Abstract
Rotavirus (RV) is the leading cause of acute gastroenteritis (AGE) in children under 5 years old worldwide, and several studies have demonstrated that histo-blood group antigens (HBGAs) play a role in its infection process. In the present study, human stool filtrates from patients diagnosed with RV diarrhea (genotyped as P[8]) were used to infect differentiated Caco-2 cells (dCaco-2) to determine whether such viral strains of clinical origin had the ability to replicate in cell cultures displaying HBGAs. The cell culture-adapted human RV Wa model strain (P[8] genotype) was used as a control. A time-course analysis of infection was conducted in dCaco-2 at 1, 24, 48, 72, and 96 h. The replication of two selected clinical isolates and Wa was further assayed in MA104, undifferentiated Caco-2 (uCaco-2), HT29, and HT29-M6 cells, as well as in monolayers of differentiated human intestinal enteroids (HIEs). The results showed that the culture-adapted Wa strain replicated more efficiently in MA104 cells than other utilized cell types. In contrast, clinical virus isolates replicated more efficiently in dCaco-2 cells and HIEs. Furthermore, through surface plasmon resonance analysis of the interaction between the RV spike protein (VP8*) and its glycan receptor (the H antigen), the V7 RV clinical isolate showed 45 times better affinity compared to VP8* from the Wa strain. These findings support the hypothesis that the differences in virus tropism between clinical virus isolates and RV Wa could be a consequence of the different HBGA contents on the surface of the cell lines employed. dCaco-2, HT29, and HT29M6 cells and HIEs display HBGAs on their surfaces, whereas MA104 and uCaco-2 cells do not. These results indicate the relevance of using non-cell culture-adapted human RV to investigate the replication of rotavirus in relevant infection models.