A chromatin-wide transition to H4K20 monomethylation impairs genome integrity and programmed DNA rearrangements in the mouse

Abstract

H4K20 methylation is a broad chromatin modification that has been linked with diverse epigenetic functions. Several enzymes target H4K20 methylation, consistent with distinct mono-, di-, and trimethylation states controlling different biological outputs. To analyze the roles of H4K20 methylation states, we generated conditional null alleles for the two Suv4-20h histone methyltransferase (HMTase) genes in the mouse. Suv4-20h-double-null (dn) mice are perinatally lethal and have lost nearly all H4K20me3 and H4K20me2 states. The genome-wide transition to an H4K20me1 state results in increased sensitivity to damaging stress, since Suv4-20h-dn chromatin is less efficient for DNA double-strand break (DSB) repair and prone to chromosomal aberrations. Notably, Suv4-20h-dn B cells are defective in immunoglobulin class-switch recombination, and Suv4-20h-dn deficiency impairs the stem cell pool of lymphoid progenitors. Thus, conversion to an H4K20me1 state results in compromised chromatin that is insufficient to protect genome integrity and to process a DNA-rearranging differentiation program in the mouse.

Figure 1.

Suv4-20h-dn mice display perinatal lethality. (A) Generation of Suv4-20h1 and Suv4-20h2 conditional knockouts. For both genes, targeting constructs (thick black bar) were generated that contain loxP sites in introns flanking the SET domain coding region. To remove the FRT-flanked neomycin resistance cassette, Suv4-20h^neo/+ mice were crossed with Flp recombinase expressing mice. Mox2-Cre was used for germline deletion (confirmed by PCR, bottom panel). Northern blot analysis of four embryonic stages indicates ubiquitous expression of Suv4-20h1 (left panel) and only low abundance of Suv4-20h2 (right panel). (B) Homozygous mutant mice were generated by intercrossing Suv4-20h1^+/− or Suv4-20h2^+/−. To generate Suv4-20h-dn mice, doubly heterozygous Suv4-20h1^+/−; Suv4-20h2^+/− mice were intercrossed. Suv4-20h1-null and Suv4-20h-dn embryos were born at sub-Mendelian ratios, are smaller than their littermates, and die perinatally.

Figure 2.

A genome-wide transition to H4K20me1 in Suv4-20h-dn pMEFs. (A) In wild-type pMEFs, H4K20me1 and H4K20me2 are broadly nuclear, H4K20me3 is focally enriched at pericentric heterochromatin. In female cells, H4K20me1 shows enrichment at the inactive X chromosome (arrows). In Suv4-20h1^−/− cells, H4K20me2 is reduced, whereas Suv4-20h2^−/− cells show loss of pericentric H4K20me3. In Suv4-20h-dn pMEFs, both H4K20me2 and H4K20me3 are strongly reduced with a concomitant increase in H4K20me1. (B) Overall H4K20 methylation levels were quantified by mass-spec of bulk histones. In wild-type cells, H4K20me2 is the most prominent mark present in ∼85% of all histone H4 molecules. In Suv4-20h-dn cells H4K20me2 and H4K20me3 are markedly reduced resulting in a genome-wide switch to H4K20me1.

Figure 3.

Suv4-20h-dn pMEFs display proliferation and cell cycle defects. (A) Proliferation of wild-type and Suv4-20h-dn pMEFs was measured by counting cell numbers of serially passaged cells. Suv4-20h-dn cells show less cumulative growth and plateau earlier than wild-type cells. (B) DNA content of wild-type and Suv4-20h-dn pMEFs was determined by FACS analysis of PI-labeled early (P1) and later passage (P3) cells. Suv4-20h-dn cells display broader G1 and G2 peaks at passage 3. (C) Exponentially growing pMEFs (P1) were pulse-labeled with BrdU and percentages of G1-, S-, and G2-phase cells were determined by FACS analysis. (D) Early passage pMEFs were synchronized in G0 by serum deprivation. Re-entry into S Phase was measured by BrdU pulse-labeling after serum add-back (one representative data-set of three independent experiments is shown). (E) Logarithmically growing pMEFs were treated with 2Gy IR. One hour post-IR treatment, either nocodazole or caffeine plus nocodazole were added to the medium, and 4–5 hours later mitotic spreads were prepared. Mitotic chromosomes of Suv4-20h-dn cells show more chromosomal gaps (arrows) than wild-type cells. (F) Summary of chromosomal gaps and breaks in >100 mitotic chromosome preparations. Asterisks indicate statistically significant differences (Wilcoxon-Mann-Whitney test, P < 0.05).

