Experiment: put the resistant and the sensitive fusion side by side - if a MEN1 interface mutant keeps the fusion in place while the wild type moves it, relocation is real and menin is what chooses the address
also called M327I discriminator · resistance mutant as an internal control · isogenic relocation test
Some patients relapse because menin changes shape so the drug cannot grip it. Those cells are a perfect built-in comparison: same cells, same drug, one with menin still working. If the fusion moves to new places only where the drug works, the moving is caused by the drug and not by anything else.
The same thing, for a scientist
Perner et al. generated **isogenic MV4;11 (KMT2A::AFF1) lines carrying MEN1 M327I and T349M at the drug interface**, and showed that these mutants retain menin on chromatin under revumenib up to 5 microM while wild-type cells lose it near-completely at 100 nM [E1]. That pair is the ideal internal control for the relocation question and it already exists. **The design.** Profile fusion occupancy (tag or MLL1n plus the ML2-style genetics control) in MEN1-WT and MEN1-M327I MV4;11 across the same revumenib ladder and timepoints, spike-in calibrated, scored for gains as well as losses. **Why this is the strongest available discriminator.** A gained-peak set in a treated sample can arise from many things that have nothing to do with the fusion - differentiation, altered accessibility, background shifts, batch. **The mutant arm controls all of them simultaneously**, because it receives the identical drug, at the identical concentration, for the identical time, in the identical genetic background, and the only thing that differs is whether menin can be displaced. If treated-only fusion peaks appear in wild-type cells and not in M327I cells at the same dose, the peaks are a consequence of menin displacement. If they appear in both, they are a drug effect that does not run through menin - which would be a finding of its own, given that MI-2's principal target has turned out to be lanosterol synthase in at least one case. **The additional prediction that makes it a test rather than a demonstration.** If menin is what selects the used 20% of the uCpG address [E2], then in wild-type cells under drug the fusion's distribution should become **less menin-like and more CFP1-like or KDM2B-like** - that is, it should broaden towards the unused uCpG - whereas in M327I cells it should not change. Correlating the treated fusion profile against menin, CFP1 and KDM2B reference profiles gives a single scalar that moves in one arm and not the other.
Arms
| arm | menin displaced by drug | prediction if relocation is real | prediction if eviction is real |
|---|---|---|---|
| MEN1-WT + DMSO | no | baseline | baseline |
| MEN1-WT + revumenib 100 nM | yes | treated-only peaks at unused uCpG; profile shifts towards CFP1/KDM2B | global loss, no gains |
| MEN1-M327I + revumenib 100 nM | no | unchanged | unchanged |
| MEN1-M327I + revumenib 5 microM | partially | intermediate | intermediate |
| MEN1-WT + MEN1 degron | yes, genetically | same gains as drug | same losses as drug |
The degron arm matters because it separates menin absent from drug present, and the base-editor screen data show the interface mutants are specific to drug binding and preserve the MLL1 interaction 1 - so the degron and the drug should diverge in exactly one respect if the drug has an off-target chromatin effect.
| design | Isogenic MV4;11 MEN1-WT versus MEN1-M327I (and T349M), revumenib dose ladder plus a MEN1-degron arm, matched timepoints, spike-in calibrated fusion and menin profiling |
|---|---|
| readout | Fusion occupancy with gains scored; menin occupancy as the manipulation check; correlation of the treated fusion profile against menin, CFP1 and KDM2B reference profiles as a single scalar relocation index |
| controls | DMSO in both genotypes; degron arm to separate menin absence from drug presence; untagged or C-terminal MLL antibody lane for wild-type MLL1 |
| cost tier | med |
| timeline | 6 months if the isogenic lines can be obtained; add 4 months to re-derive them by base editing |
| falsifying result | treated-only fusion peaks appearing equally in MEN1-M327I cells would show they are not caused by menin displacement |
| shown for fusion | proposed for KMT2A::AFF1 in MV4;11; the isogenic mutant pair exists only in that background at present |
- Using a clinical resistance mutation as a laboratory control normalises it; the same cells that make the experiment clean are the ones that make patients relapse, and any combination strategy derived from this design should be tested against the mutant arm for efficacy as well as for mechanism.
- If treated-only fusion peaks appear in both genotypes, the drug has a chromatin effect that does not run through menin, which would raise a selectivity question about a compound already approved and dosed in children.
- The isogenic pair is menin-mutant, not fusion-mutant. It can say whether menin displacement causes a redistribution; it cannot say whether the fusion's own tethers determine where it lands, which needs CXXC and LEDGF separation-of-function alleles.
