AAR2 (C20orf4) is the human ortholog of yeast Aar2p and functions as an assembly factor for the U5 small nuclear ribonucleoprotein particle, controlling the timing of spliceosome maturation. It is a component of a cytoplasmic precursor U5 snRNP containing PRPF8, EFTUD2/Snu114, U5 snRNA and the Sm proteins, but is excluded from the mature tri-snRNP and from the assembled spliceosome. AAR2 binds the RNase H-like domain of PRPF8 and acts as a placeholder and checkpoint: it occupies the surface that the SNRNP200/Brr2 helicase must ultimately engage and blocks loading of U4/U6 di-snRNA, so premature spliceosome activation is prevented. In yeast, where the mechanism was solved structurally, Aar2p and Brr2p are mutually exclusive binders of Prp8 and Aar2p is exchanged for Brr2p to yield the mature particle, with the handoff governed by phosphorylation. The human protein appears to differ in an important respect: human AAR2 co-purifies with all four members of the RNA-free heterotetrameric complex, which suggests it stays associated until that complex is fully formed rather than being displaced at the SNRNP200 step, and the human AAR2-PRPF8 interaction itself is structurally distinct from the yeast one. Structural work on the human complex also indicates a role beyond simple steric placeholding, since AAR2 locks the PRPF8 RNase H domain in a conformation compatible only with the first transesterification step of splicing and blocks the switch to the step-2 conformation.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005682 U5 snRNP | IBA GO_REF:0000033 | ACCEPT | Summary: Well-founded phylogenetic inference. The WITH/FROM field cites SGD:S000000170, which is budding yeast AAR2 - the ortholog in which this protein's role was originally defined - together with the PANTHER node PTN000294070. Yeast Aar2p is an established component of the cytoplasmic precursor U5 snRNP, and the human protein has since been shown directly to be a true ortholog that binds the same PRPF8 RNase H domain. Core cellular component. Supporting Evidence: PMID:26527271 as a true Aar2 orthologue which binds to the RH domain (hsRH) of Prp8 and file:human/AAR2/AAR2-uniprot.txt FUNCTION: Component of the U5 snRNP complex that is required for |
| GO:0045292 mRNA cis splicing, via spliceosome | IBA GO_REF:0000033 | ACCEPT | Summary: Phylogenetic inference from PomBase:SPAC3H5.04 (fission yeast) via PANTHER node PTN000294070. Correct: AAR2 is required for pre-mRNA splicing through its role in producing a mature, catalytically competent U5 snRNP. Note this is an indirect contribution - AAR2 is excluded from the assembled spliceosome and does not participate in the splicing reaction itself - but "involved_in" is appropriate for an assembly factor that the process requires. Supporting Evidence: file:human/AAR2/AAR2-uniprot.txt FUNCTION: Component of the U5 snRNP complex that is required for |
| GO:0005515 protein binding | IPI PMID:16189514 Towards a proteome-scale map of the human protein-protein in... | MARK AS OVER ANNOTATED | Summary: Interaction with EAPP (UniProtKB:Q56P03) from an early proteome-scale two-hybrid map. EAPP is best known as the E2F-associated phosphoprotein, but it is NOT unconnected to this gene's biology: PMID:34131137 - the same paper that establishes TSSC4 as a U5 chaperone - describes EAPP as "a putative chaperone EAPP that has been shown to interact with U5 proteins PRPF8 and EFTUD2". So the AAR2-EAPP interaction is at least compartmentally and functionally plausible, and it should not be dismissed as a cross-pathway artefact. It is nonetheless marked over-annotated, on narrower grounds: EAPP's own U5 role is explicitly "putative", the AAR2-EAPP pair specifically has never been followed up in any focused study, and bare "protein binding" conveys nothing about AAR2's activity regardless of whether the partner is real. This is a weaker verdict than "wrong partner" - if EAPP's U5 chaperone role is confirmed, these annotations would deserve upgrading to an informative term rather than discounting. Reason: Uninformative term for a pair that has never been characterised. Not dismissed as cross-pathway: EAPP is a putative U5 chaperone, so this may be real and merely under-annotated. Supporting Evidence: PMID:34131137 a putative chaperone EAPP that has been shown to interact with U5 proteins PRPF8 and EFTUD2 |
