Turn menin off for a defined window, then back on: the missing experiment that would bound reversibility
also called Men1 restoration mouse · reversible Men1 knockout · menin off-on experiment
Mice that lose menin in their insulin cells get overgrowth in two months and tumours by six. Nobody has tried switching menin back on after one, two or three months to see whether the damage undoes itself - which is exactly the question for a child who takes the drug for a while and stops.
The same thing, for a scientist
Every existing **endocrine** Men1 conditional model removes menin permanently (narrowed 2026-09-19: a restoration arm does exist outside the endocrine system - Men1 deleted in hippocampal CA1 and then re-expressed by AAV recovers the behavioural deficit [E5] - and menin re-expressed in Men1-deficient Leydig tumour cell lines arrests the cell cycle and induces apoptosis in vitro [E6]; neither titrates the off-window, which is what this experiment is for). Beta-cell-specific ablation gives hyperplastic islets at 2 months and insulinomas at 6 with full penetrance [E1], and the human germline curve gives 7% penetrance at age 10 [E2] - but neither system has a **restoration** arm, so the interval between a harmless and a committed exposure is unconstrained from below. This proposes a reversible conditional Men1 allele in beta cells and in parathyroid, with menin removed for 2, 4, 8 or 12 weeks and then restored, and animals followed to 12 months. The primary question is whether islet or parathyroid hyperplasia present at the end of the off-window regresses, persists, or progresses after restoration. Whatever the answer, it converts the central reversibility argument of this beat from an assertion into a number. **Correction, 2026-09-19:** the cheap pharmacological arm this design nominates as its first step has substantially been run. Twelve weeks of a menin inhibitor at 70 mg/kg/day in mice produced increased beta-cell proliferation with improved fasting glucose and glucose tolerance and no reported islet tumour, and inducible Men1 deletion in the beta cell is reported as increasing beta-cell mass and ameliorating diabetes [counter-evidence to E1]. The genetic restoration arms therefore have to be justified by something other than the assumption that drug reproduces the ablation phenotype, because at 12 weeks it does not.
Design
Allele. A doxycycline- or tamoxifen-controlled reversible Men1 system - either a degron-tagged endogenous menin or an inducible shRNA against Men1 - in which menin protein can be lowered and then restored on a defined schedule. A degron is preferable because it produces partial, graded loss of protein, which is a closer model of pharmacological occupancy than a genetic deletion.
Tissues. Pancreatic beta cell (RIP-Cre equivalent) as the primary, because the latency is established there 1; parathyroid as the secondary, because it is the first human MEN1 target.
Arms. Off-windows of 2, 4, 8 and 12 weeks, plus a never-restored positive control and a never-depleted negative control. A 12-week window is deliberately matched to the 13-week rat toxicity studies that produced the irreversible findings on the revumenib label.
Readouts. Islet and parathyroid histology with proliferation index at the end of the off-window and at 3, 6, 9 and 12 months after restoration; serum insulin, glucose, calcium and PTH; and RNA-seq of sorted cells for the JunD/AP-1 signature to test whether the transcriptional programme returns to baseline even where histology does.
A parallel pharmacological arm. The same schedule using a clinical menin inhibitor instead of the genetic system, in wild-type mice, which tests whether the drug reproduces any part of the genetic phenotype at all. This arm is cheaper than the genetic one and was originally listed as the first step - superseded, see the correction immediately below: it has substantially been run and did not reproduce the genetic phenotype, so what remains is a narrower version of it (parathyroid histology, calcium and PTH, a juvenile cohort, exposure beyond 12 weeks) rather than the arm as written.
Correction 2026-09-19 - this arm is partly done, and it did not reproduce the genetic phenotype. Ma et al. ran 12 weeks of menin inhibitor at 70 mg/kg/day by gavage in mice and saw increased beta-cell proliferation with improved fasting glucose and glucose tolerance, with no islet tumour reported [counter-evidence to E1]. What that study did not measure is exactly what a new pharmacological arm should add: parathyroid histology, serum calcium and PTH, a juvenile cohort, and exposure beyond 12 weeks. Note also that every rodent menin-inhibitor exposure in the literature is 12-13 weeks while the clinical question spans years, so the scaling assumption underneath both the pharmacological and the genetic design is untested.
What it would settle
- If hyperplasia regresses after restoration at every window up to 12 weeks: the reversibility argument that currently carries this beat's counter-case becomes an experimentally supported claim rather than a plausibility argument, and the endocrine concern can be de-prioritised with confidence.