- Do MEN1 interface mutant cells still respond transcriptionally at high drug concentrations, and if so through what? A partial menin decrease at 5 microM with no transcriptional change would suggest a threshold effect worth quantifying.
- Is the relocation index - correlation of the fusion profile against menin versus CFP1 and KDM2B - stable enough between replicates to serve as a primary endpoint, or does it need a null distribution from shuffled peak sets?
- Are the patient-derived MEN1 mutant samples from AUGMENT-101 available for chromatin profiling? A human sample with a documented interface mutation would turn this from a cell-line experiment into a clinical one.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Isogenic MV4;11 KMT2A::AFF1 lines carrying MEN1 M327I or T349M at the revumenib-menin interface retain menin on chromatin under revumenib, showing only a partial decrease even at 5 microM, while wild-type cells show near-complete global menin displacement at 100 nM; the mutations attenuate drug binding without disrupting the interaction with MLL1, and the same mutations arise in patients on the AUGMENT-101 trial.
MEN1 mutations mediate clinical resistance to menin inhibition. · Perner F, Stein EM, Wenge DV, et al. · Nature 2023What this evidence cannot show
The mutant lines were made by base editing in MV4;11, an adult KMT2A::AFF1 line, and the published chromatin readouts are menin and MLL1 in TSS windows with no gain analysis. The MLL1 antibody is shared-epitope and MV4;11 retains a wild-type KMT2A allele, so this experiment as proposed needs either a tag or the ML2-style genetics control to attribute signal to the fusion. Whether the mutant lines are available outside the originating laboratories is not established here.
Evidence that argues against this (1)
- MEN1 interface mutations were detected in 12 of 31 patients (38.7%) treated for more than two cycles, including non-responders, and were not detected pre-treatment - so they are a genuine on-treatment selection, but their frequency means the isogenic pair models a common rather than an exotic resistance state, which strengthens the clinical relevance and weakens any claim that it is a special case. Perner F et al. 2023
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin, not unmethylated CpG availability, appears to select which sites the fusion uses: MLL-AF4 occupies only 20% of unmethylated CpG sites while CFP1 and KDM2B occupy 50% and 89%, and MLL(N) signal correlates with menin at r-squared 0.96 at fusion peaks against 0.27 for menin versus CFP1.
MLL-AF4 Spreading Identifies Binding Sites that Are Distinct from Super-Enhancers and that Govern Sensitivity to DOT1L Inhibition in Leukemia. · Kerry J, Godfrey L, Repapi E, et al. · Cell Rep 2017What this evidence cannot show
SEM cells, human infant KMT2A::AFF1. Correlation at a peak set defined by the same antibody, between two complexed proteins, so it demonstrates co-occupancy rather than causal selection. The proposed experiment's profile-shift readout depends on CFP1 and KDM2B reference profiles existing in the test line, which for MV4;11 they do not.
Evidence that argues against this (1)
- The same paper's siRNA data show menin knockdown reduces fusion binding at target genes, which predicts loss rather than broadening. The profile-shift prediction is therefore the relocation hypothesis's distinctive claim and is exactly where it is most likely to fail. Kerry J et al. 2017
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A menin inhibitor phenotype can arise from a target other than menin: MI-2's principal target has been reported as lanosterol synthase with largely menin-independent activity, which is why a drug-versus-degron comparison belongs in this design.
A Menin-MLL Inhibitor Induces Specific Chromatin Changes and Eradicates Disease in Models of MLL-Rearranged Leukemia. · Krivtsov AV, Evans K, Gadrey JY, et al. · Cancer Cell 2019What this evidence cannot show
The lanosterol-synthase finding concerns MI-2, a tool compound, and is recorded elsewhere in this graph; the citation attached here is to the pivotal VTP-50469 paper, which addressed selectivity by showing that MEN1 genetic deletion phenocopies the drug's transcriptional effects - the positive control that makes the degron arm interpretable. Readers wanting the lanosterol-synthase primary source should follow the graph's existing entry rather than this citation.
Evidence that argues against this (1)
- Because the selectivity argument here rests on genetic phenocopy rather than on a direct off-target panel for revumenib, an off-target chromatin effect shared by drug and degron would be invisible to this design. Only a chemical-proteomics selectivity panel on the clinical compound would close that. Perner F et al. 2023
- depends on →MEN1 drug-interface mutations — the first proof that a chromatin drug can drive Darwinian escape in patients
uses the published isogenic M327I and T349M lines as the internal control
dose ladder in wild-type and interface-mutant MEN1 backgrounds
discriminates a drug-caused redistribution from any other treated-sample artefact
the mutations used as the control arose in patients on this trial