Experiment: long-hold degron depletion of endogenous KMT2A::AFF1, to ask whether the leukaemic program recovers without the fusion
also called hit-and-run test · degron washout · dTAG hold experiment
Tag the fusion protein so it can be destroyed on command, destroy it, and then keep it destroyed for a month while watching whether the leukaemia programme comes back by itself. If it does, drugging the fusion will not work.
The same thing, for a scientist
Knock a dTAG (FKBP12-F36V) cassette into the endogenous KMT2A allele 5' of the breakpoint in a t(4;11) line (SEM or RS4;11) so that the degron tags the fusion protein and not wild-type AFF1. Deplete with dTAG-13 and, critically, **maintain** depletion for 28 days with confirmed protein loss at every timepoint, rather than reading out at 24-72 h as all published work does. The question is not whether the program falls (it does) but whether it climbs back while the fusion is still absent. Parallel arms compare degradation against menin-interface disruption at matched levels of chromatin displacement, which is the comparison that decides between displacement and degradation as the preferred modality for this project.
Arms and readouts1 deeper
Arms: (1) vehicle; (2) dTAG-13 continuous, 28 days; (3) dTAG-13 for 72 h then withdrawal, to measure re-establishment kinetics from a transient hit; (4) revumenib at a dose titrated to match arm 2 for loss of fusion occupancy at HOXA9; (5) degron in a KMT2A-germline line, as a specificity control.
Readouts, at days 0, 1, 3, 7, 14, 21, 28: RNA-seq with a HOXA9/MEIS1 program score; CUT&RUN for fusion occupancy, menin, H3K79me2 and H3K4me3 at HOXA and other core targets; EM-seq for the DNA-methylation signature that is the leading candidate 'memory' carrier; colony replating; and at day 28 a limiting-dilution xenograft transplant to test whether self-renewal capacity recovered.
Controls that matter
Degron leakiness and escape are the main failure modes: any cell that silences the degron or loses the tag will outgrow and masquerade as program recovery. Amplicon-sequence the tagged allele at day 28 in every recovering population before interpreting recovery as biology. A second control is the reciprocal: confirm the degron does not deplete wild-type KMT2A from the untranslocated allele, or the experiment measures KMT2A loss rather than fusion loss.
| design | CRISPR knock-in of FKBP12-F36V degron at the endogenous KMT2A locus 5' of the breakpoint in SEM and RS4;11; 5 arms; continuous 28-day depletion with interval multi-omic readout and terminal limiting-dilution xenograft |
|---|---|
| readout | HOXA9/MEIS1 program score by RNA-seq; fusion, menin, H3K79me2 and H3K4me3 occupancy by CUT&RUN; DNA methylation by EM-seq; colony replating; limiting-dilution transplant at day 28 |
| controls | KMT2A-germline line with the same degron; vehicle; degron-allele amplicon sequencing on any recovering population to exclude escape; verification that wild-type KMT2A is not depleted |
| cost tier | med |
| timeline | 9-12 months including line generation and xenograft readout |
- A negative result (program recovers) would not only fail the menin hypothesis, it would fail every fusion-directed modality including degraders, so the experiment should be run before further modality investment rather than after.
- Is 28 days long enough? A stable DNA-methylation memory could outlast it. A parallel arm carried to 60 days in a slower-growing PDX would be the insurance policy.
- Should the degron tag the fusion or the reciprocal? Tagging only the derivative-11 product leaves the reciprocal AFF1::KMT2A untouched, which is the right design for a clean interpretation but forgoes a test of the reciprocal's contribution.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article
Depleting an MLL fusion protein in t(4;11) cells with fusion-directed siRNA reduces clonogenicity and xenograft engraftment over days, establishing that the dependency is real in this context and motivating a tool with acute, reversible control.
What this evidence cannot show
siRNA, not degron; acts over days; does not address recovery. **Correction 2026-09-20: this claim previously read that depletion 'changes elongation at core target genes within minutes and regresses established leukaemia' and cited this paper for it.** PMID 16046533 makes no minute-scale elongation claim - its readouts are proliferation, clonogenicity, HOXA7/HOXA9/MEIS1, CD133 and SCID xenograft engraftment. The minute-scale elongation result belongs to PMID 35245435, which is now a separate evidence item below, and is in **MLL::AF9**, not KMT2A::AFF1.
