Arabidopsis inositol polyphosphate multikinase delays flowering time through mediating transcriptional activation of FLOWERING LOCUS C.

Timely flowering is critical for successful reproduction and seed yield in plants. A diverse range of regulators have been found to control flowering time in response to environmental and endogenous signals. Among these regulators, FLOWERING LOCUS C (FLC) acts as a central repressor of floral transition by blocking the expression of flowering integrator genes. Here, we report that Arabidopsis inositol polyphosphate multikinase (AtIPK2β) functions in flowering time control by mediating transcriptional regulation of FLC at the chromatin level. The atipk2β mutant flowers earlier, and AtIPK2β overexpressing plants exhibit late-flowering phenotypes. Quantitative reverse transcription-PCR (qRT-PCR) revealed that AtIPK2β promotes FLC expression. We performed chromatin immunoprecipitation-qPCR (ChIP-qPCR) assays and found that AtIPK2β binds to FLC chromatin. Further analysis showed that AtIPK2β interacts with FVE, a key repressor required for epigenetic silencing of FLC. qRT-PCR, ChIP-qPCR, and genetic analysis demonstrated that AtIPK2β is involved in FVE-mediated transcriptional regulation of FLC by repressing the accumulation of FVE on FLC. Moreover, we found that AtIPK2β associates with HDA6, an interaction partner of FVE mediating FLC chromatin silencing, and attenuates HDA6 accumulation at the FLC locus. Taken together, these findings suggest that AtIPK2β negatively regulates flowering time by blocking chromatin silencing of FLC.

Introduction

In Arabidopsis, timely flowering is ensured by an intricate regulatory network that has evolved in responde to environmental conditions and internal cues. The mainly genetic pathways are well defined, including photoperiod, vernalization, thermosensory, gibberellins (GA), age, and autonomous pathways (Blázquez ; Srikanth and Schmid, 2011; Wang, 2014). Among these pathways, the vernalization and autonomous pathways converge on FLOWERING LOCUS C (FLC), which encodes a MADS-box transcription factor that blocks flowering by transcriptional repression of floral integrator genes including FLOWERING LOCUS T (FT) and SUPPRESSOR OF OVEREXPRESSION OF CO 1 (SOC1) (Sheldon ; Simpson, 2004; Parcy, 2005; Lee and Lee, 2010). FLC acts as a floral repressor by antagonizing the activity of pathways promoting flowering in a dose-dependent manner. Its expression is regulated at the transcriptional and post-transcriptional levels by diverse factors (Amasino, 2010; He, 2012). Although previous investigations have revealed different regulators in the control of FLC expression, more studies are needed to provide better understanding of the complex regulation in FLC-mediated flowering.

Chromatin modification plays a crucial role in transcriptional regulation of FLC. Active modifications at FLC chromatin, such as histone H3 acetylation (H3Ac), lysine-4 methylation (H3K4me2/H3K4me3), lysine-36 methylation (H3K36me2/H3K36me3), and histone H2B monoubiquitination (H2Bub1), induce gene expression, whereas repressive modifications, such as histone deacetylation, histone lysine-9 methylation, and histone H3 lysine-27 trimethylation (H3K27me3), result in silencing of FLC (Choi ; He, 2012). Histone modification at the FLC locus is mediated by various multiprotein complexes, such as the Polycomb Repressive Complex 2 (PRC2)-like complex and histone deacetylase complexes, which silence FLC by depositing the repressive H3K27me3 mark or removing the active H3Ac mark at the FLC locus, and the FRIGIDA (FRI) complex, which activates FLC expression by accumulating the active H3Ac, H3K4me3, H3K36me2, and H3K36me3 marks on the chromatin (De Lucia ; Ko ; Choi ; He, 2012). A number of proteins functioning in the chromatin-regulating complexes have been identified in both the vernalization and autonomous pathways to mediate histone modifications at FLC (He, 2012; Simpson, 2004).

Among these proteins, FVE, a key component of the autonomous flowering pathway containing histone-binding and WD40-repeat motifs, is required for silencing FLC by mediating chromatin modifications on the gene locus (Jeon and Kim, 2011; Pazhouhandeh ). FVE is homologous to the mammalian retinoblastoma-associated proteins RbAp46/RbAp48 and is also known as Arabidopsis MULTICOPI SUPPRESSOR OF IRA1 4 (AtMSI4), which belongs to the MSI1-like protein family that is conserved in yeast and plants (Ausín ; Hennig ; Gu ). FVE associates with CULLIN4-Damaged DNA Binding Protein1 (CUL4-DDB1) and PRC2-like complexes, and is required for effective H3K27 trimethylation on both FLC and FT chromatin (Pazhouhandeh ). Additionally, FVE interacts with Histone Deacetylase 6 (HDA6) to mediate histone deacetylation and DNA methylation, resulting in silencing of the target genes (Gu ; Xu ). The complex formed by FVE, HDA6, and FLOWERING LOCUS D (FLD) represses FLC expression by reducing accumulation of the active H3K4me3 and H3Ac marks and promoting deposition of the repressive H3K27me3 mark (Yu ). Moreover, the interaction between FVE and HIGH EXPRESSION OF OSMOTICALLY RESPONSIVE GENE1 (HOS1), an E3 ligase that mediates protein degradation, blocks the function of HDA6 by inhibiting the binding of HDA6 to FLC chromatin in response to short-term cold stress, resulting in delayed floral transition (Jung ). Collectively, these studies have shown pleiotropic functions of FVE in the regulation of FLC expression at the chromatin level by interacting with different proteins.

