Histone H3 Lys36 (H3K36) methylation and its associated modifiers are crucial for DNA double-strand break (DSB) repair, but the mechanism governing whether and how different H3K36 methylation forms impact repair pathways is unclear. Here, we unveil the distinct roles of H3K36 dimethylation (H3K36me2) and H3K36 trimethylation (H3K36me3) in DSB repair via non-homologous end joining (NHEJ) or homologous recombination (HR). Yeast cells lacking H3K36me2 or H3K36me3 exhibit reduced NHEJ or HR efficiency. yKu70 and Rfa1 bind H3K36me2- or H3K36me3-modified peptides and chromatin, respectively. Disrupting these interactions impairs yKu70 and Rfa1 recruitment to damaged H3K36me2- or H3K36me3-rich loci, increasing DNA damage sensitivity and decreasing repair efficiency. Conversely, H3K36me2-enriched intergenic regions and H3K36me3-enriched gene bodies independently recruit yKu70 or Rfa1 under DSB stress. Importantly, human KU70 and RPA1, the homologs of yKu70 and Rfa1, exclusively associate with H3K36me2 and H3K36me3 in a conserved manner. These findings provide valuable insights into how H3K36me2 and H3K36me3 regulate distinct DSB repair pathways, highlighting H3K36 methylation as a critical element in the choice of DSB repair pathway.

Epigenetic dysregulation is an important contributor to carcinogenesis. This is not surprising, as chromatin-genomic DNA organized around structural histone scaffolding-serves as the template on which occurs essential nuclear processes, such as transcription, DNA replication and DNA repair. Histone H3 lysine 36 (H3K36) methyltransferases, such as the SET-domain 2 protein (SETD2), have emerged as critical tumor suppressors. Previous work on mammalian SETD2 and its counterpart in model organisms, Set2, has highlighted the role of this protein in governing genomic stability through transcriptional elongation and splicing, as well as in DNA damage response processes and cell cycle progression. A compendium of SETD2 mutations have been documented, garnered from sequenced cancer patient genome data, and these findings underscore the cancer-driving properties of SETD2 loss-of-function. In this review, we consolidate the molecular mechanisms regulated by SETD2/Set2 and discuss evidence of its dysregulation in tumorigenesis. Insight into the genetic interactions that exist between SETD2 and various canonical intracellular signaling pathways has not only empowered pharmacological intervention by taking advantage of synthetic lethality but underscores SETD2 as a druggable target for precision cancer therapy.

A nucleosome contains two copies of each histone H2A, H2B, H3 and H4. Histone H3 K4me0 and K36me3 are two key chromatin marks for de novo DNA methylation catalyzed by DNA methyltransferases in mammals. However, it remains unclear whether K4me0 and K36me3 marks on both sister histone H3s regulate de novo DNA methylation independently or cooperatively. Here, taking advantage of the bivalent histone H3 system in yeast, we examined the contributions of K4 and K36 on sister histone H3s to genomic DNA methylation catalyzed by ectopically co-expressed murine Dnmt3a and Dnmt3L. The results show that lack of both K4me0 and K36me3 on one sister H3 tail, or lack of K4me0 and K36me3 on respective sister H3s results in a dramatic reduction of 5mC, revealing a synergy of two sister H3s in DNA methylation regulation. Accordingly, the Dnmt3a or Dnmt3L mutation that disrupts the interaction of Dnmt3aADD domain-H3K4me0, Dnmt3LADD domain-H3K4me0, or Dnmt3aPWWP domain-H3K36me3 causes a significant reduction of DNA methylation. These results support the model that each heterodimeric Dnmt3a-Dnmt3L reads both K4me0 and K36me3 marks on one tail of sister H3s, and the dimer of heterodimeric Dnmt3a-Dnmt3L recognizes two tails of sister histone H3s to efficiently execute de novo DNA methylation.

