Oncofusion

start / graph / hyp-q140-h3k4me3-survives-menin-eviction-at-fusion-targets-because-a-second-complex-holds-it

Hypothesisspeculative confidence 1 of 5: Our own inference or analogy. Not shown to be true.Q140-sem-menin-arm

Hypothesis: the three quarters of H3K4me3 that survive menin displacement at fusion target sites are held by a menin-independent complex, and identifying it names the combination partner for a menin inhibitor

When the drug pries menin off, most of the chemical mark at the leukaemia genes stays put. Something else must be holding it there. Finding out what would point at a second drug to give alongside the first.

The same thing, for a scientist

**Statement.** At menin-bound sites in SEM, VTP-50469 removes ~23% of the H3K4me3 share (CI 13-33%) and leaves ~77%. If H3K4me3 at these sites were deposited solely by the menin-tethered KMT2A::AFF1 complex, menin displacement should remove far more of it. The residual is therefore maintained by a **menin-independent** activity - candidates in rough order of prior plausibility: (i) wild-type KMT2A and SETD1A/B acting at the same CpG islands, which would make the residual biologically inert; (ii) the fusion itself retained by a second chromatin anchor (LEDGF/PSIP1-PWWP via the KMT2A CXXC-PSIP1 module, or the AFF1 side through the super elongation complex), which would make the residual fusion-driven and therefore a target; (iii) mark persistence - H3K4me3 simply not yet turned over at 72 h, in which case the residual is a kinetic artefact. **These three have different combination-therapy consequences and the experiments below separate them.**

Go deeper
Why this is worth stating as a hypothesis rather than a caveat

The field's working model treats menin inhibition as removing the fusion from chromatin, and reads the leukaemia's death as the consequence. The size of the residual here is the measurable gap between that model and the data in the index fusion's own cell type. Whichever of (i), (ii) or (iii) is right, the answer changes practice: (i) says stop using H3K4me3 as a pharmacodynamic marker; (ii) names a combination partner; (iii) says dose duration matters more than dose.

Prior probability and why it is not higher

0.3 for option (ii), the only one that is a genuine discovery. Options (i) and (iii) are mundane and between them probably account for most of the residual - which is exactly why the experiment has to be able to distinguish them rather than just measuring the residual again. This project has repeatedly been burned by treating an unexplained remainder as evidence for the interesting explanation.

Facts and figures
statementThe ~77% of menin-bound-site H3K4me3 that survives VTP-50469 in SEM is maintained by a menin-independent activity; identifying it distinguishes an inert residual from a combination-therapy target.
origindata-pattern
prior probability0.3
predictions
[
  "If option (ii): a PSIP1/LEDGF degron or an AFF1-side perturbation added to VTP-50469 removes a further substantial fraction of H3K4me3 at menin-bound sites, while neither alone does.",
  "If option (i): the surviving H3K4me3 at menin-bound sites is indistinguishable in amount and shape from H3K4me3 at matched menin-free CpG-island promoters, and a SETD1A degron removes both equally.",
  "If option (iii): the residual shrinks monotonically with exposure time, approaching the fusion-dependent fraction by 7 days."
]
experiments
[
  "exp-q140-spike-in-normalised-h3k4me3-and-menin-cut-and-tag-in-sem-under-vtp-50469",
  "exp-q140-a-time-course-in-sem-that-separates-a-lagging-mark-from-a-partial-one"
]
analyses
[
  "ana-q140-vtp-50469-depletes-h3k4me3-at-menin-bound-sites-by-23-percent-in-sem-and-epz-5676-does-not"
]
falsifiers
[
  "If a spike-in-normalised experiment in SEM shows H3K4me3 at menin-bound sites falling by more than 60% at 72 h under 30 uM VTP-50469, the residual this hypothesis is about does not exist and the hypothesis is dead.",
  "If the surviving H3K4me3 at menin-bound sites is quantitatively matched, in amount and in domain width, by H3K4me3 at menin-free CpG-island promoters in the same cells, option (i) is sufficient and no second complex need be invoked.",
  "If the VTP-50469 arm of GSE159608 is shown to be a low-quality library pair - for example by a spike-in re-analysis or by the depositors' own QC - the 23% itself is unreliable and this hypothesis has no observation behind it."
]
How this could bite us back
  • If option (ii) is right, the named combination partner is likely to be an essential chromatin factor, and the therapeutic window for hitting two anchors at once may be narrower than for either alone.
  • If option (i) is right, this hypothesis will have spent effort proving that normal chromatin is normal - and the cost is the opportunity, not the experiment.
  • A combination inferred from a 23% chromatin effect, with no transcriptional or survival readout tying it to the drug's action, risks designing around a mark that is not on the causal path at all.
Still open
  1. Is there any published measurement of what fraction of H3K4me3 at KMT2A-fusion target promoters is fusion-deposited, in any system? Without it, '77% survives' has no baseline.
  2. Does PSIP1/LEDGF loss reduce H3K4me3 at fusion targets in a human KMT2A::AFF1 line? The CXXC-PSIP1 module is the best-characterised menin-independent anchor and the experiment appears not to have been done in this line.
Every source behind this entry (3)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    VTP-50469 removes 23% of the menin-bound H3K4me3 share in SEM (95% CI 13-33%), leaving roughly three quarters in place.

