Oncofusion

start / graph / ana-q140-the-kmt2a-n-terminus-and-menin-are-each-others-nearest-neighbour-in-sem-with-no-peak-calling-and-no-target-list

Analysismoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q140-sem-menin-arm

In an unsupervised 5 kb correlation of 41 SEM libraries, both KMT2A N-terminal AutoCUT&RUN libraries have menin as their single nearest neighbour - the menin-interface hypothesis appearing as the top-level structure of an analysis that was not looking for it

We grouped 41 experiments on infant-leukaemia cells purely by how similar their data looked, with no gene lists and no assumptions. The experiments that tracked the front half of the leukaemia-driving protein came out closest to the experiments that tracked menin, the partner it is supposed to hold on to - which is what the leading hypothesis predicts, arrived at without asking for it.

The same thing, for a scientist

In the 41-library SEM correlation matrix of GSE159608, **GSM4842309 (KMT2A N-terminal, AutoCUT&RUN) and GSM4842311 (KMT2A N-terminal, replicate)** both have **GSM5139719 (menin AutoCUT&Tag)** as their nearest neighbour, at r=0.396 and r=0.634. GSM4842307 (KMT2A C-terminal) also nearest-neighbours to menin (r=0.443); GSM4842305 (KMT2A C-terminal replicate) goes to H3K4me3 (r=0.540). The KMT2A libraries are AutoCUT&RUN from a 2020-03-31 batch and the menin library is AutoCUT&Tag from 2021-01-26 - **different protocol, different run, ten months apart**, so the co-location is not a batch effect. No peaks were called, no promoter set was used, and the correlation is over uniform 5 kb bins across chr1-22,X.

Go deeper
Why this is worth an entry when the interface is already established

The menin-KMT2A contact is established biochemistry and structural biology. What is usually asserted rather than shown is that it governs where the protein sits genome-wide in a human infant-derived index-fusion line. Most occupancy evidence for that is either in adult-derived lines, in mouse, or at a curated target list. Here it falls out of an unsupervised whole-genome analysis in SEM, which is the cell type the question is about. 12

What it does not show

Correlation at 5 kb between a bait and a partner does not establish co-occupancy of the same molecule at the same site, and it does not establish dependence. Menin and the KMT2A N-terminus could co-distribute because both track active promoters. The test that separates these is the drug arm, and it exists: under VTP-50469 the H3K4me3 signal at menin-enriched sites falls selectively (0.766) and at KMT2A-N-enriched sites falls in the same direction (0.871). Correlation plus a perturbation is a much stronger claim than either alone, and this deposit supplies both in the same cell line. 1

The N/C asymmetry, honestly stated

The N-terminal libraries behave consistently (both to menin); the C-terminal ones split. That is the expected direction - the N-terminal fragment carries the menin-binding motifs and is the part retained in the fusion - but with two libraries per terminus this is an observation, not a measurement of asymmetry. The C-terminal contrast fails its own IgG control and is reported as uninterpretable (see ana-q140-the-kmt2a-c-terminal-site-class-fails-its-own-igg-control).

Facts and figures
scriptanalyses/sem-arm-identity-from-signal/run.py
statusrun
menin replicate pair r0.749
protocol noteKMT2A AutoCUT&RUN 2020-03-31; menin AutoCUT&Tag 2021-01-26
unsupervisedtrue
n libraries41
pairs
GSM4842309 (KMT2A N-term)0.396GSM4842309 (KMT2A N-term): 0.396GSM4842307 (KMT2A C-term)0.443GSM4842307 (KMT2A C-term): 0.443GSM4842305 (KMT2A C-term)0.540GSM4842305 (KMT2A C-term): 0.54GSM4842311 (KMT2A N-term)0.634GSM4842311 (KMT2A N-term): 0.634
pairs — the numbers behind the chart
{
  "GSM4842309 (KMT2A N-term)": {
    "nearest": "GSM5139719 (menin)",
    "r": 0.396
  },
  "GSM4842311 (KMT2A N-term)": {
    "nearest": "GSM5139719 (menin)",
    "r": 0.634
  },
  "GSM4842307 (KMT2A C-term)": {
    "nearest": "GSM5139719 (menin)",
    "r": 0.443
  },
  "GSM4842305 (KMT2A C-term)": {
    "nearest": "GSM4842299 (H3K4me3)",
    "r": 0.54
  }
}
How this could bite us back
  • If the N-terminal signal is substantially wild-type KMT2A rather than the fusion, a menin inhibitor's effect on it is an effect on normal KMT2A function, which is the toxicity axis rather than the efficacy axis.
  • A correlation this modest (r 0.40-0.63) would not survive being promoted to 'menin determines fusion localisation'. Treated as such it would license a biomarker that is really measuring promoter activity.
Still open
  1. Does the N-terminal / menin co-distribution survive at 500 bp bins and restricted to non-promoter sites, where a shared CpG-island preference cannot explain it?
  2. Menin CUT&Tag under VTP-50469 in SEM would convert this correlation into a dependency claim in one experiment and does not exist (see the bounded-absence entry).
  3. In SEM the KMT2A N-terminal antibody cannot distinguish the fusion from wild-type KMT2A's N-terminus. What fraction of this signal is the fusion?
Every source behind this entry (3)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    Both SEM KMT2A N-terminal AutoCUT&RUN libraries have a SEM menin AutoCUT&Tag library as their nearest neighbour in an unsupervised 5 kb genome-wide correlation across 41 libraries.

