H3K79 methylation is what keeps Polycomb off - and that, not gene silencing, is what a DOT1L inhibitor actually undoes
also called DOT1L-PRC1.1 antagonism · H3K79me-H2AK119ub crosstalk · DOT1L transcriptional memory
Genes that a leukaemia keeps switched on are protected from a silencing system called Polycomb, and the H3K79 tag is the protection. When a DOT1L drug strips the tag away, Polycomb moves in and shuts the genes down - and it does this gradually, as the tag is lost, rather than the moment the drug binds.
The same thing, for a scientist
A 2026 study identifies non-canonical **PRC1.1** as the mediator of the cellular response to both DOT1L and menin inhibitors, and separates the two drugs' mechanisms at the chromatin level [E1]. Menin inhibition induces **PRC1.1-dependent H2AK119ub deposition** that silences a subset of MLL-fusion targets. DOT1L inhibition instead produces a **genome-wide increase** in H2AK119ub. The critical mechanistic statement is that enhanced PRC1.1 activity arises specifically from the **progressive loss of DOT1L-mediated H3K79 methylation, independent of MLL-fusion displacement or transcriptional repression**, and is driven by **direct biochemical antagonism** between H3K79 methylation and PRC1 activity. The authors frame DOT1L as a component of transcriptional memory co-opted in leukaemia - the missing link balancing the MLL-Polycomb axis. Three consequences for this project. **The word 'progressive' ties the pharmacology to the mark's slow kinetics**, so the effect of a DOT1L inhibitor is set by integrated exposure, not occupancy. **The two drugs are not redundant**: one produces targeted silencing, the other a genome-wide Polycomb shift, which reframes the reported DOT1L-menin synergy [E2][E3] as complementarity rather than dose-addition. And **the genome-wide H2AK119ub increase is an off-target liability** in a way a mark-depletion model does not predict.
Why 'independent of MLL-fusion displacement' is the load-bearing clause
A simpler model was available and would have been wrong: DOT1L inhibition removes the mark, the fusion complex falls off, genes go quiet, Polycomb fills the vacancy. The paper explicitly excludes that ordering - the PRC1.1 response arises from loss of H3K79 methylation itself, not from fusion displacement and not as a consequence of transcriptional repression 1.
That makes H3K79 methylation an actively antagonistic mark rather than a passive elongation correlate. It is doing work: excluding a repressive complex. And it means the quantity to measure in a trial is not whether the fusion is still on chromatin but how much H3K79 methylation is left, which is the one thing bulk pharmacodynamics can in principle report.
It also gives the cleanest available explanation for why DOT1L inhibition is slow. The mark is lost slowly (mechanism-h3k79-turnover-and-dilution), so the Polycomb takeover is slow, so the transcriptional consequence is slow, so a trial with a short exposure and an unescalated dose sees very little.
What it does to the menin-DOT1L combination argument
The graph already holds two reports of complementary or synergistic activity between DOT1L and menin inhibitors 23. This work supplies a mechanism and changes what kind of combination it is 1.
If menin inhibition produces targeted PRC1.1-dependent silencing at a subset of fusion targets and DOT1L inhibition produces a genome-wide H2AK119ub increase, then the two agents are not two ways of doing the same thing. Their combination is a targeted silencing event superimposed on a global shift in Polycomb tone. That predicts synergy on efficacy and it also predicts that the toxicity of the pair is not the sum of two on-target toxicities - a genome-wide Polycomb shift in normal haematopoietic cells is a different kind of risk from menin displacement.
It also bears on resistance. If the endpoint of DOT1L inhibition is Polycomb-mediated silencing, then any lesion that impairs PRC1.1 confers resistance to a DOT1L inhibitor without touching DOT1L or the mark - a resistance route that no H3K79 pharmacodynamic assay would detect.
Compartment qualification: the direction is not universal, and an independent 2026 paper reports the opposite sign
This entry states that DOT1L inhibition raises H2AK119ub, and it is cited in 25 entries. The statement is compartment-unqualified, and it should not be.
An independent 2026 study in a different disease reports the opposite sign. In germinal-centre B-cell-like DLBCL, DOT1L shapes the composition of chromatin-bound ncPRC1 complexes and regulates, in part, the deposition of H2AK119ub1 at promoters co-occupied by H3K27me3 - defined there as PRC1/2 targets. The axis replicates; the direction at those promoters does not.
Two things follow, and both are practical.
- Say which compartment. 'DOT1L inhibition raises H2AK119ub' is safe as a genome-wide statement in MLL-fusion leukaemia and unsafe as a statement about PRC1/2-target promoters. The 25 citing entries inherit the unqualified form.
- The assay differs with it. Menin inhibition gives a subset-restricted H2AK119ub gain while DOT1L gives a genome-wide one - so a bulk immunoblot is an adequate pharmacodynamic readout for DOT1L and useless for menin, where only a locus-resolved measurement will do.