- If there is a window beyond which hyperplasia persists or progresses: that window is the number the field needs, and it maps directly onto a maximum safe continuous exposure - which matters now that menin inhibitors are moving into maintenance and FDA's postmarketing requirement specifies at least three years of treatment 3.
- If the transcriptional signature fails to normalise even where histology does: that is a hysteresis result, and it connects to this graph's existing work on pulsed menin inhibition and agency-revocation hysteresis.
| design | reversible conditional Men1 system (degron-tagged endogenous menin preferred over inducible shRNA) in pancreatic beta cells and parathyroid, with menin depleted for 2, 4, 8 or 12 weeks and then restored, plus never-restored and never-depleted controls, and a parallel pharmacological arm using a clinical menin inhibitor in wild-type mice on the same schedule |
|---|---|
| readout | islet and parathyroid histology with proliferation index at end of off-window and at 3, 6, 9 and 12 months after restoration; serum insulin, glucose, calcium and PTH; RNA-seq of sorted cells for JunD/AP-1 signature and beta-catenin target genes |
| controls | never-restored positive control; never-depleted negative control; vehicle arm in the pharmacological experiment |
| cost tier | high |
| timeline | 18-24 months for the genetic arms; 6-9 months for the pharmacological arm, which should be run first |
| rationale for 12 week window | matches the 13-week rat toxicity studies that produced the irreversible growth-plate and pancreatic findings on the revumenib label |
| primary question | is there a duration of menin depletion beyond which endocrine hyperplasia does not regress on restoration |
| gap addressed | No published conditional Men1 study with a RESTORATION arm was located, and that remains true as of 2026-09-19. What is no longer true, and was asserted here before, is that the pharmacological arm is unrun: a 12-week menin-inhibitor arm in mice exists (PMID 33821572) and produced a benign, functionally beneficial islet phenotype, and tamoxifen-inducible Men1 deletion in the beta cell is established and reported as increasing beta-cell mass and ameliorating diabetes (PMID 28270438). |
| pharmacological arm status | partly pre-empted: 70 mg/kg/day menin inhibitor by gavage, high-fat-diet mice, 12 weeks, improved glucose tolerance and increased beta-cell proliferation, no tumour reported (PMID 33821572). What that arm did NOT include: parathyroid histology, serum calcium or PTH, a juvenile cohort, or exposure beyond 12 weeks. Those are the four things a new pharmacological arm should add. |
- This is the most expensive experiment in the beat and the one most likely to be skipped. Its usual justification - that the cheap pharmacological arm should be run first and may make the genetic arms unnecessary - has partly been overtaken: a 12-week pharmacological arm exists and showed a benign islet phenotype (PMID 33821572). The genetic arms now need a justification that does not rest on that comparison.
- Mouse reversibility may not translate: mouse islets are more regenerative than human, which would make a mouse regression result over-optimistic. The same caveat cuts the other way against the reassuring 12-week pharmacological result.
- Every rodent menin-inhibitor exposure in the literature is 12-13 weeks. The clinical question concerns years. An 18-24 month genetic programme built on a 12-week comparator risks answering a question nobody asked, and the honest cheap first move is a longer pharmacological arm with parathyroid histology, calcium and PTH - none of which the existing 12-week study measured.
- If the beneficial islet phenotype is real and durable, the endocrine framing of this whole beat inverts: menin inhibition would be an endocrine-sparing or even endocrine-beneficial intervention over months, and the risk would sit entirely in the multi-year tail. An experiment designed around 12-week windows cannot see that tail.
- Does a reversible Men1 degron allele exist? Tamoxifen-inducible Men1 deletion lines clearly do (PMID 28270438), so the real rate-limiting step is restoration, not depletion - and the cheapest route may be an inducible transgene rescue on an inducible-knockout background rather than a degron built from scratch.
- Why does 12 weeks of pharmacological menin inhibition in mouse islets give compensatory proliferation with improved glucose tolerance (PMID 33821572) while genetic beta-cell Men1 ablation gives hyperplasia at 2 months and insulinoma at 6 [E1]? Partial occupancy versus complete loss is the obvious answer, but nobody has measured percentage menin occupancy against percentage protein loss, and until someone does, the latency curve in E1 cannot be applied to a drug.
- Is partial menin depletion by degron a better model of pharmacological occupancy than complete conditional deletion? It should be, but the correspondence between percentage protein depletion and percentage pocket occupancy is not established.
- Should the parathyroid arm come first given that the human signal is a parathyroid one? The islet arm has the established latency AND an existing 12-week pharmacological dataset, but the published pharmacological work measured no calcium, no PTH and no parathyroid histology, so the parathyroid arm is where the genuinely new information is.
