Oncofusion

start / graph / gene-elp4

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q96-sarcoma-mbm-partners

ELP4 (PAXNEB) - a required subunit of Elongator, the one partner of the seven with a genuine histone-acetyltransferase connection, and the one whose fusion supplies only its last 43 residues

also called ELP4 · PAXNEB · C11orf19 · hELP4 · Elongator complex protein 4 · PAX6 neighbor gene protein · Q96EB1

ELP4 is one of six parts of Elongator, a machine that rides with the gene-reading enzyme and chemically tags both transfer RNAs and, in some experiments, histones. In one patient's angiosarcoma KMT2A broke at exon 9 and joined ELP4's very last exon, so the fusion carries only the final 43 of ELP4's 424 amino acids.

The same thing, for a scientist

ELP4 (Q96EB1, 424 aa) is one of the six subunits of **Elongator** (ELP1-ELP6), which assembles as two subcomplexes, ELP1/2/3 and ELP4/5/6 [E1, E2]. Two things make it the most interesting of the seven partners and the most easily over-claimed. **First, it is the only one with a real histone connection.** Intact Elongator is a **histone H3 and H4 acetyltransferase** whose predominant sites are H3K14 and H4K8, it acetylates both free and nucleosomal substrates, and **ELP4, ELP5 and ELP6 are explicitly required for that HAT activity** while isolated ELP3 is not specific [E3]. Human Elongator purified from HeLa carries H3-directed HAT activity, binds RNA polymerase II directly, and facilitates Pol II transcription in a **chromatin- and acetyl-CoA-dependent** manner [E4]. That puts ELP4 inside the exact mechanism this graph is built around - a Pol II elongation complex that writes an acetyl mark. **Second, the field has moved and the entry has to say so.** By 2018 the ELP3 HAT domain is described as **putative**, Elongator's functions are listed as transcription regulation, alpha-tubulin acetylation *and* tRNA modification, and the ELP4/5/6 subcomplex's own measured activity is **ATP-modulated tRNA binding** [E5]. Measured 2026-09-22: `Elongator[tiab] AND tRNA[tiab]` returns **334** records against **86** for `Elongator[tiab] AND acetyltransferase[tiab]`. UniProt's own FUNCTION comment for ELP4 is tRNA-only, and it is `ECO:0000303` - an author statement, not an experiment [E1]. **And the fusion carries almost none of it.** In case S15 the fusion is `KMT2A exon 9 :: ELP4 exon 10` [E6]. ELP4's transcript NM_019040 has **ten** exons and **exon 10 is the last one, coding aa 382-424** - 43 residues, 10% of the protein [E7]. KMT2A supplies aa 1-1406, the complete leukaemic N-terminal module, with the break **inside the leukaemic breakpoint cluster region** [E8]. So this is a KMT2A N-terminal module with a 43-residue tail: **ELP4 supplies essentially no ELP4 moiety**, which is the mirror image of the YKY problem this graph already records for KMT2A.

Go deeper
What ELP4 normally does

ELP4 is a 424-residue subunit of the six-protein Elongator complex (ELP1, ELP2, ELP3, ELP4, ELP5, ELP6), which forms a ~850 kDa assembly of two copies of each subunit organised into an ELP1/2/3 catalytic lobe and an ELP4/5/6 lobe 25. UniProt annotates no domain, motif, repeat or zinc finger anywhere in ELP4 - the feature table for this protein is empty - so it is a subunit defined by its complex, not by a module 1.

UniProt's FUNCTION for ELP4 is entirely tRNA modification: Elongator is required for mcm5U, mcm5s2U and ncm5U, and catalyses carboxymethyluridine at the wobble position 34 1. That annotation is ECO:0000303 - a statement taken from an author's own review rather than an experiment, and worth recording as such.

The gene's other identity is positional. ELP4 sits immediately telomeric to PAX6 at 11p13, and its original name was PAXNEB (also C11orf19) - the gene was discovered because chromosomal rearrangements causing aniridia disrupt its unusually large final intron while leaving PAX6 intact 9. That intron contains the SIMO enhancer, and this is where the database gets it wrong: see the evidence note on E10.

What the fusion joins - and why ELP4's biology does not come with it
sidegeneexonsresiduesmodules supplied
5-primeKMT2A1-9aa 1-1406MBM 6-25, IBM1 123-134, IBM2 147-152, AT-hooks 169-309, CXXC 1147-1195
3-primeELP410 onlyaa 382-424 (43 residues, 10%)none - no annotated feature lies in that span

The source reports KMT2A exon 9 to ELP4 exon 10 (NM_019040) 6. Fetching that record: NM_019040 has ten exons, CDS 19..1293, and exon 10 spans nt 1162-8093 - it is the terminal exon, and it codes aa 382-424 before 7 kb of 3'-UTR 7. The junction is predicted in frame (exon 10 begins 1,143 nt after the ATG, divisible by three, and KMT2A exon 9 ends at a complete codon), but in-frame here buys almost nothing: 43 residues cannot assemble into Elongator, cannot bind ELP5/ELP6, and carry no annotated feature.

This is the same shape as the YKY problem in this graph's own reasoning. The king log records that exp-menin-pulldown-on-yky-versus-kmt2a-aff1 uses YAP1-KMT2A-YAP1, which inserts only KMT2A exons 4-5/6 and therefore supplies essentially no MLL1 moiety, making the framework indeterminate for it. Case S15 is the same failure on the other side of the junction: the partner supplies essentially no partner. An entry that led with Elongator's histone acetyltransferase activity would be describing a complex this fusion cannot join.

