ACSL4 - a CRISPR-screen-derived, functionally validated KMT2A-selective vulnerability in myeloid disease, and a gene whose loss normally protects cells from ferroptosis
also called ACSL4 · ACS4 · FACL4 · LACS4 · acyl-CoA synthetase long-chain family member 4 · O60488
ACSL4 activates long polyunsaturated fats so the cell can use them, and KMT2A-rearranged myeloid leukaemia needs it much more than other leukaemias do. Turning it off slows those cells and delays the disease in mice - and feeding the cells the fats back rescues them, which is how the authors showed the fat handling is really the reason.
The same thing, for a scientist
ACSL4 (UniProt O60488, 711 aa, primary gene name read back 2026-09-20) ligates long-chain and in particular polyunsaturated fatty acids to coenzyme A. It was identified from large-scale CRISPR-Cas9 screening data as a **selective** vulnerability of KMT2A-rearranged AML and then validated functionally: CRISPR interference and shRNA knockdown impaired growth of KMT2Ar but not non-KMT2Ar AML cells, reduced colony formation in patient-derived KMT2Ar AML cells and in murine MLL-AF9 cells, and delayed leukaemia onset in vivo in MLL-AF9 mice; multi-omics showed depletion of polyunsaturated lipid species with compensatory activation of lipid metabolic pathways, and **exogenous PUFA supplementation rescued the growth defect**, which is the experiment that makes the dependency a lipid dependency rather than a generic fitness effect [E1]. The authors derive a KMT2Ar-ACSL4 dependency signature (KRADS12) that correlates with KMT2Ar status and predicts poor survival in AML [E1]. **The most important thing to hold in mind is the sign inversion against the wider literature**: ACSL4 is the canonical ferroptosis-sensitising enzyme, and its loss makes most cells *resistant* to ferroptosis. Here loss is antileukaemic. The rescue experiment is what adjudicates, and the ferroptosis literature is what a design team would otherwise reason from and get backwards.
This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.
I have never read one of these pictures before
A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.
Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.
Anything in yellow sticks is a bound small molecule, which is usually the drug.
Why this is a gene entry and not only a line in the lipid state entry
This graph already carries cell-state-lipid-pufa-dependence, which describes the ACSL4-versus-ELOVL1 split across lineages. That entry is a state; this one is the enzyme, and it exists because the PI's standing granularity rule says an enzyme inside a mechanism gets its own entry, and because ACSL4 carries specifics a state entry cannot: an accession, a paralogue family (ACSL1/3/4/5/6), a CRISPR-screen provenance, a rescue experiment, and a ferroptosis sign problem that belongs to the protein rather than to the state.
ACSL4 was named in 4 entries of this graph on 2026-09-20 (word-boundary count) and had none of its own.
The ferroptosis inversion, stated plainly
In the general cancer-biology literature ACSL4 incorporates polyunsaturated fatty acids into membrane phospholipids, which are the substrate for lipid peroxidation, so ACSL4 loss confers ferroptosis resistance. A reader who knows that literature and reads 'ACSL4 is a vulnerability' will predict the wrong direction for any combination involving a ferroptosis inducer: adding a GPX4 inhibitor to ACSL4 knockdown should work less well, not more.
The paper does not resolve this; its abstract does not mention ferroptosis at all 1. What it does provide is the PUFA rescue, which says the growth defect is caused by defective PUFA utilisation rather than by altered peroxidation susceptibility 1. Those are compatible but they are not the same claim, and this entry records the tension rather than smoothing it.
There is one KMT2A-adjacent ferroptosis paper in the literature - matrine enhancing RSL3-induced ferroptosis in MLL-rearranged AML cells - and it is a natural-product combination study that does not address ACSL4 3. It is cited here as the only bounded contact point between the two literatures in this disease, not as support.
