A menin inhibitor does not only take something away - it causes PRC1.1 to put something on, and the drug needs the cell's own silencing machinery to work
also called menin inhibition requires PRC1.1 · active silencing versus passive eviction · the menin arm of PRC1.1 antagonism
The standard story is that a menin inhibitor works by prising the leukaemia's fusion protein off the DNA, after which the genes it was switching on simply fall silent. The newer picture is that silence is not automatic: the cell has to actively bolt the genes shut using a machine called PRC1.1, and if that machine is broken the drug does much less.
The same thing, for a scientist
This graph's mechanistic model of menin inhibition has been **eviction-based**: disrupt the MBM interface, displace the fusion from chromatin, lose the elongation machinery, lose the target programme. Two papers converge on a second, active step. Neville et al. state that **menin inhibition induces PRC1.1-dependent deposition of H2AK119ub to silence a subset of MLL-fusion targets** [E1]. Zhou et al. show independently that **loss of PRC1.1 subunits confers menin-inhibitor resistance**, with non-canonical menin targets reactivating [E2]. Read together: the drug's transcriptional output is partly **executed by the host cell's Polycomb machinery**, not produced directly by the loss of the fusion. Three consequences. **(1) Efficacy is contingent on an intact host complex**, so PRC1.1 status is a candidate predictive biomarker for a drug now approved and in wide use. **(2) Anything that lowers PRC1.1 function pharmacologically is a candidate antagonist** of menin inhibition - see `gene-usp7`. **(3) The DOT1L comparison sharpens**: Neville et al. report menin inhibition producing a subset-restricted H2AK119ub gain and DOT1L inhibition a genome-wide one [E1], so the two drugs are not two routes to one endpoint. Critically, the PRC1.1 step is reported as **independent of MLL-fusion displacement**, which is the clause that makes this a second mechanism rather than a description of the first.
Eviction and deposition are different claims with different predictions
| eviction model | deposition model | |
|---|---|---|
| what the drug does | removes the fusion from chromatin | removes the fusion and triggers PRC1.1 to mark the site |
| who does the silencing | nobody - transcription stops for lack of an activator | the host cell's PRC1.1 |
| effect of PRC1.1 loss | none | resistance |
| durability of silencing | until the fusion returns | a deposited mark can persist beyond drug exposure |
| predictive biomarker | fusion occupancy | PRC1.1 integrity |
The deposition model's distinctive prediction - that PRC1.1 loss produces resistance - is the one that has been measured, and it holds 2. That is the strongest available evidence for the model and it is genetic rather than chromatin-level: Zhou et al. did not show H2AK119ub arriving after drug, they showed that removing the machine that deposits it blunts the response.
The two papers agree on the dependency and differ on the timing of the mark. Zhou et al. describe the non-canonical targets as already 'bivalently co-occupied by active menin and repressive H2AK119ub' before treatment; Neville et al. describe menin inhibition as inducing the deposition. Pre-existing versus drug-deposited is not a detail: if the mark is pre-existing, a baseline H2AK119ub map could predict response, and if it is drug-induced, only an on-treatment map can. That distinction is recorded rather than resolved here.
What this does to the project's menin-interface hypothesis
It does not threaten the interface hypothesis - the interface is still the drug's target and the fusion is still displaced - but it inserts a step the project has not been modelling, and the step is in the patient's cells rather than in the drug's chemistry.
The practical translation is that a patient who fails a menin inhibitor may have a perfectly engaged target. resistance-myc-bypass-prc1 already records PRC1.1 loss as a named resistance route and notes that it 'has never been looked for in a patient, which is an immediately fixable gap'. This entry states why that gap is more urgent than a typical unlooked-for mechanism: the mechanism is not a bypass around the drug, it is a failure of the drug's own effector arm, so it should be invisible to every target-engagement assay - pharmacokinetics, menin occupancy, fusion displacement - that a trial would run to explain a non-response.
