Oncofusion

start / graph / gene-jazf1

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q126-the-partners-named-but-never-written

JAZF1 - a 243-residue zinc-finger protein that turned out to be a NuA4 subunit, and the 5' donor of the commonest endometrial-stromal-sarcoma fusion

also called JAZF zinc finger 1 · TIP27 · TAK1-interacting protein 27 · ZNF802 · JJAZ1 partner (in the JAZF1-JJAZ1 nomenclature of the founding paper, where JJAZ1 = SUZ12)

JAZF1 is a small zinc-finger protein that sits inside a machine which switches genes on by acetylating histones. When it is accidentally joined to a component of the opposite machine - the one that switches genes off - the result is the commonest fusion in a rare uterine cancer.

The same thing, for a scientist

**JAZF1 is UniProt Q86VZ6 (JAZF1_HUMAN, 243 aa), HGNC:28917 `JAZF zinc finger 1` at 7p15.2-p15.1, primary Ensembl ENSG00000153814; `gene_exact:JAZF1 AND organism_id:9606 AND reviewed:true` returned exactly one row** [E1][E2]. At 243 residues it is the smallest of the ESS 5' donors, and its only UniProt-coded localisation is 'Nucleus', from PMID 15302918. **It was discovered twice, in two literatures that barely reference each other.** In 2001 it was named from the 7p15 breakpoint of the recurrent t(7;17)(p15;q21) of endometrial stromal tumours, as the 5' half of a `JAZF1/JJAZ1` fusion - `JJAZ1` being the name then used for **SUZ12** - present in all types of endometrial stromal tumour but rarer in tumours considered high grade [E3]. In 2004, independently, it was characterised as **TIP27**, a 27-kDa two-zinc-finger repressor of the nuclear orphan receptor TAK1/TR4, interacting through residues Asp39-Lys79 and co-localising with TAK1 in the nucleus [E4]. **Its complex membership was only established eighteen years after the fusion.** Piunti et al., working on ESS fusion biochemistry, **identified JAZF1 as a new subunit of the NuA4 acetyltransferase complex** [E5]. That assignment is what makes `concept-low-grade-ess-is-a-nua4-to-polycomb-fusion-family` sayable at all - and, as that entry already flags, the rule and this instance were established in the same paper. **Dependency.** Over 1,258 DepMap screens JAZF1's mean Chronos gene effect is **-0.3954 with 27.7% of lines below -0.5 and 0.6% below -1.0** - a middling, EPC1-like profile, not a dependency [E6].

Go deeper
Two names, two literatures, and a retrieval trap the graph should record

JAZF1 and TIP27 are the same protein. The fusion literature calls it JAZF1 and never TIP27; the nuclear-receptor literature calls it TIP27 and the 2004 paper's title does not contain the string 'JAZF1' at all 4. ZNF802 is a third listed alias 2.

And the founding fusion paper names the 3' partner something nobody uses now. It is JAZF1/JJAZ1 3; JJAZ1 is SUZ12. A query for JAZF1[tiab] AND SUZ12[tiab] therefore cannot reach the paper that discovered the fusion, and a query for JAZF1[tiab] AND JJAZ1[tiab] cannot reach the modern literature. This is rule 69's shape - a superseded name is where the founding observation lives - applied to a fusion partner rather than to a gene under study. Anyone bounding an absence about JAZF1 fusions must run JJAZ1 and TIP27 as well; the corpus's data/symbol-collisions.json has no row for any of the three today.

One thing this entry deliberately does not claim. JAZF1 is a well-known type 2 diabetes GWAS locus. This beat did not read that literature and takes no position on whether it is relevant here; it is recorded as a bounded rejection so that a later cycle can tell a considered omission from an unnoticed one.

What the NuA4 assignment is and is not

Piunti et al. state the assignment plainly - JAZF1 is 'a new subunit of the NuA4 acetyltransferase complex' - and reach it by biochemical characterisation of the ESS fusion proteins 5. Two limits, both of which concept-low-grade-ess-is-a-nua4-to-polycomb-fusion-family already carries and which are repeated here so the gene entry does not read as firmer than the concept:

  1. The rule and the instance came from the same paper. The family rule 'N-terminal NuA4 component, C-terminal PcG subunit' and the JAZF1-is-NuA4 assignment were published together, so JAZF1 cannot serve as independent confirmation of the rule.
  2. The paper's own experiments concern JAZF1 and CXORF67/EZHIP. EPC1-PHF1 and MEAF6-PHF1 appear in its enumeration; no experiment on either is reported there.

