Oncofusion

start / graph / hyp-index-fusion-h3k79-is-entirely-endogenous

Hypothesisspeculative confidence 1 of 5: Our own inference or analogy. Not shown to be true.Q-partner-provenance

In KMT2A::AFF1 there is no oncogenic DOT1L unit: all H3K79 methylation is the endogenous MLLT10-directed activity, working on genes the fusion has opened

also called no fusion-directed DOT1L in MLL-AF4 · one-unit hypothesis for the index fusion

If the two-machine picture of the H3K79 tag is right, and if the infant fusion has no hook for the tag-writing enzyme, then the tag on its target genes must be put there by the cell's normal machinery. That predicts a menin drug should hardly change the tag at all in infant leukaemia - the opposite of what the project currently expects.

The same thing, for a scientist

**Statement.** The CRISPR-ChIP partition defines an oncogenic DOT1L unit whose catalysis is *directed by the MLL fusion protein* and an endogenous unit directed by MLLT10. In the system where it was measured - murine MLL-AF9 - the fusion carries an AHD and 'directs' has a molecular referent. **KMT2A::AFF1 carries neither an AHD nor an OM-LZ**, and the one AHD contact it has excludes DOT1L in reconstitution. The hypothesis is therefore that **in KMT2A::AFF1 the oncogenic unit does not exist**: H3K79me2/3 at index-fusion targets is deposited by the endogenous, MLLT10-directed, transcription-coupled DOT1L complex acting on loci the fusion has activated by other means (AEP/P-TEFb and menin-directed targeting). **This is not a claim that the mark is unimportant there.** It is well established in the index fusion and index disease [E4][E5]. It is a claim about *who puts it there*, and the claim has an uncomfortable consequence for this project's central bet: if there is no fusion-directed unit, **menin inhibition should not appreciably reduce H3K79me2 in t(4;11) cells**, because the mechanism by which it does so in MLL-AF9 - displacing an AHD-bearing fusion that carries DOT1L - has no counterpart. Prior probability 0.35: the interface argument is strong but the reconstitution negative is a tube experiment, and a route through endogenous AF9/ENL remains open.

Go deeper
Predictions, sharpest first
  1. Menin inhibition at full target engagement reduces H3K79me2 at fusion targets substantially in a KMT2A::MLLT3 line and minimally in a KMT2A::AFF1 line. This is one experiment with two cell lines and a calibrated ChIP; it is the whole hypothesis.
  2. MLLT10 knockout reduces H3K79me2 at KMT2A::AFF1 target genes by as much as it reduces it genome-wide - no preferential sparing of fusion targets, because there is no separate fusion-directed pool.
  3. DOT1L does not co-immunoprecipitate with KMT2A::AFF1 under conditions where it co-immunoprecipitates with MLL-AF9.
  4. Menin inhibition and DOT1L inhibition are additive rather than synergistic in t(4;11) cells, while being synergistic in MLL-AF9 cells. The synergy the CRISPR-ChIP authors set out to explain is a property of the fusions that carry a DOT1L interface.
  5. H3K79me2 gain at index-fusion targets is quantitatively proportional to nascent transcription gain, as a transcription-coupled mark should be.
Falsifiers, and what each would mean
  • Menin inhibition removes H3K79me2 at fusion targets in a KMT2A::AFF1 line as efficiently as in MLL-AF9. Kills the hypothesis. Would mean either that the index fusion reaches DOT1L by an unmapped route, or that menin inhibition acts on H3K79 upstream of any fusion-DOT1L contact - and the second reading would also weaken the CRISPR-ChIP partition's own interpretation.
  • DOT1L co-immunoprecipitates with KMT2A::AFF1. Kills it, and identifies the missing interface as the next thing to map.
  • MLLT10 knockout spares H3K79me2 at index-fusion targets while depleting it elsewhere. Kills it: a fusion-directed pool would then exist in KMT2A::AFF1 without an identified director.
  • Menin and DOT1L inhibitors are synergistic in t(4;11) cells to the same degree as in MLL-AF9. Strongly disfavours it.

All four are measurable in existing human cell lines with existing reagents.

What it would cost the menin-interface hypothesis if true

Less than it first appears, and in a specific way worth stating.