    What this evidence cannot show

    r=0.396 and 0.634 are modest; nearest-neighbour identity is a weak statistic and a matrix with 41 members has many ways to produce a given pairing by chance. n=2 libraries per terminus. The claim is about co-distribution at 5 kb, not co-occupancy.

    Evidence that argues against this (1)
    • Both menin and the KMT2A N-terminus concentrate at CpG-island promoters, so a shared promoter preference could produce this pairing with no direct relationship. The deposit cannot separate these by correlation alone; only the drug arm can, and it is available for H3K4me3 and not for menin itself. Krivtsov AV et al. 2019
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset

    The KMT2A antibodies in this deposit are separately directed at the N- and C-terminal portions, which is what makes the N/C distinction available at all; the series' stated purpose is mapping fusion occupancy regardless of partner.

    GSE159608: Automated CUT&Tag profiling of chromatin heterogeneity in mixed-lineage leukemia (419 samples, hg19 fragment BEDs) · Janssens DH, Meers MP, Wu SJ, Babaeva E, Meshinchi S, Sarthy JF, Ahmad K, Henikoff S · NCBI Gene Expression Omnibus 2021
    What this evidence cannot show

    Read from the sample titles and characteristics (KMT2A_N1/N2, KMT2A_C1/C2, AutoCUT&RUN, batch 20200331_JS). Epitope-level validation of those antibodies was not performed by us.

    Evidence that argues against this (1)
    • The N/C distinction is available in the antibodies but is NOT clean in the biology: wild-type KMT2A is cleaved by taspase 1 into N- and C-terminal fragments that remain associated, so an N-terminal antibody reports fusion plus wild-type N-fragment and a C-terminal antibody reports wild-type only. Similarity to menin therefore cannot distinguish the fusion from wild-type KMT2A, and this deposit contains no fusion-specific reagent.
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The same group's later work shows the oncoprotein and menin are reduced together in a sample that lineage-switched under revumenib, while ENL persists at oncoprotein target loci - direct evidence that menin and the KMT2A fusion track together on chromatin and can be separated from other cofactors.

    MLL oncoprotein levels influence leukemia lineage identities · Janssens DH, Duran M, Otto DJ, Wu W, Xu Y, Kirkey D, Mullighan CG, Yi JS, Meshinchi S, Sarthy JF, Ahmad K, Henikoff S · Nature Communications 15:9341 2024
    PMID 39472576DOI
    What this evidence cannot show

    A single lineage-switching patient sample; the observation is descriptive and not a controlled perturbation.

    Evidence that argues against this (2)
    • That result is a SINGLE patient sample that lineage-switched under revumenib - the most selected context available - and co-reduction of two proteins in cells that are changing identity is expected whether or not they are on chromatin together. It is consistent with co-occupancy and does not establish it.
    • This entry's own beat found the opposite coupling in the same cycle: q141 measured that a DOT1L inhibitor removes menin from fusion promoters about as well as destroying the fusion does (0.467 vs 0.437). If menin can be stripped without touching the fusion, menin-and-fusion 'tracking together' is a weaker inference than it appears.
Connected to