What the gene entry adds that the concept entry does not have is the accession-level identity, the TIP27/JJAZ1 retrieval trap, the pre-fusion nuclear-receptor function, and the dependency profile - and one structural observation: at 243 residues JAZF1 is short enough that a fusion retaining its N-terminal portion retains most of it. The graph should not assume a JAZF1 fusion delivers a cleanly separable 'NuA4-docking module'; the breakpoint geometry in the individual fusions was not resolved in this beat and is not asserted.

Facts and figures
uniprotQ86VZ6
length AA243
gene primaryJAZF1
hgncHGNC:28917
hgnc nameJAZF zinc finger 1
ensemblENSG00000153814
cytoband7p15.2-p15.1
accession resolutiongene_exact:JAZF1 AND organism_id:9606 AND reviewed:true returned exactly one row, Q86VZ6, primary gene name JAZF1; HGNC:28917 independently carries Q86VZ6 and ENSG00000153814. Both fetched live 2026-09-23. Rule 62 does not fire.
complex membershipNuA4 acetyltransferase complex - a NEW subunit assignment made in 2019 (PMID 31281901), eighteen years after the fusion was described. Not annotated as a NuA4 subunit in the UniProt location block.
subcellular localizationnucleus
subnuclear localizationNOT ANNOTATED anywhere this beat looked. UniProt Q86VZ6 carries a bare 'Nucleus'; no nuclear body, no chromatin call, no focus. For a claimed NuA4 subunit that is a gap.
retrieval trapTHREE names for one protein and a fourth for its commonest partner. JAZF1 = TIP27 = ZNF802; and the founding fusion paper calls the 3' partner JJAZ1, which is SUZ12. A query on JAZF1[tiab] AND SUZ12[tiab] cannot reach PMID 11371647; a query on JAZF1[tiab] AND JJAZ1[tiab] cannot reach the modern literature. Neither JAZF1, TIP27, ZNF802 nor JJAZ1 has a row in data/symbol-collisions.json as of 2026-09-23. SEPARATELY: the TAK1 in PMID 15302918 is the nuclear receptor TR4/NR2C2, not the MAP3K7 kinase, and that is a genuine collision with no row either.
relationship to existing entriesconcept-low-grade-ess-is-a-nua4-to-polycomb-fusion-family (Q123) holds the family rule and names JAZF1 in its table; gene-epc1 (Q123) holds the other NuA4 5' donor. This entry holds JAZF1 as a gene: accession, the TIP27/JJAZ1 retrieval trap, the pre-fusion TR4-repressor function that predates the NuA4 assignment, the founding fusion description, and the dependency profile.
coverage check verdicts pasted not paraphrasedBEFORE writing, scripts/beat-inventory.py --symbols ... JAZF1 ... returned the row `| `JAZF1` | **absent** | 11 |` (absent from the id set, named in 11 node files by word-boundary match), 2026-09-23. AFTER writing, `python scripts/is-symbol-covered.py JAZF1` returns: JAZF1 COVERED `gene-jazf1` (matched via the symbol itself) Recorded in both directions deliberately: the second run is the control showing the instrument sees this entry, which is the check the cycle-86 order-dependence bug would have failed.
alias symbol
[
  "TIP27",
  "DKFZp761K2222",
  "ZNF802"
]
prev symbol
[]
fusions donating this gene as the 5 prime partner
[
  "JAZF1-SUZ12 (JAZF1-JJAZ1), t(7;17)(p15;q21), about half of low-grade ESS",
  "JAZF1-PHF1, t(6;7)"
]
essentiality
{
  "depmap_source": "Open Targets GraphQL target.depMapEssentiality, fetched 2026-09-23",
  "depmap_release": null,
  "open_targets_dataVersion": "year 26, month 06, iteration null",
  "n_screens": 1258,
  "chronos_mean": -0.3954,
  "chronos_median": -0.3907,
  "frac_below_minus_0_5": 0.2774,
