CRYAA

Summary: IBA annotation for negative regulation of apoptotic process. CRYAA has well-documented anti-apoptotic activity. PMID:14752512 demonstrated that alphaA- and alphaB-crystallins bind to pro-apoptotic Bax and Bcl-X(S) proteins via GST pulldown and coimmunoprecipitation, preventing their translocation from cytosol to mitochondria during staurosporine-induced apoptosis. The R116C cataract mutant shows much weaker affinity to Bax and Bcl-X(S) (PMID:14752512). PMID:14512969 showed that the R49C mutant failed to protect lens epithelial cells from staurosporine-induced apoptosis. The IBA annotation is phylogenetically appropriate for sHSP family members with anti-apoptotic activity. This is a secondary/non-core function for CRYAA beyond its primary structural and chaperone roles.

Reason: Anti-apoptotic activity is well-supported by direct experimental evidence from PMID:14752512 and PMID:14512969, and the IBA annotation is phylogenetically sound. However, this is a secondary function compared to the core structural and chaperone holdase roles in the lens.

Summary: IBA annotation for cytoplasm localization. CRYAA is abundantly expressed in the cytoplasm of lens fiber cells. Multiple IDA-level studies confirm cytoplasmic localization including PMID:19464326, PMID:14752512, PMID:29259299, PMID:19503744, PMID:26004348, and PMID:30340470. UniProt also lists cytoplasm as a confirmed subcellular location. The IBA annotation is consistent with direct experimental evidence and is phylogenetically appropriate.

Reason: Cytoplasmic localization is the primary location of CRYAA, confirmed by multiple independent IDA studies using fluorescence microscopy in lens epithelial cells and other cell types. The IBA annotation is well supported.

Summary: IBA annotation for nucleus localization. UniProt notes that CRYAA (HSPB4) translocates to the nucleus during heat shock and resides in SC35 splicing speckles. This is a stress-induced secondary localization rather than the primary constitutive location. The IBA annotation is phylogenetically appropriate for sHSP family members that show nuclear translocation under stress.

Reason: Nuclear localization is real but conditional (stress-induced). UniProt records CRYAA translocation to SC35 splicing speckles during heat shock. This is not the primary constitutive localization of CRYAA, which is cytoplasmic. Keeping as non-core reflects that this is a secondary, stress-dependent localization.

Summary: IBA annotation for response to heat. CRYAA is a member of the small heat shock protein (sHSP) family and its chaperone-like activity was explicitly demonstrated using heat-induced aggregation assays (PMID:8943244). The protein suppresses heat-induced aggregation of aldose reductase and other lens proteins. CRYAA also translocates to nuclear SC35 speckles during heat shock (PMID:19464326). The IBA annotation is phylogenetically appropriate for the sHSP family whose hallmark is response to heat stress.

Reason: Response to heat is a core function of sHSP family members. CRYAA suppresses heat-induced protein aggregation (PMID:8943244) and shows heat-shock-dependent nuclear translocation (PMID:19464326). The IBA annotation is well justified for the sHSP family.

