Hypothesis: the menin-MLL complex writes the same mark onto two address lists with opposite cell-state outputs, and which one a tissue reads decides whether blocking menin is a therapy or a lesion
also called the two-address-list hypothesis
The same molecular machine switches on growth-blocking genes in hormone glands and identity-freezing genes in blood. If that is right, a drug that jams the machine helps in one tissue and harms in the other, for exactly the same reason.
The same thing, for a scientist
Claim: menin-directed MLL H3K4 methylation is a single activity deployed onto at least two distinct address lists - CDKN1B/CDKN2C in endocrine cells [E1][E2], the HOXA cluster and the wider Mll-dependent self-renewal network in haematopoietic cells [E3][E4] - and the cell-state consequence of interrupting it is therefore tissue-determined rather than drug-determined. The prediction that makes it useful is quantitative and directional: a menin inhibitor should REDUCE H3K4 methylation and CDKN1B/CDKN2C transcript at endocrine promoters in the same animal and at the same exposure at which it reduces HOXA9/MEIS1 in marrow. That paired experiment has never been done. Two isolated pharmacological measurements exist and they disagree in sign: a menin-MLL inhibitor DOWN-regulated CDKN1A, CDKN1B and CDKN2C in primary human beta cells [E6], and MI-503 suppressed H3K4 methylation while RAISING p27 in osteosarcoma [E5]. Neither measured H3K4 methylation at a CDK-inhibitor promoter, so the shared-ACTIVITY core of the hypothesis is untested even by the study that confirms its transcript prediction.
The structure of the claim
| endocrine cell | haematopoietic cell | |
|---|---|---|
| complex | menin + MLL | menin + MLL (or KMT2A fusion) |
| activity | H3K4 methylation | H3K4 methylation |
| address list | CDKN1B, CDKN2C 12 | HOXA cluster, wider Mll-dependent network 34 |
| cell-state output | G1 restraint | differentiation block / self-renewal |
| consequence of losing the activity | proliferation, tumour | differentiation, therapy |
If the table is right, the therapeutic index of menin inhibition is not a property of the molecule at all. It is the ratio of two address lists' dosage sensitivities in two tissues.
What would make this boring, and what would make it wrong
Boring: if the endocrine address list turns out to be reached only by KMT2B rather than KMT2A, or only by a menin function that does not involve an H3K4 methyltransferase at all, then the two limbs share a protein and not an activity, and the hypothesis collapses into a restatement of pleiotropy. That is currently the live risk, because the one study that confirms the endocrine transcript prediction 6 measured no chromatin at all.
Wrong: if a clinical-grade menin inhibitor at an efficacious marrow exposure leaves H3K4 methylation and transcript at CDKN1B and CDKN2C unchanged in endocrine tissue, the shared-activity claim fails.
Where the two existing pharmacological results leave it. In primary human beta cells a menin-MLL inhibitor DOWN-regulated CDKN1A, CDKN1B and CDKN2C, which is the predicted direction 6. In osteosarcoma MI-503 suppressed H3K4 methylation and RAISED p27, which is the opposite 5. Two measurements, opposite signs, different tissues, neither with a chromatin readout at the CDK-inhibitor promoters. The prior is set at 0.55 on that basis - raised from an initial 0.35 written before 6 was found, and not raised further because the activity half of the claim has never been measured.
| statement | Menin-directed MLL H3K4 methylation is one activity deployed onto at least two address lists with opposite cell-state outputs - CDK-inhibitor genes giving G1 restraint in endocrine cells, and HOXA plus the wider self-renewal network giving a differentiation block in haematopoietic cells - so the sign of the effect of blocking menin is set by the tissue's address list, not by the drug. |
|---|---|
| origin | literature-gap |
| prior probability | 0.55 |
| what this would change | If confirmed, the endocrine safety of menin inhibition becomes a dosing and duration question with a measurable pharmacodynamic biomarker (endocrine CDKN1B transcript), rather than an untestable latency argument. If refuted, the endocrine and haematopoietic arms are mechanistically independent and the reassurance in gene-men1 strengthens. |
predictions
[ "A clinical-grade menin inhibitor at a marrow-efficacious exposure will reduce H3K4 methylation at the CDKN1B and CDKN2C promoters in parathyroid, pituitary and islet tissue in the same animal, by at least 1.5-fold relative to vehicle, within 14 days.", "CDKN1B and CDKN2C transcript will fall by at least 30% in those tissues over the same exposure.", "The magnitude of the endocrine effect will track the tissue's baseline MLL and p27/p18 expression, which is highest in pituitary (approx 17-fold and 12-fold over liver in mouse).", "In marrow from the same animals, HOXA9 and MEIS1 transcript will fall, i.e. both limbs move together in one experiment." ]
experiments
[ "exp-paired-endocrine-and-marrow-readout-under-a-menin-inhibitor" ]
analyses
[ "ana-q127-the-pi-dimensions-were-absent-from-the-entries-not-from-the-corpus" ]
