Viperin has antiviral function against many viruses, including dengue virus (DENV), when studied in cells in culture. Here, the antiviral actions of viperin were defined both in vitro and in a mouse in vivo model of DENV infection. Murine embryonic fibroblasts (MEFs) derived from mice lacking viperin (vip−/−) showed enhanced DENV infection, accompanied by increased IFN-β and induction of ISGs; IFIT1 and CXCL-10 but not IRF7, when compared to wild-type (WT) MEFs. In contrast, subcutaneous challenge of immunocompetent WT and vip−/− mice with DENV did not result in enhanced infection. Intracranial infection with DENV resulted in body weight loss and neurological disease with a moderate increase in mortality in vip−/− compared with WT mice, although this was not accompanied by altered brain morphology, immune cell infiltration or DENV RNA level in the brain. Similarly, DENV induction of IFN-β, IFIT1, CXCL-10, IRF7 and TNF-α was not significantly different in WT and vip−/− mouse brain, although there was a modest but significant increase in DENV induction of IL-6 and IfI27la in the absence of viperin. NanoString nCounter analysis confirmed no significant difference in induction of a panel of inflammatory genes in WT compared to vip−/− DENV-infected mouse brains. Further, polyI:C stimulation of bone marrow-derived macrophages (BMDMs) induced TNF-α, IFN-β, IL-6 and Nos-2, but responses were not different in BMDMs generated from WT or vip−/− mice. Thus, while there is significant evidence of anti-DENV actions of viperin in some cell types in vitro, for DENV infection in vivo a lack of viperin does not affect systemic or brain susceptibility to DENV or induction of innate and inflammatory responses.

Zika virus (ZIKV) infection causes ocular and neurological pathologies with ZIKV-induction of developmental abnormalities following in utero infection a major concern. The study here has compared ZIKV and the related dengue virus (DENV) infection in the eye and brain. In vitro, both ZIKV and DENV could infect cell lines representing the retinal pigmented epithelium, endothelial cells, and Mueller cells, with distinct innate responses in each cell type. In a 1-day old mouse challenge model, both ZIKV and DENV infected the brain and eye by day 6 post-infection (pi). ZIKV was present at comparable levels in both tissues, with RNA increasing with time post-infection. DENV infected the brain, but RNA was detected in the eye of less than half of the mice challenged. NanoString analysis demonstrated comparable host responses in the brain for both viruses, including induction of mRNA for myosin light chain-2 (Mly2), and numerous antiviral and inflammatory genes. Notably, mRNA for multiple complement proteins were induced, but C2 and C4a were uniquely induced by ZIKV but not DENV. Consistent with the viral infection in the eye, DENV induced few responses while ZIKV induced substantial inflammatory and antiviral responses. Compared to the brain, ZIKV in the eye did not induce mRNAs such as C3, downregulated Retnla, and upregulated CSF-1. Morphologically, the ZIKV-infected retina demonstrated reduced formation of specific retinal layers. Thus, although ZIKV and DENV can both infect the eye and brain, there are distinct differences in host cell and tissue inflammatory responses that may be relevant to ZIKV replication and disease.