Every published MLL-fusion degron tags an overexpressed construct, not the endogenous allele - and that distinction decides what the washout kinetics mean
also called dTAG fusion allele versus construct · degron attribution · FKBP12-F36V MLL fusion
Degron systems let you destroy a protein in minutes and watch what happens, and the numbers they give are the fastest and cleanest in this field. But in every published case the destroyed protein was an extra copy put into the cell artificially, so the speed of recovery is set by the artificial copy rather than by the patient's own gene.
The same thing, for a scientist
A degron gives two things an inhibitor cannot: **attribution** (the tag is unique to the tagged species) and **speed** (minutes rather than days). This graph cites degron kinetics in several places and has not so far separated the two designs that produce them. **Design A - degron on an overexpressed construct.** The only published MLL-fusion degron. Murine LSK cells and human CD34+ cells were transduced with HA-FKBP12(F36V)-tagged MLL::AF9 and depleted with dTAG-13 or dTAG-VHL; transcriptional elongation changed within 15 minutes and the active chromatin landscape was lost subsequently [E1]. Attribution is perfect - no wild-type species carries FKBP12. But abundance is set by a retroviral promoter, **two wild-type MLL1 alleles remain**, and the authors say so explicitly and name endogenous degron tagging as the needed follow-up [E1]. **Design B - degron knocked into the endogenous fusion allele.** Not published for any KMT2A fusion (`concept-epitope-tagged-knock-in-at-the-endogenous-fusion-allele`). **Why the difference changes the interpretation of kinetics.** | quantity | Design A gives | Design B would give | |---|---|---| | time to protein loss | rate of the degrader on the tagged pool - roughly design-independent | same | | **time to transcriptional response** | depends on the size of the pool being cleared, so on the promoter | the endogenous number | | **recovery time after washout** | set by the retroviral promoter's resynthesis rate | set by the endogenous KMT2A promoter | | compensation by wild-type MLL1 | present, and measured as absent at 3 h [E1] | present, and the physiological ratio | The first row transfers. **The second and third do not**, and they are the rows this graph's hysteresis and washout-recovery modelling depends on (`ana-fusion-halflife-degron-vs-menin`, `exp-degron-washout-recovery-kmt2a-aff1`). A recovery half-time measured from an MSCV-driven construct is a property of MSCV. **What Design A does establish, and it is a lot.** That fusion degradation changes elongation within 15 minutes; that chromatin changes follow rather than precede; that at 3 hours there is no compensatory change in wild-type MLL1 occupancy or H3K4me3; and that combined DOT1L and menin inhibition approaches the degron phenotype where either single agent does not [E1]. None of those depends on the absolute expression level, because they are statements about ordering and about relative effects within one system.
How to cite a degron result correctly in this graph
Three rules that follow directly.
- Orderings transfer; absolute times do not. 'Elongation changes before chromatin' is a Design A result that survives transplantation to the endogenous setting. '15 minutes' is a number from a specific system and should always carry the system.
- Recovery and washout numbers do not transfer at all. They are the resynthesis rate of whatever promoter drives the construct. Any entry in this graph quoting a degron washout recovery time must say which promoter.
- A degron result is fusion-attributable even in Design A. This is the half people get wrong in the other direction. The FKBP12 tag is unique, so the degradation and everything measured against it is attributable to the tagged fusion. Design A's weakness is stoichiometry, not attribution.
The degron as a chromatin-attribution instrument
The most transferable thing in the Design A paper is not a kinetic number at all. It is that degron subtraction turns a shared-epitope antibody into a discriminating one: the MLL-N ChIP signal that disappears within three hours of degradation, while the MLL-C signal and H3K4me3 do not, is fusion signal 1. The discriminator is the perturbation, not the reagent.
That logic works at any expression level and in any line, and it is the argument for putting a degron into the endogenous allele even if one does not care about kinetics: it would let the field re-analyse every existing MLL-N track in a patient-derived background.
| reagent class | FKBP12(F36V) dTAG degron on an MLL fusion |
|---|
published designs
[
{
"design": "degron on a retrovirally overexpressed construct",
"instances": [
"PMID 35245435, MLL::AF9, murine LSK and human CD34+"
],
"attribution": "yes, tag is unique",
"stoichiometry": "promoter-set, not endogenous"
},
{
"design": "degron knocked into the endogenous fusion allele",
"instances": [],
"attribution": "would be yes",
"stoichiometry": "endogenous"
}
]transfers across designs
[ "ordering of events (elongation before chromatin)", "attribution of degraded species", "degron-subtraction logic for ChIP" ]
does not transfer
[ "absolute time to transcriptional response", "washout and recovery half-times", "the fusion-to-wild-type abundance ratio" ]
kinetic facts
[
{
"quantity": "time to elongation change after degradation",
"value": "within 15 minutes",
"system": "retrovirally overexpressed HA-FKBP12-MLL::AF9",
"source_pmid": "35245435"
}
]- A degrader developed against a KMT2A fusion would be dosed on a schedule inferred from washout kinetics measured on an overexpressed construct; if the endogenous protein resynthesises faster, the drug holiday is longer than the biology allows and the schedule under-doses.
