AKAP8::ZNF384 - the same protein that MLL-rearranged leukaemia depends on is itself a rare B-ALL fusion partner, on the 5' side, donating its condensate-forming IDR
also called AKAP8-ZNF384 · AKAP95-ZNF384
AKAP95 turns up twice in leukaemia, in two completely different ways. In MLL-rearranged leukaemia it is a helper the cancer needs. In a rare kind of B-ALL it is itself half of the fusion gene that causes the disease.
The same thing, for a scientist
**AKAP8::ZNF384 is a genuine, if rare, recurrent fusion of B-cell precursor ALL.** In the UKALL14 whole-genome-sequencing series of 52 adult B-other BCP-ALL cases, five carried a ZNF384 fusion: **EP300::ZNF384 (n = 4) and AKAP8::ZNF384 (n = 1)** [E1]. It sits in a partner list of nineteen or more, which is dominated by chromatin and scaffold proteins - EP300, CREBBP, TCF3, TAF15, EWSR1, ARID1B, SMARCA4, SMARCA2, NIPBL, EHMT1 [E2]. **The architecture is the interesting part for this graph.** ZNF384 fusions almost invariably retain the **complete ZNF384 coding sequence** and replace its 5' end with the partner's N-terminus, so the partner contributes a transactivating or scaffolding module and ZNF384 contributes DNA binding. **The AKAP8 N-terminus is exactly the module independently shown to be a condensate-forming IDR (101-390) that co-condenses with the MLL1 N-terminal fragment** - the same region, in a different disease, doing the same kind of job [E3]. ZNF384-rearranged ALL generally carries a good prognosis [E2].
The same protein, two opposite roles - and why the graph should hold both
| MLL-rearranged leukaemia | ZNF384-rearranged B-ALL | |
|---|---|---|
| AKAP8's role | a dependency - the leukaemia needs the wild-type protein, and losing it impairs transformation 3 | the oncogene itself - AKAP8's 5' portion is fused to ZNF384 1 |
| region that matters | IDR1, aa 101-390 (co-condensation with MLL1T) | the N-terminal portion retained 5' of the breakpoint |
| direction of intervention | remove or dissolve AKAP95 | the AKAP8 moiety is a structural component of the driver |
| prognosis | KMT2A-r is adverse | ZNF384-r is favourable 2 |
A protein that is a dependency in one leukaemia and a fusion partner in another is not a coincidence in this class of biology - it is the signature of a transferable module. This graph already records the same shape for other proteins: a scaffold with a disordered region that can either be recruited to an oncogenic condensate or be welded into one. The testable consequence is that the AKAP8 breakpoint should fall 3' of IDR1, so the fusion keeps the condensation module - and that is checkable from a single published breakpoint, which this beat did not obtain.
Do not over-read this. The UKALL14 case is n = 1. Nothing is known about this particular fusion's biology: no functional study of AKAP8::ZNF384 exists, no breakpoint coordinates were located, and the assignment of the retained AKAP8 portion is an inference from the general ZNF384-fusion architecture, not from this case.
The cross-family analogy, stated at the level the evidence supports
The PI asks for the analogous biology in other fusion families. For AKAP95 the analogy runs two ways and only one of them is supported.
Supported: AKAP8 is itself a fusion partner, in a different lineage and a different family, contributing an N-terminal scaffolding module to a DNA-binding transcription factor 12. That is a structural analogy with EWSR1::FLI1 and with the NUP98 fusions, where a disordered low-complexity region is welded to a DNA-binding or homeodomain module - and this graph holds both comparators (localization-ewsr1-fli1-condensate-comparator, mechanism-nup98-fg-condensate-assay-audit).
Not supported, and worth saying so: there is no published example of AKAP95 acting as a recruited coactivator in the NUP98, EWSR1::FLI1, RUNX1::RUNX1T1 or PML::RARA systems. The 2026 paper's own framing places AKAP95 alongside ENL mutants, EWS-FLI1 and NUP98-HOXA9 as examples of oncogenic phase separation, not as an interactor of any of them 3. A beat looking for 'the AKAP95 of NUP98 fusions' will not find one, and the honest bound is that nobody has looked - the AKAP8/AKAP95 title-and-abstract literature is 92 records in total and contains no NUP98, EWSR1, RUNX1T1 or PML::RARA paper.
