BCAT1 in core-binding-factor AML - the closest analogue in another fusion family, where TWO different fusions converge on one metabolic enzyme by rewiring the chromatin at its locus
also called BCAT1 in CBF-AML · branched-chain amino acid dependency in RUNX1::ETO and CBFB::MYH11
A different family of leukaemia fusions - the ones that break the core-binding-factor genes - switches on an enzyme that handles branched-chain amino acids, and the leukaemia then depends on it. Two unrelated fusions do this the same way, by changing how the DNA is folded at that gene, and an old epilepsy drug that blocks the enzyme helped mice live longer.
The same thing, for a scientist
The PI's standing priority is analogous biology in other fusion families. For metabolism the best available analogue is **BCAT1 in core-binding-factor AML**: BCAT1 expression is higher in CBF-AML patients than in other subtypes; high BCAT1 reprogrammes branched-chain amino acid metabolism and associates with a sphingolipid pathway relating to leukaemic cell fitness; **BCAT1 is directly activated by both RUNX1::ETO and CBFB::MYH11 in a RUNX1-dependent manner through rewiring chromatin conformation at the BCAT1 locus**; BCAT1 inhibition blunts the cell cycle, enhances apoptosis and induces myeloid differentiation in vitro and alleviates leukaemia burden and prolongs survival in vivo; and pharmacological inhibition with **gabapentin** delayed leukaemia progression and improved survival in vivo [E1]. **The structural parallel to KMT2A is exact and that is what makes it useful.** A chromatin-acting fusion oncoprotein directly rewires the regulatory architecture of a metabolic gene, and the metabolic gene then becomes a dependency. That is precisely the shape of the SAM/MAT2A and PTBP1/serine mechanisms in KMT2A-rearranged disease - a case-A 'consequence of the fusion programme' dependency in the taxonomy of `concept-metabolic-dependency-consequence-or-independent-kmt2a`.
The mapping, and where it holds
| CBF-AML | KMT2A-rearranged leukaemia |
|---|---|
| two distinct fusions (RUNX1::ETO, CBFB::MYH11) converge on one metabolic enzyme | multiple KMT2A fusions converge on a HOXA/MEIS1 programme; metabolic convergence is NOT established |
| mechanism is chromatin conformation rewiring at the metabolic gene's locus | the KMT2A fusion acts on chromatin at target loci; PTBP1 is a named direct target with a metabolic output 2 |
| the dependency is RUNX1-dependent, i.e. it runs through the fusion's own DNA-binding partner | the equivalent would be a menin- or LEDGF-dependent metabolic target gene - nobody has looked for one |
| a repurposable approved drug exists (gabapentin) | the repurposable candidates are immunosuppressants, which is a much worse fit for an infant |
Where the analogy is strongest: the demonstration that the mechanism of the metabolic dependency is direct transcriptional/architectural activation of the enzyme's locus by the fusion. That is a specific, testable template for KMT2A, and it suggests an experiment nobody has run: take published KMT2A-fusion chromatin occupancy data and ask which metabolic enzyme genes carry a fusion-bound element. This graph holds several such datasets already.
Where it is weakest: CBF-AML is an adult-and-older-child myeloid disease with a comparatively favourable prognosis, driven by fusions that act mainly by dominant-negative repression of RUNX1 function. KMT2A fusions act by ectopic transcriptional elongation. The cell of origin, the lineage and the mechanism class all differ, so a specific transfer of BCAT1 itself would be a stretch.
Is BCAT1 itself relevant to KMT2A-rearranged disease?
No, on the available evidence, and the absence is bounded. Measured 2026-09-20 with the instrument in ana-metabolic-enzyme-enumeration-kmt2a-bounded-absences: BCAT1 AND (MLL OR KMT2A) AND leukemia returns 114 Europe PMC records, of which 0 have BCAT1 in title or abstract together with KMT2A/MLL in the title. The same instrument on the same day returned 3 for IMPDH2, 2 for ACSL4 and 2 for MAT2A, so it is not a broken query.
