Eviction is a claim about menin. The field, and this graph, transfer it to the fusion without a measurement that supports the transfer
also called evicting the fusion from chromatin · the eviction vocabulary · menin eviction versus fusion eviction
Papers and reviews say menin inhibitors evict the leukaemia fusion protein from DNA. What has actually been measured, over and over, is that they evict menin. The fusion's own fate has been checked far less often, and when it was checked it mostly stayed put.
The same thing, for a scientist
Three claims are routinely compressed into one sentence, and they have very different evidence behind them. **(1) The drug evicts menin.** Directly measured, genome-wide, repeatedly, in several human lines and by two laboratories: near-complete global loss of menin chromatin occupancy at 100 nM revumenib and at 330 nM VTP-50469 [E1][E2]. This is as solid as chromatin pharmacology gets. **(2) The drug evicts MLL1.** Measured, and the measurement is scoped: MLL1 is lost from promoters where menin displacement is greatest [E1], and it is explicitly NOT globally displaced [E2]. **(3) The drug evicts the fusion.** The MLL1 antibody used in (2) recognises wild-type MLL1 and the fusion equally; only one published experiment isolates the fusion by genetics (the ML2 line), and there the pattern matched the lines that retain wild-type MLL1 [E2]. The only *fusion* protein directly reported as evicted is a NUP98 fusion, which has no menin motif of its own [E3]. **So (1) is established, (2) is scoped, and (3) rests on an antibody that cannot see the difference.** The compression is not malicious and it is not confined to reviews - Perner's own abstract says *eviction of menin and MLL1 from chromatin* while the body scopes MLL1 to the most menin-depleted promoters [E1]. This graph inherited the compression: `hyp-apparent-authority-residual-fusion-occupancy` opens *menin inhibitors act by evicting menin and MLL1 from chromatin*, and `localization-menin-inhibition-fusion-fate` is titled *where does the fusion go when menin inhibition evicts it*. Both are defensible readings of the literature's phrasing and both should now say **menin** where they say the fusion, or say that the fusion's fate is the open question.
A vocabulary that decides the experiment
The word chosen determines what gets measured next.
| word | what it asserts | what would test it |
|---|---|---|
| evicted | the protein has left this locus | occupancy at that locus, before and after |
| displaced | it has left AND is somewhere else | occupancy genome-wide, gains as well as losses |
| degraded | it is gone from the cell | protein level, half-life, proteasome rescue |
| inactivated | it is still there and no longer works | occupancy plus nascent transcription plus complex composition |
The published readouts answer column two for menin and, partially, for MLL1. Nothing in the class answers the other three for the fusion, and the differences are not semantic: they predict different resistance mechanisms, different combination partners and a different meaning for 'response'.
Where the graph should be corrected
These are reports, not edits (this beat does not own the files):
hyp-apparent-authority-residual-fusion-occupancy,summary_expertfirst sentence, Menin inhibitors act by evicting menin and MLL1 from chromatin, and resistance mutants work by preventing that eviction. The second clause is exact. The first should read evicting menin from chromatin, and MLL1 from the promoters where menin loss is greatest, per the body of PMID 36922589, which I read in full.- The same entry's core hypothesis - a defined subset of fusion binding sites survives saturating menin inhibition - is largely already answered in the direction it predicts, and more strongly: PMID 31821784 reports that MLL1n is not globally displaced and names the small loss set (MEIS1, MEF2C, JMJD1C) against retained HOXA, MYB and GAPDH. The hypothesis should be upgraded to cite that, and re-pointed at the residual fraction and at whether the residual sites are transcriptionally competent.
localization-menin-inhibition-fusion-fate, title, where does the fusion go when menin inhibition evicts it. Presupposes the eviction. Suggest when menin inhibition removes menin.
Each of these is a wording change with a primary source attached, not a dispute about biology.
| shown for fusion | KMT2A::AFDN (ML2) only; the phrase 'evicts the fusion' has never been demonstrated for KMT2A::AFF1 |
|---|
claims compressed
[ "menin is evicted - established genome-wide", "MLL1 is lost at the most menin-depleted promoters - measured and scoped", "the fusion is evicted - not established; one genetics-isolated line says largely retained" ]
graph entries to correct
[ "hyp-apparent-authority-residual-fusion-occupancy", "localization-menin-inhibition-fusion-fate" ]
- The vocabulary is load-bearing for trial interpretation: a correlative endpoint described as 'fusion eviction' but measured as menin displacement will saturate early and lose the ability to discriminate responders, which matters in paediatric dose-finding where every correlative sample is precious.
- If clinicians and reviewers believe the fusion has been removed, a relapse with retained fusion occupancy looks like re-expression rather than reactivation, and the wrong resistance mechanism gets targeted.
- Correcting the vocabulary weakens nothing about the drug's efficacy, which is measured in patients, but it does weaken the mechanistic story used to justify combination choices.
