Oncofusion

start / graph / localization-menin-inhibition-fusion-fate

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q4b-localization

Where does the fusion go when menin inhibition evicts it? Almost nobody has looked - the published readouts are chromatin occupancy, the one protein whose fate was measured properly is menin, and the single human observation of fusion loss is confounded by a lineage switch

also called menin inhibitor eviction · fusion protein fate · revumenib chromatin eviction · VTP-50469 · MI-induced menin degradation

Menin inhibitors work by knocking the fusion protein off the DNA. Everyone measures that it has left the DNA; nobody has taken a picture to see where it went - whether it floats free in the nucleus, piles up somewhere else, or gets destroyed. In one patient treated with the drug the fusion became undetectable, which hints that it is destroyed, but that patient's leukaemia was changing type at the same time, so it does not settle it.

The same thing, for a scientist

A dedicated search of the menin-inhibitor literature found **no study that images the fate of the KMT2A fusion protein after menin inhibition**. **Narrowed 2026-09-19:** the previous wording, 'no study that images or measures the fate', is withdrawn. One human measurement exists and was already cited elsewhere in this graph - Janssens et al., profiling oncoprotein target sites across 36 KMT2A-rearranged samples, report that in a sample that lineage-switched during revumenib treatment the oncoprotein and menin were both reduced to **undetectable** levels while ENL persisted at many target loci [E6]. That is a measurement of the fusion's fate under menin inhibition, in a patient, and it points at option 4 below (degradation or loss of expression) rather than option 2 (a soluble pool waiting to rebind). It is one sample, it is confounded by the lineage switch (which is itself associated with reduced oncoprotein levels independent of drug), and it is an occupancy assay rather than a half-life or imaging assay - so the question stays open, but it is no longer unaddressed and the entry's own title overstated it. Every pivotal paper reads out chromatin occupancy, complex composition, transcription, differentiation and survival. Krivtsov et al. showed by co-immunoprecipitation that VTP-50469 displaces menin from protein complexes and by ChIP that MLL occupancy changes at select genes, with differentiation, apoptosis and PDX survival endpoints [E1]. Klossowski et al. used proliferation, differentiation, MEIS1 as a pharmacodynamic biomarker, primary patient samples and in-vivo remission [E2]. Perner et al. identified somatic MEN1 mutations at the revumenib-menin interface in patients on trial and showed these mutants prevent inhibitor-induced **eviction of menin and MLL1 from chromatin** — again an occupancy readout [E3]. The one place a *fusion protein* has been shown to be evicted is NUP98-rearranged AML, where VTP-50469 evicts MLL1 and the NUP98 fusion protein itself from chromatin at a critical gene set, still by ChIP [E4]. And the only protein-fate measurement anywhere in this literature is about menin: menin-MLL inhibitors reduce **menin** protein levels and shorten **menin** half-life with no change in mRNA, rescued by proteasome or E1 inhibitors, through inhibitor-promoted menin-CHIP (Hsp70-associated ubiquitin ligase) interaction and increased menin ubiquitination [E5]. This is a clean, cheap, entirely open question that connects directly to the graph's existing question about where blasts go when the differentiation block is released — because a fusion that is evicted but intact, and a fusion that is evicted and degraded, predict different relapse kinetics, different rebound behaviour on drug withdrawal, and different combination strategies.

Go deeper
Eviction from chromatin is not a statement about location1 deeper

'Evicted from chromatin' is a ChIP or CUT&RUN result: less fusion signal at a set of loci. A protein can leave chromatin and do at least five different things, and the literature does not distinguish them.