    sem-arm-identity-from-signal: run3.py and run4_bootstrap.py (this repository) · Q140-sem-menin-arm · Oncofusion knowledge graph 2026
    What this evidence cannot show

    Two libraries, one time point. The residual is a difference from an unstated expectation - nobody has measured what fraction of H3K4me3 at these sites the fusion deposits, so '77% survives' is only surprising against an assumption.

    Evidence that argues against this (1)
    • The most economical explanation is option (i): these are CpG-island promoters where SETD1A/B deposit H3K4me3 regardless of the fusion, and the residual is simply normal chromatin. That explanation requires no new biology and should be excluded before any combination-therapy inference. Krivtsov AV et al. 2019
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Menin displacement by VTP-50469 reduces MLL chromatin occupancy at select genes rather than globally, establishing that occupancy loss is itself partial and site-restricted.

    A Menin-MLL Inhibitor Induces Specific Chromatin Changes and Eradicates Disease in Models of MLL-Rearranged Leukemia · Krivtsov AV, Evans K, Gadrey JY, Eschle BK, Hatton C, Uckelmann HJ, et al., Armstrong SA · Cancer Cell 36:660-673.e11 2019
    PMID 31821784DOI
    What this evidence cannot show

    If occupancy loss is itself only partial at these sites, part of the H3K4me3 residual is explained without invoking a second complex - which is why the experiment below measures menin, fusion occupancy and the mark in the same cells.

    Evidence that argues against this (2)
    • Site-restricted occupancy loss is equally consistent with the simpler hypothesis this entry is meant to be tested against: that menin is only removed where it was bound, with no second complex required to explain anything. 'Partial and site-restricted' describes the observation, not a mechanism, and it does not discriminate between the two accounts.
    • The hypothesis requires a second complex holding H3K4me3; the deposit contains no candidate measurement - no SETD1A/B, no WDR5, no CFP1 library in SEM - so the entity doing the holding is unobserved in this system and is inferred entirely from the residual.
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In a patient sample that lineage-switched during revumenib treatment, the oncoprotein and menin fell to undetectable levels while ENL persisted at numerous oncoprotein target loci.

    MLL oncoprotein levels influence leukemia lineage identities · Janssens DH, Duran M, Otto DJ, Wu W, Xu Y, Kirkey D, Mullighan CG, Yi JS, Meshinchi S, Sarthy JF, Ahmad K, Henikoff S · Nature Communications 15:9341 2024
    PMID 39472576DOI
    What this evidence cannot show

    This is the closest existing evidence for option (ii) - a cofactor remaining at fusion targets after menin and the oncoprotein have gone - but it is one patient sample, a different drug, and ENL persistence is not the same as H3K4me3 persistence.

    Evidence that argues against this (2)
    • ENL persisting while the oncoprotein and menin fall is n=1 and comes from a sample selected for having lineage-switched, so the persistence may be a property of the switched state rather than of the drug response. It is suggestive of a fusion-independent cofactor but cannot support the quantitative claim this hypothesis makes.
    • ENL is not a H3K4 methyltransferase and its persistence does not explain a surviving H3K4me3 mark. Using it as support requires an unstated intermediate - that ENL retains or recruits a SET-domain activity - which is not measured here or cited.
Connected to