- Conversely, if the endogenous protein resynthesises more slowly, intermittent dosing may be sufficient and continuous exposure adds toxicity for nothing - the two errors point in opposite directions and only an endogenous measurement distinguishes them.
- Degron phenotypes are the benchmark against which menin and DOT1L inhibitors are judged 'insufficient'; if that benchmark is set by an overexpressed fusion, the bar for a small molecule may be set higher than the disease actually requires.
- What is the washout recovery half-time of an MLL fusion expressed from its endogenous promoter? Every published number comes from a retroviral construct, and this graph's hysteresis modelling needs the endogenous one.
- How long does degron subtraction remain a valid ChIP attribution method before wild-type MLL redistribution confounds it? The three-hour timepoint was chosen for elongation kinetics, not validated as an attribution window.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The published MLL-fusion degron system uses HA-FKBP12(F36V)-tagged MLL::AF9 delivered by retroviral transduction into murine LSK cells and human CD34+ cells, with two intact wild-type MLL1 alleles remaining; degradation changes transcriptional elongation within 15 minutes and causes subsequent loss of active chromatin, no compensatory change in wild-type MLL1 occupancy or H3K4me3 was seen at three hours, and the authors state in their Limitations that future studies should include endogenously degron-tagged MLL::AF9.
MLL::AF9 degradation induces rapid changes in transcriptional elongation and subsequent loss of an active chromatin landscape. · Olsen SN, Godfrey L, Healy JP, et al. · Molecular cell 2022What this evidence cannot show
Read verbatim from the PMC body (PMC9044330, efetch db=pmc, <body> present, 108,959 characters of stripped text). MLL::AF9 in myeloid systems; there is no equivalent for KMT2A::AFF1 in a lymphoid context, and the elongation machinery the two fusions engage is shared but not identical. The 15-minute figure is for a core subset of target genes, not genome-wide.
Evidence that argues against this (1)
- The absence of wild-type MLL1 compensation at three hours does not exclude compensation on a longer timescale, and three hours is short relative to the days over which menin and DOT1L inhibitors act. A long-hold degron experiment - the design this graph proposes in exp-degron-washout-recovery-kmt2a-aff1 - would be needed to exclude it, and no published work holds depletion past a few days. Olsen SN et al. 2022
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database
No degron has been knocked into an endogenous KMT2A fusion allele: a title/abstract search of KMT2A or MLL against 'endogenously tagged', 'endogenous tagging' and dTAG returns zero records, and a search of four KMT2A::AFF1 spellings against knock-in, degron and dTAG terms returns only a murine fusion-gene knock-in.
What this evidence cannot show
Bounded absence with the bounds stated in concept-epitope-tagged-knock-in-at-the-endogenous-fusion-allele. Title/abstract only; the parser collapsed both double-colon spellings to hyphenated form. A degron knock-in described only in a methods section, or distributed as an unpublished resource, would not appear.
Evidence that argues against this (1)
- The stronger evidence for this absence is not the query but the 2022 Limitations statement by the group that built the only MLL-fusion degron. If another group had since published one, the query would most likely catch it, but a null query is weak evidence on its own and is offered here as corroboration. Olsen SN et al. 2022
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Degron subtraction converts a shared-epitope antibody into a discriminating one: MLL-N ChIP signal lost within three hours of degron-induced fusion degradation, with no matching loss in MLL-C signal and no change in H3K4me3, is attributable to the fusion by perturbation rather than by reagent.
MLL::AF9 degradation induces rapid changes in transcriptional elongation and subsequent loss of an active chromatin landscape. · Olsen SN, Godfrey L, Healy JP, et al. · Molecular cell 2022What this evidence cannot show
Read verbatim from the PMC body. The subtraction attributes lost signal to the degraded protein; it does not by itself distinguish direct binding from loss of a co-factor that required the fusion, and three hours is long enough for the latter.
Evidence that argues against this (1)
- If wild-type MLL1 redistributes onto vacated fusion sites on a timescale longer than three hours, then a degron-subtracted map made at three hours is correct and a map made at 24 hours would not be - so the subtraction has a validity window that nobody has measured. Liu H et al. 2007
that analysis fits washout kinetics whose published values come from an overexpressed construct
that experiment proposes the knock-in design this entry finds unpublished
the same construct solves both problems; a degron cassette is an epitope tag with a function
degron subtraction is one of the two surviving occupancy attribution designs
a titration design inherits the same construct-versus-allele distinction
junction-directed siRNA is the third perturbation-based attribution route alongside tag and degron
HCB1 carries MLL-AF9-HA-FKBP12 as a TRANSDUCED construct in human CD34+ cord blood cells, not a degron knocked into an endogenous fusion allele - the distinction that entry draws, applied to the one degron dataset this project has scored.