Inositol polyphosphate multikinases (IPMKs) play important roles in inositol phosphate metabolism and signal transduction. The conserved catalytic activities of IPMKs are phosphorylating inositol 1,4,5-triphosphate (IP3) to generate inositol 1,4,5,6-tetrakisphosphate (IP4) and inositol 1,3,4,5,6-pentakisphosphate (IP5) (Chang and Majerus, 2006; Resnick and Saiardi, 2008). IPMKs and their products are involved in the regulation of gene expression. IP4 and IP5 produced by yeast IPMK (also known as IPK2) act as stimulators in modulating the activities of chromatin-remodeling complexes, such as INO80 and SWI/SNF, resulting in activation of target genes (El Alami ; Steger ). IP4 also functions as a key component of mammalian class I histone deacetylase (HDAC) complexes (Millard ). Independent of their catalytic activities, IPMKs function as co-activators in transcriptional regulation. The yeast IPMK/IPK2 interacts with a MADS-box protein, Mcm1, and stabilizes the complex Mcm1-ArgR, which recognizes the specific DNA sequences termed the ‘arginine box’ in the promoter region and activates the target gene (El Bakkoury ; Bosch and Saiardi, 2012). In mammalian cells, IPMK interacts with histone acetyltransferase CBP and histone acetyltranferase complex p53-p300 to mediate transcriptional activation (Xu ; Xu and Snyder, 2013; Kim ). Moreover, a class II HDAC identified in Tetrahymena contains an IPMK domain homologous to IPK2 (Smith ). AtIPK2β, an IPMK identified in Arabidopsis, exhibits conserved 6-/3-kinase activity and is homologous to yeast IPK2, as expressing AtIPK2β in the IPK2 deletion yeast partially compensates its growth defects (Stevenson-Paulik ; Xia ; Bosch and Saiardi, 2012). A recent study showed that the kinase activity of AtIPK2β is essential for its function in plant sexual reproduction, including in pollen development, pollen tube guidance, and embryogenesis (Zhan ). Several investigations have also suggested a potential role of AtIPK2β in transcriptional regulation. AtIPK2β is enriched in the nucleus and regulates the expression of auxin-responsive genes (Xia ; Zhang ). Moreover, ectopic expression of AtIPK2β in tobacco enhances the tolerance of transgenic plants to abiotic stresses through promoting the transcription of stress-responsive genes (Yang ). However, the molecular mechanism through which AtIPK2β mediates transcriptional regulation is still unknown.

In this study, we report the function of AtIPK2β in flowering time control via mediating the transcriptional regulation of FLC. AtIPK2β acts as a negative regulator of flowering time by promoting the expression of FLC. Chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) analysis revealed that AtIPK2β mediates transcriptional regulation of FLC via binding to FLC chromatin. In addition, AtIPK2β interacts with FVE and represses the accumulation of FVE at the FLC locus. Gene expression, genetic, and ChIP-qPCR assays confirmed that AtIPK2β is involved in FVE-mediated transcriptional regulation of FLC by affecting chromatin modifications including histone H3K27 trimethylation and H3 deacetylation at FLC. Moreover, AtIPK2β associates with HDA6, an interaction partner of FVE, and attenuates the accumulation of HDA6 on FLC chromatin. In general, these results suggest that AtIPK2β functions as a repressor in flowering time control through promoting FLC expression at the chromatin level.

Materials and methods

Plant materials

All the plants used in this study were Arabidopsis thaliana Columbia type (Col-0) background. The atipk2β-1 (SALK_025091) and atipk2β-2 (SALK_104995) mutant lines were described previously by Zhang and Zhan . The flc-6 null mutant, obtained from the Arabidopsis Biological Resource Center, was originally described by Bouveret . The fve T-DNA line (SALK_013789) was provided by Dr Ligeng Ma, and genotyping PCR was performed for identifying the T-DNA insertion. Double mutants of flc/atipk2β and fve/atipk2β were generated by crossing the fve and flc-6 lines, respectively, with atipk2β-2. The homozygous double mutants of F3 progenies were selected by genotyping PCR and used in this study. The primers used in identification of the mutants, with detailed sequences, are listed in Supplementary Table S1 at JXB online.

The complementary lines of the atipk2β mutant were generated by introducing a construct encoding an AtIPK2β-Green Fluorescent Protein (GFP) fusion protein under the control of the native promoter into atipk2β-2, as described by Zhan . For overexpression transgenic lines, the AtIPK2β or FVE coding sequence was inserted into the pCAMBIA1302 vector after the CaMV 35S promoter region and followed by the coding sequence of GFP. Col-0 and the atipk2β-2 mutant plants were agro-transformed with the resulting vectors using the floral dip method (Clough and Bent, 1998), and T3 generation plants selected with hygromycin B were used in further analyses. HDA6-MYC/atipk2β overexpression plants were generated by crossing the atipk2β mutant with a HDA6-MYC transgenic line, which is in the Col-0 background and was provided by Dr Keqiang Wu (Liu ). The F3 progenies of HDA6-MYC/atipk2β were identified and used in ChIP-qPCR assays.

Flowering time analysis

Plants were grown on soil in a controlled culture room at 22 °C with cold fluorescent light (100 μmol m−2 s−1) under long-day (LD; 16 h light) or short-day (SD; 8 h light) photoperiods. Seeds were stratified for 4–5 days at 4 °C on Murashige and Skoog (MS) agar plates before being transferred on to soil. Flowering time was measured by counting the number of rosette leaves and the number of days at bolting after germination. At least 10 plants were counted and averaged for each genotype.

Gene expression analysis

For analyzing gene expression levels, seeds were sterilized with ethanol and then stratified for 4–5 days at 4 °C on solid MS agar plates. The plates were kept in a controlled culture room at 23 °C with cold fluorescent light (100 μmol m−2 s−1) under the LD or SD photoperiod for 15 to 20 days before being harvested. RNA was extracted using a TRIzol (Molecular Research Center) method and treated with DNase I (Promega) before being transcribed into cDNA by M-MLV Reverse Transcriptase (Promega). Quantitative reverse transcription-PCR (qRT-PCR) was performed with gene-specific primers (listed in Supplementary Table S1) on an ABI Step-one Plus real-time PCR system using SYBR Green Realtime PCR Master Mix (TOYOBO). The quantification results were normalized to two internal reference genes (ACT2 and UBQ10) and then compared with the wild-type (WT) control plants. Data are presented as means±SD of at least two independent repeats with similar results.

