Is IRX1 expressed in normal blood at all? Two of the three papers say no, one says exclusively yes
Before you can say a leukaemia has switched IRX1 on abnormally, you need to know whether healthy blood cells use it. The published answers flatly contradict each other and nobody has resolved it.
The same thing, for a scientist
An unresolved contradiction in the primary literature, recorded rather than adjudicated. **Kühn 2016:** "Irx1 and its human counterpart IRX1 are usually not expressed in hematopoietic tissues" [E1]. **Symeonidou 2021**, from single-cell RNA-seq of adult bone marrow and fetal liver: IRX1 "exhibited very little expression in the hematopoietic system", and in mouse gastrulation datasets Irx1 is predominantly **mesodermal** while Hoxa9 and Meis1 sit in hematoendothelial progenitors [E2]. **Nagel 2022:** IRX1 and MEIS1 are expressed **exclusively in pro-B-cells** in normal B-cell differentiation [E3]. These cannot all be right, and which is right changes the interpretation of the entire iALL-IRX1 subgroup.
What hangs on it
- If IRX1 is not expressed in blood (Kühn, Symeonidou): iALL-IRX1 is an ectopic developmental program, which is Symeonidou's argument for a developmentally earlier, possibly pre-haematopoietic cell of origin — she notes MLL-AF4 has been reported in bone-marrow mesenchymal stromal cells of infant patients, and that mesoderm has stromal progeny.
- If IRX1 is a normal pro-B gene (Nagel): iALL-IRX1 is a retained normal program of an ordinary B-cell progenitor, and the subgroup difference is developmental timing within the B lineage rather than a different germ layer.
The first reading makes iALL-IRX1 a different disease; the second makes it a different stage. Those imply different models, different targets and different expectations of lineage plasticity.
Why it may not be a simple contradiction
Three non-exclusive reconciliations, none of them tested:
- Detection floor. A gene expressed in a rare, transient pro-B window could be "very little expression" in a whole-marrow scRNA-seq summary and "exclusive to pro-B" in a stage-resolved analysis. These are compatible descriptions of the same data at different resolutions.
- Species and stage. Kühn's statement is largely about mouse Irx1 developmental expression; Symeonidou's mouse data is gastrulation, her human data adult marrow and fetal liver. Nagel's is human B-cell stages. No two of the three examine the same tissue at the same stage.
- Nagel's design. Nagel's normal-development conclusions lean on expression profiling and cell lines used as stage models, and the paper's own aberrant-IRX calls in BCP-ALL are IRX2 and IRX3, not IRX1 — so IRX1 is not the gene that paper is built to resolve.
Reading 1 is the one I would test first and the cheapest: re-interrogate a stage-resolved human B-progenitor scRNA-seq atlas for IRX1 with an explicit detection threshold stated.
| question | Is IRX1 expressed in normal human haematopoiesis, and specifically in pro-B cells? |
|---|---|
| status | unresolved; no experiment in this corpus adjudicates it |
| consequence | decides whether iALL-IRX1 is an ectopic program (earlier/possibly pre-haematopoietic origin) or a retained pro-B program |
positions
{
"Kuhn 2016 (PMID 27175594)": "not expressed in haematopoietic tissues",
"Symeonidou 2021 (PMID 33069783)": "very little expression in the haematopoietic system; mesodermal in mouse gastrulation",
"Nagel 2022 (PMID 36233173)": "expressed exclusively in pro-B-cells"
}- If iALL-IRX1 arises from a pre-haematopoietic or stromal-competent cell, then a therapy that clears the blood compartment may leave a reservoir that nothing in current protocols reaches.
- What does a stage-resolved human B-progenitor scRNA-seq atlas show for IRX1 with a stated detection threshold? This is the cheapest decisive analysis in the whole beat.
- Does IRX1 protein, as opposed to transcript, appear in any normal haematopoietic compartment? No protein-level data was found for any position.
- If IRX1 is pro-B-restricted, does its occupancy in p318 - an AML - mean the switched AML retains a pro-B chromatin feature?
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
IRX1 and its mouse orthologue are usually not expressed in hematopoietic tissues.
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
A background statement in the introduction of a mechanism paper, supported by citation to mouse developmental expression studies rather than by a measurement in that paper. It is an assertion about the field, not a result.
Evidence that argues against this (1)
- Nagel & Meyer report IRX1 expressed exclusively in pro-B-cells within normal human B-cell development, which directly contradicts this. Nagel S et al. 2022
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In single-cell RNA-seq of human adult bone marrow and fetal liver, IRX1 shows very little expression, while HOXA9 and MEIS1 are expressed in haematopoietic stem and progenitor cells; in mouse gastrulation Irx1 is predominantly mesodermal.
HOXA9/IRX1 expression pattern defines two subgroups of infant MLL-AF4-driven acute lymphoblastic leukemia · Symeonidou V, Ottersbach K · Exp Hematol 2021What this evidence cannot show
Reinterrogation of previously published atlases, not new data. No detection threshold is stated in the text available to me, and 'very little expression' is not quantified. Whole-marrow and fetal-liver atlases may under-sample a narrow pro-B window.
Evidence that argues against this (1)
- A stage-resolved analysis restricted to the B-progenitor trajectory could recover IRX1 where a whole-tissue summary does not, which would reconcile this with Nagel rather than contradict it. Nagel S et al. 2022
- E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
In normal B-cell differentiation IRX1 and MEIS1 are expressed exclusively in pro-B-cells.
Normal and Aberrant TALE-Class Homeobox Gene Activities in Pro-B-Cells and B-Cell Precursor Acute Lymphoblastic Leukemia · Nagel S, Meyer C · Int J Mol Sci 2022What this evidence cannot show
Expression profiling plus BCP-ALL cell lines used as stage models with siRNA knockdown and reporter assays. The paper's own aberrant BCP-ALL findings concern IRX2, IRX3 and MEIS1 - not IRX1 - so IRX1 in normal pro-B is a supporting observation rather than the paper's central result.
Evidence that argues against this (1)
- Kuhn states IRX1 is not expressed in haematopoietic tissues and Symeonidou finds very little IRX1 in adult marrow or fetal liver scRNA-seq, both contradicting this. Symeonidou V et al. 2021
- related to →IRX1 — Iroquois homeobox 1, the gene that names the HOXA-negative half of infant t(4;11) ALL
unresolved: is IRX1 a normal pro-B gene or not expressed in blood at all?
Nagel places IRX1 exclusively in pro-B; two other papers deny it
decides whether the state is an ectopic or a retained program
Symeonidou found very little IRX1 in fetal liver scRNA-seq
bears on the dispute without resolving it: the dichotomy is real but says nothing about normal pro-B cells
either side of the dispute changes which permissive cell this hypothesis needs; neither is assumed
the pro-B TALE-code claim that IRX2 is silent in normal B-cell differentiation rides on the same contested instrument