What the case is good for is the KMT2A side, and there it is one of the strongest in the cohort: exon 9 is inside the leukaemic breakpoint cluster region (exons 8-14), so the retained module - MBM, IBM1, IBM2, three AT-hooks, CXXC - is indistinguishable from an infant-leukaemia fusion's 8.

Epigenetics (PI dimension 3) - a complex-level histone acetyltransferase, and a contested one

ELP4 itself has no catalytic domain - no SET, JmjC, HAT, HDAC, bromodomain, chromodomain, Tudor, PHD or CXXC module, and in fact no annotated feature at all 1. The epigenetic content is entirely at the level of the complex, and it is real:

  • Intact Elongator acetylates the N-terminal tails of histone H3 and H4, on core and nucleosomal substrates, with H3K14 and H4K8 predominant; ELP4, ELP5 and ELP6 are required for the HAT activity; Elongator binds naked and nucleosomal DNA; and multiply-acetylated H3 and H4 fall in vivo in ELP3-null yeast 3. Yeast.
  • Human Elongator purified from HeLa has HAT activity specific to H3 and much weaker on H4, interacts directly with RNA polymerase II, and facilitates Pol II transcription in a chromatin- and acetyl-CoA-dependent manner 4. Human - but that purification recovered three polypeptides, so it does not itself place ELP4 in the HAT-active species.

So the honest statement is the gene-aff1 form: ELP4 writes, erases and reads nothing itself; it is a required subunit of a complex reported to write H3K14 and H4K8 acetylation, and that assignment is contested. The 2018 structural review calls ELP3's HAT domain putative, lists alpha-tubulin acetylation and tRNA modification alongside transcription, and gives the ELP4/5/6 lobe's own measured activity as ATP-modulated tRNA binding 5. The literature has voted with its feet: 334 records for Elongator + tRNA against 86 for Elongator + acetyltransferase, measured 2026-09-22 5.

The one-line version: of these seven partners ELP4 has the strongest epigenetic credential and the weakest route for it to reach the fusion, because the fusion keeps 43 of its 424 residues.

Localization, cell type and cell state (PI dimensions 1 and 2)

Subcellular: cytoplasm and nucleus - and the nuclear call is a rule-91 defect. UniProt gives ELP4 Cytoplasm (three ECO:0000269s) and Nucleus (one, PubMed:11889558). That paper describes no localisation experiment of any kind. It is the gene-discovery and synteny-conservation study: it maps ten exons, notes a 134 kb final intron, reports a 2 kb transcript expressed in all cell lines tested, and concludes with a transgenic mouse that the aniridia phenotype is not due to heterozygous PAXNEB loss 9. The same paper is cited for Tissue specificity: widely expressed, which it does support. So ELP4's nuclear compartment rests on a citation that is about the gene's structure and about PAX6.

Where does a nuclear pool actually come from? The functional case is indirect but better: human Elongator interacts directly with RNA polymerase II and acts on chromatin templates 4, which is a nuclear activity - and the cytoplasmic pool is well supported and is where the tRNA modification happens 1.

Subnuclear: no measurement. No speckle, nucleolar, lamina or PML-body assignment, and no ELP4 chromatin-occupancy profile. For a complex that travels with elongating Pol II, the address that would mean something is an occupancy coordinate along a gene rather than a compartment - the same argument this graph makes for gene-supt6h.

Cell type: widely expressed, with no resolution. Widely expressed in human, broadly expressed in mouse embryos 9. Cell state: the only state-linked result is in a different lineage - ELP5 and ELP6 knockdown reduces the tumorigenicity and migration of melanoma cells as Elongator subunits 10. Nothing in ELP4's literature is haematopoietic and nothing is mesenchymal or sarcoma.

Analogous biology in other fusion families (PI dimension 5)

ELP4 is not a fusion partner in any tumour. ELP4[tiab] AND fusion[tiab] = 1 (a case report of congenital lacrimal-gland agenesis with aniridia) and ELP4[tiab] AND translocation[tiab] = 4, all four being aniridia cytogenetics: PAX6 intrachromosomal translocation, aniridia-associated keratopathy, an insertional 15q25-q26 translocation into 11p13, and a chromosome-aberration survey in aniridia patients 9. Controls through the identical shape: EWSR1 + fusion 1,208, EWSR1 + translocation 462.

So every rearrangement in ELP4's literature is at the ELP4 locus and about the PAX6 gene next door. That is a genuinely instructive analogy for this project, and it is not the usual one: it is an example of a gene whose rearrangement literature is entirely about a neighbouring gene's regulation. A KMT2A breakpoint in ELP4's terminal exon would, on that precedent, be worth checking for a positional effect on PAX6 - not a claim, an untested question.

Within this cohort the structural analogy is the reverse of case S16: in KMT2A ex15::AHDC1 the partner exon is untranslated and contributes nothing, and in KMT2A ex9::ELP4 the partner exon is coding but contributes only its last 43 residues. Two of the seven cases supply essentially no partner protein, which is a material qualification on reading these as chimeric oncoproteins.

