Skip to main content
NCBI home page
The Plant Cell logoLink to The Plant Cell
. 2008 Mar;20(3):580–588. doi: 10.1105/tpc.108.058172

ARABIDOPSIS TRITHORAX1 Dynamically Regulates FLOWERING LOCUS C Activation via Histone 3 Lysine 4 Trimethylation[W]

Stéphane Pien a,b,c,1, Delphine Fleury c, Joshua S Mylne d, Pedro Crevillen d, Dirk Inzé c, Zoya Avramova e, Caroline Dean d, Ueli Grossniklaus a
PMCID: PMC2329943  PMID: 18375656

Abstract

Trithorax function is essential for epigenetic maintenance of gene expression in animals, but little is known about trithorax homologs in plants. ARABIDOPSIS TRITHORAX1 (ATX1) was shown to be required for the expression of homeotic genes involved in flower organogenesis. Here, we report a novel function of ATX1, namely, the epigenetic regulation of the floral repressor FLOWERING LOCUS C (FLC). Downregulation of FLC accelerates the transition from vegetative to reproductive development in Arabidopsis thaliana. In the atx1 mutant, FLC levels are reduced and the FLC chromatin is depleted of trimethylated, but not dimethylated, histone 3 lysine 4, suggesting a specific trimethylation function of ATX1. In addition, we found that ATX1 directly binds the active FLC locus before flowering and that this interaction is released upon the transition to flowering. This dynamic process stands in contrast with the stable maintenance of homeotic gene expression mediated by trithorax group proteins in animals but resembles the dynamics of plant Polycomb group function.

INTRODUCTION

Epigenetic mechanisms play crucial roles in shaping and maintaining cell identity and in patterning the body plan during development. Epigenetic information is partly carried by histone proteins in the form of reversible covalent modifications at their N-terminal tails. In Drosophila melanogaster, the Polycomb group (PcG) and trithorax group (trxG) proteins form higher-order complexes, which antagonistically repress and maintain the expression of homeotic genes (HOX genes), respectively (Simon and Tamkun, 2002). PcG and trxG complexes contain SET (for Suppressor of variegation 3-9, Enhancer of zeste, TRX) domain proteins that have histone methyltransferase (HMT) activity. They posttranslationally modify lysines on histones H3 and H4 (Lachner et al., 2004), thereby regulating the accessibility of the transcription machinery to the HOX gene clusters. These Lys methylation states have been classified as repressive and activating marks, depending on their effect on gene expression.

In recent years, several Arabidopsis thaliana PcG complexes were shown to repress their target genes via deposition of H3K27me3 marks (reviewed in Pien and Grossniklaus, 2007). This supports a conservation of the PcG function between plants and animals. Consequently, if trithorax functions were also conserved during evolution, trxG proteins may antagonistically regulate PcG target genes. Consistently, two Arabidopsis PcG target genes, the flowering time regulator FLOWERING LOCUS C (FLC) and the floral homeotic gene AGAMOUS (AG), show an enrichment of H3K4me2 and H3K4me3 marks at their chromatin, which correlates with active transcription (Bastow et al., 2004; He et al., 2004; Schubert et al., 2006). In Drosophila, such marks are deposited by the trxG protein Trithorax (TRX) and in mouse by the mixed-lineage leukemia (MLL) protein. The presence of H3K4me marks suggests that trithorax homologs and their associated functions exist in Arabidopsis.

Five close homologs of TRX and MLL were identified in the Arabidopsis genome and named ARABIDOPSIS TRITHORAX (ATX1-5) (Alvarez-Venegas and Avramova, 2001; Baumbusch et al., 2001). ATX1 is predicted to contain a SET domain (Alvarez-Venegas et al., 2003) that has both histone binding and HMT activity (Rea et al., 2000; Katsani et al., 2001). In vitro assays demonstrated that H3K4 is a substrate for ATX1's HMT activity, while mutant isoforms of ATX1 lacking part of the SET domain have no activity (Alvarez-Venegas et al., 2003). Loss of ATX1 leads to flower homeotic defects and affects leaf morphogenesis (Alvarez-Venegas et al., 2003). Recently, ATX1 was shown to bind AG chromatin and to be required for H3K4me3 deposition at this locus (Saleh et al., 2007).

Transcriptional profiling in the atx1 mutant allowed us to identify FLC, a flowering time regulator, as a putative ATX1 target gene. FLC encodes a MADS domain transcription factor that functions as a repressor of the floral transition. Transcriptional regulation of FLC has been well studied and shown to be associated with chromatin modifications, but little was known about the activation and maintenance of expression of this central gene. Therefore, we studied FLC regulation as a model to decipher the molecular mechanism of trithorax function in plants and to gain insight into novel functions of the ATX1 gene. We showed that both single and double mutation of ATX1 and its closest homolog ATX2 (Alvarez-Venegas and Avramova, 2001) lead to early flowering, correlating with a reduction of FLC transcripts levels. ATX1 and its target gene FLC are coexpressed in a spatio-temporal manner. Using chromatin immunoprecipitation (ChIP), we showed that ATX1 binds at the FLC locus and its presence correlates with H3K4me3 modifications. Furthermore, ChIP analyses revealed that ATX1 not only activates FLC but it also prevents its repression, since H3K27me2 repressive marks are deposited in the absence of ATX1 function. Finally, our study identified ATX1 as a direct transcriptional activator of FLC.

