DELLA-GAF1 Complex Is a Main Component in Gibberellin Feedback Regulation of GA20 Oxidase 2

Abstract

Gibberellins (GAs) are phytohormones that regulate many aspects of plant growth and development, including germination, elongation, flowering, and floral development. Negative feedback regulation contributes to homeostasis of the GA level. DELLAs are negative regulators of GA signaling and are rapidly degraded in the presence of GAs. DELLAs regulate many target genes, including AtGA20ox2 in Arabidopsis (Arabidopsis thaliana), encoding the GA-biosynthetic enzyme GA 20-oxidase. As DELLAs do not have an apparent DNA-binding motif, transcription factors that act in association with DELLA are necessary for regulating the target genes. Previous studies have identified GAI-ASSOCIATED FACTOR1 (GAF1) as such a DELLA interactor, with which DELLAs act as coactivators, and AtGA20ox2 was identified as a target gene of the DELLA-GAF1 complex. In this study, electrophoretic mobility shift and chromatin immunoprecipitation assays showed that four GAF1-binding sites exist in the AtGA20ox2 promoter. Using transgenic plants, we further evaluated the contribution of the DELLA-GAF1 complex to GA feedback regulation. Mutations in four GAF1-binding sites abolished the negative feedback of AtGA20ox2 in transgenic plants. Our results showed that GAF1-binding sites are necessary for GA feedback regulation of AtGA20ox2, suggesting that the DELLA-GAF1 complex is a main component of the GA feedback regulation of AtGA20ox2.

Figure 1.

A 1,003-bp region of the AtGA20ox2 promoter is regulated by the DELLA-GAF1 complex and responds to GA. A, Schematic representation of the reporter and effector. A 1,003-bp fragment of the AtGA20ox2 promoter was fused to the LUC gene. The effector plasmid expressed the full-length GAF1 or RGAΔ17 under the control of the CaMV 35S promoter with a viral translation enhancer (Ω). B, Transactivation assay of GAF1 and RGAΔ17. The effector, reporter, and internal control constructs were cobombarded into Arabidopsis leaves. The transfected leaves were incubated for 20 h, and then, LUC and rLUC activities were measured. The results are shown as LUC/rLUC activity. Error bars indicate sd (n = 3). C, Fluorometric GUS assay for GA negative feedback in transgenic Arabidopsis plants carrying the 1,003-bp AtGA20ox2 promoter::GUS construct. Gray and black bars represent the GUS activities of plants treated for 1 week with GA₃ and uniconazole (Uni), respectively. The striped bar represents the GUS activity of untreated control plants, which was arbitrarily set to 1. The mean activities of five independent transgenic lines for each construct are shown. Error bars indicate sd.

Figure 2.

Identification of the DELLA-GAF1-regulated region of the AtGA20ox2 promoter using a transactivation assay of GAF1 and RGAΔ17. A and C, Reporter, effector, and internal control constructs used in the assay are shown in A. The constructs were cobombarded into Arabidopsis leaves. B and D, The transfected leaves were incubated for 20 h, and then, LUC and rLUC activities were measured. The results are shown as LUC/rLUC activity. Error bars indicate sd (n = 3).

Figure 3.

Identification of GAF1-binding regions in the AtGA20ox2 promoter in vitro. A and B, Gel retardation assays using recombinant GAF1 protein. A, Oligonucleotides containing cisB (–243 to –224, wild type; lanes 1–4) or mtcisB (mt; lanes 5 and 6) were used as probes. B, Oligonucleotides containing cisC (–229 to –210, wild type; lanes 1–4) or mtcisC (mt; lane 5) were used as probes. Red letters indicate mutated bases. Wild type and mt indicate competition with a 200-fold excess of unlabeled wild-type and mutated probe, respectively. The specific GAF1-DNA complexes are indicated by arrowheads. +, Addition to the reaction mixtures; –, omission from the reaction mixtures. C, Gel retardation assay using recombinant GAF1 protein. Thirty-base-pair regions of the AtGA20ox2 promoter were used as probes. D, Oligonucleotides containing cisE (–871 to –852, wild type; lanes 1–4) or mtcisE (mt; lane 5) were used as probes. Red letters indicate mutated bases. Wild type and mt indicate competition with a 200-fold excess of unlabeled wild-type and mutated probe, respectively. The specific GAF1-DNA complexes are indicated by arrowheads. +, Addition to the reaction mixtures; –, omission from the reaction mixtures.

Figure 4.