- Does any published menin-inhibitor rodent study extend beyond 13 weeks? The revumenib label's juvenile-rat work is 13 weeks and the beta-cell work is 12 weeks. FDA postmarketing requirement 4730-1 contemplates at least 3 years of human treatment [E3]. Nothing in the animal literature spans that ratio.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Mice with Men1 disrupted specifically in pancreatic beta cells developed hyperplastic islets as early as 2 months and insulinomas at 6 months with full penetrance, with multistage progression and disturbed blood insulin and glucose.
Pancreatic beta-cell-specific ablation of the multiple endocrine neoplasia type 1 (MEN1) gene causes full penetrance of insulinoma development in mice · Bertolino P, Tong WM, Herrera PL, Casse H, Zhang CX, Wang ZQ · Cancer Res 63:4836-41 2003What this evidence cannot show
Read as abstract. Permanent ablation with no restoration arm, which is the gap this experiment addresses. The abstract does not report animal numbers or the proportion hyperplastic at the 2-month timepoint, both of which would be needed to power a restoration study.
Evidence that argues against this (2)
- The 'parallel pharmacological arm' this entry says should be run first has substantially been run, and its result argues against the premise. Ma et al. gave a menin inhibitor at 70 mg/kg daily by gavage to high-fat-diet mice and followed them for 12 weeks - the same window this design nominates as its longest off-period, chosen to match the 13-week rat toxicity studies. The observed islet phenotype was increased beta-cell proliferation with LOWER fasting glucose and improved glucose tolerance from week 6 through week 12, mediated through relief of JunD/menin/HDAC3 repression of Pbk; MI-503 and MI-463 abolished menin-JunD co-IP at 30-300 nM, and MI-503 at 10-100 nM raised Ki67 in primary human islets. No insulinoma or hyperplastic lesion is reported. Twelve weeks of pharmacological menin inhibition therefore does not reproduce the genetic beta-cell-ablation phenotype, which weakens the assumption that the Bertolino latency curve translates to drug exposure at all, and means the genetic restoration arms are being justified by a comparison that has already partly failed. Ma J et al. 2021
- Time-restricted, inducible menin loss in the beta cell has also already been done, and it was reported as therapeutic rather than oncogenic. Muhammad et al. state that tamoxifen-inducible deletion of Men1 increases beta-cell mass in multiple murine models of diabetes and ameliorates diabetes, and that ex vivo treatment of both mouse and human pancreatic islets with a small-molecule menin inhibitor increases Ki67. That is not a restoration experiment - menin is not brought back - but it does show that inducible, post-developmental menin loss in the islet has an established literature in which the phenotype is framed as regenerative. The entry's open question 'does a reversible Men1 allele already exist in any laboratory' should therefore start from the tamoxifen-inducible Men1 lines that clearly do exist, rather than from a degron that would have to be built. Muhammad AB et al. 2017
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Age-related penetrance of MEN1 in germline carriers is 7% at 10 years and 52% at 20 years, establishing that permanent monoallelic menin loss from conception produces disease slowly.
Characterization of mutations in patients with multiple endocrine neoplasia type 1 · Bassett JH, Forbes SA, Pannett AA, Lloyd SE, Christie PT, Wooding C, Harding B, Besser GM, Edwards CR, Monson JP, Sampson J, Wass JA, Wheeler MH, Thakker RV · Am J Hum Genet 62:232-44 1998What this evidence cannot show
Read as abstract. Human germline data cannot bound a partial reversible exposure, which is why a mouse restoration experiment is needed rather than more human epidemiology.
Evidence that argues against this (1)
- Germline penetrance establishes that permanent monoallelic loss is tumourigenic over decades. A restoration experiment is asking a different question - whether an off-window is reversible - and the penetrance curve provides no prior for how long an off-window may be, because it never measures one. American Journal of Human Genetics 1998
- E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
FDA postmarketing requirement 4730-1 specifies that all patients in the required revumenib trial should have at least 3 years of treatment or have discontinued earlier, to assess serious potential risk of endocrine-related disorders due to chronic menin inhibition.
NDA 218944 approval letter for REVUFORJ (revumenib), including postmarketing requirements 4730-1 to 4730-6 · US Food and Drug Administration, Center for Drug Evaluation and Research · Drugs@FDA, NDA 218944 approval letter 2024What this evidence cannot show
Read verbatim from the NDA 218944 approval-letter PDF on 2026-09-18. It sets the exposure duration the field expects to be clinically relevant, which is why the mouse off-windows should extend to at least the equivalent of a full human treatment course.
