Small RNAs regulate development, stem cell identity, genome integrity, and defense against viruses in animals and plants. These """"""""RNA silencing"""""""" phenomena involve proteins that are conserved from fission yeast to humans: RNA-dependent RNA polymerases (RDRs), which convert single-stranded RNA into double-stranded RNA (dsRNA), Dicer endonucleases that cleave dsRNA into short-interfering RNA (siRNA) duplexes, and Argonaute-containing complexes that use siRNA strands to program silencing effects. In the model plant Arabidopsis, protein duplication and sub-functionalization has resulted in distinct biochemical pathways that generate siRNA classes with different functions. One pathway involving RDR2 and DCL3 produces siRNAs that program repressive histone modifications and DNA methylation at transposable elements and tandem repeats. Another pathway involving RDR6 and DCL4 generates siRNAs that cleave mRNAs encoding proteins that regulate development. Keeping these pathways separate is biologically important, because loading siRNAs of incorrect sequence into downstream complexes would result in anomalous gene silencing. A central, unresolved question is: How are RNA intermediates for specific siRNA classes channeled through these distinct pathways? Answering this question is my overall research goal. Production of siRNAs is likely a highly channeled process from the outset, i.e., precursor RNA synthesis. In fact, known RDR6 substrates derive from RNA polymerase II (Pol II) transcription, whereas repeat- associated siRNA biogenesis requires a plant-specific RNA polymerase (Pol IV) upstream of RDR2. I hypothesize that proteins occupying adjoining steps in each pathway specifically interact to pass RNA intermediates down the chain of enzymatic activities. Mass spectrometry and a yeast two-hybrid system will be used to determine subunit compositions of complexes containing RDR2, RDR6, DCL3 and DCL4;furthermore, I will test RDR interactions by co-immunoprecipitation with candidate proteins. Based on these data, biochemical assays will test whether each RDR or DCL-interacting protein is functionally required for conversion of RNA precursors to products in either siRNA biogenesis pathway. This work will uncover biochemical mechanisms that support divergent, yet parallel pathways of siRNA biogenesis in Arabidopsis. Public Health Relevance: Because siRNAs anneal to and target nucleic acids with exquisite specificity, RNA silencing has potential for gene therapy and antiviral applications. The diversity of RNA silencing systems in Arabidopsis, loss-of-function mutants in key proteins, and ability to rescue mutations using tagged proteins allows us to scrutinize molecular interactions that generate and channel siRNAs.