Oncofusion

start / graph / gene-jade3

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q84-pan-fusion-core

JADE3 / PHF16 (Q92613) - the HBO1-complex scaffold with the strongest pan-fusion evidence in this beat, and an independent published contact with MLL

also called JADE3 · PHF16 · KIAA0215 · Jade family PHD finger protein 3 · Protein Jade-3

JADE3 is one of the scaffolding subunits that tells the HBO1 acetyltransferase which piece of chromatin to act on. It co-purifies with all seven MLL fusions - more strongly than almost anything else in this set - and a separate laboratory has independently shown that MLL proteins contact the HBO1 complex partly through this very subunit.

The same thing, for a scientist

JADE3 (Q92613, 823 aa; also PHF16) is a paralogue of the JADE scaffolding subunits of the HBO1/KAT7 histone acetyltransferase complex, whose PZP (PHD1-zinc-knuckle-PHD2) domain engages the nucleosome through histone H3 and DNA and directs the complex's activity toward H4 acetylation when H3K4 is unmethylated. In Skucha 2018 it co-purified with all seven MLL fusions and is, on the combined per-bait evidence, the **strongest** protein in this beat: MLL-ENL 2.11e-12 (43|41), MLL-GAS7 9.52e-10 (37|42), MLL-AF9 7.76e-9 (40|31), MLL-EEN 2.88e-9 (39|43), MLL-AF1p 8.36e-6 (17|9), MLL-CBP 2.9e-5 (23|12) and MLL-AF4 (= KMT2A::AFF1) 0.00529 (5|5), the only one of the twelve whose index-fusion p-value is below 0.01 apart from SF3A3 and MCM2. Crucially, it is also the one protein in this beat with an INDEPENDENT published contact with MLL: leukemic MLL fusions associate with the HBO1 complex through their trithorax homology domain 2, and that association is mediated mainly by the ING4/5 and PHF16 subunits.

Go deeper
The Skucha row - the strongest in this beat

7 of 7 fusions. MLL-AF1p 8.36e-6 (17|9), MLL-AF4 0.00529 (5|5), MLL-AF9 7.76e-9 (40|31), MLL-CBP 2.9e-5 (23|12), MLL-EEN 2.88e-9 (39|43), MLL-ENL 2.11e-12 (43|41), MLL-GAS7 9.52e-10 (37|42); negative control 0|0 1. Four baits below 1e-8 with 37-43 spectra per replicate. The index-fusion call, 5|5 spectra at p = 0.00529, is modest in absolute terms but is the third strongest MLL-AF4 value in this set.

The independent contact, which is the reason this entry is not weak

A separate group showed that leukemic MLL fusion proteins associate with the HBO1 histone acetyltransferase complex through their trithorax homology domain 2 (THD2) in multiple human cell lines, that the association is chromatin-context-dependent (requiring H3K4me3) and is mediated mainly by the ING4/5 and PHF16 subunits, and that MLL-ELL in particular depends on this association for leukemic transformation; the HBO1 complex promotes loading of the AF4/ENL/P-TEFb (AEP) complex onto target promoters 2. PHF16 is JADE3. That is an independent laboratory, an independent method and an independent conclusion converging on the same subunit-level contact that the Skucha table reports as a co-purification - the only such convergence among the twelve proteins in this beat. Note carefully what it is not: the MLL fusions tested there are led by MLL-ELL, and AFF1 (AF4) appears as part of the recruited AEP complex rather than as the fusion partner.

How the JADE subunit works

The PZP domain of JADE engages the nucleosome by binding histone H3 and DNA, and is necessary for the HBO1 complex to associate with chromatin targets; recognition of unmethylated H3K4 directs the complex's activity toward H4 acetylation, whereas H3K4 hypermethylation changes substrate selectivity. The same work shows the PZP domain augments the transforming activity of the NUP98-JADE2 fusion 3. A follow-up paper resolves paralogue specificity: only JADE1's PZP has strong intrinsic DNA binding and engages nearly the whole H3 tail, while BRPF1's is a thousand-fold weaker, so JADE and BRPF scaffolds are not interchangeable 4. That matters here because the MLL contact is attributed to PHF16/JADE3 specifically, not to the JADE family as a whole.

Localization: a genuine blank

UniProt Q92613 carries no subcellular location comment at all - not an uncoded one, none. Fetched 2026-09-21. This is the only protein of the twelve for which the database offers nothing, and no primary paper found in this beat assigns JADE3 to a subnuclear compartment. Recorded as a bounded absence rather than inferred from the HBO1 complex's chromatin association, which would be an inference about the complex and not a measurement of this protein.