Status: what this case is and is not
  • One patient. Case S15, angiosarcoma - one of the three cases in this cohort breaking inside the leukaemic breakpoint cluster region 6.
  • The partner contributes 43 of 424 residues, so this case is evidence about KMT2A's N-terminus and very little about ELP4 7.
  • No fusion protein, no transcript sequence, no reciprocal product, no functional assay.
  • DepMap: the most essential of the seven, and still not selective. Pan-essentiality -0.459 across 1,208 lines - genuinely required, consistent with a core tRNA-modification complex - but arm means -0.353 (KMT2A-rearranged) against -0.380 (other leukaemia), difference +0.026, selectivity rank 12,422 of 17,728 11. A shared requirement, not a fusion-driven one.
  • Not a treatment rationale. See risk-a-menin-inhibitor-prescribed-off-a-kmt2a-gene-name-in-a-solid-tumour.
Facts and figures
uniprotQ96EB1
gene symbolELP4
length AA424
accession resolutiongene_exact:ELP4 AND organism_id:9606 AND reviewed:true returned exactly one row, primary gene name ELP4, entry ELP4_HUMAN, length 424 aa, synonyms C11orf19 and PAXNEB. Primary name read back and length asserted.
refseq transcriptNM_019040.4 (10 exons, CDS 19..1293, 424 aa; exon 10 is terminal, nt 1162-8093, coding aa 382-424)
retired symbol controlRun 2026-09-22 because the brief requires it: PAXNEB[tiab] = 1 and C11orf19[tiab] = 1, both the same record (PMID 11889558), against ELP4[tiab] = 78. UNLIKE the SEPTIN row in this project's collision registry, where the approved symbol is the low-recall one, the retired symbols here add nothing. Recorded as a negative so the next beat does not repeat it.
domain noteUniProt annotates NO domain, motif, repeat, zinc finger, coiled coil or DNA-binding feature anywhere in ELP4's 424 residues - the feature table is empty. Residue ranges are therefore deliberately omitted rather than guessed, per this graph's rule that a domain entry with a null boundary is invisible to every figure.
complexElongator (ELP1, ELP2, ELP3, ELP4, ELP5, ELP6), ~850 kDa, two copies of each subunit, organised as an ELP1/2/3 lobe and an ELP4/5/6 lobe
epigenetic readoutELP4 ITSELF WRITES, ERASES AND READS NOTHING - it has no SET, JmjC, HAT, HDAC, bromodomain, chromodomain, Tudor, PHD or CXXC module, and in fact NO annotated feature at all. BUT THIS IS THE STRONGEST EPIGENETIC CREDENTIAL OF THE SEVEN AND IT IS COMPLEX-LEVEL, in the gene-aff1 'scaffold that may deliver a writer' category. INTACT ELONGATOR IS A HISTONE H3 AND H4 ACETYLTRANSFERASE: predominant sites H3K14 and H4K8, active on core AND nucleosomal substrates, ELP4/ELP5/ELP6 explicitly REQUIRED for that HAT activity where isolated ELP3 is not specific, Elongator binds naked and nucleosomal DNA, and multiply-acetylated H3/H4 fall in vivo in ELP3-null cells (PMID 11904415 - YEAST). HUMAN Elongator purified from HeLa has H3-directed HAT activity, binds RNA POLYMERASE II directly, and facilitates Pol II transcription in a CHROMATIN- and ACETYL-CoA-DEPENDENT manner (PMID 11818576) - but that preparation is described as three polypeptides, so it does NOT place ELP4 in the HAT-active human species. AND THE ASSIGNMENT IS CONTESTED: by 2018 ELP3's HAT domain is called PUTATIVE, Elongator's functions are listed as transcription regulation, alpha-tubulin acetylation and tRNA modification, and the ELP4/5/6 lobe's own measured activity is ATP-modulated tRNA BINDING (PMID 29332244). Measured 2026-09-22 with one tag per term: Elongator[tiab] AND tRNA[tiab] = 334 against Elongator[tiab] AND acetyltransferase[tiab] = 86, and ELP3[tiab] AND histone[tiab] = 70. UniProt's own FUNCTION comment for ELP4 is tRNA-only and is ECO:0000303, an author statement. THE DECISIVE QUALIFICATION FOR THIS BEAT: none of this reaches the fusion, because case S15 retains only ELP4's terminal exon, 43 of 424 residues, which cannot assemble into Elongator.
mark writtenH3K14ac and H4K8ac, AT THE LEVEL OF THE INTACT ELONGATOR COMPLEX and in yeast; ELP4 is a required subunit rather than the catalytic one, and the assignment of the catalysis to ELP3's HAT domain is described as putative in the current structural literature. Also reported for the complex: alpha-tubulin acetylation - a NON-HISTONE substrate.
non histone substratealpha-tubulin acetylation is attributed to Elongator, and the complex's best-supported activity is tRNA wobble-base modification (mcm5U, mcm5s2U, ncm5U) - an EPITRANSCRIPTOMIC rather than chromatin activity, the same distinction this graph draws for gene-igf2bp3.
subnuclear localizationNO MEASUREMENT OF ANY KIND - no speckle, nucleolar, lamina or PML-body assignment, and no ELP4 chromatin-occupancy profile in any system. For a complex that travels with elongating RNA polymerase II the meaningful address is an OCCUPANCY COORDINATE ALONG A GENE rather than a compartment, which is the argument this graph already makes for gene-supt6h: the missing measurement is a ChIP, not an image.
cell typeWIDELY EXPRESSED WITH NO RESOLUTION. 'Widely expressed' in human (PMID 11889558), a 2 kb transcript in all cell lines tested, and broad expression of the mouse orthologue in embryos. No cell-type or lineage-resolved measurement exists, so for this graph's purposes ELP4's cell-type dimension is genuinely unpublished rather than merely unrecorded.
