The menin inhibitor removes three-quarters of the fusion from its own promoters, replicated in two batches
The clinical strategy assumes that blocking the menin-MLL handshake knocks the cancer fusion protein off the genes it controls. In this dataset it does: four days of a menin inhibitor leaves about a quarter of the fusion at its target genes, and the same result appears in two separately sequenced experiments.
The same thing, for a scientist
The most direct test of the menin-KMT2A interface hypothesis available inside this deposit, and it is not the experiment the deposit is usually cited for. Median HA (MLL::AF9) TSS signal ratio after 4 days of **VTP-50469 100 nM** at the authors' published MLL::AF9 target genes is **0.254** (batch 1912x2, n=201 target genes) and **0.270** (batch 200908, n=200) - a 73-75% displacement of the fusion from its own promoters, **replicated across two independent sequencing batches**, which is more replication than any other quantity in this beat. At non-target genes the same drug gives 0.432 and 0.438, so the displacement is preferential but far from exclusive. For comparison in the same tables, the DOT1L inhibitor EPZ-5676 250 nM gives 0.362 / 0.439 at target genes and the dTAG degron gives 0.015 / 0.050.
Why this matters for the project's central hypothesis
The menin-interface hypothesis is that menin inhibitors work by displacing menin and thereby disabling the fusion. This measurement is the middle link of that chain - drug in, fusion off chromatin - measured genome-wide at gene resolution, in the same cells where the degron arm defines what complete removal looks like. It holds. The published clinical-resistance work makes the same link from the other end: MEN1 interface mutations confer clinical resistance to revumenib and act by preventing inhibitor-induced eviction of menin and MLL1 from chromatin 3.
The part that qualifies it
73-75% is not 99%. The degron arm in the same table removes 95-98.5%. So the menin inhibitor at this dose and duration is a partial displacement of the fusion, which is consistent with the reciprocal finding in this beat that about half the menin at fusion promoters does not depend on the fusion. The two observations describe one incompletely coupled interface, not two contradictions.
Scale and limits
No spike-in exists anywhere in GSE173599, so every number is a within-sample share, and the treated arms are scaled by a 21-gene housekeeping panel whose own raw 3 h H3K27ac ratio averages 0.809 and is forced to 1. A housekeeping normalisation DEFINES the global change to be zero, so only the DIFFERENCE between strata is measurable; a uniform genome-wide loss of any of these proteins would be invisible here. n = 1 per arm for every channel except HA, which has two independent 3 h batches. The cells are HCB1, transformed by KMT2A::MLLT3 (MLL::AF9) - NOT the index fusion of this project, KMT2A::AFF1 - and a prior beat in this corpus found that per-gene ordering transfers between MLLT3 and AFF1 lines while magnitude does not. These are 4-day arms, so secondary and differentiation-driven changes are not excluded, unlike the 3 h degron arm. VTP-50469 is a tool compound; revumenib and ziftomenib are the clinical agents and are not tested here.
| script | analyses/menin-retention-by-ha-occupancy/analyze_controls.py |
|---|---|
| status | run |
| result summary | VTP-50469 4 d: HA at target genes 0.254 (batch 1912x2) and 0.270 (batch 200908); at non-target genes 0.432 and 0.438. EPZ-5676: 0.362 / 0.439 at targets. dTAG-VHL: 0.015 / 0.050. |
| peak width ratio note | Signal instrument; no peak-width ratio applies. The corresponding HA peak-based arms all exceed the 1.3 width trigger and are deliberately not used here. |
inputs
[ "GSE173599_TSS_signal_Figure4.xlsx", "GSE173599_TSS_signal_Figure5.xlsx" ]
- A drug that leaves a quarter of the fusion on its own promoters after four days leaves a reservoir from which the programme can be re-established, which is a mechanistic route to the incomplete and non-durable responses seen clinically.
- MEN1 interface mutations abolish the eviction entirely and arise under treatment pressure, so this endpoint is directly convertible into a resistance assay - and directly vulnerable to resistance.
- The same interface is used by wild-type MLL1, so displacement is not fusion-selective and the therapeutic window comes from dependency, not from selectivity of binding.
- Does revumenib, the approved agent, produce the same 73-75% displacement at the same gene set, or is this a property of VTP-50469's potency and exposure?
