Tag endogenous KMT2B in an index-fusion line and measure what revumenib does to it - the experiment the reagents have blocked
also called KMT2B CUT&RUN under menin inhibition · endogenous KMT2B knock-in
Nobody has ever watched what a menin inhibitor does to KMT2B on the DNA, because the antibodies do not work. Putting a small tag onto the gene itself in a leukaemia cell line removes that obstacle entirely.
The same thing, for a scientist
CRISPR knock-in of a tandem epitope tag at the endogenous *KMT2B* locus in SEM (KMT2A::AFF1, infant t(4;11)) and MV4-11 (KMT2A::AFF1, myeloid), with a parallel tag on *KMT2A* in the same cells, followed by CUT&RUN or CUT&Tag time-courses under revumenib. **This is the only design that answers the question**, because the obstacle is explicitly a reagent one: the group that mapped the Menin-KMT2A/B unit stated that commercially available reagents to monitor KMT2B chromatin binding are inadequate, and that their KMT2A-only analyses therefore underestimate the consequences of liberating KMT2A/B after Menin loss [E1]. An endogenous tag bypasses antibody quality, avoids the overexpression that compromises the one quantitative complex dataset [E2], and lets KMT2A and KMT2B be measured with the *same* antibody against the *same* tag - which removes the antibody-efficiency confound that makes two-antibody ChIP comparisons uninterpretable. This graph already carries the template design in `exp-endogenous-tag-and-degron-knock-in-kmt2a-aff1`; this entry applies it to the paralogue.
Design
Lines. line-sem-kmt2a-aff1 (infant-type t(4;11), B-lymphoid - the index system) and line-mv4-11-kmt2a-aff1-myeloid (same fusion, myeloid, and the line most menin-inhibitor pharmacology is done in). A fusion-negative comparator - K-562, to connect directly to the published bivalency work - is the third arm and is what makes the result interpretable for hyp-menin-inhibitor-normal-tissue-effect-is-kmt2b.
Knock-in. Tandem tag (3xFLAG-V5, or a HaloTag if degron follow-on is wanted) at the KMT2B C-terminus by HDR; separately, the same tag at the KMT2A C-terminus in a parallel clone. C-terminal, deliberately: the KMT2B antibody that does work for immunoblot is C-terminal-antigen 1, and a C-terminal tag on KMT2A is not carried by the fusion, which makes the KMT2A clone a clean wild-type-only reporter - the same logic the field already uses with the 180 kDa MLL1 C-terminal antibody 3.
Perturbation. Revumenib at a concentration and schedule matched to the published VTP50469 work 1, with a DMSO arm and a time-course at 0, 2, 6, 24 and 72 hours. The early timepoints matter: displacement is a physical event and 72 hours is downstream biology.
Readout. CUT&RUN against the tag, spike-in normalised, in all arms. Paired RNA-seq at 24 and 72 hours. In the fusion-bearing lines, also anti-menin CUT&RUN so that the fusion's own menin-dependent sites are visible in the same experiment.
What each outcome means, decided in advance
| outcome | reading |
|---|---|
| KMT2B is displaced from active genes with kinetics like KMT2A's, and accumulates at bivalent promoters | the Sparbier redistribution model is a two-enzyme model and their stated underestimate is real; every menin entry in this graph needs KMT2B named |
| KMT2B is not displaced at all | menin-KMT2B is not the load-bearing anchor for KMT2B chromatin binding - consistent with the published observation that the KMT2B CXXC domain can mediate recruitment - and hyp-menin-inhibitor-normal-tissue-effect-is-kmt2b is badly damaged |
| KMT2B is displaced in K-562 but not in SEM | the fusion outcompetes wild-type KMT2B for menin, which would be a genuinely new statement about how a menin inhibitor differs between a fusion-bearing and a normal cell |
| KMT2B displaced, KMT2A not | inverts the field's assumption and would make the existing KMT2A-only literature not merely incomplete but wrong about which enzyme moves |
Controls that decide whether the experiment is believable. A tag-negative parental line processed identically (the CUT&RUN background). A clone with the tag on KMT2A, run in the same batch, so the two enzymes are compared with one antibody. Western blot confirming the tagged allele is expressed at parental levels and that tagging has not destabilised the protein - this is the commonest way an endogenous knock-in silently fails. Sanger and long-read confirmation of a clean single-copy insertion at both alleles.
