The Hepatitis C Virus-Induced Membranous Web and Associated Nuclear Transport Machinery Limit Access of Pattern Recognition Receptors to Viral Replication Sites

Neufeldt, Christopher J.; Joyce, Michael A.; Van Buuren, Nicholas; Levin, Aviad; Kirkegaard, Karla; Gale Jr., Michael; Tyrrell, D. Lorne J.; Wozniak, Richard W.

The results presented above imply that HCV infection induces the formation of the MW to sequester viral processes from RLRs found in the surrounding regions of the cytoplasm, a condition that would limit the innate immune response. In positive-strand RNA virus infection, On day 4 after infection, the localization of HCV proteins and RLRs in uninfected and HCV-infected cells was evaluated by indirect immunofluorescence confocal microscopy using antibodies specific for the indicated HCV protein or double-strand viral RNA and either the V5 or FLAG epitope tag. DNA was detected with DAPI (blue) and scale bars represent 5 μm. Pearson's correlation coefficients shown in the merge images of HCV-infected cells in panels A and B were calculated using Coloc2 software in ImageJ and represent overlap between the red and green fluorescent channels in the indicated image. viral RNA is involved is several different processes including genome replication, translation, and virion assembly. To better understand the spatial relationship of these processes relative to the MW and RLRs, we used fluorescence microscopy to examine the localization of viral RNA and viral proteins, and we compared these to ectopically expressed RLRs. The localization of viral RNAs was detected using branched DNA probes directed against either the positivestrand or negative-strand of the HCV genome [46] [47] [48] . This methodology has previously been used for single molecule detection of host cell mRNA transcripts, as well as for detection of viral RNAs [48] [49] [50] [51] . For our studies, cells harbouring the FLAG-tagged RIG-I-K270A construct were co-stained with antibodies directed against the FLAG epitope and either HCV core or NS5A. The cells were then examined by in situ hybridization using positive-strand or negative-strand RNA probes (Fig 3 and S2 Fig) . Manders Overlap Coefficients were used to determine the percent overlap of positive-or negative-strand HCV RNA with immunofluorescence signals produced by antibodies against NS5A, RIG-I-K270A, or core. In all cases, percent overlap is calculated from~15 cells. In cells analyzed for the localization of NS5A, RIG-I-K270A, and negative-strand HCV RNA, we observed that the majority of the negative-strand viral RNA (82% of the fluorescence signal) overlapped with NS5A signal. Moreover, regions containing negative-strand RNA lacked RIG-I, as only 16% fluorescence signal from the negativestrand probe overlapped with the RIG-I signal ( Fig 3A) . We infer from these data that RIG-I is present in regions of the cytoplasm largely distinct from negative-strand viral RNA containing replication centers within the MW. Similarly, in HCV-infected cells co-stained for core protein, RIG-I-K270A, and positive-strand RNA, both positive-strand RNA and core protein were predominantly localized to the same MW regions, with 55% of the positive-strand RNA fluorescence signal overlapping with the core protein signal. By contrast, only 28% of positive-strand RNA signal overlapped with the RIG-I signal ( Fig 3B) . The presence of a less abundant pool of positive-strand RNA (28% of the fluorescent signal) in regions that overlapped with the RIG-I signal implies that although the majority of positive-strand HCV RNA is present within MW compartments that lack RIG-I, another pool exists outside these regions, which is consistent with the need for translation of viral RNA.