Oncofusion

start / graph / mark-h3k79me2

Mechanismstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q3b-dot1l-h3k79

H3K79me2 - the di-methyl state the leukaemia actually runs on, and the one menin turns out to govern only half of

also called H3K79 dimethylation · H3K79me2 · histone H3 lysine 79 dimethylation

H3K79me2 is two methyl groups on the same histone position, and it is the specific state that sits in broad blocks over the genes a KMT2A fusion switches on. Measuring it is how the field decided DOT1L was a drug target, and a 2023 screen showed that only part of it is controlled by menin - the rest is put there by the cell's own machinery.

The same thing, for a scientist

H3K79me2 is the state this disease is defined by. In an MLL-AF9 leukaemia-stem-cell model, profiling H3K79me2, H3K4me3, H3K27me3 and H3K36me3 found abnormal profiles **only** for H3K79me2 at fusion target loci, and Dot1l inactivation collapsed the fusion's direct targets while leaving global expression largely unaffected [E1]. The level is graded rather than binary: structure-guided point mutations that weaken the DOT1L-AF9 interaction cause graded loss of H3K79me2/me3 at target genes and a correspondingly graded loss of transformation [E2]. Functionally it is not merely a gene-body mark - a subset of enhancers (KEEs) require H3K79me2/3 for accessibility, acetylation, transcription-factor binding and enhancer-promoter contact in MLL-AF4 cells [E3] - and the broadest H3K79me2 domains in leukaemia lines cover the oncogenic loci, with a 100 kb domain over *HOXA9* in MOLM-13 [E4]. The finding that reorganises the field is that CRISPR-ChIP partitions H3K79 methylation into **two regulatory units**: an oncogenic DOT1L complex directed by the MLL fusion in a **menin-dependent** manner, and a separate endogenous DOT1L complex whose catalysis is directed by **MLLT10** [E5]. That is simultaneously the strongest mechanistic support the menin-interface hypothesis has received from an unbiased screen and the clearest statement of what menin inhibition will not reach.

Go deeper
Two regulatory units, and what each one implies

CRISPR-ChIP couples a pooled CRISPR screen to chromatin immunoprecipitation on an integrated reporter, so it asks which genes are required to maintain a histone mark rather than which genes are near it. Applied to H3K79me2 in MLL-translocated cells it splits the mark in two 5:

  • an oncogenic DOT1L complex, directed to the fusion's targets by the MLL fusion protein, requiring menin;
  • an endogenous DOT1L complex whose catalytic direction comes from MLLT10/AF10.

For the menin-interface hypothesis this cuts both ways and the graph should hold both halves. It supports the hypothesis, because an unbiased functional screen recovered menin as a requirement for the oncogenic half of the very mark the disease runs on, in a system that did not assume the menin-KMT2A interaction. It also bounds the hypothesis: a menin inhibitor is predicted to leave the MLLT10-directed endogenous unit intact, and the endogenous unit is exactly what normal haematopoiesis uses. A residual H3K79me2 signal after menin inhibition is therefore an expected observation and not a treatment failure, which is a useful thing to know before designing a pharmacodynamic endpoint around it.

The partition also explains a fusion-specific asymmetry the graph already has to police. KMT2A::MLLT10 fuses the AF10 half of the endogenous unit onto KMT2A; KMT2A::AFF1 contains no AF10 at all and reaches DOT1L differently. A claim that begins "H3K79me2 in KMT2A-rearranged leukaemia is menin-dependent" is not automatically transferable across partners.

Graded, not switch-like - and why the pinometostat schedule matters

The single most under-used fact about this mark is that the amount matters. Kuntimaddi et al. made point mutations across three DOT1L binding sites on AF9 and found graded reduction in recruitment produced graded loss of H3K79me2 and me3 at MLL-AF9 targets, with the degree of recruitment tracking transformation potential 2.

A graded input-output relationship has a direct clinical consequence. It means the clinically relevant quantity is the time-integrated depth of mark depletion, not whether an inhibitor is bound at a moment. Pinometostat was given as a continuous intravenous infusion and its maximum tolerated dose was never reached 6, so the trial sampled a part of a dose-response curve whose shape, from this biology, is the part that matters least.

This is also why the mark's slow kinetics belong in the same argument: see mechanism-h3k79-turnover-and-dilution.

