Oncofusion

start / graph / finding-irx1-binds-the-mll-af4-complex-without-displacing-menin

Mechanismmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q145-irx1-in-hoxa-negative-infants

IRX1 binds the MLL-AF4 complex non-competitively — menin, LEDGF and DOT1L are all unchanged

IRX1 shuts down the leukaemia protein's output not by knocking its partners off, but by sitting alongside them. Menin, the drug target, stays exactly where it was.

The same thing, for a scientist

Kühn et al. immunoprecipitated MLL-AF4 from HEK293T cells with and without co-expressed IRX1 and blotted for known complex members: **"No changes could be seen for MEN1, LEDGF and DOT1L"**, and BMI-1 and CYP33 — the factors that repress wild-type MLL — were absent in both conditions. IRX1 co-precipitated with MLL-AF4, and the composition and amount of bound proteins were unchanged, so **"IRX1 binds to MLL-AF4 in a non-competitive manner"** [E1]. This is directly load-bearing for the menin-interface hypothesis: it identifies an endogenous factor that suppresses the fusion's transcriptional output **without acting through the menin interface at all**.

Go deeper
Why this is the menin-relevant result in the IRX1 literature

The menin-interface hypothesis holds that the fusion's oncogenic output depends on the menin-KMT2A contact, and that breaking that contact is therefore sufficient to collapse the program. IRX1 is a counter-example in the mechanism, not in the outcome: it collapses HOXA transcription while leaving menin in the complex 1. That does not refute the hypothesis — two mechanisms can both work — but it establishes that menin retention is compatible with a dead HOXA program, which is the same conclusion this beat reaches from a completely different direction at chromatin.

What would refute the non-competitive model: a co-IP or proximity assay in a t(4;11) leukaemia cell (not HEK293T) showing menin reduced in the fusion complex when IRX1 is present, or a menin CUT&RUN in an iALL-IRX1 blast showing menin absent from the fusion's remaining targets. Neither has been done; see exp-menin-cut-and-run-in-an-irx1-high-infant-t411-blast.

What the beat could not test

GSE252378 contains no menin track in any KMT2A::AFF1 sample. Its only menin tracks are in the menin-inhibitor-treated KMT2A::ENL pair. So nothing measured in this beat bears on menin occupancy in the IRX1 subgroup, and the falsifier above is registered as unrunnable on existing public data 2.

Facts and figures
systemHEK293T, doxycycline-inducible Sleeping Beauty transgenes
baitMLL-AF4-mCherry, anti-mCherry IP
interpretationnon-competitive binding of IRX1 to the MLL-AF4 complex
unchanged on irx1 binding
[
  "MEN1",
  "LEDGF (PSIP1)",
  "DOT1L"
]
absent in both conditions
[
  "BMI-1",
  "CYP33"
]
How this could bite us back
  • An endogenous non-competitive suppressor of the fusion is a resistance mechanism in reverse: if IRX1 expression is what makes those blasts HOXA-dead, then a menin inhibitor arriving later has less left to collapse and may look inactive for reasons that have nothing to do with target engagement.
  • Kühn's own therapeutic suggestion is an HDAC inhibitor, on the basis that TSA blocks both IRX1 and MLL-AF4 - but TSA is a pan-HDACi in HEK293T and that is a long way from a clinical proposal.
Still open
  1. Is the IRX1-MLL-AF4 contact direct, and if so which surface? No structure, no mapped interface, no affinity measurement exists.
  2. If IRX1 suppresses fusion output without touching menin, does an iALL-IRX1 blast respond to a menin inhibitor at all? Nothing in the literature or this corpus answers it.
Every source behind this entry (2)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Co-immunoprecipitation of MLL-AF4 with and without IRX1 shows MEN1, LEDGF and DOT1L unchanged in amount and composition, with IRX1 co-precipitating - i.e. IRX1 binds MLL-AF4 non-competitively.

    The IRX1/HOXA connection: insights into a novel t(4;11)-specific cancer mechanism · Kuhn A, Loscher D, Marschalek R · Oncotarget 2016
    PMID 27175594DOI
    What this evidence cannot show

    HEK293T with transfected mCherry-tagged MLL-AF4 and inducible IRX1; precipitation was via anti-mCherry, so this is a tagged-bait experiment in an embryonic kidney line, not endogenous fusion in a leukaemia cell. Western blot band intensities were reported qualitatively ('no changes could be seen'), not quantified. Full text fetched and read this cycle.

    Evidence that argues against this (1)
    • A Western blot read as unchanged can miss a two-fold change in bound menin, which would be biologically large at a stoichiometric interface; the experiment as reported cannot exclude it. Kuhn A et al. 2016
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    GSE252378 contains no menin CUT&RUN in any KMT2A::AFF1 sample; its menin tracks are confined to the menin-inhibitor-treated KMT2A::ENL pair.

    What this evidence cannot show

    Established by reading the series matrix characteristics for all 171 samples. It is an absence in one deposit, not in the literature - a menin ChIP in t(4;11) material may exist elsewhere and was not exhaustively searched.

    Evidence that argues against this (1)
    • If a menin ChIP in primary t(4;11) blasts exists in another deposit, the unrunnable falsifier becomes runnable immediately and this entry's framing is too pessimistic. Janssens DH et al. 2024
Connected to