Yeast two-hybrid assays

The pGBKT7 and pGADT7 vectors of the BD Matchmaker system (Clontech) and the yeast strain AH109 were used for yeast two-hybrid assays according to the manufacturer’s instructions. The full-length coding sequence of FVE or HDA6 was subcloned into pGADT7, and AtIPK2β was inserted into pGBKT7. The final constructs were transformed into yeast strain AH109. Transformants were spotted on to plates of selective medium without Leu and Trp (DDO), and further on medium without Leu, Trp, His, and Ade (QDO). Plates were incubated at 30 °C for 3–5 days. The filter β-galactosidase assays were conducted according to the system protocol (Clontech).

Bimolecular fluorescence complementation assays

Arabidopsis protoplasts from rosette leaves of plants grown for 3 weeks under the LD photoperiod were isolated as described by Yoo . The coding sequence of AtIPK2β or FVE was inserted between the CaMV 35S promoter and terminator regions and fused to the sequences encoding the N-terminal fragment or C-terminal region of Yellow Fluorescent Protein (YFP) to form the YN-AtIPK2β or YC-FVE fusion, respectively. The DNA fragments including the CaMV 35S promoter, coding sequence of YN-AtIPK2β or YC-FVE fusion, and the terminator regions were amplified by PCR and transformed into the protoplasts according to the method described previously by Lu . An irrelevant YC-SnRK2.6 fusion employed for testing the efficiency of the bimolecular fluorescence complementation (BiFC) system by Lu was used as a negative control, as well as empty YN or YC constructs. Fluorescence of YFP was visualized under an Olympus FV 1000 confocal microscope within 18 hours after transformation.

Immunoblot assays

Arabidopsis rosette leaves of indicated genotypes grown on MS agar plates for 15 days under the LD photoperiod were homogenized in extraction buffer containing 50 mM Tris-HCl (ph 7.5), 100 mM NaCl, 10% glycerol, 0.1% Nonidet P-40, 1 mM phenylmethylsulfonyl fluoride, and 1× complete protease inhibitor (Roche). Total protein was extracted by centrifugation at 4 °C. The protein concentration was determined by using a Bradford assay kit (BCA, Thermo Scientific). Western blot was performed by blotting protein samples on polyvinylidene fluoride membranes (Millipore) after being boiled and run on SDS-PAGE gels. Different primary antibodies were used for probing the blots. A polyclonal AtIPK2β antibody was raised in rabbits as described by Yang .

Co-immunoprecipitation assays

Co-immunoprecipitation (Co-IP) assays were performed as described previously (Chen ). Approximately 1–2 mg total proteins extracted from leaves of transgenic Arabidopsis were incubated with an anti- GFP antibody or anti-AtIPK2β antibody overnight at 4 °C, and then Protein A magnetic beads (NEB) were added and incubated for another 2 hours. Subsequently, the beads were washed according to the immunoprecipitation protocol (NEB) and the eluted precipitates were analyzed by western blot using anti-MYC, anti-GFP, or anti-AtIPK2β antibodies.

Chromatin immunoprecipitation-quantitative PCR assays

ChIP experiments were performed according to the method described previously (Saleh ). Plants 10–15 days old grown on MS agar plates under a LD photoperiod were collected for chromatin extraction. Antibodies used in ChIP assays included anti-GFP, anti-AtIPK2β, anti-MYC, anti-trimethylated-H3 (Lys 27) (Millipore, 07-449), and anti-acetyl-H3 (Abcam, ab47915). The procedure and data calculation of ChIP-qPCR were conducted as described previously (Yamaguchi ). DNA fragments of ACTIN2 (ACTIN) and FUSCA3 (FUSCA) were used for normalization (ACTIN for H3Ac and FUSCA for H3K27me3) in the ChIP-qPCR assays to detect enrichments of H3Ac and H3K27md3 marks on FLC as described previously (Gu ; Pazhouhandeh ). ACTIN was also used for normalization in the ChIP-qPCR assays conducted to examine the accumulation of AtIPK2β, AtIPK2β-GFP, FVE-GFP, and HDA6-MYC on FLC, as the indicated proteins were not found to be enriched in this region at the ACTIN2 locus. Primers used in the ChIP-qPCR and their detailed sequences are listed in Supplementary Table S2.

Results

AtIPK2β delays flowering time

The AtIPK2β gene encodes a conserved IPMK in higher plants with diverse functions in growth and development, including abiotic stress response, axillary branching, and sexual reproduction (Stevenson-Paulik ; Xia ; Zhang ; Yang ; Zhan ). However, it is unclear whether AtIPK2β plays a role in flowering time regulation. We explored the possible function of AtIPK2β in the regulation of flowering time using the atipk2β mutant and AtIPK2β overexpression plants. Two atipk2β RNA-null mutant lines, atipk2β-1 and atipk2β-2, which were described previously (Zhang ), were identified again by detecting the expression of AtIPK2β (Supplementary Fig. S1A, B). We observed that both of the atipk2β mutant lines flowered earlier and had reduced numbers of rosette leaves at bolting compared with WT plants under LD growth conditions (Supplementary Fig. S1C). The atipk2β mutant we used in the following studies was the atipk2β-2 line. We next confirmed that the early flowering of the atipk2β mutant was due to loss of AtIPK2β expression by performing genetic complementation experiments. An AtIPK2β-GFP fusion construct under the control of the native promoter was introduced into the atipk2β mutant; both of the complemented lines (Com1 and Com2) exhibited similar or slightly delayed flowering compared with WT (Supplementary Fig. S1C).

As the atipk2β mutant plants flowered earlier under both LD and SD conditions (Fig. 1A, C), we further confirmed the function of AtIPK2β by analyzing the flowering phenotype of the AtIPK2β overexpression lines generated by overexpressing AtIPK2β-GFP in the atipk2β mutant background. Two of the AtIPK2β overexpression lines (named OE1 and OE2) were used for further analysis. The AtIPK2β overexpression lines exhibited delayed flowering (Fig. 1B) and produced significantly more rosette leaves at bolting under both LD and SD photoperiods (Fig. 1C). These results suggested that AtIPK2β acts as a negative regulator of flowering time independent of photoperiod.