Figure 4.

Increased DNA damage sensitivity in H4K20me1 chromatin. (A) Wild-type and Suv4-20h-dn pMEFs were treated with 2 Gy IR, fixed after different time-points (1–10 min), and stained with antibodies against γH2A.X and 53BP1. (B) Numbers of 53BP1-positive cells (more than one focus per cell) were counted at different time-points post-treatment with 2 Gy IR. At early time-points (2 min), Suv4-20h-dn cells display reduced focal enrichment of 53BP1. At later time points (10 min), wild-type and Suv4-20h-dn cells show comparable 53BP1 foci. (C) Colony formation assays to assess sensitivity of Suv4-20h-dn and 53BP1-null cells to various types of DNA damage. Wild-type, 53BP1-null, and Suv4-20h-dn MEFs were plated at low density and treated with indicated doses of IR, etoposide, hydroxyurea, and UV. The percentage of surviving colonies relative to the nontreated samples is plotted.

Figure 5.

Suv4-20h-dn B cells are defective in CSR. (A) Wild-type and Suv4-20h-dn^vav B cells were labeled with CFSE and stimulated with LPS/IL-4. Comparable proliferation rates of wild-type and Suv4-20h-dn^vav B cells are indicated by a similar decrease of CFSE intensity. (B) Number of IgG₁-expressing B cells was determined 4 d after stimulation with LPS/IL-4 by FACS analysis. (C) Box plot representations for the percentage of switched B cells from eight (IgG₁) or four (IgG₃) independent CSR experiments. (D) Relative expression levels of GLTs, circle transcript, and Aid were analyzed by quantitative RT–PCR 4 d after stimulation with LPS/IL-4 (normalized to the average of Gapdh, HPRT, and tubulin expression levels). The bottom diagram shows a schematic representation of CSR between Sμ and Sγ1 constant regions in the IgH locus.

Figure 6.

Altered chromatin structure and increased chromosomal aberrations in Suv4-20h-dn^vav B cells. (A) Histone modifications at the Igμ and Igγ1 promoter and their associated switch regions of the IgH locus were determined by directed ChIP in resting and activated B cells. The top diagram indicates the approximate positions of the primer pairs. Histone modifications for major satellite repeats and tubulin served as controls. (B) IgH-associated chromosomal abnormalities were analyzed in metaphase spreads of activated wild-type and Suv4-20h-dn^vav B cells using chromosome painting in combination with IgH-FISH. In wild-type B cells, the IgH locus (green dots) is located on chromosome 12 (blue), proximal to the telomeres (red dots). Elevated levels of IgH-associated translocations (T) or Deletions (Del) were observed in Suv4-20h-dn^vav B cells. The box plot shows percentages of IgH-associated abnormalities in six independent wild-type (362 spreads) and Suv4-20h-dn^vav (379 spreads) B-cell preparations.

Figure 7.

Suv4-20h deficiency impairs the stem cell potential for lymphoid cells. (A) Bone marrow, spleen, and thymus of eight independent 2-mo-old wild-type and Suv4-20h-dn^vav mice were analyzed by FACS for different B-cell populations, myeloid lineages (macrophages, granulocytes), and DP T cells. Asterisks indicate statistically significant differences (t-test, P < 0.05). (B) Competitive reconstitution experiments (1:1) were performed with equal numbers of wild-type (CD45.1) and Suv4-20h-dn^vav (CD45.2) bone marrow cells, which were injected into lethally irradiated mice. The ratio between the CD45.1 and CD45.2 markers was analyzed 11 mo post-transplantation in the different hematopoietic lineages. Whereas myeloid cells show significant (20%–30%) contribution of Suv4-20h-dn^vav cells, <1% of the lymphoid lineages (B and T cells) are derived from Suv4-20h-dn^vav.