What could go wrong. KMT2B is 2,715 residues and a C-terminal HDR knock-in at a large locus is not trivial; clone recovery may take two attempts. A tag adjacent to the SET domain could impair catalysis - hence the western and, ideally, a bulk H3K4me3 check in the tagged clone. And CUT&RUN on a low-abundance chromatin protein is itself hard; the K-562 arm doubles as the assay-development arm because that is where the published H3K4me3 CUT&Tag comparison data already exist 1.
| design | CRISPR HDR knock-in of a tandem C-terminal epitope tag at endogenous KMT2B, and separately at endogenous KMT2A, in SEM and MV4-11 (both KMT2A::AFF1) plus fusion-negative K-562; spike-in-normalised CUT&RUN against the tag across a revumenib time-course at 0, 2, 6, 24 and 72 h, with paired RNA-seq at 24 and 72 h and anti-menin CUT&RUN in the fusion-bearing arms. |
|---|---|
| readout | genome-wide KMT2B occupancy change under menin inhibition, measured with the same antibody as tagged KMT2A; redistribution from active to bivalent promoters; comparison of fusion-bearing against fusion-negative cells |
| controls | tag-negative parental line processed identically; paired KMT2A-tagged clone in the same batch and same antibody; western confirming parental-level expression of the tagged allele; bulk H3K4me3 check that tagging has not impaired catalysis; long-read confirmation of clean biallelic insertion; DMSO arm at every timepoint |
| cost tier | med |
| timeline | 6-9 months, dominated by clone generation at a large locus |
| prerequisite | check first whether a chromatin-grade KMT2B antibody has been validated since 2023; if one exists this becomes a three-month experiment |
| template entry | exp-endogenous-tag-and-degron-knock-in-kmt2a-aff1 |
cell lines
[ "line-sem-kmt2a-aff1", "line-mv4-11-kmt2a-aff1-myeloid", "K-562 (fusion-negative comparator)" ]
tests
[ "hyp-menin-inhibitor-normal-tissue-effect-is-kmt2b", "mechanism-menin-binds-two-kmt2-proteins-not-one" ]
- A null result is genuinely ambiguous here. Menin-independent CXXC/CpG recruitment is documented, so 'KMT2B does not move under the drug' is compatible both with the drug not engaging menin-KMT2B and with menin-KMT2B engagement having no consequence for occupancy.
- Tagging a 2,715-residue chromatin enzyme at its C-terminus, next to the SET domain, risks measuring a subtly crippled protein - and the controls that would detect that (bulk H3K4me3, expression level) are insensitive to a partial defect.
- Cell lines are not the tissue the hypothesis is about. The normal-tissue question this experiment serves ultimately needs a primary human progenitor or a mouse tissue, and a K-562 arm is the weakest possible proxy for a child's islets or neurons.
- Has a chromatin-grade KMT2B antibody been validated since the 2023 statement? This is a catalogue search and it decides whether the experiment costs three months or nine.
- Should the tag be a degron? A HaloTag or dTAG knock-in would make the same clone usable for acute KMT2B depletion, which is the separate experiment nobody has done in a KMT2A-rearranged cell and which would cost almost nothing extra at the design stage.
- Is SEM the right index line? It is the graph's index t(4;11) line but it is from a five-year-old, not an infant; PER-485, PER-490 or an infant PDX would be closer to the disease and harder to engineer.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Commercially available reagents to monitor KMT2B binding at chromatin are inadequate, so analyses based only on KMT2A accumulation at bivalent genes likely underestimate the consequences of liberating KMT2A/B from active genes after Menin loss; the same study treated K-562 with VTP50469, performed H3K4me3 CUT&Tag in control and KMT2A/B knockout cells, and used a C-terminal-antigen KMT2B antibody for immunoblot.
