The experiment that separates lineage from ontogeny: one KMT2A fusion, two cell-of-origin contexts, three metabolic perturbations, and a limiting-dilution readout alongside the viability one
also called fetal versus adult progenitor metabolic experiment · isogenic ontogeny metabolic test
Put the same leukaemia gene into blood cells from before birth and into blood cells from an adult, then measure how each one burns fuel and how well each one can restart the disease in a mouse. If the two behave differently, the difference in metabolism that everyone attributes to cell type is really about when the cell was born.
The same thing, for a scientist
Design to discriminate `hyp-metabolic-split-is-ontogeny-not-lineage` from the default lineage explanation, and simultaneously to resolve the assay confound that may underlie both. **Arms:** one KMT2A fusion (KMT2A::AFF1 preferred, KMT2A::MLLT3 as the better-established alternative) introduced by CRISPR-induced rearrangement into (i) fetal-liver or cord-blood-derived CD34+ progenitors and (ii) adult bone-marrow-derived CD34+ progenitors, each differentiated toward both B-lymphoid and myeloid output so that ontogeny and lineage are **crossed rather than confounded**. **Perturbations:** complex I inhibition, MAT2A inhibition, and IMPDH inhibition with and without a menin inhibitor. **Readouts, run in parallel on every arm:** Seahorse extracellular flux, stable-isotope tracing of glucose and glutamine, and **limiting-dilution xenotransplantation**. The last readout is the point. This beat's central contradiction - complex I inhibition helps in KMT2A-rearranged AML [E1] and promotes leukaemia-initiating cell emergence in MLL-rearranged B-ALL [E2] - may be a *lineage* difference, an *ontogeny* difference, or simply an **assay** difference between bulk proliferation and initiating capacity. Running both readouts on every arm tests all three explanations in one experiment.
The 2x2 that does not currently exist anywhere in the literature
| fetal/neonatal origin | adult origin | |
|---|---|---|
| B-lymphoid | infant t(4;11) - where all the lymphoid metabolic data come from | adult t(4;11) B-ALL - the missing control, and clinical material exists |
| myeloid | infant KMT2A-rearranged AML - exists clinically, metabolically unstudied | MLL-AF9 models - where all the myeloid metabolic data come from |
The published evidence occupies one diagonal. Both off-diagonal cells are obtainable: adult t(4;11) B-ALL is a recognised clinical entity with biobanked material, and infant KMT2A-rearranged AML appears in the paediatric PDX platforms this beat cites 1. A cheap version of this whole experiment is therefore available without building a single new model: run Seahorse and a tracer panel on banked adult t(4;11) B-ALL and infant KMT2A-rearranged AML samples and see which diagonal each falls on.
That cheap version should be done first. The isogenic CRISPR version is the confirmatory experiment, not the screening one.
Controls and failure modes this design must carry
- The engineered-model objection. CRISPR-induced KMT2A rearrangement in primary CD34+ cells is established in this field 3 but the resulting cells are not patient leukaemia. The banked-sample arm above is the corrective and should be run in parallel, not afterwards.
- Culture normalises metabolism. Extracellular flux measured after weeks of culture in one medium measures the medium as much as the cell. Tracer work should be done as early as the model permits, and xenograft-derived cells should be assayed ex vivo within hours.
- SAMHD1 protein must be blotted in every arm. The IMPDH perturbation's effect modifier is SAMHD1, not the fusion, and the simplest explanation of the published ALL negative is low SAMHD1 in those lines - which nobody has reported 4. Without this blot the IMPDH arm is uninterpretable.
- Complex I inhibitor identity must be stated. Neither of the two key lymphoid and mechanistic studies names its OXPHOS inhibitor in its abstract 2, and metformin, rotenone and IACS-010759 are not equivalent probes.
