ARF5/MONOPTEROS directly regulates miR390 expression in the Arabidopsis thaliana primary root meristem

Abstract

The root meristem is organized around a quiescent center (QC) surrounded by stem cells that generate all cell types of the root. In the transit-amplifying compartment, progeny of stem cells further divides prior to differentiation. Auxin controls the size of this transit-amplifying compartment via auxin response factors (ARFs) that interact with auxin response elements (AuxREs) in the promoter of their targets. The microRNA miR390 regulates abundance of ARF2, ARF3, and ARF4 by triggering the production of trans-acting (ta)-siRNA from the TAS3 precursor. This miR390/TAS3/ARF regulatory module confers sensitivity and robustness to auxin responses in diverse developmental contexts and organisms. Here, we show that miR390 is expressed in the transit-amplifying compartment of the root meristem where it modulates response to exogenous auxin. We show that a single AuxRE located in miR390 promoter is necessary for miR390 expression in this compartment and identify that ARF5/MONOPTEROS (MP) binds miR390 promoter via the AuxRE. We show that interfering with ARF5/MP-dependent auxin signaling attenuates miR390 expression in the transit-amplifying compartment of the root meristem. Our results show that ARF5/MP regulates directly the expression of miR390 in the root meristem. We propose that ARF5, miR390, and the ta-siRNAs-regulated ARFs modulate the response of the transit-amplifying region of the meristem to exogenous auxin.

Keywords: Arabidopsis; auxin; meristem; miRNA; root.

Figure 1

MIR390a is expressed in the root meristem and modulates response to auxin. (a, a’) Expression of miR390 detected by whole‐mount in situ hybridisation (WMISH) on wild type plants. miR390. (b, b’) expression pattern of the pMIR390a::GUS transcriptional reporter for MIR390a. miR390 and the MIR390a promoter are expressed in the primary root meristem (a, b) and in lateral root primordia (a’, b’). The arrowhead indicates the transition zone marking the end of the meristem. Images were taken at 5 days after germination (DAG) and scale bars are 25 μm. (c, d) Expression of miR390 by WMISH (c) and pMIR390a::GUS (d) in response to IAA treatment. Plants (5DAG) were treated by 1 μM IAA for 24 hr before fixation. Images were taken at 6DAG and scale bars are 25 μm. (e) Length of the primary root in response to IAA concentrations. Plants were grown for 3 days on control medium before transfer to medium containing the indicating amount of IAA for another 3 days. Root length after 6 days was normalised to its value in the absence of exogenous IAA. The lines represent the average root length of 20 plants and the shaded ribbons the standard error to the mean. Comparison between samples was performed using ANOVA and Tukey's HSD. Samples with identical letters do not significantly differ (α = 0.05)

Figure 2

An auxin response element containing enhancer controls MIR390a expression in the root meristem. (a) Nested deletions in the region located upstream of the MIR390a stem loop region. The segment used to drive the expression of the GUS:GFP reporter is represented by a line, the MIR390a transcription start by an arrow and the stem loop by a thick line. The position of the primary root enhancer is indicated in blue and its sequence as well as the putative auxin response element (ARE) are shown in the dashed blue box. (b) Meristem expression of GUS reporter driven by the indicated segment of MIR390a promoter. The numbers indicate the proportion of independent primary transformants displaying the phenotype and scale bars are 25 μm

Figure 3

Five auxin response factors (ARFs) interact with MIR390a primary root enhancer (PRE) in yeast. (a) Yeast co‐expressing a trimerized version of the PRE as bait and a fusion between the indicated ARF and the Gal4 activation domain (AD) were grown on selective media in the presence or not of 3AT. ARF5, ARF9, ARF18, ARF4, or ARF8 interacted with PRE resulting in yeast growth on selective medium (arrowheads). (b) Yeast expressing as bait a trimerized version of the PRE lacking the auxin response element (ARE) did not grow on selective medium

Figure 4

ARF5/MP controls the abundance of miR390 in the primary root meristem. (a) RNA gel blot analysis of 15 μg of root RNA from 7DAG wild type (Col) or the indicated ARF mutant plants hybridized with miR390. U6 snRNA served as a loading control, and numbers are the ratios of miR390 to U6 signal. (b) Expression of pMIR390a::GUS in heterozygous (mpS319/+) and homozygous (mpS319) monopterous mutants. Scale bars are 50 μm. (c) Distribution of the pMIR390a::GUS intensity in heterozygous (mpS319/+) and homozygous (mpS319) monopterous mutants. (d) Expression of pMIR390a::GUS in F1 of a cross between pRBCS5A::bdl‐GR and pMIR390a::GUS upon treatment with 10 μM dexamethasone (DEX) for 48 hr to induce the nuclear translocation of bdl and inhibition of MP activity or in control conditions (DMSO). The proportion of plants with the depicted signal is indicated. Scale bars are 50 μm. (e) RNA gel blot analysis of 15 μg of root RNA from 7‐day‐old pRBCS5A::bdl‐GR treated for 48 hr with 10 μM DEX or DMSO and hybridized with miR390. U6 snRNA served as a loading control, and numbers are the ratios of miR390 to U6 signal

Figure 5

ARF5/MP activates the MIR390a promoter via the auxin response element located in the primary root enhancer. (a) Fusions between the VP16 domain and ARF4 (ARF4), ARF5 (ARF5) or empty plasmid (NO) were co‐expressed in tobacco leaves with different segment of the MIR390a promoter (−94 bp, −519 bp or −555 bp see Figure 2) driving the expression of the firefly luciferase (Luc). In the ‐555∆ARE promoter, a 5 bp deletion removes the ARE in the −555 bp segment (Figure 2). A 35S‐driven renilla luciferase (Ruc) is used to normalise activity (Luc/Ruc, expressed in arbitrary units). Each dot is a biological replicate and the horizontal bars the mean of the four replicates. (b) Annotated screen shot of genome browser at the MIR390a locus with a ARF5 DAP‐seq peak (O'Malley et al., 2016) located at the same position as the PRE (blue box). Position of the primers used for ChIP‐qPCR (C) is depicted by red arrows. (c) Enrichment of ARF5‐GFP at three loci was assessed by ChIP‐qPCR. Chromatin prepared from wild type (WT) or ARF5‐GFP plants was immunoprecipitated by an anti‐GFP antibody and enrichment of ARF5‐GFP in the region PRE of MIR390a (MIR390a PRE), in the second exon of TMO5 or upstream of the HSF1 loci was assessed by qPCR. Fold enrichment in the immunoprecipitated fraction over input fraction in ARF5‐GFP sample was normalised to the one in WT plants and expressed as arbitrary unit. Each dot is a biological replicate and the horizontal bars the mean of the three replicates. Statistical significance was evaluated by the Kruskal‐Wallis test and p values are indicated