Facts and figures
uniprotQ92613
length AA823
accession resolutionResolved live 2026-09-21: gene_exact:JADE3 AND organism_id:9606 AND reviewed:true returns exactly one row, gene_primary JADE3, Q92613, JADE3_HUMAN, 823 aa, synonyms KIAA0215 and PHF16. Matches Skucha's table. The synonym PHF16 is load-bearing here - the independent MLL contact is published under PHF16 and is unreachable from the string JADE3.
localizationNot recorded by UniProt.
subnuclear localizationNot recorded. BOUNDED ABSENCE, not an inference.
localization evidenceUniProt Q92613 carries NO subcellular location comment of any kind - neither experimental (ECO:0000269) nor by-similarity - as fetched on 2026-09-21. The only protein of the twelve in this beat with a completely empty location field. Its complex partner HBO1/KAT7 is nuclear and chromatin-associated, but that is a statement about the complex; inferring JADE3's localisation from it would be the fallback-constructs-the-answer failure this project has already documented.
shown for fusionKMT2A::AFF1 (MLL-AF4) by co-purification in Jurkat at p = 0.00529, plus all six other Skucha baits, four of them below 1e-8. Independently, PHF16/JADE3 is named as one of the subunits mediating the MLL-HBO1 complex contact, with the transformation arm shown for MLL-ELL - so: co-purification for the index fusion, an independent subunit-level contact for MLL proteins generally, and a functional requirement demonstrated only for MLL-ELL.
graph contextJADE3 is named in eight node files in this graph, all of them DepMap-rank or expression tables in the acetyltransferase and HBO1/KAT7 beats (e.g. a DepMap rank of 11,635 and an expression figure of 4.5 nTPM are recorded there); none is a gene entry and none connects it to the fusion interactome. gene-kat7 (HBO1) has an entry. Numbers quoted in those entries are theirs, not re-measured here.
skucha 2018
{
  "bait": "MLL-AF4 (= KMT2A::AFF1)",
  "p_value": 0.00529,
  "spectral_counts": "5|5",
  "n_fusions_of_seven": 7,
  "negative_control_counts": "0|0",
  "per_bait": {
    "MLL-AF1p": "8.36e-06 (17|9)",
    "MLL-AF4": "0.00529 (5|5)",
    "MLL-AF9": "7.76e-09 (40|31)",
    "MLL-CBP": "2.9e-05 (23|12)",
    "MLL-EEN": "2.88e-09 (39|43)",
    "MLL-ENL": "2.11e-12 (43|41)",
    "MLL-GAS7": "9.52e-10 (37|42)"
  },
  "note": "Every per-bait p-value is below 0.01 - the only protein in this beat for which that is true.",
  "source": "analyses/skucha-mll-fusion-interactome/results.json, re-derived from the cached MOESM4 xlsx on 2026-09-21"
}
cell types and states
[
  "Jurkat (human T-ALL) - AP-MS host, nuclear lysate",
  "multiple human cell lines - MLL/HBO1 association via THD2 and the ING4/5 and PHF16 subunits",
  "murine haematopoietic progenitors - MLL-ELL transformation dependent on the HBO1 association",
  "in vitro nucleosome binding - PZP domain specificity",
  "no KMT2A::AFF1 cell of any kind"
]
bounded literature
{
  "query": "JADE3[tiab] AND (KMT2A[tiab] OR MLL[tiab] OR \"MLL-AF4\"[tiab] OR \"MLL-AF9\"[tiab])",
  "querytranslation": "\"JADE3\"[Title/Abstract] AND (\"KMT2A\"[Title/Abstract] OR \"MLL\"[Title/Abstract] OR \"MLL-AF4\"[Title/Abstract] OR \"MLL-AF9\"[Title/Abstract])",
  "count": 0,
  "alias_query": "PHF16[tiab] AND (KMT2A[tiab] OR MLL[tiab]) returns 1 - querytranslation \"PHF16\"[Title/Abstract] AND (\"KMT2A\"[Title/Abstract] OR \"MLL\"[Title/Abstract]) - and that single record is PMID 34431785, the HBO1-MLL paper. THE CURRENT SYMBOL RETURNS ZERO AND THE RETIRED-STYLE SYNONYM RETURNS THE ONLY RELEVANT PAPER IN THE LITERATURE. This is the standing alias rule biting on a real case: a beat that searched only JADE3 would have written 'no independent evidence' and been wrong.",
  "names_tried": [
    "JADE3",
    "PHF16"
  ],
  "date": "2026-09-21",
  "tool": "NCBI esearch db=pubmed",
  "positive_control": "the identical query shape with MEN1 returns 85",
  "bound": "Titles and abstracts only; supporting figures not searched."
}
functionally screened in the source paper
{
  "_added": "king, cycle 48, 2026-09-21, after Q86 flagged that cycle-47 entries understate the record. Verified by me directly from the supplementary file, not from the paper's prose.",
  "screened": true,
  "library_symbol": "PHF16  (matched via alias; the entry is filed under the current HGNC symbol JADE3)",
  "shrnas_for_this_gene": 6,
  "design": "Supplementary Table 6 / MOESM5 is 768 data rows, 128 unique gene symbols, EXACTLY 6 shRNAs each -- counted with openpyxl on 2026-09-21. Negative-selection RNAi screen in MLL-AF9-expressing MOLM-13.",
  "what_this_DOES_establish": "Every one of the 128 conserved MLL-fusion interactors, this protein included, was functionally tested against an MLL fusion. An entry saying 'nothing in the paper validates this prey' is true of BINDING and understates the record.",
  "what_this_does_NOT_establish": "A per-gene result. Scores are presented as heatmaps (Fig. 2c/2d) and no per-gene depletion value is tabulated, so for any individual gene the outcome is measured-not-reported. It is also MLL-AF9 MOLM-13, a myeloid line, not the index fusion and not B-lineage.",
  "absence_class_for_the_per_gene_score": "measured-not-reported"
}
How this could bite us back
  • The HBO1 complex is required for normal haematopoietic stem cell function, so blocking the MLL-HBO1 contact risks the graft as well as the blast.
  • The independent functional dependency is for MLL-ELL, not for KMT2A::AFF1; designing against the THD2-HBO1 interface without first checking that the index fusion retains THD2 would repeat this project's commonest error.
  • JADE paralogues are not interchangeable at the nucleosome, so a JADE3-directed reagent may be compensated by JADE1 or JADE2 - the HBO1 complex has a built-in redundancy that a single-subunit strategy has to defeat.
Still open
  1. Does KMT2A::AFF1 retain the THD2 through which MLL contacts the HBO1 complex? The answer is determined by the breakpoint and would decide whether the independent PHF16 contact can apply to the index fusion at all - this is the single most valuable next question in this beat.
  2. Are ING4, ING5, KAT7/HBO1, MEAF6 or JADE1/JADE2 in the Skucha 960-protein network? An HBO1 complex signature would make JADE3 the best-evidenced pan-fusion module here.
  3. Is JADE3 rather than JADE1 or JADE2 the paralogue in the MLL-associated HBO1 complex in a t(4;11) B-ALL cell?
  4. Where is JADE3 in the nucleus? Nobody has published an answer, and UniProt has no entry for it.
Every source behind this entry (4)
  1. E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    JADE3 co-purified with all seven Strep-HA-tagged MLL fusions in isogenic Jurkat cells, reaching p = 2.11e-12 with 43|41 spectra for MLL-ENL, with the MLL-AF4 (= KMT2A::AFF1) call at p = 0.00529 and 5|5 spectra against a 0|0 negative control.

