Oncofusion

start / graph / finding-sem-the-standard-t411-model-has-no-irx1-signal

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

SEM, the standard t(4;11) model line, has zero signal at IRX1 — every model of this disease is a model of one subgroup

Almost everything we know about this infant leukaemia comes from one cell line. That line is firmly in the HOXA half and has nothing at IRX1 — so the worse-prognosis half of patients has effectively never been modelled.

The same thing, for a scientist

In the SEM cell line, AFF1 occupancy is **5.08x and 5.16x** at HOXA9 and HOXA10 and **0.000x at IRX1** with both AFF1 antibodies [E1]. This is the cleanest reciprocal pair in the deposit and it is the expected direction: Symeonidou states explicitly that the disease's knowledge base rests on "mouse models and cell lines, such as SEM, that express HOXA9", and calls for models that mirror iALL-IRX1 [E2]. Kühn had to *engineer* IRX1 into SEM with a Sleeping Beauty transgene to study it at all [E3]. So SEM is a positive control for the HOXA arm of this beat and **cannot serve as the IRX1 arm of anything**.

Go deeper
The measurement, and its one weakness

GSE252378 contains only AFF1 tracks for SEM — no KMT2A-N, no KMT2A-C, no IgG, no H3K4me3. AFF1 is a weak fusion-occupancy readout: q142 measured AFF1 ONCO/WT at 1.03-1.45 (mouse) and 0.56-0.76 (rabbit) against an H3K4me3 negative control at 0.97-1.06, and the deposit paper states in its own words that AF4 does not distinguish oncoprotein-target from wild-type sites 4. So the SEM result is a weaker instrument than the patient results, and it is reported as corroboration of a literature claim rather than as an independent measurement.

What survives that caveat is the contrast within one library: AFF1 at 5.08x on HOXA9 and 0.000x at IRX1 in the same SEM library, with no IgG available to set a floor. Every number is a WITHIN-SAMPLE share: an interval's mean coverage divided by that same library's mean promoter coverage, so 1.0 = an average promoter in that library. GSE252378 has no spike-in and no drug arm, so no between-sample absolute statement is licensed; the only cross-sample quantity used is a ratio of two loci measured inside one library, where the unknown per-library scale cancels.

A GEO metadata error found while reading the series matrix

GSM8371032 (SEM_AFF1_2) carries group: KMT2A-AFF2 in its !Sample_characteristics_ch1, while its pair GSM8371031 (SEM_AFF1_1) carries group: KMT2A-AFF1. SEM is a t(4;11)/KMT2A::AFF1 line, so the AFF2 label is a typographical error in the deposit metadata, not a second fusion. Recorded so a later automated harvest of that field does not create a spurious KMT2A::AFF2 sample. This is a report about a file this beat does not own and no correction was applied.

Facts and figures
cell lineSEM
fusionKMT2A::AFF1, t(4;11)
geo metadata errorGSM8371032 characteristics say 'group: KMT2A-AFF2'; SEM is KMT2A::AFF1
aff1 hoxa9
[
  5.075,
  0.836
]
aff1 hoxa10
[
  5.158,
  0.507
]
aff1 irx1
[
  0,
  0
]
aff1 meis1
[
  16.94,
  3.55
]
tracks available in GSE252378
[
  "AFF1_1 (mouse MBS190886)",
  "AFF1_2 (rabbit PA5-77068)"
]
How this could bite us back
  • Every drug that reached the clinic for KMT2A::AFF1 was optimised on HOXA-expressing models. If iALL-IRX1 blasts do not depend on the HOXA program, the whole preclinical pipeline has been selecting for activity in the better-prognosis half.
  • A menin inhibitor's pharmacodynamic biomarker in these diseases is HOXA9/MEIS1 collapse. In a HOXA-dead blast that biomarker reads 'no target engagement' whether or not the drug is working.
Still open
  1. Is there any patient-derived xenograft or cell line anywhere that is IRX1-high and HOXA-low? Establishing one is the field's stated priority and nothing in this corpus records an attempt.
  2. Would an iALL-IRX1 model even be viable in culture? iALL-IRX1 blasts are the more proliferative subgroup by GSEA, which argues they should be easier, not harder, to establish - and yet none exists.
Every source behind this entry (4)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    In SEM, AFF1 occupancy is 5.08x and 5.16x at HOXA9 and HOXA10 and 0.000x at IRX1 with both AFF1 antibodies.

    q145 re-analysis of GSE252378: IRX1 and HOXA oncoprotein occupancy across seven KMT2A::AFF1 patients (analyses/q145-irx1-in-hoxa-negative-infants) · Iqbal Lab oncofusion knowledge graph, beat q145 · internal analysis over deposited GSE252378 bigWigs 2026
    What this evidence cannot show

    SEM has AFF1 tracks ONLY in this deposit - no KMT2A-N/C, no IgG, no H3K4me3 - so there is no oncoprotein readout and no background floor for this sample. AFF1 is a poor fusion proxy. n = 1 cell line. Every number is a WITHIN-SAMPLE share: an interval's mean coverage divided by that same library's mean promoter coverage, so 1.0 = an average promoter in that library. **GSE252378 has no spike-in and no drug arm**, so no between-sample absolute statement is licensed; the only cross-sample quantity used is a ratio of two loci measured inside one library, where the unknown per-library scale cancels.

    Evidence that argues against this (1)
    • Without an IgG track for SEM the zero at IRX1 cannot be distinguished from a sparse-library zero; the second AFF1 library has weaker dynamic range throughout (HOXA9 0.836 vs 5.075), so its own IRX1 zero adds little. Janssens DH et al. 2024
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The knowledge base for MLL-AF4 infant ALL derives from mouse models and cell lines such as SEM that express HOXA9, and models mirroring iALL-IRX1 are called for as a priority.

    PMID 33069783DOI
    What this evidence cannot show

    A statement of the field's state by one group in a short-format paper, not a systematic survey of available models. Full text fetched and read this cycle.

    Evidence that argues against this (1)
    • If an iALL-IRX1-like line has been established since 2021 this claim is stale; no such line was found in the searches run for this beat, but a cell-line repository search was not exhaustive. Nagel S et al. 2022
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    IRX1 had to be introduced into SEM by a doxycycline-inducible Sleeping Beauty transgene to study its effects, and its induction downregulated MLL-AF4-driven HOXA transcription.

    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

    A transgene overexpression experiment with Q-PCR readout and no ChIP in SEM. The paper notes this was the first non-viral stable modification of a t(4;11) line, so the system was novel and unreplicated at the time.

    Evidence that argues against this (1)
    • A transgene expressed from a heterologous promoter reaches levels that need not resemble the endogenous IRX1 of an iALL-IRX1 blast, so dominant-negative effects observed may be dose artefacts. Kuhn A et al. 2016
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    AF4/AFF1 does not effectively distinguish oncoprotein-target sites from wild-type sites in patient samples.

    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 · Nat Commun 2024
    PMID 39472576DOI
    What this evidence cannot show

    The deposit paper's own conclusion about its own AFF1 data. It bounds what the SEM AFF1 tracks can support and is the reason KMT2A-N was used for every patient claim in this beat.

    Evidence that argues against this (1)
    • The statement is about patient samples; a clonal cell line with a single dominant fusion species could in principle give AFF1 better discrimination than primary material does. Janssens DH et al. 2024
Connected to