  1. Redistribute to lower-affinity chromatin sites genome-wide, producing a flatter occupancy profile rather than an absent one — which the graph's hyp-apparent-authority-residual-fusion-occupancy entry treats as a live possibility.
  2. Enter the soluble nucleoplasmic pool and remain there, available to rebind the instant drug is withdrawn.
  3. Partition into a different compartment — the nucleolus being the obvious candidate given the reported nucleolar affinity of MLL-AF4, and one where, on the EWS::FLI1 precedent, sequestration would be inactivating.
  4. Be degraded. Menin is 5. Whether the fusion is has never been tested with a half-life or proteasome-rescue assay, and the CHIP/Hsp70 machinery that degrades menin is generic enough that a destabilised, partnerless fusion is a plausible substrate. Updated 2026-09-19: the one human observation is consistent with this option - oncoprotein undetectable in a revumenib-treated sample 6 - but with a lineage switch in the same sample it cannot separate degradation from loss of expression, and it carries no proteasome-rescue arm. This option is now the leading one, not an untouched one.
  5. Be exported or aggregate. The APL precedent has arsenic moving PML::RARA to cytoplasmic assemblies, so cytoplasmic relocalisation is not an exotic hypothesis for a fusion oncoprotein under targeted therapy.

These are not equivalent. Option 2 predicts rapid, complete rebound on drug holiday and argues against intermittent dosing; option 4 predicts hysteresis and makes pulsed dosing rational, which is exactly the question hyp-agency-revocation-hysteresis-pulsed-menin and exp-menin-pulse-chase-hysteresis already pose in this graph without knowing the answer.

The menin degradation result and what it does not license

Wu et al. showed menin-MLL inhibitors reduce menin protein levels, shorten menin half-life measured by cycloheximide chase, do not change menin mRNA, and act through inhibitor-promoted menin-CHIP interaction and increased menin ubiquitination; proteasome or E1 inhibition rescues both the degradation and the downstream reduction in H3K4me3 and target-gene expression 5.

This is important and under-cited: the drug class is partly a menin degrader, not purely an interface blocker. But it says nothing about the fusion. Reasoning 'menin is degraded, therefore the complex is degraded, therefore the fusion is degraded' is exactly the kind of inference chain this project has been burned by. The fusion's half-life under menin inhibition is an unmeasured quantity, and measuring it takes one cycloheximide chase with a fusion-specific antibody or tag.

Why the answer changes clinical reasoning, not just mechanism

Combination logic. If the fusion is evicted but intact and abundant in the soluble pool, a degrader added to a menin inhibitor is strongly synergistic in principle: the inhibitor makes the fusion available, the degrader removes it. If the inhibitor already destabilises the fusion, that combination is redundant.

Dosing. Continuous dosing is the current standard. A fusion that is destroyed on eviction supports intermittent schedules and reduces the cumulative differentiation-syndrome and QTc burden in infants; a fusion that simply floats free argues against holidays.

Resistance. MEN1 interface mutations sustain occupancy 3. But if part of the drug's effect runs through menin degradation 5, then resistance could also arise through the degradation machinery — CHIP/STUB1 or HSP70 alterations — with a wild-type MEN1 sequence. Nobody is sequencing STUB1 in menin-inhibitor relapses, and this entry predicts they should.

Biomarker. If the fusion is degraded, fusion protein level itself becomes a pharmacodynamic biomarker measurable in blasts, complementing MEIS1 transcript. That is testable in samples already banked from AUGMENT-101.

