Hepatitis B virus (HBV) is a major global health concern, and the development of curative therapeutics is urgently needed. Such efforts are impeded by the lack of a physiologically relevant, pre-clinical animal model of HBV infection. Here, we report that expression of the HBV entry receptor, human sodium-taurocholate cotransporting polypeptide (hNTCP), on macaque primary hepatocytes facilitates HBV infection in vitro, where all replicative intermediates including covalently closed circular DNA (cccDNA) are present. Furthermore, viral vector-mediated expression of hNTCP on hepatocytes in vivo renders rhesus macaques permissive to HBV infection. These in vivo macaque HBV infections are characterized by longitudinal HBV DNA in serum, and detection of HBV DNA, RNA, and HBV core antigen (HBcAg) in hepatocytes. Together, these results show that expressing hNTCP on macaque hepatocytes renders them susceptible to HBV infection, thereby establishing a physiologically relevant model of HBV infection to study immune clearance and test therapeutic and curative approaches.

Cytomegaloviruses (CMV) are highly species-specific due to millennia of co-evolution and adaptation to their host, with no successful experimental cross-species infection in primates reported to date. Accordingly, full genome phylogenetic analysis of multiple new CMV field isolates derived from two closely related nonhuman primate species, Indian-origin rhesus macaques (RM) and Mauritian-origin cynomolgus macaques (MCM), revealed distinct and tight lineage clustering according to the species of origin, with MCM CMV isolates mirroring the limited genetic diversity of their primate host that underwent a population bottleneck 400 years ago. Despite the ability of Rhesus CMV (RhCMV) laboratory strain 68–1 to replicate efficiently in MCM fibroblasts and potently inhibit antigen presentation to MCM T cells in vitro, RhCMV 68–1 failed to productively infect MCM in vivo, even in the absence of host CD8+ T and NK cells. In contrast, RhCMV clone 68–1.2, genetically repaired to express the homologues of the HCMV anti-apoptosis gene UL36 and epithelial cell tropism genes UL128 and UL130 absent in 68–1, efficiently infected MCM as evidenced by the induction of transgene-specific T cells and virus shedding. Recombinant variants of RhCMV 68–1 and 68–1.2 revealed that expression of either UL36 or UL128 together with UL130 enabled productive MCM infection, indicating that multiple layers of cross-species restriction operate even between closely related hosts. Cumulatively, these results implicate cell tropism and evasion of apoptosis as critical determinants of CMV transmission across primate species barriers, and extend the macaque model of human CMV infection and immunology to MCM, a nonhuman primate species with uniquely simplified host immunogenetics.

DNA-interstrand cross-links (ICLs) can be repaired by biochemical pathways requiring DNA polymerases that are capable of translesion DNA synthesis (TLS). The anticipated function of TLS polymerases in these pathways is to insert nucleotides opposite and beyond the linkage site. The outcome of these reactions can be either error-free or mutagenic. TLS-dependent repair of ICLs formed between the exocyclic nitrogens of deoxyguanosines (N2-dG) can result in low-frequency base substitutions, predominantly G to T transversions. Previously, we demonstrated in vitro that error-free bypass of a model acrolein-mediated N2-dG ICL can be accomplished by human polymerase (pol) κ, while Rev1 can contribute to this bypass by inserting dC opposite the cross-linked dG. The current study characterized two additional human DNA polymerases, pol η and pol ι, with respect to their potential contributions to either error-free or mutagenic bypass of these lesions. In the presence of individual dNTPs, pol η could insert dA, dG, and dT opposite the cross-linked dG, but incorporation of dC was not apparent. Further primer extension was observed only from the dC and dG 3'; termini and the amounts of products were low relative to the matched undamaged substrate. Analyses of bypass products beyond the adducted site revealed that dG was present opposite the cross-linked dG in the majority of extended primers and short deletions were frequently detected. When pol ι was tested for its ability to replicate past this ICL, the correct dC was preferentially incorporated, but no further extension was observed. Under the steady-state conditions, the efficiency of dC incorporation was reduced ∼500-fold relative to the undamaged dG. Thus, in addition to pol κ-catalyzed error-free bypass of N2-dG ICLs, an alternative, albeit low-efficiency, mechanism may exist. In this pathway, either Rev1 or pol ι could insert dC opposite the lesion, while pol η could perform the subsequent extension.