| GO:0005515 protein binding | IPI PMID:28514442 Architecture of the human interactome defines protein commun... | MARK AS OVER ANNOTATED | Summary: Second large-scale recovery of the AAR2-EAPP (UniProtKB:Q56P03) interaction. Repetition across screens makes the physical interaction more likely to be real, and EAPP is a putative U5 chaperone reported to interact with PRPF8 and EFTUD2, so the pairing is plausible. It remains uninformative as annotated: the AAR2-EAPP pair itself has not been pursued, and the term conveys nothing. Reason: Reproducible but uncharacterised interaction under an uninformative term. Supporting Evidence: PMID:34131137 a putative chaperone EAPP that has been shown to interact with U5 proteins PRPF8 and EFTUD2 |
| GO:0005515 protein binding | IPI PMID:33961781 Dual proteome-scale networks reveal cell-specific remodeling... | MARK AS OVER ANNOTATED | Summary: Third recovery of AAR2-EAPP (UniProtKB:Q56P03), this time by BioPlex affinity-purification mass spectrometry - methodologically orthogonal to the two-hybrid screens, which strengthens the case that the interaction is physically real. Combined with EAPP's reported association with PRPF8 and EFTUD2, the interaction is quite likely genuine; what is missing is a study characterising it, and a term that says anything. Reason: Reproducible but functionally uncharacterised high-throughput hit under an uninformative term. Supporting Evidence: PMID:34131137 a putative chaperone EAPP that has been shown to interact with U5 proteins PRPF8 and EFTUD2 |
| GO:0000387 spliceosomal snRNP assembly | NAS PMID:36322420 Structural and functional investigation of the human snRNP a... | ACCEPT | Summary: Correct and core: AAR2 is a U5 snRNP assembly factor, which is the whole of its known function. The evidence code understates the source badly, though. PMID:36322420 is a structure-function study of the human AAR2-PRPF8 RNase H complex - it reports a crystal structure, designed interface variants that fail to bind PRPF8 in vitro, and size-exclusion interaction mapping against U5 proteins. Those are direct experimental results on the human protein, not a non-traceable author statement. Accepted as a core process, with the evidence-code question raised in suggested_questions. Supporting Evidence: PMID:36322420 Protein-interaction studies of AAR2 with U5 proteins using size-exclusion chromatography reveal similarities and marked differences in the interaction patterns compared with yeast Aar2p file:human/AAR2/AAR2-uniprot.txt FUNCTION: Component of the U5 snRNP complex that is required for |
| GO:0005682 U5 snRNP | NAS PMID:36322420 Structural and functional investigation of the human snRNP a... | ACCEPT | Summary: Correct core localisation, and the same evidence-code observation applies as for the snRNP-assembly annotation from this reference. AAR2 is a component of the precursor U5 snRNP, though notably not of the mature tri-snRNP or of the assembled spliceosome. When it leaves is species-dependent: in yeast Aar2p is exchanged for Brr2p, whereas human AAR2 co-purifies with all four members of the RNA-free heterotetrameric complex, indicating it stays bound until that complex is fully assembled. Supporting Evidence: PMID:26527271 as a true Aar2 orthologue which binds to the RH domain (hsRH) of Prp8 and PMID:34131137 In yeast, Aar2p is exchanged with Brr2p13,14, while in humans AAR2 co-purifies with all four RHC members12, which indicates that AAR2 stays associated until the whole RHC is formed. |
| GO:0048025 negative regulation of mRNA splicing, via spliceosome | NAS PMID:36322420 Structural and functional investigation of the human snRNP a... | ACCEPT | Summary: Counter-intuitive for a splicing assembly factor, but correct, and worth stating why. AAR2's function is genuinely inhibitory at the molecular level: it occupies the PRPF8 surface that the SNRNP200/Brr2 helicase must engage and blocks U4/U6 di-snRNA loading, holding the particle in an immature state until the handoff (phosphorylation-triggered in yeast; the human exchange step is less clear, since AAR2 co-purifies with the whole heterotetrameric complex). The human structural work sharpens this further - AAR2 locks the PRPF8 RNase H domain in a conformation compatible only with the first transesterification step and blocks the switch to the step-2 conformation. The negative regulation is the mechanism by which AAR2 enforces assembly order, not an incidental side effect. Supporting Evidence: PMID:36322420 seems to lock PRPF8 RH in a conformation that is only compatible with the first file:human/AAR2/AAR2-deep-research-affinage.md sterically occludes the binding sites for the Brr2/SNRNP200 helicase while also occupying the RNase H RNA-binding surface to block U4/U6 di-snRNA loading, thereby preventing premature spliceosome activation |
