Dot1 histone methyltransferases share a distributive mechanism but have highly diverged catalytic properties

Abstract

The conserved histone methyltransferase Dot1 establishes an H3K79 methylation pattern consisting of mono-, di- and trimethylation states on histone H3 via a distributive mechanism. This mechanism has been shown to be important for the regulation of the different H3K79 methylation states in yeast. Dot1 enzymes in yeast, Trypanosoma brucei (TbDot1A and TbDot1B, which methylate H3K76) and human (hDot1L) generate very divergent methylation patterns. To understand how these species-specific methylation patterns are generated, the methylation output of the Dot1 enzymes was compared by expressing them in yeast at various expression levels. Computational simulations based on these data showed that the Dot1 enzymes have highly distinct catalytic properties, but share a distributive mechanism. The mechanism of methylation and the distinct rate constants have implications for the regulation of H3K79/K76 methylation. A mathematical model of H3K76 methylation during the trypanosome cell cycle suggests that temporally-regulated consecutive action of TbDot1A and TbDot1B is required for the observed regulation of H3K76 methylation states.

Figure 1. Quantitative H3K79 methylation analysis on a series of Dot1 alleles.

A) Schematic overview of yeast Dot1, trypanosome Dot1A and Dot1B, and human Dot1L. The enzymes share a conserved catalytic domain (grey box, CD). In addition, yDot1 contains an N-terminal domain with a lysine-rich domain (light grey and white shaded box, respectively). Human Dot1L was expressed with part of its C-terminal domain that has weak similarity with the yDot1 N-terminal domain. B) Western-blot analysis of H3K79 methylation and yDot1 expression using specific antibodies. H3 and Pgk1 were used as loading controls. A wild-type yeast strain (NKI6061; WT) was used as a reference throughout the manuscript; its H3K79 methylation levels were determined by mass spectrometry. A dot1Δ (–) was included to determine antibody specificity. Dot1 enzymes were expressed from the following promoters at the endogenous yeast DOT1 locus: DOT1pr (D), TEFpr + yDOT1 5’UTR (T*), GPDpr + yDOT1 5’UTR (G*), ADHpr (A), TEFpr (T) or GPDpr (G). H3K79 methylation patterns confirmed by mass spectrometry (MS) are indicated with ♦. C) Western-blot analysis of a series of reference yeast strains (Y1-Y7) containing a range of known H3K79 methylation levels (see below) to determine the linearity of the H3K79me1, -me2 and -me3 home-made (hm) antibodies and the H3K79me2 antibody from Millipore (mp). Y1-Y7 refers to strains: NKI6081, NKI6077, NKI6084, NKI6099, NKI6100, NKI6085 and NKI6083. D) Samples described in (C) were quantified using an Odyssey scanner and by MS. To determine the linearity of the H3K79 methylation antibodies, MS data were plotted against the quantified western-blot data and a non-linear regression fit was performed (see Supplemental Methods). The linear function resulting from the fit was used to correct the quantified western-blot data described in this manuscript and to subsequently estimate the unmethylated H3 fraction. No linear fit was obtained for the H3K79me2 home-made antibody; the low correlation between MS and western-blot data of the H3K79me2 home made antibody was caused by cross-reactivity of this antiserum with H3K79me3. Unless stated otherwise, H3K79me2 was quantified using the Millipore antibody, since this showed very little cross-reactivity to H3K79me3.

Figure 2. Quantitative assessment of H3K79/K76 methylation by TbDot1A, TbDot1B and hDot1L.

Western-blot analysis of H3K79 methylation states generated by TbDot1A (A), TbDot1B (B) and hDot1L (C) upon expression from the yDOT1 locus in yeast. H3K79 methylation states were determined using specific H3K79 methylation antibodies and the Dot1 enzymes were detected using their Myc and/or Flag tags. H3 and Pgk1 were used as loading controls. H3K79 methylation patterns that were confirmed by mass spectrometry are indicated with ♦. A-B) TbDot1A (A) and TbDot1B (B) enzymes were expressed from the same promoter series as described in Fig. 1B. C) hDot1L was expressed in yeast from the GAL1 promoter at the endogenous yDOT1 locus (Flag-hDot1L-9xMyc) upon induction by 2% galactose as the sole carbon source in rich medium. D) hDot1L was expressed in yeast from the GAL1 promoter at the endogenous yDOT1 locus (Flag-hDot1L-9xMyc) or at a multi-copy plasmid (2µ-hDot1L-Flag) upon induction by 2% galactose as the sole carbon source in the minimal medium. Human Dot1L protein expression levels are shown in Supplemental Fig. S4.

Figure 3. Dot1 enzymes share a distributive mode of action.

Experimental data of H3K79 methylation patterns generated by yDot1, TbDot1A, and TbDot1B in yeast and simulations using the models for distributive and processive enzymes as described in De Vos et al (2011). For each enzyme, the experimental H3K79 methylation pattern data that passed the quality test (presented in Fig. S3 and Supplemental Table SIII) were used to model the kinetic mechanism of the Dot1 enzymes. For yDot1, data from Frederiks et al were included. A subset of this dataset was confirmed by mass spectrometry analysis (see Supplemental Table SIII for numeric data). Simulations for hDot1L are shown in Supplemental Fig. S4. Each graph represents the H3K79 methylation patterns of yeast strains expressing a Dot1 enzyme at different concentrations. Each tick on the x-axis represents a single yeast strain. Simulations with the distributive model generated a better fit with the experimental data than the simulations with the processive model, as shown by the lower sum of squares values (SS).

