When MYC falls after a menin inhibitor, there are three mechanisms it could be, and no published experiment separates them
also called is MYC a direct menin target · MYC pharmacodynamic biomarker problem · MYC direct versus differentiation readout
MYC goes down when leukaemia cells are treated with a menin inhibitor, and people would like to use that as a way of checking the drug is working. But there are at least three different reasons MYC could fall, and telling them apart matters: one of them means MYC is a real drug target, and another means MYC is just what happens when a cell grows up.
The same thing, for a scientist
Three mechanisms are individually supported and mutually unseparated. **(A) Direct transcriptional target.** MYC is among the non-canonical menin targets whose chromatin is bivalently co-occupied by active menin and repressive H2AK119ub, and PRC1.1 loss reactivates MYC to produce menin-inhibitor resistance [E1] - genetics that place MYC downstream of menin through a PRC1.1-gated chromatin route. **(B) Direct protein partner of MYC, not of the fusion.** Menin interacts directly with the MYC transactivation domain, co-localises with MYC at E-boxes, and enhances transcription of MYC target genes in a P-TEFb-dependent manner **independent of H3K4me3** [E2]. On this model a menin inhibitor could collapse MYC *output* without changing MYC *level* at all. **(C) Differentiation read-out.** Menin inhibition's dominant phenotype is differentiation, and MYC falls during normal myeloid differentiation for reasons that have nothing to do with the drug; combination with a DOT1L inhibitor suppresses MYC targets more profoundly precisely as differentiation and killing are enhanced [E3]. **No experiment in the literature separates A, B and C**, and until one does, every MYC-based pharmacodynamic biomarker proposed for this disease rests on an unassigned mechanism - a point sharpened by the fact that the closest analogous claim, in SWI/SNF, is itself an unresolved contradiction [E4][E5].
What distinguishes the three, concretely
| MYC transcript | MYC protein | MYC target genes | requires differentiation | time to effect | |
|---|---|---|---|---|---|
| A direct chromatin target | falls | falls | fall | no | hours |
| B menin as a MYC coactivator | unchanged | unchanged | fall | no | hours |
| C differentiation read-out | falls | falls | fall | yes | days |
Three cheap discriminators follow and none of them has been run:
- Blot MYC protein and measure MYC transcript at early timepoints. Model B predicts both are flat while a MYC target signature collapses. This alone separates B from A and C and takes a western blot.
- Block differentiation and re-measure. If MYC still falls in cells held at an immature state - by a differentiation-blocking genetic lesion, or by comparing a menin-inhibitor-resistant differentiated line - then C is excluded. This is the discriminator that matters most for biomarker use.
- Compare kinetics. A is hours; C follows the differentiation programme and is days. A dense early time course distinguishes them without any perturbation at all.
The graph already holds the analogous unresolved question one layer up: mechanism-baf-myc-contradiction-two-papers-one-model records two papers disagreeing about whether MYC is the node downstream of SWI/SNF in the same disease, and one of the open questions there - whether MYC protein rather than transcript moves - is exactly discriminator 1 above, asked of a different upstream factor. Two independent chromatin dependencies in this disease have converged on the same unmeasured quantity.
Why model B is the one this project has been ignoring, and why it is the most awkward
Model B deserves separate attention because it is the only one under which MYC transcript is a useless biomarker while MYC biology is entirely real.
Wu et al. report menin binding the MYC transactivation domain directly and enhancing MYC target-gene transcription at E-boxes in a P-TEFb-dependent manner, explicitly independent of H3K4me3 2. That is menin acting as a MYC coactivator, not as a KMT2A-fusion scaffold. This graph is built around the menin/KMT2A MBM1/MBM2 pocket - the interface every menin inhibitor in the clinic occupies - and the obvious and unanswered question is whether the MYC TAD contacts that same pocket. If it does, every menin inhibitor is also a MYC-coactivator inhibitor and nobody has said so. If it does not, then menin inhibitors leave the MYC axis intact and model B is irrelevant to this project's drugs. Neither has been tested, and the experiment is a competition binding assay that this field has the reagents for.