The records returned are reviews of leukaemic-stem-cell metabolism and epigenetics that mention both topics, plus a branched-chain paper (GPRC5C) that is not KMT2A-specific. So the honest statement is: the analogy is structural, not molecular. BCAT1 is a template for how to look, not a target to transfer.
This matters because the easy error here would be to write 'BCAT1 is a candidate in KMT2A-rearranged leukaemia' on the strength of the analogy. It is not a candidate; it is an unexamined gene in this disease, which is a different and weaker statement.
| domain | ecology |
|---|---|
| bcat1 in kmt2a | BOUNDED ABSENCE: BCAT1 AND (MLL OR KMT2A) AND leukemia returns 114 Europe PMC records on 2026-09-20, 0 passing the title check; instrument positive controls IMPDH2=3, ACSL4=2, MAT2A=2 the same day |
| transferable experiment | re-analyse existing KMT2A-fusion chromatin occupancy datasets in this graph for fusion-bound elements at metabolic enzyme loci - the BCAT1 question asked of the index fusion |
| caution on the source | the source abstract calls gabapentin a specific BCAT1 inhibitor; gabapentin's principal pharmacological target is the CACNA2D subunit, and this entry does not adopt that characterisation |
mapping
[
{
"bio": "KMT2A fusion directly activates a target gene whose output is metabolic (PTBP1 -> PKM2 -> glycolytic outflow)",
"analog": "RUNX1::ETO and CBFB::MYH11 directly activate BCAT1 by rewiring chromatin conformation at its locus"
},
{
"bio": "multiple KMT2A fusions converge on a shared HOXA/MEIS1 transcriptional programme",
"analog": "two structurally unrelated CBF fusions converge on one metabolic enzyme, RUNX1-dependently"
},
{
"bio": "the KMT2A repurposing candidates are immunosuppressants, a poor fit for an infant",
"analog": "the CBF repurposing candidate is an anticonvulsant already used in children"
},
{
"bio": "no KMT2A-fusion-bound regulatory element at a metabolic enzyme gene has been reported",
"analog": "the CBF work located exactly such an element and made it the mechanism"
}
]generated hypotheses
[ "hyp-metabolic-split-is-ontogeny-not-lineage" ]
- The most tempting misuse of this entry is to propose BCAT1 or gabapentin in KMT2A-rearranged leukaemia. There is no KMT2A evidence for either, and the counted absence is recorded here specifically to prevent that inference.
- Branched-chain amino acids are essential nutrients required for growth and protein synthesis. A BCAA-directed strategy in an infant carries the same nutritional objection as methionine restriction and a worse one, because leucine is a direct mTORC1 signal.
- The analogy's strength is architectural, and architecture transfers badly: CBF fusions act by dominant-negative interference with RUNX1, KMT2A fusions by ectopic elongation. Two different mechanisms can produce the same-looking regulatory rewiring and respond completely differently to intervention.
- Which metabolic enzyme genes carry a KMT2A-fusion-bound regulatory element? This graph already holds fusion occupancy datasets; the BCAT1 paper shows what the answer would look like, and nobody has asked.
- Is there a menin- or LEDGF-dependent metabolic target gene, as BCAT1 is RUNX1-dependent in CBF-AML? A metabolic gene whose activation requires the menin interface would be directly on this lab's hypothesis.
- Do NUP98-fusion leukaemias have a comparable direct metabolic target? Neither this beat nor the existing NUP98 entries in this graph identify one, and the NUP98 metabolic literature returned mostly nucleoporin-trafficking and review material on 2026-09-20.