Evidence that argues against this (4)
- Do not assume the degron works at the depth the design requires; measure it. Liu et al. showed that wild-type MLL is cleared by SCF(Skp2) at S phase and APC(Cdc20) at late M phase, that degradation initiates within the N-terminal approximately 1400 amino acids retained by every fusion, and that the prevalent MLL fusions tested are RESISTANT to that degradation, which they proposed as a universal mechanism underlying MLL leukaemias. A dTAG/FKBP12-F36V system recruits a different E3 ligase, so this does not formally block the approach, but it means the fusion is a protein with a documented, disease-relevant history of escaping ubiquitin-proteasome turnover, and the 28-day maintained-depletion arm could silently become a 28-day PARTIAL-depletion arm. A partial depletion that drifts upward over four weeks would be read as programme recovery, producing a false positive for the hit-and-run model. The controls section should therefore add quantitative fusion-protein measurement at every interval timepoint, not just confirmation of loss, with a pre-specified depletion floor below which the timepoint is discarded. Read: abstract only. Liu H et al. 2007
- The outcome this experiment treats as the surprising branch has already happened in another system, so power the design for it rather than against it. Green et al. showed that transient Bcl6 expression in murine HSPCs produces mature B-cell lymphomas that lack Bcl6 and its target repression yet retain HSPC-stage epigenetic changes. If KMT2A::AFF1 behaves the same way, the readouts that will show it are the DNA-methylation (EM-seq) and H3K79me2 CUT&RUN arms rather than the HOXA9/MEIS1 transcript score, because the transcript score is what would recover while the chromatin state never changed. Read: abstract only. Green MR et al. 2014
- Searched and empty, recorded so the gap is a fact rather than an absence. PubMed searches for an endogenous-locus FKBP12-F36V/dTAG knock-in at KMT2A in a t(4;11) human line, and for dTAG or degron studies of MLL/KMT2A fusion proteins generally, returned nothing in this pass (queries combining dTAG, degron, FKBP12 F36V, KMT2A, MLL-AF4, MLL-AF9, SEM, acute depletion). The entry's claim that the decisive experiment has not been published survives this check. No citation is attached because there is nothing to cite; this note records the search, not a source.
- The two papers were doing different experiments on different fusions and were merged into one sentence. The merged version made the tool class look better characterised in t(4;11) than it is: **nobody has run acute degradation on KMT2A::AFF1 at all** - the minute-scale data are MLL::AF9 in a different system. Molecular Cell 2022
- E1bstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.journal-article
Acute degron-mediated degradation of MLL::AF9 changes transcriptional elongation at target genes within 15 minutes, with loss of the active chromatin landscape following afterwards - establishing the kinetic separation this experiment is designed to exploit.
MLL::AF9 degradation induces rapid changes in transcriptional elongation and subsequent loss of an active chromatin landscape · Molecular Cell 2022What this evidence cannot show
**MLL::AF9, not KMT2A::AFF1**, and in a system with an HA-FKBP12 tag knocked in at the endogenous allele. The 15-minute figure is the elongation readout; chromatin changes are slower. Whether KMT2A::AFF1 behaves the same way is unmeasured - this entry's whole premise is that it has not been done.
Evidence that argues against this (1)
- Carrying an MLL::AF9 kinetic result into a KMT2A::AFF1 experiment design assumes the two fusions share turnover and re-loading behaviour. They differ in partner, in complex composition and in disease lineage, and this graph has repeatedly found partner-specific behaviour where it expected class behaviour. Blood 2005
the proposed knock-in degron is the design this audit finds nobody has published
that experiment proposes the knock-in design this entry finds unpublished
the same knock-in supplies the endogenous degron that experiment assumes
that entry cites the same siRNA paper for a claim about minute-scale elongation change, which the siRNA paper does not make
Counter-evidence (Green 2014, PMID 24887457; Liu 2013, PMID 24150221): hit-and-run is documented for a transcription-factor oncogene initiated in HSPCs, and an MLL-AF9 leukaemic genome is fully reprogrammable to pluripotency -- so the degron washout arm must be powered to detect programme recovery, with the chromatin and methylation readouts rather than the transcript score as the discriminating measurements
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
Counter-evidence (Liu 2007, PMID 17908926): prevalent MLL fusions resist the SCF(Skp2)/APC(Cdc20) degradation that clears wild-type MLL, so the 28-day maintained-depletion arm needs quantitative fusion-protein measurement at every timepoint and a pre-specified depletion floor, or a drifting partial depletion will be misread as programme recovery
the 28-day degron hold is the decisive test
EM-seq arm tests methylation as the candidate memory carrier
matched-occupancy arm compares menin inhibition with degradation
editing's correct role in this project is as a laboratory reagent, e.g. degron knock-in
recovery kinetics after washout, never measured for any KMT2A fusion
shares the endogenously engineered line and the washout logic