MicroRNA directs mRNA cleavage of the transcription factor NAC1 to downregulate auxin signals for arabidopsis lateral root development

Abstract

Although several plant microRNAs (miRNAs) have been shown to play a role in plant development, no phenotype has yet been associated with a reduction or loss of expression of any plant miRNA. Arabidopsis thaliana miR164 was predicted to target five NAM/ATAF/CUC (NAC) domain-encoding mRNAs, including NAC1, which transduces auxin signals for lateral root emergence. Here, we show that miR164 guides the cleavage of endogenous and transgenic NAC1 mRNA, producing 3'-specific fragments. Cleavage was blocked by NAC1 mutations that disrupt base pairing with miR164. Compared with wild-type plants, Arabidopsis mir164a and mir164b mutant plants expressed less miR164 and more NAC1 mRNA and produced more lateral roots. These mutant phenotypes can be complemented by expression of the appropriate MIR164a and MIR164b genomic sequences. By contrast, inducible expression of miR164 in wild-type plants led to decreased NAC1 mRNA levels and reduced lateral root emergence. Auxin induction of miR164 was mirrored by an increase in the NAC1 mRNA 3' fragment, which was not observed in the auxin-insensitive mutants auxin resistant1 (axr1-12), axr2-1, and transport inhibitor response1. Moreover, the cleavage-resistant form of NAC1 mRNA was unaffected by auxin treatment. Our results indicate that auxin induction of miR164 provides a homeostatic mechanism to clear NAC1 mRNA to downregulate auxin signals.

Figure 1.

Expression Patterns of NAC1 and miR164 in Wild-Type Arabidopsis RNA. Gel blot analyses of NAC1 mRNA and miR164 from wild-type (Col) leaves (L), inflorescences (I), 2-week-old seedling rosette leaves (Ro), and roots (Rt). A gel blot containing total RNA (10 μg per lane) was hybridized with a NAC1 3′-specific fragment (nucleotides 682 to 1063; top gel). Intact NAC1 mRNA and the NAC1 3′ fragment resulting from miR164-guided cleavage are indicated. 28S rRNA visualized by methylene blue staining was used as a loading control (second gel). A gel blot containing low-molecular-weight RNA (30 μg per lane) was hybridized with an end-labeled DNA oligonucleotide complementary to miR164 (third gel). The membrane was stripped and reprobed with an end-labeled DNA oligonucleotide complementary to miR163 for a loading control (bottom gel).

Figure 2.

Determination of NAC1 mRNA Cleavage by miR164 Using a Transient Expression System. (A) Alignment of miR164 with NAC1 (wild type) and NAC1 mutant (NAC1m) mRNA. The NAC1/miR164 pair contains two mismatches in base pairing, whereas there are eight mismatches in the NAC1 mutant (NAC1m/miR164). The mismatched nucleotides are underlined, and the predicted G-U base pairing is shown with dots. The arrow indicates the cleavage site between nucleotides 763 and 764 of the NAC1 mRNA. Ten clones were sequenced, and an identical 5′ end was obtained in all clones. (B) miR164 sequence and the predicted structure of the MIR164a and MIR164b synthetic precursors. A deletion mutant, MIRΔ164b (contains only 5 nucleotides of miR164), was created by digestion of the precursor MIR164b sequence with PmlI to remove 16 nucleotides of the miR164 sequence. (C) Effect of miR164 on Myc-NAC1 (left) and Myc-NAC1m (right) transcripts in coinfiltration assays. Ten micrograms of low-molecular-weight RNAs (top gels) or 5 μg of total RNA (middle gels) from N. benthamiana leaf samples collected at 48 h after infiltration was loaded and hybridized with the appropriate probes as described for Figure 1. 28S rRNA was used as a loading control (bottom gels). The constructs used for coinfiltration are indicated above each lane. Myc-NAC1 and Myc-NAC1m mRNA are indicated. Asterisks indicate additional bands presumably derived from cleavage of Myc-NAC1 transcripts by N. benthamiana endogenous miRNA.

Figure 3.

RNA Analysis and Phenotypes of 35S-NAC1 Transgenic Plants. (A) Myc-NAC1 but not Myc-NAC1m transcripts were cleaved by endogenous miR164 (top). RNAs were extracted from roots of 2-week-old seedlings of the wild type (Ler), 35S-Myc-NAC1 (Ler), the wild type (Col), 35S-Myc-NAC1 (Col), and 35S-Myc-NAC1m (Col). Each lane contained 5 μg of RNA, and blots were hybridized to a NAC1 gene-specific probe as described for Figure 1. Full-length endogenous NAC1 and Myc-NAC1 mRNA and their respective 3′ fragments resulting from miR164-guided cleavage are indicated. At bottom is a scheme of endogenous NAC1 and Myc-NAC1m transcript constructs and the positions complementary to miR164. (B) Effects of overexpression of 35S-Myc-NAC1 and 35S-Myc-NAC1m on lateral root development. Wild-type (Col) and transgenic (Col) seedlings expressing 35S-Myc-NAC1 or 35S-Myc-NAC1m transgene were grown on MS medium with 2% sucrose. Seedlings were photographed 12 d after germination. Lateral root initials were identified by staining whole seedlings with a mixture of toluidine blue and basic fuchsin and counted with a binocular microscope. Numbers of lateral roots per centimeter of primary root of each seedling (average ± se) for wild-type Col, 35S-Myc-NAC1 (12 independent lines), and 35S-Myc-NAC1m (18 independent lines) are presented at bottom (n = 20). Bar = 1 cm.