RESULTS

atx1 and atx2 Are Early-Flowering Mutants Affecting the FLC Expression Level

The atx1-1 mutation was previously reported to delay the transition to flowering (Alvarez-Venegas et al., 2003). These studies, however, did not quantify the changes in flowering time or leaf number at bolting. Therefore, we investigated the atx1-1 mutant in more detail to decipher how the flowering transition was affected. In contrast with previously published work (Alvarez-Venegas et al., 2003), under our growth conditions atx1mutations led to an early-flowering phenotype in the rapid-flowering accessions Wassilewskija (Ws) and Columbia (Col), as measured by the number of rosette leaves at bolting (Figures 1A and 1B). atx1-1 mutants flowered early under both long-day and short-day conditions, showing that ATX1 is involved in repressing the transition to flowering, independently of the photoperiod.

Figure 1.

Figure 1.

Open in a new tab

Characterization of atx Mutants.

(A) At flowering, the atx1 (Ws) rosette If ([A], right) is smaller than the wild type ([A], left), with a reduction in the leaf number in atx1 compared with the wild type.

(B) The average number of rosette leaves at flowering, a measure of flowering time, is reduced in atx1 and atx2 mutants. Open bars, leaf number at flowering under long-day conditions; gray bars, short-day conditions. All data are presented as means ± se (n = 15 to 20; P < 0.05 using Student's t test).

(C) RT-PCR quantification of FLC transcripts in the wild type and in atx1 and atx2 mutants. Left panel, wild type (Ws) and atx1-1 (Ws) mutant; right panel, wild type (Col), atx1-2 (Col), and atx2-1 (Col) mutants. Numbers (±se) refer to FLC transcript level relative to the wild type of three independent biological replicate experiments. ACT, actin loading control.

(D) Average number of rosette leaves at flowering. atx1-1 (Ws) mutants were crossed with Col plants (ColSf-2), into which the wild-type San Feliu-2 (Sf-2) flowering-time locus FRIGIDA (FRISf-2) had been introgressed (Lee and Amasino, 1995) (white columns). atx1-2 atx2-1 (Col) double mutants were crossed with ColSf-2 (gray columns). Columns 1 to 4 represent segregating F2 populations; the origin of the relevant alleles is indicated. Columns 5 to 7 are from a homogeneous background resulting from crosses between ColSf-2 and Col. 1, FRISf-2 (ColSf-2) ATX1 (ColSf-2) (n = 12); 2, FRISf-2/fri (ColSf-2/Ws) ATX1 (ColSf-2) (n = 46); 3, FRISf-2/fri (ColSf-2/Ws) ATX1/atx1 (ColSf-2/Ws) (n = 67); 4, FRISf-2/fri (ColSf-2/Ws) atx1-1 (Ws) (n = 38); 5, FRISf-2 (ColSf-2) ATX1 (Col) ATX2 (Col); 6, FRISf-2 (ColSf-2) atx1-2 (Col) ATX2 (Col); 7, FRISf-2 (ColSf-2) atx1-2 (Col) atx2-1 (Col). In a heterozygous FRISf-2/fri background, the introduction of one atx1 copy results in a reduced number of leaves at flowering time. The suppression of FRI had a greater effect on flowering time in the atx1 (Ws) background. In a ColSf-2 background carrying the FRISf-2 allele, the atx1-1 atx2-1 double mutant suppressed the late-flowering phenotype more dramatically than the atx1-2 single mutant. All data are presented as means ± se (n = 12 to 67; P < 0.05 using Student's t test).

To investigate the molecular basis of the early-flowering phenotype of atx1-1 mutant plants, we performed RNA profiling experiments with 10-d-old atx1-1 and wild-type seedlings using the Affymetrix ATH1 GeneChip. To reveal whether trxG and PcG proteins play similar antagonistic roles as they do in animals (Simon and Tamkun, 2002), we looked for differentially expressed genes that had already been classified as PcG target genes. With these criteria, we found that the floral regulator FLC, a PcG target gene, had greatly reduced steady state transcript levels in the atx1-1 mutant (3.4-fold decrease; see Supplemental Table 1 and Supplemental Table 2 online). This reduction was confirmed by RT-PCR in both atx1-1 and atx1-2 (Col) homozygous mutants (Figure 1C).

We further investigated the impact of atx1 mutations on flowering time by crossing atx1-1 and atx1-2 mutants with a line containing an active FRIGIDA (FRI) allele (Lee and Amasino, 1995). The presence of the active FRI allele is associated with Arabidopsis late-flowering accessions and results in a high level of FLC expression. Loss of ATX1 strongly suppressed the late-flowering effect of FRI in a dosage-dependent manner (Figure 1D). Thus, ATX1 is required for the increased expression of FLC that results from overexpression of FRI in the line containing an active FRI allele.

Mutation of the closest homolog of ATX1, ATX2 (Alvarez-Venegas and Avramova, 2001), revealed a role for ATX2 in FLC regulation (Figure 1C). In a FRI background, the atx1-2 atx2-1 double mutant suppressed the late-flowering phenotype more dramatically than in the atx1-2 single mutant (Figure 1D), suggesting that ATX1 and ATX2 play a partially redundant role in activating FLC.