Evidence that argues against this (1)
- A postmarketing requirement specifies a REGULATORY minimum follow-up, not a biological off-window, and the two are easily confused. Three years of continuous treatment is the opposite of what this experiment studies: it is sustained exposure, whereas the experiment asks whether a period OFF the target is reversible. The PMR justifies the question without constraining the design. FDA Drugs@FDA 2024
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In MEN1-deficient pancreatic neuroendocrine tumours, beta-catenin signalling is activated and conditional beta-catenin knockout suppresses tumorigenesis and prolongs survival in Men1-deficient mice, while beta-catenin blockade does not impair physiological beta-cell function.
Targeting beta-catenin signalling for therapeutic intervention in MEN1-deficient pancreatic neuroendocrine tumours · Jiang X, Cao Y, Li F, Su Y, Li Y, Peng Y, Cheng Y, Zhang C, Wang W, Ning G · Nat Commun 5:5809 2014What this evidence cannot show
Read as abstract. Identifies beta-catenin activation as a candidate molecular marker of commitment, which gives the restoration experiment a mechanistic readout beyond histology.
Evidence that argues against this (1)
- Beta-catenin dependence is a property of established MEN1-deficient tumours. A latency experiment measures time to initiation, so a maintenance mechanism neither predicts the off-window nor provides a readout for it. Nature Communications 2014
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
MEN1 deletion restricted to hippocampal CA1 in a conditional knockout mouse produces contextual fear-conditioning deficits that are recovered by subsequently re-injecting to overexpress MEN1 - an existing deletion-then-restoration design.
Selective Menin Deletion in the Hippocampal CA1 Region Leads to Disruption of Contextual Memory in the MEN1 Conditional Knockout Mouse: Behavioral Restoration and Gain of Function following the Reintroduction of MEN1 Gene · Ulfat AK, Batool S, Iqbal F, Syed NI · Cells 2022What this evidence cannot show
Neurons, behaviour, viral overexpression rather than restoration of the endogenous allele, single fixed off-interval, no neoplasia endpoint. It shows the design is buildable, not that the endocrine question is answered.
Evidence that argues against this (1)
- The hippocampal restoration result is the closest thing to a precedent and it is in NEURONS, on a behavioural readout, with no tumour endpoint. Recovery of contextual fear conditioning after AAV re-expression shows a menin-dependent function can be restored in one tissue; it says nothing about whether an endocrine proliferative lesion regresses. Cells 2022
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Reconstituted menin expression in Men1-deficient mouse Leydig tumour cell lines inhibits growth, blocks G0/G1-to-S transition, raises p18INK4C and p27Kip1 and increases apoptosis.
Reconstituted expression of menin in Men1-deficient mouse Leydig tumour cells induces cell cycle arrest and apoptosis · Hussein N, Casse H, Fontaniere S, Morera AM, Asensio MJ, Bakeli S, Lu JL, Coste I, Di Clemente N, Bertolino P, Zhang CX · European Journal of Cancer 2007What this evidence cannot show
In vitro, transfected re-expression, cells already transformed, Leydig lineage. Establishes that menin-loss-driven growth stays menin-responsive after transformation in a dish; says nothing about whether an established lesion regresses in vivo.
Evidence that argues against this (1)
- Reconstituted menin arrests growth in Men1-deficient Leydig tumour CELL LINES in vitro. That is restoration into an already-transformed line rather than into a normal tissue within an off-window, and cell-cycle arrest in culture is a weaker endpoint than tumour regression in an animal. European Journal of Cancer 2007
the titration this risk entry says is missing; both were narrowed 2026-09-19 because off-then-on Men1 designs exist outside endocrine tissue
existing restoration designs (PMIDs 36552783, 17184987) and the onset bound (PMID 28420716)
supplies the restoration arm that no published conditional Men1 study has
would produce the duration threshold the reversibility argument currently asserts without a number
a transcriptional signature that fails to normalise after menin restoration would be a hysteresis result in endocrine tissue
Counter-evidence (Ma 2021, PMID 33821572; Muhammad 2017, PMID 28270438): the 'cheap pharmacological arm to run first' has substantially been run - 12 weeks of menin inhibitor at 70 mg/kg/day gave increased beta-cell proliferation with improved glucose tolerance and no tumour - and tamoxifen-inducible Men1 deletion lines already exist. The drug does not reproduce the ablation phenotype at 12 weeks.
Provides the pharmacological arm that experiment nominates, and leaves its central question - does it reverse off drug - entirely open, since no withdrawal arm was run