Comparison of human intestinal enteroid and zebrafish larva models for replication of human norovirus

Human intestinal enteroid (HIE) and zebrafish larva (ZF) models support the replication of certain strains of human norovirus (HuNoV), the leading cause of viral gastroenteritis worldwide. The replication of 17 HuNoV-positive stool specimens (nine genotypes) from patients ranging from 5 months to 83 years old was evaluated in both HIE and ZF models. The yolk of 3-day post-fertilization ZF larvae was microinjected with 3 nL 10% clarified stool suspension. Each day post-infection (dpi), 10 larvae were pooled as one sample, and two specimens were collected daily until 5 dpi. Viral RNA was extracted from harvested larvae and quantified using reverse transcription (RT) droplet digital PCR. For the HIE model, J4FUT2 K1 enteroid monolayers were inoculated with 100-fold dilution of each stool specimen. Viral RNA levels were quantified at 1 and 72 h post-infection using RT real-time PCR. All genogroup (G) I specimens (n = 5) replicated in both models showing a 1-2.5 log increase in HuNoV RNA. Of the 12 GII specimens tested, 8 replicated in the ZF model, and 7 replicated in the HIE model. Specimens that failed to replicate in each model were not the same, although most of them were GII.4 Sydney [P16]. One additional strain, GII.17[P17], replicated only in the ZF model. Across GII strains, viral RNA increases ranged from 1.6 to 3.3 log in the ZF model and from 0.4 to 3.68 log in the HIE model. These data highlight variability in HuNoV replication efficiency across models, potentially indicating the influence of model-specific factors on replication.IMPORTANCEHuman intestinal enteroid and zebrafish larva models have been independently shown to support replication of different strains of human norovirus (HuNoV), the leading cause of viral gastroenteritis worldwide. However, whether the same HuNoV-positive stool specimens replicate similarly in both systems has not been examined. In this study, stool specimens positive for a range of HuNoV genotypes were tested in both models. While specimens from the genogroup I (GI) genogroup consistently replicated in both systems, specimens from the GII genogroup showed variable replication, with some replicating in only one model and a couple of specimens not replicating in either model. These findings demonstrate that model-specific biological factors influence HuNoV replication and highlight the significance of using complementary model systems to better understand HuNoV pathogenesis.