ATX1 and FLC Are Spatio-Temporally Coexpressed

Since we found that ATX1 and ATX2 regulate FLC, we analyzed the spatio-temporal expression of these three genes by in situ hybridization. If ATX1 and ATX2 directly regulate FLC during the plant life cycle, then the spatio-temporal expression patterns of ATX1, ATX2, and FLC are expected to overlap. In situ hybridization analyses for ATX1 and FLC in 10-d-old seedlings revealed expression of both genes in the vasculature of the cotyledons, hypocotyls, and the first pair of leaves (Figures 2A to 2D). The pattern of FLC mRNA accumulation reproduces the expression pattern of the reporter gene uidA, encoding β-glucuronidase (GUS), translationally fused to FLC (FLC-GUS) (Bastow et al., 2004). Both ATX1 and FLC transcripts were present in overlapping patterns during embryogenesis (Figures 2E and 2F). FLC transcript was not detected in atx1-1 embryos, suggesting that ATX1 is necessary to activate FLC well before the floral induction pathways are active (Figures 2G and 2H). Just prior to flowering, a strong reduction of both ATX1 and FLC transcript levels occurred in the vasculature of wild-type plants (Figures 2I and 2J). The overlap of the spatio-temporal expression patterns of the two genes is consistent with a direct regulation of FLC by ATX1. These data were confirmed by a cross between a FRI line containing an FLC-LUCIFERASE (FLC-LUC) translational fusion (Mylne et al., 2004) and the atx1-2 mutation (Figures 3A and 3B). In FRI FLC-LUC plants, FLC was highly expressed, as indicated by the FLC-LUC signal in the vasculature and the shoot apex (Figure 3A). In FRI FLC-LUC atx1-2 lines, FLC expression was strongly reduced in the vasculature (Figure 3B). Interestingly, ATX1 expression could not be detected in the shoot apical meristem of wild-type plants (Figure 2A), and mutations in ATX1 did not lead to a loss of FLC expression in that tissue (Figure 3B), suggesting that FLC expression in the shoot apical meristem is positively regulated by some other factor(s). We investigated the expression pattern of ATX2 at the same developmental stage. ATX2 was expressed in the vasculature and, unlike ATX1, was also detected in the shoot apical meristem (see Supplemental Figure 1 online). In FRI FLC-LUC atx1-2 atx2-1 lines, FLC expression was strongly reduced in the vasculature and in the shoot apex (Figure 3C). However, FLC expression was still detectable in the shoot apical meristem of FRI FLC-LUC atx1-2 atx2-1 lines, confirming that ATX1 and ATX2 play their major role in the vasculature.

Figure 2.

Figure 2.

Open in a new tab

Spatio-Temporal Expression Patterns of ATX1 and FLC in Wild-Type and atx1-1 Tissue as Assayed by in Situ Hybridization.

(A), (C), (E), (G), and (I) Sections probed with an antisense ATX1 probe.

(B), (D), (F), (H), and (J) Sections probed with an antisense FLC probe. Wild-type tissue sections hybridized with sense probes for ATX1 and FLC gave no signal at any developmental stage.

(A) and (B) Ten-day-old seedlings with ATX1 and FLC transcripts accumulating in the vasculature and the hypocotyl.

(C) and (D) Ten-day-old seedlings. Cross sections of the first pair of leaves, with ATX1 and FLC transcripts accumulating in the vasculature (arrows).

(E) and (F) Wild-type globular embryos showing expression of both ATX1 and FLC.

(G) and (H) In atx1-1 globular embryos, neither ATX1 nor FLC transcripts are detectable.

(I) and (J) At flowering, neither ATX1 nor FLC message is detectable in the wild-type vasculature.

Figure 3.

Figure 3.

Open in a new tab

FLC-LUC Expression Pattern in FRI, FRI atx1-2, and FRI atx1-2 atx2-1 Plants.

(A) In FRI FLC-LUC lines, FLC is highly expressed in the vasculature and the shoot apex (arrow).

(B) In FRI FLC-LUC atx1-2 lines, FLC expression is strongly reduced in the vasculature.

(C) In FRI FLC-LUC atx1-2 atx2-1 lines, FLC expression is strongly reduced in the vasculature and in the shoot apex (arrow).

Twenty-five-day-old plants were analyzed for FLC-LUC expression for all lines analyzed. Two biological replicates were performed, growing side by side 10 plants of each genotype.

ATX1 Is Required for the Deposition of H3K4me3 Marks at the FLC Locus

To address whether ATX1-dependent histone modifications are involved in the regulation of FLC, we analyzed chromatin modifications at the FLC locus by ChIP at three regions surrounding the translational start codon (Figure 4A). These regions are essential for FLC transcription and function (Bastow et al., 2004; Kim et al., 2005), and H3K4me3 marks in these regions correlate with FLC transcription (He et al., 2004). We found that H3K4me3 levels were reduced in region B and undetectable in region A in atx1-1 mutants (Figure 4B) compared with wild-type plants at the same developmental stage. Just prior to floral induction, H3K4me3 levels decreased in wild-type plants to levels similar to those in atx1-1 mutant seedlings. Taken together, these findings suggest that ATX1 is required for the establishment of the H3K4me3 mark at the FLC locus to promote and/or maintain a transcriptionally active state.

Figure 4.

Figure 4.

Open in a new tab

Histone Modifications and ATX1 Binding at the FLC Locus.

(A) Genomic structure of the FLC promoter and regions investigated by ChIP. The thick lines represent the 5′ untranslated region and intron 1, while the black box represents the translated region of exon 1. Regions amplified by PCR are labeled A to C.

(B) Relative levels of histone modifications in FLC chromatin were analyzed by PCR from at least three replicate ChIP assays using H3K4me2-, H3K4me3-, H3K27me2-, and H3K27me3-specific antibodies. Black bars, 10-d-old Ws seedlings; open bars, 10-d-old atx1-1 mutant seedlings; gray bars, 16-d-old wild-type plants prior to flowering. Means are calculated based on at least three independent experiments and are given with bars indicating 1 se.

(C) ChIP assay using an ATX1-specific antibody. Regions A, B, and C were examined for ATX1 enrichment in FLC chromatin. +, ATX1 antibody; −, no antibody controls. Regions A and B showed enrichment of ATX1 in Ws seedlings. PF, 16-d-old Ws plants prior to flowering.

