DOT1L was not refuted - the wrong thing was blocked, at a dose nobody finished escalating
also called wrong modality hypothesis · pinometostat failure reinterpretation
The usual story is that DOT1L was tested in patients and failed, so the target is dead. This hypothesis says the drug blocked only the enzyme's chemistry while leaving its other job intact, and was stopped before anyone found the highest safe dose - so the target was never really tested.
The same thing, for a scientist
**Statement.** The modest clinical activity of pinometostat reflects the modality and the exposure, not the validity of DOT1L as a target in KMT2A-rearranged leukaemia. Three independent facts support the reading and none of them is decisive on its own. **(1) Catalysis is not all of DOT1L.** Knockout and catalytic-dead knock-in diverge in cell-fate determination and in transcription elongation, and embryonic erythropoiesis needs the protein and not the activity [E1][E2]; a catalytic inhibitor cannot reach any of that. **(2) The dose was never finished.** The adult phase 1 never reached a maximum tolerated dose [E3], so the exposure-response curve was sampled only at its lower end - and at low exposure the drug acts through a different pathway (FLT3) than at high exposure (HOXA9/MEIS1) [E4]. **(3) The alternative modality works preclinically where the drug did not**: DOT1L PROTACs achieve complete regression in an orthotopic model and retain activity in menin-inhibitor-resistant cells [E5]. **The competing hypothesis is not that the target is dead but that DOT1L is a moderate, graded dependency whose maximal achievable benefit is small** - a reading the graph's own DepMap synthesis supports, with a lineage-consistent raw effect of -0.178 Chronos and a rank of 61 of 8,269 rather than 1. Prior probability 0.45: the evidence for a catalysis-independent function is strong, the evidence that it matters *in this disease* is absent.
What would have to be true
- A substantial part of DOT1L's contribution to the KMT2A-fusion programme survives catalytic inactivation in a leukaemic cell. This has been shown in ESCs and in mouse embryos and never in this disease - the single largest hole in the hypothesis.
- Removing DOT1L protein produces more anti-leukaemic effect than removing its activity, at matched target coverage and matched marrow toxicity. Both degrader papers report efficacy, neither reports that comparison.
- The exposure achieved in the trial was below what is needed for the canonical mechanism. Plausible given the unreached MTD, and unquantified - nobody has related intracellular pinometostat concentrations in the trial to the in vitro dose thresholds.
If 1 is false, the correct reading is under-dosing, which is a different and simpler failure with a different fix (a better-scheduled catalytic inhibitor rather than a degrader).
Why this is not special pleading
The reason to write this down rather than accept the refutation is that the two readings have opposite consequences for the menin programme, which is this project's central bet.
If DOT1L is a refuted target, then the lesson for menin is 'excellent target validation and excellent chemistry do not guarantee clinical benefit in this disease', which is a warning about the whole approach.
If pinometostat was a wrong-modality result, then the lesson is narrower and more useful: a catalytic inhibitor of a chromatin enzyme may not remove the enzyme's scaffolding role, and menin inhibitors - which act by displacing a protein-protein interaction rather than by blocking catalysis - are not subject to the same failure. Menin has no catalytic activity to spare, so there is no analogous scaffold left behind when its interface is blocked.
That asymmetry is real and it is the reason this hypothesis is worth resolving rather than assuming.
