AP2B1 (beta-2-adaptin) - the only one of AFF4's four two-hybrid partners that is not a hub, and the only one with an echo in this graph's endocytosis biology
also called AP2B1 · ADTB2 · CLAPB1 · beta-2-adaptin · beta-adaptin · AP-2 complex subunit beta · P63010
AP2B1 is part of the machinery that pulls cargo into cells from the cell surface. It was reported as a partner of AFF4 in one large automated yeast screen, and unlike the other three proteins from that screen it does not stick to everything, which makes it the least dismissible of them.
The same thing, for a scientist
AP2B1 (UniProt **P63010, 937 aa**, primary gene name read back 2026-09-21) is the beta subunit of the AP-2 clathrin adaptor complex. It reaches this graph as a **yeast two-hybrid partner of AFF4** in Rolland et al. 2014, recorded as three IntAct rows that are three assay annotations of **one screen**, in host *S. cerevisiae*, MI-score 0.56 [E1]. **Provenance first: a proteome-scale two-hybrid map, never replicated, and nobody has ever studied this pair.** `AP2B1[tiab] AND (leukemia OR leukaemia)` returns **1** record and it title-checks off-beat; the AP2B1/beta-2-adaptin query against MLL, KMT2A, CALM-AF10 or PICALM returns **1**, also off-beat [E3]. So the on-topic literature is **zero**. What distinguishes AP2B1 from GOLGA2, MTUS2 and TRAF2 - the three other proteins fished by AFF4 in the same screen - is measurable: **AP2B1 has 10 distinct partners in that screen against GOLGA2's 249, MTUS2's 153 and TRAF2's 97**, with AFF4 itself at 5 [E2]. That removes the commonest artefact explanation and establishes nothing positive. **One adjacency is worth naming and is not evidence.** This graph already carries `gene-picalm` - the clathrin assembly protein that is the 5' partner of `fusion-picalm-mllt10` (CALM-AF10) - and `mechanism-calm-af10-impaired-endocytosis-and-iron-uptake`, which is a KMT2A-like leukaemia fusion whose mechanism involves clathrin-mediated endocytosis. AP-2 is the adaptor PICALM works alongside in that process [E4]. **An AP-2 subunit turning up beside an elongation scaffold in a yeast assay is not a finding about CALM-AF10**, and I am not proposing one; it is recorded because a reader of this graph will notice the coincidence and should find it already noticed and already discounted.
This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.
I have never read one of these pictures before
A protein is one long chain of amino acids that folds up. The ribbon traces that chain from one end to the other; it is not a wire, it is a simplification of where the backbone runs. Drag to rotate, scroll to zoom.
Domains are the stretches of chain that do a particular job, and they are what the coloured blocks above the picture mark out. Click one and the view flies to it. In this project the important question about any domain is usually whether the fusion still has it — the labels say lost where it does not.
Anything in yellow sticks is a bound small molecule, which is usually the drug.
What the evidence is
AFF4 as DB-fusion bait and AP2B1 as AD-fusion prey in a yeast two-hybrid screen, scored in the primary pooled screen, retested on an array, and retested pairwise. No mammalian cell, no endogenous protein, no region mapped, no affinity, no structure, no reciprocal result in any other assay class.
The bait/prey roles that made AP2B1 "bait-supported" in the recount are the DB/AD construct orientation. In an atlas every protein is a bait; the label does not mean anyone set out to find what binds AFF4.
The bait is wild-type AFF4, not AFF1 and not any KMT2A fusion. There is no experiment placing AP2B1 near any fusion protein in any system.
The PICALM adjacency, stated so that it can be dismissed properly
PICALM (CALM) is a clathrin assembly protein and the 5' partner of the PICALM::MLLT10 fusion that this graph treats as the control experiment for DOT1L-dependent leukaemia. Its normal job is in clathrin-mediated endocytosis, where it works with the AP-2 adaptor complex of which AP2B1 is the beta subunit. This graph carries mechanism-calm-af10-impaired-endocytosis-and-iron-uptake, in which the fusion's disruption of that pathway is a documented consequence.
