The cold signaling attenuator HIGH EXPRESSION OF OSMOTICALLY RESPONSIVE GENE1 activates FLOWERING LOCUS C transcription via chromatin remodeling under short-term cold stress in Arabidopsis

Abstract

Exposure to short-term cold stress delays flowering by activating the floral repressor FLOWERING LOCUS C (FLC) in Arabidopsis thaliana. The cold signaling attenuator HIGH EXPRESSION OF OSMOTICALLY RESPONSIVE GENE1 (HOS1) negatively regulates cold responses. Notably, HOS1-deficient mutants exhibit early flowering, and FLC expression is suppressed in the mutants. However, it remains unknown how HOS1 regulates FLC expression. Here, we show that HOS1 induces FLC expression by antagonizing the actions of FVE and its interacting partner histone deacetylase 6 (HDA6) under short-term cold stress. HOS1 binds to FLC chromatin in an FVE-dependent manner, and FVE is essential for the HOS1-mediated activation of FLC transcription. HOS1 also interacts with HDA6 and inhibits the binding of HDA6 to FLC chromatin. Intermittent cold treatments induce FLC expression by activating HOS1, which attenuates the activity of HDA6 in silencing FLC chromatin, and the effects of intermittent cold are diminished in hos1 and fve mutants. These observations indicate that HOS1 acts as a chromatin remodeling factor for FLC regulation under short-term cold stress.

Figure 1.

HOS1 Regulates FLC Transcription at the Chromatin Level. (A) Isolation of hos1 mutants. In the hos1-3 and hos1-5 mutants, one copy of the T-DNA element is inserted in the 5th and 7th exons of the HOS1 gene, respectively (top panel). Black boxes represent exons, and white boxes represent untranslated regions. Black arrows indicate the primer pair used for RT-PCR. HOS1 expression was examined by RT-PCR using total RNA samples extracted from 10-d-old whole plants grown on MS-agar plates (bottom panel). The eIF4A gene was used as the RNA quality control. (B) Flowering times of hos1 mutants. Five-week-old plants grown in soil under LDs were photographed (left panel). The leaf numbers of 20 plants were averaged for each plant genotype and statistically treated using a Student’s t test (*P < 0.01) (right panel). Bars indicate se. (C) Structure of the FLC gene. The sequence regions marked by P1 to P5 indicate regions examined in the ChIP assays. Numbers indicate residue positions relative to the transcription start site. Black boxes indicate exons, and white boxes denote untranslated regions. (D) to (F) Relative levels of histone modifications in FLC chromatin. ChIP assays were performed using anti-H3Ac (D), anti-H3K4 trimethylation (E), and anti-H3K27Me3 (F) antibodies. Plants grown on MS-agar plates for 10 d under LDs were used for chromatin preparation. An eIF4A DNA fragment was used for normalization. Four measurements were averaged for each plant genotype and statistically treated using a Student’s t test (*P < 0.01). Bars indicate se. [See online article for color version of this figure.]

Figure 2.

Chromatin Binding of HOS1 Induces FLC Expression under Cold Stress. For the ChIP assays, biological triplicates were averaged for each measurement and statistically treated using a Student’s t test (*P < 0.01). Bars indicate se. (A) ChIP assays of HOS1 binding to FLC chromatin using MYC-specific antibody. ChIP assays were conducted as described in Figure 1, using an anti-MYC antibody. The HOS1-ox transgenic plants grown on MS-agar plates at 23°C for 10 d were further grown at either 23°C or 4°C for 2 d before harvesting plant materials. The genomic DNA sequences examined in the ChIP assays were identical to those described in Figure 1C. Transgenic plants harboring the 6xMYC-pBA vector alone were used as control plants. (B) Schematic illustration of anti-HOS1 antibody production. A peptide region covering residues 796 to 810 of HOS1 protein was selected for the epitope (black box). Nuclear proteins extracted from Col-0 plants and hos1-3 mutants were used for examination of anti-HOS1 antibody specificity. HOS1 proteins were immunologically detected using the anti-HOS1 antibody. Arrow indicates endogenous HOS1 protein. Parts of Coomassie blue–stained gels are shown at the bottom as loading control. aa, amino acids. (C) ChIP assays of HOS1 binding to FLC chromatin using HOS1-specific antibody. ChIP assays were conducted as described in Figure 1, using an anti-HOS1 antibody. Chromatin extracts from Col-0 and hos1-3 plants were used for the ChIP assays. The conditions for the cold treatments were identical to those described in (A). (D) ChIP assays using HOS1_pro:HOS1-MYC transgenic plants. ChIP assays were performed using an anti-MYC antibody as described in Figure 1. The conditions for the cold treatments were identical to those described in (A). Transgenic plants harboring the 6xMYC-pBA vector were used as control plants. (E) Effects of intermittent cold on FLC expression. Plants treated with intermittent cold for 6 h at dawn for 15 d after germination were used for extraction of total RNA. Levels of FLC mRNA were determined by qRT-PCR. Biological triplicates were averaged for each plant genotype and statistically treated using a Student’s t test (*P < 0.01). Bars indicate se. (F) Effects of intermittent cold on flowering time. Plants were treated with intermittent cold for 6 h at dawn, each day until flowering. Rosette leaves of 15 plants were counted for each plant genotype and statistically treated using a Student’s t test (*P < 0.01). Bars indicate se.