| gene5 | AKAP8 |
|---|---|
| gene3 | ZNF384 |
| cytoband | AKAP8 19p13.12; ZNF384 12p13.31 |
| uniprot5 | O43823 |
| uniprot3 | Q8TF68 |
| prognosis | ZNF384-rearranged ALL as a class carries a good prognosis and a younger age distribution; nothing is known specifically about the AKAP8 partner, n = 1. |
| shown for fusion | This IS a fusion entry. No functional study of AKAP8::ZNF384 exists in any system. |
diseases
[ "B-cell precursor acute lymphoblastic leukaemia (adult B-other)" ]
breakpoints
[]
frequency
{
"ukall14_b_other_n52_cases": 1,
"ukall14_znf384_fusions_n5_cases": 1
}measurements — the numbers behind the chart
[
{
"quantity": "count",
"subject": "AKAP8::ZNF384 cases in the UKALL14 B-other whole-genome series",
"value": 1,
"units": "cases",
"n": 52,
"universe": "52 adult B-other BCP-ALL cases with paired tumour-normal WGS from UKALL14, PMID 36867579",
"null": null,
"note": "1 of 52 B-other cases and 1 of 5 ZNF384 fusions in that series; EP300::ZNF384 accounts for the other 4"
},
{
"quantity": "count",
"subject": "reported ZNF384 fusion partners in ALL",
"value": 19,
"units": "partners",
"n": null,
"universe": "review enumeration, PMID 37306722; stated as 'more than 19'",
"null": null,
"note": "a lower bound taken from a review; not independently enumerated here"
},
{
"quantity": "count",
"subject": "papers in the complete AKAP8/AKAP95 title-abstract literature mentioning NUP98, EWSR1, RUNX1T1 or PML-RARA",
"value": 0,
"units": "records",
"n": 92,
"universe": "all 92 records matching AKAP8[tiab] OR AKAP95[tiab], efetched and keyword-scanned 2026-09-24",
"null": null,
"note": "bounds the cross-family analogy: AKAP95 has never been studied in another fusion family"
}
]- If the AKAP8 N-terminus is oncogenic when fused, an agent that dissolves AKAP95 condensates is not obviously safe in a ZNF384-rearranged patient - it could be active, inert or harmful depending on whether the fusion's condensate uses the same chemistry, and nothing is known.
- This entry is a single WGS case in an adult cohort. Treating a one-case fusion as a mechanistic parallel to a dependency established in mouse genetics is the kind of analogy that reads better than it evidences. The structural argument stands; the biology behind it is entirely unmeasured.
- The favourable prognosis of ZNF384-rearranged ALL is a warning for any argument that AKAP8-containing condensates are intrinsically aggressive. The same module sits in the good-risk disease.
- Where is the AKAP8 breakpoint, and does the fusion retain IDR1 (101-390)? If it does, AKAP8::ZNF384 and the AKAP95/MLL1T co-condensate use the same module, and the fusion is a natural experiment in what that module does when it is welded to a DNA-binding protein rather than recruited to one. The answer is in the UKALL14 supplementary data.
- Does AKAP8::ZNF384 phase-separate? No functional study of this fusion exists. It is a one-construct experiment and the prediction is specific: condensates in cells, abolished by mutating the six IDR1 tyrosines that abolish AKAP95-MLL1T co-condensation.
- Is AKAP95 a dependency in ZNF384-rearranged ALL that does NOT carry the AKAP8 partner? If the AKAP95 module matters to ZNF384 biology generally, the EP300::ZNF384 cases should need wild-type AKAP95 - and that is a DepMap-scale question nobody has asked.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In whole-genome sequencing of 52 adult B-other BCP-ALL cases from UKALL14, five cases carried a ZNF384 fusion: EP300::ZNF384 (n = 4) and AKAP8::ZNF384 (n = 1). Four of the five were aged 40 or under, compared with 18 of 47 in the remainder of the cohort, consistent with the reported younger age and good outcome of ZNF384-rearranged disease.
Diagnostic utility of whole genome sequencing in adults with B-other acute lymphoblastic leukemia. · Leongamornlert D, Gutierrez-Abril J, Lee S, Barretta E, et al. · Blood Adv 2023What this evidence cannot show
Full text retrieved and the AKAP8 sentence read directly; only the one paragraph was read, not the whole paper. n = 1 case. The paper reports the fusion call from WGS and RNA-seq classification and does NOT give breakpoint coordinates for it, name the retained AKAP8 exons, or report anything about its biology. The cohort is ADULT B-other ALL, not paediatric, which is this project's population.
