Cellular mRNA of higher eukaryotes and many viral RNA are methylated at the N-7 and 2′-O positions of the 5′ guanosine cap by specific nuclear and cytoplasmic methyltransferases (MTases), respectively. Whereas N-7 methylation is essential for RNA translation and stability 1, the function of 2′-O methylation has remained uncertain since its discovery 35 years ago 2-4. Here, we show that a West Nile virus (WNV) mutant (E218A) that lacks 2′-O MTase activity was attenuated in wild type primary cells and mice but was pathogenic in the absence of type I interferon (IFN) signaling. 2′-O methylation of viral RNA did not affect IFN induction in WNV-infected fibroblasts but instead modulated the antiviral effects of IFN-induced proteins with tetratricopeptide repeats (IFIT), which are interferon-stimulated genes (ISG) implicated in regulation of protein translation. Poxvirus and coronavirus mutants that lacked 2′-O MTase activity similarly showed enhanced sensitivity to the antiviral actions of IFN and specifically, IFIT proteins. Our results demonstrate that the 2′-O methylation of the 5′ cap of viral RNA functions to subvert innate host antiviral responses through escape of IFIT-mediated suppression, and suggest an evolutionary explanation for 2′-O methylation of cellular mRNA: to distinguish self from non-self RNA. Differential methylation of cytoplasmic RNA likely serves as a paradigm for pattern recognition and restriction of propagation of foreign viral RNA in host cells.

The innate immune response is essential for controlling West Nile virus (WNV) infection but how this response is propagated and regulates adaptive immunity in vivo are not defined. Herein, we show that IPS-1, the central adaptor protein to RIG-I-like receptor (RLR) signaling, is essential for triggering of innate immunity and for effective development and regulation of adaptive immunity against pathogenic WNV. IPS-1−/− mice exhibited increased susceptibility to WNV infection marked by enhanced viral replication and dissemination with early viral entry into the CNS. Infection of cultured bone-marrow (BM) derived dendritic cells (DCs), macrophages (Macs), and primary cortical neurons showed that the IPS-1-dependent RLR signaling was essential for triggering IFN defenses and controlling virus replication in these key target cells of infection. Intriguingly, infected IPS-1−/− mice displayed uncontrolled inflammation that included elevated systemic type I IFN, proinflammatory cytokine and chemokine responses, increased numbers of inflammatory DCs, enhanced humoral responses marked by complete loss of virus neutralization activity, and increased numbers of virus-specific CD8+ T cells and non-specific immune cell proliferation in the periphery and in the CNS. This uncontrolled inflammatory response was associated with a lack of regulatory T cell expansion that normally occurs during acute WNV infection. Thus, the enhanced inflammatory response in the absence of IPS-1 was coupled with a failure to protect against WNV infection. Our data define an innate/adaptive immune interface mediated through IPS-1-dependent RLR signaling that regulates the quantity, quality, and balance of the immune response to WNV infection.