This is not a retraction of the entry. The antagonism is replicated across two diseases by two groups, which is stronger evidence for the axis than this graph usually has. What is withdrawn is the licence to quote the direction without naming where it was measured.
| mechanism class | chromatin crosstalk / drug mechanism |
|---|---|
| menin inhibitor effect | PRC1.1-dependent H2AK119ub deposition silencing a subset of MLL-fusion targets (targeted) |
| dot1l inhibitor effect | genome-wide increase in H2AK119ub (global) |
| causal ordering | PRC1.1 activation follows progressive H3K79me loss, not MLL-fusion displacement and not transcriptional repression |
| prc1 1 targeting subunit | KDM2B (CxxC) - not named in the source abstract; noted here because KDM2B is separately and contestedly claimed as an H3K79 demethylase |
| predicted resistance route | PRC1.1 loss-of-function would confer DOT1L-inhibitor resistance without any change in DOT1L or in H3K79 methylation |
| source pmid | 41634341 |
antagonists
[ "mark-h3k79me2", "H2AK119ub deposited by non-canonical PRC1.1" ]
- A drug whose therapeutic effect is a genome-wide increase in Polycomb activity is one whose dose-limiting toxicity may be a genome-wide increase in Polycomb activity, in normal cells that need the same balance.
- If the endpoint is PRC1.1-mediated silencing, resistance can arise entirely downstream of the target - a PRC1.1 lesion - and would be invisible to every H3K79 methylation assay used to confirm target engagement.
- Because the PRC1.1 response tracks the progressive loss of the mark, intermittent dosing that allows the mark to re-accumulate resets the effect rather than merely reducing it.
- Does PRC1.1 loss confer resistance to DOT1L inhibition in a KMT2A-rearranged model? If it does, PRC1.1 integrity becomes a predictive biomarker no pharmacodynamic assay currently measures.
- Is the genome-wide H2AK119ub increase after DOT1L inhibition tolerated by normal haematopoietic progenitors, or is it the mechanism of the cytopenias seen with Dot1l loss?
- KDM2B is the CxxC targeting subunit of PRC1.1 and has separately been claimed as an H3K79me2/3 demethylase by one laboratory. Is the apparent locus-level loss of H3K79me during Polycomb takeover competitive exclusion or enzymatic erasure?
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Non-canonical PRC1.1 mediates cellular responses to DOT1L and menin inhibitors: menin inhibition induces PRC1.1-dependent deposition of H2AK119ub silencing a subset of MLL-fusion targets, whereas DOT1L inhibition produces a genome-wide increase in H2AK119ub; enhanced PRC1.1 activity arises specifically from the progressive loss of DOT1L-mediated H3K79 methylation, independent of MLL-fusion displacement or transcriptional repression, and is driven by direct biochemical antagonism between H3K79 methylation and PRC1 activity, conserved across cell types.
DOT1L provides transcriptional memory through PRC1.1 antagonism. · Neville D, Ferguson DT, Heikamp EB, et al. · Nat Cell Biol 2026What this evidence cannot show
Abstract read from the eutils record; full text not read. Published 2026, single report, not independently replicated. The cell systems, the MLL fusion used and the inhibitors' identities and concentrations are not stated in the abstract, and 'conserved across cell types' is asserted without the types being named. 'Direct biochemical antagonism' implies an in vitro reconstitution whose details are not visible from the abstract, which matters because the strength of the whole model rests on that experiment. **Resolved on all five open clauses.** Systems: mouse MLLAF9 plus human MOLM13, MV4;11, K562, HUDEP-2, mouse ES cells and H9. Dominant fusion: MLL-AF9. Inhibitors and doses: **SGC0946 3-5 microM; VTP-50469 300 nM at 8 h and 48 h**. And the phrase 'conserved across cell types' is now bounded: the span is **myeloid, erythroid and pluripotent, with zero B-lymphoid samples**. GSE181829, GSE260456, GSE260742. Resolved 2026-09-19 by the Q18b unstated-systems sweep from the study's deposition, not from its text; bounded as a statement about the deposited data rather than about the whole study.
Evidence that argues against this (2)
- An equally documented consequence of H3K79me2 loss in this disease is non-canonical, Polycomb-independent: low-dose pinometostat kills some KMT2A-rearranged leukaemias through transcriptional downregulation of FLT3, and although downregulation of polycomb components accounted for SOME of the proliferation defect, FLT3 was the major pathway. A Polycomb-centred account of the drug's mechanism is therefore incomplete. Richter WF et al. 2021
- The conservation claim is the exposed one. If DOT1L-PRC1.1 antagonism is a general chromatin-memory property of DOT1L-high cells rather than something the fusion directs, then the 'oncogenic DOT1L unit' the menin interface is supposed to command loses its specificity. The same deposition pairs SGC0946 with VTP-50469, so the combination this project contemplates is already present in the data and unexamined. NCBI Gene Expression Omnibus 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
DOT1L and menin inhibitors have complementary activities in MLL-rearranged leukaemia.