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. The reagent statement is one group's assessment in 2022-2023 and a new chromatin-grade antibody could make the knock-in unnecessary; that should be checked before building clones. The published system is K-562 with VTP50469, not a KMT2A-rearranged line with revumenib, so the proposed experiment is an extension rather than a replication.
Evidence that argues against this (1)
- The same study's genetic data suggest KMT2B's contribution at these loci is smaller than KMT2A's - KMT2B knockout alone only modestly impaired IFN-gamma-induced MHC-I while the double knockout impaired it strongly - so the experiment may return a real but small effect, which spike-in-normalised CUT&RUN on a low-abundance protein is poorly powered to distinguish from none. Chen Y et al. 2019
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The one quantitative dataset assigning menin to MLL1 and MLL2 complexes used GFP-tagged catalytic subunits expressed from cDNA in HeLa, so its conclusions rest on overexpressed constructs in an epithelial line - which an endogenous knock-in in a leukaemia line would replace with a physiological measurement.
Quantitative dissection and stoichiometry determination of the human SET1/MLL histone methyltransferase complexes. · van Nuland R, Smits AH, Pallaki P, Jansen PW, Vermeulen M, Timmers HT · Mol Cell Biol 2013What this evidence cannot show
Full text read via the PMC HTML route (PMC3647974) on 2026-09-20. This is a methodological criticism and not a claim that the dataset is wrong; label-free quantification with stoichiometry was a considerable advance and remains the best complex-composition measurement in this family.
Evidence that argues against this (1)
- An endogenous tag introduces its own artefact: a C-terminal tag sits near the SET domain and could impair complex assembly or catalysis, in which case the physiological measurement is of a slightly non-physiological protein. The overexpression criticism does not make the knock-in automatically better, only differently limited. Sparbier CE et al. 2023
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
An antibody detecting only the 180 kDa MLL1 C-terminal fragment cannot detect MLL-AF9 and was used to show that wild-type MLL1 and WDR5 co-occupy H3K4me1/2-enriched enhancers while MLL-AF9 binds regions with no H3K4me1 enrichment - establishing the precedent that a C-terminal epitope isolates the wild-type allele from the fusion.
MLL1 and MLL1 fusion proteins have distinct functions in regulating leukemic transcription program. · Xu J, Li L, Xiong J, et al. · Cell Discov 2016What this evidence cannot show
Abstract-level use in this entry: the identifiers were re-fetched from PubMed on 2026-09-20 rather than copied from the sibling entry in this graph that also cites this paper. The design precedent is the relevant part; the biological claim is contested by a 2024 study reporting identical MLL1 and MLL-AF9 occupancy, which this graph records elsewhere.
Evidence that argues against this (1)
- A C-terminal tag on KMT2A reports only the wild-type allele, which is a strength for this design and a limitation for interpretation: the experiment will not see the fusion's own chromatin behaviour unless anti-menin or an N-terminal reagent is added, which is why the proposed design includes menin CUT&RUN. Sparbier CE et al. 2023
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
Menin-independent KMT2B chromatin recruitment exists as a documented alternative: in mouse embryonic stem cells the CXXC domain of KMT2B can mediate recruitment to chromatin, and CpG islands can drive KMT2A/B recruitment - so a null result in this experiment has a ready mechanistic explanation rather than being uninformative.
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
This statement appears in the discussion of the cited paper as a summary of prior work in mouse ES cells, with the primary sources not identified here. It is cited as a pre-registered alternative explanation for a possible null, not as an established fact about human leukaemia cells.
Evidence that argues against this (1)
- If CXXC/CpG binding dominates KMT2B recruitment, then the menin interface may be dispensable for occupancy while still mattering for complex composition or activity - in which case chromatin occupancy is the wrong readout and a composition readout (tag co-IP under drug) would be needed instead. Bach C et al. 2009
the cellular arm; fusion-bearing versus fusion-negative comparison is the discriminating design
index t(4;11) line for the KMT2B tag knock-in
bypasses the stated chromatin-reagent gap that blocks every other route