Glycan Attachment Specificity, Toward ROtavirus Vaccine IMprovement – GASTROVIM

Rotaviruses (RVAs) are the most common cause of acute gastroenteritis in children. The disease frequently requires hospitalization and annually causes the death of several hundred thousands children in low-income countries. Two live vaccines have been developed and both show remarkable efficacy in high-income countries, but unfortunately and for unknown reasons they prove much less efficient in several regions of lower income. Nonetheless, recent data on the glycan-binding properties of human RVAs unexpectedly offered a potential clue to understanding of this limitation. RVAs are known to interact with glycans of the host cell membrane through their VP4 spikes, more specifically by the VP8* outermost domain. Until recently it was considered that, depending on the strains P-genotype characterized by genetic variation of the VP4 protein, interacting glycans contained sialic acid motifs, either sensitive or resistant to sialidase treatment. The new data indicate that human RVA strains additionally bind to neutral fucosylated glycans of the histo-blood group type (HBGAs). These carbohydrate antigens are located at the termini of either O-linked or N-linked glycan chains of proteins and of lipids and are expressed mainly on epithelial cells. They are characterized by an extensive genetic variation caused by common polymorphisms at the ABO, FUT2 and FUT3 loci. These genes encode glycosyltransferases that contribute in concert to the synthesis of the ABH and Lewis antigens that define the ABO, secretor and Lewis phenotypes. The VP8* of some strains appears to recognize the Leb difucosylated motif, whilst others bind to the A blood group antigen. Based on a limited number of cases, we and others observed that individuals lacking a functional FUT2 enzyme (nonsecretors) were never found among children with severe rotavirus gastroenteritis, suggesting resistance to infection by the common P[8] strains. This is consistent with the requirement of a functional FUT2 enzyme in the synthesis of the Leb ligand. Others additionally observed that children infected with P[6] strains were found among Lewis positive (FUT3+) only, suggesting that different strains may have different HBGAs specificities and may recognize distinct subgroups of the population. The vaccine strains contain a VP8* of the P[8] type and may not infect FUT2- or FUT3-null children. Since the frequencies of HBGAs polymorphisms greatly vary across geographical locations such refractory individuals may represent up to 50% of the population in some areas. We hypothesize that the lack of vaccine efficacy in many children of those areas could be due to an absence of HBGA ligand and that due to the higher diversity of circulating virus strains in tropical areas, the vaccine refractory children may remain susceptible to strains other than those of the common P[8] type.
The aim of our project is to document this hypothesis through the combination of a prospective genetic study, analyses of the glycan specificity of the VP8* from clinical and vaccine strains and of in vitro experiments aimed at characterizing the role of HBGAs binding in the infection process. The prospective study will be carried both on a western European population in Nantes and a population from a tropical area at Cayenne (French Guyana). RVA infected children admitted at the pediatric emergency units at both locations will be enrolled alongside non-infected control groups. Virus genotypes as well as patients and controls HBGAs subtypes will be determined in order to uncover a relationship between them. Glycan specificity of the VP8* of strains of diverse P-genotypes will be determined using glycan microarrays and saliva mucins of diverse HBGA phenotypes in order to control for a match with the infection data. HBGA involvement in the infection process will be assessed using cultivable strains through manipulation of the glycan expression of susceptible cells lines under conditions mimicking in vivo infection.