Menin inhibition recruits PRC1 to the BENC MYC enhancer cluster and evicts BRG1, exactly as CBFbeta-SMMHC inhibition does in inv(16)
also called the menin BENC swap hypothesis · does menin inhibition do what AI-10-49 does
In one leukaemia, a drug that breaks a fusion protein causes a silencing machine to replace an activating machine on the DNA switches that control MYC. This entry proposes that the menin inhibitors used in this project do the same thing at the same switches, and sets out exactly what result would prove it wrong.
The same thing, for a scientist
Three established facts converge on one untested prediction. **(1)** In KMT2A-rearranged cells the MYC enhancer cluster at +1.7 Mb (BENC/BDME) is occupied by BRG1 and BRD4, Brg1 is required there for transcription-factor occupancy and looping to the Myc promoter, and the cluster is required for MLL-AF9 leukaemia maintenance in mice [E1][E4]. **(2)** In inv(16) AML, disrupting the fusion's obligate partner interface causes RING1B to replace BRG1 at element E3 of that same cluster and at two proximal MYC enhancers, repressing MYC and triggering apoptosis; the exchange is time-ordered, physically verified by ChIP-re-ChIP, and partially reversed by RNF2 knockdown [E2]. **(3)** Menin inhibition induces PRC1.1-dependent H2AK119ub deposition that silences a subset of MLL-fusion targets [E3]. Nobody has joined them. **Hypothesis: within 2-8 hours of menin inhibition in a KMT2A-rearranged cell, RING1B occupancy rises and BRG1 occupancy falls at BENC, H2AK119ub rises there, and MYC transcription falls - and PRC1.1 subunit loss blunts all four.** Prior probability is set at 0.3, deliberately below even odds, because the causal shapes differ: inv(16) releases a repressor, menin inhibition removes an activator, and nothing requires the downstream chromatin event to be the same.
Why this is worth doing even if it is wrong
Both outcomes are informative, which is unusual for a mechanism hypothesis.
If it holds, this project acquires a locus-resolved pharmacodynamic assay for menin inhibitors - a named element, a named factor pair, and a time window - where at present it has none. mark-h2ak119ub records that the reported menin-inhibitor effect is subset-restricted and therefore invisible in bulk; a BENC-anchored CUT&RUN readout would be the first place a menin inhibitor's chromatin action could be measured in a defined locus in a patient sample. It would also predict menin-plus-BET synergy at that element for the same mechanistic reason JQ1 was required in inv(16), which is a specific clinical hypothesis rather than a general combination guess.
If it fails, the failure discriminates between two live models of the menin-inhibitor response. mechanism-myc-after-menin-inhibition-direct-or-differentiation-readout sets out the alternative: MYC falls because the cell differentiates, not because anything happens at the MYC locus. A clean negative at BENC with MYC still falling is strong evidence for the differentiation-readout model, and it would mean that every MYC-based pharmacodynamic biomarker proposed for this disease is measuring a consequence of response rather than the drug's action.
And the null has a boring version that must be excluded first. Cruickshank et al. found SWI/SNF is not required for MYC-regulated gene expression in human KMT2A-rearranged cells 5. If there is no BRG1 at BENC in the chosen human line to begin with, the swap cannot be observed for an uninteresting reason. The baseline BRG1 CUT&RUN is therefore not a control, it is the first experiment.
What would make this hypothesis wrong, and what would make it unfalsifiable
The falsifiers are listed in data.falsifiers. Two of them deserve comment.
A partial result is the likely outcome and must not be read as support. RNF2 knockdown only partially rescued MYC in inv(16) 2; a partial effect in KMT2A-rearranged cells would be entirely consistent with a PRC1-independent mechanism plus noise. The hypothesis is graded on the time-ordering prediction, which is much harder to produce by accident: RING1B up before BRG1 down. A simultaneous change, or BRG1 leaving first, is a vacancy model and refutes the replacement claim even if both factors move.
The unfalsifiability risk is the flavour question. If RING1B arrives and no PCGF1/BCOR/KDM2B arrives with it, the result is a PRC1 result and not a PRC1.1 result, and it does not connect to Neville et al. The experiment must include at least one PRC1.1-defining subunit or it cannot distinguish the hypothesis from canonical PRC1 recruitment - which is the exact limitation that keeps the inv(16) paper from being a PRC1.1 result today.