- Is the residual 25-27% of fusion occupancy the pool tethered by CXXC-DNA binding rather than by menin, and does it predict the depth of clinical response?
- Nothing here is measured in KMT2A::AFF1, the fusion this project is about.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
After 4 days of VTP-50469, median HA TSS signal at the authors' MLL::AF9 target genes falls to 0.254 of DMSO in batch 1912x2 and 0.270 in batch 200908.
GSE173599_TSS_signal_Figure4.xlsx - per-gene HA, DOT1L and MENIN TSS signal after 4 days of EPZ-5676, VTP-50469 or dTAG-VHL (batch 1912x2) · Olsen SN et al. · NCBI GEO supplementary file 2022What this evidence cannot show
Floored and housekeeping-normalised; n=1 per arm within each batch, but the two batches are independent experiments and agree to within 1.6 points. 4-day treatment admits secondary effects.
Evidence that argues against this (1)
- A 4-day treatment that induces differentiation changes the cell population, so reduced fusion occupancy could partly reflect a different cell state rather than direct displacement; the 3 h degron arm has no such confound but is not a drug arm. Olsen SN et al. 2022
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
The same measurement in the second batch, which also carries the combination arm, gives 0.270 at target genes and 0.438 at other genes.
GSE173599_TSS_signal_Figure5.xlsx - per-gene HA and H3K27ac TSS signal after 4 days of EPZ-5676, VTP-50469, the combination or dTAG-VHL (batch 200908) · Olsen SN et al. · NCBI GEO supplementary file 2022What this evidence cannot show
618 genes in this sheet after the authors' floor, of which 200 are target genes; the 'other' comparator here is the non-target genes in the sheet, not the genome.
Evidence that argues against this (1)
- The 618-gene sheet in this batch is heavily floored, so the non-target comparator is a set of genes with substantial HA signal that are nonetheless not on the target list - a narrow and possibly atypical comparator. Olsen SN et al. 2022
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MEN1 mutations at the drug-menin interface confer clinical resistance to revumenib and act by preventing inhibitor-induced eviction of menin and MLL1 from chromatin.
MEN1 mutations mediate clinical resistance to menin inhibition · Perner F, Stein EM, Wenge DV, Singh S, Kim J, Apazidis A, et al., Levine RL, Armstrong SA, Cai SF · Nature 615(7954):913-919 2023What this evidence cannot show
Clinical samples plus xenograft and base-editor models; the mechanism claim is about menin and MLL1 eviction, and MLL1 here includes the fusion. Human, KMT2A-rearranged and NPM1-mutant leukaemia.
Evidence that argues against this (1)
- That resistance mechanism shows the drug's chromatin effect is necessary for response, but not that displacing the fusion is sufficient - responses are partial and relapse is common. Perner F et al. 2023
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
VTP-50469 was developed as a selective menin-MLL1 inhibitor and was reported to displace menin from protein complexes and reduce chromatin occupancy of MLL at select genes.
A Menin-MLL Inhibitor Induces Specific Chromatin Changes and Eradicates Disease in Models of MLL-Rearranged Leukemia · Krivtsov AV, Evans K, Gadrey JY, Eschle BK, Hatton C, Uckelmann HJ, et al., Armstrong SA · Cancer Cell 36(6):660-673.e11 2019What this evidence cannot show
The original report's occupancy claim is at 'select genes' and is the founding observation this analysis extends to gene resolution genome-wide.
Evidence that argues against this (1)
- The founding report's occupancy claim is for MLL at 'select genes', not genome-wide and not for the fusion specifically, so the quantitative extension made here goes beyond what that paper established. Krivtsov AV et al. 2019
blocking the menin-MLL1 interface removes 73-75% of the fusion from its own promoters at 4 d, in two independent batches
- related to →KMT2A::MLLT3 (MLL-AF9), t(9;11)(p21.3;q23.3)
VTP-50469 4 d leaves 25-27% of fusion occupancy at target genes
the clinical agent's resistance mutations act by preventing the same eviction of menin and MLL1 from chromatin
the reciprocal half of the same interface: menin half-leaves when the fusion goes, and the fusion three-quarters-leaves when menin is blocked
class read-across: the interface-blocking mechanism measured here with the tool compound VTP-50469