The WD40 Domain Protein MSI1 Functions in a Histone Deacetylase Complex to Fine-Tune Abscisic Acid Signaling

Abstract

MSI1 belongs to a family of histone binding WD40-repeat proteins. Arabidopsis thaliana contains five genes encoding MSI1-like proteins, but their functions in diverse chromatin-associated complexes are poorly understood. Here, we show that MSI1 is part of a histone deacetylase complex. We copurified HISTONE DEACETYLASE19 (HDA19) with MSI1 and transcriptional regulatory SIN3-like proteins and provide evidence that MSI1 and HDA19 associate into the same complex in vivo. These data suggest that MSI1, HDA19, and HISTONE DEACETYLATION COMPLEX1 protein form a core complex that can integrate various SIN3-like proteins. We found that reduction of MSI1 or HDA19 causes upregulation of abscisic acid (ABA) receptor genes and hypersensitivity of ABA-responsive genes. The MSI1-HDA19 complex fine-tunes ABA signaling by binding to the chromatin of ABA receptor genes and by maintaining low levels of acetylation of histone H3 at lysine 9, thereby affecting the expression levels of ABA receptor genes. Reduced MSI1 or HDA19 levels led to increased tolerance to salt stress corresponding to the increased ABA sensitivity of gene expression. Together, our results reveal the presence of an MSI1-HDA19 complex that fine-tunes ABA signaling in Arabidopsis.

Figure 1.

HDA19 Associates into the Same Complex with MSI1 in Vivo. (A) MSI1 and HDA19 are components of the same histone deacetylase complex. HDA19:GFP and MSI1 were found to be part of 500- to 600-kD complexes (red arrows) detected by anti-GFP and anti-MSI1 using native PAGE. No anti-GFP signal was detected in wild-type Col. Specificity of the anti-MSI1 antibody was shown before (Hennig et al., 2003). (B) MSI1 copurifies with HDA19. HDA19:GFP (red arrows) was precipitated from inflorescences of wild-type and Pro35S:HDA19:GFP plants. Precipitates were analyzed by immunoblotting using anti-MSI1 antibodies. (C) MSI1 interacts with HDA19 in a yeast two-hybrid assay. Only yeast cotransformed with bait (pGBKT7:MSI1) and prey (pGADT7:HDA19) plasmids grow on selective medium (GDO, quadruple dropout, SD/–Ade/–His/–Leu/–Trp) and give a positive X-Gal signal (DDO, double dropout, SD/–Leu/–Trp without or with X-α-Gal).

Figure 2.

Both msi1-as and hda19 Mutants Show Upregulation of ABA-Responsive Genes over Time in an ABA Dose-Dependent Manner. (A) to (C) Expression kinetics of RD29B, ANACO19, and COR15A. Relative change in expression over 8 h of RD29B, ANAC019, and COR15A in the wild type (gray circles), msi1-as (orange triangles), and hda19 (blue squares), respectively. Differences from the wild type are significant (two-sided t test, P ≤ 0.05) for all genes for time points at 3 h and longer and for ANAC019 also at 1 h. (D) to (F) ABA dose dependency of RD29B, ANACO19, and COR15A expression. Relative change in expression of RD29B, ANAC019, and COR15A in the wild type (gray circles), msi1-as (orange triangles), and hda19 (blue squares), respectively, treated with different concentrations of ABA for 4 h. Differences from the wild type are significant (P ≤ 0.05) for RD29B above 0.5 µM ABA, for ANAC019 at all concentrations, and for COR15A above 10 µM ABA. Expression levels are relative to PP2A. Graphs show the mean ± se of three biological replicates.

Figure 3.

Increased Transcript Levels of ABA Receptors in msi1-as and hda19. Relative expression of PYL4, PYL5, and PYL6 in the wild type (gray), msi1-as (orange), hda19 (blue), and msi1-as hda19 (green). Expression levels are relative to PP2A and further normalized to the wild type. Graphs show the mean ± se of three biological replicates. Asterisks indicate values that are significantly larger (P ≤ 0.05) than in the wild type. n.s. indicates no significant difference between msi1-as hda19 and hda19.

Figure 4.

MSI1 and HDA19 Bind to Chromatin at ABA Receptor Genes. ChIP was performed using 14-d-old Pro35S:GFP, ProMSI1:MSI1:GFP, and Pro35S:HDA19:GFP plants and anti-GFP antibodies. The graphs show MSI1 (yellow) and HDA19 (green) recovery at PYL4, PYL5, and PYL6 compared with a Pro35S:GFP control (black). Schemes above the diagrams represent the recovery at PYL4 (A), PYL5 (B), and PYL6 (C) loci. Black lines with Roman numerals represent the regions probed by PCR. Results of IgG control ChIP assays were plotted separately (Supplemental Figure 2). Values are recovery as percent of input; shown are mean ± se of three biological replicates. Asterisks indicate significant (P ≤ 0.001) difference from Pro35S:GFP controls.

Figure 5.

MSI1 and HDA19 Affect the H3K9 Acetylation Status at ABA Receptor Genes. The graphs show increased H3K9ac levels in msi1-as (orange) and hda19 (blue) compared with the wild type (gray) at the PYL4, PYL5, and PYL6 locus, respectively. Fragment labels below the graphs are as in Figure 4. Anti-H3K9ac ChIP-qPCR was performed using 14-d-old plants. Values represent recovery expressed as relative enrichment of H3K9ac compared with H3 levels. Results of IgG control ChIP assays were plotted separately (Supplemental Figure 3). Shown are means ± se of three biological replicates. Asterisks indicate significant (P ≤ 0.01) difference from the wild type.

Figure 6.

msi1-as and hda19 Have Increased Tolerance to Salt Stress. (A) Performance of plants grown on NaCl. msi1-as and hda19 plants remain green longer on 150 mM NaCl compared with the wild type. (B) Change in chlorophyll content over time on 150 mM NaCl. msi1-as (orange), hda19 (blue), and ABA-treated wild type (dark gray) have substantially more chlorophyll after 64 h than the wild type (gray). Differences from the wild type are significant (two-sided t test, P ≤ 0.05) for time points at 16 h and longer. (C) Effect of pretreatment with TSA for 48 h on changes in chlorophyll content during exposure to salt (150 mM). Wild type treated with TSA (dark gray) show delayed loss of chlorophyll compared with the untreated wild type (gray). Differences from the wild type are significant (P ≤ 0.05) for time points at 8 h and longer. Shown are means ± se for three biological replicates.

Figure 7.

Proposed Model of MSI1 Function in ABA Signaling. In the wild type, a MSI1-HDA19 SIN3-like complex deacetylates chromatin at ABA receptor genes such as PYL4 and thus maintains low transcription of the receptor genes (gray arrow). Increased expression of ABA receptor genes (green arrow) in msi1 or hda19 mutant plants causes increased expression of ABA-responsive genes in the presence of ABA (blue hexagons). Reduction of MSI1 or HDA19 levels has no strong effect on the expression of ABA-responsive genes in the absence of ABA.