    MLL-fusion-driven leukemia requires SETD2 to safeguard genomic integrity. · Skucha A, Ebner J, Schmoellerl J, Roth M, Eder T, Cesar-Razquin A, Stukalov A, Vittori S, Muhar M, Lu B, Aichinger M, Jude J, Mueller AC, Gyorffy B, Vakoc CR, Valent P, Bennett KL, Zuber J, Superti-Furga G, Grebien F · Nature Communications 2018
    PMID 29777171DOI
    What this evidence cannot show

    SYSTEM: the AP-MS host is JURKAT, a T-ALL line, whereas KMT2A::AFF1 disease is B-lineage infant leukaemia, so a partner that needs a B-cell or infant proteome cannot appear here and a T-lineage partner can appear spuriously. The bait is a Strep-HA-tagged, SINGLE-COPY, doxycycline-INDUCED transgene, not the endogenous fusion. Prey was recovered from bulk nuclear lysate, so co-purification is not evidence of a direct contact. INCLUSION RULE: the paper kept the top 300 interactions per bait by combined p-value rather than applying a p threshold (Methods, verbatim: "The 300 most significant interactions per bait were retained"). Measured by this beat by reading the cached Supplementary Table 5 xlsx directly on 2026-09-21: 545 of the 2,049 per-bait p-values in the table (26.6%) exceed 0.05, to a maximum of 0.175. This graph's own king analysis ana-the-index-fusion-interactome-is-not-distinctive-and-two-earlier-answers-were-artifacts found the MLL-AF4 column reaches deepest into the weak tail of the seven, so an MLL-AF4 p-value is systematically weaker than the same protein's p-value under another bait and should not be read as a weaker interaction. This is the strongest pan-fusion profile in this beat, but the index-fusion call still rests on five spectra per replicate. JADE3 is not discussed anywhere in the paper's body text; it exists only as a row in the supplementary table. Values re-read from the cached xlsx on 2026-09-21.