So the shape of the coincidence is: an AP-2 subunit binds an elongation scaffold in yeast, and a different clathrin adaptor is fused to an elongation-complex partner in a real leukaemia. Stated plainly it dissolves. The two facts share a pathway name and nothing else. PICALM's leukaemic role is as a fusion N-terminus contributing an oligomerisation domain, not as an endocytic adaptor recruiting elongation machinery, and no experiment connects AP-2 to AFF1, AFF4 or any KMT2A fusion.
It is written down because a graph that makes adjacencies visible will surface this one, and the useful output of surfacing it is a recorded rejection rather than a new hypothesis.
| uniprot | P63010 |
|---|---|
| gene symbol | AP2B1 |
| length AA | 937 |
| alphafold ID | AF-P63010-F1 |
| protein class | alpha-solenoid adaptin; beta subunit of the heterotetrameric AP-2 clathrin adaptor |
| uniprot resolution | gene_exact:AP2B1 AND organism_id:9606 AND reviewed:true returned ONE reviewed row on 2026-09-21; primary gene name read back as AP2B1; entry AP2B1_HUMAN; 937 aa; synonyms ADTB2, CLAPB1. |
| domains note | No domain boundaries asserted; the adaptin trunk, hinge and ear ranges were not read back from UniProt with residue numbers this cycle. |
| shown for fusion | Nobody has looked in any fusion. The assay is wild-type AFF4 against AP2B1 in yeast; there is no experiment in AFF1, in any KMT2A fusion, or in any human cell. |
domains
[]
interaction with the core complex
{
"partner": "AFF4",
"partner_accession": "Q9UHB7",
"our_protein_role": "prey (AD fusion)",
"bait": "AFF4 (DB fusion)",
"methods": [
"psi-mi:\"MI:1112\"(two hybrid prey pooling approach)",
"psi-mi:\"MI:0397\"(two hybrid array)",
"psi-mi:\"MI:1356\"(validated two hybrid)"
],
"interaction_type": "psi-mi:\"MI:0915\"(physical association)",
"expansion": "none - these are binary assays",
"host_system": "Saccharomyces cerevisiae (taxid:4932)",
"source_publications": [
"25416956"
],
"distinct_experiments": 1,
"source_class": "proteome-scale yeast two-hybrid atlas",
"intact_acs": [
"intact:EBI-10627110",
"intact:EBI-10321207",
"intact:EBI-10473943"
],
"intact_miscore": 0.56,
"replicated_in_huri_luck_2020": false,
"replication_caveat": "IntAct holds no AFF4 row from PMID 32296183 at all; whether AFF4 was in HuRI's search space is not resolvable from IntAct",
"region_mapped": false,
"affinity_measured": false,
"structures": []
}promiscuity controls 2026 09 21
{
"distinct_partners_in_rolland_2014": 10,
"comparators_in_the_same_screen": {
"GOLGA2": 249,
"MTUS2": 153,
"TRAF2": 97,
"AFF4": 5
},
"distinct_partners_all_of_intact": 158,
"distinct_publications_all_of_intact": 55,
"source": "IntAct PSICQUIC MITAB, retrieved 2026-09-21"
}counted absences
{
"leukaemia": {
"query": "AP2B1[tiab] AND (leukemia[tiab] OR leukaemia[tiab])",
"records_raw": 1,
"records_on_topic": 0,
"pmid": "35678671"
},
"fusion_families": {
"query": "(AP2B1[tiab] OR beta2-adaptin[tiab] OR \"beta 2 adaptin\"[tiab]) AND (MLL[tiab] OR KMT2A[tiab] OR CALM-AF10[tiab] OR PICALM[tiab])",
"records_raw": 1,
"records_on_topic": 0,
"pmid": "23811784"
},
"aliases_searched": [
"AP2B1",
"beta2-adaptin",
"beta 2 adaptin"
],
"aliases_not_searched": [
"ADTB2",
"CLAPB1"
],
"complex_form_not_searched": "\"AP-2 complex\" - collides with the TFAP2 transcription factor family and was not attempted (brief rule 64 gap)",
"date": "2026-09-21",
"bound": "zero in titles and abstracts over three aliases; supporting figures not searched; complex-level phrase not run"
