Experiment: endogenously tagged KMT2A::AFF1 imaged through a menin-inhibitor time course, with a five-compartment segmentation and a half-life arm
also called HaloTag fusion junction knock-in · menin inhibitor imaging time course · fusion fate experiment
Put a fluorescent tag on the fusion protein at its own gene in a leukemia cell line, then film the cells while the menin inhibitor is applied and score where the protein goes and whether it disappears. Nobody has filmed it. One study has measured the amount of fusion protein left in a treated patient — it was undetectable — but that patient's leukaemia had also switched lineage, which lowers the protein independently, so what the drug did to it is still open.
The same thing, for a scientist
Knock a HaloTag (or a self-labelling tag plus HA epitope) into the KMT2A::AFF1 junction at the endogenous locus in SEM and RS4;11 by CRISPR HDR, so only the fusion is labelled and neither wild-type KMT2A nor wild-type AFF1 contributes signal — the limitation that made the 1998 endogenous imaging uninterpretable. Verify the tagged allele's expression is within 1.5-fold of untagged by western blot and that target-gene expression and revumenib sensitivity are unchanged. Then run three arms in parallel on the same clone. **Arm 1, localisation:** live confocal or lattice light-sheet across a 0-72 h revumenib time course, co-imaged with fibrillarin-mScarlet (nucleolus), SRSF2 (speckle), lamin B1 (periphery), RPB1 (Pol II clusters) and Hoechst-dense foci (chromocentres); score the fraction of total nuclear fusion signal in each compartment at each time point, plus total nuclear signal and nucleocytoplasmic ratio. **Arm 2, stability:** cycloheximide chase with and without revumenib, western blot for the 240-kDa species, with menin as an internal positive control since its degradation is published, and a proteasome-inhibitor rescue. **Arm 3, kinetics:** FRAP and single-molecule tracking before and during drug to measure chromatin-bound fraction and residence time, which converts 'evicted' from a population-average ChIP statement into a per-molecule one.
Design decisions that matter
Tag at the junction, not at either terminus. A junction knock-in labels only the fusion. An N-terminal tag labels the fusion and wild-type KMT2A; a C-terminal AFF1 tag labels the fusion and wild-type AFF1. This is the whole reason the existing endogenous dataset cannot be interpreted.
Two lines, not one. SEM and RS4;11 differ in breakpoint and in co-mutations; a compartment shift seen in one line only is a clone property.
Live, not fixed. Paraformaldehyde fixation can create or erase droplet-like puncta relative to the live cell, so a fixed-only design cannot distinguish a real redistribution from a fixation artefact. If fixed imaging is needed for patient material, include the glycine-modulated fixation control.
Total signal is a primary endpoint, not a normalisation. Most imaging pipelines normalise per cell and would discard exactly the measurement that distinguishes degradation from redistribution.
Add RPB1. The Pol II channel costs nothing and simultaneously addresses the transcription-factory question in the same dataset.
What each outcome would mean
| observation | interpretation | which graph entry it settles |
|---|---|---|
| total signal falls, no compartment gains | fusion degraded on eviction | hyp-evicted-fusion-is-degraded-not-displaced confirmed |
| total signal flat, nucleolar fraction rises | nucleolar sequestration as an off state | localization-nucleolus-kmt2a-fusion, localization-ewsr1-fli1-condensate-comparator |
| total signal flat, bound fraction falls, no compartment gains | soluble nucleoplasmic pool | pulsed dosing disfavoured; hyp-agency-revocation-hysteresis-pulsed-menin weakened |
| bound fraction falls only partially, residence time unchanged | a resistant subset of loci retains occupancy | hyp-apparent-authority-residual-fusion-occupancy supported |
| cytoplasmic signal appears | APL-like relocalisation | localization-pml-nuclear-bodies-pmlrara analogy strengthened |
Every outcome is informative, which is the mark of a well-posed experiment and the reason this should be prioritised over further compartment cataloguing.