The secondary consequence is a toxicity one. Menin's MYC-coactivator function is described in cancers generally, not in leukaemia specifically, so if the clinical menin inhibitors hit it, their activity in KMT2A-wild-type tissue is not fully accounted for by the fusion mechanism - which bears on the on-target effects seen in normal tissues that gene-men1 and this project's menin-normal-tissue entries track.
The bounded absences
Measured 2026-09-20 on PubMed with querytranslation read back on each query.
(menin inhibitor terms) AND MYC[ti]returns 1 record, and it is a 2005 paper about MTA1 as a c-MYC effector with no menin content. No paper has put a menin inhibitor and MYC in its title. Positive control through the identical code path: the same clause withMYC[tiab]instead ofMYC[ti]returns 13 records, so the instrument returns ones.(RING1B OR RNF2 OR PCGF1 OR BCOR OR KDM2B)[tiab] AND (menin inhibitor terms)returns 1 record - Zhou et al. 1. The entire published intersection of PRC1.1 and menin inhibitors is one paper. Same positive control applies.
Both bounds are symbol- and phrasing-limited in the way brief rule 64 describes: a paper framing its result at the level of 'Polycomb' or 'the menin-MLL interaction' without these tokens would be missed.
| mechanism class | unassigned causal route for a candidate pharmacodynamic biomarker |
|---|---|
| parallel unresolved case | mechanism-baf-myc-contradiction-two-papers-one-model - the same unmeasured quantity (MYC protein) is the open discriminator for a different upstream factor in the same disease |
models
[
{
"label": "A",
"name": "direct chromatin target",
"prediction": "MYC transcript and protein both fall within hours; no differentiation required",
"support": "PMID 39158067 (PRC1.1-gated non-canonical menin target, genetics)"
},
{
"label": "B",
"name": "menin as a MYC coactivator",
"prediction": "MYC transcript and protein UNCHANGED; MYC target signature collapses",
"support": "PMID 28474697 (direct menin-MYC TAD interaction, E-box co-localisation, P-TEFb-dependent, H3K4me3-independent)",
"unresolved": "whether the MYC TAD contacts the MBM1/MBM2 pocket occupied by clinical menin inhibitors"
},
{
"label": "C",
"name": "differentiation read-out",
"prediction": "MYC falls on the timescale of the differentiation programme (days) and does not fall if differentiation is blocked",
"support": "PMID 27840424 (MYC-target suppression reported together with enhanced differentiation, no separating arm)"
}
]discriminating experiments none of which have been run
[ "blot MYC protein and measure MYC transcript at 2, 4, 8 and 24 h - separates B from A and C", "block differentiation and re-measure MYC - excludes C", "dense early time course - separates hours (A) from days (C) with no perturbation", "competition binding of the MYC TAD against an MBM peptide for the menin pocket - decides whether model B is even pharmacologically reachable by the clinical agents" ]
bounded absences
[
{
"claim": "no paper places a menin inhibitor and MYC in its title",
"query": "(revumenib OR \"menin inhibitor\" OR \"menin inhibitors\" OR VTP-50469 OR SNDX-5613 OR ziftomenib)[tiab] AND MYC[ti]",
"count": 1,
"on_topic": 0,
"positive_control": "same clause with MYC[tiab] returns 13",
"tool": "NCBI E-utilities esearch db=pubmed",
"date": "2026-09-20"
},
{
"claim": "the entire published intersection of PRC1.1 subunits and menin inhibitors is one paper",
"query": "(RING1B OR RNF2 OR PCGF1 OR BCOR OR KDM2B)[tiab] AND (revumenib OR \"menin inhibitor\" OR \"menin inhibitors\" OR SNDX-5613 OR VTP-50469)[tiab]",
"count": 1,
"on_topic": 1,
"record": "PMID 39158067",
"positive_control": "each clause alone returns hundreds; the PRC1.1 clause is not empty",
"tool": "NCBI E-utilities esearch db=pubmed",
"date": "2026-09-20"
}
]- If model C is right, a MYC-based pharmacodynamic biomarker measures differentiation and will therefore be positive in any patient who responds for any reason, including responses that have nothing to do with target engagement - a biomarker that cannot fail is not a biomarker.