In plants and animals, H3K4 can be either, mono-, di-, or trimethylated. These three epigenetic marks can be interpreted differently by the transcription machinery depending on the organism (Fuchs et al., 2006). Therefore, we quantified in parallel H3K4me2 and H3K4me3 marks at the FLC locus in wild-type and atx1-1 plants. Surprisingly, region B, which covers the transcription start site of FLC, displayed elevated levels of H3K4me2 in 10-d-old atx1-1 seedlings compared with the wild type (Figure 4B). In regions A and C, this mark was almost at the same level as in wild-type plants. At later developmental stages, prior to the flowering transition, a similar pattern could be observed. Altogether, the loss of ATX1 activity strongly impaired the deposition of H3K4me3 marks but did not suppress the deposition of H3K4me2 marks at FLC chromatin. These findings suggest that the function of ATX1 seems to be specific for the deposition of the H3K4me3 mark.

Loss of H3K4me3 marks at the FLC locus induces a gain of H3K27me2 marks

Since in the atx1-1 mutant, FLC is depleted in the H3K4me3 activation mark and is transcriptionally repressed, we measured the levels of H3K27me2 and H3K27me3, two marks previously shown to correlate with FLC repression (Bastow et al., 2004; Sung et al., 2006). In the regions A and B, an increased level of H3K27me2 was observed in atx1-1 mutant seedlings compared with wild-type seedlings at the same stage. H3K27me2 levels also increased, although less dramatically, prior to flowering. In wild-type plants, the level of H3K4me3 inversely correlated with the level of H3K27me2. However, H3K27me2 marks were still present on FLC chromatin in regions A and B during active transcription of the gene prior to the flowering transition. Similarly, low H3K27me3 levels could be detected in wild-type seedlings at the FLC chromatin in all regions investigated (Figure 4B). It is worth noting that the presence of this repressive mark together with the activating mark H3K4me3 does not prevent active transcription of FLC. The level of the H3K27me3 mark, however, substantially increased in plants prior to flowering, which correlates with FLC repression at that developmental stage. This is consistent with the observation that H3K27me3 deposition by plant PcG proteins correlates with transcriptional repression of the Arabidopsis PcG target genes MEDEA (MEA), PHERES1 (PHE1), AG, SHOOTMERISTEMLESS (STM), and AGAMOUS-LIKE 19 (AGL19) (Gehring et al., 2006; Jullien et al., 2006; Makarevich et al., 2006; Schönrock et al., 2006; Schubert et al., 2006).

ATX1 Directly Interacts with FLC Chromatin to Regulate Its Transcription

The ability of ATX1 to promote H3K4me3 deposition suggests that ATX1 directly interacts with the FLC locus to modify its chromatin state via its HMT activity (Alvarez-Venegas et al., 2003). We investigated this possibility by ChIP (Figure 4C) using an antibody raised against ATX1. We found that ATX1 was enriched at the FLC chromatin (regions A and B) in wild-type seedlings relative to atx1-1 mutants. In the wild type, ATX1 was not enriched in region C, which is consistent with the absence of the H3K4me3 mark in that region (Figure 4B). These data show that ATX1 binding to the FLC chromatin correlates with the deposition of H3K4me3 marks. In wild-type plants at the floral transition, when FLC is downregulated, ATX1 binding at the FLC locus could not be detected. This observation suggests that ATX1 dynamically binds the FLC locus to regulate its transcription.

DISCUSSION

The trxG Genes ATX1 and ATX2 Are Required to Activate FLC Expression

Our study provides strong evidence that ATX1, a homolog of the Drosophila trx protein, is required to control flowering transition and acts to upregulate FLC expression. ATX1 acts downstream of, or in parallel with, FRI in an interdependent manner. It also acts directly on FLC and binds to its promoter and transcription start site regions. Many regulators of FLC transcription have been described (reviewed in He and Amasino, 2005), but, unlike ATX1, these do not appear to modify FLC chromatin directly. The putative HMT EARLY FLOWERING IN SHORT DAYS (EFS), a homolog of the trxG Drosophila protein Absent small homeotic disks1 (ASH1) (Tripoulas et al., 1994), was shown to be necessary for the deposition of H3K4me3 marks at FLC chromatin of winter annual accessions (He et al., 2004). In such accessions, vernalization (extended exposure to cold) is required to activate the VRN Polycomb Repressive Complex 2 (VRN-PRC2) (Levy et al., 2002; Kim et al., 2005), which represses FLC, leading to flowering. However, in the commonly studied rapid-flowering accession Col, vernalization is not required for flowering, and mutations in EFS do not affect the level of H3K4me3 at the FLC locus (He et al., 2004). In this context, our data provide evidence that ATX1 regulates FLC transcription in the rapid-flowering accessions Ws and Col. Additionally, in lines containing an active FRI allele, which mimic winter annual accessions, the atx1-1 atx2-1 double mutant suppressed the late-flowering phenotype, suggesting that genes of the ATX family are also central for FLC-mediated regulation in winter annual accessions. How ATX1, ATX2, and EFS collaborate in this process remains unclear and would require additional investigations. Our data suggest that ATX1 and ATX2 are involved in the same pathway since the early-flowering phenotype observed in single mutants is not more severe in the atx1-2 atx2-1 double mutant. However, in the FRI background, mutation of both genes leads to a shorter vegetative phase compared with atx1 single mutants in the same background. This may be explained by delayed transcription of ATX2, whose expression is detected later than ATX1 during the vegetative phase and which may have a stronger impact in winter annual accessions than in rapid-flowering accessions.