| design | CRISPR HDR knock-in of HaloTag + HA at the KMT2A::AFF1 junction in SEM and RS4;11; three arms (live compartment imaging across a revumenib time course, cycloheximide-chase stability with proteasome rescue, FRAP/single-molecule tracking) on the same validated clone |
|---|---|
| readout | fraction of total nuclear fusion signal per compartment; total nuclear signal; nucleocytoplasmic ratio; fusion protein half-life; chromatin-bound fraction and residence time |
| controls | untagged parental line; tagged allele expression within 1.5-fold of untagged by western; unchanged target-gene expression and revumenib IC50; menin as internal positive control for drug-induced degradation; glycine-modulated fixation control for any fixed imaging; DMSO time-matched vehicle; a menin-inhibitor-resistant MEN1-mutant clone as a specificity control |
| cost tier | med |
| timeline | 9-12 months (4-5 months to validated knock-in clones, 3 months imaging, 2 months analysis) |
compartment markers
[ "fibrillarin (nucleolus)", "SRSF2/SC35 (speckles)", "lamin B1 (periphery)", "RPB1 (Pol II clusters)", "Hoechst-dense foci (chromocentres)" ]
settles entries
[ "hyp-evicted-fusion-is-degraded-not-displaced", "localization-menin-inhibition-fusion-fate", "localization-nucleolus-kmt2a-fusion", "localization-endogenous-mll-af4-punctate-negative", "localization-transcription-factories-pol2-clusters" ]
prior human observation
{
"pmid": "39472576",
"what": "oncoprotein and menin undetectable in one revumenib-treated, lineage-switched patient sample",
"implication": "the imaging arm should be powered to detect a decrease as well as no change; the earlier design assumed no change on the basis of PMID 17908926",
"recorded": "2026-09-19",
"title": "MLL oncoprotein levels influence leukemia lineage identities.",
"authors": "Janssens DH, Duran M, Otto DJ, et al.",
"venue": "Nat Commun",
"year": 2024,
"title_provenance": "Title/authors/venue backfilled from PubMed esummary on 2026-09-20 because this citation stored a bare PMID with no title, which made it unscoreable by verify-citations.py. BACKFILLING IS NOT VERIFICATION: it confirms the PMID resolves to a real paper, not that this is the right paper for the claim. That still requires reading."
}summary narrowed 2026 09 19
{
"was": " single experiment answers most of the open questions in this beat and nobody has done it.",
"reason": "The expert-level claim was narrowed this cycle from 'images or measures' to imaging and half-life, because PMID 39472576 measured fusion abundance in a revumenib-treated patient. The lay summary still said 'nobody has done it' and was left behind by that correction.",
"found_by": "scripts/check-stale-corrections.py, on its first run"
}- A tag at the junction is precisely where a junction-epitope therapeutic would bind; if the tag perturbs function, it also suggests that the junction epitope is functionally sensitive, which is informative but would complicate interpreting the parent line as wild type.
- Cell lines adapted to culture may have altered chaperone and proteasome capacity relative to primary blasts, so a negative stability result in SEM does not exclude degradation in a patient.
- Is a junction knock-in tolerated at the KMT2A::AFF1 junction without disrupting the fusion's function? A degron knock-in at the same locus has been done in this field, so the locus is editable, but the junction is the functional seam and an internal tag could disturb it — a small linker screen may be needed.
- Can the same tagged line be xenografted so the imaging endpoints can be repeated on cells recovered from marrow under in-vivo drug exposure, where the pharmacology is real?
- Would a split-fluorophore design (one half knocked into KMT2A, the other into AFF1) be a lower-risk alternative that labels only the fusion by complementation and avoids a bulky junction insert?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
No published study **images** the fate of a KMT2A fusion protein after menin inhibition, and none measures its abundance or half-life with a controlled biochemical assay; the pivotal papers read out chromatin occupancy, complex composition, transcription, differentiation and survival, and the closest result in this citation is chromatin eviction of the NUP98 fusion protein by ChIP. **Narrowed 2026-09-19:** the previous version of this claim said no study 'images or measures' the fusion's fate, which is too strong - see [E4], where oncoprotein occupancy was profiled in a patient sample during revumenib treatment and the oncoprotein was reduced to undetectable levels.
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 2022What this evidence cannot show
Used to document the state of the art the experiment is designed to exceed; the absence of imaging studies is established by search rather than by any single citation. Corrected 2026-09-19: the absence claim is now restricted to imaging and to controlled abundance/half-life measurement. A human occupancy measurement under menin inhibition exists (PMID 39472576, [E4]) and was already cited elsewhere in this graph.
Evidence that argues against this (1)
- Prior expectation for the proposed time course, from the only direct measurement of MLL protein turnover. Liu, Cheng and Hsieh show wild-type MLL is degraded in defined cell-cycle windows by SCF(Skp2) and APC(Cdc20), that degradation initiates within the N-terminal approximately 1400 amino acids retained in every fusion, and that prevalent MLL fusions tested are resistant to that degradation - stability being, on their reading, part of the oncogenic mechanism. The experiment should therefore be powered to detect NO change in total fusion signal, and should carry an internal positive control for the degradation assay working (menin itself, whose level does fall under these inhibitors). Read: abstract only. Liu H et al. 2007
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Paraformaldehyde fixation can both enhance and diminish apparent phase-separation behaviour and can create droplet-like puncta absent from the live cell, with glycine co-fixation reversing the direction of the artefact.