cell stateThe only state-linked result for the ELP4/5/6 lobe is in a different lineage: knockdown of ELP5 or ELP6 reduces the TUMORIGENICITY AND MIGRATION of melanoma cells as Elongator subunits (PMID 22854966). ELP4 itself has no loss-of-function phenotype reported in any tumour type.
localization shown for fusionNO - 'nobody has looked in any fusion' (rule 86). ELP4[tiab] AND KMT2A[tiab] = 0 and AND MLL[tiab] = 0, and no sarcoma KMT2A fusion protein has been visualised by any method.
fusion partner in other tumoursNOT A FUSION PARTNER ANYWHERE - and the shape of the negative is itself the analogy. ELP4[tiab] AND fusion[tiab] = 1 (PMID 40226226, a congenital lacrimal-gland agenesis and aniridia case report) and ELP4[tiab] AND translocation[tiab] = 4, and ALL FOUR are aniridia cytogenetics about the PAX6 gene next door: PMID 42133303 (copy-number-neutral intrachromosomal PAX6 translocation), 34102310 (aniridia-associated keratopathy), 22991255 (insertional 15q25-q26 translocation into 11p13) and 12386836 (chromosome aberrations in aniridia patients). Controls: EWSR1[tiab] AND fusion[tiab] = 1,208; AND translocation[tiab] = 462. So every rearrangement in ELP4's literature is at the ELP4 LOCUS and about a NEIGHBOURING gene's regulation - an instructive and untested question for case S15, not a claim. Inside this cohort the structural analogy is case S16 (KMT2A ex15::AHDC1), where the partner exon is untranslated: TWO of the seven cases supply essentially no partner protein.
absence classnever-measured
aliases and retired symbols searched
[
  "ELP4",
  "PAXNEB",
  "C11orf19",
  "hELP4"
]
domains
[]
functions
[
  "subunit of the Elongator complex",
  "required for tRNA wobble modifications mcm5U, mcm5s2U and ncm5U (ECO:0000303, author statement)",
  "required for the histone H3/H4 acetyltransferase activity of intact Elongator (yeast)",
  "ATP-modulated tRNA binding as part of the ELP4/5/6 subcomplex"
]
sarcoma case
{
  "patient": "S15",
  "diagnosis": "angiosarcoma",
  "fusion": "KMT2A exon 9 :: ELP4 exon 10 (NM_019040)",
  "kmt2a_polarity": "5-prime - the LEUKAEMIC orientation, opposite to the recurrent sarcoma arms",
  "kmt2a_exons_retained": "1-9",
  "kmt2a_residues_retained": "aa 1-1406",
  "kmt2a_modules_retained": [
    "MBM 6-25",
    "IBM1 123-134",
    "IBM2 147-152",
    "AT-hooks 169-309",
    "CXXC 1147-1195"
  ],
  "mbm_retained": true,
  "cxxc_retained": true,
  "in_leukaemic_breakpoint_cluster_region": true,
  "partner_residues_predicted": "aa 382-424 - 43 residues, 10% of the protein, the terminal exon only",
  "partner_modules_predicted_retained": [],
  "predicted_frame": "IN FRAME (ELP4 exon 10 starts 1,143 nt after the ATG, 1143 % 3 == 0; KMT2A exon 9 ends at a complete codon) - but in-frame buys little when the contribution is 43 residues",
  "frame_derivation": "computed 2026-09-22 from NM_001197104.2 and NM_019040.4 exon coordinates - a derivation, not an observation",
  "the_point": "ELP4 SUPPLIES ESSENTIALLY NO ELP4 MOIETY. 43 residues cannot assemble into Elongator, cannot bind ELP5/ELP6 and carry no annotated feature. This is the mirror image of the YKY problem this graph already records - YAP1-KMT2A-YAP1 inserts only KMT2A exons 4-5/6 and supplies essentially no MLL1 moiety, which is why the framework is indeterminate for it. Here the DEFICIENT side is the partner.",
  "source": "Supplemental Table 2 of PMID 32461620",
  "fusion_protein_observed": false,
  "reciprocal_transcript_reported": false,
  "functional_assay_done": false,
  "authors_own_verdict": "of unclear significance"
}
localization
{
  "CONSTITUTIVE": "CYTOPLASM (three ECO:0000269s: PubMed 11818576, 11889558, 22854966) and NUCLEUS (one ECO:0000269: PubMed 11889558). RULE 91 FAILS ON THE NUCLEAR CALL: PMID 11889558 describes NO localisation experiment in its abstract - it is the gene-structure and synteny study that maps ten exons and a 134 kb final intron and concludes with a transgenic mouse that aniridia is NOT caused by heterozygous PAXNEB loss. The best support for a nuclear pool is functional and indirect: human Elongator interacts directly with RNA polymerase II and acts on chromatin templates (PMID 11818576).",
  "systems": "HeLa; unspecified human cell lines; mouse embryos; Saccharomyces cerevisiae. No haematopoietic and no mesenchymal or sarcoma cell."
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.3534,
  "chronos_mean_other_leukaemia": -0.3795,
  "difference": 0.0261,
  "pan_essentiality_all_lines": -0.4592,
  "n_all_lines": 1208,
  "selectivity_rank": 12422,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-22 from analyses/depmap-kmt2a-selective-scan-full/results.json (generated 2026-09-19T10:12:06Z), this 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, including that the 11 KMT2A-rearranged lines are drawn from nine individuals and that the arm is myeloid-enriched. The scan records a generation timestamp and NOT a DepMap RELEASE, which is rule 83 exposure carried forward rather than fixed here.",