PcG and trxG Proteins Dynamically Regulate FLC Expression

In contrast with animals, where PcG and trxG proteins play a role in the permanent repression or activation of genes whose expression state was determined by other factors, in plants, PcG and trxG proteins dynamically interact in the regulation of target genes, such as FLC, during the plant life cycle (reviewed in Pien and Grossniklaus, 2007). In wild-type plants at the floral transition, when FLC transcripts are no longer detectable in the apex and neighboring vasculature, ATX1 binding at the FLC locus could not be detected (Figure 4C). This indicates that ATX1 dynamically binds the FLC locus to regulate its transcription. In Drosophila, TRX together with members of the PRC1/2 complexes are constitutively bound to the HOX Ultrabithorax (Ubx) locus independent of whether the Ubx gene is actively transcribed or not (Papp and Müller, 2006). In contrast with the situation in Drosophila and mammals, ATX1 binding is not stable, which points to a dynamic function of trxG complexes in plants. This dynamic process is reflected by the removal of previously deposited H3K4me3 at the FLC promoter during the transition from vegetative to reproductive development (Figure 4B). Recently, dynamic regulation was also demonstrated for PcG proteins (Baroux et al., 2006). Together, these data suggest that plant PcG and trxG proteins affect a wide range of gene expression programs and potentially contribute to plant developmental plasticity.

FLC Is Regulated through Dosage-Dependent Interactions of Activating and Repressive Histone Modifications

The results presented here highlight the importance of the H3K4me3 modification mediated by ATX1 for the transcriptional activation of FLC: the levels of H3K4me2 in the atx1 mutants are clearly not sufficient in this context. However, we cannot rule out that the deposition of H3K4me2 marks does not play a role in active FLC transcription. A recent study provided evidence on the requirement of FCA together with FLOWERING LOCUS D (FLD) to mediate H3K4 demethylation of FLC in its central region and, thus, to silence the gene (Liu et al., 2007). By contrast, at several Arabidopsis loci, the H3K4me2 mark was shown to be associated with the H3K27me2 mark, independent of whether the associated genes were actively transcribed or not (Alvarez-Venegas and Avramova, 2005).

Surprisingly, our data showed that repressive H3K27me2 and H3K27me3 modifications were present at the FLC locus during active transcription and FLC silencing. This observation is in agreement with the whole-genome analysis of H3K27me3 distribution in the Arabidopsis Ws accession, where this mark was detected at FLC chromatin during active transcription (Zhang et al., 2007). However, it is notable that the levels of H3K27me2 and H3K27me3 marks at FLC are always lower than the levels observed in plants prior to flowering, where FLC is silenced (Figure 4B). The presence of repressive marks at the FLC locus during its active expression can be interpreted as basal levels that are not sufficient to repress FLC expression in that context. Therefore, our results suggest a mechanism where the active or repressive state of FLC expression depends on the accumulation of repressive and activating marks in a dosage-dependent manner (i.e., the expression of FLC correlates with the deposition of H3K4me3 marks and basal levels of H3K27me2 and H3K27me3 marks at the FLC promoter). Conversely, repression of FLC is associated with the removal of H3K4me3 marks and a substantial increase of H3K27me3 marks at the FLC promoter (see model in Figure 5).

Figure 5.

Figure 5.

Open in a new tab

Model for Dosage-Dependent Regulation of FLC Expression by Chromatin Modifications.

(A) In rapid-flowering accessions (fri background), FLC is activated by ATX1 via the deposition of H3K4me3 marks at the FLC 5′ untranslated region during the vegetative phase.

(B) The H3K27me3 repressive mark is present but does not prevent FLC expression. EFS is required to prevent early flowering but does not modify the level of H3K4me3 marks at the FLC locus (Kim et al., 2005). The removal of H3K4me3, together with an increased level of H3K27me3 mark deposited by a still unknown PRC2 complex, leads to FLC repression and subsequent flowering.

(C) In winter annual accessions (FRI background), ATX1 together with EFS activates FLC expression via the deposition of H3K4me3 marks.

(D) A prolonged cold treatment (vernalization) induces the VRN2-PRC2 complex, which in turn represses FLC via the deposition of H3K27me3 marks.

The absence of H3K4me3 marks at FLC chromatin in the atx1-1 mutant is correlated with an accumulation of H3K27me2 (Figure 4B), a mark associated with FLC gene repression (Bastow et al., 2004). This suggests the presence of a default mechanism that represses FLC transcription in the absence of H3K4me3 marks. A similar mechanism was described in Drosophila, where the trxG proteins ASH1 and TRX have been proposed to counteract PcG repression, either by histone binding and/or H3K4-trimethylation, which subsequently prevents the binding of PcG proteins to HOX genes (Klymenko and Müller, 2004). The simultaneous binding of TRX and PcG proteins at the Ubx locus challenged this hypothesis (Papp and Müller, 2006). Recently, ASH1 binding at the Ubx locus was shown to correlate with H3K4me3 deposition and to occur only when Ubx is transcribed (Papp and Müller, 2006). Therefore, ASH1 was proposed to counteract PcG repression via the deposition of the H3K4me3 marks, which subsequently restricts H3K27 methylation in the promoter and coding regions. Whether or not this mechanism is conserved in plants will require more investigations; however, our study provides evidence for a similar mechanism in Arabidopsis using different histone marks.

In the atx1-1 mutant background, the accumulation of H3K27me3 marks was reduced in the promoter and the first intron, arguing for the requirement of ATX1 and/or the presence of H3K4me3 marks for the deposition of this repressive mark. A similar result was recently observed at the AG locus, where ATX1 is required for the trimethylation of H3K27 in the promoter and the downstream coding region (Saleh et al., 2007). By contrast, the atx1-1 mutation results in an increased level of another repressive mark at FLC, H3K27me2, showing that ATX1 activity is not required for the deposition of this repressive mark. This suggests that in the absence of the H3K4me3 mark in this region, the H3K27me3 mark is not required to repress FLC.

