The histone H3K79 methyltransferase Dot1L is essential for mammalian development and heterochromatin structure

Abstract

Dot1 is an evolutionarily conserved histone methyltransferase specific for lysine 79 of histone H3 (H3K79). In Saccharomyces cerevisiae, Dot1-mediated H3K79 methylation is associated with telomere silencing, meiotic checkpoint control, and DNA damage response. The biological function of H3K79 methylation in mammals, however, remains poorly understood. Using gene targeting, we generated mice deficient for Dot1L, the murine Dot1 homologue. Dot1L-deficient embryos show multiple developmental abnormalities, including growth impairment, angiogenesis defects in the yolk sac, and cardiac dilation, and die between 9.5 and 10.5 days post coitum. To gain insights into the cellular function of Dot1L, we derived embryonic stem (ES) cells from Dot1L mutant blastocysts. Dot1L-deficient ES cells show global loss of H3K79 methylation as well as reduced levels of heterochromatic marks (H3K9 di-methylation and H4K20 tri-methylation) at centromeres and telomeres. These changes are accompanied by aneuploidy, telomere elongation, and proliferation defects. Taken together, these results indicate that Dot1L and H3K79 methylation play important roles in heterochromatin formation and in embryonic development.

Figure 1. Generation of mutant Dot1L alleles in mice.

(A) Schematic depiction of the strategy used to generate the Dot1L^3lox and Dot1L^1lox alleles. The exons are numbered. The locations of the Southern probe and PCR primers (DF1, DR1, and DR2) used for genotyping, as well as the sizes of the diagnostic fragments recognized by the Southern probe, are indicated (E, EcoRI). loxP sites are shown as triangles. (B) Southern blot analysis of EcoRI-digested genomic DNA probed with an 860-bp 5′ probe external to the targeting vector. The presence of the 8.3-kb band confirms homologous recombination. (C) PCR genotyping of DNA from ES cells. WT, 485 bp; 1lox, 233 bp.

Figure 2. Essential role for Dot1L in mouse embryonic development.

(A) A representative X-gal stained 7.5-dpc Dot1L^3lox/+ embryo demonstrating ubiquitous Dot1L transcription throughout the embryo. (B) A representative X-gal stained 9.5-dpc Dot1L^3lox/+ embryo demonstrating ubiquitous Dot1L transcription throughout the embryo with elevated Dot1L expression in the indicated regions. (C) A representative X-gal stained 9.5-dpc Dot1L^3lox/+ yolk sac demonstrating Dot1L transcription in visceral endoderm, visceral mesoderm, and primitive erythrocytes. (D) Representative pictures of 9.5-dpc Dot1L^1lox/+ and Dot1L^1lox/1lox embryos. Dot1L^1lox/+ embryos (left) were indistinguishable from wild-type embryos. Most Dot1L^1lox/1lox embryos were undersized, had an enlarged heart (cardiac dilation) and stunted tail (center), while approximately 15% exhibited developmental arrest at E8.5 (right). (E) Representative pictures of a 10.5-dpc Dot1L^1lox/1lox embryo (right) and a heterozygous littermate (left). (F) Representative pictures showing the yolk sac vasculature of 9.5-dpc Dot1L^1lox/+ (left) and Dot1L^1lox/1lox (right) embryos. The vasculature of the Dot1L^1lox/1lox yolk sac is thinner and less organized than that of the heterozygous littermate.

Figure 3. Phenotypic analysis of Dot1L mutant ES cells.

