DELLA-mediated PIF degradation contributes to coordination of light and gibberellin signalling in Arabidopsis

Abstract

Light and gibberellins (GAs) antagonistically regulate hypocotyl elongation in plants. It has been demonstrated that DELLAs, which are negative regulators of GA signalling, inhibit phytochrome-interacting factors 3 and 4 (PIF3 and PIF4) by sequestering their DNA-recognition domains. However, it is unclear whether there are other mechanisms of regulatory crosstalk between DELLAs and PIFs. Here, we demonstrate that DELLAs negatively regulate the abundance of four PIF proteins through the ubiquitin–proteasome system. Reduction of PIF3 protein abundance by DELLAs correlates closely with reduced hypocotyl elongation. Both sequestration and degradation of PIF3 by DELLAs contribute to a reduction in PIF3 binding to its target genes. Thus, we show that promotion of PIF degradation by DELLAs is required to coordinate light and GA signals, and the dual regulation of transcription factors by DELLAs by both sequestration and degradation may be a general mechanism.

Introduction

Light promotes plant photomorphogenesis, giving rise to open and expanded cotyledons, and short hypocotyls in light-grown Arabidopsis seedlings. In the dark, seedlings undergo skotomorphogenesis/etiolation, as characterized by closed cotyledons and elongated hypocotyls¹. A subset of basic helix-loop-helix (bHLH) transcription factors, known as phytochrome-interacting factors (PIFs), has been reported to have a key role in etiolation and light-regulated plant development. PIF3 was the first characterized member of this gene family, identified by yeast two-hybrid screen using phytochrome B as the bait². Mutation of PIF3 resulted in short hypocotyls in red light, indicating that PIF3 is a negative regulator of light signal transduction^3,4. Several other homologous PIF proteins, including PIF1, PIF4, and PIF5, have also been reported to regulate photomorphogenesis^5,6,7. A quadruple pif mutant (lacking PIF1, PIF3, PIF4, and PIF5) exhibits a striking constitutively photomorphogenic phenotype in the dark, which indicates that these four PIF factors act redundantly to promote etiolated growth^8,9,10. In the dark, PIF proteins accumulate and directly regulate thousands of genes to maintain skotomorphogenesis^11,12,13. Upon illumination, the photoactived phytochromes trigger PIFs’ rapid phosphorylation and subsequently proteasome-mediated degradation^14,15,16,17, leading to a cascade of transcriptional changes to promote photomorphogenesis^12,18,19,20.

During seedling development, gibberellins (GAs) promote etiolated growth, thus showing the opposite effect to light on photomorphogenic programme²¹. Deficiency of GA (as in ga1-3 mutant) induces partial constitutive photomorphogenesis phenotype in darkness, resulting in a loss of apical hook, open cotyledons and shortened hypocotyls²¹. DELLA proteins are the key repressors of almost all GA responses²². There are five DELLA proteins in Arabidopsis GA INSENSITIVE (GAI), REPRESSOR OF ga1-3 (RGA), RGA-LIKE 1 (RGL1), RGL2 and RGL3, which have both distinct and overlap** functions^{22,23,24,25,26,27,28,29,30}. In darkness, constitutively photomorphogenic phenotypes of ga1-3 mutants can be almost fully suppressed by rga and gai null alleles, indicating that RGA and GAI are the two main DELLA members involved in GA-dependent repression of photomorphogenic growth in seedlings^21,31,32. When GA is present, GA receptor GID1 binds to the DELLAs to form GID1–GA–DELLA complex, which triggers the ubiquitination and subsequent degradation of DELLA proteins by the 26S proteasome^33,34,35. DELLAs have a conserved DELLA domain at the N terminus that is essential for GA-triggered protein degradation²². Deletion of the DELLA domain results in stabilization of these proteins and leads to a GA-unresponsive dwarf phenotype²².

