The interaction between DCL1 and HYL1 is important for efficient and precise processing of pri-miRNA in plant microRNA biogenesis

Abstract

It has been reported that some double-stranded RNA (dsRNA) binding proteins interact with small RNA biogenesis-related RNase III enzymes. However, their biological significance is poorly understood. Here we examine the relationship between the Arabidopsis microRNA- (miRNA) producing enzyme DCL1 and the dsRNA binding protein HYL1. In the hyl1-2 mutant, the processing steps of miR163 biogenesis were partially impaired; increased accumulation of pri-miR163 and reduced accumulation of short pre-miR163 and mature miR163 as well as misplaced cleavages in the stem structure of pri-miR163 were detected. These misplaced cleavages were similar to those previously observed in the dcl1-9 mutant, in which the second double-stranded RNA binding domain of the protein was disrupted. An immunoprecipitation assay using Agrobacterium-mediated transient expression in Nicotiana benthamiana showed that HYL1 was able to form a complex with wild-type DCL1 protein, but not with the dcl1-9 mutant protein. We also examined miR164b and miR166a biogenesis in hyl1-2 and dcl1-9. Increased accumulation of pri-miRNAs and reduced accumulation of pre-miRNAs and mature miRNAs were detected. Misplaced cleavage on pri-miR164b was observed only in dcl1-9 but not in hyl1-2, whereas not on pri-miR166a in either mutant. These results indicate that HYL1 has a function in assisting efficient and precise cleavage of pri-miRNA through interaction with DCL1.

FIGURE 1.

Impaired miR163 processing and determination of the cleavage sites on pri-miR163. (A) Northern blot analysis for detection of pri-miR163 and/or pre-miR163s. Bands a, b, c, and d correspond to pri-miR163, long pre-miR163, short pre-miR163, and the remnant, respectively. The asterisk indicates an additional band between the long pre-miR163 (b) and short pre-miR163 (c) in hyl1-2. (B) Northern blot analysis for detection of small RNA UL and miR163. 5S rRNA and tRNAs were used as loading controls. (C) The sequence of the stem region of pri-miR163. The three correct cleavage sites are indicated by arrows b–d, which correspond to the roots of structures b–d in A. The sequence of miR163 is underlined. The misplaced cleavage sites in hyl1-2 and/or dcl1-9 mutants are shown by b′ and c′, respectively. Although the miR163 gene of dcl1-9 has a 12-nt deletion and 60-nt insertion in its loop structure compared with those in Col-0 and Ler (Kurihara and Watanabe 2004), the sequence and structure of the stem region were the same as those in Col-0 and Ler. (D) Cleavage sites categorized into “others” are shown by lines. The numbers shown on the line indicate total numbers of clones at that position.

FIGURE 2.

Western blot analysis to detect the interaction between DCL1 and HYL1. (A) Coimmunoprecipitation of HYL1-HA with Flag-DCL1 using the anti-Flag antibody. RNase A was added (RNase A+) at a concentration of 50 μg/mL. (B) HYL1 was coimmunoprecipitated with DCL1 but not dcl1-9 protein. Each Agrobacterium culture was infiltrated, respectively, at the following densities: Flag-DCL1, FLAG-dcl1-7, FLAG-dcl1-9 OD₆₀₀ = 0.5; HYL1-HA, OD₆₀₀ = 0.3. (WT) wild-type, (IP) immunoprecipitate.

FIGURE 3.

(A) Northern blot analysis for detection of miR164 and miR166. (B) Northern blot analysis for detection of miR164b and miR166a precursors. mir164b-1 mutant (SALK_136105) was used as a negative control (Mallory et al. 2004). 5S rRNA and tRNAs was used as a loading control. The mean relative accumulation levels for pri-miRNAs and pre-miRNAs are indicated under the panels, respectively. The accumulation levels in wild-type plants were normalized to 1.0. (C) Cycle-RT-PCR analysis of pre-miR164b and pre-miR166a.