  "frac_below_minus_1_0": 0.0064,
  "leukaemia_strata": {
    "Acute Myeloid Leukemia": {
      "n": 35,
      "mean": -0.3448
    },
    "B-Cell Acute Lymphoblastic Leukemia": {
      "n": 13,
      "mean": -0.2759
    },
    "T-Lymphoblastic Leukemia/Lymphoma": {
      "n": 8,
      "mean": -0.4059
    }
  }
}
bounded rejections
[
  "JAZF1 is a well-known type 2 diabetes GWAS locus. This beat did NOT read that literature and takes no position on its relevance; recorded so a later cycle can distinguish a considered omission from an unnoticed one.",
  "The breakpoint geometry of JAZF1-SUZ12 and JAZF1-PHF1 - which residues of the 243 are retained - was NOT resolved in this beat and is not asserted anywhere in this entry."
]
How this could bite us back
  • Nothing in this entry supports an intervention, and the graph should not let the NuA4 framing suggest one. There is no chemical matter against JAZF1, no structure, and no evidence that the JAZF1 moiety of either fusion is a druggable surface rather than a short disordered docking region.
  • If anyone reasons from the family rule toward acetyltransferase inhibition, the target would be NuA4/TIP60 itself, whose catalytic subunit KAT5 is far closer to pan-essential than JAZF1 is. The dispensability of the fusion's own 5' partner is not evidence that the complex it docks into is dispensable, and this is the same trap as reading NSD1's -0.1231 as though it licensed anything about NUP98::NSD1.
  • The whole family rule rests on a small number of tumours - the founding JAZF1-PHF1 and EPC1-PHF1 description is a three-case report - so any therapeutic inference built on membership is built on an enumeration that could still change.
Still open
  1. Does JAZF1's TR4/NR2C2-repressor function survive in the fusion, and does anyone care? The TAK1-interaction domain is Asp39-Lys79 - very close to the N-terminus, so almost certainly retained in a 5' JAZF1 fusion. If it is, JAZF1-SUZ12 carries a nuclear-receptor-repressor module into a PRC2 context and nobody has said so. This beat found no paper that connects the 2004 TIP27 work to the fusion literature at all.
  2. Which residues of JAZF1 does each fusion retain? At 243 aa the whole protein is short enough that 'the NuA4-docking module' may not be separable from the rest of it. Without breakpoint geometry the family rule's mechanical claim - that the 5' half donates a complex address - is an assertion about an unmapped region.
  3. Is JAZF1 a stoichiometric NuA4 subunit or a sub-stoichiometric associate? 'A new subunit of the NuA4 acetyltransferase complex' is the authors' phrasing (PMID 31281901); this beat did not establish whether the assignment rests on stoichiometric recovery or on detection. The distinction decides whether removing JAZF1 should perturb NuA4 at all.
  4. Does any haematopoietic malignancy carry a JAZF1 fusion? The gene is here only because the ESS family is the clearest published complex-defined fusion family. If JAZF1 never appears in leukaemia, that is a clean negative worth stating; this beat did not run the query and does not state it.
Every source behind this entry (7)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    JAZF1 is UniProt Q86VZ6 (JAZF1_HUMAN), a 243-residue protein named 'Juxtaposed with another zinc finger protein 1' with gene-name synonyms TIP27 and ZNF802, and a subcellular-location block reading 'Nucleus {ECO:0000269|PubMed:15302918}'.

