As mentioned above dsRNA is a major marker of viral infection. Consequently, the detection of dsRNA is a critical function of innate immunity. The function of TLR3 may, however, be somewhat limited by its subcellular localisation. It may be simplest to consider TLR3 (like TLR7, 8 and 9) as having a role in sensing viruses that achieve entry into the host cell by endocytosis and so may come in contact with TLR3 in endosomal compartments. Naturally, this is not true for all viruses. A large number of viruses replicate in the cytoplasm of host cells and so generate vast quantities of dsRNA in the cytoplasm during their replicative cycle. Intuitively, other intracellular sensors of viral dsRNA are being identified and it is becoming clear that their importance in initiating an antiviral response is paramount.
One such receptor is retinoic acid inducible gene I (RIG-I), a DExD/H box RNA helicase molecule containing tandem caspase recruitment domains (CARD). Full-length RIG-I protein and a truncated form, comprising the CARD motifs alone (denoted ARIG-I), hugely enhanced transcription from an IRF regulatory sequence and the native IFN-P promoter in response to poly(I:C) transfection and infection with Newcastle disease virus (NDV) . This study elegantly demonstrated the antiviral action of RIG-I using siRNA to specifically knock-down RIG-I expression. While NDV infection was associated with a clear induction of IRF-3 dimerisation in cells stably transfected with a control siRNA sequence, targeted siRNA knockdown of RIG-I expression specifically inhibited this activation. Furthermore, IFN-a, IFN-P, IP10 and RANTES gene induction in response to NDV infection was completely blocked in cells where RIG-I protein expression was impaired. In this study, experiments showing that full-length RIG-I could substantially impede the formation of any visible cytopathic effect on infection with both vesicular stomatitis virus (VSV) and encephalomyocarditis virus (EMCV) emphatically underlie the importance of RIG-I in controlling virus infection. Thus, RIG-I would appear to play a central role in establishing an antiviral state and preventing the spread of invading viruses.
The RNA-binding activity of RIG-I has been mapped to its helicase domain, which has intrinsic ATPase activity. Critically, the RIG-I helicase domain bound the HCV 5' and 3' nontranslated regions . Thus, the interaction of RIG-I with these highly structured regions of the HCV genome represents an affinity of RIG-I for a natural dsRNA-like viral PAMP. Unstructured single-stranded RNA from HCV was not recognised by RIG-I, confirming the specificity of RIG-I for dsRNA. Melanoma differentiation associated gene-5 (Mda-5), a closely related CARD domain containing DExD/H box RNA helicase, similarly enhanced IRF-3 activation in response to both poly(I:C) and NDV infection . Interestingly, V proteins encoded by a range of paramyxoviruses, including SeV, human parainfluenza virus 5 and mumps virus, could antagonise the induction of type I IFNs by mda-5 .
As mentioned above PKR is a long-standing intracellular receptor for dsRNA. Direct binding of dsRNA induces PKR dimerisation and the resulting activation of PKR via auto-phosphorylation allows this kinase to catalyse the phosphorylation of its primary target, eukaryotic initiation factor (eIF) 2a . PKR is central to the induction of an antiviral state by type I IFNs, blocking the translation of viral mRNAs and hence, preventing de novo synthesis of viral proteins. PKR can also signal to activate NF-kB and MAP kinases [44, 45] but does not appear to regulate IRF-3 activation . Furthermore, signalling through TLR3, TLR4 and TLR9 appears to induce autophosphorylation of PKR [47, 48]. Thus, the antiviral functions of PKR are many fold.
In summary, although TLR3 does appear to function as an antiviral innate immune receptor and can mount a strong inflammatory response on detection of viral dsRNA, it may be that other cytoplasmic pattern recognition receptors, such as RIG-I, are more suitably located to act as major sensors of dsRNA released into the cytoplasm during viral infection. The cell type-specific subcellular localisation of TLR3 might reflect a divergence in TLR3 function in different tissue types. Human fibroblasts and epithelial cells display both a cell surface and an intracellular pattern of TLR3 expression while monocyte-derived immature dendritic cells (iDCs) and CD11c+ blood DCs express TLR3 solely in intracellular vesicles . Thus, it may be that, where TLR3 is exposed at the cell surface, it can serve to detect the presence of dsRNA of both cellular and viral origin released upon lysis of virally infected cells. This is supported by the fact that anti-TLR3 antibodies could block activation of fibroblasts in response to poly(I:C) treatment , indicating that this is a cell surface event. The observation that TLR3-mediated IFN induction by DCs was inhibited by chloroquine or bafilomycin strongly implies that this signalling cascade is initiated in endosomal compartments . While endosome-localised TLR3 may serve to alert host innate immunity to invading endocytic viruses, a growing body of evidence suggests a major role for TLR3 expressed within the endosomal compartments of DCs in activating the adaptive cellular immune response to viral infection.
Was this article helpful?