| statement | Within 2-8 hours of menin-inhibitor treatment of a KMT2A-rearranged leukaemia cell, RING1B occupancy rises and BRG1 occupancy falls at the BENC MYC enhancer cluster, H2AK119ub rises at the same elements, MYC transcription falls, and loss of a PRC1.1-defining subunit (PCGF1, BCOR or KDM2B) blunts all four. |
|---|---|
| origin | analogy |
| prior probability | 0.3 |
| prior reasoning | Set below even odds because the causal shapes differ. In inv(16) the drug RELEASES a repressor (RUNX1) which actively recruits RING1B; in KMT2A-rearranged disease the drug REMOVES an activator and no released repressor is named. PRC1.1 arrival at freed chromatin is a plausible but different mechanism, and Neville et al. report the menin-inhibitor PRC1.1 effect as subset-restricted without naming MYC. |
| cheapest discriminating measurement | a BRG1 CUT&RUN at BENC in an untreated human KMT2A-rearranged line; it costs one experiment and determines whether the whole hypothesis is testable |
predictions
[ "BRG1 is detectable at BENC in an untreated human KMT2A-rearranged line (if not, the hypothesis is untestable in that line and a different model is required)", "RING1B occupancy at BENC rises before BRG1 occupancy falls - a replacement, not a vacancy", "H2AK119ub rises at BENC modules and not at a non-myeloid control enhancer such as N-ME", "MYC transcript falls with kinetics compatible with the occupancy change, not before it", "PCGF1, BCOR or KDM2B knockout blunts the H2AK119ub gain and the MYC fall - which is the locus-level version of the published whole-cell resistance epistasis", "adding a BET inhibitor deepens MYC repression at BENC, mirroring the JQ1 requirement in inv(16)" ]
experiments
[ "exp-brg1-ring1b-timecourse-at-benc-under-menin-inhibition" ]
analyses
[ "ana-prc1-myc-enumeration-and-bounded-absences" ]
falsifiers
[ "RING1B occupancy at BENC does not change at any timepoint from 1 to 24 hours after menin inhibition, while a positive control locus (a known menin-inhibitor-silenced target) shows the expected H2AK119ub gain - the instrument works and the element does not respond", "BRG1 occupancy falls BEFORE or simultaneously with RING1B arrival - a vacancy model, which refutes the replacement claim specifically", "MYC transcript falls with normal kinetics while neither BRG1 nor RING1B occupancy at BENC changes - supports the differentiation-readout model instead", "PCGF1, BCOR and KDM2B knockouts leave the BENC chromatin response intact - the event is PRC1 but not PRC1.1, and does not connect to the menin/DOT1L axis", "no BRG1 is detectable at BENC in the untreated human line - the premise fails and the experiment cannot be run there" ]
- If the hypothesis holds, PRC1.1 loss becomes an enhancer-local resistance mechanism as well as a global one, and enhancer-local resistance is harder to detect in any bulk pharmacodynamic assay - the finding would make resistance monitoring harder, not easier.
- A BENC-anchored pharmacodynamic assay would go quiet in exactly the situation where MYC is being sustained from an alternative enhancer, producing a false reassurance at the moment resistance emerges.
- The hypothesis, if confirmed, argues for adding a BET inhibitor to a menin inhibitor. BET inhibitors have their own documented resistance route in this same model through WNT-driven MYC restoration, so the proposed combination inherits a known escape mechanism.
- If the hypothesis is refuted and MYC falls anyway, every MYC-based pharmacodynamic biomarker for this disease is measuring differentiation rather than drug action, which would invalidate a class of proposed clinical assays rather than just this one.
- Is there a released repressor in the menin axis at all? If the mechanism is PRC1.1 arrival at vacated chromatin rather than active recruitment by a transcription factor, the predicted time-ordering may be reversed and the hypothesis should be restated rather than discarded.
- Which KMT2A-rearranged line has BRG1 at BENC? Nobody has published the baseline, and the answer selects the model system.
- Should the experiment be run in a B-lineage KMT2A::AFF1 model even though every supporting result is myeloid? The project's index disease is lymphoid and the hypothesis has no lymphoid support at all.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The MYC enhancer cluster 1.7 Mb downstream is occupied by SWI/SNF and BRD4 in MLL-AF9 leukaemia cells, and Brg1 is required there to maintain transcription-factor occupancy and long-range looping to the Myc promoter.
Role of SWI/SNF in acute leukemia maintenance and enhancer-mediated Myc regulation. · Shi J, Whyte WA, Zepeda-Mendoza CJ, Milazzo JP, Shen C, Roe JS, et al. · Genes Dev 2013What this evidence cannot show
Mouse MLL-AF9 in the RN2 model, which carries a cooperating NrasG12D allele. The MYC dependency conclusion is contested in human cells; the occupancy and looping observations are the part this hypothesis leans on.