Summary: IBA annotation for protein refolding. This is problematic for CRYAA. While some sHSP family members can assist in protein refolding (e.g., HSPB1/Hsp27 cooperates with Hsp70 to refold substrates), CRYAA specifically acts as a holdase that prevents aggregation but does NOT refold proteins. This is documented by the NOT annotation to GO:0042026 (ISS from bovine ortholog, GO_REF:0000024). The IBA propagation from the broader sHSP family is overly broad for CRYAA, which lacks foldase activity. PMID:19464326 showed that HSPB1 and HSPB5 kept heat-unfolded substrates folding-competent, but HSPB7 did not support refolding, highlighting functional divergence within the family. CRYAA has holdase but not foldase activity. A focused OpenScientist function-assignment run on this hypothesis (file:human/CRYAA/CRYAA-hypotheses/function-hypothesis-go-0042026/openscientist.md) independently reached the holdase-not-foldase conclusion on mechanistic and architectural grounds: the UniProt record for P02489 has "no nucleotide-binding site and no ATPase domain", so CRYAA cannot be an autonomous foldase, and the productive refolding step is executed by downstream ATP-dependent Hsp70/Hsp100 chaperones (PMID:34055885, PMID:35281256). That run nonetheless recommended RETAINing the IBA as a non-core cooperative process. That recommendation is not adopted here, for two reasons that the run got factually wrong (see the reference_review on its report): it did not register that GOA already carries a NOT|involved_in ISS annotation on this very term, and its headline recommendation to "add the missing GO:0051082" is a no-op -- GOA already carries three GO:0051082 annotations for P02489 (IBA, IPI from PMID:8943244, IMP from PMID:18407550), and GO:0051082 is now formally obsolete, with the GO obsoletion note directing replacement to GO:0044183 or GO:0140309. On provenance: this IBA's actual ancestral evidence set is the WITH/FROM column of the GOA row, which is six Drosophila sHSP genes plus the PANTHER node -- FB:FBgn0001224 (Hsp23), FB:FBgn0001225 (Hsp26), FB:FBgn0001226 (Hsp27), FB:FBgn0011296 (l(2)efl), FB:FBgn0031037 (CG14207), FB:FBgn0035817 (CG7409), PANTHER:PTN000897708. Those fly seeds do carry direct refolding evidence at the source: the local DROME/Hsp23, DROME/Hsp26 and DROME/Hsp27 reviews each retain GO:0042026 on IDA-level cellular refolding assays (PMID:26705243, PMID:16572729). So the source supports the term; the question is whether it transfers to human alphaA, and a Drosophila-sHSP-to-human-alphaA transfer is a long reach. GOA's own NOT|involved_in ISS from the bovine ortholog (UniProtKB:P02470) is the direct answer that it does not. Separately, PMID:8093612 and PMID:1438232 are family-level literature context for the sHSP refolding property rather than part of this IBA's evidence chain, and neither is alphaA-specific. PMID:8093612 (Jakob 1993) assayed "murine Hsp25, human Hsp27, and bovine alpha-B-crystallin" -- alphaA-crystallin is not among the proteins named in the abstract (only the abstract is cached). PMID:1438232 (Horwitz 1992) reported refolding of GdnHCl-denatured gamma-crystallin using unfractionated alpha-crystallin (alphaA + alphaB) with a circular-dichroism readout, which as the run notes can reflect secondary-structure recovery once aggregation is suppressed rather than chaperone-catalyzed folding.

Reason: CRYAA is a holdase, not a foldase. It prevents aggregation of denatured proteins but does not refold them. The NOT annotation to GO:0042026 (ISS, GO_REF:0000024) explicitly documents this. The IBA propagation from the broader sHSP family is incorrect for CRYAA because not all sHSP members have refolding activity. There is functional divergence within the family (PMID:19464326). The IBA's seeds are six Drosophila sHSPs (see summary) which do have their own refolding evidence, so this is a transfer-failure rather than a bad source: the property does not reach human alphaA, as GOA's own NOT|involved_in ISS from bovine UniProtKB:P02470 states directly. The family-level literature context sometimes cited for refolding (PMID:1438232, PMID:8093612) is mixed-oligomer or non-alphaA and does not establish autonomous alphaA foldase activity; both the domain architecture (single alpha-crystallin domain, no nucleotide-binding domain) and the modern mechanistic consensus that Hsp70/Hsp100 execute the refolding step (PMID:34055885, PMID:35281256) argue against it.

Summary: GO:0051082 "unfolded protein binding" is now formally obsolete (go-ontology#30962). CRYAA/HSPB4 does interact with partially unfolded/destabilized proteins, but the term "unfolded protein binding" is problematic because it implies a simple binding function rather than the active chaperone holdase activity that CRYAA performs. CRYAA suppresses nonspecific aggregation of destabilized proteins (PMID:8943244) but does NOT refold them -- it is a holdase, not a foldase. The IBA annotation is phylogenetically propagated from sHSP family members across Drosophila, zebrafish, and mammals, which is appropriate for the family-level chaperone function. However, the term itself needs replacement, and as of 2026-07-25 no adequate replacement exists. GO:0051082 is now FORMALLY OBSOLETE ("obsolete unfolded protein binding"), and its obsoletion comment offers only two replacements: "protein folding chaperone (GO:0044183) or unfolded protein holdase activity (GO:0140309)". Neither fits CRYAA. GO:0044183 is for foldases that "assist the protein folding process", which CRYAA does not do. GO:0140309, despite having been relabelled to "unfolded protein holdase activity", retains an unchanged CARRIER-SPECIFIC definition -- "A protein carrier activity that binds to a protein in an unfolded state and escorts it to an acceptor molecule or to a specific location" -- and remains a child of GO:0140597 "protein carrier chaperone". CRYAA is an in-situ holdase that prevents aggregation without escorting substrates between compartments, so it does not satisfy that definition. Per projects/UNFOLDED_PROTEIN_BINDING.md, GO:0140309 was created (go-ontology#30552) for TIM carrier-holdases, and CRYAA is listed there as "retain GO:0051082; in-situ holdase, not carrier". The row is therefore modified to the proposed "holdase chaperone activity" term (see proposed_new_terms), consistent with the sibling sHSP reviews. GO:0140309 does not fit because it is carrier-specific. As a biological process annotation, GO:0050821 "protein stabilization" (defined as "Any process involved in maintaining the structure and integrity of a protein and preventing it from degradation or aggregation") is also relevant and is already annotated for CRYAA via PMID:12235146.

