(A) Genome browser view of P4R2 RNAs (shades of blue) and 24 nt small interfering RNA (siRNAs) (shades of gray) at a representative locus, an AtSN1 retrotransposon on chromosome 3. Each horizontal bar represents a specific RNA sequence (RNA-seq), with arrows depicting their direction relative to the Arabidopsis reference genome sequence (TAIR10). The intensity of shading reflects the abundance of each RNA species in the RNA-seq dataset. Brightly colored nucleotides, color coded for A, G, C, or U (see inset), represent nucleotides that do not match the corresponding DNA sequence of the locus. The dotted line highlights the coincident 5’ ends of the most abundant P4R2 RNAs at the locus (colored deep purple) and the most abundant siRNAs (colored black). (B) Heat map depicting the frequency of mismatched nucleotides at each position of RNAs ranging in size from 15 to 76 nt in dcl2 dcl3 dcl4 triple mutant plants. To correct for the frequency of errors inherent to sequencing, mismatch values for each position of 15–76 nt RNAs in wild-type plants were subtracted prior to plotting the data. Only read sequences with single mismatches or perfect matches to the reference genome were utilized for this analysis. (C) Over-expression and purification of recombinant RDR2. The image on the left shows a 7.5% sodium dodecyl sulfate-polyacrylamide gel electrophoresis(SDS-PAGE) gel, stained with Coomassie blue, showing molecular weight markers (M), proteins of un-infected High Five cells (lane 1), proteins of High Five cells 72 hr after infection with baculovirus expressing recombinant RNA-dependent RNA polymerase 2 (RDR2) (lane 2), and purified recombinant V5-tagged RDR2 after affinity purification and elution with V5 peptide. The image at right shows anti-RDR2 and anti-V5 immunoblots of the same three protein samples. For RDR2 detection, rabbit anti-RDR2 primary antibody was used in conjunction with donkey anti-rabbit HRP-conjugated secondary antibody. Detection of V5-tagged RDR2 involved anti-V5 HRP conjugate antibody. (D) RDR2 terminal transferase activity. Recombinant RDR2 or an active-site mutant form of RDR2 (RDR2-ASM) was incubated with alpha-labeled 32P-CTP and 51 nt RNA substrates bearing 3’ hydroxyl or 3’ dideoxy termini. Reaction products were subjected to denaturing polyacrylamide gel electrophoresis (PAGE) and autoradiography. For gel lane 4, reaction products were treated with RNase One, which degrades single-stranded RNAs, prior to PAGE. RNA size markers were run in lane M. The 51 nt RNA template, 5’ end-labeled using T4 polynucleotide kinase, was run as a size marker in the lane at far right.