Is the metabolic phenotype a consequence of the fusion programme or an independent dependency? The answer is that BOTH directions are published, and one of them points the other way entirely
also called consequence versus dependency · metabolism upstream or downstream of the KMT2A fusion
When a leukaemia has an unusual metabolism, there are two possible explanations: the cancer gene switched the metabolism on, or the cancer happens to need that metabolism for its own reasons. In KMT2A-rearranged leukaemia both turn out to be true for different pathways - and in one case the arrow points backwards, with the metabolism controlling how much of the cancer machine the cell can build.
The same thing, for a scientist
This entry exists because the question the metabolism beat was dispatched with has a three-part answer, and collapsing it to one part would misdirect the lab. Sorting every mechanism in this beat by the direction its evidence supports gives: **(A) Consequence of the fusion programme.** The fusion's chromatin output creates a downstream metabolic requirement. Best case: SAM. The methionine dependency of KMT2A-rearranged ALL is stated by its authors to be *driven by* an increased need for SAM to maintain the hypermethylated state [E1]. Second case, and better specified mechanistically: **PTBP1**, a *direct* MLL-ENL target gene, whose knockdown is rate-limiting for proliferation and produces reduced glucose consumption, reduced lactate and loss of the PKM2 isoform [E2]. Third case: **AK_7_3**, a fusion-derived circular RNA that restrains mitochondrial respiration in the index fusion itself [E5]. **(B) Independent dependency.** The requirement exists in the cell for reasons the fusion did not create, and is merely shared. Best case: **IMPDH**, where the most rigorous test of genotype-linkage found sensitivity **independent of KMT2A status** and the operative modifier to be SAMHD1 [E4]. **(C) The inverse - metabolism upstream of the fusion complex.** IMPDH2 blockade lowers rRNA transcription and with it the **protein levels of LEDGF and menin**, impairing formation and chromatin binding of the MLL-fusion complex [E3]. Here the metabolic supply determines how much of the oncogenic machine the cell can assemble. **The practical consequence is that 'target the metabolism of KMT2A-rearranged leukaemia' is not a strategy.** Each direction implies a different trial design, a different biomarker and a different resistance profile.
Why the three cases are not interchangeable in the clinic
| direction | example | what it implies for design | resistance profile |
|---|---|---|---|
| A. consequence | SAM/MAT2A 1; PTBP1/PKM2/serine 2; AK_7_3 5 | a second route to the same output. Combine with, or substitute for, an interface drug. Biomarker is the fusion itself. | not defeated by MEN1 pocket mutation; defeated by anything that restores the chromatin output |
| B. independent | IMPDH/SAMHD1 4 | a separate agent that happens to help. Do not enrich the trial for KMT2A. Biomarker is the modifier - here SAMHD1 protein. | ordinary pharmacological resistance; genotype irrelevant |
| C. inverse | IMPDH2 -> LEDGF/menin protein 3 | a sensitiser. Only meaningful in combination with a menin inhibitor. Biomarker is menin-inhibitor exposure. | specifically NOT defeated by MEN1 binding-site mutation, which is the dominant clinical resistance mechanism |
Case C is the most valuable and the least replicated. It is a single laboratory in myeloid models, and it has never been tested in a MEN1-mutant menin-inhibitor-resistant line, which is the one experiment that would establish its clinical niche 3.
The cross-cutting confound: lineage, not fusion
Every direction above is complicated by a split that runs through the whole beat. Where the same metabolic axis has been measured in both lineages, the answers oppose each other:
- OXPHOS. MLL/AF9 AML cells depend mainly on OXPHOS and are repressed by metformin 6. In MLL-rearranged B-ALL, inhibiting oxidative phosphorylation blunts blast proliferation but promotes leukaemia-initiating cell emergence, while inhibiting hypoxia and glycolysis impairs the initiating cells 7.
- Lipids. The myeloid dependency is ACSL4, on polyunsaturated substrates; the infant lymphoid requirement is ELOVL1, on saturated very-long-chain substrates. Opposite sides of the saturation divide.
- IMPDH. Synergy with nucleoside analogues was present in a subset of AML cells and absent in ALL cell lines 4.
So the honest summary is: the fusion is not the variable that best predicts the metabolic phenotype; the lineage is. That is a claim this graph can test, and hyp-metabolic-split-is-ontogeny-not-lineage states the sharper version - that the split may be ontogeny (fetal versus adult origin) rather than lineage as such.
