Pablo Menendez & Clara Bueno (Josep Carreras Leukemia Research Institute, Barcelona) — cellular origin, fetal models and infant-specific drug resistance
also called Menendez Lab · Bueno Lab · Josep Carreras Leukemia Research Institute · IJC Barcelona · Stem Cell Biology, Developmental Leukemia and Immunotherapy Group
This Barcelona group asks which cell the t(4;11) translocation has to happen in for leukemia to follow, and builds human stem-cell and fetal models to test it. They are also the group most willing to publish results that do not work.
The same thing, for a scientist
Pablo Menendez and Clara Bueno run the Stem Cell Biology, Developmental Leukemia and Immunotherapy Group at the Josep Carreras Leukemia Research Institute, Barcelona, with ICREA and CIBERONC affiliations and a School of Medicine, University of Barcelona appointment. Their programme treats infant t(4;11) B-ALL as a developmental rather than a multi-hit adult-type cancer: prenatal single-cell origin, a near-silent genome, and a critical window of stem-cell vulnerability. They build human embryonic stem cell, cord blood and fetal models, publish negative results on MLL-AF4 biology, and have more recently moved into CAR immunotherapy and infant-specific glucocorticoid resistance.
What they own that the Iqbal Lab does not1 deeper
Human fetal and pluripotent model systems for the cell-of-origin question. Their central argument is that MLL-AF4 modelling has repeatedly failed because the models use the wrong cell at the wrong developmental stage 1. They have spent fifteen years building human embryonic stem cell, cord blood CD34+ and fetal-tissue systems to attack that. A lab that wants to test a fusion-biology hypothesis in a developmentally correct human context has to either build this or borrow it.
A cohort and assay pipeline for detecting preleukemic fusions in newborns. The IMPACT-BCN work screened 741 cord-blood samples from a randomised pregnancy-intervention trial using GIPFEL and found 6.5% positive for ETV6::RUNX1, with corticosteroid use for lung maturation during pregnancy significantly associated with positivity (adjusted OR 3.9, 95% CI 1.6-9.8) 3. That is a prospective, interventional, prenatal-exposure platform with a molecular readout. Nothing in the KMT2A literature comes close to it, and the obvious extension is to run the same design for KMT2A rearrangements.
Infant-specific pharmacology. The NG2 paper shows NG2 is a direct MLL-AF4 target that drives glucocorticoid resistance in MLLr B-ALL by regulating NR3C1 expression 2. Glucocorticoid response is the single strongest early clinical predictor in infant ALL, so this connects fusion target biology directly to the treatment backbone rather than to a hypothetical new drug.
Why they are a good-faith collaborator
Two signals. First, they publish negative results: the DNA-repair study concluded flatly that expressing MLL-AF4 or the reciprocal AF4-MLL does not alter double-strand-break repair kinetics, NHEJ or HR efficiency, and they published it with Marschalek as a co-author even though it undercuts a mechanism the Frankfurt group had floated 4. A group that will co-publish a result against its collaborator's hypothesis is a safe partner.
Second, they are already networked into the groups that matter: Menendez is a co-author on Milne's PROM1 paper [Milne node E1], Bueno is on the prenatal cohort work with Borkhardt in Dusseldorf, and Marschalek and Stam appear on their papers. Partnering with Barcelona is partnering with a hub.
Overlap risk is low. They do cellular origin and models; the Iqbal Lab does interfaces and computation. The natural trade is: we supply interface predictions and the analysis layer, they supply the developmentally correct human cells to test them in.
