The two co-IPs that settle whether menin's SUV39H1 window and its gamma-H2A.X motif are one surface: swap each laboratory's construct into the other's assay
also called reciprocal deletion co-IP · delta-371-400 versus D418N/W436R
Two labs made the constructs needed to answer this and neither tested the other's. Put the deletion mutant from the 2026 repair paper into the 2013 enzyme-binding assay and the two patient mutants from 2013 into the 2026 repair assay, and the question closes.
The same thing, for a scientist
Two co-immunoprecipitations in HEK-293T, both using plasmids described in published work. Arm 1: menin delta-371-400 (validated gamma-H2A.X-null in PMID 42578372) against GFP-SUV39H1, the exact assay of PMID 23579270. Arm 2: menin D418N and W436R (validated SUV39H1-binding-impaired in PMID 23579270) against gamma-H2A.X after 10 Gy, the exact assay of PMID 42578372. Internal negatives: menin A411P and L413P, SvID mutants that do NOT impair SUV39H1 binding. Mandatory fold control on every construct: co-IP of the KMT2A MBM, because menin 363-371 line the drug pocket and a construct that has lost KMT2A binding has lost its fold rather than a specific contact. Readout is co-IP band intensity normalised to input and to the KMT2A control, three biological replicates.
Design, and the one control that makes it interpretable
| arm | construct | bait/prey | prediction if the surfaces are SEPARATE | prediction if they are ONE |
|---|---|---|---|---|
| 1 | menin delta-371-400 | GFP-SUV39H1, 293T co-IP | SUV39H1 binding retained, >= 70% of wild type | SUV39H1 binding lost, <= 30% of wild type |
| 2 | menin D418N | endogenous gamma-H2A.X, 293T or A549, 1 h after 10 Gy | gamma-H2A.X binding retained, >= 70% | lost, <= 30% |
| 2 | menin W436R | as above | retained, >= 70% | lost, <= 30% |
| neg | menin A411P, L413P | GFP-SUV39H1 | retained (they already are) | retained |
| fold control, every construct | KMT2A MBM co-IP | must be >= 70% of wild type, or the construct is discarded as misfolded | same |
Why the fold control is not optional. Menin is a compact alpha-helical protein and both source mappings are deletion or point mutations inside its core. A W436R substitution replaces a buried tryptophan on a helix that contacts histone H3 in 8GPN; a 30-residue internal deletion removes a loop whose flanking helix is at the cavity rim. Either could destabilise the fold, and the published experiments have no fold control. If the KMT2A co-IP also fails, the construct proves nothing about any specific partner.
Add one cheap third arm if the reagents allow: the same panel plus a 4 h pre-treatment with 1 uM revumenib. That converts the experiment from a geometry test into a drug-relevance test and answers the question risk-a-menin-degrader-is-not-a-pocket-inhibitor-for-the-hr-scaffold is built on.
| design | Two reciprocal co-immunoprecipitation arms in HEK-293T (arm 1) and A549 or 293T (arm 2), transiently transfecting Myc- or FLAG-tagged menin constructs: wild type, delta-371-400, D418N, W436R, A411P, L413P. Arm 1 co-transfects GFP-SUV39H1 and blots GFP after anti-tag IP. Arm 2 irradiates at 10 Gy, harvests at 1 h, and blots endogenous gamma-H2A.X after anti-tag IP. Every construct is additionally tested for KMT2A MBM co-IP as a fold control and discarded if below 70% of wild type. Optional third arm: the whole panel with 4 h 1 uM revumenib pre-treatment. |
|---|---|
| readout | co-IP band intensity normalised to input and expressed as a percentage of wild-type menin in the same blot; three biological replicates; report each construct's KMT2A fold-control value alongside |
| controls | menin A411P and L413P as SvID mutants known NOT to impair SUV39H1 binding (internal negatives); KMT2A MBM co-IP on every construct as a fold control with a pre-specified discard threshold; IgG and untransfected lanes; for arm 2, an unirradiated matched sample, since the gamma-H2A.X interaction is damage-induced |
| cost tier | low |
| timeline | 4-6 weeks with existing plasmids; 8-10 weeks if constructs must be resynthesised, which is a two-gene synthesis order |
| system | HEK-293T and A549 - deliberately the systems the two source papers used, so a negative cannot be blamed on the cell line. A leukaemia arm (MV4-11 or SEM) is the obvious follow-up and is NOT part of the decisive test, because neither original result exists in a haematopoietic cell. |
| decision value | This is the only experiment in this beat that changes an interpretation the graph currently relies on. A 'separate' result restores the SUV39H1 entry's protective reading on a residue basis; a 'one surface' result means a menin ligand or competing partner at 360-445 disturbs both the H3K9me3 delivery arm and the HR arm, and the entry currently used to protect the menin-interface hypothesis is doing the opposite work. |
| who could run it | either source laboratory - Cho/Youn (Sungkyunkwan/Seoul National) for arm 1's assay, Jin/Gao (Xiamen) for arm 2's - or any laboratory with menin expression plasmids, which is most menin-inhibitor groups |
- A negative result in 293T does not travel to a KMT2A-rearranged blast, and this graph's commonest documented failure is exactly that generalisation. The experiment settles the structural question and not the clinical one.