What a decisive experiment looks like
One design settles most of it. Take an isogenic pair - the same fusion expressed in a fetal-derived B-lymphoid progenitor and in an adult-derived myeloid progenitor, which this field can now construct with CRISPR-induced rearrangements - and run: Seahorse, a glucose/glutamine tracer panel, and limiting-dilution transplantation, under (i) complex I inhibition, (ii) MAT2A inhibition, (iii) IMPDH inhibition with and without a menin inhibitor.
The predictions differ sharply. If the metabolic phenotype is a consequence of the fusion programme, the two cell-of-origin contexts should behave the same. If it is set by ontogeny, they should diverge and the divergence should match the published myeloid/lymphoid split. No published study has run the index fusion in both contexts with a metabolic readout, which is why this entry is a concept rather than a finding.
| answer in one line | Both, plus a third direction nobody expected: for SAM, serine and the fusion circRNA the metabolism is DOWNSTREAM of the fusion programme; for IMPDH the dependency is measurably INDEPENDENT of KMT2A status; and for the LEDGF/menin arm the metabolism is UPSTREAM of the fusion complex itself. |
|---|---|
| best predictor of the metabolic phenotype | lineage, not fusion - and possibly ontogeny rather than lineage; see hyp-metabolic-split-is-ontogeny-not-lineage |
| decisive experiment | isogenic fusion in fetal-derived B-lymphoid versus adult-derived myeloid progenitors, with Seahorse, tracer metabolomics and limiting-dilution transplantation under complex I, MAT2A and IMPDH inhibition |
| never done | the index fusion KMT2A::AFF1 has never been run in both cell-of-origin contexts with a metabolic readout |
case a consequence
[ "mechanism-sam-supply-sustains-kmt2a-r-hypermethylation", "mechanism-serine-glycine-auxotrophy-mll-fusion", "mechanism-kmt2a-aff1-fcircrna-ak73-mitochondrial-restraint" ]
case b independent
[ "gene-samhd1", "gene-impdh2 (translational claim only)" ]
case c inverse
[ "mechanism-guanine-nucleotide-supply-sustains-ledgf-menin-mll-complex" ]
- The temptation this entry exists to resist is a single 'metabolic strategy' for KMT2A-rearranged leukaemia. Three of the mechanisms here point in incompatible directions, and one of them - OXPHOS inhibition - has been measured to help in the myeloid disease and to expand the relapse-capable compartment in the lymphoid one.
- The most valuable finding in the beat is supported by one laboratory in one lineage with no independent replication. Building a programme on case C before it is replicated would repeat a failure mode this graph has logged before.
- Case B is a warning about trial design specifically: if IMPDH-inhibitor sensitivity is independent of KMT2A status, a trial enriched for KMT2A-rearranged patients on the strength of the mechanism would dilute rather than concentrate the effect.
- Does case C survive replication? A second laboratory showing that GTP depletion lowers menin protein would make it the most important result in the metabolic layer; failure to replicate would remove the only mechanism in this beat that speaks directly to the lab's central hypothesis.
- Is the myeloid/lymphoid metabolic split a lineage effect or an ontogeny effect? Fetal-origin infant disease and adult-origin myeloid disease differ in more than lineage.
- Are there case-A dependencies specific to KMT2A::AFF1 rather than to KMT2A rearrangement generally? Every case-A example except the fusion circRNA was found in a non-AFF1 fusion.
- Which direction should a first trial take? Case C is the highest-value and least replicated; case A is the best replicated and the least specific; case B is the most rigorously tested and says not to enrich for KMT2A at all.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
KMT2A-rearranged leukaemia is particularly vulnerable to methionine-cycle perturbation, and this methionine dependency is driven by an increased need for S-adenosylmethionine to maintain the hypermethylated state of KMT2A-rearranged leukaemias; pro-survival KMT2A-r target genes are repressed under methionine restriction.
