Biochemical activities of Arabidopsis RNA-dependent RNA polymerase 6

Abstract

In Arabidopsis, genetic evidence demonstrates that RNA-dependent RNA polymerase 6 (RDR6) plays a fundamental role in at least four RNA silencing pathways whose functions range from defense against transgenes or viruses to endogene regulation in development and in stress responses. Despite its critical role in RNA silencing, the biochemical activities of RDR6 have yet to be characterized. In this study, we transiently expressed Arabidopsis RDR6 in Nicotiana benthamiana and investigated the biochemical activities of immunopurified RDR6 in vitro. We showed that RDR6 possesses terminal nucleotidyltransferase activity as well as primer-independent RNA polymerase activity on single-stranded RNAs. We found that RDR6 cannot distinguish RNAs with or without a cap or poly(A) tail. We also demonstrated that RDR6 has strong polymerase activity on single-stranded DNA. All these activities require the conserved catalytic Asp(867) residue. Our findings have important implications on the processes involving RDR6 in vivo and provide new biochemical insights into the mechanisms of RNA silencing in Arabidopsis.

FIGURE 1. Complementation of the rdr6–11 mutation by the 35S::HA-RDR6 transgene

A, 23-day-old Columbia wild type (Col-0), rdr6–11 mutant and two rdr6–11 transgenic lines containing either 35S::HA-RDR6 or the 35S::HA-RDR6m transgene. rdr6–11 mutants showed downward curling of rosette leaves. This phenotype was rescued by 35S::HA-RDR6 but not 35S::HA-RDR6m. B, Northern blot analysis of two ta-siRNAs (ASRP255 and ASRP1511). The analysis was performed with 40 μg of total RNAs extracted from influorescences of wild type (Col-0), rdr6–11 mutants, and six independent 35S::HA-RDR6 rdr6–11 and two independent 35S::HA-RDR6m rdr6–11 transgenic lines. The same membrane was used (after stripping) for the 3 hybridizations. miR173, a microRNA, served as a loading control.

FIGURE 2. RNA polymerase and nucleotidyltransferase activities of purified RDR6

A, schematic representation of ssRNAs used as templates in various RDR assays. A dsDNA fragment (sequence in supplemental Fig. S2A) of a luciferase/nopaline synthase chimeric gene with an engineered miR173 sequence was used as the template for PCR amplification of various segments, which in turn served as templates for in vitro transcription to generate the ssRNA transcripts. The name of each ssRNA referred to the size (in nucleotides) and orientation (F, forward; or R, reverse) compared with the DNA sequence. B and C, reaction products from RDR assays resolved on 6% polyacrylamide, 7 M urea gels and visualized by phosphorimaging. B, assays were performed with immunopurified HA-RDR6 or HA-RDR6m and different quantities of the 185F RNA template (1 to 27 pmol) in the presence of all four cold NTPs and [α-³²P]UTP. The image of the stained gel at the bottom illustrates the amounts of templates used. C, assays were performed with HA-RDR6, 15 pmol of the 246R template, and [α-³²P]UTP, in the presence (+) or absence (-) of the four cold NTPs. After the reaction, the products were treated (+)or not (-) with RNase I for 20 min at 37 °C. The lower panels were images from ethidium bromide (EtBr) staining of the PAGE gels before phosphorimaging. The fact that no RNAs were visible by EtBr staining in lane 9 suggested that the amount of RDR6 products was very small.

FIGURE 3. RDR6 activity on RNA templates with different 5′ and 3′ end structures

All enzymatic assays were performed with HA-RDR6 in the presence of [α-³²P]UTP and four cold NTPs. A, assays were performed with 15 pmol of four 246R templates possessing different structures at the 5′ ends, a triphosphate (246R), a hydroxyl group (OH-246R), a monophosphate (P-246R), or a m⁷G cap (CAP-246R). B, assays were performed with 5 or 10 pmol of 206R-A⁴⁰, a 206-nucleotide RNA with 40 additional adenosines at the 3′ end. The 246R template lacked an A tail and was used as a control.

FIGURE 4. RDR6 activity on double-stranded RNAs and the absence of a priming activity

All enzymatic assays were performed with HA-RDR6 in the presence of [α-³²P]UTP and all four cold NTPs. A, assays were performed with 10 pmol of the 435F template that had been hybridized with an increasing quantity (0–20 pmol) of its complementary strand, 435R. B, assays were formed with ssRNAs (lanes 1 and 2) and equal amounts of 435F and 435R annealed together and treated with RNase I to remove any residual ssRNAs (lane 3). A portion of the RNA templates used for the reactions were resolved on a denaturing gel and stained with ethidium bromide to demonstrate the near equal amounts of templates used. C, priming activity was tested on 15 pmol of the 246R template in the presence (+) or absence (-) of 30 pmol of miR173 as a primer. The expected product synthesized from the primer (∼80 nucleotide) would migrate to the position indicated by the arrowhead.

FIGURE 5. DNA-dependent RNA polymerase activity of RDR6

A, all enzymatic assays were performed with HA-RDR6 or HA-RDR6m in the presence of α-³²P]UTP and all four cold NTPs. The same PCR product was used to synthesize the ssDNA, by asymmetric PCR, and the ssRNA, by in vitro transcription. 5 pmol of a ssDNA template (d246R) and the same amount of a ssRNA template (246R) were tested with HA-RDR6 (lanes 1–2) or HA-RDR6m (lane 3). B, assays were conducted with a ssDNA (d246R) template, HA-RDR6, and [α-³²P]UTP in the presence (+) or absence (-) of cold NTPs. After the RDR6 reactions, the reaction mixture was treated (+) or not (-) with RNase H or RQ1 (a DNase). C, assays were conducted with HA-RDR6, [α-³²P]UTP, and all four cold NTPs on a ssDNA (d246R) and a dsDNA (ds-d246R) template. The stained gel at the bottom illustrates near equal amounts of templates used. The larger size of the dsDNA template as compared with the ssRNA template is due to the presence of the T7 promoter sequence in the dsDNA.