| uniprot | O60488 |
|---|---|
| length AA | 711 |
| alphafold ID | AF-O60488-F1 |
| ec number | 6.2.1.3 |
| pathway | long-chain and polyunsaturated fatty acid activation to acyl-CoA |
| resolution note | gene_exact:ACSL4 AND organism_id:9606 AND reviewed:true returned 1 reviewed row on 2026-09-20; gene_primary read back as ACSL4; length 711 aa asserted, with the caveat that ACSL1 (698 aa) and ACSL3 (720 aa) sit close enough that length is a weak discriminator in this family. |
| evidence lineage | myeloid only (KMT2Ar AML lines, patient KMT2Ar AML cells, MLL-AF9 mice) |
| provenance of finding | large-scale CRISPR-Cas9 screening data, then CRISPRi and shRNA validation, then in vivo |
| rescue experiment | exogenous PUFA supplementation restores growth - the experiment that makes this a lipid dependency |
| signature | KRADS12, correlates with KMT2Ar status and predicts poor survival in AML (internal derivation, no external validation) |
| direction in disease | pro-leukaemic; loss is antileukaemic in KMT2Ar AML, which is the opposite of the ferroptosis-sensitisation direction reported for ACSL4 elsewhere |
retired symbols for literature search
[ "FACL4", "LACS4", "ACS4" ]
depmap kmt2a selectivity
{
"chronos_mean_kmt2a_rearranged": -0.7172,
"chronos_mean_other_leukaemia": -0.2574,
"difference": -0.4598,
"pan_essentiality_all_lines": -0.1175,
"selectivity_rank": 22,
"rank_denominator": 17728,
"provenance": "Joined 2026-09-21 from analyses/depmap-kmt2a-selective-scan-full (generated 2026-09-19T10:12:06Z), the project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies here, including that the KMT2A-rearranged arm is a small number of lines drawn from nine individuals, that the rank denominator is 17,728 genes, and that a difference is not a dependency. Added because only 26 of 308 gene entries carried this number, which made any KMT2A-selectivity question answerable for 8% of the corpus.",
"how_to_read": "A negative difference means MORE required in KMT2A-rearranged lines than in other leukaemia lines. A rank near 17,728 is ANTI-selective. Neither is evidence of a fusion-driven dependency on its own: check pan_essentiality_all_lines before calling anything selective, and note that a paralogue-redundant gene can score near zero for reasons that have nothing to do with this disease - see ana-the-paralogue-redundancy-excuse-is-not-tested-and-my-proxy-decided-the-answer."
}- ACSL4 loss is the canonical route to ferroptosis resistance. An ACSL4-directed therapy could select for a leukaemia that is harder to kill with the oxidative-stress mechanisms several existing agents depend on.
- The cells compensate: the study itself reports compensatory activation of lipid metabolic pathways on ACSL4 loss, so a partial pharmacological inhibition is likely to be buffered in a way a genetic knockdown is not.
- The rescue is dietary. Exogenous polyunsaturated fatty acids restore growth, and PUFAs are present in enteral and parenteral nutrition given to children on intensive chemotherapy. A patient's lipid emulsion could antagonise the drug.
- No isoform-selective inhibitor exists. Delivering this would mean either an undeveloped chemistry programme or a pan-ACSL agent whose effects on normal marrow, brain and liver lipid handling are not characterised in children.
- Does ACSL4 loss make KMT2A-rearranged AML resistant to ferroptosis inducers while making it grow more slowly? If so, any combination of an ACSL4 strategy with a GPX4 inhibitor is antagonistic by construction and should not be attempted.
- Is ACSL4 a dependency in infant KMT2A::AFF1 B-lineage cells? The lymphoid fatty-acid requirement identified to date is ELOVL1, on the other side of the saturation divide, and nobody has tested ACSL4 there.
- Is there any isoform-selective ACSL4 chemistry? Without it the finding is a target-validation result with no tool, and the available pan-ACSL compounds cannot test the hypothesis.
- Does the DepMap-derived selectivity survive a lineage split? The KMT2A-rearranged arm in DepMap is myeloid-enriched, and this project has already had two ranked hits turn out to be lineage artefacts.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Analysis of large-scale CRISPR-Cas9 screening data identified ACSL4 as a selective vulnerability in KMT2A-rearranged AML; CRISPR interference and shRNA knockdown confirmed that ACSL4 loss impairs growth of KMT2Ar but not non-KMT2Ar AML cells, reduced colony formation in cells from patients with KMT2Ar AML and in murine MLL-AF9 cells, and delayed leukaemia onset in vivo in MLL-AF9 mice; transcriptomics, proteomics and lipidomics revealed depletion of polyunsaturated lipid species with compensatory activation of lipid metabolic pathways on ACSL4 loss; supplementation with exogenous polyunsaturated fatty acids rescued the growth defect; a KMT2Ar-ACSL4 dependency signature (KRADS12) correlates with KMT2Ar status and predicts poor survival in AML patients.