S-adenosylmethionine addiction confers sensitivity to methionine restriction in KMT2A-rearranged acute lymphoblastic leukemia. · Tee T, Ruiter TJJ, Wu S, ..., Van Leeuwen FN · Haematologica 2025What this evidence cannot show
Abstract read this session. The causal direction ('driven by') is the authors' interpretation; the abstract does not describe the experiment that establishes direction rather than correlation. Fusion partners in the PDX panel are not named. Countervailing internal evidence: MAT2A is depleted in 95.0% of DepMap lines, which is floor compression and means the screen cannot see selectivity here.
Evidence that argues against this (1)
- Every proliferating cell requires SAM, and methionine restriction has broad antiproliferative effects across tumour types, so a consequence-of-the-programme reading has to be distinguished from a generic proliferation effect. The distinguishing observation offered is that specific pro-survival KMT2A-r target genes are repressed - a transcriptional readout, not a growth readout - and that observation has not been independently replicated. Secker KA et al. 2020
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Knockdown of the direct MLL-ENL target gene PTBP1 was rate-limiting for proliferation and caused a metabolic phenotype of reduced glucose consumption and lactate production with reduction of the PKM2 splice isoform of pyruvate kinase; inhibition of serine biosynthesis, glucose starvation and energetic decoupling each affected transformed cells significantly more than normal controls, and a serine/glycine-free diet significantly extended survival in transplantation experiments.
Exploitable metabolic dependencies in MLL-ENL-induced leukemia. · Garcia-Cuellar MP, Lawlor J, Bottcher M, ..., Slany RK · Blood Adv 2020What this evidence cannot show
Abstract read this session. MLL-ENL (KMT2A::MLLT1), not the index fusion. PTBP1 is a splicing factor, so attributing the metabolic phenotype to the PKM2 isoform shift specifically is an interpretation; many other splicing changes are expected. This is nonetheless the best-specified consequence chain in the beat because the intermediate is a named DIRECT target gene of the fusion.
Evidence that argues against this (1)
- One of the three perturbations used, rotenone, is a complex I inhibitor, and OXPHOS inhibition has been separately measured to promote leukaemia-initiating cell emergence in MLL-rearranged B-ALL. The three perturbations in that study are not interchangeable and one of them points the opposite way in the lymphoid disease. Morris V et al. 2022
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Targeting IMPDH2 reduces guanine nucleotides and rRNA transcription, leading to reduced protein expression of LEDGF and menin, impaired formation and chromatin binding of the MLL-fusion complex and reduced MLL target gene expression; combining guanine-nucleotide or rRNA-transcription blockade with a menin inhibitor further suppresses MLLr AML.
Guanine nucleotide biosynthesis blockade impairs MLL complex formation and sensitizes leukemias to menin inhibition. · Shi X, Li M, Liu Z, ..., Nakada D · Nat Commun 2025What this evidence cannot show
Abstract read this session. Single laboratory, myeloid only, no independent replication found on 2026-09-20. The chain runs through rRNA transcription, a general growth axis, so a specific relationship to the MLL complex has not been separated from a global reduction in protein synthesis. This is the load-bearing citation for the entire 'case C' argument in this entry and it is the least replicated claim in the beat.
Evidence that argues against this (1)
- A study designed to test the KMT2A link found IMPDH-inhibitor sensitivity independent of KMT2A status and attributed the therapeutically relevant effect to SAMHD1 rather than to the MLL complex; and an AML leukaemia-stem-cell study reaches the antileukaemic endpoint through ribosome biogenesis, TP53 and MYC with no MLL complex in the account at all. Kawano G et al. 2026
- E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Sensitivity to IMPDH inhibitors was independent of KMT2A status in AML cell lines and primary AML samples; synergy with cytarabine and fludarabine was SAMHD1-dependent and present in a subset of AML cells but absent in acute lymphoblastic leukaemia cell lines.
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. An absence of a detected difference in a panel of unstated size, not a demonstrated equivalence. Endpoint is viability and nucleoside-analogue synergy, not complex assembly, so it does not directly test case C. Cell lines and primary samples; no in vivo arm; ALL negative is in cell lines only.
Evidence that argues against this (1)
- The earlier MLL-fusion work reports in vitro susceptibility of MLL-fusion AMLs and in vivo suppression of MLL-AF9-driven AML development. A single-agent viability panel is not the assay that would detect a mechanism whose prediction is menin-inhibitor sensitisation. Liu X et al. 2023
- E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
A recurrent KMT2A::AFF1-derived fusion circular RNA (AK_7_3) is present in the majority of tested t(4;11)-positive B-ALL patients and absent in t(4;11)-negative patients; its knockdown in KMT2A::AFF1 cell lines increases reactive oxygen species, mitochondrial membrane potential and mitochondrial respiration, produces dysmorphic mitochondria and induces apoptosis - so a fusion-derived product directly sets a mitochondrial set-point in the index fusion.