(A) Western blot analysis using extracts from ES cell lines of the indicated Dot1L genotypes and antibodies specific for di-, and tri-methylated H3K79. Total histone H3 was used as a loading control. (B) Analysis of H3K79 methylation by mass spectrometry. Quantification of different forms of H3K79 methylation was obtained by comparing the extracted ion chromatogram (EIC) intensity of the ion signals corresponding to the unmodified (Me0), mono-methylated (Me1), di-methylated (Me2), and tri-methylated (Me3) K79-containing peptides. (C) The proliferation of Dot1L^+/+, Dot1L^1lox/+ and Dot1L^1lox/1lox ES cells was determined by doing cell counts every 24 hours for five days. Cells were grown in triplicate, and data shown is representative of three independent experiments. (D) The percentages of apoptotic cells in Dot1L^+/+, Dot1L^1lox/+ and Dot1L^1lox/1lox ES cell cultures. The asterisk indicates P<0.05 (Student t-test). ES cells were stained with propidium iodide (PI) and PE conjugated anti-annexin V antibodies and analyzed by FACS. Apoptotic cells were annexin V positive and PI negative. Cells were grown in triplicate, and data shown are representative of two independent experiments. (E) The percentages of each cell cycle stage in Dot1L^+/+, Dot1L^1lox/+ and Dot1L^1lox/1lox ES cell cultures as determined by PI staining and FACS.

Figure 4. Telomere elongation and aneuploidy in Dot1L-deficient ES cells.

(A) Telomere restriction fragment (TRF) analysis upon MboI digestion of genomic DNA from two independent ES cell clones of each of the genotypes: Dot1L^+/+, Dot1L^1lox/+ and Dot1L^1lox/1lox. Note the presence of high molecular weight TRFs in Dot1L^1lox/1lox cells, which correspond to longer telomeres. (B) Representative images generated during the Q-FISH assay showing metaphase spreads from Dot1L^+/+ and Dot1L^1lox/1lox ES cells labeled with a telomere-specific fluorescent probe. (C) Telomere length distribution of two independent ES cell clones of each of the Dot1L genotypes as determined by Q-FISH. Twenty metaphases of each ES cell clone were analyzed. Note the increase in mean telomere length (Mtl) in both clones of Dot1^1lox/1lox cells and the intermediate phenotype of Dot1^1lox/+ lines. The percentages of telomeres below 50 kb and above 100 kb in length are indicated. (D) Scatter plot of the chromosome number of a Dot1L^+/+ ES cell line and two Dot1L^1lox/1lox cell lines. Chromosome number was determined by manually counting chromosomes in chromosome spreads. Each point represents the chromosome number of a single cell (n represents the number of metaphase cells counted).

Figure 5. Increased APBs in Dot1L-deficient cells.

(A) Confocal microscopy images showing either TRF1 (telomere marker, green), PML (marker for PML bodies, red), or combined fluorescence (yellow if colocalize, indicated by arrows) in wild-type (+/+) and Dot1L-deficient (1lox/1lox) ES cells. Late-passage (p120) Dnmt3a/3b-deficient (3a−/−3b−/−) ES cells were used as a positive control. Circled are nuclei of cells. (B) Quantification of percentage of cells showing colocalization of telomeres with PML bodies. A cell was considered positive when it showed 2 or more colocalization events. An increased frequency of cells showing APBs was observed in Dot1^1lox/1lox cultures compared to wild-type controls (χ² test, P<0.001). (C) Quantification of the number of APBs per cell. Dot1^1lox/1lox cells showed a significant increase in the number of APBs compared to wild-type cells (χ² test, P<0.001).

Figure 6. Changes of heterochromatin structure in Dot1L-deficient ES cells.

(A) Quantitative real-time PCR results using DNA from Dot1L mutant and wild-type ES cells immunoprecipitated with antibodies specific for the indicated histone modifications or without an antibody (No Ab) and normalized using input DNA values. PCR primers specific for major satellite repeats, minor satellite repeats, the subtelomere region of chromosome 1 or the subtelomere region of chromosome 2 were used. (B) Dot blot analysis of ChIP DNA using either a telomere-specific probe or a major satellite repeat-specific probe. Input DNA at 1∶10, 1∶100 and 1∶1000 dilutions was used as a positive control. DNA precipitated from 2.5×10⁶ cells were used for each assay.