Figure 4. Kinetics of H3K79 methylation of new histones.

Dot1 methylation activity on newly synthesized histones was tracked using a modified version of the Recombination-Induced Tag Exchange (RITE) assay in yeast. A) The only gene present encoding for histone H3 (HHT2) was tagged with a RITE cassette (LoxP-T7-Hyg-LoxP-HAHis), which resulted in expression of H3 tagged with T7, also referred to as the “old” histone. Addition of β-estradiol in log phase induced recombination between the LoxP sites (and deletion of the T7-hygromycin (HYG) cassette), resulting in the production of “new” H3 tagged with HAHis. Since recombination occurs asynchronously and takes a few hours before it is completed in a population of cells, the percentage of cells in the population that underwent RITE and produced H3-HAHis increased during the course of the experiment. As a consequence, H3 with a new tag represented a mix of histones of different ages; this age distribution changed over time. B-D) Western-blot analysis of H3K79 methylation levels on new (upper band) and old (lower band) histone H3 in the presence of yDot1 (B), TbDot1A-HA-TAP (C) and TbDot1B-HA-TAP (D). Samples were taken upon one, two or three population doublings (Doubling = 1, 2, or 3). As controls, samples were taken immediately after the addition of β-estradiol (100% “old”, Doubling = 0), and from strains that expressed only the H3-HAHis (100% “new”, ∞). Antibodies against H2A (yDot1) or H2B (TbDot1A and -B) were used as loading controls. The percentage of recombination was determined for each sample at the different time points. The graphs represent the average H3K79 methylation levels on newly synthesized histone H3 of which at least two replicates were available. H3K79 methylation states (relative to H2A/H2B and normalized to the new H3 signal of the LoxP blot) were converted into estimated absolute methylation levels by using as a reference the 100% new strain (∞; see Supplemental Fig. S5).

Figure 5. Simulation of H3K76 methylation throughout the trypanosome cell cycle.

Simulation of H3K76 methylation in a 13-hour procyclic trypanosome cell cycle divided in an S-phase (20%), G2/M-phase (40%) and G1-phase (40%). TbDot1A and -B expression was manually set to approximately fit the reported experimental data (see main text). Furthermore, adjustments were made in the model to simulate H3K76 methylation patterns in TbDot1A or -B depleted or overexpressing trypanosomes. A) Expression of TbDot1A in G2/M and G1 with 1500 copies and expression of TbDot1B only during G1 with 3000 copies resulted in a model that simulated the experimental data best. B) TbDot1B was excluded from the model to simulate a TbDot1B knockout. C) To simulate TbDot1A overexpression studies, TbDot1A expression was increased to 3000 copies per cell and continuously expressed (c) during the cell cycle. D) To simulate TbDot1A knockdown studies, TbDot1A expression was lowered to 600 copies per cell. E) TbDot1A was excluded from the model to simulate a TbDot1A knockout. F) To simulate TbDot1B misregulation studies, TbDot1B was continuously expressed (c) during the cell cycle.

Figure 6. Expression of TbDot1A or TbDot1B in yeast affects silencing but not replication.

A) Cell cycle progression of yeast strains overexpressing yDot1, TbDot1A or TbDot1B was determined by flow cytometry (FACS) analysis of DNA content and compared to the wild-type strain NKI6061 (WT). X-axis depicts DNA signal, Y-axis depicts cell count. The promoter from which the Dot1 enzymes were expressed is indicated behind the dash: A = ADHpr, G = GPDpr, D = DOT1pr. Strain Y7092 was used as the WT control for yDot1-G. B) To test whether the altered H3K79 methylation profiles affected silencing of the silent HMLα mating locus yeast strains expressing yDot1, TbDot1A or TbDot1B in a MATa ADE2 + leu2Δ background were mated with a WT strain with a MATα ade2Δ LEU2 + background (BY4726). Loss of HMLα silencing in MATa cells leads to loss of mating type identity and loss of mating ability. Cells were plated in 10-fold serial dilutions on selective synthetic media. Diploids that result from effective mating were ADE2 + LEU2 + and able to grow on –ADE –LEU media. C) Telomeric silencing was examined in a strain (UCC7164) containing a telomeric URA3 reporter (TEL-VII-L) and telomeric ADE2 reporter (ADE2-TEL-VR). Plasmid-based Dot1 expression was induced with 3% galactose. Cells were plated in 10-fold serial dilutions on selective synthetic media with or without 5-FOA. Cells that silence URA3 can grow on 5-FOA media whereas cells that express URA3 cannot. Cells that silence ADE2 accumulate a red pigment whereas cells that express ADE2 are white. TbDot1A G138A and TbDot1B G121A show H3K79 methylation activity, while TbDot1A 138R and TbDot1B G121R have lost their ability to methylate H3K79. D) Yeast strains expressing yDot1, no Dot1, partially active yDot1 (ΔN and G401A), TbDot1A or TbDot1B were plated in 10-fold dilution series on rich media with or without 100 mM hydroxyurea (HU). Loss of mating type identity by TbDot1A or –B expression (see panel B) was eliminated by deletion of the HMLα1/2 genes.