Menin inhibitors would still work in t(4;11) disease - the clinical activity is not in question and does not depend on this. What would fail is the chromatin narrative: the account in which menin inhibition works by displacing the fusion, thereby removing fusion-directed H3K79 methylation, thereby collapsing the oncogenic programme. If there is no fusion-directed H3K79 unit in the index fusion, the middle of that chain is empty, and the drug's effect must run through something else - most plausibly direct loss of fusion occupancy and the AEP/P-TEFb output, with H3K79 change as a slow downstream consequence of reduced transcription rather than a cause.

That matters practically because it changes what a pharmacodynamic assay should measure: fusion occupancy or nascent transcription, not H3K79me2.

Facts and figures
statementIn KMT2A::AFF1 there is no fusion-directed oncogenic DOT1L unit; H3K79 methylation at index-fusion targets is endogenous, MLLT10-directed and transcription-coupled, acting on loci the fusion has opened through AEP/P-TEFb and menin-directed targeting.
originliterature-gap
prior probability0.35
predictions
[
  "Menin inhibition at full target engagement reduces H3K79me2 at fusion targets substantially in KMT2A::MLLT3 and minimally in KMT2A::AFF1.",
  "MLLT10 knockout reduces H3K79me2 at index-fusion target genes as much as genome-wide, with no sparing of fusion targets.",
  "DOT1L does not co-immunoprecipitate with KMT2A::AFF1 under conditions where it co-immunoprecipitates with MLL-AF9.",
  "Menin and DOT1L inhibitors are additive rather than synergistic in t(4;11) cells while remaining synergistic in MLL-AF9 cells.",
  "H3K79me2 gain at index-fusion targets is proportional to nascent transcription gain."
]
experiments
[
  "exp-crispr-chip-h3k79-partition-in-human-kmt2a-aff1",
  "exp-ahd-occupancy-competition-in-t411-cells"
]
analyses
[
  "ana-crispr-chip-fusion-identity-resolved",
  "ana-index-fusion-h3k79-evidence-that-exists"
]
falsifiers
[
  "Menin inhibition removes H3K79me2 at fusion targets in a KMT2A::AFF1 line as efficiently as in MLL-AF9.",
  "DOT1L co-immunoprecipitates with KMT2A::AFF1 from a human t(4;11) line.",
  "MLLT10 knockout spares H3K79me2 at index-fusion targets while depleting it elsewhere.",
  "Menin and DOT1L inhibitors are synergistic in t(4;11) cells to the same degree as in MLL-AF9."
]
How this could bite us back
  • If H3K79 methylation in t(4;11) disease is entirely endogenous, DOT1L inhibition there removes an activity normal lymphopoiesis also uses, so the therapeutic index is narrower than a fusion-directed model implies and the modest clinical performance of pinometostat becomes expected rather than puzzling.
  • A pharmacodynamic endpoint built on total or even locus-specific H3K79me2 would report a correctly engaged menin inhibitor as a failure in infant disease.
  • If the hypothesis is wrong because an unmapped AFF1-DOT1L motif exists, that motif is a fusion-specific surface and therefore a better drug target than anything currently in this graph - so the hypothesis is worth testing whichever way it resolves.
Still open
  1. Has any published experiment measured H3K79me2 by calibrated ChIP after menin inhibition in a human KMT2A::AFF1 line? This beat found none, and it is a one-week experiment.
  2. If the index fusion has no DOT1L interface, why is H3K79me2 enriched specifically where MLL-AF4 spreads rather than uniformly across all transcribed genes? A transcription-coupled model must account for the spatial specificity Kerry et al. report.
  3. Does AFF4, rather than the fusion, occupy the AHD in t(4;11) cells? If so the competition argument applies to the wrong molecule.
Every source behind this entry (5)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.preprint

    CRISPR-ChIP partitions H3K79 methylation into an oncogenic DOT1L complex in which the malignant driver directs DOT1L catalysis at MLL-fusion target genes and a separate endogenous DOT1L complex directed by MLLT10, and offers this as the explanation for menin/DOT1L inhibitor synergy and for menin inhibition controlling H3K79 methylation at a subset of genes.

    CRISPR-ChIP delineates a Menin-dependent oncogenic DOT1L complex in MLL-leukaemia · Gilan O, Bell C, Talarmain L, Neville D, Knezevic K, Ferguson D, Boudes M, Chan Y-c, Davidovich C, Lam E, Dawson MA · bioRxiv 2023
    DOI
    What this evidence cannot show

    The preprint abstract was read via the Europe PMC record on 2026-09-19; the preprint full text and PDF were unreachable (HTTP 429, Cloudflare error 1015, four attempts, two clients) and the published version is paywalled. The preprint abstract is quoted here rather than the published one because it states the synergy rationale explicitly. Preprint, not peer reviewed at the version read; the published version exists as PMID 37679565.

