Measure the menin-KMT2B affinity nobody has measured, and the four clinical inhibitors' IC50 against it
also called menin-KMT2B Kd · KMT2B MBM peptide competition
Nobody has ever measured how tightly menin grips KMT2B, or how well the leukaemia drugs block that grip. The measurement is a standard one and would take weeks.
The same thing, for a scientist
A fluorescence-polarisation or ITC titration of menin against a synthetic KMT2B N-terminal peptide, run head-to-head against the KMT2A MBM1 peptide whose Kd is approximately 10 nM [E1], followed by competition IC50 determination for revumenib, ziftomenib, bleximenib and enzomenib against both peptides. **The reagents all exist**: recombinant menin is a standard construct from two decades of this chemistry, the KMT2A MBM1 assay is the assay those compounds were optimised in, and the KMT2B peptide is a synthesis order. The only new element is the peptide sequence, which this beat derived from the fetched UniProt canonical sequence: KMT2B aa 12-30 spanning R17-G18-R19-F20-P21-G22-R23-P24, aligning to KMT2A R6-W7-R8-F9-P10-A11-R12-P13 with the alanine-scan-critical Phe/Pro/Pro conserved and Trp7 replaced by Gly [E2][E3]. **Why it has not been done**: the field's interest in KMT2B is recent, the therapeutic question was posed as efficacy rather than toxicity, and the answer was assumed from the complex-level pulldown [E4].
Design
Peptides. Synthesise four, all N-terminally fluorescein-labelled for FP plus unlabelled counterparts for ITC and competition:
- KMT2A MBM1, aa 1-20 (positive control, Kd approximately 10 nM known 1)
- KMT2B aa 8-30, containing the RGRFPGRP motif (the test)
- KMT2B aa 8-30 F20A (negative control; if Phe20 is the KMT2A Phe9 equivalent, this should abolish binding)
- KMT2B aa 8-30 G18W (the reversion; asks directly whether the Trp-to-Gly change is what separates the two affinities)
Peptide 4 is the informative one and the reason to run this rather than assume. It converts a sequence observation into a causal test of a single residue.
Assay 1 - direct binding. FP titration of recombinant human menin against each labelled peptide; confirm the two strongest by ITC, which gives stoichiometry and is not subject to the fluorophore artefacts FP is.
Assay 2 - competition. Displacement of labelled KMT2A MBM1 and of labelled KMT2B peptide by revumenib, ziftomenib, bleximenib and enzomenib across a concentration series; report IC50 for each compound against each peptide and the ratio. The ratio is the deliverable.
Assay 3 - orthogonal, optional. Menin pulldown from a fusion-negative human cell lysate with and without each compound, blotted for KMT2A and KMT2B with the C-terminal-antigen KMT2B antibody the Sparbier study used for immunoblot (chromatin reagents are inadequate but lysate detection works) 4. This tests whether the peptide result predicts displacement of the full-length protein.
Readout, controls and interpretation
Primary readout. Kd(menin-KMT2B) / Kd(menin-KMT2A), and IC50 ratios for four compounds.
Interpretation rule, set before the experiment:
| result | conclusion |
|---|---|
| KMT2B Kd within 10-fold of KMT2A | the two-target picture holds at the interface; hyp-menin-inhibitor-normal-tissue-effect-is-kmt2b survives and every menin entry in this graph needs a second partner named |
| 10-100 fold weaker | partial engagement; the arm matters at high exposure and the trial pharmacokinetics decide |
| more than 100-fold weaker | the hypothesis is falsified and the complex-level pulldown result must be explained by an avidity or second-site contact rather than by an MBM |
| F20A abolishes, G18W restores KMT2A-like affinity | Trp7 is the discriminating residue and the whole family's menin selectivity has a one-residue explanation |
Controls. JUND peptide (binds the same menin pocket 5) as a specificity control; KMT2A MBM1 F9A as the established negative; a scrambled KMT2B peptide. Run menin preparations from two sources, because this pocket's literature has a history of construct-dependent results.