Reason: GO:0051082 is now formally obsolete. CRYAA has well-documented chaperone-like holdase activity: it suppresses aggregation of heat-denatured aldose reductase and singlet-oxygen-damaged gamma-crystallin (PMID:8943244). However, it does NOT refold proteins (NOT annotation to GO:0042026 protein refolding, ISS). The correct MF replacement does not yet exist: GO:0044183 is foldase-specific and GO:0140309 is carrier-specific (escorting between cellular components, per go-ontology#30552), which does not describe CRYAA's in-situ holdase activity. Replace with the proposed holdase chaperone activity term (NTR; see proposed_new_terms). GO:0050821 "protein stabilization" is appropriate as a BP term.

Proposed replacements: holdase chaperone activity

Summary: IBA annotation for lens development in camera-type eye. CRYAA is the most abundantly expressed protein in the ocular lens and is essential for normal lens development and transparency. UniProt states it is expressed in the eye lens (PubMed:12356833, PubMed:23255486). Knockout studies in zebrafish show abnormal differentiation of lens fiber cells when alpha-crystallins are absent (cited in PMID:29259299). Numerous CRYAA mutations cause congenital cataracts (PMID:9467006, PMID:14512969, PMID:18407550), demonstrating the essential role in lens development. The IBA annotation is appropriate for crystallin family members involved in lens development.

Reason: CRYAA is a core lens protein essential for normal lens development. Mutations consistently cause congenital cataracts (PMID:9467006, PMID:14512969, PMID:18407550). The IBA annotation appropriately captures this conserved developmental role.

Summary: IEA annotation for structural molecule activity. CRYAA is a major structural protein of the eye lens contributing to transparency and refractive index (UniProt FUNCTION section). While correct, the more specific child term GO:0005212 "structural constituent of eye lens" is also annotated and is more informative. This broader IEA term is acceptable as a parent term but less informative than the specific one.

Reason: While structural molecule activity is correct for CRYAA as a major structural lens protein, the more specific child term GO:0005212 "structural constituent of eye lens" is already annotated and is more informative. Modifying to the specific term for consistency with the IDA annotation review of the same GO term.

Proposed replacements: structural constituent of eye lens

Summary: IEA annotation for structural constituent of eye lens. This is one of the most appropriate terms for CRYAA. As alphaA-crystallin, it is the most abundant soluble protein in the lens, contributing to transparency and refractive index (UniProt FUNCTION: "Contributes to the transparency and refractive index of the lens"). The IEA mapping is accurate and captures a core function.

Reason: CRYAA is the defining structural constituent of the eye lens. This is a core function annotation. UniProt documents that CRYAA contributes to transparency and refractive index of the lens, confirmed by the fact that mutations cause cataracts (PMID:9467006, PMID:18407550).

Summary: IEA annotation for nucleus based on UniProt subcellular location mapping. UniProt lists nucleus as a confirmed subcellular location, noting that CRYAA translocates to the nucleus during heat shock and resides in SC35 speckles (PMID:19464326). The IEA mapping is correct and consistent with the IBA and IDA annotations for the same term. However, nuclear localization is stress-induced and secondary, not constitutive.

Reason: The IEA mapping from UniProt subcellular location is accurate. Nuclear localization is supported by PMID:19464326 showing heat-shock-induced translocation to SC35 speckles. However, this is a stress-dependent secondary localization, not the primary constitutive location of CRYAA. Marked as KEEP_AS_NON_CORE for consistency with the IBA and IDA annotations for the same GO term.

Summary: IEA annotation for cytoplasm. Cytoplasmic localization is the primary constitutive location of CRYAA, confirmed by multiple IDA-level studies (PMID:19464326, PMID:29259299, PMID:19503744, PMID:14752512, etc.) and UniProt subcellular location annotation. The IEA mapping is correct.