- If the revumenib arm shows reduced gamma-H2A.X co-IP, the immediate clinical question - whether menin inhibitors impair repair in children receiving concurrent TOP2 poisons - becomes urgent and is not answerable from any existing trial dataset, because nobody collected a repair readout.
- Overexpression co-IP is the assay both original claims rest on, so reproducing it reproduces its weaknesses; the experiment is designed to settle a disagreement between two overexpression results, not to establish either as physiological.
- Should the revumenib arm be run first? It is the same blot with one extra condition and it answers the clinically relevant question directly, whereas the geometry question is of interest mainly to this graph.
- Is 70% the right fold-control threshold? It is a judgement; the defensible alternative is to require the KMT2A co-IP to be statistically indistinguishable from wild type, which is stricter and will discard more constructs.
- Should a thermal-shift assay replace the KMT2A co-IP as the fold control? It is more direct and needs purified protein, which most laboratories doing co-IP do not have.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin delta-371-400 is an existing, validated construct: it abolishes gamma-H2A.X binding, fails to reduce IR-induced gamma-H2A.X foci on reconstitution into Men1-knockdown A549 cells, and fails to rescue HR after Men1 siRNA - so arm 1 requires no new reagent design.
The tumor suppressor menin is a key scaffold mediator for homologous recombination repair. · Lian A, Zhu X, Jin B, Gu G, Xu B, Zhang L, Lin X, Zhu X, Jin G · Nucleic Acids Res 2026What this evidence cannot show
The paper does not report a fold control or a KMT2A-binding control for delta-371-400, so its own validation establishes loss of function rather than retention of structure. Full text read 2026-09-22.
Evidence that argues against this (1)
- If delta-371-400 is partly misfolded, arm 1 returns a false 'one surface' answer. That is why the KMT2A co-IP control is specified as a discard criterion rather than a reported covariate. Huang J et al. 2012
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Menin A411P, L413P, D418N and W436R are existing constructs tested side by side in one assay, with D418N and W436R impaired in SUV39H1 binding and A411P and L413P not - so arm 2 has both its test articles and its internal negatives already built and published.
Menin mediates epigenetic regulation via histone H3 lysine 9 methylation. · Yang YJ, Song TY, Park J, Lee J, Lim J, Jang H, Kim YN, Yang JH, Song Y, Choi A, Lee HY, Jo CH, Han JW, Kim ST, Youn HD, Cho EJ · Cell Death Dis 2013What this evidence cannot show
The four mutants were assayed by in-vitro translation of 35S-menin against partially purified FLAG-SUV39H1, read by autoradiography - a semi-quantitative readout with no replicate statistics stated in the readable text and no fold control. The effect sizes in this experiment's falsifiers are therefore specified against a fresh wild-type reference rather than against the 2013 band intensities. Full text read 2026-09-22.
Evidence that argues against this (1)
- UniProt records W436R as having 'no effect on histone methylation; almost no effect on JUND-binding; modest repression of JUND transactivation' from four independent studies, so this substitution is not globally destabilising - which is reassuring for the fold-control concern and simultaneously makes the 2013 SUV39H1 loss more specific and more interesting. UniProt Consortium 2026
the two co-IPs that would settle the adjudication
does menin delta-371-400 still bind SUV39H1?
do menin D418N and W436R still bind gammaH2A.X after 10 Gy?