The mark is not only in gene bodies

Two results move H3K79me2 out of the 'passive elongation mark' category. Godfrey et al. define KEEs - putative enhancers carrying H3K79me2/3 - in multiple cell types, and show with EPZ-5676 in MLL-AF4 cells that H3K79me2/3 is required for accessibility, histone acetylation and transcription-factor binding specifically at KEEs and not at non-KEE enhancers, and for maintaining enhancer-promoter interactions at a subset of them 3. Sharma et al. show that the genes with the broadest H3K79me2 domains in leukaemic lines are known leukaemia genes, that HOXA9 in MOLM-13 sits inside a ~100 kb domain shared with other HOXA and oncogenic lncRNA genes, and that this broad domain has enhanced chromatin accessibility throughout 4.

Broad-domain logic is familiar from H3K4me3, where breadth marks cell-identity genes in normal cells and oncogenes in cancer. That the same geometry appears for H3K79me2 suggests the leukaemic feature is domain breadth rather than peak height, and breadth is not what a bulk western-blot pharmacodynamic assay reports.

Facts and figures
markH3K79me2
histoneH3
residueK79, nucleosome lateral surface (not the N-terminal tail)
writergene-dot1l
dose responsegraded - graded loss of DOT1L-AF9 interaction gives graded loss of H3K79me2/3 and graded loss of transformation
erasers knownnone established; see mechanism-h3k79-demethylation-contested
pharmacodynamic useH3K79 methylation reduction was the pharmacodynamic endpoint of the pinometostat adult phase 1 (PMID 29724899)
normal tissue roleaccumulates during murine B-cell maturation and is required for pre-B expansion and marginal-zone development (PMID 40972012, see mechanism-dot1l-normal-b-lymphopoiesis)
regulatory units
[
  {
    "unit": "oncogenic",
    "director": "MLL fusion protein",
    "requirement": "gene-men1",
    "source_pmid": "37679565"
  },
  {
    "unit": "endogenous",
    "director": "gene-mllt10",
    "requirement": "MLLT10-directed catalysis",
    "source_pmid": "37679565"
  }
]
recruitment routes to fusion
[
  "MLLT1/MLLT3 AHD",
  "MLLT10 OM-LZ via the DOT1L coiled coil"
]
genomic features
[
  "broad gene-body domains over HOXA (~100 kb in MOLM-13)",
  "H3K79me2/3 enhancer elements (KEEs)"
]
How this could bite us back
  • The endogenous MLLT10-directed DOT1L unit is the one normal haematopoiesis uses, so any intervention that removes total H3K79me2 rather than the oncogenic fraction takes normal progenitors with it - which is the mechanistic basis of the cytopenias seen when Dot1l is deleted in adult marrow.
  • H3K79me2 is required for normal B-cell development, including pre-B expansion and BCR repertoire breadth, so in B-ALL the host compartment being treated is the compartment that needs the mark.
  • A pharmacodynamic endpoint defined as 'H3K79 methylation reduction' measures the sum of two regulatory units with different therapeutic meanings, and will report success for depleting the wrong one.
Still open
  1. Does the menin-dependent oncogenic H3K79me2 unit exist in KMT2A::AFF1, where there is no AF10 and the DOT1L route runs through AFF1-associated complexes? The CRISPR-ChIP partition was established in an MLL-translocated reporter system and the partner is not stated in the abstract.
  2. Is the leukaemic feature domain breadth or peak height? No experiment has tested breadth as a causal variable, and the only functional selectivity shown so far is at enhancers rather than at the broadest gene-body domains.
  3. After a menin inhibitor, how much H3K79me2 remains, and is the residue the MLLT10-directed endogenous unit? If so, residual H3K79me2 is a predicted finding rather than incomplete response, and a pharmacodynamic endpoint built on total H3K79me2 will misread it.
Every source behind this entry (6)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    In an MLL-AF9 leukaemia model, epigenetic profiling of H3K79me2, H3K4me3, H3K27me3 and H3K36me3 in haematopoietic progenitors and leukaemia stem cells found abnormal profiles only for H3K79me2 at MLL-AF9 fusion target loci; Dot1l inactivation downregulated direct targets and the MLL-translocation gene signature while global gene expression remained largely unaffected, and MLL-AF9 leukaemia depended on Dot1l in vivo.