Facts and figures
questionsubcellular and molecular fate of the KMT2A fusion protein after menin inhibition
only protein fate measuredmenin itself — reduced level and shortened half-life via CHIP/Hsp70 ubiquitination (PMID 31497350)
only fusion eviction shownNUP98 fusion protein, by ChIP (PMID 34582559)
confidence note 2026 09 19Lowered from strong to moderate on 2026-09-19: the entry's central assertion is an absence, and that absence is now partial - one human measurement of the fusion's fate under revumenib exists (PMID 39472576), confounded but real.
published readouts available
[
  "co-immunoprecipitation complex displacement",
  "ChIP / CUT&RUN chromatin occupancy",
  "gene expression",
  "differentiation markers",
  "xenograft survival",
  "oncoprotein occupancy/level in a revumenib-treated patient sample (PMID 39472576, n=1, lineage-switch confounded)"
]
published readouts absent
[
  "live or fixed imaging of fusion localisation",
  "fusion protein half-life",
  "subnuclear fractionation of the fusion",
  "nucleocytoplasmic partition of the fusion",
  "proteasome-rescue test of fusion degradation"
]
candidate fates
[
  "redistribute to lower-affinity chromatin",
  "soluble nucleoplasmic pool",
  "nucleolar or other compartment sequestration",
  "proteasomal degradation",
  "cytoplasmic relocalisation"
]
search terms returning zero
[
  "menin inhibitor AND MLL fusion protein half-life",
  "revumenib AND MLL fusion protein imaging",
  "HaloTag MLL-AF9 live imaging leukemia",
  "MLL fusion protein nuclear puncta menin inhibitor relocalization"
]
narrowed
{
  "date": "2026-09-19",
  "was": "no study images or measures the fate of the fusion after menin inhibition",
  "now": "no study images it and none measures its half-life; one confounded human occupancy observation exists",
  "source": "39472576"
}
How this could bite us back
  • If the fusion is merely displaced into a soluble pool, drug holidays and missed doses in a paediatric outpatient setting permit immediate rebinding, and adherence becomes a pharmacodynamic variable rather than a compliance footnote.
  • If part of the drug's action is degradation through the CHIP/Hsp70 axis, then co-administered HSP90 or proteasome inhibitors — plausible in combination trials — could antagonise the menin inhibitor rather than add to it.
  • A fusion that redistributes to lower-affinity sites rather than leaving chromatin would produce a partial transcriptional response that looks like inadequate dosing, prompting dose escalation into toxicity when the real problem is a redistribution the assay cannot see.
Still open
  1. Is the KMT2A::AFF1 protein degraded, stabilised or unchanged in abundance after menin inhibition? One cycloheximide chase with a fusion-specific tag answers a question the entire field has assumed away.
  2. Where is the evicted fusion? Live imaging of an endogenously tagged fusion through a menin-inhibitor time course, with nucleolar, speckle and nucleoplasmic segmentation, distinguishes five mutually exclusive fates that currently all sit behind the single word 'evicted'.
  3. If part of the menin inhibitor's effect is CHIP/Hsp70-mediated menin degradation, can resistance arise through STUB1 or HSPA alterations with a wild-type MEN1 sequence? Nobody is sequencing those genes in menin-inhibitor relapse samples.
  4. Does the fusion rebind chromatin faster than it is resynthesised after drug washout? That ratio, not the on-rate alone, determines whether intermittent dosing is viable and connects directly to exp-menin-pulse-chase-hysteresis.
Every source behind this entry (6)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The menin-MLL inhibitor VTP-50469 displaces menin from protein complexes by co-immunoprecipitation and alters MLL chromatin occupancy at target genes, inducing differentiation and apoptosis and eradicating disease in MLL-rearranged AML and ALL patient-derived xenografts; no measurement of fusion protein abundance, stability or localisation is reported.

    PMID 31821784DOI
    What this evidence cannot show

    Cited here partly for what it does not contain. Co-IP displacement is a solution-phase biochemical result and ChIP is a locus-occupancy result; neither is spatial. The absence of a fusion-fate measurement is an observation about the paper's design, not a criticism of its conclusions.

    Evidence that argues against this (1)
    • Searched and empty, recorded explicitly. Independent PubMed searches on 2026-09-18 (menin inhibitor + MLL/KMT2A fusion protein level / western blot / degradation / half-life; menin inhibition + fusion + eviction + imaging / single molecule / live-cell; dTAG or degron + MLL-AF9/MLL-AF4) found no study that images or quantifies the fusion protein after menin inhibition. This confirms rather than contradicts the entry's central negative claim; it is logged here so that a later reader knows the gap was looked for and not merely assumed. No citation is attached because there is nothing to cite: this note is a bounded absence recording searches that returned nothing, not a sourced fact. Uncited and unciteable look identical in this field, so it is said here explicitly (Q42 citation-integrity repair, 2026-09-20).
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The menin inhibitor MI-3454 induces differentiation, suppresses proliferation, reduces MEIS1 as a pharmacodynamic biomarker and induces remission or regression in MLL1-rearranged and NPM1-mutated leukemia models including patient-derived xenografts, while being tolerated by normal haematopoiesis; fusion protein fate and localisation were not measured.

