Regulation of arabidopsis flowering by the histone mark readers MRG1/2 via interaction with CONSTANS to modulate FT expression

Abstract

Day-length is important for regulating the transition to reproductive development (flowering) in plants. In the model plant Arabidopsis thaliana, the transcription factor CONSTANS (CO) promotes expression of the florigen FLOWERING LOCUS T (FT), constituting a key flowering pathway under long-day photoperiods. Recent studies have revealed that FT expression is regulated by changes of histone modification marks of the FT chromatin, but the epigenetic regulators that directly interact with the CO protein have not been identified. Here, we show that the Arabidopsis Morf Related Gene (MRG) group proteins MRG1 and MRG2 act as H3K4me3/H3K36me3 readers and physically interact with CO to activate FT expression. In vitro binding analyses indicated that the chromodomains of MRG1 and MRG2 preferentially bind H3K4me3/H3K36me3 peptides. The mrg1 mrg2 double mutant exhibits reduced mRNA levels of FT, but not of CO, and shows a late-flowering phenotype under the long-day but not short-day photoperiod growth conditions. MRG2 associates with the chromatin of FT promoter in a way dependent of both CO and H3K4me3/H3K36me3. Vice versa, loss of MRG1 and MRG2 also impairs CO binding at the FT promoter. Crystal structure analyses of MRG2 bound with H3K4me3/H3K36me3 peptides together with mutagenesis analysis in planta further demonstrated that MRG2 function relies on its H3K4me3/H3K36me3-binding activity. Collectively, our results unravel a novel chromatin regulatory mechanism, linking functions of MRG1 and MRG2 proteins, H3K4/H3K36 methylations, and CO in FT activation in the photoperiodic regulation of flowering time in plants.

Figure 1. Chromodomains of MRG1 and MRG2 specifically bind to tri-methylated H3K4 and H3K36 in vitro.

A. Binding assays of N-terminal His-tagged chromodomains of MRG1 (His-MRG1N) and MRG2 (His-MRG2N) with H3 peptides containing varying degrees of methylation at K4, K9, K27, or K36. B. ITC measurements of binding between the MRG2 chromodomain and histone peptides. Top panel, MRG2 and H3K4me3 peptide; bottom panel, MRG2 and H3K36me3 peptide.

Figure 2. MRG1 and MRG2 act redundantly in flowering time control of Arabidopsis in the photoperiodic flowering pathway.

A. Gene structure of mrg1 and mrg2 mutant alleles. Dark boxes represent exons; lines represent introns; red triangles indicate T-DNA insertions. B. RT-PCR analysis of MRG1 and MRG2 expression in leaves of mrg1, mrg2, and mrg1 mrg2 plants. ACTIN2 was used as the internal control. C. Phenotypes of the wild-type (WT), the single mutants mrg1 and mrg2, and the double mutant mrg1 mrg2 grown under long-day photoperiods (LD; 16 h light: 8 h dark). D. Phenotypes of WT and mrg1 mrg2 plants grown under short-day photoperiods (SD; 8 h light: 16 h dark). E. Flowering time, as measured by rosette leaf number at bolting, in plants grown under LD and SD conditions. The mean value from 20 plants is shown. Error bars represent standard deviations. F. Tissue expression pattern analyses of MRG1 and MRG2 by histochemical GUS staining in P_MRG1::MRG1-GUS and P_MRG2::MRG2-GUS transgenic plants. Staining was performed at different times after seed germination (day 5 or 12), and in inflorescences.

Figure 3. MRG1/2 are required for normal activation of FT expression.

A. Diurnal expression of CO and FT genes in wild-type (WT) and mrg1 mrg2 double mutant. Time is expressed as hours from dawn (ZT, zeitgeber time). Values are normalized to ACTIN2. Error bars show standard deviation from three biological replicates. B. Top panel, flowering time of the wild-type (WT), mrg1 mrg2 double mutant, transgenic 35S::FT plants in wild-type (35S::FT) or in mrg1 mrg2 (35S::FT/mrg1 mrg2), transgenic 35S::MYC-CO plants in wild-type (35S::MYC-CO), in heterozygous for mrg1 mrg2 (35S::MYC-CO/mrg1(+/−) mrg2(+/−)), or in homozygous for mrg1 mrg2 (35S::MYC-CO/mrg1 mrg2) grown in LDs. The mean value from 20 plants is shown. Error bars represent standard deviations. Bottom panel, Relative expression of CO and FT in indicated genotypes at ZT16. Values are presented as logarithmic mode (lg) of relative expression normalized to ACTIN2. Error bars show standard deviation from three biological replicates. A single asterisk indicates the statistically significant difference between 35S::MYC-CO/mrg1 mrg2 plants and the wild-type (P<0.05), and double asterisks indicate the statistically significant difference between 35S::MYC-CO/mrg1 mrg2 plants and the mrg1 mrg2 double mutants (P<0.05).