Intron retention (IR), the most prevalent alternative splicing form in plants, plays a critical role in gene expression during plant development and stress response. However, the molecular mechanisms underlying IR regulation remain largely unknown.
Knockdown of SDG725, a histone H3 lysine 36 (H3K36)-specific methyltransferase in rice, leads to alterations of IR in more than 4700 genes. Surprisingly, IR events are globally increased at the 5\' region but decreased at the 3\' region of the gene body in the SDG725-knockdown mutant. Chromatin immunoprecipitation sequencing analyses reveal that SDG725 depletion results in a genome-wide increase of the H3K36 mono-methylation (H3K36me1) but, unexpectedly, promoter-proximal shifts of H3K36 di- and tri-methylation (H3K36me2 and H3K36me3). Consistent with the results in animals, the levels of H3K36me1/me2/me3 in rice positively correlate with gene expression levels, whereas shifts of H3K36me2/me3 coincide with position-specific alterations of IR. We find that either H3K36me2 or H3K36me3 alone contributes to the positional change of IR caused by SDG725 knockdown, although IR shift is more significant when both H3K36me2 and H3K36me3 modifications are simultaneously shifted.
Our results revealed that SDG725 modulates IR in a position-specific manner, indicating that H3K36 methylation plays a role in RNA splicing, probably by marking the retained introns in plants.

The Drosophila Ash1 protein is a trithorax-group (trxG) regulator that antagonizes Polycomb repression at HOX genes. Ash1 di-methylates lysine 36 in histone H3 (H3K36me2) but how this activity is controlled and at which genes it functions is not well understood. We show that Ash1 protein purified from Drosophila exists in a complex with MRG15 and Caf1 that we named AMC. In Drosophila and human AMC, MRG15 binds a conserved FxLP motif near the Ash1 SET domain and stimulates H3K36 di-methylation on nucleosomes. Drosophila MRG15-null and ash1 catalytic mutants show remarkably specific trxG phenotypes: stochastic loss of HOX gene expression and homeotic transformations in adults. In mutants lacking AMC, H3K36me2 bulk levels appear undiminished but H3K36me2 is reduced in the chromatin of HOX and other AMC-regulated genes. AMC therefore appears to act on top of the H3K36me2/me3 landscape generated by the major H3K36 methyltransferases NSD and Set2. Our analyses suggest that H3K36 di-methylation at HOX genes is the crucial physiological function of AMC and the mechanism by which the complex antagonizes Polycomb repression at these genes.

Chromatin modification through histone H3 lysine 36 methylation by the SETD2 tumor suppressor plays a key role in maintaining genome stability. Here, we describe a role for Set2-dependent H3K36 methylation in facilitating DNA replication and the transcriptional responses to both replication stress and DNA damage through promoting MluI cell-cycle box (MCB) binding factor (MBF)-complex-dependent transcription in fission yeast. Set2 loss leads to reduced MBF-dependent ribonucleotide reductase (RNR) expression, reduced deoxyribonucleoside triphosphate (dNTP) synthesis, altered replication origin firing, and a checkpoint-dependent S-phase delay. Accordingly, prolonged S phase in the absence of Set2 is suppressed by increasing dNTP synthesis. Furthermore, H3K36 is di- and tri-methylated at these MBF gene promoters, and Set2 loss leads to reduced MBF binding and transcription in response to genotoxic stress. Together, these findings provide new insights into how H3K36 methylation facilitates DNA replication and promotes genotoxic stress responses in fission yeast.

Missense mutations in Ten-eleven translocation 2 (TET2) gene are frequently found in leukaemia patients. Although mutations span the entire coding region, they tend to cluster in the C-terminal enzymatic domain and a cysteine-rich (CR) domain of unknown function. Herein, we found the CR domain binds chromatin preferentially at the histone H3 tail by recognising H3 lysine 36 mono- and dimethylation (H3K36me1/2). Importantly, missense mutations in the CR domain perturbed TET2 recruitment to the target locus and its enzymatic activities. Our findings identify a novel H3K36me recognition domain and uncover a critical link between histone modification and DNA hydroxylation in leukaemogenesis.