}adjacency recorded and rejected
{
"observation": "AP-2 is the clathrin adaptor complex that PICALM works alongside in clathrin-mediated endocytosis, and PICALM is the 5' partner of PICALM::MLLT10 (CALM-AF10), a fusion this graph treats as the DOT1L control experiment.",
"related_graph_entries": [
"gene-picalm",
"fusion-picalm-mllt10",
"mechanism-calm-af10-impaired-endocytosis-and-iron-uptake"
],
"verdict": "coincidence of pathway vocabulary, not evidence. PICALM's leukaemic contribution is as a fusion N-terminus, not as an endocytic adaptor, and no experiment connects AP-2 to AFF1, AFF4 or any KMT2A fusion. Recorded so the coincidence is found already discounted."
}- Nothing in this entry supports an intervention. Its risk is in the other direction: the PICALM adjacency is exactly the kind of pathway-name coincidence that generates a plausible-sounding mechanism in a summary, and the entry exists partly to stop that happening.
- If this pair were real, AP-2 is a core housekeeping complex present in every cell, so any strategy built on disrupting an AFF4-AP2B1 contact would have no therapeutic window - the essential-function toxicity is total before resistance is even considered.
- Does endogenous AP2B1 co-purify with AFF4 or AFF1 in any mammalian cell? A single AP-MS experiment would either promote this pair or retire it, and no such measurement exists in BioPlex despite AFF4 having been a bait there twice.
- Is beta-2-adaptin's AFF4-binding surface accessible in the assembled AP-2 heterotetramer, or only in the isolated subunit the two-hybrid assay presents? If only the free subunit binds, the interaction cannot occur at physiological stoichiometry and the pair should be retired regardless of how well it replicates.
- Was AFF4 in HuRI's search space in 2020? If it was and the pair did not score, this becomes a failed replication rather than an untested one.
- E1weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.database
IntAct records AFF4 and AP2B1 as interacting in three rows, all from PMID 25416956 / IMEx IM-23318, in host Saccharomyces cerevisiae, annotated two hybrid prey pooling approach, two hybrid array and validated two hybrid, MI-score 0.56.
IntAct molecular interaction database, PSICQUIC MITAB 2.7, query id:P63010 AND id:Q9UHB7, 3 rows retrieved 2026-09-21 · IntAct (EMBL-EBI) 2026What this evidence cannot show
Three rows, one publication, one screen. Yeast two-hybrid in S. cerevisiae detects binary contact between fusion constructs in a heterologous nucleus, without mammalian post-translational modification, stoichiometry or compartment. Not recovered in HuRI (Luck 2020). No region mapped, no affinity, no structure, no mammalian-cell confirmation, and the bait is wild-type AFF4 rather than any fusion.
Evidence that argues against this (1)
- Rolland et al. argue their systematic map is at least as accurate as the literature-curated interactome, so the discount here should be for the absence of follow-up on this particular pair rather than for the assay class. That argument cuts against dismissing the pair outright and is the reason this entry is weak rather than speculative. Rolland T et al. 2014
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.analysis
Within PMID 25416956, AP2B1 has 10 distinct partners while the three other AFF4 preys from the same screen have 249 (GOLGA2), 153 (MTUS2) and 97 (TRAF2), and AFF4 itself has 5.
