Normal blood needs DOT1L too - lose it in an adult mouse and the marrow fails within weeks
also called Dot1l in normal hematopoiesis · DOT1L HSC requirement · DOT1L bite-back
DOT1L is not a leukaemia-only protein. Delete it in an adult mouse and blood production collapses: progenitors disappear within a week and the long-lived stem cells follow, so any drug that removes DOT1L is working against the tissue it is trying to save.
The same thing, for a scientist
The therapeutic-index question for this target is settled in the unhelpful direction and the graph should say so plainly. Conditional deletion of *Dot1l* in postnatal mice causes **pancytopenia and failure of haematopoietic homeostasis**, and Dot1l-deficient cells minimally reconstitute recipient marrow in competitive transplantation; in the same study MLL-AF9 cells required Dot1l for transformation while Hoxa9/Meis1- and E2A-HLF-driven cells did not [E1]. Acute inactivation in adult marrow reduces myeloid progenitors within **7 days** and then depletes long-term HSCs, and the rescue experiment localises the requirement: wild-type and **PPI-deficient** DOT1L restore sustained haematopoiesis, an **enzymatically inactive** mutant does not [E2]. So normal adult blood needs DOT1L's **catalytic activity** and does not need its AF9 interaction - the exact inverse of the leukaemic cell, which is the whole therapeutic-index argument and the reason recruitment-interface inhibition and degradation point in opposite directions (`mechanism-dot1l-ppi-inhibition-modality`, `drug-dot1l808`). Two partial exceptions bound the requirement: embryonic **primitive erythropoiesis** needs DOT1L protein and not its catalytic activity [E3], and pharmacological inhibition with EPZ-5676 in vivo minimally affects erythropoiesis [E3] - consistent with the clean safety profile pinometostat showed in adults [E4].
The gap between genetic ablation and a drug, stated honestly
Everything above about marrow failure comes from complete, irreversible genetic deletion. A drug gives partial, reversible, dose-dependent target coverage for weeks. This project's standing rule about database and genetic facts applies here in full: a developmental or knockout phenotype says the gene matters, it is not by itself a safety prediction for pharmacological inhibition.
The empirical check supports the caution. Pinometostat had a clean safety profile in 51 adults and never reached its maximum tolerated dose 4, and EPZ-5676 in vivo minimally perturbed mouse erythropoiesis 3. So the achievable degree of catalytic inhibition in a patient was apparently well short of the genetic phenotype.
That cuts both ways, and this is the uncomfortable part of the beat. If the drug was tolerable because target coverage was incomplete, then the trial tested neither the safety limit nor the efficacy limit of catalytic DOT1L inhibition, and the modest activity is uninformative about the target. The two explanations - wrong modality and insufficient exposure - are not distinguished by anything published.
What a degrader does to this picture
A degrader removes the catalytic activity that adult marrow demonstrably requires 2, in addition to the scaffold. The prediction from these genetics is therefore that DOT1L degradation is more myelosuppressive than catalytic inhibition, not less, and that its dose-limiting toxicity will be haematological in exactly the population being treated.
Neither published degrader paper reports a haematological comparison against a catalytic inhibitor at matched target coverage, and one reports 'no toxicity on normal cells' without stating which cells. This is the single largest unaddressed risk in the post-pinometostat programme and it is written as exp-dot1l-degrader-vs-inhibitor-matched-exposure.
| mechanism class | normal-tissue requirement / bite-back |
|---|---|
| gene | gene-dot1l |
| adult marrow phenotype | myeloid progenitor loss within 7 days of acute inactivation, then long-term HSC depletion; pancytopenia and failed reconstitution in competitive transplant |
| function required in adult marrow | catalytic activity (enzymatically inactive mutant fails to rescue) |
| function NOT required in adult marrow | the AF9-DOT1L protein-protein interaction (PPI-deficient mutant rescues) |
| function required in embryonic primitive erythropoiesis | protein presence, not catalytic activity |
| pharmacological counterpoint | EPZ-5676 in vivo minimally affected mouse erythropoiesis; pinometostat was well tolerated in 51 adults with the MTD never reached |
| therapeutic index implication | recruitment-interface inhibition is favoured by these genetics; degradation is disfavoured by them |
| lymphoid requirement | see mechanism-dot1l-normal-b-lymphopoiesis |
- Adult marrow requires DOT1L's catalytic activity, so an inhibitor that fully engages the enzyme should be myelosuppressive; the clean safety profile observed clinically may therefore be evidence of incomplete target engagement rather than of a real therapeutic window.
- A degrader removes both the catalytic and the scaffolding function and is predicted by these genetics to be the more haematotoxic modality, in a population whose marrow is already compromised.
- Because the AF9 interface is dispensable in normal blood and essential in MLL-AF9 leukaemia, the safest modality is the one with the weakest chemical matter - micromolar peptide-derived interface inhibitors against a picomolar catalytic benchmark.
- What degree of H3K79 methylation reduction did EPZ-5676 achieve in the mouse erythropoiesis experiment? Without that number, 'minimally affected' cannot be distinguished from 'minimally engaged', and the same ambiguity sits over the human safety data.