How light and GAs antagonistically regulate hypocotyl elongation has been extensively analyzed in the past two decades. Two independent studies demonstrated that DELLAs can physically interact with and block PIF3 and PIF4 activities by sequestering the transcription factors from binding to their targets, which ultimately results in inhibition of hypocotyl elongation^36,37. Further study showed that DELLAs also interact with the bHLH proteins PIF1 (also known as PIL5), PIF6 (also known as PIL2) and SPATULA(SPT)³⁸. The sequestration of PIFs by DELLAs provided an important molecular link between light and GA signalling in regulating photomorphogenesis. Later, the model of sequestration was demonstrated to be a general mechanism for DELLAs to regulate transcription factors involved in other signalling pathways^{39,40,41,42,40, 41, 42, 43, 44). These results suggest that sequestration is a general mechanism by which DELLAs inhibit transcription factors. Among these transcription factors, BZR1 was the only one for which DELLA was reported to promote a reduction in protein levels41. Here, we showed that DELLAs can promote the degradation of PIF1, PIF3, PIF4 and PIF5. Extending from this finding, we propose that controlling protein abundance of transcription factors represents another general mechanism of how DELLAs may regulate their targets. To test this hypothesis, we examined EIN3 from the ethylene signalling pathway. A previous study showed that DELLAs sequester EIN3, but GA₃ did not markedly affect EIN3 protein abundance42. Consistently, GA treatments did not noticeably affect the abundance of EIN3-GFP and EIN3-FLAG proteins in our experiment; however, PAC treatments dramatically reduced their abundance (Supplementary Fig. 9a,b). It is possible that EIN3 accumulates to a high level in darkness, and is not sensitive to exogenous GA, which decreases DELLA levels but could not further increase EIN3 abundance. On the other hand, highly elevated levels of DELLAs after PAC treatment may promote EIN3 degradation. Taken together, our data suggests that the inhibition of transcription factors by DELLAs is mediated through both sequestration and degradation.}

Methods

Plant materials and growth conditions

The wild-type Arabidopsis ecotypes used in this study were Columbia-0 (Col-0), Landsberg erecta (Ler) and Wassilewskija (WS). The mutants and transgenic lines were described previously: pif3-3 (ref. 4), 35S: PIF1-Myc, 35S: PIF3-Myc, 35S: PIF4-Myc and 35S: PIF5-Myc (ref. 48), rga-24 (ref. 25), rga-Δ17 (ref. 26), della (gai-t6 rga-t2 rgl1-1 rgl2-1 rgl3-1) (ref. 37), sly1-10 gai-t6, sly1-10 rga-24, sly1-10 gai-t6 rga-24 (ref. 34), phyB-9 (ref. 51), lrb123 (ref. 16), cop1-4 (ref. 52), det1-1(ref. 53), 35S: EIN3-GFP/ein3 eil1 (ref. 54) and 35S: EIN3-FLAG/ein3 eil1 (ref. 55). Since homozygous rga-Δ17 is sterile²⁶, the homozygous rga-Δ17 seedlings segregated from the progenies of heterozygous rga-Δ17 were used for experiments. gai (Col background) is a gift from **angdong Fu of IGDB. All seeds were surface-sterilized and sown on Murashige and Skoog (MS) medium containing 1% sugar. Seeds were cold-treated at 4 °C for 4 days in the dark before germination. The seedlings were grown in the dark, under continuous red light (0.5 μmol m⁻²s⁻¹) or under short-day conditions (8 h white light (85 μmol m⁻²s⁻¹) +16 h dark) for the indicated times, unless indicated otherwise. Manipulation of seedlings in darkness was performed under dim green light.