In summary, we demonstrate that ATX1 directly regulates the floral regulator FLC by mediating the H3K4me3 modification. Additionally, we show that H3K4me3 deposition is accompanied by a decrease in H3K27me2 levels at the FLC locus. Thus, we propose that the developmentally regulated binding of ATX1 and trimethylation of H3K4 at FLC chromatin counteract FLC silencing. Our study also shows that transition to flowering correlates with the release of ATX1 from the FLC locus and an increase of the level of H3K27me3 repressive marks, of which a critical level is required to achieve full repression of FLC (Shindo et al., 2006). This time- and dosage-dependent regulation resembles the vernalization process, where prolonged exposure to cold leads to progressive silencing of FLC (Chouard, 1960; Lang, 1965). Chromatin-mediated regulation of FLC, and probably other genes, is not an all-or-nothing process and fine-tuning may be achieved through different levels of histone modifications.

METHODS

Plant Material and Growth Conditions

Seeds, wild-type Ws, and atx1-1 (Ws) (Alvarez-Venegas et al., 2003), wild-type Col, atx1-2 (Col) (SALK_149002), atx2-1 (Col) (SALK_074806), and FRI (ColSf-2) plants, in which the flowering-time locus FRI has been introgressed from the Sf-2 accession into a Col background (Lee and Amasino, 1995), were grown on Murashige and Skoog media with 15 g/L of sucrose at 4°C for 2 d under short-day conditions (8/16 h day/night) with 10 μmol photons m−2 s−1 white light and then transferred to 20°C under either long-day (16/8 h day/night) or short-day conditions with 57 μmol photons m−2 s−1 white light. Luciferase imaging was as described by Mylne et al. (2004), and the images were obtained using a NightOwl imaging system (Berthold Technologies).

atx1-1 plants were genotyped using SP26 (forward) 5′-TCTATGCAGCTCTTTGCTAATTGG-3′ and TDNA-LB SP11 (reverse) 5′-GATGCACTCGAAATCAGCCAATTTTAGAC-3′ or SP26 (forward) and SP27 (reverse) 5′-AGCCCAGAGCATGAGCTTACC-3′ for the wild-type ATX1 gene. atx1-2 plants were genotyped using JM341 (forward) 5′-GGTATAGCTCATGCTCTGGGC-3′ and SALK-LB (reverse) 5′-CCAAACTGGAACAACACTCAAC-3′ or JM341 (forward) and JM340 (reverse) 5′-TCTCTTTTGTGGACTTGCTGTG-3′ for the wild-type ATX1 gene. atx2-1 plants were genotyped using JM345 (forward) 5′-GCTGCAAAGAACAAACTCTTCC-3′ and SALK-LB (reverse) or JM345 (forward) and JM346 (reverse) 5′-AGGCCACCAATAGCTGACAAG-3′ for the wild-type ATX2 gene.

RT-PCR Analyses

RT-PCR quantifications were performed with 10-d-old seedlings. RNA was isolated with the Trizol reagent (Invitrogen) according to the manufacturer's instructions. RT-PCR for FLC was performed using FLC-specific primers SP135 (forward) 5′-TTGGATCAGTCAAAAGC-3′ and SP136 (reverse) 5′-AGTAGTGGGAGAGTCACGGG-3′, and ACTIN2 (ACT2) control primers were SP105 (forward) 5′-GCCCTCGTTTGTGGGAATGG-3′ and SP106 (reverse) 5′-AAGCCTTTGATCTTGAGAGC-3′. Signal intensities using ethidium bromide staining (0.4 μg/mL) were normalized relative to ACTIN2 PCR products with ImageQuant software (Molecular Dynamics). These results are representative of three independent biological replicate experiments. Fold changes are the mean of three independent quantifications from three independent RNA extractions. Primer efficiency was tested to quantify FLC and ACT2 PCR product in the logarithmic phase.

In Situ Hybridization

Fixation and hybridization were performed as previously described (Köhler et al., 2003). Primers used to make the probes were as follows: ATX1, SPG63 (forward) 5′-AGCTGGATCCAGTCTGATGTCTAAGAAGG-3′ and SPG64 (reverse) 5′-ACGTGAATTCCCTTACACCTTCTTAAACC-3′; ATX2, SPG54 (forward) 5′-ATGCGGATCCGGAAGATCAGTCCTCGTAC-3′ and SPG55 (reverse) 5′-AGCTGAATTCTTTCTGAAGTTGATCCATC-3′; FLC, SPB1 (forward) 5′-AGCTGGATCCTTGGATCATCAGTCAAAAGC-3′ and SPB2 (reverse) 5′-AGCTGAATTCAGTAGTGG GAGAGTCACCGG-3′.

ChIP

ChIP was performed on 10-d-old seedlings and 16-d-old plants prior to flowering, grown under long-day conditions, as previously described (Köhler et al., 2003). Antibodies used were H3K4me2 (Upstate), H3K4me3 (Upstate), H3K27me2 (Upstate), H3K27me3 (Upstate), and ATX1 (GenScript). Primers for ACTIN2/7 and for FLC regions A (Bastow et al., 2004), B (He et al., 2004), and C (Bastow et al., 2004) were as previously described. PCR conditions were similar to the ones used by Bastow et al. (2004) and He et al. (2004), where analysis to show the amplification efficiency of all primer pairs used in the chromatin immunoprecipitation analysis has been published. Signal intensities using ethidium bromide staining (0.4 μg/mL) were normalized relative to ACTIN 2/7 PCR products with ImageQuant software (Molecular Dynamics), and the fold changes are expressed relative to the value of wild-type seedlings. Means are given with bars indicating 1 se.

