SET8-mediated methylations of histone H4 lysine 20 mark silent heterochromatic domains in apicomplexan genomes

Abstract

Posttranslational histone modifications modulate chromatin-templated processes in various biological systems. H4K20 methylation is considered to have an evolutionarily ancient role in DNA repair and genome integrity, while its function in heterochromatin function and gene expression is thought to have arisen later during evolution. Here, we identify and characterize H4K20 methylases of the Set8 family in Plasmodium and Toxoplasma, two medically important members of the protozoan phylum Apicomplexa. Remarkably, parasite Set8-related proteins display H4K20 mono-, di-, and trimethylase activities, in striking contrast to the monomethylase-restricted human Set8. Structurally, few residues forming the substrate-specific channel dictate enzyme methylation multiplicity. These enzymes are cell cycle regulated and focally enriched at pericentric and telomeric heterochromatin in both parasites. Collectively, our findings provide new insights into the evolution of Set8-mediated biochemical pathways, suggesting that the heterochromatic function of the marker is not restricted to metazoans. Thus, these lower eukaryotes have developed a diverse panel of biological stages through their high capacity to differentiate, and epigenetics only begins to emerge as a strong determinant of their biology.

FIG. 1.

A new SET8 family. (A) Schematic dendrogram showing the relationships among some of the better-characterized SET domain proteins. For a more detailed phylogenetic tree, see Fig. S1 in the supplemental material. Major lineages of the SET domain and T. gondii TgSet proteins are indicated next to the tree. (B) Expanded representation of the new SET8 family branch. (C and D) Costaining with anti-HA antibody and Hoechst 33258 of human fibroblast cells infected with transgenic T. gondii strain RH::HA-Flag-TgSet8. The transgene is driven by the parasite-specific strong promoter GRA1. aa, amino acids; Acc., accession no. (E) T. gondii whole-cell extract was fractionated by chromatography, and the fraction shown by an arrow was used for histone H4 affinity purification, followed by Western blotting with a specific antibody raised against TgSet8.

FIG. 2.

Methylation of T. gondii H4K20 is mediated by TgSet8. (A) Histones were extracted from the indicated organisms and fractionated by SDS-polyacrylamide gel electrophoresis. (Top) Western blot analysis with antibodies against H4K20me1 (39), H4K20me2 (Ab1, Upstate 07-441; Ab2, reference ; Ab3, Abcam ab9052), H4K20me3 (Abcam ab9053), and unmodified histone H4 (Abcam ab4559). (Bottom) Coomassie staining of histones. AcH4, acetylated H4. (B) Schematic representation of N-terminal deletion mutant forms of TgSet8. The amino acid (aa) numbers for each construct and the SET domain are indicated. (C) HMTase assays with TgSet8ΔN1 following incorporation of ³H-labeled CH₃ into recombinant histone substrates rH3, rH4, rH2A, and rH2B. (D) HMTase assays of TgSet8ΔN1 with recombinant His₆-H4 were analyzed by Western blotting. No cross-reaction with recombinant H4 was observed. (E) Time course HMTase assays of TgSet8ΔN1 and rHsSet8 (Upstate). Reactions of TgSet8ΔN1 or rHsSet8 with recombinant His₆-H4 were stopped at sequential incubation time points and analyzed by Western blotting. (F) Product specificity of rTgSet8ΔN1 determined by mass spectrometry. Deconvoluted electrospray mass spectrum of the SGRGKGGKGLGKGGAKRHRK₂₀VLR(D) peptide following AspN digestion as described in Materials and Methods. The molecular masses of the unmethylated and mono-, di-, and trimethylated peptides are marked above the peaks. (G) Gel filtration chromatographic analysis of recombinant TgSet8ΔN1 with a Superdex 200 column. The eluted fractions were analyzed by Coomassie staining, Western blot analysis, and HMTase assay.

FIG. 3.

Structural modeling of TgSet8ΔN5(1759-1893). (A) Primary sequence alignment of HsSet8(225-352) and TgSet8ΔN5(1759-1893) used for modeling with the program Modeler. The secondary-structure element annotations have been extracted from the HsSet8 1zkk Protein Data Bank structure (8). The image shown was produced with the program ESPRIT (14). (B) Superposition of the three-dimensional SET domains of HsSet8 (blue, 1zkkA) and the TgSet8ΔN5 Modeler model (pink). The TgSet8ΔN5 insertion sequences are indicated by arrows and the one-letter amino acid code. (C) The geometry and molecular environment of the lysine 20 binding site are shown for HsSet8 and TgSet8. (D) HsSet8 (blue), TgSet8 (pink), and DIM-5 (green) lysine 20 binding site residue superposition. The water molecule (red) belongs to the HsSet8 Protein Data Bank structure (1zkk).

FIG. 4.

Structural features of the TgSet8 catalytic site. The geometry and molecular environment of the lysine 20 binding site of wild-type (W.T.) (A) and mutant (B and C) TgSet8 are shown. (B) Time course HMTase assays of mutant TgSet8ΔN1(V1875Y) versus wild-type TgSet8 with recombinant His₆-H4 and analysis by Western blotting. A weak H4K20me2 signal is observed with long exposure times. (C and D) Time course HMTase assays of mutant TgSet8ΔN1(F1808Y) versus wild-type TgSet8 with recombinant His₆-H4 and analysis by Western blotting. (E) Influence of pH on the multiplicity of TgSet8ΔN1 on a His₆-H4 substrate.