GAF1 binds to multiple regions of the AtGA20ox2 promoter in vivo. ChIP assays were performed with anti-GST or anti-GFP in RGApro:RGA-GFP transgenic plants. Green bars indicate GAF1-binding sites of the AtGA20ox2 promoter. The coprecipitated level of each DNA fragment was quantified by real-time PCR and normalized to the input DNA. The relative coprecipitated levels of each DNA fragment using anti-GST were set to 1, and relative enrichments of each DNA fragment using anti-GFP compared with a DNA fragment using anti-GST are shown in the graph. ACT, ACTIN. Error bars indicate the sd of three biological replicates (n = 3). *, P < 0.05 versus preimmune DNA fragment enrichment by Student’s t test.

Figure 5.

Mutation of four GAF1-binding sites completely abolishes DELLA-GAF1 activation of the GA20ox2 promoter. A, Comparison between ID1-cis and GAF1-binding sites in the AtGA20ox2 promoter. Green letters show conserved sequence among these GAF1-binding sites. Arrows indicate the direction of each GAF1-binding site. B, Transactivation assays of GAF1 and RGA. Reporter constructs of each AtGA20ox2 promoter mutant fused with LUC used in the assay are shown. Red bars indicate mutations in GAF1-binding sites of the AtGA20ox2 promoter. C and D, The constructs were cobombarded into Arabidopsis leaves. The transfected leaves were incubated for 20 h, after which the LUC and rLUC activities were measured. The results are shown as LUC/rLUC activity. WT, Wild type. Error bars indicate sd (n = 6). *, P < 0.05 versus control by Student’s t test.

Figure 6.

Mutation of four GAF1-binding sites completely abolishes the GA responsibility of the AtGA20ox2 promoter. A, Reporter constructs of each AtGA20ox2 promoter mutant fused with GUS used in the assay. Red bars indicate mutations in GAF1-binding sites of the AtGA20ox2 promoter. B, Fluorometric GUS assay for GA negative feedback in transgenic Arabidopsis plants carrying 1,003 bp of AtGA20ox2 promoter::GUS, mtABCAtGA20ox2 promoter::GUS (including three mutations), or mtABCEAtGA20ox2 promoter::GUS (including four mutations). The red and green bars represent the GUS activities of plants treated for 1 week with GA₃ and uniconazole (Uni), respectively. The blue bars represent the GUS activity of untreated control plants, which was arbitrarily set to 1 (control = 1). The mean activities of eight independent transgenic lines for each construct are shown. Error bars indicate sd. *, P < 0.05 versus control by Student’s t test. C, Comparison of GUS staining patterns in transgenic plants carrying the wild-type (WT), mtAtGA20ox2 promoter::GUS (mtABC), and mtAtGA20ox2 promoter::GUS (mtABCE) constructs with those of the mutant versions. Meristematic regions of shoot and root tip in seedlings were observed at 12 d. Bars = 0.05 mm (shoot) or 0.2 mm (root tip).

Figure 7.

DELLAs bind to the GAF1-binding region of the AtGA20ox2 promoter in a GA-dependent manner. The DELLA complex binds to the AtGA20ox2 promoter in vivo. ChIP assays were performed with anti-GST or anti-GFP using transgenic plants carrying with RGApro:RGA-GFP that were treated with paclobutrazol (PAC) or PAC + GA₃ or left untreated for 1 week. The coprecipitated level of each DNA fragment T1, including AtGA20ox2, cisA, and cisE, T2, including AtGA20ox2, cisB, and cisC, or AtGA20ox2 coding, was quantified by real-time PCR and normalized to the input DNA. The relative coprecipitated levels of each DNA fragment using anti-GST were set to 1, and relative enrichments of each DNA fragment using anti-GFP versus a DNA fragment using anti-GST are shown in the graphs. Error bars indicate the sd of three biological replicates (n = 3). *, P < 0.05 versus control by Student’s t test.

Figure 8.

Proposed model of GA feedback regulation by the DELLA-GAF1 complex in a GA-dependent manner. In GA feedback regulation, endogenous GA levels affect the amount of DELLA protein, which is reflected in the expression level of AtGA20ox2. A decrease in GAs promotes DELLA accumulation, resulting in increased DELLA binding to GAF1-binding sites on the AtGA20ox2 promoter and the activation of AtGA20ox2 expression, whereas exogenous GA treatment suppresses the accumulation of DELLA, thus decreasing the amount of DELLA on the AtGA20ox2 promoter and AtGA20ox2 expression. DELLAs associate with GAF1 on the AtGA20ox2 promoter and might control the feedback regulation of AtGA20ox2.