What this cannot do. It is peptide biochemistry. It cannot tell you occupancy in a cell, and the composite menin-MLL1 surface that binds LEDGF 5 means a peptide Kd is not the whole interface even for KMT2A. Assay 3 partially bridges that and exp-kmt2b-endogenous-tag-occupancy-under-menin-inhibitor is the cellular successor.
| design | Fluorescence-polarisation and ITC titration of recombinant human menin against synthetic KMT2A MBM1 and KMT2B N-terminal peptides plus F20A and G18W variants, followed by competition IC50 determination for revumenib, ziftomenib, bleximenib and enzomenib against both peptides; optional menin co-IP from fusion-negative lysate blotted for KMT2A and KMT2B. |
|---|---|
| readout | Kd(menin-KMT2B)/Kd(menin-KMT2A) and per-compound IC50 ratios; the effect of F20A and of the G18W reversion |
| controls | KMT2A MBM1 (positive, Kd ~10 nM), KMT2A MBM1 F9A (established negative), JUND peptide (same-pocket specificity), scrambled KMT2B peptide, two independent menin preparations |
| cost tier | low |
| timeline | 6-10 weeks including peptide synthesis; step zero is a full-text read of PMID 30315824 and its references, which could find the data already published |
| decision rule | Kd within 10-fold: hypothesis survives; 10-100 fold: exposure-dependent; >100-fold weaker: hypothesis falsified |
peptides
[
{
"name": "KMT2A MBM1",
"span": "aa 1-20",
"note": "positive control"
},
{
"name": "KMT2B MBM-like",
"span": "aa 8-30, contains R17-G18-R19-F20-P21-G22-R23-P24",
"note": "the test"
},
{
"name": "KMT2B F20A",
"span": "aa 8-30",
"note": "tests whether Phe20 is the KMT2A Phe9 equivalent"
},
{
"name": "KMT2B G18W",
"span": "aa 8-30",
"note": "reversion to the KMT2A Trp; the single-residue causal test"
}
]tests
[ "hyp-menin-inhibitor-normal-tissue-effect-is-kmt2b" ]
- A negative result here does not exonerate the drug class in normal tissue: menin has many non-KMT2 partners and the pocket is promiscuous, so ruling out KMT2B narrows the question without closing it.
- A positive result creates a problem with no engineering solution, because the compounds bind the menin pocket and the pocket is shared. The honest response would be a KMT2B-sparing strategy that does not exist.
- Peptide biochemistry has repeatedly disagreed with cellular measurement in this exact system - the LEDGF-prefers-KMT2B result came from cells and conflicts with fragment-based work - so this experiment should be scoped as one input, not as the answer.
- Do the pharmaceutical sponsors already have this data? Menin-KMT2B counter-screening would be routine selectivity work for a clinical programme and may sit unpublished in four companies' files. Asking is cheaper than measuring.
- Is the canonical KMT2B N-terminus the one expressed in the relevant tissues? Isoform usage would move the motif and this experiment assumes the canonical record.
- Should the same assay be run for SETD1A and SETD1B? The pulldown says menin is absent from those complexes, and a peptide panel would convert a negative pulldown into a quantitative selectivity profile across the whole family for the price of two more syntheses.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
MLL binds menin with a Kd of approximately 10 nM through MBM1 and MBM2 within the N-terminal 43 amino acids, MBM1 being the high-affinity motif, with alanine scanning identifying Phe9, Pro10 and Pro13 as most critical and an MBM1 peptide efficiently dissociating the menin-MLL complex in competition experiments.
Molecular basis of the mixed lineage leukemia-menin interaction: implications for targeting mixed lineage leukemias. · Grembecka J, Belcher AM, Hartley T, Cierpicki T · J Biol Chem 2010What this evidence cannot show
Abstract fetched and read 2026-09-20. This is the methodological template for the proposed experiment and it is entirely a KMT2A study - KMT2B appears nowhere in it. NMR and peptide competition on a disordered fragment; the reported Kd is for the peptide, not the full-length protein.
Evidence that argues against this (1)
- Peptide affinity systematically overstates the isolated contribution of a linear motif when the real interface is composite. For menin-MLL1 the LEDGF-binding surface is formed jointly by menin and MLL1, so even the 10 nM benchmark is an incomplete description of the interaction this experiment would compare against. Huang J et al. 2012
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt canonical sequences fetched 2026-09-20 give KMT2A beginning MAHSCRWRFPARPG (MBM1 residues F9, P10, P13) and KMT2B containing R17-G18-R19-F20-P21-G22-R23-P24, so the KMT2B peptide for this experiment is defined by fetched sequence and not by recall.
