HISTONE DEACETYLASE19 interacts with HSL1 and participates in the repression of seed maturation genes in Arabidopsis seedlings

Abstract

The seed maturation genes are specifically and highly expressed during late embryogenesis. In this work, yeast two-hybrid, bimolecular fluorescence complementation, and coimmunoprecipitation assays revealed that HISTONE DEACETYLASE19 (HDA19) interacted with the HIGH-LEVEL EXPRESSION OF SUGAR-INDUCIBLE GENE2-LIKE1 (HSL1), and the zinc-finger CW [conserved Cys (C) and Trp (W) residues] domain of HSL1 was responsible for the interaction. Furthermore, we found that mutations in HDA19 resulted in the ectopic expression of seed maturation genes in seedlings, which was associated with increased levels of gene activation marks, such as Histone H3 acetylation (H3ac), Histone H4 acetylation (H4ac), and Histone H3 Lys 4 tri-methylation (H3K4me3), but decreased levels of the gene repression mark Histone H3 Lys 27 tri-methylation (H3K27me3) in the promoter and/or coding regions. In addition, elevated transcription of certain seed maturation genes was also found in the hsl1 mutant seedlings, which was also accompanied by the enrichment of gene activation marks but decreased levels of the gene repression mark. Chromatin immunoprecipitation assays showed that HDA19 could directly bind to the chromatin of the seed maturation genes. These results suggest that HDA19 and HSL1 may act together to repress seed maturation gene expression during germination. Further genetic analyses revealed that the homozygous hsl1 hda19 double mutants are embryonic lethal, suggesting that HDA19 and HSL1 may play a vital role during embryogenesis.

Figure 1.

HDA19 Interacts with HSL1 in Yeast Two-Hybrid Assays. Top: Test for autonomous activation of HDA19, HSI2, and HSL1. Bottom: Interaction assay with HSI2 and HSL1 as preys (fused to the AD) and HDA19 as a bait (fused to the BD). Interaction was determined by growth assay on medium lacking adenine. Dilutions (1, 10⁻¹, and 10⁻²) of saturated cultures were spotted onto the plates. [See online article for color version of this figure.]

Figure 2.

BiFC Visualization and Co-IP Experiments Show Interaction between HDA19 and HSL1. (A) Subcellular localization of HSL1. Protoplasts were isolated from Arabidopsis seedlings. Confocal images of transgenic protoplasts (35S_pro:YFP and 35S_pro:HSL1-YFP). Bars = 7.5 µm. (B) BiFC in Arabidopsis protoplasts showing the interaction between HDA19 and HSL1 in living cells. HDA19 fused with the C terminus of YFP (YC) and HSL1 fused with the N terminus of YFP (YN) were cotransfected into protoplasts and visualized using confocal microscopy. As a negative control, HSL1 fused with YN and empty vector YC as well as HDA19 fused with YC and empty vector YN were cotransfected into protoplasts. Bars = 7.5 μm. (C) Co-IP assay showing HDA19 interaction with HSL1. Tobacco leaves coexpressing 35S:HA-HSL1 and 35S:Myc-HDA19 or only 35S:Myc-HDA19 was used to immunoprecipitate with the anti-HA antibody, and the immunoblot was probed with the anti-Myc antibody.

Figure 3.

Different Regions Required for Interaction between HSL1 and HDA19 in Yeast Two-Hybrid Assays. (A) Left: Diagrams of HSL1 constructs for interaction studies. Quadrate boxes represent B3 domains, elliptical boxes represent ZF domains, and triangular boxes represent EAR motifs. Right: Diagrams of HDA19 constructs for interaction studies. Rounded rectangles represent HD domains of HDA19. In the front of the HD domain is unknown region 1, designated as UR1, and at the back of the HD domain is unknown region 2, designated as UR2. (B) Top: Interaction assay with HDA19 as a bait and different deletions of HSL1 as putative preys. Schemes of HSL1 domains and the different protein deletions are shown. Bottom: Interaction assay with the ZF domain of HSL1 as a prey and different deletions of HDA19 as putative baits. Schemes of HDA19 domains and the different protein deletions are shown. Interaction was determined by growth assays on medium lacking adenine. Dilutions (1, 10⁻¹, and 10⁻²) of saturated cultures were spotted onto the plates. [See online article for color version of this figure.]

Figure 4.

Gene Expression in Various Embryo Developmental Stages. (A) Expression of seed maturation genes including 2S2, 7S1, OLE1, and ABI3 in six stages of the embryo development (3 d, globular; 4 d, heart; 5 d, torpedo; 7 d, walking stick; 9 d, curled cotyledons; 10 d, green cotyledons). Data used for the analysis were retrieved from the database in TAIR (http://bbc.botany.utoronto.ca/efp/cgi-bin/efpWeb.cgi). The values shown are means + sd. (B) Expression of HSL1 and HDA19 in six stages of embryo development (3 d, globular; 4 d, heart; 5 d, torpedo; 7 d, walking stick; 9 d, curled cotyledons; 10 d, green cotyledons). Data used for the analysis were retrieved from the database in TAIR (http://bbc.botany.utoronto.ca/efp/cgi-bin/efpWeb.cgi). The values shown are means + sd.