The cheapest available test is not a new experiment. BCOR and BCORL1 are recurrently mutated in myeloid malignancy and are on standard sequencing panels; PCGF1 and KDM2B are not. A retrospective look at BCOR status against menin-inhibitor response in the revumenib and ziftomenib trial populations would be an analysis of data that already exists.
| mechanism class | drug mechanism, second step |
|---|---|
| distinctive prediction of deposition model | PRC1.1 loss confers resistance - measured and confirmed (PMID 39158067) |
| unresolved between the two sources | whether H2AK119ub at the relevant loci is pre-existing (Zhou) or drug-induced (Neville); baseline versus on-treatment biomarker strategies follow from the two answers |
| biomarker implication | PRC1.1 integrity is a candidate PREDICTIVE biomarker for menin inhibitors; BCOR and BCORL1 are already on routine myeloid sequencing panels, PCGF1 and KDM2B are not |
| immediately available analysis | retrospective BCOR/BCORL1 mutation status against menin-inhibitor response in existing revumenib and ziftomenib trial populations - data that already exist |
| why this resistance is invisible to standard assays | it is a failure of the drug's effector arm, not a bypass around the drug, so pharmacokinetics, menin occupancy and fusion-displacement assays all read as normal in a non-responder |
| contrast with dot1l arm | menin inhibition: subset-restricted H2AK119ub gain. DOT1L inhibition: genome-wide gain. Not two routes to one endpoint. |
models compared
{
"eviction": "MBM interface disrupted -> fusion displaced -> target programme lapses for want of an activator",
"deposition": "MBM interface disrupted -> fusion displaced -> PRC1.1 deposits H2AK119ub at a subset of targets -> active silencing"
}affects
[ "drug-revumenib", "drug-ziftomenib", "drug-bleximenib", "drug-enzomenib" ]
- The drug depends on machinery in the patient's own cells, which means efficacy can fail with the target fully engaged, and every standard target-engagement assay would report success in that patient.
- PRC1.1 loss as a resistance route is genetic and could be selected for during treatment, so a mechanism that is rare at diagnosis may be common at relapse - and nobody is sequencing PCGF1 or KDM2B.
- Any co-administered drug that reduces PRC1.1 function - a USP7 inhibitor is the concrete case - is predicted to antagonise the menin inhibitor. This is a combination that would look rational from either drug's own literature.
- If the silencing is actively deposited by PRC1.1, then the same complex is depositing repressive marks genome-wide in normal haematopoietic cells under DOT1L inhibition, which is an untested toxicity mechanism for the combination the field is enthusiastic about.
- Is H2AK119ub at the silenced subset pre-existing or drug-deposited? The two published accounts differ and the answer decides whether a baseline biomarker is possible.
- Has anyone looked at BCOR or BCORL1 mutation status against menin-inhibitor response in the trial populations? The data exist and the analysis has not been reported.
- Does PRC1.1 loss blunt the menin-inhibitor response in a B-lineage KMT2A::AFF1 model? The mechanism is entirely myeloid, and infant KMT2A-rearranged ALL has an exceptionally low mutation burden, so the route may not be available there.
- If silencing is actively deposited, does it persist after drug withdrawal? A deposited mark could outlast exposure, which would be an argument for intermittent dosing rather than against it - the opposite of what the transcriptional-memory framing suggests.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin inhibition induces PRC1.1-dependent deposition of H2AK119ub to silence a subset of MLL-fusion-protein targets, whereas DOT1L inhibition produces a genome-wide increase in H2AK119ub; the enhanced PRC1.1 activity arises specifically from progressive loss of DOT1L-mediated H3K79 methylation, independent of MLL-fusion displacement or transcriptional repression.
DOT1L provides transcriptional memory through PRC1.1 antagonism. · Neville D, Ferguson DT, Heikamp EB, Lai Z, Magor GW, Lam C, et al. · Nat Cell Biol 2026What this evidence cannot show
Abstract-level; the full text was not reachable through the routes tried on 2026-09-20. Note a subtlety in the abstract's structure: the 'independent of MLL-FP displacement' clause is attached to the DOT1L arm, where enhanced PRC1.1 activity follows H3K79me loss. Whether the menin arm's PRC1.1 deposition is likewise displacement-independent is not stated and is not asserted here. The silenced subset is not enumerated, so the loci are not nameable.
Evidence that argues against this (1)
- Zhou et al. describe H2AK119ub at the non-canonical menin targets as a pre-existing bivalent feature co-occupying chromatin with active menin, and report those targets as downregulated after menin inhibition - a model in which PRC1.1 is already in place and the drug removes the opposing activity, rather than one in which the drug recruits PRC1.1. Both are compatible with PRC1.1 dependence and they predict different biomarker strategies. Zhou X et al. 2024
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Chromatin-focused CRISPR screens plus genetic, epigenetic and pharmacological studies in human and mouse KMT2A-rearranged AML models show that loss of PRC1.1 subunits such as PCGF1 or BCOR leads to menin-inhibitor resistance by epigenetic reactivation of non-canonical menin targets including MYC, and that MYC inhibition resensitises PRC1.1-deficient cells.