Fixation can change the appearance of phase separation in living cells · Irgen-Gioro S, Yoshida S, Walling V, Chong S · eLife 2022What this evidence cannot show
Demonstrated for a panel of disordered proteins in cultured cells; the magnitude is protein-specific, which is why the design specifies a live primary arm rather than relying on the control alone.
Evidence that argues against this (1)
- The fixation caveat should not be read as making the experiment infeasible. Cho et al. imaged Mediator and RNA Pol II in living mouse embryonic stem cells by superresolution and light-sheet microscopy and resolved small transient and large stable clusters, their chromatin association and their sensitivity to transcriptional inhibitors - i.e. the class of measurement this experiment needs, on endogenous machinery, in live cells, already exists and is tractable. The design should commit to live-cell imaging of an endogenously knocked-in tag rather than treating fixed imaging plus a glycine control as the fallback. Read: abstract only. Cho WK et al. 2018
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The 240-kDa MLL-AF4 fusion protein is resolvable by western blot from the 125 and 145 kDa wild-type AF4 species in t(4;11) cell lines, making a protein-stability arm feasible with standard reagents.
AF4 encodes a ubiquitous protein that in both native and MLL-AF4 fusion types localizes to subnuclear compartments · Li Q, Frestedt JL, Kersey JH · Blood 1998What this evidence cannot show
1998 polyclonal reagents; the tagged knock-in removes the dependence on antibody quality for the imaging arm but the stability arm still benefits from an epitope tag on the same allele.
Evidence that argues against this (1)
- A size-resolved western is only as transferable as the fusion's size, and the fusion does not have one size. Meyer et al. sequenced MLL genomic breakpoints in 2,345 acute leukaemia patients, catalogued 135 distinct rearrangements with 94 partner genes characterised at the molecular level, and showed the MLL-side breakpoint shifts systematically with age and lineage - intron 11 in ALL and younger patients, intron 9 in AML and older patients. The 240-kDa species was defined in particular t(4;11) lines; a knock-in clone with a different junction, or a patient sample, need not run there. For the stability arm this matters less than it sounds - the epitope tag is on the knock-in allele - but any attempt to extend the assay to primary material must type the junction first rather than assume the 1998 size. Meyer C et al. 2018
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Profiling oncoprotein target sites across 36 KMT2A-rearranged leukaemia samples found oncoprotein genomic enrichment 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 many 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 2024What this evidence cannot show
One patient sample under revumenib, read out by CUT&RUN-style occupancy rather than by imaging or by a half-life assay; the sample also lineage-switched, and lineage switching is independently associated with reduced oncoprotein levels, so drug effect and switch effect are confounded. It establishes that the fusion's fate under menin inhibition has been observed once in human material - not that it has been measured properly, which is what this experiment proposes.
Evidence that argues against this (1)
- The time course assumes the cells are drug-sensitive and that on-target pharmacology reports the biology, and both assumptions have now failed in patient material. Mahdavi et al. found that serial samples from two patients with multiply relapsed ALL - including KMT2A::AFF1 disease - were far less sensitive to menin inhibition than earlier samples from the same patients, with resistance acquired in the complete ABSENCE of menin-inhibitor exposure, and that transcriptomic analysis showed sustained on-target efficacy against the canonical menin-inhibitor target genes in the resistant cells. So a KMT2A::AFF1 line can show the predicted transcriptional response and the predicted fusion-protein fate while being functionally resistant. The design needs a resistant arm - an isogenic MEN1-mutant clone or a late-passage resistant derivative - or its imaging readout cannot be connected to whether the drug works. Mahdavi L et al. 2025
- related to →ENL stays after the fusion leaves — an epigenetic lesion that outlives the protein that made it
the one human measurement of fusion fate under revumenib (PMID 39472576) is the same sample that carries the persistent-ENL observation
the experiment designed to answer this entry, now powered for a decrease as well as no change
Now powered to detect a null: MLL fusions are reported resistant to the degradation pathway that reads their retained N-terminus, so the experiment needs menin itself as an internal positive control
imaging, stability and kinetics arms on one endogenously tagged clone
the fusion-specific label the 1998 experiment lacked
shares the endogenously engineered line and the washout logic
a cheaper, attributable alternative to the imaging time course, answering the same question without a knock-in
the imaging time course now needs a resistant arm, because a sensitive-looking transcriptional response no longer implies a working drug