  "how_to_read": "A rank is not an effect size and a difference is not a dependency (this graph's DepMap rules). Check pan_essentiality_all_lines before reading anything as selective, and note the anti-selective half is 50% of any gene set by construction, with this graph's own base rate at 45.3% - so a tally of genes in that half is the null (rule 84). The decisive limitation here is lineage: the screen is leukaemia lines, and for a sarcoma fusion that is the wrong tissue. Cellosaurus holds 592 KMT2A-commented cell lines and ZERO are sarcoma or any mesenchymal tumour, so no cell-line screen can address these cases at all. ELP4 has the deepest pan-essentiality of the seven partners at -0.459, which is consistent with a core tRNA-modification complex and is exactly the regime where a near-zero arm-to-arm difference is uninterpretable rather than small."
}
queries run 2026 09 22
[
  "ELP4[tiab] -> 78; PAXNEB[tiab] -> 1; C11orf19[tiab] -> 1 (retired-symbol control, adds nothing)",
  "ELP4[tiab] AND fusion[tiab] -> 1 (aniridia case report)",
  "ELP4[tiab] AND translocation[tiab] -> 4 (all aniridia/PAX6 cytogenetics)",
  "ELP4[tiab] AND KMT2A[tiab] -> 0",
  "ELP4[tiab] AND MLL[tiab] -> 0",
  "ELP4[tiab] AND histone[tiab] -> 13",
  "ELP4[tiab] AND acetylation[tiab] -> 6",
  "Elongator[tiab] AND tRNA[tiab] -> 334",
  "Elongator[tiab] AND acetyltransferase[tiab] -> 86",
  "ELP3[tiab] AND histone[tiab] -> 70",
  "ELP4[tiab] AND cancer[tiab] -> 3",
  "controls: EWSR1[tiab] AND fusion[tiab] -> 1208; EWSR1[tiab] AND translocation[tiab] -> 462; AFF1[tiab] AND KMT2A[tiab] -> 104; MLLT3[tiab] AND MLL[tiab] -> 84; KMT2A[tiab] AND histone[tiab] -> 218"
]
graph neighbours already present
[
  "gene-kmt2a",
  "gene-supt6h",
  "gene-paf1",
  "gene-ep300",
  "gene-crebbp",
  "disease-kmt2a-r-sarcoma"
]
How this could bite us back
  • This entry must not be read as a menin-inhibitor rationale. Menin is a TUMOUR SUPPRESSOR in mesenchyme, and the only non-haematologic menin trial with a result, NCT05731947, was terminated by the sponsor after phase 1b for not meeting pre-defined efficacy signals. An MBM-retaining KMT2A fusion in a sarcoma is a test of a hypothesis, not an indication.
  • If menin inhibition were tried here off the gene name alone, the predicted direction of effect is not obviously anti-tumour: displacing menin from chromatin in a mesenchymal cell removes a tumour suppressor from its targets. See risk-a-menin-inhibitor-prescribed-off-a-kmt2a-gene-name-in-a-solid-tumour.
  • Every quantitative statement about the fusion in this entry is a coordinate derivation from transcript records. If the panel's exon numbering differs from RefSeq's, the frame prediction, the retained-residue count and the MBM claim all move together - one systematic error, not several independent ones.
  • Do not read the Elongator HAT literature as a druggable node here. The fusion keeps 43 of ELP4's 424 residues, the HAT assignment belongs to ELP3 and is called putative in the current structural literature, and the complex's best-supported activity is tRNA wobble modification - which is required in most cells (pan-essentiality -0.459 across 1,208 lines).
  • Elongator loss of function in humans causes familial dysautonomia and neurodevelopmental disease, so systemic inhibition of this complex has a known and severe phenotype rather than an unknown one.
  • The tRNA-modification function makes Elongator a general translational-fidelity factor, which means any agent hitting it would have off-target effects distributed across the whole proteome rather than at identifiable targets - the worst shape for a therapeutic window.
Still open
  1. Does a 43-residue ELP4 tail do anything to a KMT2A N-terminal module? The span carries no annotated feature and has never been characterised in isolation, so a degron, a phosphosite or an interaction epitope cannot be ruled out - and if it is inert, case S15 is functionally a KMT2A truncation, which is a different lesion class from a chimera.
  2. Is the ELP4 locus's OTHER tenant affected? Every rearrangement in ELP4's literature is about PAX6 regulation next door, and nobody has asked whether a KMT2A breakpoint in ELP4's terminal exon has a positional effect on PAX6 in an angiosarcoma.
  3. Does Elongator acetylate histones in a human haematopoietic or mesenchymal cell? The H3K14/H4K8 assignment is yeast, the human HAT preparation had three subunits, and the current structural literature calls ELP3's HAT domain putative - so a calibrated ICeChIP or a clean human reconstitution would settle a 24-year-old question this graph now depends on.
  4. Is Elongator anywhere in a KMT2A fusion interactome? Elongator binds RNA polymerase II directly and this graph's whole mechanism is forced elongation, yet ELP4[tiab] AND KMT2A[tiab] returns 0 and no Elongator subunit appears in this repository - exp-turboid-map-of-kmt2a-aff1 would answer it as a by-product.
  5. Does the retained KMT2A 1-1406 module in case S15 bind menin? Exon 9 is inside the leukaemic breakpoint cluster region, so this is one of the three strongest natural tests in the cohort.
Every source behind this entry (11)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    ELP4 (Q96EB1) is a 424-residue subunit of the Elongator complex with no annotated domain, motif or repeat of any kind, whose UniProt FUNCTION comment is tRNA wobble modification (mcm5U, mcm5s2U, ncm5U, carboxymethyluridine at position 34) and is coded ECO:0000303 - an author statement rather than an experiment.