ACSL4-associated lipid metabolism is a distinct therapeutic vulnerability in KMT2A-rearranged acute myeloid leukemia. · Schafer S, Rahimian E, Schmitz-Hubsch L, ..., Wurm AA · Cell Rep 2026What this evidence cannot show
Abstract read this session; full text not retrieved, so the effect sizes, the number of KMT2Ar versus non-KMT2Ar lines and the depth of the in vivo delay were not checked. **Myeloid throughout** - AML cell lines, patient KMT2Ar AML cells and MLL-AF9 mice; no lymphoid or infant KMT2A::AFF1 data, and the companion enzyme in infant ALL is a different one (ELOVL1). DepMap-derived selectivity claims in this disease carry a documented lineage confound, because the KMT2A-rearranged arm of DepMap is myeloid-enriched; the functional validation here is what carries the claim, not the screen. KRADS12 is a signature derived and evaluated in the same study and has no external validation. No selective ACSL4 inhibitor suitable for clinical use is described.
Evidence that argues against this (1)
- In the infant KMT2A::AFF1 lymphoid disease the identified fatty-acid requirement is ELOVL1, a very-long-chain elongase acting on saturated and monounsaturated substrates - the opposite side of the saturated/polyunsaturated divide from ACSL4. The same founding lesion therefore does not produce the same lipid dependency across lineages, so ACSL4 must not be generalised to KMT2A-rearranged disease as a class. Symeonidou V et al. 2021
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
UniProt reviewed human record O60488 (ACSL4_HUMAN) has primary gene name ACSL4, gene synonyms ACS4, FACL4 and LACS4, length 711 aa, and recommended protein name 'Long-chain-fatty-acid--CoA ligase 4' with alternative name 'Arachidonate--CoA ligase'.
UniProtKB entry O60488 (ACSL4_HUMAN), Long-chain-fatty-acid--CoA ligase 4 · UniProt Consortium · UniProt 2026What this evidence cannot show
Resolved 2026-09-20 with gene_exact:ACSL4 AND organism_id:9606 AND reviewed:true; exactly one reviewed row and gene_primary read back as ACSL4. The ACSL family has six human members (ACSL1, ACSL3, ACSL4, ACSL5, ACSL6, and the bubblegum-family ACSBG enzymes) with overlapping substrate preferences, and the retired symbols FACL4/LACS4 are the forms used in the older literature - a PubMed enumeration keyed only to 'ACSL4' will under-recover pre-2005 work. Length 711 aa separates ACSL4 comfortably from ACSL1 (698) and ACSL3 (720) only by a few tens of residues, so the length assertion is a weak defence here and the primary symbol is the real one.
Evidence that argues against this (1)
- A reviewed record cannot establish which ACSL isoform carries the activity measured in leukaemic cells; the cited study used genetic knockdown of ACSL4 specifically, which is stronger evidence than any pharmacology in this family, because the available ACSL inhibitors (for example triacsin C) are not isoform-selective. Schafer S et al. 2026
- E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper
The only KMT2A/MLL-titled ferroptosis paper recoverable on 2026-09-20 reports that matrine enhances RSL3-induced ferroptosis at non-toxic doses in acute myeloid leukaemia cells with MLL rearrangement; it does not address ACSL4.
Matrine enhances RLS3-induced ferroptosis at non-toxic doses in acute myeloid leukemia cells with MLL rearrangement. · see PubMed record; title fetched from Europe PMC 2026-09-20 · journal per PubMed record 2026What this evidence cannot show
**Title only.** This record was surfaced by a Europe PMC title-check sweep on 2026-09-20 and its abstract was not fetched and the paper was not read; the author string is therefore deliberately not asserted, per this project's rule against writing an identifier or author from anything other than a fetched record. Note also that the title as indexed reads 'RLS3', which is almost certainly a typographical error for the GPX4 inhibitor RSL3 - a further reason to treat this record as unread rather than as evidence. It is recorded here only to bound the statement that the ACSL4 and ferroptosis literatures barely touch in this disease, and it should be read before being cited for anything.
Evidence that argues against this (1)
- Because this record is title-only and the title itself contains a probable transcription error in the compound name, nothing in it should be treated as established. The correct disposition is a reading-queue item, not a claim. Schafer S et al. 2026
the myeloid enzyme inside the lipid-dependent state entry
selective vulnerability of KMT2Ar AML, validated by CRISPRi, shRNA and in vivo MLL-AF9
- depends on →KMT2A::MLLT3 (MLL-AF9), t(9;11)(p21.3;q23.3)
ACSL4 knockdown reduced colony formation in murine MLL-AF9 cells and delayed leukaemia onset
opposite side of the saturated/polyunsaturated divide; the lineage split in lipid dependency