Generation of transgenic plants

DNA fragments containing two deletion mutants (RGAΔ17 and GAIΔ17) without stop codons were amplified and inserted into the XmaI and PstI restriction sites of a pBSK-derived plasmid containing triple HA-tag to make constructs pBSK-RGAΔ17-HA and pBSK-GAIΔ17-HA. Further, the fragments RGAΔ17-HA and GAIΔ17-HA were amplified from pBSK-RGAΔ17-HA or pBSK-GAIΔ17-HA, and these fragments were digested with the SalI and SpeI, and then ligated into the XhoI and SpeI digested binary vector pTA7002 (ref. 56) to generate pTA7002-RGAΔ17-HA and pTA7002-GAIΔ17-HA constructs. These binary constructs were introduced into the GV3101 strain of Agrobacterium and transformed into della mutant plants, using the floral dip transformation method⁵⁷. The transformants were selected on MS medium containing 50 mg ml⁻¹ hygromycin B (Sigma-Aldrich), and named as RGAΔ17-HA and GAIΔ17-HA, respectively. The primers are listed in Supplementary Table 2.

Plant treatments

GA₃ (GA), paclobutrazol (PAC) and DEX were dissolved in ethanol. The proteasome inhibitor MG132 was dissolved in DMSO. For continuous GA₃ and PAC treatments, after the seeds were induced by white light for germination on regular MS medium, they were transferred to medium containing GA₃ (10 μM), PAC (0.5 μM) and ethanol (0.01% (v/v), as control), and then grown in the dark for 4 days. For the transient treatment by GA₃, the seedlings were collected and vacuum-infiltrated with liquid MS medium containing 100 μM GA₃ or ethanol alone (as control) for 10 min and then kept immersed in the same solution for the indicated times. For continuous DEX treatment, Arabidopsis seedlings were grown on the MS medium containing 1 μM DEX or ethanol alone (as control). For the transient induction by DEX, the 4-day-old seedlings were vacuum-infiltrated with liquid MS medium containing 10 μM DEX or ethanol alone (as control) for 10 min and then kept immersed in the same solution for the indicated times. For MG132 treatment, the 4-day-old seedlings were vacuum-infiltrated with liquid MS medium containing 100 μM MG132 (dissolved in DMSO) or DMSO alone (as control) for 10 min and kept immersed in the same solution for 4 h unless indicated otherwise.

Protein extraction and Immunoblots

Total proteins were extracted by homogenizing seedlings using denaturing buffer (100 mM NaH₂PO4, 10 mM Tris–HCl pH 8.0, 8 M urea, 1 mM phenylmethylsulfonyl fluoride, and × 1 complete protease inhibitor mixture (Roche). Seedlings were growing in the dark for 4 days unless specific indications. Extracts were centrifuged at 16,000g for 10 min at 4 °C, and protein concentration in the supernatants was quantified by the Bradford assay. Aliquots of denatured total protein were separated on 8% SDS–PAGE gels and transferred to PVDF membranes. Anti-PIF3 purified antibody at 1:500 (v/v) dilution (ref. 48), anti-Myc polyclonal antibody (Sigma-Aldrich, Cat. No: C3956) at 1:1,000 (v/v) dilution, anti-HA antibody (Sigma-Aldrich, Cat. No: H9658) at 1:1,000 (v/v) dilution, anti-RGA antibody (Agrisera, Cat. No: AS111630) at 1:1,000 (v/v) dilution, anti-FLAG (Sigma-Aldrich, Cat. No: F3165) at 1:1,000 (v/v) dilution, anti-GFP (Abmart, Cat. No: M20004) at 1:1,000 (v/v) dilution, anti-RPN6 polyclonal antibody (ref. 58) at 1:2,000 (v/v) dilution, and anti-PRT5 polyclonal antibody (ref. 58) at 1:2,000 (v/v) dilution were used as primary antibodies. Each experiment was repeated at least three times, and one representative result was shown. Quantification results of immunoblots in Figs 3f, 4d, 6b, 6d-6f, were quantified by Image J software. Supplementary Figures 10–22 are original full versions of the immunoblot images.

