Arabidopsis KANADI1 acts as a transcriptional repressor by interacting with a specific cis-element and regulates auxin biosynthesis, transport, and signaling in opposition to HD-ZIPIII factors

Abstract

The formation of leaves and other lateral organs in plants depends on the proper specification of adaxial-abaxial (upper-lower) polarity. KANADI1 (KAN1), a member of the GARP family of transcription factors, is a key regulator of abaxial identity, leaf growth, and meristem formation in Arabidopsis thaliana. Here, we demonstrate that the Myb-like domain in KAN1 binds the 6-bp motif GNATA(A/T) and that this motif alone is sufficient to squelch transcription of a linked reporter in vivo. In addition, we report that KAN1 acts as a transcriptional repressor. Among its targets are genes involved in auxin biosynthesis, auxin transport, and auxin response. Furthermore, we find that the adaxializing HD-ZIPIII transcription factor REVOLUTA has opposing effects on multiple components of the auxin pathway. We hypothesize that HD-ZIPIII and KANADI transcription factors pattern auxin accumulation and responsiveness in the embryo. Specifically, we propose the opposing actions of KANADI and HD-ZIPIII factors on cotyledon formation (KANADI represses and HD-ZIPIII promotes cotyledon formation) occur through their opposing actions on genes acting at multiple steps in the auxin pathway.

Figure 1.

EMSA Reveals KAN1 DNA Binding Characteristics in Vitro. The in vitro KAN1 DNA binding site (KBX; boxed) was identified using affinity-purified KAN1db-GST protein in EMSA-based oligonucleotide selection (Supplemental Figure 1). The height of each nucleotide letter is proportional to its representation. Effects of mutating individual sites within the consensus DNA binding site are shown immediately below each position in the consensus KBX site. The mean fraction of bound DNA in three independent replicates was calculated relative to the consensus (GAATAA, lane 1), which was arbitrarily set to 1.0. EMSA of KAN1db-GST bound to a perfect palindrome of KBX (GAATATT, lane 8) was similar to that of that of a single site. The KAN1db-GST showed little affinity for the consensus binding site for the GARP protein ARR10 (AGATT, lane 9) (Hosoda et al., 2002).

Figure 2.

KBX Confers Tissue-Specific Repression of the Downstream Gene in Planta. (A) GAL4 GFP enhancer trap line E100 exhibits GFP expression in the SAM and young leaf primordia. (B) The GUS expression pattern in E100>>UAS-mKBX:GUS mirrors the GFP pattern in E100. (C) In E100>>UAS-KBX:GUS, the GUS stain was only detected on the adaxial side of the leaf primordia and absent in the SAM and on the abaxial side of leaves. (D) KAN1:GUS is expressed on the abaxial side of leaf primordia and in the SAM, a pattern complementary to the E100>>UAS-KBX:GUS expression. Bars = 20 μm.

Figure 3.

Posttranslational Activation of KAN1-GR Produces Defects in Leaf Polarity and Meristem Function. Continuous exposure of KAN1-GR seedlings to 10 μM DEX (B) on media for 9 d led to loss of cotyledon blade expansion, formation of partially radialized leaf primordia, and inhibition of further shoot meristem activity consistent with strong KAN1 overexpression. Mock-treated KAN1-GR seedlings (A) resembled mock- or DEX- treated Col seedlings (Supplemental Figure 2). By contrast, soil-grown KAN1-GR seedlings exposed every other day to 10 μM DEX ([D] and [F]) displayed reduced petiole and blade expansion with strong epinasty leading to leaves with an asymmetric appearance ([G]; bottom) that was not evident in mock-treated plants ([C], [E], and [G]; top). Plants were photographed at 14 ([C] and [D]) and 29 d old ([E] to [G]).

Figure 4.

Validation of Microarray/RT-PCR Identified KAN1 Target Genes by RNA-Seq. (A) Transcripts that tested positive by RT-PCR test. Data plotted as number of normalized counts after DEX treatment. (B) Transcripts that tested positive by RT-PCR test. Data plotted as ratio of number of counts in DEX treated KAN1-GR samples versus DEX-treated Col samples. (C) Transcripts that tested negative by RT-PCR test. Data plotted as number of normalized counts after DEX treatment. (D) Transcripts that tested negative by RT-PCR test. Data plotted as ratio of number of counts in DEX treated KAN1-GR samples versus DEX-treated Col samples. Minutes = minutes of DEX treatment.

Figure 5.

ChIP Confirms DEX-Dependent Association of KAN-GR with the Promoters of KANT Genes. ChIP was performed on 9-d-old transgenic Arabidopsis seedlings using antibodies specific for GR. Immunoprecipitated genomic DNA from mock (M) and DEX (D) treated KAN-GR and wild-type control seedlings (Supplemental Figure 4) was amplified with primers specific for the indicated promoters. Fold enrichment was calculated by normalizing PCR product intensities to a negative control, the RIBOSOMAL PROTEIN L4D (RPL4D) coding region, followed by calculating the ratio DEX IP/input to mock IP/input. The mean of at least two independent IP experiments (Supplemental Table 2) with technical replicates is reported as fold enrichment. Schematics of the gene promoters are shown with the positions and orientations of KBX sites indicated by < or > and the amplified region represented by a gray bar.

Figure 6.

Cotyledon Numbers Are Altered in HD-ZIPIII Gain- and Loss-of-Function Mutants. (A) Frequency of tricots in self progeny of gain of function mutants of REV, PHABULOSA, PHAVOLUTA, and INCURVATA4. (B) incurvata4-d mutant tricot with normal cotyledon blades. (C) phabulosa-1d tricot with tube formed adaxialized cotyledons. (D) ph phb rev triple mutants with no (left), one (middle), or two (right) radialized, abaxialized cotyledons.

Figure 7.

Differential Regulation of Members of the NPY/MEL Gene Family by REV and KAN. (A) Phylogenetic tree of members of the NPH3-like family of genes. Values are probabilities for genotype by time of treatment interaction in a two-way ANOVA (microarray experiment) comparing GR- REV lines treated with DEX to KAN1-GR lines treated with DEX. NA, not assayed. (B) Graphs of transcript levels for wild-type (blue), GR-REV (red), and KAN1-GR (green) lines treated with DEX. M, data from microarray experiment. y axis is normalized expression in log₂ units. S, data from RNA-seq experiment. y axis is normalized counts. Error bars are se. (C) qRT-PCR experiments on cDNAs made from DEX-treated seedlings for 1 h in the presence and absence of CHX. Three biological replicates were tested for each bar. (Three technical replicates were tested for each biological replicate.) Expression is relative to actin. Asterisk indicates significant difference relative to Col.

Figure 8.

Model for Embryonic Patterning of Auxin Biosynthesis, Transport, and Reception by the Ad/Abaxial Regulators REV and KAN. (A) Summary of REV- and KAN-regulated auxin pathway genes. Arrow indicates direction of regulation. i = indirectly regulated. Data for direct regulation of TAA1 and YUC5 by REV is from Brandt et al. (2012) (1), Stepanova et al. (2008) (2), Friml et al. (2002) (3), Ploense et al. (2009) (4), and Hamman et al. (1999) (5). (B) Location of KAN- and REV-regulated components of the auxin pathway during embryogenesis.