Figure 4.

RNA Analysis and Phenotypes of Transgenic Plants with Inducible Expression of miR164. (A) Induction of miR164 and effect on target NAC1 mRNA accumulation. Seedlings containing either 35S-Myc-NAC/pX7-164 or 35S-Myc/pX7-164 were germinated on selective medium for 6 d before being transferred to MS medium or MS + 17-β-estradiol (2 μM). After 10 d, root samples were collected for analysis. Ten micrograms of low-molecular-weight RNAs (top gel) or 20 μg of total RNA (middle gel) was loaded and hybridized with the appropriate probes as described for Figure 1. 28S rRNA was used as a loading control (bottom gel). miR164, full-length Myc-NAC1, and NAC1 mRNAs and their derived 3′ fragments are indicated. Plus and minus signs indicate with or without inducer treatment, respectively. Numbers below the middle gel indicate relative endogenous NAC1 mRNA levels. (B) Effect of inducible expression of miR164 on lateral root development. Representative plants from transgenic lines in (A) were photographed 10 d after transfer to medium with (+) or without (−) inducer. Numbers of lateral roots per centimeter of primary root (average ± se) were determined as described for Figure 3B. Bars = 1 cm. (C) Effect of a shorter time course of induction on endogenous NAC1 mRNA levels. Two independent lines of 35S-Myc/pX7-164 seedlings were germinated on selective medium for 9 d before being transferred to MS + 17-β-estradiol (2 μM). Root samples were collected 24 and 48 h after treatment. Each lane contained 10 μg of total RNA (top gel). Blots were hybridized with NAC1 3′-specific probes as described for Figure 1. 28S rRNA was used as a loading control (bottom gel).

Figure 5.

RNA Gel Blot Analysis and Root Phenotypes of mir164a and mir164b Mutants, dcl1-9 Mutant, and P1/HC-Pro Transgenic Plants. (A) Root phenotypes of wild-type Col, mir164a, and mir164b mutants (top) and of complemented lines (bottom). Seedlings were photographed 10 d after germination on MS medium with 2% sucrose. Bars = 1 cm. (B) T-DNA insertion sites of mir164b-1, mir164a-1, mir164a-2, and mir164a-3 relative to the annotated genes. (C) RNA analysis of mir164a and mir164b mutants and complemented lines. Ten micrograms of total RNA (top gel) or low-molecular-weight RNAs (fourth gel) from roots of 10-d-old seedlings was loaded and hybridized with the appropriate probes as described for Figure 1. The filters were then stripped and rehybridized with a CUC3 3′ gene-specific sequence (nucleotides 1273 to 1863; second gel) or with a 5′ end-labeled DNA oligonucleotide complementary to miR163 (fifth gel) to serve as loading controls. 28S rRNA was used as a total RNA loading control (third gel). For each plant line, the relative accumulation levels of full-length NAC1 mRNA, miR164, and miR163 are indicated below the lanes. Levels in wild-type Col were arbitrarily designated as 1.00. (D) and (E) RNA analysis and root phenotypes of the dcl1-9 mutant and P1/HC-Pro transgenic plants. dcl1-9 (Ler) and P1/HC-Pro (Col) transgenic seedlings and their respective wild-type controls were grown on MS medium with 2% sucrose for 2 weeks. (D) Root RNA was analyzed as described for Figure 1 using 10 μg of low-molecular-weight RNAs (top gel) or 3 μg of total RNA (middle gel). 28S rRNAs bands were used as loading controls (bottom). (E) Representative plants were photographed. dcl1-9 and P1/HC-Pro seedlings were identified by their developmental phenotypes. Bars = 1 cm. Numbers below (A) and (E) indicate the number of lateral roots per centimeter of primary root (average ± se) as described for Figure 3B (n = 20).

Figure 6.

RNA Analysis of Wild-Type (Col) Plants, 35S-Myc-NAC1m Transgenic Plants, and axr1-12, axr2-1, and tir1-1 Plants. Ten-day-old seedlings were treated with NAA (10 μM), and root samples were collected at the times indicated at top. Each lane contained 10 μg of total RNA (top two gels) or low-molecular-weight RNAs (bottom two gels). Blots were hybridized with the appropriate probes as described for Figure 1. Stripped low-molecular-weight RNA filter was rehybridized with a miR163 probe (bottom gels). 28S rRNA was used as a loading control (second gel). Numbers indicate relative levels of NAC1 mRNA (top gel), miR164 (third gel), and miR163 (bottom gel). Relative mRNA levels (average ± se) are 1.89 ± 0.03 (2 h), 1.86 ± 0.03 (4 h), 0.11 ± 0.02 (6 h), and 0.12 ± 0.02 (8 h) for NAC1 and 1.03 ± 0.05 (2 h), 1.15 ± 0.05 (4 h), 1.57 ± 0.05 (6 h), and 1.57 ± 0.10 (8 h) for miR164. Positions of full-length endogenous NAC1, transgenic Myc-NAC1 mRNA, their respective 3′ fragments, and miR164 and miR163 are shown.