SHH1, a homeodomain protein required for DNA methylation, as well as RDR2, RDM4, and chromatin remodeling factors, associate with RNA polymerase IV

Abstract

DNA methylation is an evolutionarily conserved epigenetic modification that is critical for gene silencing and the maintenance of genome integrity. In Arabidopsis thaliana, the de novo DNA methyltransferase, domains rearranged methyltransferase 2 (DRM2), is targeted to specific genomic loci by 24 nt small interfering RNAs (siRNAs) through a pathway termed RNA-directed DNA methylation (RdDM). Biogenesis of the targeting siRNAs is thought to be initiated by the activity of the plant-specific RNA polymerase IV (Pol-IV). However, the mechanism through which Pol-IV is targeted to specific genomic loci and whether factors other than the core Pol-IV machinery are required for Pol-IV activity remain unknown. Through the affinity purification of nuclear RNA polymerase D1 (NRPD1), the largest subunit of the Pol-IV polymerase, we found that several previously identified RdDM components co-purify with Pol-IV, namely RNA-dependent RNA polymerase 2 (RDR2), CLASSY1 (CLSY1), and RNA-directed DNA methylation 4 (RDM4), suggesting that the upstream siRNA generating portion of the RdDM pathway may be more physically coupled than previously envisioned. A homeodomain protein, SAWADEE homeodomain homolog 1 (SHH1), was also found to co-purify with NRPD1; and we demonstrate that SHH1 is required for de novo and maintenance DNA methylation, as well as for the accumulation of siRNAs at specific loci, confirming it is a bonafide component of the RdDM pathway.

Figure 1. Complementation of DNA methylation defects with epitope tagged RdDM components and co-immunoprecipitation analyses.

(A and B) Southern blot analysis of DNA methylation at the MEA-ISR locus using genomic DNA digested with the Msp I methylation sensitive restriction enzyme. Bands representing the presence or absence of DNA methylation are indicated as ME or unME, respectively. In (A), genomic DNA was isolated from either wild-type plants of the Columbia (Col) ecotype, homozygous nrpd1-4 mutant plants, or T₄ plants homozygous for both the nrpd1-4 allele and a transgene containing the indicated carboxy-terminal epitope tagged versions of the NRPD1 gene under the control of its endogenous promoter (pNRPD1). BLRP; Biotin Ligase Recognition Peptide. In (B), genomic DNA was isolated from either Col plants, shh1-1, rdr2-2, or clsy1-7 single mutant plants, or T₂ transgenic plants expressing the indicated epitope tagged versions of SHH1, RDR2, or CLSY1 in their respective homozygous mutant backgrounds. (C–F) Co-immunoprecipitation analyses confirming the interactions between (C) NRPD1 and RDR2, (D) NRPD1 and CLSY1, (E) NRPD1 and SHH1, and (F) NRPD1 and RDM4. For each experiment, lanes containing protein extracted from each of the parental lines and the resultant F₁ line are indicated below the “input” and “immunoprecipitation (IP)” headings and the antibody (α) used for each western blot is indicated next to each panel (upper left corner). The position of each epitope tagged or endogenous protein is indicated with a closed arrowhead and background bands are indicated by an asterisk (*).

Figure 2. Characterization of the shh1-1 allele and genomic complementation.

(A) Cartoon representation of the SHH1 locus. The promoter and UTRs are shown in grey and the exons are shown as black arrows. The location of the T-DNA insertion in the shh1-1 mutant is indicated and the three regions, A, B, and C, amplified by Reverse-Transcriptase PCR are indicated below. (B) Analysis of SHH1 expression in the shh1-1 allele by Reverse Transcriptase followed by gene specific PCR. Three primer sets, (A), (B), and (C), within the SHH1 gene were assessed (see Table S2). As a loading control, primers amplifying a ubiquitin gene (UBQ) were also used (see Table S2). Reactions lacking (-) the Reverse Transcriptase (RT) were included to ensure that no DNA contamination was present and reactions containing (+) the RT were conducted in duplicate. (---) indicates the PCR reaction was conducted in the absence of cDNA. (C) MEA-ISR Southern blot as described in Figure 1, showing complementation of the shh1-1 mutant phenotype with a genomic SHH1 transgene (pSHH1::SHH1) using genomic DNA extracted from wild-type Col plants, shh1 mutants, or individual T₁ transformants in the shh1-1 mutant background.

Figure 3. Characterization of methylation defects in the shh1 mutant.

Assessment of DNA methylation levels at the (A) MEA-ISR and (B) Ta3 loci by Southern blotting using Msp I digested genomic DNA extracted from the genotypes indicated above each lane and a locus-specific radiolabeled probe. ME; methylated DNA, unME; unmethylated DNA. (C) Assessment of the levels of DNA methylation at the endogenous FWA locus by bisulfite sequencing. The y axis is the percent of methylation and the x-axis is the context of methylation. (D) Assessment of the level of DNA methylation at the AtSN1 locus using a methyl-cutting assay in which genomic DNA extracted from the indicated genotypes was either untreated (unCut) or digested with the methylation sensitive Hae III enzyme prior to amplification of the AtSN1 locus. (E and F) Assessment of the role of SHH1 in de novo DNA methylation by flowering time (E) and bisulfite sequencing (F). In (E), the graph shows the leaf number (y axis) of individual T₁ plants of the indicated genotype (x axis) either untransformed or transformed with a transgene carrying the FWA gene. The black bars represent standard error. In (F), T₂ tissue from individual Col or shh1 T₁ transformants was used to assess the level of DNA methylation on the FWA transgene. The y axis is the percent of methylation observed and the x-axis is the context of methylation. The number in parentheses next to the plant genotype indicates the flowering time of the T₁ parent.

Figure 4. Characterization of siRNAs and Pol-V dependent transcripts in the shh1 mutant.

(A) Assessment of siRNA levels at the indicated loci (right) by northern blotting using RNA extracted from the genotypes indicated above. (B–C) Assessment of Pol-V dependent transcript levels (y axis) using RNA isolated from the genotypes indicated (x axis) relative to ACTIN and normalized to Col at the MEA-ISR and IGN5 loci. Thick blue bars represent the average of three biological replicates and thin black bars are the standard error of the three experiments.