FIG. 5.

H4K20me1 and H4K20me3 mark apicomplexan heterochromatin domains. (A) Costaining with anti-H4K20me1 antibody (green), anti-IMC1 antibody (red), and Hoechst 33258 (blue) of human fibroblast cells infected with T. gondii strain RH (tachyzoites). (B) Nuclear areas marked by H4K20me1 at a higher magnification. (C) Costaining with anti-H4K20me3 antibody (green), anti-IMC1 antibody (red), and Hoechst 33258 (blue) of human fibroblast cells infected with T. gondii strain RH (tachyzoites). (D) Nuclear areas marked by trimethyl H4K20me3 at a higher magnification. The cross-reactions at the apical end of Toxoplasma cells are frequently observed with antibodies raised against any trimethylated lysine residue. (E) Schizont stage P. falciparum parasites (strain 3D7) costained with anti-H4K20me3 antibody (green) and Hoechst 33258 (blue).

FIG. 6.

The H4K20me1 marker is cell cycle regulated. (A) T. gondii cell cycle scheme. S, S phase; M, mitosis; C, cytokinesis. (B) Costaining with anti-HA antibody (red) and anti-H4K20me1 antibody (green) of human fibroblast cells infected with transgenic T. gondii strain RH::HA-Flag-TgSet8. (C) Fluorescence-activated cell sorter analysis of HA-Flag-TgSet8ΔN1 protein level in relation to the cell cycle. Tachyzoites were labeled with anti-HA and anti-Toxoplasma antibodies. DNA contents were detected with propidium iodide (PI). The intensity of the antibody fluorescence signal is plotted against the propidium iodide signal. The circle indicates an increase in the intensity of TgSet8ΔN1 expression in the tachyzoite G₁ phase. (D) Costaining with anti-H3S10ph antibody (red), anti-H4K20me1 antibody (green), and Hoechst 33258 (blue) of human fibroblast cells infected with T. gondii strain RH. (E) Costaining with anti-H3S10ph antibody (red), anti-HA antibody (green), and Hoechst 33258 (blue) of human fibroblast cells infected with strain RH::HA-Flag-TgSet8. (F) Costaining with anti-IMC1 antibody (red), anti-H4K20me1 antibody (green), and Hoechst 33258 (blue) of human fibroblast cells infected with T. gondii strain RH (tachyzoites in mitosis).

FIG. 7.

H4K20 methylation states index heterochromatin in T. gondii. (A) Representative view of Toxoplasma ChIP-on-chip analysis. High-resolution mapping of both myc-Set8 binding sites and acetylated histone H4 (K5, K8, K12, and K16) locations on chromosome IX. The blue (H4ac) and green (myc-TgSet8) lines plot the median logarithmic ratio (log₂ R) of hybridization intensities in a 500-base window of genomic versus immunoprecipitated DNA. The x axis denotes the genomic position of each probe, and the y axis shows the normalized log ratio of Cy5 and Cy3 signals, and therefore the signal strength is represented by the height of the rectangle. The data were visualized by the SignalMap software (NimbleGen Systems, Inc.). (B) Close-up view of chromosome IX regions. The horizontal bars represent possible binding regions identified by the SignalMap software. The height of the bar is the signal strength of the probe. Thresholds of 1.3 and 2 were applied to myc-Set8 (yellow) and acetylated histone H4 (blue), respectively. rRNA gene loci are shown in red. Green boxes indicate ToxoDB gene annotations. The data were visualized by the GenoBrowser software. (C) Chromatin from T. gondii strain RH::HAMyc-TgSet8ΔN1 was immunoprecipitated with anti-myc, anti-H4K20me1, and anti-H4K20me3 antibodies; immunoprecipitated DNA was analyzed by PCR with specific primers selected in T. gondii rRNA gene loci (5S, 17S, and ITS1; see Table S1 in the supplemental material) and the gene for dihydrofolate reductase (DHFR) as a negative control. (D) Detailed view of myc-Set8 (yellow) and H4ac (blue) ChIP-on-chip data at the chromosome XI and Ib telomeric regions. (E to G) DNA obtained by ChIP was analyzed by real-time PCR with repeat-specific primer sets (see Table S1 in the supplemental material). These particular loci were chosen for study because their amplification by real-time PCR gave specific amplicons. Assays were done under conditions in which the product accumulated linearly with respect to the input DNA (see Materials and Methods in the supplemental material). Relative intensity is shown with standard deviations. Chromatin from T. gondii strain RH::HAMyc-TgSet8ΔN1 was immunoprecipitated with anti-myc (E), anti-H4K20me1 (F), and anti-H4K20me3 (G) antibodies. (H) Statistical analyses of the HAMyc-TgSet8ΔN1 ChIP signal repartition across chromosomes Ia and Ib (see Materials and Methods in the supplemental material). The enzyme is more frequently enriched in the intergenic region than within gene-rich and promoter-proximal regions. IgG, immunoglobulin G; Chr, chromosome.