| statement | The modest clinical activity of pinometostat reflects modality (catalytic inhibition leaves DOT1L's scaffolding function intact) and exposure (the MTD was never reached), not refutation of DOT1L as a target in KMT2A-rearranged leukaemia. |
|---|---|
| origin | literature-gap |
| prior probability | 0.45 |
| what it would mean for menin | If confirmed, the failure is specific to catalytic inhibition of a chromatin enzyme and does not transfer to menin inhibitors, which act by blocking a protein-protein interface on a protein with no catalytic activity to spare. If refuted, the failure is a general warning that target validation quality does not predict clinical benefit in this disease. |
predictions
[ "A catalytically dead DOT1L knock-in will NOT phenocopy DOT1L knockout in a KMT2A-rearranged line; the knockout will be more lethal.", "A DOT1L degrader will outperform a catalytic inhibitor at matched target occupancy in the same model.", "The degrader's advantage will be accompanied by greater marrow toxicity, because adult haematopoiesis requires DOT1L's catalytic activity.", "Pinometostat exposures achieved in the adult trial will map to the FLT3-pathway regime rather than to the HOXA9/MEIS1 regime." ]
experiments
[ "exp-dot1l-degrader-vs-inhibitor-matched-exposure", "exp-h3k79-state-resolved-icechip-dose-series" ]
analyses
[ "ana-dot1l-dependency-lineage-split-raw-effect" ]
falsifiers
[ "A catalytically dead DOT1L knock-in phenocopies DOT1L knockout in a KMT2A-rearranged line - then catalysis is the whole contribution in this disease and the modality argument fails.", "A DOT1L degrader gives no efficacy advantage over a catalytic inhibitor at matched target coverage in the same model.", "Pharmacokinetic bridging shows the trial achieved intracellular exposures above the in vitro threshold for HOXA9/MEIS1 collapse - then the dose arm fails and the result is a genuine efficacy ceiling.", "Degrader advantage disappears when the catalytic inhibitor is given on a schedule matched for time-integrated target coverage - then the difference was pharmacokinetic, not mechanistic." ]
- If the hypothesis is right, the fix - a degrader - removes the DOT1L function that normal adult marrow requires, so the modality that restores efficacy is the one predicted to import haematological toxicity.
- Reinterpreting a negative clinical result as a wrong-modality result is the standard way a field keeps a dead target alive, and this hypothesis is stated with a prior below 0.5 and explicit falsifiers for exactly that reason.
- Even if fully confirmed, the graph's own dependency synthesis puts DOT1L at a lineage-consistent raw effect of -0.178 Chronos and rank 61 of 8,269 - a moderate dependency whose ceiling may be modest regardless of modality.
- Has anyone made a catalytically dead DOT1L knock-in in a KMT2A-rearranged human leukaemia line? This beat's reading found the genotype in ESCs and in mouse embryos and nowhere in this disease.
- What intracellular pinometostat exposure did the trial achieve, relative to the in vitro thresholds for the FLT3 and HOXA9 regimes?
- Why did dose escalation stop without reaching an MTD? The answer is likely to be in the trial protocol or sponsor record rather than in the literature, and it decides how much weight the exposure arm of this hypothesis can bear.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In mouse embryonic stem cells, DOT1L loss rather than catalytic inactivation causes defects in glial cell specification, and elongation defects under super-elongation-complex inhibition are exacerbated in DOT1L knockout but not in catalytically dead DOT1L cells, revealing a role in productive transcription elongation independent of H3K79 methylation.
DOT1L-controlled cell-fate determination and transcription elongation are independent of H3K79 methylation. · Cao K, Ugarenko M, Ozark PA, et al. · Proc Natl Acad Sci U S A 2020What this evidence cannot show
Abstract read from the eutils record; full text not read. Embryonic stem cells and neural differentiation - not haematopoietic, not leukaemic, and not a KMT2A fusion. The elongation phenotype requires a sensitising SEC inhibitor to appear.
Evidence that argues against this (1)
- The leukaemic dependency was established with an SAM-competitive catalytic inhibitor that extended survival in MLL-fusion xenografts, which is hard to reconcile with catalysis being a minor part of DOT1L's contribution in this disease specifically. Daigle SR et al. 2013
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A Dot1l methyltransferase-mutant mouse is not overtly anaemic at E10.5 where the Dot1l knockout is, and EPZ-5676 administered in vivo minimally affects erythropoiesis, indicating that early embryonic erythropoiesis requires a DOT1L function independent of methyltransferase activity.
