Proteolysis-independent downregulation of DELLA repression in Arabidopsis by the gibberellin receptor GIBBERELLIN INSENSITIVE DWARF1

Abstract

This article presents evidence that DELLA repression of gibberellin (GA) signaling is relieved both by proteolysis-dependent and -independent pathways in Arabidopsis thaliana. DELLA proteins are negative regulators of GA responses, including seed germination, stem elongation, and fertility. GA stimulates GA responses by causing DELLA repressor degradation via the ubiquitin-proteasome pathway. DELLA degradation requires GA biosynthesis, three functionally redundant GA receptors GIBBERELLIN INSENSITIVE DWARF1 (GID1a, b, and c), and the SLEEPY1 (SLY1) F-box subunit of an SCF E3 ubiquitin ligase. The sly1 mutants accumulate more DELLA proteins but display less severe dwarf and germination phenotypes than the GA biosynthesis mutant ga1-3 or the gid1abc triple mutant. Interestingly, GID1 overexpression rescued the sly1 dwarf and infertility phenotypes without decreasing the accumulation of the DELLA protein REPRESSOR OF ga1-3. GID1 rescue of sly1 mutants was dependent on the level of GID1 protein, GA, and the presence of a functional DELLA motif. Since DELLA shows increasing interaction with GID1 with increasing GA levels, it appears that GA-bound GID1 can block DELLA repressor activity by direct protein-protein interaction with the DELLA domain. Thus, a SLY1-independent mechanism for GA signaling may function without DELLA degradation.

Figure 1.

The Effect of GID1 Overexpression on Plant Growth and RGA Protein Accumulation. (A) Three chimeric constructs (HA:GID1a, HA:GID1b, and HA:GID1c) were introduced into wild-type Ler, sly1-10, sly1-2, ga1-3, and rga-Δ17(+/−) backgrounds. Heterozygous (+/−) plants were used because homozygous rgaΔ-17 (−/−) plants are infertile. Representative 36-d-old T3 transgenic plants in which HA:GID1a, HA:GID1b, and HA:GID1c are overexpressed are shown. Plants were treated with (+GA) or without (−GA) 10 μM GA₄ every 3 d. Bars = 5 cm. (B) Protein blot analysis of HA:GID1 fusions and RGA protein accumulation in independent T3 lines was performed using HA and RGA antibodies. Total protein (150 μg) from wild-type Ler and 40 μg total protein isolated from all other genotypes in the absence of GA application was loaded. Representative pictures shown in (A) correspond to lanes shown in bold in (B). Equal protein loading was confirmed by Ponceau staining (bottom panels).

Figure 2.

GID1 Overexpression in ga1-3 Enhanced Vegetative Sensitivity to GA Treatment. The ga1-3 mutant and the transgenic ga1-3 mutant plants overexpressing each HA:GID1 fusion were grown on MS-agar containing different concentrations of GA₄ (0, 10⁻¹⁰, 10⁻⁹, and 10⁻⁸ M). After 10 d of incubation at 22°C, the leaf area and root elongation of these seedlings was measured. Error bars are se (n = 10). A significant difference from untransformed ga1-3 is indicated: a, P < 0.05; b, P < 0.01, as determined by t test.

Figure 3.

The sly1 Intermediate Phenotype Is Dependent on the Presence of a Functional DELLA Motif in RGA and GAI Proteins. (A) Shown are 50-d-old sly1-10, sly1-10 rga-Δ17, and rga-Δ17 mutants. Enlarged magnification of the sly1-10 rga-Δ17 double mutant is shown in the right panel. (B) Shown are 50-d-old sly1-10, gai-1, and sly1-10 gai-1 mutants. Bars = 1 cm. (C) GAI and RGA protein accumulation in wild-type Ler, ga1-3, sly1-2, sly1-2 gai-t6, gai-1, sly1-10 gai-1, and sly1-10 was determined by protein blot analysis of 40 μg of total protein extracted from 50-d-old rosette leaves. The asterisk denotes nonspecific bands.