The four AFF4 two-hybrid partners are three rows of one screen, not three experiments · Oncofusion knowledge graph, ana-the-aff4-two-hybrid-quartet-is-one-screen-and-three-of-the-four-preys-are-hubs 2026What this evidence cannot show
Degrees computed from IntAct's import of that publication on 2026-09-21, deduplicated to distinct accessions, not from the paper's supplementary tables. A low prey degree removes an artefact explanation; it is not evidence of specificity, and a protein can be a specific contaminant of one bait class while having few partners overall.
Evidence that argues against this (1)
- AP2B1's low degree is specific to this screen. Across all of IntAct it has 158 distinct partners over 55 publications, including 32 in BioPlex 3.0, so it is a well-connected protein in general and its quietness in Rolland 2014 may reflect the screen's coverage rather than the protein's behaviour. Huttlin EL et al. 2021
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
No literature connects AP2B1 to leukaemia or to the KMT2A/PICALM fusion families: AP2B1[tiab] AND (leukemia[tiab] OR leukaemia[tiab]) returns 1 record and (AP2B1[tiab] OR beta2-adaptin[tiab] OR 'beta 2 adaptin'[tiab]) AND (MLL[tiab] OR KMT2A[tiab] OR CALM-AF10[tiab] OR PICALM[tiab]) returns 1, and both title-check off-beat.
PubMed counted-absence searches for AP2B1 against leukaemia and the KMT2A/PICALM fusion families, run 2026-09-21 · PubMed (NCBI) 2026What this evidence cannot show
Both querytranslations expanded as written, with no [UID] rewriting. The two hits are PMID 35678671 (luteolin gene-expression array in breast cancer lines) and PMID 23811784 (clozapine effects on clathrin-mediated endocytosis gene expression in SH-SY5Y cells). Aliases searched: AP2B1, beta2-adaptin, 'beta 2 adaptin'; ADTB2 and CLAPB1 were not searched, and the complex-level phrase 'AP-2 complex' was not run - brief rule 64 makes that a real gap for a complex subunit. Zero in titles and abstracts; supporting figures not searched.
Evidence that argues against this (1)
- A subunit-symbol query cannot reach work framed at the complex level, and this project has twice recorded a false absence for exactly that reason. 'AP-2' is also an ordinary-word collision with the AP-2 transcription factor family (TFAP2A-E), which would make the complex-level query hard to title-check rather than impossible - it was not attempted, and the absence is bounded accordingly. Rolland T et al. 2014
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
AP2B1 is UniProt P63010, 937 aa, primary gene name AP2B1, synonyms ADTB2 and CLAPB1, recommended name AP-2 complex subunit beta, the beta subunit of the clathrin-associated AP-2 adaptor.
UniProtKB entry P63010 (AP2B1_HUMAN, AP-2 complex subunit beta) · UniProtKB/Swiss-Prot 2026What this evidence cannot show
Resolved 2026-09-21 with gene_exact:AP2B1 AND organism_id:9606 AND reviewed:true, fields accession,id,gene_primary,gene_synonym,protein_name,length; exactly one reviewed record, primary gene name read back. The AP-2 complex is a heterotetramer and this accession names one subunit; nothing in this entry is a statement about the assembled complex, whose behaviour in a two-hybrid assay of a single subunit is not tested.
Evidence that argues against this (1)
- A two-hybrid assay presents an isolated subunit out of its heterotetramer. Beta-2-adaptin's physiological binding surfaces are occupied by alpha-adaptin and by clathrin in the assembled complex, so a free-subunit interaction may be inaccessible in the cell - an objection that applies to this pair with unusual force and that no experiment has addressed. Huttlin EL et al. 2017
yeast two-hybrid hit in Rolland 2014; one screen, never replicated, but the only one of the four preys that is not a hub
the non-hub control that makes the degree argument informative rather than circular
AP-2 is the clathrin adaptor PICALM works alongside; recorded as a pathway-name coincidence and rejected, not proposed