IRX1 CUT&RUN in primary t(4;11) blasts — testing Kühn's HEK293T mechanism where it matters
The idea that IRX1 shuts down HOXA by sitting on it comes entirely from kidney cells in a dish. Nobody has checked whether IRX1 sits there in an actual patient's leukaemia.
The same thing, for a scientist
Kühn's central mechanism — IRX1 protein occupying the HOXA9 and HOXA10 promoters, only in the presence of MLL-AF4, and repressing them dominant-negatively — rests on ChIP in **HEK293T cells with transfected inducible transgenes** [E1]. No IRX1 ChIP or CUT&RUN in primary t(4;11) material was found in the searches run for this beat. This experiment is the direct test, and it is the cleanest discriminator between the two mechanisms in `hyp-irx1-is-a-second-oncoprotein-address-in-infant-t411`.
Design and the reagent problem
Material. >=3 IRX1-high primary or PDX KMT2A::AFF1 blasts, with >=3 HOXA-high blasts as the negative arm (IRX1 should be absent or unbindable there).
Tracks. IRX1; KMT2A-N1/N2; KMT2A-C1/C2; H3K4me3; deep IgG.
The reagent is the risk. Kühn used Abcam ab72642 for both ChIP and Western. A homeobox transcription factor antibody validated in an overexpression system may have no usable signal-to-noise at endogenous levels — and in a CUT&RUN, a poor antibody produces a sparse library whose zeros look like absence. Validate first in an IRX1-overexpressing SEM line (Kühn's own construct) against parental SEM, and report the dynamic range before profiling any patient. If validation fails, the experiment does not proceed and that is reported, rather than a patient profile with an uninterpretable track.
Pre-declare the control loci: HOXA9 and HOXA10 promoters (Kühn's positives), HOXB4 promoter (Kühn's one ChIP-confirmed direct IRX1 target 1, and measured in this beat as carrying no oncoprotein in any of seven patients 2), MEIS1 promoter (Kühn's own experimental control), GAPDH, and per-locus IgG.
Outcomes
- IRX1 at HOXA9/HOXA10 in a primary blast → Kühn's mechanism transfers out of HEK293T, and mechanism (B) is supported: the subgroup is IRX1 shutting HOXA down.
- IRX1 at HOXB4 but not at HOXA → IRX1 is running its own program without touching HOXA, and the HOXA silence has another cause.
- IRX1 nowhere above the floor, with a validated antibody → the protein is not a chromatin factor at these loci in patient blasts, and the HEK293T result is a system artefact.
- Antibody fails validation → report unmeasured. A sparse IRX1 track scored as "IRX1 is not at HOXA" would be exactly the absence-as-evidence error this beat's controls were built to prevent 3.
| design | IRX1 + KMT2A-N1/N2 + KMT2A-C1/C2 + H3K4me3 + deep IgG CUT&RUN on >=3 IRX1-high and >=3 HOXA-high primary or PDX KMT2A::AFF1 blasts |
|---|---|
| readout | IRX1 within-sample share at HOXA9, HOXA10, HOXB4 and MEIS1 promoters versus a per-locus IgG floor |
| controls | antibody validation in IRX1-overexpressing SEM versus parental SEM BEFORE any patient sample; HOXB4 positive; GAPDH; per-locus IgG; dynamic range reported |
| gate | if the antibody does not clear validation the experiment stops and reports unmeasured |
| cost tier | med |
| timeline | 4-8 months including antibody validation |
| key reagent | anti-IRX1 Abcam ab72642 (Kuhn's reagent), validation status at endogenous levels unknown |
- A negative from an unvalidated antibody would be recorded as 'IRX1 does not bind HOXA in patients' and would be wrong in the most consequential direction, which is why validation is a gate here rather than a courtesy.
- Does any validated ChIP-grade IRX1 antibody exist? Only one is named in this literature and only in an overexpression context.
- If IRX1 binding requires MLL-AF4 co-presence, is a patient blast with the fusion away from HOXA even a permissive context for the HEK293T mechanism?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
IRX1 occupies the HOXA9 and HOXA10 promoters only in the presence of MLL-AF4 and shows ChIP enrichment at the HOXB4 promoter, all in HEK293T cells with inducible transgenes, using Abcam ab72642.
The IRX1/HOXA connection: insights into a novel t(4;11)-specific cancer mechanism · Kuhn A, Loscher D, Marschalek R · Oncotarget 2016What this evidence cannot show
HEK293T, transfected transgenes, one representative experiment measured in triplicate. The antibody was used in an overexpression context throughout; no endogenous-level validation is reported. Full text fetched and read this cycle.
Evidence that argues against this (1)
- IRX1 binding at HOXA promoters required MLL-AF4 co-expression, so it is a property of the pair; in a patient blast where the fusion is already leaving HOXA, the conditions for IRX1 binding there may not exist. Kuhn A et al. 2016
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
HOXB4 carries no KMT2A::AFF1 oncoprotein (0.00-0.93x mean promoter) in any of seven infant patients, consistent with Kuhn's statement that HOXB4 is not an MLL-AF4 target.
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 2026What this evidence cannot show
A promoter-scale negative in seven patients; neighbouring HOXB promoters were not separately profiled. 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)
- One deposit, one lab, one assay; seven patients, no RNA-seq, no menin track in any KMT2A::AFF1 sample. Janssens DH et al. 2024
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
In this deposit, P(a promoter reads zero in both KMT2A-N antibodies by chance) ranges from 0.019 to 0.260 across libraries, so an absence from a shallow library is not evidence.
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 2026What this evidence cannot show
Computed as the product of per-library promoter zero-rates, assuming the two antibodies fail independently; if depth is a shared cause the true joint rate is higher, making shallow libraries even less informative.
Evidence that argues against this (1)
- The same logic applies to the proposed IRX1 track and is the reason antibody validation is a gate rather than a step. Janssens DH et al. 2024
HOXB4 is the pre-declared positive control for that experiment
- tests →IRX1 is a second address for the KMT2A::AFF1 oncoprotein, not merely the gene that turns HOXA off
the cleanest discriminator between the two mechanisms
whether IRX1 protein occupies HOXA promoters outside HEK293T
Kuhn's HOXA9 promoter ChIP, in patient material