- Does any primary paper in the class state, in its results rather than its abstract or discussion, that a KMT2A fusion protein is evicted from chromatin genome-wide? I found none in the fifteen full texts read for this beat, and a systematic sentence-level audit of the class would settle it.
- How far has the compressed sentence travelled? A count of reviews in the 61-record menin-inhibitor chromatin set that assert fusion eviction while citing only menin-occupancy data would quantify the drift.
- If the field standardised on 'menin eviction' as the pharmacodynamic claim, what would it then use as the efficacy-linked chromatin endpoint, given that menin displacement is already near-complete at 100 nM and cannot grade depth of response?
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The abstract of the clinical-resistance paper states that MEN1 interface mutations prevent inhibitor-induced eviction of menin and MLL1 from chromatin, while the body of the same paper reports near-complete global displacement of menin at 100 nM revumenib and scopes MLL1 loss to promoters with the highest degree of menin displacement (over 80%).
MEN1 mutations mediate clinical resistance to menin inhibition. · Perner F, Stein EM, Wenge DV, et al. · Nature 2023What this evidence cannot show
This is a claim about what the paper says, verified by reading the full text of PMC10157896, not an independent measurement. The abstract wording is a fair summary of a headline and is not an error by the authors; the point is that the abstract sentence is what propagates. The MLL1 antibody is shared-epitope, and MV4;11 retains a wild-type KMT2A allele, so 'MLL1' in that paper is not the fusion.
Evidence that argues against this (1)
- A reader could reasonably argue the compression is harmless because menin and the fusion leave together at the loci that matter clinically. That is exactly the claim the retained HOXA and MYB occupancy in PMID 31821784 contradicts, and HOXA is not a locus anyone would call clinically unimportant in this disease. Krivtsov AV et al. 2019
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
The founding pharmacology paper states that VTP-50469 did not globally displace MLL1 from chromatin in MOLM13, RS4;11 or ML2 cells, and that in ML2 - which has lost the wild-type KMT2A allele, so all MLL1n signal is fusion - the sites and magnitude of MLL1n loss were similar to the lines that retain wild-type MLL1.
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
A similarity statement between lines, not a residual-occupancy measurement. Quantification is TSS-anchored signal, one timepoint (3 days), one drug concentration. ML2 carries KMT2A::AFDN in adult AML; the index fusion is KMT2A::AFF1 in infant B-ALL.
Evidence that argues against this (1)
- If wild-type MLL1 dominated the MLL1n ChIP signal in MOLM13 and RS4;11, a global retention driven by the wild-type protein could mask fusion-specific eviction in those two lines. The ML2 arm is the control for exactly that, but it is one line, one fusion partner and one experiment, with no replicate deposited. Krivtsov AV et al. 2019
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The only published statement that a fusion oncoprotein is evicted from chromatin by a menin inhibitor concerns NUP98 fusions, and it is scoped to a critical set of genes rather than genome-wide.
The menin-MLL1 interaction is a molecular dependency in NUP98-rearranged AML. · Heikamp EB, Henrich JA, Perner F, et al. · Blood 2022What this evidence cannot show
Mouse cell lines driven by NUP98-HOXA9 and NUP98-JARID1A plus PDX models; the NUP98 fusion lacks a menin-binding motif and reaches chromatin through wild-type MLL1, so it is the case in which eviction is most expected and least informative about a KMT2A fusion.
Evidence that argues against this (1)
- One could read the NUP98 result as proof of principle that the assay can detect fusion eviction when it happens, which weakens any claim that KMT2A-fusion eviction was simply never lookable-for. That is fair, and it is why the absence entry in this beat is bounded on what was ANALYSED rather than on what was assayable. Krivtsov AV et al. 2019
this graph's existing caution is exactly right here: the number that replicates is menin's, not the fusion's
the MLL1 half of a caution this graph already holds for the menin half
Sharpens that entry from the other direction: not only does 'eviction' often name menin rather than the fusion, but menin's own residence is NOT a proxy for the fusion's - the two dissociate completely when the fusion is removed.
the measurement behind the vocabulary correction
the protein the eviction claim is actually about
- related to →KMT2A / MLL1 — the gene that breaks
the shared N-terminal epitope is why MLL1 and the fusion cannot be separated in these assays
- related to →MEN1 drug-interface mutations — the first proof that a chromatin drug can drive Darwinian escape in patients
the resistance mutations prevent menin eviction, which is the exact half of the claim that is established
its opening sentence transfers eviction to the fusion; correction owed, with PMID 36922589 and 31821784 as the sources
- related to →Necessary is not explanatory: four different questions that a knockout answers as if they were one
same family of claim inflation: a scoped result quoted as an unscoped one
the vocabulary outruns the measurement because the measurement has one dimension