UniProtKB canonical sequences Q03164 (KMT2A, 3969 aa) and Q9UMN6 (KMT2B, 2715 aa), fetched 2026-09-20 · UniProt Consortium; sequences fetched and aligned by this project · Database 2026What this evidence cannot show
The sequences are database facts; the alignment and the choice of peptide boundaries are this project's. Isoform choice matters: the canonical KMT2B record is used and an alternative N-terminal isoform would shift the numbering. A synthesiser should be given the sequence string, not the residue numbers.
Evidence that argues against this (1)
- If the physiological menin contact on KMT2B is not this motif, the experiment measures the wrong peptide and returns a clean, wrong negative. The complex-level pulldown that motivates the whole question does not localise the contact, so there is no independent evidence that aa 17-24 is where menin binds. van Nuland R et al. 2013
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Menin-MLL interaction compounds inhibit menin-MLL1 and menin-MLL2 as well as menin-fusion in biochemical and structural assays, according to the study that otherwise concludes the predominant antileukaemic action is direct fusion inhibition - so the compounds' KMT2B engagement is asserted in the literature without a published IC50.
Distinct pathways affected by menin versus MLL1/MLL2 in MLL-rearranged acute myeloid leukemia. · Chen Y, Jones KL, Anastassiadis K, et al. · Exp Hematol 2019What this evidence cannot show
Abstract fetched and read 2026-09-20; full text not obtained, so the primary sources behind the polypharmacology statement are not identified here and it is possible a published menin-MLL2 IC50 exists in one of them. **A full-text read of this paper is step zero of this experiment** and could conceivably make it unnecessary.
Evidence that argues against this (1)
- If the menin-MLL2 inhibition data already exist in the medicinal-chemistry literature behind this statement, this experiment is a literature search rather than a bench project. That possibility is not excluded and is the cheapest thing to check first. Grembecka J et al. 2010
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A chromatin-grade KMT2B reagent gap is stated in the literature, while a KMT2B antibody against a C-terminal antigen (Cell Signaling 63735, clone D6X2E) was used successfully for immunoblot in the same study - so lysate-level detection of KMT2B, as proposed in assay 3, is feasible where chromatin occupancy is not.
Targeting Menin disrupts the KMT2A/B and polycomb balance to paradoxically activate bivalent genes. · Sparbier CE, Gillespie A, Gomez J, et al. · Nat Cell Biol 2023What this evidence cannot show
Full text and methods read via Europe PMC fullTextXML (PMC7614190) on 2026-09-20. That an antibody works for denatured immunoblot says nothing about whether it works for native immunoprecipitation, which is what assay 3 requires; assay 3 should be regarded as exploratory and its failure should not be read as a biological negative.
Evidence that argues against this (1)
- Assay 3 has the reagent problem the whole field has. If the pulldown fails, the experiment still delivers assays 1 and 2, which is why the peptide biochemistry is the core of this proposal and the cellular arm is optional. Chen Y et al. 2019
- E5established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper
Crystal structures of human menin free and in complex with MLL1, with JUND, and with an MLL1-LEDGF heterodimer show that menin contains a deep pocket binding short peptides of MLL1 or JUND in the same manner, with LEDGF held at a distinct surface formed by both menin and MLL1.
The same pocket in menin binds both MLL and JUND but has opposite effects on transcription. · Huang J, Gurung B, Wan B, et al. · Nature 2012What this evidence cannot show
Abstract fetched and read 2026-09-20. Provides the JUND specificity control for this experiment and establishes that the pocket is promiscuous for short motifs - which cuts both ways, since a pocket that binds two unrelated peptides may bind a KMT2B motif without that meaning much biologically.
Evidence that argues against this (1)
- If the menin pocket accepts short motifs relatively promiscuously, a nanomolar KMT2B peptide Kd would be less informative than it looks, and the cellular experiment becomes necessary rather than optional. Sparbier CE et al. 2023
the measurement that decides the structural possibility
Kd and IC50 ratios; >100-fold weaker KMT2B binding falsifies the hypothesis
runs the KMT2A MBM1 peptide as the head-to-head positive control