Human Rotaviruses of Multiple Genotypes Acquire Conserved VP4 Mutations during Serial Passage

Abstract

Human rotaviruses exhibit limited tropism and replicate poorly in most cell lines. Attachment protein VP4 is a key rotavirus tropism determinant. Previous studies in which human rotaviruses were adapted to cultured cells identified mutations in VP4. However, most such studies were conducted using only a single human rotavirus genotype. In the current study, we serially passaged 50 human rotavirus clinical specimens representing five of the genotypes most frequently associated with severe human disease, each in triplicate, three to five times in primary monkey kidney cells then ten times in the MA104 monkey kidney cell line. From 13 of the 50 specimens, we obtained 25 rotavirus antigen-positive lineages representing all five genotypes, which tended to replicate more efficiently in MA104 cells at late versus early passage. We used Illumina next-generation sequencing and analysis to identify variants that arose during passage. In VP4, variants encoded 28 mutations that were conserved for all P[8] rotaviruses and 12 mutations that were conserved for all five genotypes. These findings suggest there may be a conserved mechanism of human rotavirus adaptation to MA104 cells. In the future, such a conserved adaptation mechanism could be exploited to study human rotavirus biology or efficiently manufacture vaccines.

Keywords: VP4; culture adaptation; genotype; rotavirus; serial passage.

Figure 1

Workflow for rotavirus serial passaging. Each human rotavirus clinical specimen was serially passaged in triplicate lineages in primary RhMK cells in roller tubes three to five times. Most rotavirus-positive lineage lysates from RhMK passage were serially passaged ten times in monkey kidney epithelial MA104 cells in tissue culture flasks. Numbers of input lineages and specimens in each series and of rotavirus positive lineages and specimens at the end of each series are indicated.

Figure 2

Replication of early- and late-passage rotaviruses in MA104 cells. MA104 cells were adsorbed for 1 h at 37 °C with trypsin-activated, rotavirus-positive P3 or P10 lysates at an MOI of 0.01 FFU/cell or with undiluted lysate if an MOI of 0.01 FFU/cell could not be reached. Cells were washed to remove unbound virus and incubated with serum-free medium plus 0.5 µg/mL trypsin at 37 °C for 0, 24, or 48 h. Plates were frozen at −80 °C and thawed three times prior to determining virus titer by FFA on MA104 cells. The mean and individual data points are shown. Virus yield was determined by dividing titer at 24 h or 48 h by titer at 0 h. *, p < 0.05; **, p < 0.01; ****, p < 0.0001 by two-way ANOVA with Šídák’s multiple comparisons.

Figure 3

Locations of conserved VP4 amino acid mutations. VP4 of human G1P[8] vaccine candidate CDC-9 is shown in an upright ((A–C); PDB ID 7UMS) or reversed ((D); PDB ID 7UMT) conformation [51]. In (C,D), VP4 is shown relative to VP7 (yellow) and VP6 (green) rotavirus capsid layers. Monomers of trimeric VP4 are shown as red and orange ribbons, with hydrophobic loops colored green, or as red and orange ovals when the structure is unresolved. Locations of VP4 polymorphisms detected following serial passage in MA104 cells that are conserved for P[8] rotaviruses (A) or across all tested genotypes (B–D) are shown as spheres and colored cyan.

Figure 4

Locations and potential interactions of conserved VP4 amino acid mutations. VP4 of human G1P[8] vaccine candidate CDC-9 is shown in an upright conformation relative to VP7 (yellow) and VP6 (green) rotavirus capsid layers ((A); PDB ID 7UMS) [51]. Areas that are magnified in panels (B–D) are indicated. Monomers of trimeric VP4 are shown as red and orange ribbons, with hydrophobic loops colored green. Locations of VP4 polymorphisms detected following serial passage in MA104 cells that are conserved across all tested genotypes are shown as cyan spheres in (A) or as ball-and-stick representations in (B–E). In (E), the VP4 head domain of CDC-9 (dark red) onto which locations of conserved VP4 mutations from this study have been mapped, has been aligned with a human P[8] VP8* domain (light gray) bound to the secretory H type-1 antigen (green) (PDB ID 6HA0) [66].