Host species-specific mutations in the thumb domain of the 3Dpol polymerase are required for efficient replication of human hepatitis A virus in mice

Hepatitis A virus (HAV) is a globally important cause of enterically-transmitted hepatitis. It is one of 9 distinct Hepatovirus species in the Picornaviridae, among which phylogenetic reconstructions suggest multiple past host species jumps. HAV readily infects mice with defective type I interferon responses, suggesting the major barrier preventing human HAV from replicating in a rodent host is an inability to overcome innate immune responses. In prior studies, only a single nonsynonymous mutation of uncertain significance (3Dpol-R468K) was identified within the genome of wild-type HAV following passage in interferon-receptor knockout mice. Here, we show that R468K and other mutations in the 3Dpol polymerase (E461D and D473G) are uniformly present in virus recovered from Ifnar1-/- mice following intrahepatic injection of HAV RNA. Reverse molecular genetics experiments confirmed RNAs with R468K or D473G mutations were more likely to initiate sustained infection than wild-type RNA in mice. In competition experiments using cell culture-adapted virus, a K468 mutant out-replicated wild-type R468 in murine cells, whereas R468 rapidly replaced K468 in human cells. These 3Dpol mutations thus promote HAV replication in a host species-specific manner. AlphaFold 3 modeling indicates E461, R468, and D473 are closely positioned on the surface of the 3Dpol thumb domain, suggesting they modulate interactions with species-specific host factor(s). Proteomics analysis of proteins co-precipitating with HA-3Dpol expressed in Huh-7.5 cells identified heat shock 70 protein HSPA8 and its co-chaperone, BAG2. HSPA8 is known to be a critical hepatovirus host factor and HAV genome replication is highly dependent upon heat shock chaperone activity. The mouse-adaptive R468K mutation enhances co-immunoprecipitation of 3Dpol with murine HSPA8 and BAG2, suggesting it facilitates chaperone-dependent acquisition of polymerase function in mouse cells. Our results identify a non-immune barrier to HAV replication in mice and enable future reverse molecular genetics studies in a small animal model.