RNA extraction and quantitative RT-PCR

Total RNA was extracted using the RNeasy plant mini kit (Qiagen). cDNA was synthesized by ReverTra Ace qPCR RT Master Mix (TOYOBO). The quantitative RT-PCR analysis was performed using SYBR Premix Ex Taq (Takara) in an ABI 7500 fast real-time instrument. Each experiment was repeated with three biological samples, and RT-PCR reactions were performed with three technical replicates for each sample. The primers are listed in Supplementary Table 2.

Hypocotyl lengths measurement

After the indicated times of growth and treatment, at least 30 seedlings were laid on the agar plates, and digital pictures were taken. Then, the hypocotyl lengths were measured using Image J software.

In vitro pull down assays

The constructs for expressing His-PIF1, His-PIF3, GST-PIF4 and GST-PIF5 were described previously⁴⁸. Full-length RGA fragment was inserted into pMal-C2X vector to fuse with maltose-binding protein (MBP). All constructs were expressed in E. coli strain BL21 under the induction of 1 mM IPTG (isopropyl-β-D-thiogalactopyranoside). Two micrograms MBP or MBP–RGA proteins were mixed with 2 μg His- or GST-tagged proteins in 500 μl binding buffer (20 mM Tris–HCl, pH 7.5, 150 mM NaCl and 0.1% Nonidet P-40), and the mixture was rotated at 4 °C for 2 h. The amylase agarose beads were washed with binding buffer for three times and then added into the mixture. Then, the mixture was rotated at 4 °C for another 2 h. After being washed for five times with binding buffer, the MBP resin was boiled with protein loading buffer and analyzed by immunoblots. Anti-His (Sigma-Aldrich, Cat. No: H1029-.2ML) at 1:1,000 dilution, anti-GST (Sigma-Aldrich, Cat. No: G1160-.2ML) at 1:5,000 dilution and anti-MBP (New England Biolabs, Cat. No: E8032S) at 1:5,000 dilution were used for the western blots.

Bimolecular fluorescence complementation assay

The full-length cDNA of RGA was amplified and inserted into the SpeI and BamHI sites of pSY736 (YFP^N) vector, resulting in plasmid YFP^N-RGA. YFP^C-PIF1, YFP^C-PIF3, YFP^C-PIF4 and YFP^C-PIF5 plasmids were described previously⁴⁸. The plasmids were extracted and concentrated to 2 mg ml⁻¹. The particle-mediated transformation using onion epidermal cells was performed⁵⁹. After 24 h of incubation, YFP signal was detected using a Zeiss LSM 710 confocal microscope. The primers used for plasmid construction were listed in Supplementary Table 2.

Chromatin immunoprecipitations assay

ChIP assays were performed as described previously³⁷. 35S:PIF3-Myc seedlings were grown for 4 days on the MS medium containing 0.5 μM PAC or EtOH in darkness, and were collected and treated with DMSO or 100 μM MG132 for 4 h. RGAΔ17-HA seedlings were grown for 4 days in the dark, and were collected and infiltrated with or without 10 μM DEX for 24 h. The samples (2 g) were treated with 15 ml of 1% formaldehyde under vacuum infiltration for 15 min, and then 1 ml 2 M glycine was added to stop crosslinking for 5 min. For the ChIP analysis used 35S: PIF3-Myc seedlings, the solubilized chromatin was immunoprecipitated by 30 μl EZview Red Anti-c-Myc Affinity Gel (Sigma-Aldrich, Cat. No: E6654) at 4 °C for 5 h. For the ChIP analysis using RGAΔ17-HA seedlings, the solubilized chromatin was mixed with 10 μl anti-HA antibody (Sigma) and incubated at 4 °C for 1 h. Then, 40 μl Dynabeads Protein G (Life Technologies, Cat. No: 10003D) was added, and the sample was incubated overnight at 4 °C. The coimmunoprecipitated DNA was recovered and analyzed by quantitative PCR. All primers used in ChIP assays were listed in Supplementary Table 2.

Data availability

The authors declare that all data supporting the findings of this study are available within the article and its Supplementary Information files or are available upon request from the corresponding authors.