to saliva as well as PGM and D-Caco-2 cells, in a dose-dependent manner. But one interesting result when comparing blocking assays in both saliva and D-Caco-2 cells was that while in salivary assays the anti-Lewis or anti-H antigens used blocked the VLP binding, the HBGA blocking on the surface of D-Caco-2 cells did not affect NoV VLP binding. Furthermore, no co-localisation of HBGA and NoV VLPs was observed by immunofluorescence. These results suggest that binding to Caco-2 cells could be mediated by other receptors different from HBGAs, in addition to these. It was also studied the IgG antibody prevalence against NoV GII.4 in a Spanish population using the recombinant P domain of the NoV GII.4-Apeldoorn_2007 variant as the coating antigen in ELISA. Baculovirus-expressed virus-like particles (VLPs) of NoV GII.4-Den Haag_2006b variant were also used as antigen to compare seroreactivity. Of the 434 serum specimens analyzed, 429 (98.6 %) had antibodies against the P domain. The comparison of reactivities of 30 serum samples to the NoV GII.4 P polypeptide and VLP showed reproducible results with a correlation coefficient of r = 0.607. Titers of antibodies to the P domain increased gradually and significantly with age, reaching the highest levels at the age group of 41-50 years. These results confirm the high prevalence of NoV GII.4 infections in our community from early childhood. Finally, it is also presented the data on the immunogenicity of the NoV in natural NoV infections in humans. The results showed that the NoV elicits a humoral immune response in people that had been naturally infected by NoV. The IgG serical antibodies developed against the NoV in individuals suffering acute NoV gastroenteritis, present cross-reactivity against more than one genotype, but not against different genogroups, and are capable of blocking the NoV VLPs binding to saliva.