Figure 4. MRG2 binds the chromatin at the FT promoter.

A. Protein levels of MRG2 in WT, mrg1 mrg2, sdg8, and atx1 plants at ZT16 verified by the antibody against MRG2. B. ChIP analysis using @MRG2, @H3K4me3 or @H3K36me3 at FT chromatin in WT, mrg1 mrg2, sdg8, and atx1 plants at ZT16. Error bars show standard deviation from three biological replicates. Asterisks indicate statistically significant differences between the indicated genotypes and the wild-type (P<0.01). C. Relative expression levels of MRG1 and MRG2 in indicated genotypes at ZT16. Values are normalized to ACTIN2. Error bars show standard deviation from three biological replicates.

Figure 5. MRG2 physically interacts with CO.

A. Top panel, flowering time of WT, mrg1 mrg2, co and co mrg1 mrg2, as measured by rosette leaf number at bolting, in plants grown under LD conditions. The mean value from 20 plants is shown. Error bars represent standard deviations. Bottom panel, relative expression of FT in indicated genotypes at ZT16. Values are normalized to ACTIN2. Error bars show standard deviation from three biological replicates. B. Interaction of purified His-tagged MRG2 and GST-fused CO from E. coli by pull-down assays. C. BiFC analysis of the interaction between MRG2 and CO in tobacco leaf cells. Bar  = 50 µm. D. Co-IP detection of MRG2 and CO interaction in planta. Total protein extracts from plants expressing YFP-MRG2 alone, MYC-CO alone, or both YFP-MRG2 and MYC-CO were immunoprecipitated with anti-MYC affinity beads and the resulting fractions analyzed by western blot using anti-GFP antibodies (top panel) or HRP-conjugated anti-MYC monoclonal antibodies (bottom panel).

Figure 6. MRG2 and CO enhance each other's binding to the FT promoter.

A. Enrichment of MRG2 at the FT promoter at ZT16 in wild-type and co mutant at ZT16. Error bars show standard deviation from three biological replicates. Asterisks indicate statistically significant differences between co and the wild-type (P<0.01). B. Enrichment of MYC-CO at the FT promoter at ZT16 in indicated genotypes at ZT16. Error bars show standard deviation from three biological replicates. Asterisks indicate statistically significant differences between 35S::MYC-CO/mrg1 mrg2 plants and 35S::MYC-CO plants (P<0.01). C. Protein levels of MYC-tagged CO in indicated genotypes at ZT16.

Figure 7. MRG1/2 binding to histone marks is required for their function in regulating flowering.

A. Structure of the MRG2 chromodomain in complex with H3K36me3. The overall structure of the MRG2 chromodomain consisting of four β strands is shown in the left panel. The tri-methylated lysine sticking in the aromatic cage was surrounded by five conserved residues (right panel). Tri-methylated lysine is shown in purple; residues forming the aromatic cage are shown in cyan. B. Binding assays of His-tagged chromodomains of MRG2 (His-MRG2N) and a Y87A substitution in MRG2 chromodomain (His-MRG2N(Y87A)) with H3K4me3 and H3K36me3. C. Phenotypes of wild-type (WT), mrg1 mrg2, P_MRG2::MRG2-YFP/mrg1 mrg2 (MRG2-YFP/mrg1 mrg2), and P_MRG2::MRG2(Y87A)-YFP/mrg1 mrg2 (MRG2(Y87A)-YFP/mrg1 mrg2) plants grown under long-day photoperiods (LD; 16 h light: 8 h dark), noting that the mrg1 mrg2 double mutants have bolted. D. Flowering time of indicated genotypes, as measured by rosette leaf number at bolting, in plants grown under LD conditions. The mean value from 20 plants is shown. Error bars represent standard deviations. E. ChIP analysis using @GFP at FT chromatin in indicated genotypes at ZT16. Error bars show standard deviation from three biological replicates. Asterisks indicate statistically significant differences between MRG2(Y87A)-YFP/mrg1 mrg2 plants and MRG2-YFP/mrg1 mrg2 plants (P<0.01).