Fig. 1.

AtIPK2β delays flowering time. (A) Flowering phenotype of wild-type (WT) and atipk2β mutant plants grown under long-day (LD) and short-day (SD) photoperiods. (B) Flowering phenotypes of WT and AtIPK2β overexpressing lines grown under LD and SD conditions. Two lines of transgenic plants (OE1 and OE2) are shown. (C) Number of rosette leaves of WT, atipk2β, and AtIPK2β overexpressing (OE1 and OE2) plants at bolting. Rosette leaf numbers of 10–15 plants at flowering were counted and averaged for each genotype. Each value for the atipk2β mutant and overexpressing lines was compared with WT separately. Statistically significant differences were determined by one-way ANOVA with Tukey’s multiple comparison test (**P<0.01, ***P<0.001). (This figure is available in colour at JXB online.)

AtIPK2β affects the expression of flowering-related genes

To investigate the possible molecular mechanism by which AtIPK2β regulates flowering time, four genes, CONSTANS (CO), FLC, FT, and SOC1, which encode key flowering-time regulators and floral integrators, were analyzed in the atipk2β mutant and AtIPK2β overexpression plants grown under both LD and SD conditions. The AtIPK2β overexpression line (OE1) exhibiting a late-flowering phenotype (Fig. 1B, C) was used in the gene expression analysis.

FLC, which encodes a key repressor of flowering, is transcriptionally regulated by multiple signaling pathways and factors (Srikanth and Schmid, 2011). To explore whether AtIPK2β regulates flowering by affecting the transcription of FLC, we examined the expression level of FLC in the atipk2β mutant and the AtIPK2β overexpression line OE1. This analysis showed that the level of FLC transcript decreased significantly in the atipk2β mutant and increased in OE1 compared with WT plants under both LD and SD photoperiods (Fig. 2A). Signaling pathways controlling flowering time converge on integrator genes including FT and SOC1, both of which function as floral activators and are repressed by FLC directly at the transcriptional level (Helliwell ; Lee and Lee, 2010; Srikanth and Schmid, 2011). We examined the expression levels of FT under both LD and SD growth conditions. As Fig. 2B shows, the transcription level of FT increased in the atipk2β mutant and decreased in AtIPK2β overexpressing plants, correlating inversely with the expression levels of FLC. Except when being repressed by FLC, SOC1 is transcriptionally activated by FT directly (Lee and Lee, 2010). Similar to FT, the expression of SOC1 was promoted in the atipk2β mutant and repressed in OE1 under both LD and SD photoperiods (Fig. 2C). These results indicate that AtIPK2β promotes FLC expression and thereby represses transcription of FT and SOC1.

Fig. 2.

Expression analysis of flowering-related genes. Relative expression levels of FLC (A), FT (B), SOC1 (C), and CO (D) in wild-type (WT), atipk2β mutant, and AtIPK2β overexpressing line (OE1) plants under long-day (LD) and short-day (SD) growth conditions. Plants of the indicated genotypes were grown under LD and SD photoperiods for 15 days and harvested at zeitgeber time (ZT) 7. The relative expression level of each gene was determined by qRT-PCR, followed by normalizing to ACT2 and UBQ10. Data presented are the mean±SD of two biological replicates.

We also examined the expression level of CO, which encodes a transcriptional activator of FT that responds to photoperiod (Suárez-López ). As AtIPK2β regulates flowering time in both LD and SD conditions, it may not be involved in the photoperiod pathway. In support of this hypothesis, the transcription level of CO was not affected in either atipk2β mutant or AtIPK2β overexpressing plants (Fig. 2D).

AtIPK2β interacts genetically with FLC

We further confirmed that AtIPK2β regulates FLC expression to control flowering time by investigating the genetic interaction between AtIPK2β and FLC. The double mutant flc/atipk2β was produced by crossing the atipk2β mutant with the flc-6 null mutant (Bouveret ). As the early-flowering phenotype of the flc-6 mutant under SD photoperiods is more significant than that under LD conditions (Steinbach and Hennig, 2014; Tsuchiya and Eulgem, 2010), we analyzed the flowering phenotypes of WT and atipk2β, flc-6, and flc/atipk2β mutants grown under SD conditions (Fig. 3A). All of the mutant plants flowered earlier than WT. Moreover, the flc/atipk2β and flc-6 mutant lines exhibited a similar flowering time, suggesting that the expression of AtIPK2β in flc-6 is unable to rescue the early-flowering phenotype of flc/atipk2β. We further examined the expression levels of FT and SOC1 in the same mutant lines. Although expression of AtIPK2β in WT resulted in decreased transcription of FT and SOC1 compared with the levels in atipk2β, flc-6 exhibited similar expression levels of the two genes as flc/atipk2β (Fig. 3B, C), demonstrating that the flowering phenotype of the double mutant is due to the loss of FLC. These results indicated that the functions of AtIPK2β in regulating flowering time may require FLC.

Fig. 3.

AtIPK2β interacts genetically with FLC. (A) Flowering time analysis of atipk2β, wild-type (WT), flc/atipk2β, and flc-6 plants grown under short-day conditions. Numbers of rosette leaves of 10–15 plants at bolting were counted and averaged. Data are presented as means±SD. Data for WT were compared with atipk2β, and flc-6 with flc/atipk2β, using one-way ANOVA with Tukey’s multiple comparison test (***P<0.001; ns, not significant). (B, C) Expression level of FT (B) and SOC1 (C) in WT and mutant lines. 20-day-old plants grown under SD conditions were collected at zeitgeber time (ZT) 5 for detecting gene expression. Data presented are the means±SD of two biological repeats.