    What this evidence cannot show

    Fetched live 2026-09-23 via gene_exact:JAZF1 AND organism_id:9606 AND reviewed:true, exactly ONE row. RULE 91 CHECK RUN: the single ECO:0000269 citation, PMID 15302918, IS a genuine JAZF1/TIP27 paper and its abstract does describe confocal microscopy showing nuclear co-localisation with TAK1 - so unlike NUP98 and MEAF6 in this same tranche, UniProt's pointer here is on-target. There is NO subnuclear annotation of any kind, and no NuA4 subunit annotation in the location block.

    Evidence that argues against this (1)
    • UniProt's record for JAZF1 is behind the literature and should not be used to bound its compartments or its complexes. A single 2004 localisation citation and a bare 'Nucleus' is what the record carried before the 2019 NuA4 assignment, and the record still carries no NuA4 subunit annotation. Absence of a complex annotation here is a curation lag, not evidence, and the same is true of the missing subnuclear detail.
  2. E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    HGNC:28917 assigns the symbol JAZF1 the name 'JAZF zinc finger 1' at 7p15.2-p15.1, with primary Ensembl gene id ENSG00000153814, UniProt Q86VZ6, alias symbols TIP27, DKFZp761K2222 and ZNF802, and no previous symbol.

    HGNC record for JAZF1 (HGNC:28917) · HUGO Gene Nomenclature Committee · HGNC REST 2026
    What this evidence cannot show

    Fetched live 2026-09-23; exactly one document returned. Note that JJAZ1 - the name the founding fusion paper uses for the 3' partner - is NOT an alias of JAZF1 and must not be confused with it; JJAZ1 is SUZ12.

    Evidence that argues against this (1)
    • HGNC reporting 'no previous symbol' gives a false sense that JAZF1's nomenclature is stable. The instability is on the PARTNER side - JJAZ1 for SUZ12 - and a nomenclature check run on JAZF1 alone will return clean and still leave a searcher unable to reach the founding paper. A gene-level symbol check cannot detect a fusion-level naming problem, which is a general limit of the checks this project runs before writing an entry.
  3. E3established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.paper

    Two zinc-finger genes, JAZF1 and JJAZ1, lie at the 7p15 and 17q21 breakpoints of the recurrent t(7;17)(p15;q21) of endometrial stromal tumours; tumour-RNA analysis shows a JAZF1/JJAZ1 fusion present in all types of endometrial stromal tumour, though the fusion appears rarer among endometrial stromal sarcomas that would be considered high grade under certain classification schemes.

    Frequent fusion of the JAZF1 and JJAZ1 genes in endometrial stromal tumors · Koontz JI, Soreng AL, Nucci M, Kuo FC, et al. · Proc Natl Acad Sci U S A 2001
    PMID 11371647DOI
    What this evidence cannot show

    Abstract read this session; full text not read, and the author list is the four names in the fetched efetch record plus 'et al.'. This is a 2001 paper using the nomenclature of its time: JJAZ1 is the gene now called SUZ12, and the paper contains no PRC2 or NuA4 framing because neither assignment existed. The high-grade/low-grade statement is explicitly conditioned by the authors on 'certain classification schemes', and endometrial stromal tumour classification has since been revised - concept-low-grade-ess-is-a-nua4-to-polycomb-fusion-family already carries that caveat and it applies here.

    Evidence that argues against this (1)
    • The finding that most undercuts a simple oncogenic reading is in the abstract and is easy to skim past: the JAZF1/JJAZ1 fusion is present in ALL types of endometrial stromal tumour, including benign stromal nodules. A fusion found in benign lesions is not sufficient for malignancy, so JAZF1-SUZ12 cannot be the whole transforming event, and any account of the ESS family that treats the fusion as the disease is incomplete. The authors themselves reach for a progression model rather than a driver model.
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    TIP27 is a 27-kDa protein containing two zinc-finger motifs, identified by yeast two-hybrid against the ligand-binding domain of the nuclear orphan receptor TAK1/TR4; the region between Asp39 and Lys79 (the TAK1-interaction domain) is critical for the interaction, confocal microscopy shows TIP27 and TAK1 co-localise in the nucleus, and TIP27 acts as a strong and TAK1-selective repressor of DR1-dependent transcriptional activation without inhibiting TAK1 homodimerisation or DR1 binding.

    TIP27: a novel repressor of the nuclear orphan receptor TAK1/TR4 · Nakajima T, Fujino S, Nakanishi G, Kim YS, Jetten AM · Nucleic Acids Res 2004
    PMID 15302918DOI
    What this evidence cannot show

    Abstract read this session; full text not read. Confidence is moderate because the work is yeast two-hybrid plus mammalian two-hybrid plus pull-down plus overexpression confocal microscopy - a 2004 interaction-biology package with no endogenous-protein arm - and because the repression mechanism is left as a suggestion ('may be due to an effect on co-activator recruitment'). Note TAK1 here is the nuclear receptor TR4/NR2C2, NOT the MAP3K7 kinase also universally called TAK1. That is a live symbol collision and it has no row in data/symbol-collisions.json today.