Evidence that argues against this (1)
- Cruickshank et al. found the same SWI/SNF subunits not required for c-MYC-regulated gene expression in human MLL-rearranged cells, explicitly contrasting with the mouse result. If BRG1 does not hold MYC in the human system, the baseline this hypothesis assumes may not exist there. Cruickshank VA et al. 2015
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In inv(16) AML, interface disruption causes RING1B to replace BRG1 at three MYC distal enhancers including element E3 of the +1.7 Mb cluster, with RUNX1 and RING1B occupancy rising from about 2.5 hours and BRG1 falling only at 4-6 hours, and MYC partially rescued by RNF2 knockdown or PRT4165.
CBFbeta-SMMHC Inhibition Triggers Apoptosis by Disrupting MYC Chromatin Dynamics in Acute Myeloid Leukemia · Pulikkan JA, Hegde M, Ahmad HM, Belaghzal H, Illendula A, Yu J, O'Hagan K, Ou J, Muller-Tidow C, Wolfe SA, Zhu LJ, Dekker J, Bushweller JH, Castilla LH · Cell 2018What this evidence cannot show
A different fusion family, a different causal shape (repressor released rather than activator removed), and largely one cell line. The repressive mark measured is H3K27me3, not H2AK119ub, so the PRC1 flavour is unidentified. Read in full from PMC6211564 on 2026-09-20.
Evidence that argues against this (1)
- The mechanism was complete and the single agent insufficient - clearing the disease required JQ1. A mechanistic analogy imported from this paper should carry the combination requirement with it rather than only the chromatin result. Rathert P et al. 2015
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Non-canonical PRC1.1 mediates the cellular response to menin and DOT1L inhibitors; menin inhibition induces PRC1.1-dependent deposition of H2AK119ub that silences a subset of MLL-fusion-protein targets.
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 silenced subset is not enumerated and MYC is not mentioned, so this supplies the mechanism's existence and not its address. Whether the subset includes MYC or any MYC enhancer is the open question this hypothesis exists to close.
Evidence that argues against this (1)
- The strongest menin-inhibitor MYC result on record runs through PRC1.1 in the opposite direction: PRC1.1 LOSS reactivates MYC as a resistance route. That places MYC among the PRC1.1-controlled non-canonical menin targets by genetics, but it is a whole-cell epistasis result with no locus resolution, so it neither supports nor excludes an enhancer-level swap. Zhou X et al. 2024
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
BENC is essential for Myc expression in haematopoietic stem cells and progenitors and is essential for maintenance of MLL-AF9-driven leukaemia in mice.
A Myc enhancer cluster regulates normal and leukaemic haematopoietic stem cell hierarchies. · Bahr C, von Paleske L, Uslu VV, Remeseiro S, Takayama N, Ng SW, et al. · Nature 2018What this evidence cannot show
Mouse genetics; deletion, not drug. Establishes the element is load-bearing in an MLL fusion context, which is what makes it the right address to interrogate, and says nothing about what a menin inhibitor does there.
Evidence that argues against this (1)
- Enhancer requirement measured by germline deletion in the untreated state does not predict behaviour under drug: in the same MLL-AF9;NrasG12D background, chromatin-drug pressure re-routes MYC transcription to an alternative focal enhancer recruiting WNT machinery. A BENC-anchored assay could therefore go quiet while MYC is sustained elsewhere. Rathert P et al. 2015
- E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In human MLL-rearranged cells, SMARCA4, SMARCD2 and DPF2 are required for expression of haematopoietic-stem-cell-associated genes but, in contrast to previous mouse results, are not required for expression of c-MYC-regulated genes.
SWI/SNF Subunits SMARCA4, SMARCD2 and DPF2 Collaborate in MLL-Rearranged Leukaemia Maintenance. · Cruickshank VA, Sroczynska P, Sankar A, Miyagi S, Rundsten CF, Johansen JV, et al. · PLoS One 2015What this evidence cannot show
The readout is a MYC-regulated gene signature, not MYC transcript, MYC protein or BRG1 occupancy at any enhancer. The paper does not report an enhancer measurement at all, so it does not directly test whether BRG1 sits at BENC in human cells - which is why the baseline occupancy check is the first step of the proposed experiment rather than an assumption.
Evidence that argues against this (1)
- Shi et al. report direct Brg1 occupancy at the element and a requirement for looping to the Myc promoter, which is a physical measurement rather than a signature inference. A signature-level negative and an occupancy-level positive are not the same measurement and can both be correct. Shi J et al. 2013
generated by transplanting the inv(16) chromatin experiment onto the menin axis
the element the hypothesis addresses
the intervention whose chromatin action is being predicted
gated time-course CUT&RUN; the discriminating quantity is the order of the two occupancy changes
- related to →Is MYC the node? Two papers, the same MLL-AF9 model, opposite answers - and the second one failed to rescue
arm 0 of the test measures the BRG1 occupancy neither Shi nor Cruickshank measured in a human line