- Limiting-dilution transplantation is the expensive part and is the part that cannot be dropped. A design that measures only viability will reproduce the existing contradiction rather than resolve it.
| design | Crossed 2x2 of cell-of-origin (fetal/cord-blood versus adult CD34+) by lineage output (B-lymphoid versus myeloid), carrying one KMT2A fusion introduced by CRISPR-induced rearrangement; run in parallel with a no-new-models arm using banked ADULT t(4;11) B-ALL and INFANT KMT2A-rearranged AML primary samples. |
|---|---|
| readout | Seahorse extracellular flux; stable-isotope tracing of glucose and glutamine; limiting-dilution xenotransplantation for leukaemia-initiating cell frequency - the last run on EVERY arm, not only on the lymphoid ones |
| cost tier | high |
| timeline | the banked-sample arm is 3-6 months and needs no new models; the isogenic CRISPR arm is 18-24 months |
| do first | the banked-sample arm - adult t(4;11) B-ALL is the single most informative missing data point in this beat and the material already exists |
| also resolves | whether the myeloid/lymphoid contradiction is an assay artefact (bulk proliferation versus initiating capacity) rather than a biological difference |
perturbations
[ "complex I inhibitor (compound identity stated, not 'an OXPHOS inhibitor')", "MAT2A inhibitor (FIDAS-5 or PF-9366)", "IMPDH inhibitor with and without a menin inhibitor" ]
controls
[ "SAMHD1 protein blot in every arm", "assay cells ex vivo within hours of xenograft harvest", "match culture medium across arms and report it", "vehicle and a pan-essential gene reference for the perturbation panel" ]
tests
[ "hyp-metabolic-split-is-ontogeny-not-lineage" ]
- This design could produce a clean negative - no ontogeny effect - and still leave the clinical contradiction unresolved, because the assay confound and the lineage explanation would both survive it.
- An engineered isogenic model's metabolism is substantially a property of its culture medium. A well-controlled experiment on models that have both been normalised by culture can give a confident, reproducible, irrelevant answer.
- The expensive readout is the indispensable one. A funder or reviewer will push to drop limiting-dilution transplantation, and dropping it converts this from an experiment that resolves the contradiction into one that reproduces it.
- Who holds viable banked adult t(4;11) B-ALL material? This is a collaboration question, not a science question, and it gates the cheap version of the whole design.
- Can limiting-dilution transplantation be run at the scale this design implies? Four arms times three perturbations times a dilution series is a large mouse commitment, and a staged design that runs it only where Seahorse shows divergence may be the realistic version.
- Is there a metabolic readout that predicts leukaemia-initiating capacity without transplantation? If one existed it would collapse the cost of this and of much else in this beat.
- Should the fusion be KMT2A::AFF1 (index, harder to model in myeloid output) or KMT2A::MLLT3 (better established, but not the index fusion)? Choosing MLLT3 makes the experiment easier and weakens the inference to the disease this graph is about.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A platform of 26 paediatric AML patient-derived xenografts spanning 14 high-risk genetic subtypes was established with preserved immunophenotypic, genomic and transcriptomic profiles, and was used to test IACS-010759 plus venetoclax in KMT2A-rearranged PDXs modelling both disease onset and relapse, with reduced AML progression; the authors conclude their results strongly support mitochondrial targeting in this genetic AML subtype.
Advancing precision therapy in pediatric acute myeloid leukemia through PDX models and mitochondrial targeting. · Da Ros A, Peloso A, Longo G, ..., Pigazzi M · Blood Adv 2026What this evidence cannot show
Abstract read this session; full text not retrieved. Cited here as evidence that the material required for the cheap version of this experiment exists in paediatric PDX platforms, and as one arm of the contradiction to be resolved. The abstract does not state the ages of the donors or whether any KMT2A-rearranged AML model is from an infant, so the claim that infant KMT2A-rearranged AML material is obtainable from such platforms is an expectation, not a verified fact.
