Osmotic stress induces phosphorylation of histone H3 at threonine 3 in pericentromeric regions of Arabidopsis thaliana

Abstract

Histone phosphorylation plays key roles in stress-induced transcriptional reprogramming in metazoans but its function(s) in land plants has remained relatively unexplored. Here we report that an Arabidopsis mutant defective in At3g03940 and At5g18190, encoding closely related Ser/Thr protein kinases, shows pleiotropic phenotypes including dwarfism and hypersensitivity to osmotic/salt stress. The double mutant has reduced global levels of phosphorylated histone H3 threonine 3 (H3T3ph), which are not enhanced, unlike the response in the wild type, by drought-like treatments. Genome-wide analyses revealed increased H3T3ph, slight enhancement in trimethylated histone H3 lysine 4 (H3K4me3), and a modest decrease in histone H3 occupancy in pericentromeric/knob regions of wild-type plants under osmotic stress. However, despite these changes in heterochromatin, transposons and repeats remained transcriptionally repressed. In contrast, this reorganization of heterochromatin was mostly absent in the double mutant, which exhibited lower H3T3ph levels in pericentromeric regions even under normal environmental conditions. Interestingly, within actively transcribed protein-coding genes, H3T3ph density was minimal in 5' genic regions, coincidental with a peak of H3K4me3 accumulation. This pattern was not affected in the double mutant, implying the existence of additional H3T3 protein kinases in Arabidopsis. Our results suggest that At3g03940 and At5g18190 are involved in the phosphorylation of H3T3 in pericentromeric/knob regions and that this repressive epigenetic mark may be important for maintaining proper heterochromatic organization and, possibly, chromosome function(s).

Keywords: abiotic stress; epigenetics; heterochromatin; histone phosphorylation; protein kinase.

Fig. 1.

Characterization of Arabidopsis mutants defective in MUT9-like kinases. (A) Three-week old plants of T-DNA insertion lines defective in individual MUT9-like kinase genes and of the mlk1 mlk2 double mutant (dm). (B) Schematic representation of gene structures indicating coding sequences (black boxes) and untranslated regions (gray boxes). T-DNA insertion sites and the locations of RT-PCR primers are indicated by arrowheads and arrows, respectively. (C) RT-PCR analysis of transcript abundance in the T-DNA lines. POLYUBIQUITIN 10 (Ubi) was used as an internal standard. (D) MLK1 transcript levels and phenotypes of the indicated strains. A construct harboring the MLK1 cDNA, driven by the 35S CAMV promoter, was introduced into the dm background to rescue the dwarf phenotype.

Fig. 2.

Plant responses to osmotic/salt stress and H3T3 phosphorylation status. (A) Phenotypes of Arabidopsis lines subject to osmotic or salt stress. (B) Root elongation under the indicated treatments. (C) Relative root elongation. Values, normalized to those of Col-0 control, are means ± SD of three independent experiments. (D) Drought stress in soil-grown plants. Survival rate was calculated as percentage of plants regaining growth upon rewatering, following 7-d water withdrawal. Values indicate means ± SD of three independent experiments. (E and F) Immunoblot analyses of H3T3ph levels in plants grown under normal conditions (E) or subject to PEG osmotic stress for different periods (F). H3T3ph signal intensity normalized to that of histone H3 is indicated.

Fig. 3.

Genome-wide distribution of H3T3ph in wild-type Arabidopsis. (A) Chromosomal distribution of genes (blue) and transposable elements (TEs) (pink). (Horizontal scale bar, 5 Mb.) (Lower) Chromosomal distribution under normal environmental conditions of H3T3ph normalized to histone H3. The binary logarithm of the fold change (FC), average of two independent experiments, is shown. (B and C) Chromosomal distribution of normalized H3T3ph in plants grown under the conditions indicated on the right axis. (D) Chromosomal distribution of histone H3 in PEG-treated plants relative to that in well-watered controls. (E) Distribution of H3T3ph within TEs of different lengths in wild-type and dm plants under normal environmental conditions. Solid and dashed lines represent TEs located in gene rich (i.e., chromosomal arms) or pericentromeric regions, respectively. (F) Distribution of H3T3ph and H3K4me3 in active protein-coding genes in wild-type plants grown under normal environmental conditions. Genes were aligned at the transcription start sites or the transcription end sites within each length group. The ratio of reads was determined at 100-bp intervals.

Fig. 4.

Localization of the MLK1-GFP protein, heterochromatin organization, and transposon expression. (A) Subcellular localization of the MLK1-GFP fusion protein. Pseudocolored images of leaf cells are shown with nuclei indicated by DAPI staining. (Scale bar, 5 µm.) (B) Chromocenter organization of nuclei, stained with Hoechst 33342, from leaves of the wild type and dm. (Scale bar, 5 µm.) (C) The relative heterochromatin fraction (RHF) corresponds to the fluorescence intensity of all chromocenters relative to that of the entire nucleus (47). Fluorescence intensity of nuclei was determined with FociCounter (33). Values shown, normalized to those of Col-0 control, are means ± SD of three independent experiments (n = 50). (D) Transcript abundance of the indicated endogenous loci examined by real-time RT-qPCR. Values indicate normalized means ± SD of three independent experiments. POLYUBIQUITIN 10 was used as the internal standard.