Histone H3K4 methylation regulates deactivation of the spindle assembly checkpoint through direct binding of Mad2

Abstract

Histone H3 methylation on Lys4 (H3K4me) is associated with active gene transcription in all eukaryotes. In Saccharomyces cerevisiae, Set1 is the sole lysine methyltransferase required for mono-, di-, and trimethylation of this site. Although H3K4me3 is linked to gene expression, whether H3K4 methylation regulates other cellular processes, such as mitosis, is less clear. Here we show that both Set1 and H3K4 mutants display a benomyl resistance phenotype that requires components of the spindle assembly checkpoint (SAC), including Bub3 and Mad2. These proteins inhibit Cdc20, an activator of the anaphase-promoting complex/cyclosome (APC/C). Mutations in Cdc20 that block Mad2 interactions suppress the benomyl resistance of both set1 and H3K4 mutant cells. Furthermore, the HORMA domain in Mad2 directly binds H3, identifying a new histone H3 "reader" motif. Mad2 undergoes a conformational change important for execution of the SAC. We found that the closed (active) conformation of both yeast and human Mad2 is capable of binding methylated H3K4, but, in contrast, the open (inactive) Mad2 conformation limits interaction with methylated H3. Collectively, our data indicate that interactions between Mad2 and H3K4 regulate resolution of the SAC by limiting closed Mad2 availability for Cdc20 inhibition.

Keywords: H3K4; HORMA domain; Set1; chromatin; lysine methylation; spindle assembly checkpoint.

Figure 1.

Loss of COMPASS-mediated lysine methylation results in benomyl resistance. (A) A serial fivefold dilution assay of yeast strains with the indicated genotypes was performed to visualize growth phenotypes. Cells were placed onto YPD plates (YPD) with or without 30 µg/mL benomyl, and these plates were placed for 2 d at 30°C. (B) Whole-cell extracts (WCEs) were isolated from strains with the indicated genotypes. The amount of total Set1 protein was assessed by immunoblot analysis using an antibody specific for Set1 (1:100; Santa Cruz Biotechnology, SC-101858). Pgk1 protein levels were used as a loading control. (C) Both wild-type Set1 and Set1G951S proteins similarly associate with the COMPASS component Bre2. Immunoprecipitation (IP) assays were used to pull down Bre2, and the associated levels of Set1 were assessed in the wild-type (WT) and set1G951S strains. Bre2-TAP (α-protein A) and Pgk1 (α-Pgk1) were used as loading controls. (D) bre1Δ, set1Δ, and the catalytically inactive mutant set1G951S were subjected to the same assay as in A. (E) COMPASS mutants were assessed for growth phenotypes using the same serial fivefold dilution assay as in A and D.

Figure 2.

The set1G951S mutant requires the SAC and Cdc20 inhibition for benomyl resistance. (A) An illustration of Cdc20 associated with the APC complex in both an active and an inactive state. Mad2 and Mad3 bind to Cdc20 to inhibit the ubiquitination of APC substrates. (B–D) Serial fivefold dilution assay of yeast strains with the indicated genotypes grown on normal YPD plates (YPD) or plates containing 30 µg/mL benomyl for 2 d at 30°C.

Figure 3.

Loss of H3K4 methylation results in a thick mitotic spindle. (A) Cells bearing a SET1 catalytic mutant allele (set1G951S) display a more robust mitotic spindle staining compared with wild-type cells (WT). Immunofluorescent confocal images of wild-type and set1G951S mitotic cells are shown. Tubulin (green) and DAPI (blue) stainings were used to identify cells undergoing mitosis in an asynchronous culture. Mitotic cells are outlined in white. (B) Cell cycle arrest and release experiments reveal that Pds1 is more stable in SET1 mutants (set1G951S) than in wild-type cells. Cells were arrested in G2/M and released into fresh medium. Samples were taken at the indicated time points, and whole-cell extracts and immunoblots were prepared and probed with α-Myc antibody to assess the protein levels of Myc-tagged Pds1. Immunoblots of Pgk1 served as a loading control. Cells from the indicated time points were taken for analysis by flow cytometry. Histograms of DNA content reveal the cell cycle profile of the indicated strains at specific times after cell cycle arrest and release. (C) Serial fivefold dilution assays of wild-type and mutants with the indicated genotype placed onto either control plates (YPD) or plates containing 30 µg/mL benomyl as in previous figures. (D) Immunofluorescent confocal images of wild type and mutants with the indicated genotypes. Staining and selection of cells were preformed as in A.

Figure 4.

The H3K4R mutant requires the SAC and Cdc20 inhibition for benomyl resistance. (A,B) Serial fivefold dilution assay of wild-type (WT) and mutants with the indicated genotypes placed onto either control plates (YPD) or plates containing 30 µg/mL benomyl, as in previous figures. (C) Immunoblots of whole-cell extracts from wild-type, set1Δ, and mad2Δ cells to compare di- and trimethylation of histone H3 levels. Total H3 was used as a loading control. (D) Immunofluorescent confocal images of wild-type and mutants with the indicated genotypes. Tubulin (green) and DAPI (blue) were used to identify cells undergoing mitosis from an asynchronous culture. Mitotic cells are outlined in white, as in previous figures. (E) Serial fivefold dilution assay of wild-type and mutants with the indicated genotypes placed onto either control plates (YPD) or plates containing 30 µg/mL benomyl, as above.

Figure 5.

Mad2 binds H3 methylated at K4 in a conformation-specific manner. (A,C) Colloidal staining of purified recombinant proteins. Pull-down assays of the indicated recombinant proteins to identify direct binding with calf thymus histones, assayed using immunoblots of membranes probed with antibodies against H2A, H2B, or H3. (B,D) Pull-down assay with recombinant GST-Mad2 fusion proteins and unmodified H3 (K4me0) or MLA full-length histones generated to mimic H3K4 monomethylation (K4me1), H3K4 dimethylation (K4me2), or H3K4 trimethylation (K4me3). Direct binding was assayed using immunoblots of membranes probed with an antibody recognizing histone H3. (E,F) GST pull-down assays using recombinant GST-yMad2 and GST-hMad2 with wild-type histone H3 or H3 with various point mutations. Direct binding was assessed by using immunoblots of membranes probed with an antibody against H3.

Figure 6.

A model showing that methylated H3K4 binds to C-Mad2. An H3 sequence required for Mad2 binding, which is similar to the Mad2-binding consensus sequence, is underlined. Our in vivo and in vitro data indicate that H3K4me acts as a negative regulator of the SAC through direct binding of methylated H3K4 to an active C-Mad2, thereby limiting C-Mad2–Cdc20 interactions.