- If model B is right, MYC transcript and protein are flat in a responding patient and a MYC assay reads as a treatment failure while the drug is working.
- Model A is the only one under which MYC is a target worth co-inhibiting, and the resistance literature already shows that MYC reactivation is the escape route - so pursuing MYC co-inhibition on the strength of an unassigned mechanism risks building a combination on the wrong causal picture.
- All three models predict MYC falls. A confirmatory experiment that only measures MYC going down cannot distinguish them and will be read as support for whichever model its authors favour.
- Does the MYC transactivation domain bind the same menin pocket as the KMT2A MBM motifs? If it does, every clinical menin inhibitor is also a MYC-coactivator inhibitor and nobody has said so; if it does not, model B is pharmacologically unreachable and can be set aside.
- Does MYC protein change after menin inhibition at early timepoints? A western blot separates the coactivator model from the other two and appears never to have been reported.
- Does MYC still fall if differentiation is blocked? This is the discriminator that decides whether a MYC-based pharmacodynamic assay measures drug action or measures response.
- Is the answer the same in KMT2A::AFF1 B-lineage infant disease as in KMT2A-r AML? Every result cited here is myeloid, and the differentiation programme a B-lineage blast would enter is different.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
MYC is among a group of non-canonical menin targets bivalently co-occupied by active menin and repressive H2AK119ub that are typically downregulated after menin inhibition; loss of PRC1.1 subunits reactivates them including MYC and confers menin-inhibitor resistance, and genetic or pharmacological 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
The resensitisation genetics establish that MYC is functionally load-bearing in the resistant state; they do not establish that the MYC fall in the SENSITIVE state is a direct chromatin event rather than a consequence of differentiation. The same paper reports that PRC1.1-deficient cells have a reduced monocytic signature, i.e. the resistance phenotype is itself partly a differentiation-state phenotype, which entangles models A and C rather than separating them.
Evidence that argues against this (1)
- Menin has a direct, H3K4me3-independent coactivator function at MYC's own transactivation domain, which predicts that MYC target genes can collapse without MYC transcript changing at all. A chromatin-level account of MYC downregulation does not exclude a post-transcriptional, protein-protein account operating in parallel, and no reported experiment separates them. Wu G et al. 2017
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Menin interacts directly with the transactivation domain of MYC and co-localises with MYC at E-boxes, enhancing transcription of MYC target genes in a P-TEFb-dependent manner and independently of H3K4me3 activity, thereby stimulating proliferation and cellular metabolism in vitro and in vivo.
Menin enhances c-Myc-mediated transcription to promote cancer progression. · Wu G, Yuan M, Shen S, Ma X, Fang J, Zhu L, et al. · Nat Commun 2017What this evidence cannot show
Not leukaemia - the systems are cancer cell lines generally, and no KMT2A-rearranged model is named in the abstract. Critically, the paper does NOT report whether the MYC TAD contacts the MBM1/MBM2 pocket that clinical menin inhibitors occupy, so it does not establish that any approved or trial menin inhibitor disrupts this interaction. Graded moderate on that basis: the interaction is supported, its pharmacological relevance to this project's drugs is not.