Figure 3.

HOS1 Interacts with FVE in the Nucleus. (A) Interactions of HOS1 with FVE in yeast cells. Autonomous flowering pathway genes and the HOS1 gene were fused in frame to the GAL4 activation domain (AD)–coding and GAL4 DNA binding domain (BD)–coding sequences, respectively. The HOS1–FVE interactions were assayed by cell growth on selective medium. –LWHA indicates Leu, Trp, His, and Ade dropout plates. –LW indicates Leu and Trp dropout plates. P, positive control. (B) Deletion constructs of HOS1 and FVE. Numbers indicate residue positions. Black box, REALLY INTERESTING NEW GENE finger domain. WD, WD40 domain. aa, amino acids. (C) and (D) Interactions of HOS1 with FVE in yeast cells. The HOS1 (C) or FVE (D) constructs were fused with GAL4 BD or GAL4 AD, respectively. Yeast two-hybrid assays were performed as described in (A). (E) In vitro pull-down assays. Recombinant MBP-FVE fusion and in vitro–translated ³⁵S-labeled HOS1 polypeptides were used (top panel). Part of a Coomassie blue–stained gel is shown (bottom panel). The input represents 5% of the radiolabeled protein used in the assays. (F) Coimmunoprecipitation assays. Total proteins were extracted from transgenic plants overexpressing a FVE-MYC gene fusion (FVE-ox) grown on MS-agar plates for 6 d. Protein complexes were immunoprecipitated using an anti-MYC antibody and detected immunologically using an anti-MYC or anti-HOS1 antibody. Transgenic plants harboring the 6xMYC-pBA vector alone were used as control plants. (G) BiFC assays. The YFP^N-FVE and HOS1-YFP^C constructs were coexpressed transiently in Arabidopsis protoplasts and visualized by differential interference contrast microscopy and fluorescence microscopy. Bar = 20 μm.

Figure 4.

HOS1 Regulation of FLC Transcription Requires FVE. Whole plant materials grown on MS-agar plates at 23°C under LDs for 10 d were used for ChIP assays. Biological quadruplicates were averaged for each measurement and statistically treated using a Student\x{2019}s t test (*P < 0.01). Bars indicate se. (A) Relative levels of H3Ac in FLC chromatin. ChIP assays were conducted using an anti-H3Ac antibody. (B) Binding of HOS1 to FLC chromatin in the fve-4 mutant. ChIP assays were conducted using an anti-HOS1 antibody.

Figure 5.

