Expression of Arabidopsis MIRNA genes

Abstract

MicroRNAs (miRNAs) are approximately 21-nucleotide noncoding RNAs that regulate target transcripts in plants and animals. In addition to miRNAs, plants contain several classes of endogenous small interfering RNAs (siRNAs) involved in target gene regulation and epigenetic silencing. Small RNA libraries were constructed from wild-type Arabidopsis (Arabidopsis thaliana) and mutant plants (rdr2 and dcl3) that were genetically enriched for miRNAs, and a computational procedure was developed to identify candidate miRNAs. Thirty-eight distinct miRNAs corresponding to 22 families were represented in the libraries. Using a 5' rapid amplification of cDNA ends procedure, the transcription start sites for 63 miRNA primary transcripts from 52 MIRNA loci (99 loci tested) were mapped, revealing features consistent with an RNA polymerase II mechanism of transcription. Ten loci (19%) yielded transcripts from multiple start sites. A canonical TATA box motif was identified upstream of the major start site at 45 (86%) of the mapped MIRNA loci. The 5'-mapping data were combined with miRNA cloning and 3'-PCR data to definitively validate expression of at least 73 MIRNA genes. These data provide a molecular basis to explore regulatory mechanisms of miRNA expression in plants.

Figure 1.

Cloning and identification of Arabidopsis miRNAs. A, Flowchart for identification miRNAs in cloned small RNA libraries. The number of small RNAs passing each filter is shown in parentheses. B, Predicted precursor foldback structure of miRNAs (red) validated in this study. C, Blot analysis of small RNAs in wild-type (Col-0) and mutant (hyl1-2, hst-15, hen1-5, dcl1-7, dcl2-1, dcl3-1, rdr1-1, rdr2-1, rdr6-15, sgs3-11, and zip1-1) plants. AtSN1-siRNAs and siR1511 represent two classes of endogenous siRNA. Relative accumulation (R.A.) of small RNAs in each mutant compared to wild-type Col-0 is shown.

Figure 2.

RLM-5′RACE on MIRNA transcripts. A, Schematic representation of a generic MIRNA transcript, and SCL6-IV mRNA (5′ cap-containing control) and miR171-guided cleavage product from SCL6-IV mRNA (noncapped control). The relative positions of oligonucleotide primers used in 5′RACE and 3′RACE reactions are shown, with the alternative primer sets shown in gray. B, RLM-5′RACE reactions using poly(A⁺)-selected RNA. RNA was either pretreated with CIP + TAP (even-numbered lanes) or with buffer (odd-numbered lanes) prior to adaptor ligation. The 5′RACE products for SCL6-IV mRNA or internal cleavage product (lanes 1–4) and three MIRNA loci (lanes 5–10) were resolved on a 2% agarose gel. Gene-specific primers used in each reaction are indicated above each lane.

Figure 3.

Transcription initiation sites of Arabidopsis MIRNA primary transcripts, and core promoter elements. A, Base composition at positions flanking MIRNA transcription initiation sites (n = 63). B, Genomic sequence of 60 nt flanking each of 63 MIRNA initiation sites (red letters) from 52 MIRNA loci. Putative TATA box-like motifs (bold) are indicated. MotifMatcher scores are given at the end of each sequence. C, Frequency of high-scoring TATA box-like motifs within a 250-nt (−200 to +50) context encompassing all mapped MIRNA transcript initiation sites. Frequency (%) was determined by count of sequences with TATA box-like motif within a single-nucleotide scrolling window.

Figure 4.

Locus-specific expression of 99 predicted MIRNA genes encoding validated miRNAs in Arabidopsis. Expression of a specific locus was considered definitive (dark green shading) if a primary transcript was detected by 5′RACE or 3′RACE, or a unique miRNA sequence was cloned or amplified from the ASRP library described here (gray shading with total clones sequenced) or from another published library (Other Refs.). The number of clones corresponding to a specific miRNA or miRNA* (in parentheses) sequence in the ASRP database is shown. Sequences that were detected only in other studies are indicated by orange in the 3′RACE and references columns. Loci for which data support expression from more than one possible gene are indicated by light green shading. nt, Not tested. References cited are as follows: 1, Allen et al. (2004); 2, Aukerman and Sakai (2003); 3, Chen (2004); 4, Jones-Rhoades and Bartel (2004); 5, Kurihara and Watanabe (2004); 6, Llave et al. (2002a); 7, Llave et al. (2002b); 8, Mette et al. (2002); 9, Palatnik et al. (2003); 10, Park et al. (2002); 11, Reinhart et al. (2002); 12, Sunkar and Zhu (2004); and 13, Arabidopsis EST clones were identified for MIR167d (GenBank accession no. AU239920) and MIR168a (H77158).