    Menin inhibitor MI-3454 induces remission in MLL1-rearranged and NPM1-mutated models of leukemia · Klossowski S, Miao H, Kempinska K, et al. · Journal of Clinical Investigation 2020
    PMID 31855575DOI
    What this evidence cannot show

    Same design limitation as E1: transcriptional and phenotypic readouts throughout, no protein-fate or imaging endpoint.

    Evidence that argues against this (1)
    • MEIS1 is used here as the pharmacodynamic biomarker, and in patient material that biomarker has been shown to stay on while the drug stops working. Mahdavi et al. tested menin inhibition in patient-derived xenografts of KMT2A-rearranged leukaemia, including KMT2A::AFF1 ALL, and found that heavily pretreated serial samples were much less sensitive than earlier samples from the same patients - resistance acquired without any menin-inhibitor exposure - while transcriptomic analysis documented SUSTAINED on-target efficacy against the canonical menin-inhibitor target genes in those resistant cells. A preclinical package built on transcriptional and phenotypic endpoints therefore cannot distinguish a working drug from a drug whose target engagement has been uncoupled from cell fate, which is an additional reason - beyond the one this entry gives - that fusion-protein fate needs measuring directly. Mahdavi L et al. 2025
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Somatic MEN1 mutations at the revumenib-menin interface arise in patients on trial and mediate clinical resistance by preventing inhibitor-induced eviction of menin and MLL1 from chromatin, a chromatin-occupancy readout rather than a measurement of protein fate.

    MEN1 mutations mediate clinical resistance to menin inhibition · Perner F, Stein EM, Wenge DV, et al. · Nature 2023
    PMID 36922589DOI
    What this evidence cannot show

    Clinical sequencing plus xenografts, base-editor screening and structural analysis. 'Prevents eviction' is measured as sustained chromatin occupancy; the paper does not report where menin or MLL1 are in the nucleus in either condition.

    Evidence that argues against this (1)
    • Interface mutation is not the only route to clinical resistance, nor the first. Mahdavi et al. found that multiply relapsed KMT2A-rearranged ALL samples, including KMT2A::AFF1, had already become insensitive to menin inhibition BEFORE any exposure to a menin inhibitor, with multiple comutations emerging (RAS-pathway and TP53) that were individually insufficient to confer resistance in vitro, and with downregulation of a KMT2 family member implicated in one patient; KMT2C-edited cells from that patient were selected for under menin-inhibitor conditions. Their conclusion is to use these drugs earlier, before genomic and epigenomic evolution. So a chromatin-occupancy account of resistance built on MEN1 interface mutants describes the on-treatment mechanism in a population that has already been filtered by prior therapy. Mahdavi L et al. 2025
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In NUP98-rearranged AML, menin-MLL1 inhibition with VTP50469 evicts both MLL1 and the NUP98 fusion protein itself from chromatin at a critical set of genes required for the malignant phenotype, upregulates differentiation markers such as CD11b, downregulates Meis1 and the HOXA cluster, and prolongs survival in NUP98-NSD1 and NUP98-JARID1A patient-derived xenografts.

    The menin-MLL1 interaction is a molecular dependency in NUP98-rearranged AML · Heikamp EB, Henrich JA, Perner F, Wong EM, Hatton C, Wen Y, Barwe SP, Gopalakrishnapillai A, Xu H, Uckelmann HJ, Takao S, Kazansky Y, Pikman Y, McGeehan GM, Kolb EA, Kentsis A, Armstrong SA · Blood 2022
    PMID 34582559DOI
    What this evidence cannot show

    The only published demonstration that a fusion oncoprotein itself is evicted by a menin inhibitor, and it is in the NUP98 family rather than KMT2A. Still a ChIP-class occupancy measurement in mouse lines and PDX, so it locates the protein on the genome, not in the nucleus.