Evidence that argues against this (1)
- PDX passaging selects for engraftable subclones and imposes a murine marrow microenvironment, both of which can alter metabolic phenotype. A metabolic comparison across PDX models is therefore a comparison of what survived xenografting, which is exactly the variable a limiting-dilution readout is sensitive to. Morris V et al. 2022
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In MLL-rearranged B-ALL, single-cell transcriptomics with quantitative xenotransplantation showed that inhibiting oxidative phosphorylation blunts blast proliferation while promoting leukaemia-initiating cell emergence, whereas inhibiting hypoxia and glycolysis impairs leukaemia-initiating cells and provides therapeutic benefit - establishing that bulk-proliferation and initiating-capacity readouts can give opposite answers for the same intervention.
Hypoxic, glycolytic metabolism is a vulnerability of B-acute lymphoblastic leukemia-initiating cells. · Morris V, Wang D, Li Z, ..., Rowe RG · Cell Rep 2022What this evidence cannot show
Abstract read this session; identifiers re-fetched from PubMed. The OXPHOS inhibitor is not named in the abstract, which is why this experiment specifies that the compound must be stated. Xenotransplantation engraftment is a proxy for leukaemia-initiating capacity. Infants are not separately reported, which is the stratification this design exists to supply.
Evidence that argues against this (1)
- If the divergence between bulk and initiating readouts is general rather than lineage-specific, then the entire myeloid literature - which is overwhelmingly proliferation-based - may be reporting the same artefact with the opposite sign, and the ontogeny hypothesis would be answering a question that was never well posed. Da Ros A et al. 2026
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
A human CRISPR/Cas9-MLL-rearranged leukaemia model generated in primary cells was used to study a metabolic enzyme (MAT2A) against healthy controls and against non-MLLr and MLLr cell lines, with an MLLr-specific enhanced response to inhibition and knockdown - demonstrating that CRISPR-induced KMT2A rearrangement in primary human haematopoietic cells is an established platform for metabolic questions.
MAT2A as Key Regulator and Therapeutic Target in MLLr Leukemogenesis. · Secker KA, Bloechl B, Keppeler H, ..., Schneidawind C · Cancers (Basel) 2020What this evidence cannot show
Abstract read this session. Cited as feasibility evidence for the engineered arm, not as a metabolic result. The abstract does not state the developmental origin of the CD34+ cells used, so it does not itself demonstrate that the fetal-versus-adult comparison has been done or is straightforward. The model is described as mimicking MLLr pathology with indefinite growth potential in vitro, which is a strength for tractability and a weakness for representing primary disease metabolism.
Evidence that argues against this (1)
- An engineered model with indefinite in vitro growth potential has by definition been selected for proliferation in culture medium, which is the single strongest determinant of measured metabolic phenotype. This is why the banked primary-sample arm of the design is not optional. Symeonidou V et al. 2021
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
IMPDH-inhibitor synergy with nucleoside analogues was SAMHD1-dependent and absent in acute lymphoblastic leukaemia cell lines, and sensitivity to IMPDH inhibitors was independent of KMT2A status - so any arm of this experiment using an IMPDH inhibitor must measure SAMHD1 protein or it cannot be interpreted.
Targeting IMPDH to inhibit SAMHD1 in KMT2A-rearranged leukaemia. · Klootsema Y, Tsesmetzis N, Sharma S, ..., Herold N · Cell Cycle 2026What this evidence cannot show
Abstract read this session. The ALL negative is in cell lines and SAMHD1 levels in those lines are not reported in the abstract, which is the gap this experiment's control list is designed to close.
Evidence that argues against this (1)
- If SAMHD1 fully explains the ALL negative, the IMPDH arm of this experiment tests nothing about ontogeny or lineage and should be dropped in favour of spending the resource on the complex I and MAT2A arms. Yague-Capilla M et al. 2026
the crossed ontogeny-by-lineage design that discriminates the two explanations
fetal-derived progenitor arm of the crossed design
adult-derived progenitor arm of the crossed design
also discriminates assay artefact from biological difference