| GO:0005515 protein binding | IPI PMID:34131137 TSSC4 is a component of U5 snRNP that promotes tri-snRNP for... | MODIFY | Summary: This one is different from the other three protein-binding annotations and should not be dismissed with them. The partner is TSSC4 (UniProtKB:Q9Y5U2), and the source is not a screen but a dedicated functional study establishing TSSC4 as a U5 snRNP component and chaperone that promotes tri-snRNP formation, interacting with the U5-specific proteins PRPF8, EFTUD2 and SNRNP200. AAR2 and TSSC4 are therefore both U5 snRNP assembly chaperones acting on the same particle with overlapping partners, so the interaction is what the biology predicts rather than an unexplained hit. On the replacement term, one caveat is recorded rather than glossed. GO:0044877 protein-containing complex binding normally expects a complex in the WITH/FROM field, whereas this row's WITH/FROM is a single protein accession, and the AAR2-specific datum in the source is IP-MS co-purification with the paper explicitly declining to establish direct contacts. The term is chosen because the interaction demonstrably occurs in the context of the U5 particle rather than as an isolated pair - which is exactly what co-purification shows and what a bare pairwise term misses - but a curator may reasonably prefer to leave the row as GO:0005515 until a direct AAR2-TSSC4 contact is demonstrated. Flagged in suggested_questions. Reason: A real, mechanistically expected interaction recorded under an uninformative term. GO:0044877 is used because TSSC4 is engaged in the context of the U5 snRNP particle, which is what the co-purification evidence supports; a term for snRNP assembly-factor binding would be better still. Proposed replacements: protein-containing complex binding Supporting Evidence: PMID:34131137 Specifically, TSSC4 interacts with U5-specific proteins PRPF8, EFTUD2 PMID:34131137 TSSC4 emerges as a specific chaperone that acts in U5 snRNP de novo |
| GO:0000244 spliceosomal tri-snRNP complex assembly | ISS GO_REF:0000024 | ACCEPT | Summary: Sequence-similarity transfer from yeast Aar2p (UniProtKB:P32357). Appropriate: AAR2 acts upstream of tri-snRNP formation by holding the precursor U5 snRNP in an immature state and then being displaced, and human AAR2 has been confirmed as a true ortholog of the yeast protein binding the same PRPF8 domain. Note the direction of AAR2's contribution - it does not build the tri-snRNP so much as gate the step that permits it. Supporting Evidence: PMID:26527271 as a true Aar2 orthologue which binds to the RH domain (hsRH) of Prp8 and file:human/AAR2/AAR2-uniprot.txt FUNCTION: Component of the U5 snRNP complex that is required for |
| GO:0005682 U5 snRNP | ISS GO_REF:0000024 | ACCEPT | Summary: Sequence-similarity transfer of U5 snRNP membership from yeast Aar2p (UniProtKB:P32357). Redundant with the IBA and NAS annotations to the same term but correct, and it rests on the same well-established orthology that UniProt uses for its own ECO:0000250 FUNCTION statement. Supporting Evidence: file:human/AAR2/AAR2-uniprot.txt FUNCTION: Component of the U5 snRNP complex that is required for |
Loading supporting content…
Download this section (compressed HTML)Q: Should the three NAS annotations from PMID:36322420 be upgraded? That paper is a structure-function study of the human AAR2-PRPF8 RNase H complex, with a crystal structure, designed interface variants that fail to bind PRPF8 in vitro, and size-exclusion interaction mapping. Its results support GO:0000387, GO:0005682 and GO:0048025 directly, so IDA or IPI would reflect the evidence better than a non-traceable author statement.