Evidence that argues against this (1)
- If menin's MYC-coactivator function were the dominant route in KMT2A-rearranged leukaemia, menin inhibitors would be expected to act in KMT2A-wild-type MYC-driven disease as well, and their activity is instead strikingly restricted to KMT2A-rearranged and NPM1-mutant contexts. The selectivity of the clinical agents argues that the fusion-scaffold route dominates in this disease whatever the coactivator route does elsewhere. Zhou X et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Combining a DOT1L inhibitor with an inhibitor of the MLL-menin interaction markedly enhances induction of differentiation and cell killing across MLL disease models including primary cells while sparing normal haematopoiesis, and gene-expression analysis shows that MLL-fusion target genes and MYC targets are 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
The MYC-target suppression is reported together with enhanced differentiation, in the same experiment, with no arm that separates them - which is precisely the entanglement this entry is about. The measurement is a gene-expression signature of MYC targets, not MYC transcript or protein, so it is compatible with all three models including the coactivator model.
Evidence that argues against this (1)
- Neville et al. reframe the same DOT1L-menin complementarity as two drugs producing differently shaped chromatin responses - subset-restricted versus genome-wide H2AK119ub gain - rather than as additive suppression of a shared target set. Under that reading, deeper suppression of MYC targets under combination is a consequence of two distinct mechanisms overlapping, not evidence that MYC is the shared node. Neville D et al. 2026
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In mouse MLL-AF9 leukaemia, SWI/SNF sustains the oncogenic programme with Myc as a key target, acting at a cluster of lineage-specific enhancers 1.7 Mb downstream of Myc where Brg1 maintains 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, RN2 model with a cooperating NrasG12D allele, and the MYC arm is directly contested in human cells. Cited here to show that the same unassigned-mechanism problem exists for a second chromatin dependency in this disease, not as independent support for any of models A, B or C.
Evidence that argues against this (1)
- Cruickshank et al. found the same subunits not required for c-MYC-regulated gene expression in human MLL-rearranged cells. The disagreement is recorded and unresolved in mechanism-baf-myc-contradiction-two-papers-one-model, and its open questions include the same discriminator this entry proposes: nobody blotted MYC protein after subunit knockdown in the human lines. Cruickshank VA et al. 2015
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Counted absence: on PubMed, 2026-09-20, the query (revumenib OR 'menin inhibitor' OR 'menin inhibitors' OR VTP-50469 OR SNDX-5613 OR ziftomenib)[tiab] AND MYC[ti] returns 1 record, which is a 2005 paper on MTA1 as a c-MYC effector with no menin content - so no paper has placed a menin inhibitor and MYC together in its title. Positive control: the identical query with MYC[tiab] returns 13 records.
Metastasis-associated protein 1 (MTA1) is an essential downstream effector of the c-MYC oncoprotein. · Zhang XY, DeSalle LM, Patel JH, Capobianco AJ, Yu D, Thomas-Tikhonenko A, et al. · Proc Natl Acad Sci U S A 2005What this evidence cannot show
Bounded by the drug-name list and by [ti]; a title using 'MLL-menin' or a compound code not in the list would be missed, and brief rule 64's complex-form failure mode applies. The citation attached is the single off-topic record returned, cited so the count is auditable rather than as support for anything. Title, authors and DOI taken from the fetched PubMed record on 2026-09-20.
Evidence that argues against this (1)
- Thirteen records name a menin inhibitor and MYC in title or abstract, so the field is not silent about the pair - only about making it the headline. An absence at title level is weaker evidence of neglect than it appears, and this bound should be read as 'nobody has framed a study around it', not 'nobody has looked'. Zhou X et al. 2024
three unseparated causal routes for the MYC fall after menin inhibition
model B: menin binds the MYC transactivation domain directly and coactivates at E-boxes, independent of H3K4me3
any MYC-based pharmacodynamic biomarker for this drug rests on an unassigned mechanism
- related to →Is MYC the node? Two papers, the same MLL-AF9 model, opposite answers - and the second one failed to rescue
the same unmeasured quantity - MYC protein - is the open discriminator in both
- related to →The differentiation block — what it actually is, operationally, in KMT2A::AFF1 leukaemia
model C: MYC falls because the cell differentiates, not because of anything at the MYC locus
- related to →MYC reactivation through PRC1.1 loss — a named, druggable bypass out of menin dependence
model A rests on the PRC1.1-gated non-canonical menin target genetics recorded there