| group | Stem Cell Biology, Developmental Leukemia and Immunotherapy Group (Pablo Menendez, Clara Bueno) |
|---|---|
| institution | Josep Carreras Leukemia Research Institute, Barcelona, Spain; Department of Biomedicine, School of Medicine, University of Barcelona; ICREA; CIBERONC-ISCIII (affiliations as printed on PubMed-indexed papers through 2025) |
| focus | Cell of origin and developmental window of infant t(4;11) B-ALL; human pluripotent, cord blood and fetal models of MLL-AF4; prenatal origin and preleukemic clone detection; NG2 and glucocorticoid resistance; CAR immunotherapy |
| partnering angle | Trade analysis for cells. Offer computational interface predictions, cross-fusion comparative analysis and target-surfaceome enumeration; ask to test predictions in their developmentally correct human fetal and pluripotent MLL-AF4 models, which is the one context where an infant-specific result is credible. A second, distinct proposal: extend their cord-blood GIPFEL prenatal screening design from ETV6::RUNX1 to KMT2A rearrangements, where no comparable prospective cohort exists. |
| relationship | natural collaborator, low overlap, well networked |
key papers — the numbers behind the chart
[
{
"title": "Revisiting the biology of infant t(4;11)/MLL-AF4+ B-cell acute lymphoblastic leukemia",
"year": 2015,
"venue": "Blood",
"pmid": "26463423",
"doi": "10.1182/blood-2015-09-667378"
},
{
"title": "NG2 is a target gene of MLL-AF4 and underlies glucocorticoid resistance in MLLr B-ALL by regulating NR3C1 expression",
"year": 2024,
"venue": "Blood",
"pmid": "39093982",
"doi": "10.1182/blood.2023022050"
},
{
"title": "Expression of MLL-AF4 or AF4-MLL fusions does not impact the efficiency of DNA damage repair",
"year": 2016,
"venue": "Oncotarget",
"pmid": "27119507",
"doi": "10.18632/oncotarget.8938"
},
{
"title": "Modulation of the ETV6::RUNX1 Gene Fusion Prevalence in Newborns by Corticosteroid Use During Pregnancy",
"year": 2025,
"venue": "Int J Mol Sci",
"pmid": "40243620",
"doi": "10.3390/ijms26072971"
}
]leverage they hold
[ "human ESC, cord blood and fetal MLL-AF4 model systems", "prospective cord-blood cohort with preleukemic fusion screening (GIPFEL) embedded in a randomised pregnancy trial", "connections to Oxford (Milne), Frankfurt (Marschalek), Rotterdam/Utrecht (Stam) and Dusseldorf (Borkhardt)" ]
- If infant KMT2A-r really is a one-hit developmental disease with a near-silent genome, there may be no second target to combine with, which removes the standard escape route from single-agent resistance.
- Their own negative results show how easily MLL-AF4 phenotypes fail to reproduce across model systems; an interface result obtained in a cell line may not survive transfer to a fetal human progenitor.
- Prenatal-exposure findings are politically hazardous: an association between antenatal corticosteroids and preleukemic clones could be misread as advice against a life-saving obstetric intervention.
- What is the prevalence of KMT2A rearrangements in unselected cord blood, measured prospectively with a GIPFEL-equivalent assay? The ETV6::RUNX1 answer exists; the KMT2A answer does not, and it decides whether the KMT2A break is rare-and-sufficient or common-and-usually-silent.
- If the fusion alone is sufficient in the right fetal cell, why has no group produced a faithful human model with short latency? Is the missing ingredient a cell state, a microenvironment, or a non-coding co-lesion?
- Does NG2-driven glucocorticoid resistance interact with menin inhibition? If menin inhibitors restore glucocorticoid sensitivity, the cheapest combination in infant ALL is with a drug already in the backbone.
- E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review
Menendez and Bueno argue infant t(4;11)/MLL-AF4 B-ALL is a developmental disease of prenatal single-cell origin with a near-silent genome, in which the fusion alone may be sufficient, and that the field's failure to build faithful mouse or human models reflects use of the wrong cell of origin and developmental window.
Revisiting the biology of infant t(4;11)/MLL-AF4+ B-cell acute lymphoblastic leukemia · Sanjuan-Pla A, Bueno C, Prieto C, Acha P, Stam RW, Marschalek R, Menendez P · Blood 2015What this evidence cannot show
A review stating the group's interpretive position, not primary data. The 'one big hit is sufficient' claim rests on whole-genome sequencing showing a low mutation burden, which is an argument from absence: it shows no recurrent coding second hit, not that no second event (epigenetic, non-coding, microenvironmental) is required. The twin and Guthrie-card evidence for prenatal origin is genuinely strong and separable from this interpretation.
Evidence that argues against this (1)
- The claim that the field has failed to build a faithful t(4;11) model, and that the failure is a cell-of-origin problem, is directly contradicted by a model published the following year. Transducing human CD34+ cells with MLL fused to murine Af4 produced a pro-B ALL that the authors report 'fully recapitulates the immunophenotypic and molecular aspects of the disease', distinct from MLL-AF9 through differential genomic target binding, and the model reproduced the clinically observed myeloid lineage switch under environmental pressure. That result argues the earlier failures were a construct problem (poor retroviral titre with human AF4) rather than a developmental-window problem, and it supports an INSTRUCTIVE role for the specific fusion over a permissive cell-of-origin account. Lin S et al. 2016
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The group showed NG2 is a direct target gene of MLL-AF4 and underlies glucocorticoid resistance in MLL-rearranged B-ALL by regulating NR3C1 (glucocorticoid receptor) expression.
NG2 is a target gene of MLL-AF4 and underlies glucocorticoid resistance in MLLr B-ALL by regulating NR3C1 expression · Lopez-Millan B, ... Bueno C · Blood 2024What this evidence cannot show
Single-group finding, not yet independently replicated as far as this pass found. NG2 (CSPG4) positivity has long been a known immunophenotypic hallmark of MLLr B-ALL, so the association is well established; the causal claim that NG2 regulates NR3C1 and thereby drives resistance is the new and less-tested part.