    MLL-rearranged leukemia is dependent on aberrant H3K79 methylation by DOT1L. · Bernt KM, Zhu N, Sinha AU, et al. · Cancer Cell 2011
    PMID 21741597DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. Mouse MLL-AF9 model - a myeloid fusion, not the KMT2A::AFF1 that dominates infant ALL. Genetic Dot1l inactivation, not pharmacological inhibition, so it removes the protein and its non-catalytic functions together with the mark. 'Abnormal profiles only for H3K79me2' is bounded by the four marks that were profiled.

    Evidence that argues against this (1)
    • Removing Dot1l is not the same experiment as removing H3K79 methylation. A catalytically dead DOT1L knock-in behaves differently from a DOT1L knockout in cell-fate determination and in transcription elongation, so in vivo dependence on Dot1l cannot be read straight across to dependence on the mark. Cao K et al. 2020
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    DOT1L has three AF9 binding sites; structure-guided point mutations produce graded reductions in DOT1L recruitment to MLL-AF9, graded differential loss of H3K79me2 and H3K79me3 at MLL-AF9 target genes by ChIP-seq, and the degree of recruitment is linked to the level of MLL-AF9 haematopoietic transformation.

    PMID 25921540DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. NMR solution structure of a DOT1L-AF9 complex plus mutational series in an MLL-AF9 system; the graded relationship is established for AF9 and not for AFF1, MLLT1 or MLLT10, and the three binding sites are a property of the DOT1L-AF9 pair. Transformation is measured in a mouse haematopoietic assay.

    Evidence that argues against this (1)
    • A graded recruitment-to-mark relationship does not license a graded mark-to-survival relationship. Some KMT2A-rearranged leukaemias die at pinometostat concentrations far below those needed to collapse HOXA9/MEIS1, via H3K79me2-dependent pathways running through FLT3, so the dose-response for killing can be steeper and left-shifted relative to the dose-response for the canonical programme. Richter WF et al. 2021
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Putative enhancers carrying H3K79me2/3 (H3K79me2/3 enhancer elements, KEEs) exist in multiple cell types, and in MLL-AF4 leukaemia cells treated with EPZ-5676 H3K79me2/3 is required for chromatin accessibility, histone acetylation and transcription-factor binding specifically at KEEs but not at non-KEE enhancers, and for maintaining enhancer-promoter interactions at a subset of KEEs.

    DOT1L inhibition reveals a distinct subset of enhancers dependent on H3K79 methylation. · Godfrey L, Crump NT, Thorne R, et al. · Nat Commun 2019
    PMID 31243293DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. The functional experiments are in MLL-AF4 cell lines with a small-molecule inhibitor, so the effect is on the catalytic output and the timing is set by inhibitor exposure. 'A subset of KEEs' is not quantified in the abstract. Enhancer calls depend on the definition used and on antibody specificity across me2 and me3.

    Evidence that argues against this (1)
    • An inhibitor experiment cannot separate loss of the mark from loss of DOT1L's non-catalytic activity at those elements, because DOT1L remains bound. Cryo-EM plus single-molecule FRET show DOT1L destabilising the nucleosome in a non-catalytic manner that H2B ubiquitination enhances, so a catalytically inhibited but chromatin-bound DOT1L is not an inert control. Jang S et al. 2019
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Genes carrying the broadest H3K79me2 domains in human leukaemic cell lines are known leukaemia genes; in the MLL-rearranged line MOLM-13 HOXA9 lies within a ~100 kb H3K79me2 domain shared with other HOXA protein-coding and oncogenic long non-coding RNA genes, and that broad domain shows enhanced chromatin accessibility throughout.

    Bioinformatic Analyses of Broad H3K79me2 Domains in Different Leukemia Cell Line Data Sets. · Sharma P, Sattarifard H, Fatemiyan N, et al. · Cells 2022
    PMID 36139405DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. Re-analysis of existing public ChIP-seq datasets from immortalised leukaemia cell lines, not primary patient blasts; no perturbation and no functional test that breadth rather than height is the causal feature. Domain-breadth calls are sensitive to sequencing depth, normalisation and peak-caller choice, and datasets generated in different laboratories are compared.