AtIPK2β binds to FLC chromatin in Arabidopsis

Next, we performed ChIP-qPCR assays to investigate the molecular mechanism by which AtIPK2β regulates the transcription of FLC. Eight sequence regions within the FLC locus (P1–P8; Fig. 4A) were examined by ChIP-qPCR. The OE1 line overexpressing the AtIPK2β-GFP fusion construct and an anti-GFP antibody were used for ChIP. A band of AtIPK2β-GFP of the expected size (61 kDa) was detected in extracts from the transgenic plants (Fig. 4B). Chromatins extracted from OE1 were immunoprecipitated with anti-GFP antibody and the precipitated DNA fragments after ChIP were quantified by qPCR. ChIP-qPCR assays revealed that AtIPK2β binds to FLC chromatin; among the regions of FLC examined, the enrichment level of AtIPK2β was higher in sequence regions P2, P6, and P7 (Fig. 4C). We further confirmed the enrichment of AtIPK2β at the FLC locus using an anti-AtIPK2β antibody in WT plants, because the ectopic expression of GFP-tagged AtIPK2β may cause potential artifacts. The specificity of the antibody was verified by performing immunological analysis using WT and the atipk2β mutant. Western blot analysis showed that a band of AtIPK2β of the expected size (33 kDa) was detected with anti-AtIPK2β antibody in protein extracts from WT plants, but no such band was detected in the extracts from the atipk2β mutant (Fig. 4D). The ChIP-qPCR results showed that AtIPK2β binds to FLC chromatin and the enrichment levels were higher in the P6 and P7 regions (Fig. 4E). These data suggest a possible regulatory mechanism through which AtIPK2β promotes FLC expression by mediating chromatin regulation of FLC.

Fig. 4.

Binding of AtIPK2β to FLC chromatin in plants. (A) Schematic diagram of the FLC gene. The sequence regions marked P1–P8 indicate the eight regions examined in the ChIP assays, and the numbers below indicate residue positions relative to the ATG start codon. The black boxes represent exons and lines between them represent introns. White boxes denote the 5ʹ and 3ʹ untranslated regions. The arrow indicates the transcription start site of FLC. (B) Immunodetection of AtIPK2β-GFP fusion protein (arrow) with anti-GFP antibody in OE1 plants. The large subunit of ribulose 1,5-bisphosphate carboxylase (Rub) stained with Ponceau S is shown as a loading control in the lower panel. The position and size of standard protein markers are indicated to the left. (C) AtIPK2β binds to FLC chromatin in AtIPK2β overexpressing plants. 15-day-old plants of the OE1 line overexpressing AtIPK2β-GFP grown under long-day (LD) conditions were collected and an anti-GFP antibody was used for chromatin immunoprecipitation (ChIP). An ACTIN2 DNA fragment (ACTIN) was used for normalizing the amount of precipitated DNA fragments quantified by qPCR. Fold enrichment levels of AtIPK2β-GFP in transgenic plants relative to the negative control [wild-type (WT)] are shown as the mean±SD of two biological repeats. (D) Immunodetection of AtIPK2β protein (arrow) with anti-AtIPK2β antibody in WT and atipk2β mutant plants. The positions and size of standard protein markers are indicated to the left. (E) Binding of AtIPK2β to the FLC locus in WT plants. ChIP-qPCR assays were performed using WT and the atipk2β mutant with an anti-AtIPK2β specific antibody, and the precipitated DNA fragments were quantified and then normalized to ACTIN. Plants were grown under LD conditions for 15 days and the atipk2β mutant was used as negative control. Fold enrichment levels of AtIPK2β at the FLC in WT relative to the atipk2β mutant are shown as the means±SD of two independent experiments.

AtIPK2β interacts with FVE in vivo

We further investigated how AtIPK2β regulates the transcription of FLC. IPMKs function in transcriptional regulation by binding to the other regulators or chromatin modifiers (El Bakkoury ; Kim ). It is possible that AtIPK2β also mediates the activation of FLC expression through a similar mechanism. Potential interaction proteins of AtIPK2β were identified by affinity purification and mass spectrometry (unpublished data) performed as described previously (Wang ), and one peptide corresponding to FVE was identified (Supplementary Fig. S2).

FVE is a key regulator in the autonomous flowering pathway, which inhibits the expression of FLC by mediating histone modification on FLC chromatin (Gu ; Pazhouhandeh ). The potential interaction between FVE and AtIPK2β may explain how AtIPK2β regulates FLC expression. We first confirmed the interaction between AtIPK2β and FVE in yeast by performing yeast two-hybrid assays (Fig. 5A). BiFC assays were conducted to test whether AtIPK2β interacts with FVE in plants. The N-terminal fragment of YFP was fused to AtIPK2β to form YN-AtIPK2β, and the C-terminal half of YFP was fused to FVE to form YC-FVE. YN-AtIPK2β and YC-FVE were simultaneously expressed in Arabidopsis protoplasts and YFP fluorescence was observed (Fig. 5B), reflecting the physical association of AtIPK2β and FVE in vivo. Co-IP assays were conducted to further determine the interaction using transgenic plants co-expressing AtIPK2β-MYC and FVE-GFP fusion proteins; transgenic plants overexpressing AtIPK2β-MYC in the WT background was used as a negative control. The FVE-GFP fusion and the interacting proteins were co-immunoprecipitated by an anti-GFP antibody from the total proteins of FVE-GFP/AtIPK2β-MYC transgenic plants, and the precipitates were analyzed by western blot using an anti-MYC antibody. The band of AtIPK2β-MYC was detected in the protein extracts from the FVE-GFP/AtIPK2β-MYC transgenic plants, but not in the extracts from the negative control (Fig. 5C). Taken together, these results evidence the interaction between AtIPK2β and FVE, indicating a functional link between the two proteins.

Fig. 5.