    What this evidence cannot show

    Accession resolved this cycle by gene_exact query with gene_primary read back: exactly one reviewed human record, primary symbol ELP4, entry ELP4_HUMAN, length 424 aa (rules 62, 65, 66; accession AND length asserted). Two things about this record are worth flagging rather than copying. The FUNCTION comment's evidence code is ECO:0000303, which is a statement attributed to an author in a publication and NOT an experimental observation - the distinction rule 91 item 4 asks for. And the feature table is EMPTY: no domain, motif, repeat, zinc finger, coiled coil or DNA-binding feature is annotated anywhere in the 424 residues, so any domain-architecture argument about ELP4 is an argument about a complex, not about a protein.

    Evidence that argues against this (1)
    • An empty feature table is as much a statement about annotation effort as about the protein: there is no experimental structure of human ELP4 alone, and the yeast Elp4/5/6 subcomplex is a RecA-like ATPase-fold hexamer, a fold UniProt does not annotate here. So 'no domain' should be read as 'no annotated domain', not as 'unstructured'. Dalwadi U et al. 2018
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    The human Elongator complex consists of ELP1, ELP2, ELP3, ELP4, ELP5 and ELP6, and ELP4 is a bona fide subunit of the purified human complex.

    Purification and characterization of the human elongator complex. · Hawkes NA, Otero G, Winkler GS et al · Journal of Biological Chemistry 2002
    PMID 11714725DOI
    What this evidence cannot show

    Biochemical purification from human cells; this and PMID 22854966 are the two citations UniProt uses for the SUBUNIT comment and both are ECO:0000269. Abstract-level reading only - the stoichiometry, the subcomplex split and the 850 kDa figure come from the later structural review (PMID 29332244), not from this paper. Note that the composition of Elongator was genuinely unsettled in 2002: a contemporaneous human purification recovered only three polypeptides, which is why the number of subunits is asserted here from the purification paper rather than from the HAT paper.

    Evidence that argues against this (1)
    • Complex composition determined by purification depends on the tag, the salt and the cell type, and the disagreement is on the record: the human HAT-activity purification of the same year describes a complex of THREE polypeptides. A subunit list is a property of a preparation until a structure settles it. Kim JH et al. 2002
  3. E3moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Intact Elongator is a histone H3 and H4 acetyltransferase acting on core and nucleosomal substrates with H3K14 and H4K8 predominant, ELP4, ELP5 and ELP6 are required for that HAT activity, Elongator binds both naked and nucleosomal DNA, and multiply-acetylated H3 and H4 levels fall in vivo in yeast lacking ELP3.

    Elongator is a histone H3 and H4 acetyltransferase important for normal histone acetylation levels in vivo. · Winkler GS, Kristjuhan A, Erdjument-Bromage H et al · Proceedings of the National Academy of Sciences of the United States of America 2002
    PMID 11904415DOI
    What this evidence cannot show

    SACCHAROMYCES CEREVISIAE, and the first clause of the claim should be read that way (rule 73): this is a yeast complex, a yeast in-vivo ChIP and yeast genetics, not a human, haematopoietic or mesenchymal system. The requirement for Elp4/5/6 is a requirement for HAT ACTIVITY OF THE INTACT COMPLEX - the paper's own contrast is between intact Elongator and isolated Elp3, so it does not show that ELP4 contacts histones or contributes catalysis. Abstract read; full text not read. And the assignment itself is now contested: see E5.

    Evidence that argues against this (1)
    • By 2018 a structural review of Elongator describes the Elp3 histone acetyltransferase domain as PUTATIVE, lists alpha-tubulin acetylation and tRNA modification alongside transcription among Elongator's implicated functions, and gives the Elp4/5/6 subcomplex's measured activity as ATP-modulated tRNA binding rather than anything histone-directed. Measured 2026-09-22, the literature has followed: Elongator[tiab] AND tRNA[tiab] = 334 against Elongator[tiab] AND acetyltransferase[tiab] = 86. Dalwadi U et al. 2018
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Human Elongator purified from HeLa cells carries histone acetyltransferase activity specific to H3 and much weaker on H4, interacts directly with RNA polymerase II, and facilitates transcription by RNA polymerase II in a chromatin- and acetyl-CoA-dependent manner.

    Human Elongator facilitates RNA polymerase II transcription through chromatin. · Kim JH, Lane WS, Reinberg D · Proceedings of the National Academy of Sciences of the United States of America 2002
    PMID 11818576DOI
    What this evidence cannot show

    HUMAN HeLa cells and in-vitro chromatin templates - the right species, the wrong cell type for this graph, and an in-vitro reconstitution rather than a cellular measurement. THE LIMITATION THAT MATTERS HERE: this preparation is described as 'composed of three polypeptides', so it does NOT place ELP4 in the HAT-active human species, and the paper cannot be cited for ELP4 having HAT activity. It is cited here for the complex's activity and for the direct Pol II contact, which is what makes Elongator relevant to this graph's elongation mechanism at all. Abstract read; full text not read.

    Evidence that argues against this (1)
    • A three-polypeptide preparation with H3-directed HAT activity is in tension with the yeast result that Elp4, Elp5 and Elp6 are REQUIRED for HAT activity - if the human active species lacks them, the requirement is not general. One of the two results has to be preparation-specific, and the entry cannot say which. Winkler GS et al. 2002
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.review

    Elongator is implicated in transcription regulation, alpha-tubulin acetylation and tRNA modification; its catalytic subunit ELP3 contains a radical-SAM domain and a PUTATIVE histone acetyltransferase domain; and the ELP4/5/6 subcomplex's own measured activity is ATP-modulated tRNA binding.