Complementary activities of DOT1L and Menin inhibitors in MLL-rearranged leukemia. · Dafflon C, Craig VJ, Mereau H, et al. · Leukemia 2017What this evidence cannot show
Cited from its title and venue via esummary; the abstract was not retrieved for this entry, so this evidence item carries only the existence and direction of the complementarity claim and no experimental detail. Preclinical. An author surname carries a diacritic in the original record that is transliterated here.
Evidence that argues against this (1)
- Complementarity in cell lines has not translated into a combination trial. The clinical development of pinometostat stopped, and the menin inhibitors were developed as single agents and approved that way, so the combination remains a preclinical proposition seven years on. Stein EM et al. 2018
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
KMT2A-rearranged acute lymphoblastic and myeloid leukaemia show distinct responses to menin inhibition and synergy with DOT1L inhibition.
Distinct Responses to Menin Inhibition and Synergy with DOT1L Inhibition in KMT2A-Rearranged Acute Lymphoblastic and Myeloid Leukemia. · Adriaanse FRS, Schneider P, Arentsen-Peters STCJM, et al. · Int J Mol Sci 2024What this evidence cannot show
Cited from its title and venue via esummary; the abstract was not retrieved for this entry, so this item carries the existence of a lineage-distinct response and a synergy claim, not the effect sizes or the models. The title's own framing - distinct responses between ALL and AML - is a warning that a single combination rationale may not hold across lineages.
Evidence that argues against this (1)
- If ALL and AML respond distinctly to menin inhibition, then a mechanism established on one MLL fusion in one lineage cannot be assumed to describe the other, and the PRC1.1 account does not state which lineage it was established in. Neville D et al. 2026
- E-compartment-inverted-in-dlbclmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In germinal-centre B-cell-like diffuse large B-cell lymphoma, DOT1L shapes the composition of chromatin-bound non-canonical PRC1 complexes and regulates, in part, the deposition of H2AK119 monoubiquitination at gene promoters co-occupied by H3K27me3, defined in that study as PRC1/2 targets.
DOT1L Shapes ncPRC1-Target Gene Repression to Maintain Germinal Center B Cell Identity of Diffuse Large B cell Lymphoma. · Gobel C, Niccolai R, Gregoricchio S, de Groot MHP, Kneefel SFK, Kreft M, et al. · Blood 2026What this evidence cannot show
Abstract fetched and read back on 2026-09-20; full text not read. **Moderate: it is GCB-DLBCL, not KMT2A-rearranged leukaemia**, so it replicates the AXIS in a second disease rather than contradicting this one directly. 'Regulates, in part, the deposition' is the paper's own hedge and the abstract does not state a signed effect size. The compartment definition - promoters co-occupied by H3K27me3 - is that study's, and this graph has not established that the same compartment is the relevant one in MLL-fusion leukaemia.
Evidence that argues against this (1)
- A second disease reporting a different direction at a differently-defined compartment is weak evidence that this entry's statement is wrong in KMT2A-rearranged cells; it is strong evidence only that the statement needs a compartment attached. Reading it as a contradiction would overclaim in the opposite direction. Neville D et al. 2026
H3K79me directly antagonises PRC1 activity; its loss licenses H2AK119ub
- related to →Pinometostat (EPZ-5676) — the DOT1L inhibitor, and the most instructive failure in this field
explains why the drug acts progressively rather than on binding
menin inhibition acts through targeted PRC1.1-dependent silencing
the two inhibitors converge on PRC1.1 by different routes
- ← related toNobody has convincingly found the H3K79 eraser - and that constraint sets the clock on every DOT1L inhibitor
KDM2B is both the contested H3K79 eraser and the PRC1.1 targeting subunit
Polycomb was the minor pathway here and the central one in the 2026 account
same source series: its PCGF1-KO and BCOR-KO arms were used here to ask whether the serine collapse needs PRC1.1, and it survives both
this mark is the measurable output of the antagonism that entry describes
the candidate KMT2A-side mechanism if the swap exists there: PRC1.1 arrival at freed fusion-target chromatin
the source of the phrase in this graph; it is used there in the inducible sense
the menin arm of the same axis, separated out because the two drugs move the mark in different geometries
replicates the axis independently in a second disease and reports the opposite sign for the mark
- ← analogous toTrithorax against Polycomb at HOXA: the textbook antagonism, and the two ways this disease breaks it
the PRC1 arm is antagonistic; the PRC2 arm is not - the asymmetry is the finding
names the deposit behind that mechanism entry as the instrument for a question its own paper did not ask
closes five clauses of the entry's caveat, including both inhibitor identities and concentrations