Figure 4.

Increased Interaction of GID1b with DELLA RGA Protein Is Dependent on GA. The co-IP experiment was performed using protein extracted from 12-d-old sly1-10 FLAG:GID1b seedlings. Protein extract was incubated with FLAG agarose in the presence of 0.1% ethanol (mock), 1 μM GA₃, or 100 μM GA₃ and loaded on an SDS-PAGE gel. Protein blot analysis was performed using anti-RGA, anti-FLAG, and anti-cullin. Forty micrograms of total sly1-10 and sly1-10 FLAG:GID1b protein were loaded (input).

Figure 5.

The GA Biosynthesis Inhibitor PAC Blocks Rescue of sly1 by GID1 Overexpression and Causes Decreased RGA and GAI Accumulation. (A) The 10-d-old Ler, ga1-3, sly1-2, and the sly1-2 GID1-OE plants were transferred to MS-agar with and without 1 μM PAC treatment and incubated for 12 d at 22°C. Bars = 1 cm. (B) The effect of PAC treatment on the RGA and GAI protein accumulation was determined by protein blot analysis using RGA antibody. Forty micrograms of total protein from (A) was loaded and equal loading confirmed by Ponceau staining. Controls include wild-type Ler, ga1-3, ga1-3 rgat-2, and the gid1a gid1b gid1c triple mutant with and without GA treatment.

Figure 6.

Effect of the ga1-3 Mutation on sly1-10 Growth and RGA Protein Accumulation. (A) The 21-d-old seedlings of ga1-3, sly1-10, and the ga1-3 sly1-10 double mutant in the absence (−GA) and presence (+GA) of 100 μM GA₄. Bar = 5 mm. (B) The effect of GA treatment (100 μM GA₃ treatment) on accumulation of RGA protein in the ga1-3, sly1-10, and ga1-3 sly1-10 mutants. Plants (35 d old) were treated with GA, and time points were taken as indicated for protein blot analysis. sly1-10 is a control for equal loading. Protein was extracted from rosette leaves. (C) RGA mRNA accumulation in rosette leaves of the ga1-3, sly1-10, and ga1-3 sly1-10 mutants was determined at time points indicated after GA treatment by quantitative RT-PCR. Mean values for at least three independent experiments are shown. Error bars show sd.

Figure 7.

Interaction with GID1 Accelerates RGA Protein Accumulation after GA Treatment. (A) Ten-day-old sly1-2 and sly1-2 GID1-OE seedlings were transferred to MS-agar plus 1 μM PAC for 12 d. Seedlings were sprayed with 100 μM GA₃ and time points (0, 2, 9, 15, 24, 72, and 120 h) taken for protein blot analysis. Untreated sly1-2 is a loading control. C, control without PAC. Equal protein loading was confirmed by Ponceau staining. (B) RGA mRNA accumulation was determined by quantitative RT-PCR during the time course indicated. Mean values for at least three independent experiments are shown. Error bars show sd. A t test was used to determine a statistically significance increase (a, P < 0.05; b, P < 0.01) or decrease (c, P < 0.05) compared with 0 h.

Figure 8.

Model for the Proteolysis-Independent Regulation of DELLA Repressor Activity. (A) In the absence of GA, GA responses are inhibited since RGA protein levels and repressor activity is high. GA treatment relieves DELLA repression of GA responses by causing the formation of the GID1-GA-DELLA complex recognized by the SCF^SLY1 E3 ubiquitin ligase. Polyubiquitination by SCF^SLY1 causes DELLA proteolysis via the 26S proteasome. (B) The sly1 mutants accumulate DELLA proteins at a higher level due to lack of DELLA ubiquitination and proteolysis. However, the DELLA protein that accumulates is a mixture of the active DELLA repressor (light gray) and inactive GID1-GA-DELLA (dark gray), resulting in an intermediate phenotype. (C) sly1 GID1-OE plants show increased GA response due to an increase in the proportion of inactive GID1-GA-DELLA complex relative to active DELLA repressor.