AtIPK2β interacts with FVE in yeast and plants. (A) Interaction of AtIPK2β with FVE in yeast. Yeast two-hybrid assays were used for testing the interaction. Yeast cells were streaked on the selective media DDO (SD/-Leu/-Trp) and QDO (SD/-Leu/-Trp/-His/-Ade). X-Gal indicates filter β-galactosidase assays. (B) BiFC assays of the interaction between AtIPK2β and FVE in Arabidopsis protoplasts. AtIPK2β and FVE were fused to the YFP N-terminal region (YN) and C-terminal region (YC) to form YN-AtIPK2β and YC-FVE, respectively. Empty YN and YC constructs or an irrelevant YC-SnRK2.6 fusion were used as negative controls. Confocal images of YFP, autofluorescence, and bright-field detections are merged and shown in the lower panels, and images of YFP are shown in the upper panels. Bars=10 μm. (C) AtIPK2β interacts with FVE in transgenic Arabidopsis. Co-immunoprecipitation (Co-IP) assays were conducted by extracting total proteins from transgenic plants co-expressing AtIPK2β-MYC and FVE-GFP fusions, which were then immunoprecipitated with an anti-GFP antibody. The precipitates were detected by western blot with an anti-MYC antibody. An AtIPK2β-MYC overexpressign line was used as the negative control. The asterisk indicates a non-specific IgG band. (This figure is available in colour at JXB online.)

Binding of FVE to FLC chromatin is repressed by AtIPK2β

FVE binds to the FLC locus by forming protein complexes with chromatin modifiers (Jeon and Kim, 2011; Pazhouhandeh ). As AtIPK2β binds to FLC chromatin and interacts with FVE, it is presumed that AtIPK2β affects the accumulation of FVE at the FLC locus. To confirm this hypothesis, we performed ChIP-qPCR assays using transgenic Arabidopsis overexpressing a FVE-GFP fusion construct in the WT and atipk2β mutant backgrounds. The transgenic lines were identified and used in the analysis. Similar levels of FVE-GFP protein were identified to be expressed in both the atipk2β mutant and WT background (Fig. 6A). The sequence regions examined in the ChIP-qPCR assays were identical to those described in Fig. 4A, and WT was used as a negative control. This experiment confirmed that FVE-GFP accumulates on FLC chromatin in both FVE-GFP and FVE-GFP/atipk2β transgenic plants (Fig. 6B), which supports the reports from previous studies that FVE is enriched at the FLC locus (Jeon and Kim, 2011; Pazhouhandeh ). Additionally, ChIP-qPCR assays revealed that knockout of AtIPK2β results in a greater accumulation of FVE-GFP on FLC chromatin in FVE-GFP/atipk2β plants compared with that in FVE-GFP transgenic plants. We found that AtIPK2β is enriched in sequence regions P1, P2, P6, and P7 of FLC in WT (Fig. 4E), and ChIP-qPCR results demonstrated that the enrichment level of the FVE-GFP fusion in the FVE-GFP overexpressing line was reduced by more than 50% in these four regions compared to that in FVE-GFP/atipk2β transgenic plants (Fig. 6B), suggesting that AtIPK2β represses the accumulation of FVE on FLC chromatin.

Fig. 6.

AtIPK2β represses the accumulation of FVE on FLC chromatin. (A) Relative levels of FVE-GFP fusion protein in FVE-GFP and FVE-GFP/atipk2β transgenic plants. Immunodetection of FVE-GFP protein was performed as described in Fig. 4B. (B) Enrichment level of FVE on FLC chromatin in the atipk2β mutant. ChIP assays were performed as described in Fig. 4C. Transgenic plants overexpressing FVE-GFP in the wild-type (WT) and the atipk2β mutant background grown under long-day (LD) conditions for 15 days were harvested for ChIP-qPCR assays. Sequence regions examined in ChIP-qPCR assays were the same as those described in Fig. 4A. Fold enrichment levels of FVE-GFP in different genetic backgrounds over the control line (WT) are shown as the means±SD of biological duplicates. (C) Relative expression level of AtIPK2β in the fve mutant. The relative transcript level of AtIPK2β was examined by RT-qPCR as described in Fig. 2. Data are presented as the means±SD of two independent repeats. (D) Enrichment level of AtIPK2β on FLC in the fve mutant. ChIP-qPCR assays were performed using anti-AtIPK2β antibody as described in Fig. 4E. 15-day-old plants grown under LD conditions were harvested for chromatin extraction. Fold enrichment levels of AtIPK2β on FLC in the fve mutant over the control line (WT) are shown as the means±SD of biological duplicates.

Next, we examined whether the accumulation of AtIPK2β at the FLC locus is affected by FVE. A T-DNA insertion line of FVE (SALK_013789) was identified (Supplementary Fig. S3A, B) and used in the analysis. As the mutant line has T-DNA inserted in the first intron of FVE, it still expresses a small amount of FVE (Supplementary Fig. S3B). Similar to the previously reported fve mutants (Gu ; Jeon and Kim, 2011; Pazhouhandeh ), this T-DNA insertion line exhibits a significantly late-flowering phenotype (Supplementary Fig. S3C). Gene expression analysis showed that the transcription of AtIPK2β is promoted in the fve mutant (Fig. 6C), and ChIP-qPCR assays demonstrated that the enrichment level of AtIPK2β at the FLC locus in the mutant is consistently increased (Fig. 6D). These data showed that FVE negatively regulates the enrichment of AtIPK2β at FLC, indicating a dynamic regulatory mechanism of FLC expression effected by FVE and AtIPK2β.

AtIPK2β functions in FVE-mediated transcriptional regulation of FLC

FVE blocks transcription of FLC by mediating histone modifications such as H3K27 trimethylation (H3K27me3) and histone H3 deacetylation of FLC chromatin (Ausín ; Pazhouhandeh ; Yu ). We have found that AtIPK2β interacts with FVE and represses the accumulation of FVE on FLC chromatin, indicating a functional link between AtIPK2β and FVE in the transcriptional regulation of FLC. To study whether AtIPK2β regulates FLC expression with FVE, we generated the fve/atipk2β double mutant and examined the expression level of FLC in WT, atipk2β, fve, and fve/atipk2β plants. Fig. 7A shows that the expression level of FLC in the double mutant was similar to that in the fve mutant, and higher than that in the atipk2β mutant. Both the fve/atipk2β double mutant and the fve mutant consistently exhibited late-flowering phenotypes compared with WT plants (Fig. 7B). These data showed that FVE is required for AtIPK2β regulation of FLC expression.

Fig. 7.