    Structural insights into the function of Elongator. · Dalwadi U, Yip CK · Cellular and Molecular Life Sciences 2018
    PMID 29332244DOI
    What this evidence cannot show

    A structural review, cited here as a map of where the field stands rather than as primary evidence (the brief's rule on reviews). It is also the paper UniProt cites for ELP4's FUNCTION under ECO:0000303, which is why its framing matters: the database's own tRNA-only annotation for ELP4 traces to this review. Its own closing statement is that how Elongator performs its multiple functions and how the subunits regulate them 'remains poorly understood'. The 334-vs-86 literature counts quoted alongside it were run 2026-09-22 with one [tiab] per term and their querytranslation read back; raw counts on the word 'Elongator' were not individually title-checked and 'elongator' is a reasonably specific token but not a perfectly clean one.

    Evidence that argues against this (1)
    • A shift in publication volume is a statement about what a field finds tractable, not a refutation: the yeast H3K14/H4K8 result and the human H3-directed HAT activity were never retracted, and tRNA modification and histone acetylation are not mutually exclusive functions of an 850 kDa two-lobed complex. 'Putative' in a review is a hedge, not a negative result. Winkler GS et al. 2002
  6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In case S15 of a pan-sarcoma KMT2A-rearrangement series, an angiosarcoma carried KMT2A exon 9 fused to ELP4 exon 10 (NM_019040), one of seven cases with KMT2A 5-prime and one of the three whose breakpoint lies inside the leukaemic breakpoint cluster region.

    PMID 32461620DOI
    What this evidence cannot show

    This is ONE patient (case S15, angiosarcoma) in a commercial targeted RNA-panel series enriched for advanced disease. The source authors call the non-recurrent KMT2A partners of UNCLEAR SIGNIFICANCE and several may be passengers. No fusion protein has been observed for this or any other sarcoma KMT2A fusion by any method, no reciprocal transcript was reported, no functional experiment has been done, and the panel targets specific KMT2A exons, which biases which breakpoints are detectable at all. Supplemental Table 2 reports EXON NUMBERS from the panel's own annotation - the junction nucleotide, the transcript sequence and the reading frame are not given - and it misspells the index gene as KTM2A in two of eighteen rows, so a reader grepping for KMT2A silently loses cases. This graph's parse is at analyses/sarcoma-nonrecurrent-partners/results.json. Naming this case makes it auditable; it does not promote the partner to a driver and it is not a treatment rationale.

    Evidence that argues against this (1)
    • The cohort's recurrent, morphology- and methylation-defined KMT2A sarcoma arms are YAP1-KMT2A-YAP1 and VIM::KMT2A, and those are what the paper's title and conclusions are about. The seven 5-prime non-recurrent cases are explicitly set aside by the authors as of unclear significance, and a targeted panel run in advanced disease will also detect passenger rearrangements. A single MBM-retaining case is a lead, not an entity. Massoth LR et al. 2020
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    ELP4's transcript NM_019040 has ten exons with a CDS of 19..1293, and exon 10 is the terminal exon spanning nt 1162-8093 and coding amino acids 382-424 - so a KMT2A exon 9 :: ELP4 exon 10 fusion contributes only the final 43 of ELP4's 424 residues, 10% of the protein, in a span carrying no annotated feature.

    What this evidence cannot show

    A derivation computed this cycle from the RefSeq record the source itself names, not an observation. The junction is predicted IN FRAME (exon 10 begins 1,143 nt after the ATG, an exact multiple of three, and KMT2A exon 9 ends at a complete codon), but in-frame is a weak fact when the contribution is 43 residues. It can be wrong if the commercial panel's exon numbering differs from RefSeq's or if the breakpoint is intra-exonic; the junction nucleotide is not reported. The independent confirmation that ELP4 has ten exons comes from the gene's founding characterisation (PMID 11889558), which mapped ten exons and a 134 kb final intron.

    Evidence that argues against this (1)
    • A 43-residue tail is not necessarily inert - a short C-terminal extension can carry a degron, a phosphosite or an interaction epitope, and this span has never been characterised in isolation. The defensible claim is that the fusion cannot supply Elongator function, not that the tail does nothing. UniProt Consortium 2026
  8. E8strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    KMT2A exons 1-9 encode aa 1-1406 and therefore retain the menin-binding motif (6-25), LEDGF IBM1 (123-134) and IBM2 (147-152), all three AT-hooks and the CXXC domain (1147-1195); exon 9 lies inside the leukaemic breakpoint cluster region (exons 8-14), so the KMT2A portion of this fusion is not distinguishable from that of an infant-leukaemia KMT2A fusion.

    What this evidence cannot show

    KMT2A exon-to-residue map, re-derived this cycle from RefSeq NM_001197104.2 (CDS 22..11940, 36 exons, NP_001184033.1) and independently reproducing this graph's own ana-kmt2a-exon-to-domain-map: exon 1 = aa 1-144 (MBM 6-25, LEDGF IBM1 123-134), exon 2 = aa 145-168 (IBM2 147-152), exon 3 = aa 168-1052 (all three AT-hooks), exon 4 = 1053-1112, exons 5-6 = 1112-1212 (CXXC 1147-1195). Exon 3 ends at aa 1052, exon 8 at 1362, exon 9 at 1406, exon 15 at 1668, exon 20 at 1888, exon 22 at 1987. All six KMT2A donor exons in this cohort end at a COMPLETE CODON (phase 0), which is why KMT2A can fuse in frame to so many acceptors. Note NM_001197104.2 encodes 3,972 aa while UniProt Q03164 is 3,969 aa; the three-residue difference lies downstream of every coordinate used here. This is a COORDINATE DERIVATION from a transcript record, not a measurement of a fusion protein: no sarcoma KMT2A fusion protein has ever been observed by any method and no menin-binding assay has been run on one.