    Evidence that argues against this (1)
    • Breadth may be a consequence rather than a cause. H3K79me2 is deposited co-transcriptionally across gene bodies, so the broadest domains are expected over the longest and most highly transcribed loci, and the HOXA cluster in an MLL-rearranged line is both. The functional test that exists - inhibitor treatment - shows selective dependence at a subset of ENHANCERS rather than at the broadest gene-body domains. Godfrey L et al. 2019
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    CRISPR-ChIP reveals a functional partitioning of H3K79 methylation into two distinct regulatory units: an oncogenic DOT1L complex directed by the MLL fusion protein in a menin-dependent manner, and a separate endogenous DOT1L complex whose catalytic activity is directed by MLLT10.

    CRISPR-ChIP reveals selective regulation of H3K79me2 by Menin in MLL leukemia. · Gilan O, Talarmain L, Bell CC, et al. · Nat Struct Mol Biol 2023
    PMID 37679565DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. The readout is an integrated synthetic reporter containing a cis-regulatory element and a single guide RNA, not the endogenous HOXA locus, so 'directed by the MLL fusion protein' is established on a chromatinised template designed for the assay. **The fusion is now resolved and it is not the index fusion: murine MLL-AF9.** It is not named in the abstract, and the full text is unreachable, but the study's GEO depositions name it per sample. In GSE192555 all eight H3K79me2 ChIP-seq samples are Mus musculus with `cell type: MLLAF9`, including both DOT1L-KO guides, both MLLT10-KO guides and **the menin-inhibitor arm** (GSM5751590). GSE192561 is 24/24 murine. Across all 65 samples in the three series, **none mentions AFF1, AF4 or any AFF-family gene**. The human arm is K562 — a CML line with no KMT2A rearrangement — carrying H3K4me3, Pol II, and five H3K79me2 samples used as a mouse/human library-mix species-spike normalisation under a DOT1L inhibitor, not as a fusion system. So this result reaches KMT2A::AFF1 only by import across fusion, species and lineage at once. Verified by the coordinator directly against the GEO SOFT records on 2026-09-19 and the finding should not be assumed to transfer to KMT2A::AFF1, which contains no AF10. A screen identifies requirement, not mechanism.

    Evidence that argues against this (2)
    • A menin requirement recovered on a reporter is compatible with menin acting upstream of fusion targeting generally rather than specifically on DOT1L-directed methylation, since displacing the fusion from chromatin would remove the mark as a side-effect. The paper's own framing - that H3K79me2 has a 'putative' oncogenic function - is weaker than the use this graph would like to make of it. Neville D et al. 2026
    • The fusion behind this result is murine MLL-AF9, not KMT2A::AFF1 — established from the study's own GEO depositions, since the full text is unreachable. Every H3K79me2 sample in GSE192555 is Mus musculus, cell type MLLAF9, and no sample in any of the three series mentions an AFF-family gene. The menin-dependence of the oncogenic unit, which is the part this graph leans on hardest, was measured in mouse myeloid MLL-AF9 cells. KMT2A::AFF1 engages DOT1L through a different architecture — AFF1 carries neither the AHD that MLLT3 and MLLT1 use nor the AF10 OM-LZ — so the partition may not transfer, and `hyp-index-fusion-h3k79-is-entirely-endogenous` is the sharper alternative. NCBI Gene Expression Omnibus 2022
  6. E6established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Pinometostat reduced H3K79 methylation in patients and produced 2 complete remissions in 51 adults with acute leukaemia, delivered by continuous intravenous infusion.

    PMID 29724899DOI
    What this evidence cannot show

    Abstract read from the eutils record; full text not read. A phase 1 dose-escalation in adults whose maximum tolerated dose was not reached, so the exposure-response relationship was not characterised over its full range; 51 patients across genotypes gives very little power for any single fusion partner. Cited here for the pharmacodynamic and response facts only.

    Evidence that argues against this (1)
    • Both complete remissions occurred in t(11;19)/KMT2A::MLLT1, so the trial is not uninformative about which biology responds - it is a two-patient signal in the partner whose AHD binds DOT1L most directly. Reading the trial as a flat negative discards the only genotype-linked observation in it. Daigle SR et al. 2013
Connected to