Q: Does AAR2 have a molecular function term it should carry? Its activity is well defined - binding the PRPF8 RNase H domain so as to exclude SNRNP200 and block di-snRNA loading - yet GOA records only bare protein binding. This is the same gap seen for AAGAB in this campaign: assembly chaperones and placeholder factors have no molecular-function vocabulary in GO beyond binding terms.
Q: Is the AAR2-TSSC4 interaction direct? The evidence is IP-MS co-purification within U5 particles and the source paper explicitly declines to establish direct contacts, describing "multiple weak contacts". That distinction determines whether GO:0044877 protein-containing complex binding is the right replacement term for that row, or whether it should stay as bare protein binding pending a demonstrated direct contact.
Q: Does human AAR2 leave the particle at the same step as yeast Aar2p? In yeast, Aar2p is exchanged for Brr2p, but human AAR2 co-purifies with all four members of the RNA-free heterotetrameric complex, indicating it stays bound until that complex is complete. The human AAR2-PRPF8 structure also differs from the yeast one. Resolving this determines whether the SNRNP200-displacement model transfers to human at all.
Q: Which kinases actually control the AAR2-PRPF8 handoff in human cells? CK2alpha1 and SGK2 have been proposed as candidates that abrogate the interaction, and the yeast S253E phospho-mimetic lowers Aar2p affinity for Prp8. Establishing the physiological kinase and site would allow a specific phosphorylation-dependent regulatory annotation.
Q: Is EAPP's putative U5 chaperone role real, and is the AAR2-EAPP interaction an assembly-context contact? PMID:34131137 describes EAPP as "a putative chaperone EAPP that has been shown to interact with U5 proteins PRPF8 and EFTUD2", which places it in the same pathway as AAR2 rather than only in the E2F pathway it is better known for. The AAR2-EAPP pair is reproducible across three screens including an orthogonal AP-MS method, but has never been followed up directly, and EAPP's U5 role is still labelled putative. Confirming it would turn three uninformative protein-binding rows into an assembly-context annotation.
Q: How do AAR2 and TSSC4 divide labour? Both are U5 snRNP assembly chaperones binding overlapping sets of U5 proteins, and they interact with each other, but whether they act sequentially, competitively or as a complex during de novo biogenesis versus post-splicing recycling is unresolved.
Experiment: Deplete AAR2 in human cells and characterise the resulting U5 particles by gradient fractionation and mass spectrometry, asking specifically whether SNRNP200 loads onto PRPF8 early and whether U4/U6 di-snRNA associates prematurely, rather than only quantifying tri-snRNP yield.
Hypothesis: AAR2 enforces the order of U5 snRNP maturation, so bypassing it produces prematurely activated particles rather than simply fewer particles.
Type: cell biology and proteomics
Experiment: Map AAR2 phosphosites by mass spectrometry in human cells, test the candidate kinases CK2alpha1 and SGK2 by targeted inhibition and knockdown, and assay phospho-mimetic and phospho-dead AAR2 variants for PRPF8 binding and for the ability to support tri-snRNP formation.
Hypothesis: The AAR2-PRPF8 handoff is triggered by phosphorylation at a defined site by a specific kinase.
Type: phosphoproteomics and biochemistry
Experiment: Determine whether AAR2 and TSSC4 occupy the same particle simultaneously or sequentially, using sequential immunoprecipitation and cryo-EM of isolated precursor particles, and test whether depletion of one is compensated by the other in de novo biogenesis versus post-splicing recycling.
Hypothesis: AAR2 and TSSC4 act at distinct steps of U5 snRNP biogenesis rather than redundantly.
Type: structural and cell biology
Experiment: Use the designed AAR2 interface variants that fail to bind PRPF8, plus variants that bind but do not impose the step-1 conformation, to test in cells whether the conformational lock is required for splicing fidelity as opposed to merely for particle assembly.
Hypothesis: The human-specific conformational lock on the PRPF8 RNase H domain is functionally important, not a crystallisation artefact.
Type: structure-function
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)