    Evidence that argues against this (1)
    • Retaining a motif is not the same as presenting it. The chimera's folding, stability and nuclear import are all unmeasured, and a fusion protein can retain a binding motif in primary sequence while failing to engage its partner. This graph's queued test exp-menin-pulldown-on-yky-versus-kmt2a-aff1 exists precisely because retention has never been shown to imply binding for any sarcoma arm. Massoth LR et al. 2020
  9. E9strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    ELP4 was discovered as PAXNEB (C11orf19), a ten-exon gene immediately telomeric to PAX6 with an unusually large 134 kb final intron that is disrupted by every aniridia-associated chromosomal rearrangement examined; a transgenic mouse model showed the aniridia phenotype of those rearrangement cases is NOT due to heterozygous loss of PAXNEB function, and the 2 kb transcript is expressed in all cell lines tested.

    PMID 11889558DOI
    What this evidence cannot show

    Human genomics plus a transgenic mouse rescue, 2002. This is the paper UniProt cites with ECO:0000269 for ELP4's NUCLEUS localisation comment, and RULE 91 FAILS HERE: the abstract describes no localisation experiment of any kind - it is a gene-structure, synteny and mouse-transgenic study. The same citation is used for 'Tissue specificity: widely expressed', which it does support. Abstract read; full text not read, so a localisation panel in a figure cannot be excluded (rule 80) - but the annotation as filed points at a paper that is about PAX6. Retired-symbol control run 2026-09-22 per the brief's rule: PAXNEB[tiab] = 1 and C11orf19[tiab] = 1, both this same record, against ELP4[tiab] = 78 - so unlike the SEPTIN case the retired symbols add nothing, and the check is recorded as a negative rather than omitted.

    Evidence that argues against this (1)
    • UniProt nonetheless files ANIRIDIA 2 as a DISEASE of ELP4, citing PMID 24290376 - a paper about a de novo point mutation in SIMO, an ultraconserved cis-element 150 kb downstream of PAX6 that disrupts a PAX6 AUTOREGULATORY binding site and abolishes enhancer activity. That is a PAX6 regulatory lesion that happens to lie inside the ELP4 locus. So the record carries a disease annotation whose cited evidence is positional, next to a citation whose explicit conclusion is that PAXNEB loss does NOT cause the phenotype - two co-cited papers on one record pointing opposite ways. Bhatia S et al. 2013
  10. E10moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    Elongator subunits are required for tumour cell behaviour in at least one lineage: knockdown of ELP5 (DERP6) or ELP6 (C3ORF75) reduces the tumorigenicity and migration of melanoma cells, acting as Elongator subunits.

    DERP6 (ELP5) and C3ORF75 (ELP6) regulate tumorigenicity and migration of melanoma cells as subunits of Elongator. · Close P, Gillard M, Ladang A et al · Journal of Biological Chemistry 2012
    PMID 22854966DOI
    What this evidence cannot show

    Human MELANOMA cells - not sarcoma, not haematopoietic - and the subunits knocked down are ELP5 and ELP6, not ELP4. It is cited here as the nearest cell-state result for the ELP4/5/6 lobe rather than as evidence about ELP4 itself, and ELP4 is the third member of that lobe. Abstract read; full text not read. No ELP4 loss-of-function experiment in any tumour type was found.

    Evidence that argues against this (1)
    • A shared-lobe argument is an inference, not a measurement: ELP4, ELP5 and ELP6 are required together for Elongator's activity in yeast but that does not make their loss-of-function phenotypes interchangeable in a human tumour. And Elongator's disease literature is dominated by familial dysautonomia and neurodevelopmental phenotypes rather than cancer, which is where a reader should expect its loss-of-function effects to be best characterised. Gaik M et al. 2023
  11. E11weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset

    ELP4 is the most broadly required of the seven partners in this project's own DepMap scan and is still not KMT2A-selective: pan-essentiality -0.459 across 1,208 lines, with mean Chronos -0.353 in 11 KMT2A-rearranged lines against -0.380 in 43 other leukaemia lines, difference +0.026 and selectivity rank 12,422 of 17,728.

    DepMap CRISPR (Chronos) gene-effect scan of 11 KMT2A-rearranged versus 43 other leukaemia lines (this project's own scan) · Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full · analyses/depmap-kmt2a-selective-scan-full 2026
    What this evidence cannot show

    Joined 2026-09-22 from analyses/depmap-kmt2a-selective-scan-full/results.json (generated 2026-09-19T10:12:06Z), this 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, including that the 11 KMT2A-rearranged lines are drawn from nine individuals and that the arm is myeloid-enriched. The scan records a generation timestamp and NOT a DepMap RELEASE, which is rule 83 exposure carried forward rather than fixed here. A rank is not an effect size and a difference is not a dependency (this graph's DepMap rules). Check pan_essentiality_all_lines before reading anything as selective, and note the anti-selective half is 50% of any gene set by construction, with this graph's own base rate at 45.3% - so a tally of genes in that half is the null (rule 84). The decisive limitation here is lineage: the screen is leukaemia lines, and for a sarcoma fusion that is the wrong tissue. Cellosaurus holds 592 KMT2A-commented cell lines and ZERO are sarcoma or any mesenchymal tumour, so no cell-line screen can address these cases at all. A pan-essentiality of -0.459 is a genuine shared requirement - consistent with a core tRNA-modification complex - and that is exactly the regime where a small arm-to-arm difference is uninterpretable rather than small.

    Evidence that argues against this (1)
    • A gene required across most lines has little dynamic range in which selectivity could appear, so the near-zero difference is the expected reading and carries no information (the floor-compression argument this graph applies to HCFC1 and KAT8). And the screen targets the wild-type ELP4 locus, which in this fusion retains its own promoter and all nine other exons. Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full 2026
Connected to