Epigenetic regulation of noncanonical menin targets modulates menin inhibitor response in acute myeloid leukemia. · Zhou X, Zhang L, Aryal S, Veasey V, Tajik A, Restelli C, et al. · Blood 2024What this evidence cannot show
Cell-line and mouse models; no patient cohort. Subunit knockouts are chronic, so the resistance phenotype could include adaptation rather than the acute loss of an effector arm. PCGF1 and KDM2B are not on routine clinical sequencing panels, so the proposed retrospective analysis is limited to BCOR and BCORL1.
Evidence that argues against this (1)
- PRC1 activity and H2AK119ub deposition are separately shown to be REQUIRED for the leukaemogenic potential of MLL fusions in mice, which predicts that PRC1.1 loss should impair the leukaemia rather than rescue it from a drug. Both results are on record in overlapping systems and nobody has reconciled the sign. Rossi A et al. 2016
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A menin inhibitor combined with a DOT1L inhibitor markedly enhances differentiation and cell killing across MLL disease models including primary cells while sparing normal haematopoiesis, with MLL-fusion target genes and MYC targets suppressed more profoundly under the combination.
Complementary activities of DOT1L and Menin inhibitors in MLL-rearranged leukemia. · Dafflon C, Craig VJ, Mereau H, Grasel J, Schacher Engstler B, Hoffman G, et al. · Leukemia 2017What this evidence cannot show
Predates the PRC1.1 mechanism by nine years and interprets the synergy as additive targeting of nodes in the MLL complex, not as two differently shaped Polycomb responses. Cited here as the empirical observation the newer mechanism reinterprets, not as support for the mechanism itself.
Evidence that argues against this (1)
- If the two drugs act through the same PRC1.1 effector arm, their combination should be limited by the capacity of that arm rather than being freely additive - a ceiling the synergy data do not obviously show. The complementarity may therefore have a component that is not Polycomb-mediated at all. Neville D et al. 2026
unresolved disagreement: that entry reports target-locus silencing under menin inhibitor, this finds no HOXA/MEIS1 suppression -- different compounds, lines and timepoints
the drug's transcriptional output is partly executed by the host cell's PRC1.1
efficacy contingent on an intact host complex; PRC1.1 integrity is a candidate predictive biomarker
subset-restricted deposition after menin inhibition, against a genome-wide gain after DOT1L inhibition
- related to →BCOR — the other side of the CBX8 conflict, and a target you would be inhibiting in the direction of disease
BCOR and BCORL1 are on routine myeloid panels, so this biomarker is retrospectively testable today
reframes that resistance route as a failure of the drug's own effector arm rather than a bypass around it
the menin arm of the same axis, separated out because the two drugs move the mark in different geometries
the whole published PRC1.1 x menin-inhibitor symbol intersection is one paper
THE DISAGREEMENT AT ISSUE: whether the H2AK119ub mark is pre-existing or drug-deposited, which the locus intersection was meant to decide.
if silencing after eviction is PRC1.1 deposition, this arm says the H3K79me2 half of the same drug's chromatin action does not need PRC1.1 - the two halves are separable
confirms the published ordering that entry records - PRC1.1 H2AK119ub arising FROM progressive H3K79me2 loss, so the mark is upstream - because a downstream PCGF1 knockout leaves the mark's drug response unchanged at 0.0166
the mark itself, graded by eviction depth and reduced in a PCGF1 knockout
the mark-level measurement that mechanism asserted from occupancy alone
the active silencing step this reading needs is already published
the H2AK119ub gain is the third of four documented arrivals
the drug's output is partly deposited by host PRC1.1, which the chain does not represent
sets the catalytic output of the complex that deposits H2AK119ub after menin inhibition; untested in that system
the effector arm of the proposed single activity
THE PARTITION IS MEANINGFUL: the H2AK119ub-only group returns the canonical Polycomb developmental signature, unsought.
THE DISAGREEMENT THE JOIN CANNOT SETTLE ALONE: Zhou's table has no condition column, so a static partition cannot distinguish pre-existing marks from drug-deposited ones.