Figure 5.

Expression of Seed Maturation Genes in Wild-Type (Ws) and hda19-1 Mutant Seedlings. The expression of 2S2, 7S1, CRA1, OLE1, LEC1, LEC2, ABI3, and FUS3 in wild-type (Ws) and hda19-1 seedlings grown on MS agar for 14 d was analyzed by qRT-PCR. Wild-type (Ws) RNA levels were designated as onefold. The expression values were normalized using ACT2 (A) and UBQ10 (B) as an internal control, respectively. Each experiment was repeated three biological replicates. Error bars represent se.

Figure 6.

Expression of Seed Maturation Genes in Wild-Type (Col) and hsl1 Mutants (hsl1-1 and hsl1-2) Seedlings. The expression of 2S2, 7S1, CRA1, OLE1, LEC1, LEC2, ABI3, and FUS3 in wild-type (Col), hsl1-1, and hsl1-2 seedlings grown on MS agar for 14 d was analyzed by qRT-PCR. Wild-type (Col) RNA levels were designated as onefold. The expression values were normalized using ACT2 (A) and UBQ10 (B) as an internal control, respectively. Each experiment was repeated three biological replicates. Error bars represent se.

Figure 7.

Levels of H3ac, H4ac, H3K4me3, and H3K27me3 in 2S2, 7S1, CRA1, LEC1, and LEC2 Chromatin. (A) Schematic structure of genomic sequences of 2S2, 7S1, CRA1, LEC1, and LEC2 and the regions examined by ChIP. (B) to (E) Relative levels of H3ac (B), H4ac (C), H3K4me3 (D), and H3K27me3 (E) at the proximal promoter (P1) and the coding region (C1) in 2S2, 7S1, CRA1, LEC1, and LEC2 in Ws wild-type and hda19-1 seedlings grown on MS agar for 14 d. The amounts of DNA after ChIP were normalized using ACT7 as an internal control. The value of Ws was designated as onefold. Each experiment was repeated with three biological replicates. Error bars represent se.

Figure 8.

Levels of H3ac, H4ac, H3K4me3, and H3K27me3 in 7S1, OLE1, ABI3, CRA1, and LEC2 Chromatin. (A) Schematic structure of genomic sequences of 7S1, OLE1, ABI3, CRA1, and LEC2 and the regions examined by ChIP. (B) to (E) Relative levels of H3ac (B), H4ac (C), H3K4me3 (D), and H3K27me3 (E) at the proximal promoter (P1) and the coding region (C1) in 7S1, OLE1, ABI3, CRA1, and LEC2 in wild-type (Col), hsl1-1, and hda19-2 seedlings grown on MS agar for 14 d. The results were normalized for the amount of input DNA. The value of Col was arbitrarily given as 1. Each experiment was repeated three biological replicates. Error bars represent se.

Figure 9.

Target Genes of HDA19 Identified by ChIP Followed by Real-Time PCR Analysis. (A) Schematic structure of genomic sequences of 7S1, LEC2, 2S2, CRA1, LEC1, and ACT7 and the regions examined by ChIP. (B) to (G) Transgenic plants expressing HDA19-GFP were subjected to ChIP analysis using an anti-GFP antibody. Wild-type (Col) plants were used as negative controls. 7S1 (B), LEC2 (C), 2S2 (D), CRA1 (E), LEC1 (F), and ACT7 (G) were chosen for target genes. The results were normalized for the amount of input DNA. The value of Col was arbitrarily given as 1. Each experiment was repeated three biological replicates. Error bars represent se.

Figure 10.

The Homozygous hsl1 hda19 Double Mutant Is Seed Lethal. (A) The dissection of immature siliques (top) and mature siliques (bottom) in hsl1-1^+/−/hda19-2^+/− plants compared with the siliques in the wild type (WT), hsl1-1, and hda19-2. Aborted seeds are highlighted by red arrows. Bars = 500 µm. (B) and (C) The phenotype of an abnormal embryo in the yellow seed (B) compared with the phenotype of a normal embryo in the green seed (C) from the same immature silique of hsl1-1^+/−/hda19-2^+/− plant. Bars = 50 µm. (D) The dissection of immature siliques (top) and mature siliques (bottom) in hsl1^−/−/hda19^+/− and hsl1^+/−/hda19^−/− plants. Aborted seeds are highlighted by red arrows. Bars = 500 µm.