- Is a DOT1L degrader more myelosuppressive than a catalytic inhibitor at matched target coverage? The adult-marrow rescue data predict yes and nothing published tests it.
- Does the adult-mouse assignment - marrow needs catalysis, not the AF9 interface - hold in the paediatric and infant haematopoietic system, where the embryonic data give the opposite answer?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In a conditional Dot1l knockout mouse, deletion of Dot1l led to pancytopenia and failure of haematopoietic homeostasis, and Dot1l-deficient cells minimally reconstituted recipient bone marrow in competitive transplantation; MLL-AF9 cells required Dot1l for oncogenic transformation whereas cells transformed by Hoxa9/Meis1 or E2A-HLF did not.
Requirement for Dot1l in murine postnatal hematopoiesis and leukemogenesis by MLL translocation. · Jo SY, Granowicz EM, Maillard I, et al. · Blood 2011What this evidence cannot show
Abstract read from the eutils record; full text not read. Mouse conditional knockout - complete genetic ablation, which is not the perturbation a drug delivers. MLL-AF9 is the only MLL fusion tested. The selectivity claim rests on two non-MLL comparators, one of which (Hoxa9/Meis1) is downstream of the same programme and so is a weak negative control for a HOX-driven mechanism.
Evidence that argues against this (1)
- Pharmacological inhibition does not reproduce the genetic phenotype: EPZ-5676 administered in vivo minimally affected mouse erythropoiesis, and the adult human trial of the same compound had a clean safety profile at doses that never reached a maximum tolerated dose. The knockout overstates what an inhibitor does to normal blood. Malcom CA et al. 2021
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Acute Dot1l inactivation in adult mouse bone marrow reduces myeloid progenitor numbers within 7 days and subsequently depletes long-term haematopoietic stem cells; wild-type and PPI-deficient DOT1L mutants, but not an enzymatically inactive mutant, rescue sustained haematopoiesis, showing the AF9-DOT1L interaction is dispensable in non-leukaemic haematopoiesis while catalytic activity is not.
Elucidating the Importance of DOT1L Recruitment in MLL-AF9 Leukemia and Hematopoiesis. · Grigsby SM, Friedman A, Chase J, et al. · Cancers (Basel) 2021What this evidence cannot show
Abstract read from the eutils record; full text not read. Mouse marrow; rescue constructs expressed at unstated levels against endogenous background. The 7-day progenitor loss is a genetic ablation timescale and does not translate to a dosing interval. 'Enzymatically inactive' is a point mutant, which retains the protein and therefore tests catalysis rather than protein presence.
Evidence that argues against this (1)
- The opposite assignment holds in embryonic primitive erythropoiesis, where a methyltransferase-mutant mouse is not overtly anaemic while the full knockout is - so which DOT1L function blood needs depends on the developmental stage, and a paediatric population is not an adult mouse. Malcom CA et al. 2021
- E3strong 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, with comparable yolk-sac vascularity, AGM red cells and primitive erythroblast colonies to wild type, although definitive erythroblast colonies were smaller; in vivo administration of EPZ-5676 minimally affected erythropoiesis. Early embryonic erythropoiesis therefore requires a DOT1L function independent of its 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; both genotypes die by E13.5, so catalytic activity is required for something later that this study does not identify. Phenotyping at E10.5 is largely morphological plus colony assay. The EPZ-5676 in vivo dose, schedule and degree of H3K79 methylation reduction achieved are not stated in the abstract, which is what would make the 'minimally affected' claim interpretable.
Evidence that argues against this (1)
- A minimal effect of EPZ-5676 on erythropoiesis is uninterpretable without the degree of target engagement achieved. The same compound in humans reduced H3K79 methylation at doses that never reached a maximum tolerated dose, so 'the drug did not hurt normal cells' and 'the drug did not fully engage the target' are the same observation in both settings. Stein EM et al. 2018
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
In 51 adults with acute leukaemia, pinometostat reduced H3K79 methylation and produced 2 complete remissions; the compound's clinical activity was modest.
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. This entry uses it for the tolerability observation, and tolerability in a relapsed/refractory adult leukaemia population with existing cytopenias is a weak test of marrow toxicity - baseline pancytopenia masks drug-induced pancytopenia. The maximum tolerated dose was never reached.
Evidence that argues against this (1)
- DOT1L's requirement in normal haematopoiesis extends to lymphoid development: H3K79me2 accumulates during murine B-cell maturation and DOT1L is essential for pre-B expansion and marginal-zone B-cell generation. In a B-ALL population the host compartment being treated is the compartment that needs the mark, which a myeloid-focused safety argument does not capture. Kealy LC et al. 2026
adult marrow failure within weeks of Dot1l inactivation
long-term HSCs depleted after acute Dot1l loss
adult marrow needs the catalytic activity a degrader removes
PPI-deficient DOT1L rescues normal blood; AF9 interface is dispensable there
the lymphoid half of the normal-tissue requirement