Search

Retinoic acid inducible gene I (RIG-I)-like receptors (RLRs) function as cytoplasmic sensors for viral RNA to initiate antiviral responses including type I interferon (IFN) production. It has been unclear how RIG-I encounters and senses viral RNA. To address this issue, we examined intracellular localization of RIG-I in response to viral infection using newly generated anti-RIG-I antibody. Immunohistochemical analysis revealed that RLRs localized in virus-induced granules containing stress granule (SG) markers together with viral RNA and antiviral proteins. Because of similarity in morphology and components, we termed these aggregates antiviral stress granules (avSGs). Influenza A virus (IAV) deficient in non-structural protein 1 (NS1) efficiently generated avSGs as well as IFN, however IAV encoding NS1 produced little. Inhibition of avSGs formation by removal of either the SG component or double-stranded RNA (dsRNA)-dependent protein kinase (PKR) resulted in diminished IFN production and concomitant enhancement of viral replication. Furthermore, we observed that transfection of dsRNA resulted in IFN production in an avSGs-dependent manner. These results strongly suggest that the avSG is the locus for non-self RNA sensing and the orchestration of multiple proteins is critical in the triggering of antiviral responses.

BACKGROUND:Despite being expensive, the standard combination of pegylated interferon (Peg-IFN)-α and ribavirin used to treat chronic hepatitis C (CH) results in a moderate clearance rate and a plethora of side effects. This makes it necessary to predict patient outcome so as to improve the accuracy of treatment. Although the antiviral mechanism of genetically altered IL28B is unknown, IL28B polymorphism is considered a good predictor of IFN combination treatment outcome. METHODOLOGY:Using microarray, we quantified the expression profile of 237 IFN related genes in 87 CH liver biopsy specimens to clarify the relationship between IFN pathway and viral elimination, and to predict patients' clinical outcome. In 72 out of 87 patients we also analyzed IL28B polymorphism (rs8099917). PRINCIPAL FINDINGS:Five IFN related-genes (IFI27, IFI 44, ISG15, MX1, and OAS1) had expression levels significantly higher in nonresponders (NR) than in normal liver (NL) and sustained virological responders (SVR); this high expression was also frequently seen in cases with the minor (TG or GG) IL28B genotype. The expression pattern of 31 IFN related-genes also differed significantly between NR and NL. We predicted drug response in NR with 86.1% accuracy by diagonal linear discriminant analysis (DLDA). CONCLUSION:IFN system dysregulation before treatment was associated with poor IFN therapy response. Determining IFN related-gene expression pattern based on patients' response to combination therapy, allowed us to predict drug response with high accuracy. This method can be applied to establishing novel antiviral therapies and strategies for patients using a more individual approach.