Evidence that argues against this (1)
- A single WGS fusion call in one patient is not by itself evidence that AKAP8::ZNF384 is recurrent. The recurrence claim rests on the review's partner list, which is a compilation of case reports and cohort calls that this beat did not trace to their primaries. Until a second independent case is located, the honest statement is one reported case in a series where four of five ZNF384 fusions had a different partner. Zhu L et al. 2023
- E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.review
More than 19 ZNF384 fusion partners have been reported in ALL, including EP300, CREBBP, TCF3, TAF15, EWSR1, ARID1B, SMARCA4, SMARCA2, SYNRG, CLTC, BMP2K, NIPBL, AKAP8, C11orf74, DDX42, ATP2C1, EHMT1 and TEX41; ZNF384 rearrangement in ALL was first reported in 2002, and patients with ZNF384-rearranged ALL commonly have a good prognosis.
ZNF384-Related Fusion Genes in Acute Lymphoblastic Leukemia. · Zhu L, Bai W, Cheng Q, Fang J · Cancer Control 2023What this evidence cannot show
A REVIEW, cited as a map of the partner list only. Its statement that ZNF384 fusions retain the full ZNF384 coding sequence with the partner supplying the 5' end is the field's general architecture and was NOT verified against a primary source in this beat - the AKAP8 breakpoint specifically is unknown here.
Evidence that argues against this (1)
- The review's architectural premise - that ZNF384 fusions retain the complete ZNF384 coding sequence with the partner supplying the 5' end - is doing all the work behind this entry's claim that AKAP8 donates IDR1, and it was verified against no primary source here. If the AKAP8 breakpoint falls 5' of residue 101, the fusion carries little or none of IDR1 and the structural analogy with the MLLr co-condensate collapses entirely. The breakpoint is not reported in the one primary series located. Leongamornlert D et al. 2023
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
In MLL-rearranged leukaemia, wild-type AKAP95 is a dependency rather than an oncogene: its IDR1 (aa 101-390) co-condenses with the MLL1 translocated fragment, knockdown impairs MLLr line growth, and Akap95 loss impairs MLL-AF9 transformation while sparing normal haematopoiesis. The same paper cites ENL mutants, EWS-FLI1 and NUP98-HOXA9 as prior examples of oncogenic proteins that acquire phase separation for persistent transcriptional activation, but reports no AKAP95 interaction with any of them.
AKAP95 condensates regulate transcription and can be targeted in MLL-fusion driven oncogenesis. · Yan X, Luan Y, Zhu Y, Cai Y, Ye F, Yang L, Zhu K, Hu J, Luo S, Lin J, Wang Y, Huang Y, Sun P, Yang Z, Shi B, Jiang H, et al. · Nat Commun 2026What this evidence cannot show
Full text read. The EWS-FLI1 and NUP98-HOXA9 mentions are an introductory citation of prior literature, not a result; this paper contains no experiment with any non-MLL fusion. The expected_absent_terms entry records that absence as part of the claim.
Evidence that argues against this (1)
- Reading a paper's introductory citations as a survey is weak evidence of absence. These authors had no reason to test NUP98 or EWSR1::FLI1 and their silence measures their scope, not the field's. The bound that actually carries weight is the 0-of-92 enumeration recorded in this entry's measurements, and that bound inherits the rule-80 blind spot: a screen reporting AKAP95 among thousands of proteins in a NUP98 system would name it in no title or abstract and would be invisible to it. this project 2026
AKAP8 is the 5' partner, donating its N-terminal condensate-forming region
a disordered low-complexity module welded to a DNA-binding factor - the same architecture, a different family
the second cross-family condensate comparator this graph holds; AKAP95 has never been studied in NUP98 biology
the same AKAP8 region is a recruited coactivator in MLLr leukaemia and a fusion moiety in ZNF384-rearranged B-ALL
the 3' partner, retaining its complete eight-finger C2H2 array at 228-455
contrast: AKAP8::ZNF384 is a canonical 5'-partner fusion, this one inverts the order and is out of frame