    Evidence that argues against this (1)
    • IntAct has never curated this dataset (pubid:29777171 returns 0 rows against a working positive control), so no independent curator has re-examined these calls, and the raw data at ProteomeXchange PXD009338 have not been reanalysed here. Nothing in the paper validates this particular prey by an orthogonal assay: of the twelve proteins in this beat, only SETD2 is mentioned anywhere in the paper's body text (counted by regex over the PMC full text, 2026-09-21; SF3A3 appears once, as a named non-selective control). ProteomeXchange/PRIDE 2018
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    Leukemic MLL fusion proteins associate with the HBO1 histone acetyltransferase complex through their trithorax homology domain 2 in human cell lines, in a chromatin-bound H3K4me3 context, mediated mainly by the ING4/5 and PHF16 (= JADE3) subunits; MLL-ELL in particular depends on that association for leukemic transformation, and the HBO1 complex promotes loading of the AF4/ENL/P-TEFb complex onto target promoters.

    HBO1-MLL interaction promotes AF4/ENL/P-TEFb-mediated leukemogenesis. · Takahashi S, Kanai A, Okuda H, Miyamoto R, et al. · eLife 2021
    PMID 34431785DOI
    What this evidence cannot show

    The transformation arm is MLL-ELL in murine haematopoietic progenitors, not KMT2A::AFF1 in human B-lineage cells; AFF1/AF4 appears in this paper as a component of the recruited AEP complex, not as the fusion partner. The subunit attribution ('mainly ING4/5 and PHF16') is from the abstract; whether JADE3 alone is sufficient, and whether the same holds for MLL-AF4, is not established there. Read from the abstract. This is nevertheless the only independent, non-Skucha evidence in this beat that places any of the twelve proteins in physical contact with an MLL protein.

    Evidence that argues against this (1)
    • An interaction shown for MLL-ELL and attributed jointly to ING4/5 and PHF16 does not establish that JADE3 contacts KMT2A::AFF1; this project's recurring failure is exactly this kind of cross-fusion transfer. Skucha A et al. 2018
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The PZP domain of the JADE subunit engages the nucleosome through histone H3 and DNA and is necessary for the HBO1 complex to associate with chromatin targets; recognition of unmethylated H3K4 directs the complex's activity toward histone H4 acetylation, while H3K4 hypermethylation alters substrate selectivity, and the PZP domain augments the transforming activity of the NUP98-JADE2 fusion.

    Guiding the HBO1 complex function through the JADE subunit. · Gaurav N, Kanai A, Lachance C, Cox KL, Liu J, et al. · Nat Struct Mol Biol 2024
    PMID 38448574DOI
    What this evidence cannot show

    Structural, genomic and in vivo assembly work; the leukaemia arm is the NUP98-JADE2 fusion, a different fusion family and a different JADE paralogue. Read from the abstract. The H3K4-methylation-dependent substrate switch is mechanistically interesting for this graph because KMT2A fusions change H3K4 methylation, but nothing here tests that connection.

    Evidence that argues against this (1)
    • Paralogue specificity is substantial in this family, so a mechanism established for 'the JADE subunit' may not hold for JADE3 in particular. Gaurav N et al. 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The PZP domains of JADE1/2/3 and BRPF1 engage the nucleosome in paralogue-specific ways: only JADE1's PZP has strong intrinsic DNA binding and binds almost the entire histone H3 tail, whereas BRPF1's association with the nucleosome is about a thousand-fold weaker.

    Distinct roles of the JADE and BRPF scaffolding subunits of the acetyltransferase HBO1 complex. · Gaurav N, Ebert LK, Lachance C, Selvam K, Chen RW, et al. · Nat Commun 2026
    PMID 42431943DOI
    What this evidence cannot show

    Structural and biochemical, in vitro nucleosome binding; no leukaemia and no fusion. Read from the abstract. Cited to bound E3 - a JADE-family mechanism is not automatically a JADE3 mechanism, and the abstract singles out JADE1 rather than JADE3 for the strong DNA binding.

    Evidence that argues against this (1)
    • In vitro nucleosome affinity need not rank the same way in a cell where the complex is pre-assembled and targeted by partners. Gaurav N et al. 2024
Connected to