- Is gabapentin's effect in CBF-AML actually BCAT1-mediated? If a widely used, well-tolerated anticonvulsant genuinely has an on-target antileukaemic mechanism, that is worth knowing precisely, and the specificity claim in the source is not one this entry accepts.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Integrating expression profile and genetic makeup in an AML cohort found higher BCAT1 expression in core-binding-factor AML patients than in other subtypes; metabolic profiling showed high BCAT1 expression leads to reprogrammed branched-chain amino acid metabolism in CBF-AML, associated with a sphingolipid pathway relating to leukaemia cell fitness; in cell lines and primary patient samples BCAT1 was directly activated by RUNX1::ETO and CBFB::MYH11 fusion proteins similarly in a RUNX1-dependent manner through rewiring chromatin conformation at the BCAT1 gene locus; BCAT1 inhibition blunted the cell cycle, enhanced apoptosis and induced myeloid differentiation of CBF-AML cells in vitro and alleviated leukaemia burden and prolonged survival in vivo; and pharmacological inhibition of BCAT1 using the specific inhibitor gabapentin demonstrated therapeutic effects with delayed leukaemia progression and improved survival in vivo.
RUNX1::ETO and CBFbeta::MYH11 converge on aberrant activation of BCAT1 to confer a therapeutic vulnerability in core-binding factor-acute myeloid leukaemia. · Wang S, Liu Y, Zhao X, ..., Wang K · Br J Haematol 2024What this evidence cannot show
Abstract read this session; full text not retrieved, so the chromatin-conformation assay, the cohort size and the in vivo model were not checked. **CBF-AML, not KMT2A-rearranged disease** - this is cited as an analogy, not as evidence in the index disease. The abstract describes gabapentin as a 'specific inhibitor' of BCAT1; gabapentin is a widely used anticonvulsant whose principal pharmacological target is the alpha-2-delta subunit of voltage-gated calcium channels, and describing it as a specific BCAT1 inhibitor is a claim this entry does not adopt. The expression difference is a cohort association and the causal chain rests on the fusion-activation experiments rather than on it.
Evidence that argues against this (1)
- BCAT1 has no KMT2A-specific evidence at all. Measured 2026-09-20 on Europe PMC: BCAT1 AND (MLL OR KMT2A) AND leukemia returns 114 records of which 0 carry BCAT1 in title or abstract together with KMT2A/MLL in the title, on an instrument that returns 3 for IMPDH2, 2 for ACSL4 and 2 for MAT2A the same day. The analogy is therefore structural and must not be read as nominating BCAT1 as a KMT2A target. Schafer S et al. 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In KMT2A-rearranged disease the structurally equivalent finding is that the direct MLL-ENL target gene PTBP1 is rate-limiting for proliferation and produces the metabolic phenotype - reduced glucose consumption, reduced lactate production and loss of the PKM2 isoform - which is the same shape of mechanism (fusion directly activates a gene whose output is metabolic) reached in a different fusion family.
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. PTBP1 is a splicing factor rather than a metabolic enzyme, so the parallel with BCAT1 is one step less direct: the CBF fusions activate the enzyme itself, whereas MLL-ENL activates a regulator whose downstream output is metabolic. MLL-ENL is not the index fusion. No study has asked the BCAT1-shaped question in KMT2A-rearranged disease - which metabolic enzyme genes carry a fusion-bound regulatory element.
Evidence that argues against this (1)
- Because the KMT2A example runs through a splicing factor rather than through direct activation of a metabolic enzyme locus, the analogy may be pointing at a gap rather than at a parallel: the direct fusion-to-enzyme architecture demonstrated in CBF-AML has simply not been looked for in KMT2A-rearranged leukaemia, and its absence from the literature is not evidence that it does not exist. Wang S et al. 2024
- analogous to →CBFB::MYH11 (CBFbeta-SMMHC), inv(16)(p13;q22)
CBFB::MYH11 directly activates BCAT1 by rewiring chromatin at its locus
- analogous to →RUNX1::RUNX1T1 (AML1-ETO), t(8;21)(q22;q22)
RUNX1::ETO converges on the same metabolic enzyme in a RUNX1-dependent manner
same architecture: fusion directly activates a locus whose output is metabolic
a worked case-A example from another fusion family
the BCAT1 bounded absence in KMT2A-rearranged disease was measured here