    Evidence that argues against this (1)
    • The TIP27 work is a 2004 interaction-biology package - yeast two-hybrid, mammalian two-hybrid, pull-down, overexpression confocal - with no endogenous-protein arm, and this beat found no replication of the TR4 repression at physiological levels. It is also in tension with the later assignment: a dedicated selective repressor of one nuclear receptor and a constitutive subunit of a general acetyltransferase complex are different jobs, and nothing reconciles them. Either could be the minor activity.
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Using biochemical characterisation of fusion proteins associated with endometrial stromal sarcoma, Piunti et al. identified JAZF1 as a new subunit of the NuA4 acetyltransferase complex, and CXORF67/EZHIP as a subunit of PRC2 whose interaction decreases PRC2-dependent H3K27me2/3 deposition.

    PMID 31281901DOI
    What this evidence cannot show

    Abstract read this session; the FULL TEXT of this paper was read by Q123 in cycle 83 and its conclusions are already recorded in concept-low-grade-ess-is-a-nua4-to-polycomb-fusion-family, which was read before this entry was written (rule 82). The limits that entry records apply here: the family rule and the JAZF1-is-NuA4 assignment come from the SAME paper, so JAZF1 cannot independently confirm the rule; the paper's own experiments concern JAZF1 and CXORF67 and its fusion work is ectopic HA-FLAG expression in HEK293T and HCT-116.

    Evidence that argues against this (1)
    • The secondary literature does not carry this assignment, which is a real hazard for anyone checking it quickly. concept-low-grade-ess-is-a-nua4-to-polycomb-fusion-family records that a 2026 review of MBTD1 states it 'is associated with PRC1', which is wrong against three primary papers - and if a reviewer can get a NuA4 subunit's complex membership backwards in print, a reader taking the family rule from a review will get a Polycomb-to-Polycomb story with different predicted mechanics. The assignment survives only at the primary-literature level.
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    JAZF1's DepMap profile over 1,258 screens is a mean Chronos gene effect of -0.3954, median -0.3907, with 27.7% of lines below -0.5 and 0.64% below -1.0.

    What this evidence cannot show

    Fetched live 2026-09-23; persisted in analyses/q126-the-partners-named-but-never-written/depmap-essentiality.json (rule 90). RULE 83: meta.dataVersion {year 26, month 06, iteration null} is the OPEN TARGETS platform release, not a DepMap release. Leukaemia strata, as counts not contrasts: AML n=35 mean -0.3448, B-ALL n=13 mean -0.2759, T-ALL n=8 mean -0.4059. No endometrial stromal sarcoma line is known to this beat to be in the panel, so this figure is about JAZF1 knockout in cancer lines generally and says nothing about the fusion. Per rule 84, no 'above the median' claim is made and none should be derived from this.

    Evidence that argues against this (1)
    • A whole-panel mean over 1,258 lines with no endometrial stromal sarcoma line in it is not a statement about JAZF1 in the disease, and the resemblance to EPC1's profile is not evidence of anything: both are small accessory subunits of the same complex with plausible buffering, so similar middling scores are what redundancy produces as much as what shared partial dependency produces. Nothing here distinguishes the two readings.
  7. E7strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    PHF1 at 6p21 is recurrently rearranged in endometrial stromal sarcomas lacking the t(7;17); in two tumours with t(6;7) PHF1 is recombined with JAZF1 from 7p15 to form a JAZF1/PHF1 fusion, and in a third with t(6p;10q;10p) PHF1 is fused to EPC1 from 10p11.

    PMID 16397222DOI
    What this evidence cannot show

    Abstract read this session; full text not read. THREE TUMOURS, and the record is typed by PubMed as 'Case Reports; Journal Article'. That is the founding description of both JAZF1-PHF1 and EPC1-PHF1 and it is a three-case series, which is the right size to keep in mind whenever the ESS family is discussed as though its membership were well enumerated.

    Evidence that argues against this (1)
    • The 2006 paper frames JAZF1-PHF1 and EPC1-PHF1 as endometrial stromal sarcoma fusions, and that framing did not survive: both, and MEAF6-PHF1, were subsequently found in ossifying fibromyxoid tumour. Three tumours in a Case-Reports-typed paper is a thin basis for a tissue assignment, and this is a worked instance of a founding description's disease scope being narrower than the entity. Antonescu CR et al. 2014
Connected to