AtIPK2β regulation of FLC transcription requires FVE. (A) Expression analysis of FLC in wild-type (WT), atipk2β, fve, and fve/atipk2β plants. Relative transcript levels of FLC in 10-day-old WT and mutant plants grown under long-day (LD) conditions were determined by qPCR, as described in Fig. 2. Data are the means±SD of three biological replicates. (B) Flowering phenotype in WT, atipk2β, fve, and fve/atipk2β plants, expressed as the number of rosette leaves at flowering and the number of days to bolting. For each line, 10–15 plants were scored; significant differences (**P<0.01, ***P<0.001) compared to WT were determined by one-way ANOVA with Tukey’s multiple comparison test. (C) Analysis of enrichment levels of the H3K27me3 mark at the FLC locus in WT, atipk2β, fve, and fve/atipk2β. ChIP-qPCR assays were conducted using H3K27me3 antibody. The amount of DNA precipitates after ChIP was quantified and normalized to FUSCA. Data from two independent experiments were averaged and are presented as means±SD; fold enrichments of the H3K27me3 mark on FLC in the mutants relative to WT are shown. (D) Analysis of enrichment levels of the H3Ac mark at the FLC locus in WT, atipk2β, fve, and fve/atipk2β. An anti-acetyl-H3 antibody was used in ChIP-qPCR assays. Chromatin was prepared from 10-day-old WT and mutant plants grown under LD conditions. Sequence regions on FLC chromatin examined in ChIP-qPCR assays were identical to those described in Fig. 4A. ACTIN was used for normalizing the quantified DNA fragments. Fold enrichments of the H3Ac mark on FLC in the mutants relative to WT are shown.

We further determined whether AtIPK2β regulates FLC transcription by involvement in FVE-mediated chromatin modification of FLC. Consistent with the reduced expression level of FLC, ChIP-qPCR assays showed that deposition of the repressive H3K27me3 mark is increased on FLC chromatin in the atipk2β mutant (Fig. 7C), and enrichment of the active H3Ac mark is decreased at the same locus, compared to that in WT plants (Fig. 7D). In contrast, both fve/atipk2β and fve mutant plants showed lower H3K27me3 and higher H3Ac levels on FLC compared with WT, consistent with the increased level of FLC transcription in the mutants. Taken together, these results suggest that AtIPK2β promotes FLC expression by blocking FVE-mediated chromatin silencing of FLC.

AtIPK2β associates with HDA6 and attenuates its accumulation on FLC chromatin

Previous studies have revealed that FVE interacts with HDA6 to regulate histone modification at the FLC locus and mediates chromatin silencing of the target genes (Gu ; Yu ). HDA6 is one of the Class I HDACs identified in Arabidopsis. It is involved in diverse signaling pathways that interact with different proteins to mediate histone deacetylation and DNA methylation, leading to transcriptional silencing of the target genes (Pandey ; Liu ). In this study, we have confirmed that AtIPK2β interacts with FVE and represses the accumulation of FVE on FLC chromatin. As HDA6 is an interaction partner of FVE, we presumed that AtIPK2β also affects the enrichment level of HDA6 at the FLC locus.

We found that HDA6 interacts with AtIPK2β directly in yeast cells (Fig. 8A). In vivo Co-IP assays were performed to confirm the interaction by using transgenic plants overexpressing a HDA6-MYC fusion protein in the WT and atipk2β mutant backgrounds. Anti-AtIPK2β antibody was applied for immunodetection of AtIPK2β protein in the precipitates. Transgenic plants overexpressing HDA6-MYC in the WT background were described previously (Liu ); HDA6-MYC/atipk2β transgenic plants were generated by crossing the HDA6-MYC line with the atipk2β mutant and used as a negative control in the Co-IP assays. Similar levels of HDA6-MYC fusion protein were detected in both HDA6-MYC and HDA6-MYC/atipk2β overexpressing plants (Fig. 8B). The Co-IP assays showed that a band of HDA6-MYC fusion protein was detected in the co-immunoprecipitated proteins from HDA6-MYC, but no such band was detected in precipitates from HDA6-MYC/atipk2β plants, demonstrating the interaction between AtIPK2β and HDA6 in vivo (Fig. 8C).

Fig. 8.

AtIPK2β interacts with HDA6 and attenuates the accumulation of HDA6 on FLC chromatin. (A) AtIPK2β interacts with HDA6 in yeast. Yeast two-hybrid assays were performed as described in Fig. 5A. (B) Relative levels of HDA6-MYC fusion protein in HDA6-MYC and HDA6-MYC/atipk2β transgenic plants. Immunodetection of HDA6-MYC was conducted as described in Fig. 4A using anti-MYC antibody. The arrow indicates the band of HDA6-MYC fusion protein. (C) In vivo interaction between AtIPK2β and HDA6. Total proteins were extracted from 3-week-old HDA6-MYC transgenic plants in the indicated genetic backgrounds and Co-IP assays were conducted as described in Fig. 5C. HDA6-MYC was detected immunologically with anti-MYC antibody. (D) Analysis of HDA6 enrichment on FLC chromatin in HDA6-MYC and HDA6-MYC/atipk2β transgenic plants. Chromatin was extracted from 15-day-old plants grown under long-day conditions. ChIP-qPCR assays were conducted using anti-MYC antibody. Sequence regions examined are identical to those described in Fig. 4A. Data shown are the means±SD of two independent experiments. (This figure is available in colour at JXB online.)

We further analyzed the effect of AtIPK2β on the accumulation of HDA6 at the FLC locus. ChIP-qPCR assays showed that HDA6-MYC is enriched in the examined regions on FLC chromatin in both of the transgenic lines overexpressing HDA6-MYC in the WT and atipk2β background; the enrichment level of HDA6-MYC in HDA6-MYC/atipk2β was higher than that in HDA6-MYC (Fig. 8D), suggesting that the expression of AtIPK2β in HDA6-MYC transgenic plants reduces the enrichment level of HDA6 at the FLC locus These results indicated that the interaction between AtIPK2β and HDA6 attenuates HDA6 accumulation at the FLC locus.