Accession Numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative database under the following accession numbers: AG, At4g18960; AGL19, At4g22950; ATX1, At2g31650; ATX2, At1g05830; EFS, At1g77300; FCA, At4g16280; FLD, At3g10390; FLC, At5g10140; FRI, At4g00650; MEA, At1g02580; PHE1, At1g65330; STM, At1g62360.

Supplemental Data

The following materials are available in the online version of this article.

  • Supplemental Figure 1. Spatio-Temporal Expression Patterns of ATX2 in Wild-Type 10-d-Old Seedlings Assayed by in Situ Hybridization.

  • Supplemental Table 1. Steady State Message Levels in the atx1-1 Homozygous Mutant Compared with Wild-Type Ws.

  • Supplemental Table 2. Downregulated Genes in the atx1-1 Homozygous Mutant Compared with Wild-Type Ws.

  • Supplemental References.

Supplementary Material

[Supplemental Data]
plntcell_tpc.108.058172_index.html (760B, html)

Acknowledgments

S.P. thanks Nikolaus Amrhein (Eidgenössische Technische Hochschule) for his support and continuous interest in this project. We thank three anonymous reviewers for helpful suggestions, Sharon Kessler, Célia Baroux, Stephen Schauer, and Mark Curtis (all at the University of Zurich, Switzerland) for comments on the manuscript, and the ABRC for provision of the SALK T-DNA lines. This work was supported by an EMBO long-term fellowship to S.P., the University of Zurich, the Eidgenössische Technische Hochschule, and grants from the Swiss National Science Foundation and the European Union's FP6 Network of Excellence “The Epigenome” to U.G.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (www.plantcell.org) is: Ueli Grossniklaus (grossnik@botinst.uzh.ch).

[W]

Online version contains Web-only data.

