In most RNA viruses, genome replication and transcription are catalysed by a viral RNA-dependent RNA polymerase. Double-stranded RNA viruses perform these operations in a capsid (the polymerase complex), using an enzyme that can read both single- and double-stranded RNA. Structures have been solved for such viral capsids, but they do not resolve the polymerase subunits in any detail. Here we show that the 2 A resolution X-ray structure of the active polymerase subunit from the double-stranded RNA bacteriophage straight phi6 is highly similar to that of the polymerase of hepatitis C virus, providing an evolutionary link between double-stranded RNA viruses and flaviviruses. By crystal soaking and co-crystallization, we determined a number of other structures, including complexes with oligonucleotide and/or nucleoside triphosphates (NTPs), that suggest a mechanism by which the incoming double-stranded RNA is opened up to feed the template through to the active site, while the substrates enter by another route. The template strand initially overshoots, locking into a specificity pocket, and then, in the presence of cognate NTPs, reverses to form the initiation complex; this process engages two NTPs, one of which acts with the carboxy-terminal domain of the protein to prime the reaction. Our results provide a working model for the initiation of replication and transcription.

SummaryRNA interference pathways use small RNAs to mediate gene silencing in eukaryotes. In addition to small interfering RNAs (siRNA) and microRNAs, several types of endogenously produced small RNAs play important roles in gene regulation, germ cell maintenance and transposon silencing 1 -4. Production of some of these RNAs requires the synthesis of aberrant RNAs (aRNAs) or pre-siRNAs, which are specifically recognized by RNA-dependent RNA polymerases (RdRPs) to make double stranded RNA (dsRNA). The mechanism for aRNA synthesis and recognition is largely unknown. Here we show that DNA damage induces the expression of the Argonaute protein QDE-2 and a novel class of small RNAs in the filamentous fungus Neurospora. This class of small RNAs, named qiRNAs for their association with QDE-2, are about 20-21 nt long (several nt shorter than Neurospora siRNAs) with a strong preference for uridine at the 5′ end and originate mostly from the ribosomal DNA locus. Production of qiRNAs requires the RdRP QDE-1, the Werner/Bloom RecQ DNA helicase homolog QDE-3 and dicers. qiRNA biogenesis also requires DNA damage-induced aRNAs as precursor, a process that is dependent on QDE-1 and QDE-3. Surprisingly, our results suggest that QDE-1 is the DNA-dependent RNA polymerase that produces aRNAs. In addition, the Neurospora RNAi mutants exhibit increased sensitivity to DNA damage, suggesting a role for qiRNAs in DNA damage response by inhibiting protein translation.In the filamentous fungus Neurospora crassa, the RNAi pathway is essential for both dsRNA and transgene-induced gene silencing (quelling) 5 . In the quelling pathway, QDE-1 and QDE-3 are thought to be involved in the generation of dsRNA 6 , 7. In addition, QDE-3 was previously shown to be involved in DNA repair 6. It has been proposed that a repetitive transgene leads to the production of transgene-specific aRNA, which is converted to dsRNA

Is it possible to meaningfully comprehend the diversity of the viral world? We propose that it is. This is based on the observation that, although there is immense genomic variation, every infective virion is restricted by strict constraints in structure space (i.e., there are a limited number of ways to fold a protein chain, and only a small subset of these have the potential to construct a virion, the hallmark of a virus). We have previously suggested the use of structure for the higher-order classification of viruses, where genomic similarities are no longer observable. Here, we summarize the arguments behind this proposal, describe the current status of structural work, highlighting its power to infer common ancestry, and discuss the limitations and obstacles ahead of us. We also reflect on the future opportunities for a more concerted effort to provide high-throughput methods to facilitate the large-scale sampling of the virosphere.