Discussion

In this study, we report that AtIPK2β promotes FLC expression by blocking FVE-mediated chromatin silencing of FLC, delaying flowering time. We reveal that AtIPK2β regulates FLC expression by enriching at the FLC locus. Furthermore, we demonstrate that AtIPK2β blocks FVE-mediated chromatin silencing of FLC by interacting with FVE and repressing the accumulation of FVE on FLC chromatin. AtIPK2β also interacts with HDA6, an interaction partner of FVE, and represses the accumulation of HDA6 on FLC chromatin. This work provides a possible molecular mechanism by which AtIPK2β functions in the transcriptional regulation of FLC and flowering time control.

We show that AtIPK2β mediates activation of FLC transcription (Fig. 2A). Previous studies of AtIPK2β indicated its potential function in transcriptional regulation in plants. Although there is no conserved nuclear localization domain found in the protein sequence, localization analysis revealed that AtIPK2β enriches in the nucleus (Xia ). Previous investigations showed that overexpressing AtIPK2β in Arabidopsis affects the transcription of genes in the auxin signaling pathway, and ectopic expression of AtIPK2β in tobacco activates the stress-responsive genes (Zhang ; Yang ). It has been reported that yeast IPMK (also known as IPK2), which is homologous to AtIPK2β, activates target genes by interacting with a MADS-box transcription factor, Mcm1, and stabilizes the Mcm1-ArgR complex (El Bakkoury ; Bosch and Saiardi, 2012). In mammalian cells, IPMK functions in transcriptional activation by interacting with transcription factors and regulators (Malabanan and Blind, 2016). We propose that the function of IPMK in transcriptional regulation is conserved in plants, yeast, and mammals.

We performed ChIP-qPCR assays and revealed that AtIPK2β binds to FLC chromatin (Fig. 4C, E). As there is no conserved DNA-binding motif found in the protein structure (Xia and Yang, 2005; Endo-Streeter ), it is unlikely that AtIPK2β binds to DNA directly. We noticed that the enrichment level of AtIPK2β on FLC chromatin is much higher in the P6 and P7 sequence regions (Fig. 4E), indicating that AtIPK2β enriches in these specific regions, probably through interacting with DNA-binding proteins.

We show in this study that AtIPK2β delays flowering time through FVE-mediated transcriptional regulation of FLC. We demonstrated that AtIPK2β interacts with FVE (Fig. 5), which belongs to the conserved MSI1-like protein family that mediates chromatin assembly and histone modification by interacting with histones directly and indirectly (Hennig ; Derkacheva ). As a positive regulator of floral transition in the autonomous pathway, FVE forms a large complex consisting of at least 16 proteins and binds to FLC chromatin, resulting in the inhibition of FLC expression (Jeon and Kim, 2011). Previous studies have revealed that FVE regulates histone H3 deacetylation, and histone H3K4 and K27 trimethylation (H3K4me3 and H3K27me3) at the FLC locus by forming complex with HDA6 and FLD (Gu ; Yu ). A study reported that FVE interacts with both CUL4-DDB1 and PRC2-like complexes, and binds to FLC and FT chromatins dependent of CUL4, leading to enhancement of histone H3K27 trimethylation and histone H3 deacetylation at the FLC and FT loci (Pazhouhandeh ). In this study, we reveal a negative regulatory role of AtIPK2β in FVE-mediated transcriptional regulation of FLC. We suggest that the interaction between AtIPK2β and FVE represses the accumulation of FVE on FLC chromatin and releases the transcription of FLC as a result. The functional link between AtIPK2β and FVE explains the reduced FLC expression level in the atipk2β mutant, and the late-flowering phenotype with higher FLC transcript level in AtIPK2β-overexpressing plants. Moreover, FVE represses AtIPK2β transcription and therefore reduces the enrichment of AtIPK2β at the FLC locus (Fig. 6C, D), indicating a dynamic regulation of FLC by both AtIPK2β and FVE.

Recent studies have shown that IPMK plays an important role in the regulation of histone modification in mammalian cells. IPMK physically interacts with CREB binding protein (CBP), which is a histone acetyl transferase, and mediates histone acetylation on the target loci (Xu ). IPMK is essential for the recruitment of CBP to chromatins, as deletion of IPMK causes incomplete histone H3 and H4 acetylation, which is dependent on CBP. IPMK is also a critical component for enhancing chromatin association of the p53-p300 histone acetyl transferase complex (Xu and Snyder, 2013). Here, we suggest that AtIPK2β interacts with histone deacetylase HDA6 and attenuates the accumulation of HDA6 on FLC chromatin (Fig. 8). Since FVE is essential for HDA6 binding to FLC chromatin (Jung ), we presumed that AtIPK2β reduces the enrichment of HDA6 on FLC by repressing the accumulation of FVE at the FLC and probably destabilizing the interaction between HDA6 and FVE. Our investigation suggests a distinguishing function of IPMK in the regulation of histone modification in plant and in mammalian cells.

Taken together, our findings provide evidence for a regulatory mechanism by which AtIPK2β functions in flowering time control. AtIPK2β enriches on FLC chromatin and represses FVE-mediated histone modification on FLC by attenuating the accumulation of FVE and HDA6 at the FLC locus, leading to transcriptional activation of FLC and delayed flowering. More widely, our study of the functional link between AtIPK2β and FVE-HDA6 will improve understanding of the FLC-dependent pathway in flowering time control. How AtIPK2β interacts with FVE and HDA6 to repress the accumulation of FVE and HDA6 on FLC chromatin will be the focus of future investigations.

Supplementary data

Supplementary data are available at JXB online.

Fig. S1. Analysis of flowering time in atipk2β mutants and complemented lines under the LD condition.

Fig. S2. Identification of FVE as a potential interactor of AtIPK2β by liquid chromatography coupled to tandem mass spectrometry.

Fig. S3. Molecular characterization of the fve mutant.

Table S1. Primer sequences used for genotyping PCR and qRT-PCR.

Table S2. Primer sequences used for ChIP-qPCR.

Click here for additional data file.