Primitive Erythropoiesis in the Mouse is Independent of DOT1L Methyltransferase Activity. · Malcom CA, Ratri A, Piasecka-Srader J, et al. · Front Cell Dev Biol 2021What this evidence cannot show
Abstract read from the eutils record; full text not read. Mouse embryo, primitive erythropoiesis; both genotypes die by E13.5 and definitive erythroblast colonies from the mutant were smaller, so the separation is partial. Normal development, not leukaemia.
Evidence that argues against this (1)
- The opposite assignment holds in adult marrow: an enzymatically inactive DOT1L fails to rescue sustained haematopoiesis while a PPI-deficient one succeeds. Which DOT1L function a tissue needs depends on the tissue and the stage, so a developmental separation does not license a general claim that catalysis is dispensable. Grigsby SM et al. 2021
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Pinometostat reduced H3K79 methylation as intended in 51 adults with acute leukaemia and produced 2 complete remissions, with modest overall clinical activity.
The DOT1L inhibitor pinometostat reduces H3K79 methylation and has modest clinical activity in adult acute leukemia. · Stein EM, Garcia-Manero G, Rizzieri DA, et al. · Blood 2018What this evidence cannot show
Abstract read from the eutils record; full text not read. The claim that the maximum tolerated dose was never reached is carried by this graph's existing pinometostat entries and by the trial report; it is the load-bearing caveat for this hypothesis and anyone relying on it should read the trial's dose-escalation section rather than the abstract.
Evidence that argues against this (1)
- Not reaching an MTD in a phase 1 usually means escalation stopped for reasons other than toxicity - sponsor decision, pharmacokinetic plateau, or a judgement that further escalation would not change the outcome. An unreached MTD is therefore not automatically an untested dose range, and the hypothesis treats it as though it were. Cancer Chemother Pharmacol 2016
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Some MLL-rearranged leukaemias are susceptible to pinometostat concentrations far below those that downregulate HOXA9 and MEIS1, acting mainly through transcriptional downregulation of FLT3; loss of FLT3 function recapitulates low-dose cytotoxicity and constitutively active STAT5A rescues it.
Non-canonical H3K79me2-dependent pathways promote the survival of MLL-rearranged leukemia. · Richter WF, Shah RN, Ruthenburg AJ · Elife 2021What this evidence cannot show
Abstract read from the eutils record; full text not read. Cell lines, with 'some MLL-rearranged leukemias' left without a denominator. Relating in vitro concentration thresholds to the exposures achieved in the trial requires pharmacokinetic bridging that nobody has done.
Evidence that argues against this (1)
- If a low exposure is already cytotoxic through FLT3, then the trial was not obviously under-dosed for effect - it was dosed into a range where an anti-leukaemic mechanism exists - and still produced 2 complete remissions in 51 patients. That weakens rather than strengthens the exposure arm of this hypothesis. Stein EM et al. 2018
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
DOT1L-targeting PROTACs developed explicitly because non-enzymatic DOT1L functions cannot be blocked by enzymatic inhibitors show MLL-rearrangement-dependent viability effects, retained activity against menin-inhibitor-resistant cells, and complete tumour regression in an orthotopic leukaemia model without overt toxicity.
Development of DOT1L-Targeted Protein Degraders for Treating MLL-r Leukemia. · Quan S, Unno K, Gupta DG, et al. · J Med Chem 2026What this evidence cannot show
Abstract read from the eutils record; full text not read; single 2026 report. The MLL-r dependence and menin-resistance results belong to one compound and the in vivo regression to another. Crucially, no comparison against a catalytic inhibitor is reported, so the paper demonstrates that degradation works and not that it works better.
Evidence that argues against this (1)
- A degrader could outperform a catalytic inhibitor for a reason that has nothing to do with the hypothesis: an event-driven pharmacology with a long duration of effect solves the same scheduling problem that the mark's slow kinetics create. Superiority of the degrader would then be evidence about exposure, not about catalysis-independent function. Chory EJ et al. 2019
- refutes →Pinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
reads the trial as a modality and exposure result, not a refuted target
the hypothesis stands or falls on catalysis-independent function existing in this disease
knockout versus catalytic-dead in a KMT2A-rearranged line is the decisive comparison