A novel KMT2A::AFF1-derived fusion circRNA regulates mitochondrial metabolism in t(4;11) B-cell acute lymphoblastic leukemia. · Tolomeo D, Bardini M, Venuto S, ..., Storlazzi CT · Haematologica 2026What this evidence cannot show
Abstract read this session; no PMCID returned, full text not retrieved. No in vivo data and no identified molecular target for the circRNA. The standard circRNA control - showing the junction-spanning knockdown reagent spares the linear fusion transcript - is not described in the abstract, and if it fails then the phenotype belongs to the fusion protein rather than the circle.
Evidence that argues against this (1)
- If the knockdown also reduces the linear KMT2A::AFF1 transcript, the mitochondrial phenotype is an effect of reducing the fusion oncoprotein, which would move this from a distinct case-A example to a restatement of the general proposition that the fusion sets the cell's state. Morris V et al. 2022
- E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper
AML cell lines show high metabolic heterogeneity; cells with MLL/AF9 have upregulated mitochondrial activity and depend mainly on oxidative phosphorylation for energy production, and metformin repressed their proliferation by inhibiting mitochondrial respiration.
High Metabolic Dependence on Oxidative Phosphorylation Drives Sensitivity to Metformin Treatment in MLL/AF9 Acute Myeloid Leukemia. · Liu L, Patnana PK, Xie X, ..., Khandanpour C · Cancers (Basel) 2022What this evidence cannot show
Abstract read this session; identifiers re-fetched from PubMed rather than copied from the sibling cell-state entry that already cites this paper. Cell lines only; no in vivo or patient data in the abstract. Metformin is a weak and pleiotropic complex I inhibitor requiring OCT1-mediated uptake, and its use as a mechanistic probe for OXPHOS dependence is a recognised weakness. 'MLL/AF9 genotype cell lines' within a heterogeneous panel is a small-n comparison.
Evidence that argues against this (1)
- The same axis measured in the lymphoid form of the same lesion gives the opposite operational answer: OXPHOS inhibition blunts blasts while promoting leukaemia-initiating cell emergence, and it is glycolysis and hypoxia, not OXPHOS, whose inhibition impairs the initiating compartment. Morris V et al. 2022
- E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In MLL-rearranged B-ALL, while inhibiting oxidative phosphorylation blunts blast proliferation, this intervention promotes leukaemia-initiating cell emergence; conversely, inhibiting hypoxia and glycolysis impairs MLL-rearranged B-ALL leukaemia-initiating cells and provides therapeutic benefit in xenotransplantation systems.
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. Xenotransplantation engraftment is a proxy for leukaemia-initiating capacity, not a direct measure. The specific OXPHOS inhibitor is not named in the abstract. MLL-rearranged B-ALL as a class; infants are not separately reported.
Evidence that argues against this (1)
- In the myeloid form, complex I inhibition combined with venetoclax reduces KMT2A-rearranged AML progression in PDX models of both onset and relapse, and a separate study makes complex I inhibition the manoeuvre that creates an exploitable MCL-1 dependency. The lineage disagreement is real and has never been adjudicated in one system. Da Ros A et al. 2026
clearest published case of a metabolic requirement that is a consequence of the chromatin programme
the case where causation runs from metabolism to the fusion complex
the case where the metabolic requirement is downstream of the chromatin programme
best-specified consequence chain: fusion to direct target gene to metabolic phenotype
a fusion product that directly sets a mitochondrial set-point, in the index fusion
the lineage split that cuts across all three directions
glycolysis inhibition, not OXPHOS inhibition, impairs MLL-r B-ALL initiating cells
states the sharper testable version of the lineage confound
- related to →KMT2A / MLL1 — the gene that breaks
frames whether the fusion programme causes or merely coexists with the metabolic phenotype
case B: the dependency measured as independent of KMT2A status
a worked case-A example from another fusion family
also discriminates assay artefact from biological difference
a menin-driven metabolic axis that is not downstream of the fusion