falsifiers
[ "In mice dosed to a marrow-efficacious exposure of revumenib or ziftomenib for 14 days, H3K4me3 by CUT&RUN at the CDKN1B and CDKN2C promoters in PITUITARY changes by less than 1.5-fold versus vehicle while HOXA9 transcript in marrow falls by at least 50% in the same animals. That dissociates the two limbs at a single exposure and kills the shared-activity claim. NOT YET RUN.", "ALREADY PARTLY RESOLVED, AND IN BOTH DIRECTIONS AT ONCE. The transcript half of prediction 2 is CONFIRMED in one endocrine tissue: a menin-MLL inhibitor down-regulated CDKN1A, CDKN1B and CDKN2C in primary human beta cells [E6]. The same prediction is REFUTED in one non-endocrine tissue: MI-503 suppressed H3K4 methylation and RAISED p27 in osteosarcoma [E5]. Two pharmacological measurements exist in the whole literature and they disagree in sign, which is why the prior was raised from 0.35 to 0.55 rather than to something higher. The discriminating variable is presumably tissue; nothing tests that.", "If chromatin immunoprecipitation in endocrine tissue shows the CDKN1B and CDKN2C promoters are occupied by KMT2B and not KMT2A, the two limbs share a cofactor and not an enzyme, and the hypothesis must be restated at the level of menin rather than of the menin-MLL activity. NOT YET RUN.", "The SHARED-ACTIVITY claim specifically requires H3K4 methylation to fall at endocrine CDK-inhibitor promoters under drug. No pharmacological experiment anywhere has measured H3K4 methylation at CDKN1B or CDKN2C in any tissue. If the endocrine transcript effect turns out to be H3K4-independent, the hypothesis is reduced to 'menin does two things' and loses all of its predictive content." ]
status of predictions
{
"prediction_1_endocrine_H3K4me3_falls_under_drug": "NOT TESTED anywhere, in any tissue",
"prediction_2_endocrine_CDKN_transcript_falls_under_drug": "CONFIRMED ex vivo in human beta cells (PMID 32646642); CONTRADICTED in osteosarcoma, where p27 rose (PMID 40016386)",
"prediction_3_effect_tracks_baseline_module_expression": "NOT TESTED",
"prediction_4_both_limbs_move_in_one_experiment": "NOT TESTED - no study has read an endocrine and a marrow limb in the same animals"
}- If the hypothesis is true, every menin inhibitor is an endocrine CDK-inhibitor suppressant at therapeutic exposure, and the currently reassuring absence of endocrine neoplasia reflects only the short exposure of the relapsed population rather than an absence of the mechanism.
- A pharmacodynamic biomarker drawn from marrow cannot report on the endocrine limb, so the very assay used to show the drug is working is blind to the liability this hypothesis predicts.
- Is there any tissue in which both address lists are active at once? If so it is the cleanest system to test the hypothesis in, and nobody has looked for one.
- Does the KMT2A fusion, which occupies menin at high copy number, deplete menin from the endocrine address list in a patient with both a leukaemia and a germline MEN1 variant? That is a rare but real genotype and the answer is unknown.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin activates CDKN1B and CDKN2C transcription by recruiting MLL to their promoters and coding regions, and loss of either menin or MLL down-regulates both and deregulates cell growth.
Menin and MLL cooperatively regulate expression of cyclin-dependent kinase inhibitors. · Milne TA, Hughes CM, Lloyd R, Yang Z, Rozenblatt-Rosen O, Dou Y, Schnepp RW, Krankel C, Livolsi VA, Gibbs D, Hua X, Roeder RG, Meyerson M, Hess JL · Proc Natl Acad Sci U S A 2005What this evidence cannot show
Cell lines and transfection; abstract read only.
Evidence that argues against this (1)
- This is transfection and cell-line promoter work. The in vivo endocrine arm was established separately, and in ISLET TUMOURS rather than normal tissue, so the clean recruitment model may not describe a normal endocrine cell. Karnik SK et al. 2005
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin associates directly with regions of the p27 and p18 promoters and increases H3 lysine 4 methylation there in vivo, and that methylation is reduced at both loci in islet tumours from Men1 mutant mice.
Menin regulates pancreatic islet growth by promoting histone methylation and expression of genes encoding p27Kip1 and p18INK4c. · Karnik SK, Hughes CM, Gu X, Rozenblatt-Rosen O, McLean GW, Xiong Y, Meyerson M, Kim SK · Proc Natl Acad Sci U S A 2005What this evidence cannot show
Mouse islet tumours rather than normal islets; methylation state not resolved; abstract read only.
Evidence that argues against this (1)
- Measured in Men1-mutant islet TUMOURS, so the reduced H3K4 methylation may follow transformation rather than cause it; and the pharmacological arm that does test causality in normal human beta cells reports transcript only, with no chromatin readout, so the mark step of the hypothesis remains untested even where the transcript step is confirmed. Pahlavanneshan S et al. 2020
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin is required for the overwhelming majority of H3K4 methylation at Hox loci, while Mll1 is required for only a subset and MLL3/MLL4 and Set1 complexes have little to no effect there.