    Evidence that argues against this (1)
    • For NUP98 fusions specifically, 'evicted from chromatin' has a competing physical reading that changes what the observation means for the KMT2A question. Terlecki-Zaniewicz et al. showed that five NUP98 fusion proteins expressed in human leukaemia cells share interactors enriched for proteins involved in biomolecular condensation, that those interactors colocalise with the fusions in nuclear puncta, that NUP98 fusions alter the global composition of condensates as measured by biotinylated-isoxazole condensome mass spectrometry, and that an artificial FG-repeat-containing fusion phenocopies both the nuclear localisation pattern and the oncogenic gene-expression programme. If NUP98-fusion chromatin association is a property of a condensate rather than of a discrete binding event, then loss of occupancy on menin inhibition may record dissolution or redistribution of that compartment rather than release of a soluble protein - and the analogy to a KMT2A fusion, which has no FG repeats, weakens accordingly. Terlecki-Zaniewicz S et al. 2021
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Menin-MLL inhibitors reduce menin protein levels and shorten menin protein half-life in MLL-fusion-expressing leukemia cells without changing menin mRNA; proteasome or E1 ligase inhibition rescues the degradation, and mechanistically the inhibitors promote interaction of menin with the Hsp70-associated ubiquitin ligase CHIP, increasing menin ubiquitination, with proteasome inhibition also rescuing the drug-induced reduction in H3K4me3 and target-gene expression.

    Disruption of the menin-MLL interaction triggers menin protein degradation via ubiquitin-proteasome pathway · Wu Y, Doepner M, Hojnacki T, Feng Z, Katona BW, He X, Ma J, Cao Y, Busino L, Zhou F, Hua X · American Journal of Cancer Research 2019
    PMID 31497350
    What this evidence cannot show

    A lower-profile journal and a single-group result that has not been widely replicated; cell lines only, with western blot and cycloheximide chase. Crucially it measures the fate of **menin**, not of the MLL fusion protein, and must not be cited as evidence about the fusion.

    Evidence that argues against this (1)
    • Counter to extrapolating menin's fate onto the fusion's. The one protein-fate measurement in this literature is about menin; the only published measurement of KMT2A/MLL protein turnover says the opposite for the fusion. Liu, Cheng and Hsieh show MLL is degraded by SCF(Skp2) at S phase and APC(Cdc20) at late M, that degradation initiates within the N-terminal approximately 1400 amino acids that every fusion retains, and that the prevalent MLL fusions tested are resistant to that degradation - stability being, on their argument, part of the oncogenic mechanism. So the default expectation for a menin-inhibitor time course should be that fusion protein level does NOT fall, and a fall would be the surprising result. Read: abstract only (PMC full text empty on retrieval). Liu H et al. 2007
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Across 36 KMT2A-rearranged leukaemia samples profiled for oncoprotein target-site occupancy, oncoprotein genomic enrichment was highly variable and dynamically regulated; in one sample that lineage-switched during revumenib treatment, the oncoprotein and menin were reduced 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 2024
    PMID 39472576DOI
    What this evidence cannot show

    n=1 for the drug-treated observation, confounded by a concurrent lineage switch which independently associates with reduced oncoprotein levels; the readout is chromatin occupancy profiling, not imaging, fractionation, half-life or a proteasome-rescue experiment. It narrows the absence claim without answering the question.

    Evidence that argues against this (1)
    • An n=1 observation in a lineage-switching sample sits inside a population where the drug's effect is evidently heterogeneous, and the size of that heterogeneity is now quantified. In AUGMENT-101, 94 patients were treated with revumenib and, among the 57 efficacy-evaluable KMT2A-rearranged patients, CR+CRh was 22.8% and overall response 63.2%, with 15 of 22 assessed responders having no detectable residual disease. So roughly three-quarters of patients did not reach CR/CRh. Whatever happens to the fusion protein in a responding, switching sample need not be what happens in the majority who do not respond, and a single-sample undetectable reading cannot be generalised into 'the fusion is destroyed under menin inhibition' without the non-responders measured alongside. Issa GC et al. 2025
Connected to