Evidence that argues against this (2)
- NG2 is a less reliable marker than its use as a KMT2A-rearrangement surrogate implies. It is malleable and does NOT enrich for leukaemia-initiating or CNS-initiating cells, so NG2 positivity marks a state rather than a compartment. That matters wherever NG2 is treated as identifying the cells that matter, and it cuts against reading NG2-directed findings as targeting the initiating population. Leukemia 2018
- It cuts the other way for therapy, which is the more useful half. Whatever NG2 does or does not enrich for, the NG2-ITGA4 axis drives leukaemic aggressiveness in KMT2A-rearranged B-ALL - the right disease rather than a murine myeloid model - and natalizumab, an approved anti-ITGA4 antibody, delayed progression and potentiated standard-of-care chemotherapy in PDX. A marker that fails as a stem-cell label can still be a good drug target. Blood 2026
- E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
Within the randomised IMPACT-BCN pregnancy-intervention trial, the group screened 741 cord-blood samples by GIPFEL and found 6.5% (48/741) positive for ETV6::RUNX1; corticosteroid use for fetal lung maturation during pregnancy was significantly associated with positivity (adjusted OR 3.9, 95% CI 1.6-9.8), with stronger associations before 26 weeks gestation and for betamethasone.
Modulation of the ETV6::RUNX1 Gene Fusion Prevalence in Newborns by Corticosteroid Use During Pregnancy · Benitez L, Fischer U, Crispi F, et al., Bueno C, Menendez P, Borkhardt A, Gratacos E · International Journal of Molecular Sciences 2025What this evidence cannot show
This is about ETV6::RUNX1, not KMT2A, and it does not transfer automatically. The corticosteroid finding rests on only 39 exposed neonates, the confidence intervals are very wide (one subgroup OR 7.7, 95% CI 1.08-50), it is a secondary analysis of a trial designed for other endpoints, and GIPFEL positivity in cord blood is a contested assay whose specificity and clinical meaning are disputed. Antenatal corticosteroids are a life-saving intervention, so this must not be presented as a clinical caution.
Evidence that argues against this (1)
- The cited study is about **ETV6::RUNX1** prevalence in newborns, not a KMT2A fusion. GIPFEL screening of cord blood is the shared method, so the entry supports a claim about the group's capability and not about KMT2A fusion prevalence — and this graph elsewhere records that cord-blood KMT2A screening returned a ZERO. The two should not be run together. International Journal of Molecular Sciences 2025
- E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
The group published a negative result: CRISPR-engineered expression of MLL-AF4 or the reciprocal AF4-MLL in HEK293 cells did not alter etoposide- or irradiation-induced double-strand-break repair kinetics, NHEJ or homologous recombination efficiency, or DNA damage checkpoint activation.
Expression of MLL-AF4 or AF4-MLL fusions does not impact the efficiency of DNA damage repair · Castano J, Herrero AB, Bursen A, Gonzalez F, Marschalek R, Gutierrez NC, Menendez P · Oncotarget 2016What this evidence cannot show
HEK293 cells are an epithelial kidney line and a poor proxy for a fetal haematopoietic progenitor; a null result in the wrong cell type is weak evidence of a general null. The study also used ectopic AAVS1-locus expression rather than an endogenous rearrangement.
Evidence that argues against this (1)
- Publishing a negative is creditable and the negative is narrower than it reads: the finding is that MLL-AF4 or AF4-MLL expression does not impair DNA damage repair efficiency, in HEK293 cells. A non-haematopoietic immortalised line is close to the least permissive setting for this fusion, so the result bounds the repair hypothesis rather than the fusion's transforming capacity. Oncotarget 2016
- ← proposed byFetal CD19+CD34+CD10-negative PreProB progenitor — a cell that exists before birth and almost not at all after
cellular origin and the primary fetal-liver models
cell-of-origin programme built on human fetal and pluripotent models of MLL-AF4
frames infant t(4;11) B-ALL as a developmental, prenatal-origin, near-silent-genome disease
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
NG2 as a direct MLL-AF4 target driving glucocorticoid resistance via NR3C1
shared prenatal-origin and preleukemic-clone research agenda, approached epidemiologically versus experimentally
- ← related toThomas Milne (MRC WIMM, Oxford) — MLL-AF4 target biology, enhancer architecture and PROM1/CD133
Menendez is a co-author on the Oxford PROM1/CD133 paper; the Oxford-Barcelona axis already exists
- related to →Rolf Marschalek & Claus Meyer (DCAL, Goethe-University Frankfurt) — the world's KMT2A breakpoint dataset
co-published, including a negative result on MLL-AF4 and AF4-MLL effects on DNA repair
- ← proposed byITGA4 (VLA-4) and NG2 — the niche-adhesion axis of KMT2A-rearranged B-ALL, and the one approved drug that hits it
both NG2 papers are from this group; the axis has no independent replication