    Evidence that argues against this (1)
    • The deposited data show every H3K79 experiment in this study is murine MLL-AF9, so the partition's own scope is one fusion; extending 'the malignant driver directs DOT1L' to fusions with no DOT1L interface is the authors' generic phrasing rather than their result. Gilan O et al. 2023
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    KMT2A::AFF1 carries neither the AHD (Pfam PF17793, present on MLLT3 and MLLT1) nor an AF10-type DOT1L-binding leucine-zipper region (UniProt P55197 703-784 and 750-778); AFF1's Pfam complement is PF05110, PF18875 and PF18876.

    What this evidence cannot show

    Accessions fetched directly on 2026-09-19 with primary gene names and lengths read back rather than resolved by a gene_exact query. Pfam absence is absence of a detected match; AFF1 is annotated as overwhelmingly disordered, where profile HMMs under-detect short linear motifs, so this negative is weaker than the same negative for a globular protein.

    Evidence that argues against this (1)
    • An unmapped short linear motif in AFF1's disordered bulk would rescue a fusion-directed unit without any of the known interfaces, and disorder is exactly where such a motif would hide. Leach BI et al. 2013
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In reconstitution, DOT1 forms a stable complex with AF9 and does not interact with AF9-AF4-P-TEFb complexes.

    Function of leukemogenic mixed lineage leukemia 1 (MLL) fusion proteins through distinct partner protein complexes. · Biswas D, Milne TA, Basrur V, Kim J, Elenitoba-Johnson KS, Allis CD, Roeder RG · Proc Natl Acad Sci U S A 2011
    PMID 21896721DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read. Purified minimal complexes 'in the absence of other factors', a sufficiency test that can miss interactions stabilised by chromatin or bridging factors in cells. The functional work in this paper is MLL-AF9.

    Evidence that argues against this (1)
    • The same study describes these complexes as dynamic and interconverting, so a steady-state negative does not exclude transient DOT1L visits to an AF4-occupied AHD - which for a distributive methyltransferase would suffice. Kuntimaddi A et al. 2015
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    H3K79 methylation profiles define murine and human MLL-AF4 leukaemias, and DOT1L suppression inhibits expression of critical MLL-AF4 target genes.

    H3K79 methylation profiles define murine and human MLL-AF4 leukemias. · Krivtsov AV, Feng Z, Lemieux ME, Faber J, Vempati S, Sinha AU, Xia X, Jesneck J, Bracken AP, Silverman LB, Kutok JL, Kung AL, Armstrong SA · Cancer Cell 2008
    PMID 18977325DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read. ChIP-chip, uncalibrated; the human arm is partner-mixed MLL-rearranged ALL; RNAi-era knockdown. Cited here to establish that the mark matters in the index fusion, which this hypothesis does not dispute.

    Evidence that argues against this (1)
    • If the mark is functionally required in KMT2A::AFF1 while being deposited by an endogenous complex, then this hypothesis predicts DOT1L inhibition should be no less effective there - which is hard to reconcile with pinometostat's modest single-agent clinical activity unless the limiting factor was exposure rather than biology. Kerry J et al. 2017
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    MLL-AF4 spreading into gene bodies co-occurs with menin spreading and increased H3K79me2/3, and spreading-gene expression is preferentially DOT1L-inhibitor sensitive.

    MLL-AF4 Spreading Identifies Binding Sites that Are Distinct from Super-Enhancers and that Govern Sensitivity to DOT1L Inhibition in Leukemia. · Kerry J, Godfrey L, Repapi E, Tapia M, Blackledge NP, Ma H, Ballabio E, O'Byrne S, Ponthan F, Heidenreich O, Roy A, Roberts I, Konopleva M, Klose RJ, Geng H, Milne TA · Cell Rep 2017
    PMID 28076791DOI
    What this evidence cannot show

    Abstract fetched from PubMed and read on 2026-09-19; full text not read. Human MLL-AF4 ChIP-seq with lines unnamed in the abstract. This is occupancy correlation; it is compatible with either the fusion-directed or the transcription-coupled model and is cited here as the observation both models must explain.

    Evidence that argues against this (1)
    • Menin spreading tracking H3K79me2/3 is the single observation most supportive of a menin-linked, fusion-directed pool in the index fusion, and it is the strongest argument against this hypothesis. Gilan O et al. 2023
Connected to