Protection against West Nile virus (WNV) infection requires rapid viralRecent studies have begun to delineate the molecular mechanisms that regulate alpha interferon (IFN-␣) and IFN-␤ induction after infection by RNA and DNA viruses (reviewed in references 13, 24, 25, and 57). Pattern recognition receptors (PRR) sense conserved structural microbial elements identified as pathogen-associated molecular patterns. PRR involved in the recognition of RNA viruses can be divided into two classes. Toll-like receptors (TLR) on the cell surface or within endosomes recognize single-and double-stranded RNA and signal through the adaptor molecules MyD88 and TRIF. In comparison, RIG-I and MDA5 helicases recognize single-and double-stranded RNA in the cytosol and signal through the adaptor protein IPS-1 (also known as MAVS, Cardif, and VISA) (reviewed in references 28, 42, and 55). Recognition by viral RNA sensors results in the downstream activation and nuclear translocation of IRF-3 and IRF-7, which transcriptionally activate IFN-␣ and -␤ gene promoters (46).Cell culture experiments defined TLR3 as a PRR that recognizes double-stranded RNA, activates IRF-3 and NF-B transcriptional pathways, and induces type I IFN (2). The role of TLR3 in vivo in protection against viral infections has been less clear (reviewed in references 5, 36, and 58). Experiments in TLR3Ϫ/Ϫ mice infected with lymphocytic choriomeningitis virus, vesicular stomatitis virus, and reovirus failed to show increased mortality or altered viral burden phenotypes (12). In contrast, TLR3 restricts replication, regulates cytokine production, and protects against infection by mouse cytomegalovirus and encephalomyocarditis virus (EMCV) (21, 54). Indeed, a deficiency in TLR3 in humans was identified as a predisposing genetic risk factor for herpes simplex virus (HSV) encephalitis (61) and influenza A virus-induced encephalopathy (23). An adverse role for TLR3 also has been proposed since TLR3 Ϫ/Ϫ mice infected with influenza, punta toro, and vaccinia viruses showed improved survival and decreased production of inflammatory cytokines (19,26,33). Analogously, conflicting results have been observed with respect to the role of TLR3 in inducing IFN after infection with the encephalitic flavivirus West Nile virus (WNV). TLR3 was largely dispensable for WNV recognition and induction of IFN responses in vitro (15, 47), whereas in mice, an absence of TLR3 protected mice from lethal infection (59). TLR3 Ϫ/Ϫ mice infected via intraperitoneal injection with WNV showed decreased systemic tumor necrosis factor alpha (TNF-␣) and interleukin-6 (IL-6) production, blood-brain barrier (BBB) permeability, and infection in the brain.Given this apparent conflict and our previous studies demonstration of an essential role for IRF-3 in controlling WNV infection (6), we reexamined the pathogenesis of virulent WNV infection in TLR3 Ϫ/Ϫ mice. Remarkably, we observed increased susceptibility of TLR3 Ϫ/Ϫ mice to WNV infection. TLR3 had a modest effect on WNV infection in peripheral tissues and was...

Interferon regulatory factor (IRF)-3 is a master transcription factor that activates host antiviral defense programs. Although cell culture studies suggest that IRF-3 promotes antiviral control by inducing interferon (IFN)-β, near normal levels of IFN-α and IFN-β were observed in IRF-3−/− mice after infection by several RNA and DNA viruses. Thus, the specific mechanisms by which IRF-3 modulates viral infection remain controversial. Some of this disparity could reflect direct IRF-3-dependent antiviral responses in specific cell types to control infection. To address this and determine how IRF-3 coordinates an antiviral response, we infected IRF-3−/− mice and two primary cells relevant for West Nile virus (WNV) pathogenesis, macrophages and cortical neurons. IRF-3−/− mice were uniformly vulnerable to infection and developed elevated WNV burdens in peripheral and central nervous system tissues, though peripheral IFN responses were largely normal. Whereas wild-type macrophages basally expressed key host defense molecules, including RIG-I, MDA5, ISG54, and ISG56, and restricted WNV infection, IRF-3−/− macrophages lacked basal expression of these host defense genes and supported increased WNV infection and IFN-α and IFN-β production. In contrast, wild-type cortical neurons were highly permissive to WNV and did not basally express RIG-I, MDA5, ISG54, and ISG56. IRF-3−/− neurons lacked induction of host defense genes and had blunted IFN-α and IFN-β production, yet exhibited only modestly increased viral titers. Collectively, our data suggest that cell-specific IRF-3 responses protect against WNV infection through both IFN-dependent and -independent programs.

PCR screening of 1,482 murid rodents from 13 genera caught in 18 different localities of Guinea, West Africa, showed Lassa virus infection only in molecularly typed Mastomys natalensis. Distribution of this rodent and relative abundance compared with M. erythroleucus correlates geographically with Lassa virus seroprevalence in humans.