The discovery of RNA interference (RNAi) in 1998 was one of the most unexpected findings in the last 30 years. The Nobel-winning observations were made in C. elegans, where RNAi is particularly potent. Two aspects of RNAi in C. elegans contribute to this potency; the double-strand (ds)RNA triggers are mobile, spreading the silencing to cells throughout the organisms, and once triggered, RNAi is self-amplifying, resulting in robust silencing. I was particularly intrigued by dsRNA mobility, an observation that lacked a ready mechanistic explanation. To investigate this, two graduate students in my lab, Bill Winston and Christina Molodowitch, isolated mutants that specifically disrupted dsRNA mobility (Science, PDF). This screen identified SID-1, which encodes a large transmembrane protein that we subsequently showed acts as a dsRNA channel essential for the import of silencing signals (Science, PDF). SID-1 homologs are present in many animals and all vertebrates. The human and mouse Sidt1 and Sidt2 proteins are remarkably well conserved, suggesting a highly selected function. To investigate, we obtained Sidt1 and Sidt2 knock-out mice.
In this month’s Immunity, PDF, we report that Sidt2, like SID-1, imports extracellular dsRNA into cells. But rather than triggering an RNAi response, the imported dsRNA triggers an interferon-mediated antiviral response. It has been known for decades that extracellular dsRNA triggers a potent anti-viral response, but inexplicably, the dsRNA receptors for this response, RIG-I and MDA-5, are cytoplasmic proteins. How extracellular dsRNA is delivered to the cytoplasm had remained mysterious. Ken Pang, a former postdoc in my lab, and I, together with his colleagues at the Walter and Eliza Hall Institute of Medical Research in Melbourne Australia, have shown that Sidt2 is localized in endo-lysosomes where it interacts with internalized dsRNA. When Sidt2 is absent, the internalized dsRNA remains trapped in these vesicles, but in wild-type cells it is released into the cytoplasm. The released dsRNA then triggers RIG-I-like receptor (RLR) signaling, including MAVS aggregation, to activate interferon production. Infection of Sidt2 knockout mice with herpes simplex virus (HSV) is lethal, while control mice survive. This likely explains the remarkable conservation of these proteins. Intriguingly this effect is triggered not in virally infected cells, but in their uninfected neighbors. This so-called bystander effect, whereby viral replication intermediates are release by infected cells, bypasses the intricate viral mechanisms that mask the presence of viral dsRNA.
These are very satisfying results, showing conserved activity (dsRNA import) and parallel outputs (RNAi in C. elegans and RLR-signaling in mouse) between these homologous proteins.