HOS1 Interacts with HDA6 in the Nucleus. (A) Effects of TSA on FLC expression. Plants were grown for 10 d on MS-agar plates containing 0.5 μM TSA. The levels of FLC mRNA were determined by qRT-PCR. Biological triplicates were averaged for each plant genotype and statistically treated (Student’s t test, *P < 0.01). Bars indicate se. (B) Interactions of HOS1 with HDA6 and HDA15 in yeast cells. The HDA6 and HDA15 genes were fused in frame to the GAL4 AD-coding sequence. The HOS1 constructs used were as described in Figure 3B. (C) In vitro pull-down assays. Recombinant MBP-FVE and MBP-HOS1 fusions as well as in vitro–translated, ³⁵S-labeled HDA6 and HDA15 polypeptides were used (top and middle panels, respectively). Part of a Coomassie blue–stained gel is shown (bottom panel). (D) BiFC assays. Partial YFP fusion constructs were coexpressed transiently in Arabidopsis protoplasts and visualized by differential interference contrast microscopy (DIC) and fluorescence microscopy. Bars = 20 μm. (E) and (F) in vivo interaction of HDA6 with FVE (E) and HOS1 (F). Total proteins extracted from the FVE-ox and HOS1_pro:HOS1-MYC transgenic plants grown for 6 d were immunoprecipitated with anti-MYC agarose beads. The FVE, HOS1, and HDA6 proteins were detected immunologically using an anti-MYC or anti-HDA6 antibody.

Figure 6.

Cold Stress Attenuates Binding of HDA6 to FLC Chromatin. (A) Effects of cold stress on HDA6 protein stability. Ten-day-old transgenic plants overexpressing a HDA6-MYC gene fusion under the control of the CaMV 35S promoter (HDA6-ox) were exposed to 4°C for the indicated time. Whole plants were harvested for preparation of protein extracts. The levels of HDA6 were compared immunologically using an anti-MYC antibody. Parts of Coomassie blue–stained gels containing the ribulose-1,5-bis-phosphate carboxylase/oxygenase protein are shown at the bottom. (B) Effects of cold stress on HDA6 ubiquitination. Ten-day-old HDA6-ox transgenic plants were used in the ubiquitination assays. Total protein was immunoprecipitated with an anti-MYC antibody and analyzed immunologically using anti-MYC and anti-ubiquitin antibodies. Transgenic plants overexpressing the ICE1-MYC gene fusion (ICE1-ox) were also analyzed for comparison. IP, immunoprecipitation assay. WB, protein gel blot analysis. (C) Binding of HDA6 to FLC chromatin. The HDA6-ox transgenic plants were grown on MS-agar plates for 2 weeks before harvesting of plant materials. ChIP assays were performed using an anti-MYC antibody, as described in Figure 1. Biological triplicates were averaged and statistically treated using Student’s t test (*P < 0.01). Bars indicate se. (D) Relative levels of HDA6 protein and HDA6 mRNA in HDA6-ox transgenic plants in different genetic backgrounds. The mRNA levels were examined by qRT-PCRs (left panel). Biological triplicates were averaged and statistically treated (t test, *P < 0.01). Bars indicate se. The levels of HDA6 were determined immunologically using an anti-MYC antibody (right panel). (E) Effects of cold stress on HDA6 binding to FLC chromatin. ChIP assays were performed as described in Figure 1, using an anti-MYC antibody. Chromatin extracts from plant materials described in (D) were used. Ten-day-old plants grown on MS-agar plates at 23°C were further grown at either 23 or 4°C for 2 d before harvesting of plant materials. The P1 sequence region of the FLC chromatin, as shown in Figure 1C, was used in the ChIP assays. Biological triplicates were averaged and statistically treated (t test, *P < 0.01). Bars indicate se. (F) Binding of HDA6 to FLC chromatin in the hos1-3 and fve-4 mutants. ChIP assays were performed as described in Figure 1, using an anti-HDA6 antibody. Chromatin extracts from Col-0 plants and hos1-3 and fve-4 mutants were used. Conditions for cold treatments were identical to those described in (E). Biological triplicates were averaged and statistically treated (t test, *P < 0.01). Bars indicate se. (G) Effects of intermittent cold on flowering time. Intermittent cold treatments and measurements of flowering times were performed as described in Figure 2. The numbers in parentheses refer to the ratios of rosette leaf numbers with and without intermittent cold treatments.

Figure 7.

Model of HOS1 Function in Arabidopsis Flowering. Under normal conditions, the FVE-HDA6 complex silences FLC chromatin. Under cold stress, HOS1 strongly binds to FLC chromatin and interacts with FVE and HDA6 to dissociate HDA6 from FLC chromatin, resulting in the activation of FLC chromatin and delayed flowering. [See online article for color version of this figure.]