Global analysis of H3K4 methylation defines MLL family member targets and points to a role for MLL1-mediated H3K4 methylation in the regulation of transcriptional initiation by RNA polymerase II. · Wang P, Lin C, Smith ER, Guo H, Sanderson BW, Wu M, Gogol M, Alexander T, Seidel C, Wiedemann LM, Ge K, Krumlauf R, Shilatifard A · Mol Cell Biol 2009What this evidence cannot show
Mouse embryonic fibroblasts, not haematopoietic; abstract read only.
Evidence that argues against this (1)
- Fibroblasts, not blood. The haematopoietic limb of this hypothesis needs menin's H3K4 footprint measured in an HSPC, and in the one haematopoietic study available the Mll-dependent network is explicitly split into menin-dependent and menin-independent targets - which is a weaker statement than 'menin carries the majority'. Artinger EL et al. 2013
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The Mll-dependent transcriptional network in haematopoietic stem cells extends well beyond Hox genes, contains many characterised regulators of self-renewal, and contains both menin-dependent and menin-independent target genes.
An MLL-dependent network sustains hematopoiesis. · Artinger EL, Mishra BP, Zaffuto KM, Li BE, Chung EKY, Moore AW, Chen Y, Cheng C, Ernst P · Proc Natl Acad Sci U S A 2013What this evidence cannot show
Mouse; abstract read only; the menin-dependent/independent partition is not resolved gene by gene in the abstract.
Evidence that argues against this (1)
- If MLL1's methyltransferase activity is dispensable for HSC maintenance, then the haematopoietic limb of a hypothesis built on a shared H3K4-METHYLATION ACTIVITY is weaker than it looks: menin may matter in blood through complex assembly and targeting rather than through the mark, in which case the two limbs share an adaptor and not an activity. Mishra BP et al. 2014
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In osteosarcoma cells the menin inhibitor MI-503 markedly suppressed H3K4 methylation and INCREASED p27 expression by Western blot, alongside increased cleaved PARP and reduced Mcl-1 and c-Myc.
Menin inhibitor MI-503 exhibits potent anti-cancer activity in osteosarcoma. · Tian S, Hao ZY, Xu DH, Wang XZ, Shi CC, Zhang Y · Sci Rep 2025What this evidence cannot show
Osteosarcoma, tool compound, abstract read only, and the p27 rise co-occurs with apoptosis so it may be a consequence of arrest rather than a transcriptional effect. It is nevertheless the only pharmacological measurement of this axis in existence and it points the wrong way.
Evidence that argues against this (1)
- This result may be an artefact of the readout rather than a counter-example. p27 rose alongside cleaved PARP in the same blot, so the rise is consistent with arrest and apoptosis rather than with a transcriptional effect, and in an endocrine tissue with a clean transcript readout the direction is the predicted one. Pahlavanneshan S et al. 2020
- E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In primary human pancreatic beta cells, the menin-MLL interaction inhibitor MI-2 together with TGF-beta signalling inhibitors synergistically increased ex vivo replication without loss of insulin production, and the cooperation was attributed to DOWN-REGULATION of the cell-cycle inhibitors CDKN1A, CDKN1B and CDKN2C.
Combined inhibition of menin-MLL interaction and TGF-beta signaling induces replication of human pancreatic beta cells. · Pahlavanneshan S, Behmanesh M, Oropeza D, Furuyama K, Tahamtani Y, Basiri M, Herrera PL, Baharvand H · Eur J Cell Biol 2020What this evidence cannot show
Ex vivo human donor islets, days in culture, MI-2 (an early tool compound, weaker than MI-463/MI-503 and not a clinical molecule), and the CDKN result is transcript-level with no chromatin readout - no H3K4 methylation was measured, so the shared-ACTIVITY half of the hypothesis is still untested even here. Abstract read, full text not opened. Already cited twice in this corpus (mechanism-menin-inhibition-replicates-human-beta-cells-ex-vivo, concept-menin-inhibition-is-an-active-therapeutic-hypothesis), and the conclusion drawn there is the same one drawn here; this entry adds the pairing with the osteosarcoma counter-result, not the observation.
Evidence that argues against this (1)
- The confirming result used MI-2, an early tool compound, ex vivo, for days, and measured transcript only. A clinical menin inhibitor at a marrow-efficacious exposure is a different perturbation, and no clinical menin inhibitor has ever been tested on human islets - so this evidence item confirms the prediction for a compound nobody gives to patients. Brown MR et al. 2025
the endocrine limb of the hypothesis
the haematopoietic limb of the hypothesis
the paired-tissue experiment that scores the hypothesis' falsifiers
the endocrine half of the two-address-list claim