References

  1. Alvarez-Venegas, R., and Avramova, Z. (2001). Two Arabidopsis homologs of the animal trithorax genes: A new structural domain is a signature feature of the trithorax gene family. Gene 271 215–221. [DOI] [PubMed] [Google Scholar]
  2. Alvarez-Venegas, R., and Avramova, Z. (2005). Methylation patterns of histone H3 Lys 4, Lys 9 and Lys 27 in transcriptionally active and inactive Arabidopsis genes and in atx1 mutants. Nucleic Acids Res. 33 5199–5207. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Alvarez-Venegas, R., Pien, S., Sadder, M., Witmer, X., Grossniklaus, U., and Avramova, Z. (2003). ATX1, an Arabidopsis homolog of trithorax, activates flower homeotic genes. Curr. Biol. 13 627–637. [DOI] [PubMed] [Google Scholar]
  4. Baroux, C., Gagliardini, V., Page, D.R., and Grossniklaus, U. (2006). Dynamic regulatory interactions of Polycomb group genes: MEDEA autoregulation is required for imprinted gene expression in Arabidopsis. Genes Dev. 20 1081–1086. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bastow, R., Mylne, J.S., Lister, C., Lippman, Z., Martienssen, R.A., and Dean, C. (2004). Vernalization requires epigenetic silencing of FLC by histone methylation. Nature 427 164–167. [DOI] [PubMed] [Google Scholar]
  6. Baumbusch, L.O., Thorstensen, T., Krauss, V., Fischer, A., Naumann, K., Assalkhou, R., Schulz, I., Reuter, G., and Aalen, R.B. (2001). The Arabidopsis thaliana genome contains at least 29 active genes encoding SET domain proteins that can be assigned to four evolutionarily conserved classes. Nucleic Acids Res. 29 4319–4333. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chouard, P. (1960). Vernalization and its relations to dormancy. Annu. Rev. Plant Physiol. 11 191–237. [Google Scholar]
  8. Fuchs, J., Demidov, D., Houben, A., and Schubert, I. (2006). Chromosomal histone modification patterns – From conservation to diversity. Trends Plant Sci. 11 199–208. [DOI] [PubMed] [Google Scholar]
  9. Gehring, M., Huh, J.H., Hsieh, T.F., Penterman, J., Choi, Y., Harada, J.J., Goldberg, R.B., and Fischer, R.L. (2006). DEMETER DNA glycosylase establishes MEDEA Polycomb gene self-imprinting by allele-specific demethylation. Cell 124 495–506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. He, Y., Doyle, M.R., and Amasino, R.M. (2004). PAF1-complex-mediated histone methylation of FLOWERING LOCUS C chromatin is required for the vernalization-responsive, winter-annual habit in Arabidopsis. Genes Dev. 18 2774–2784. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. He, Y., and Amasino, R.M. (2005). Role of chromatin modification in flowering-time control. Trends Plant Sci. 10 30–35. [DOI] [PubMed] [Google Scholar]
  12. Jullien, P.E., Katz, A., Oliva, M., Ohad, N., and Berger, F. (2006). Polycomb group complexes self-regulate imprinting of the Polycomb group gene MEDEA in Arabidopsis. Curr. Biol. 16 486–492. [DOI] [PubMed] [Google Scholar]
  13. Katsani, K.R., Arredondo, J.J., Kal, A.J., and Verrijzer, C.P. (2001). A homeotic mutation in the trithorax SET domain impedes histone binding. Genes Dev. 15 2197–2202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Kim, S.Y., He, Y., Jacob, Y., Noh, Y.S., Michaels, S., and Amasino, R. (2005). Establishment of the vernalization-responsive, winter-annual habit in Arabidopsis requires a putative histone H3 methyltransferase. Plant Cell 17 3301–3310. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Klymenko, T., and Müller, J. (2004). The histone methyltransferases Trithorax and Ash1 prevent transcriptional silencing by Polycomb group proteins. EMBO Rep. 5 373–377. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Köhler, C., Hennig, L., Spillane, C., Pien, S., Gruissem, W., and Grossniklaus, U. (2003). The Polycomb-group protein MEDEA regulates seed development by controlling expression of the MADS-box gene PHERES1. Genes Dev. 17 1540–1553. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Lachner, M., Sengupta, R., Schotta, G., and Jenuwein, T. (2004). Trilogies of histone lysine methylation as epigenetic landmarks of the eukaryotic genome. Cold Spring Harb. Symp. Quant. Biol. 69 209–218. [DOI] [PubMed] [Google Scholar]
  18. Lang, A. (1965). Physiology of flower initiation. In Encyclopedia of Plant Physiology, Vol. 15, W. Ruhland, ed (Berlin: Springer-Verlag), pp. 1380–1536.
  19. Lee, I., and Amasino, R. (1995). Effect of vernalization, photoperiod, and light quality on the flowering phenotype of Arabidopsis plants containing FRIGIDA gene. Plant Physiol. 108 157–162. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Levy, Y.Y., Mesnage, S., Mylne, J.S., Gendall, A.R., and Dean, C. (2002). Multiple roles of Arabidopsis VRN1 in vernalization and flowering time control. Science 297 243–246. [DOI] [PubMed] [Google Scholar]
  21. Liu, F., Quesada, V., Crevillén, P., Bäurle, I., Swiezewski, S., and Dean, C. (2007). The Arabidopsis RNA-binding protein FCA requires a lysine-specific demethylase 1 homolog to downregulate FLC. Mol. Cell 28 398–407. [DOI] [PubMed] [Google Scholar]
  22. Makarevich, G., Leroy, O., Akinci, U., Schubert, D., Clarenz, O., Goodrich, J., Grossniklaus, U., and Köhler, C. (2006). Different Polycomb group complexes regulate common target genes in Arabidopsis. EMBO Rep. 7 947–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mylne, J., Greb, T., Lister, C., and Dean, C. (2004). Epigenetic regulation in the control of flowering. In Epigenetics, Vol. LXIX, B. Stillman and D. Stewart, eds (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 457–464. [DOI] [PubMed]
  24. Papp, B., and Müller, J. (2006). Histone trimethylation and the maintenance of transcriptional ON and OFF states by trxG and PcG proteins. Genes Dev. 20 2041–2054. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Pien, S., and Grossniklaus, U. (2007). Polycomb group and trithorax group proteins in Arabidopsis. Biochim Biophys Acta. 1769 375–382. [DOI] [PubMed] [Google Scholar]
  26. Rea, S., Eisenhaber, F., O'Carroll, D., Strahl, B.D., Sun, Z.W., Schmid, M., Opravil, S., Mechtler, K., Ponting, C.P., Allis, C.D., and Jenuwein, T. (2000). Regulation of chromatin structure by site-specific histone H3 methyltransferases. Nature 406 593–599. [DOI] [PubMed] [Google Scholar]
  27. Saleh, A., Al-Abdallat, A., Ndamukong, I., Alvarez-Venegas, R., and Avramova, Z. (2007). The Arabidopsis homologs of trithorax (ATX1) and Enhancer of zeste (CLF) establish 'bivalent chromatin marks' at the silent AGAMOUS locus. Nucleic Acids Res. 35 6290–6296. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Schönrock, N., Bouveret, R., Leroy, O., Borghi, L., Köhler, C., Gruissem, W., and Hennig, L. (2006). Polycomb-group proteins repress the floral activator AGL19 in the FLC-dependent vernalization pathway. Genes Dev. 20 1667–1678. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Schubert, D., Primavesi, L., Bishopp, A., Roberts, G., Doonan, J., Jenuwein, T., and Goodrich, J. (2006). Silencing by plant Polycomb-group genes requires dispersed trimethylation of histone H3 at lysine 27. EMBO J. 25 4638–4649. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Shindo, C., Lister, C., Crevillen, P., Nordborg, M., and Dean, C. (2006). Variation in the epigenetic silencing of FLC contributes to natural variation in Arabidopsis vernalization response. Genes Dev. 20 3079–3083. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Simon, J.A., and Tamkun, J.W. (2002). Programming off and on states in chromatin: mechanisms of Polycomb and trithorax group complexes. Curr. Opin. Genet. Dev. 12 210–218. [DOI] [PubMed] [Google Scholar]
  32. Sung, S., Schmitz, R.J., and Amasino, R.M. (2006). A PHD finger protein involved in both the vernalization and photoperiod pathways in Arabidopsis. Genes Dev. 20 3244–3248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Tripoulas, N.A., Hersperger, E., La Jeunesse, D., and Shearn, A. (1994). Molecular genetic analysis of the Drosophila melanogaster gene absent, small or homeotic discs1 (ash1). Genetics 137 1027–1038. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Zhang, X., Clarenz, O., Cokus, S., Bernatavichute, Y.N., Pellegrini, M., Goodrich, J., and Jacobsen, S.E. (2007). Whole-genome analysis of histone H3 lysine 27 trimethylation in Arabidopsis. PLoS Biol. 5 e129. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

[Supplemental Data]
plntcell_tpc.108.058172_index.html (760B, html)
plntcell_tpc.108.058172_1.pdf (97.7KB, pdf)

Articles from The Plant Cell are provided here courtesy of Oxford University Press

RESOURCES