Reason: Cytoplasm is the primary constitutive localization of CRYAA, well supported by multiple independent studies. The IEA is consistent with IBA and IDA annotations.

Summary: IEA annotation for visual perception based on UniProt keyword mapping. CRYAA is essential for lens transparency, and mutations cause cataracts that impair vision (PMID:9467006, PMID:14512969). However, CRYAA does not directly participate in the visual perception signaling pathway (phototransduction). Rather, it maintains the structural integrity of the lens through which light passes. This term is somewhat over-annotated as it implies a role in visual signaling rather than lens maintenance.

Reason: While CRYAA is essential for lens transparency and mutations cause cataracts that impair vision, the term "visual perception" implies involvement in the phototransduction/visual signaling pathway. CRYAA contributes to vision indirectly by maintaining lens structure, not by participating in the perception process itself. GO:0002088 "lens development in camera-type eye" and GO:0005212 "structural constituent of eye lens" better capture the actual role.

Summary: IEA annotation for metal ion binding from UniProt keyword mapping. UniProt documents that CRYAA binds zinc ions, with the zinc-binding motif created from residues of 3 different molecules (His-100, Glu-102 from one molecule; His-107 and His-154 from additional molecules) to create tetrahedral coordination geometry (PMID:22890888). Inter-subunit bridging via zinc ions enhances oligomer stability. This is a legitimate but generic annotation; a more specific term like "zinc ion binding" (GO:0008270) would be more informative.

Reason: Metal ion binding is too generic. The specific metal bound is zinc, which is important for oligomer stability (PMID:22890888 via UniProt). GO:0008270 "zinc ion binding" would be more informative.

Proposed replacements: zinc ion binding

Summary: IPI annotation for protein binding from PMID:11700327. This study used a mammalian two-hybrid system to detect interactions between alphaA-crystallin and other lens crystallins (alphaB-, betaB2-, and gammaC-crystallin) and Hsp27 in HeLa cells. The interactions between alpha- crystallins and beta/gamma-crystallins were about one-third the intensity of alphaA-alphaB interactions. While the interactions are real and functionally relevant (crystallin-crystallin interactions maintain lens transparency), "protein binding" is uninformative.

Reason: "Protein binding" is too vague. The actual interactions detected are crystallin-crystallin interactions, not direct evidence for a specific molecular-function replacement. CRYAA holdase activity is supported by aggregation-suppression assays elsewhere; this two-hybrid interaction record should not be used to infer a chaperone term.

Summary: IPI annotation for protein binding from PMID:12601044. This study used the mammalian two-hybrid system to show that cataract-causing R116C alphaA-crystallin mutant has altered protein-protein interactions with crystallins. The wild-type CRYAA interactions with betaB2- and gammaC-crystallin decreased in the R116C mutant, while interactions with alphaB-crystallin and Hsp27 increased. "Protein binding" is uninformative for the functional context.

Reason: "Protein binding" is too generic. The crystallin-crystallin interactions documented are functionally relevant to lens protein organization, but altered two-hybrid interactions in a cataract mutant are not direct evidence for a chaperone molecular-function replacement. Keep the mechanistic context without treating generic protein binding as a core function.

Summary: IPI annotation for protein binding from PMID:19651604. This study analyzed quaternary structures of alpha-crystallins and showed that alphaA-crystallin forms defined oligomers of 24 subunits as well as smaller oligomers and large clusters. The protein-protein interactions documented here are oligomerization (homo- and hetero-oligomerization with CRYAB), which is intrinsic to the structural and chaperone function. "Protein binding" is uninformative.

Reason: "Protein binding" is too vague. The interactions are homo-oligomerization and hetero- oligomerization relevant to CRYAA structural and chaperone function. GO:0042802 "identical protein binding" is already annotated from this reference, which is more specific.

Proposed replacements: identical protein binding

Summary: IPI annotation for protein binding from PMID:22085609, which studied the F71L mutant of alphaA-crystallin. This study examined structural and functional properties of the cataract- causing mutant. The interactions documented are likely crystallin oligomerization relevant to the chaperone function. "Protein binding" is uninformative.

Reason: "Protein binding" is too generic. The interactions documented in this paper are crystallin self-interactions (oligomerization) relevant to chaperone function. GO:0042802 "identical protein binding" is already annotated from this reference.

Proposed replacements: identical protein binding

Summary: IPI annotation for protein binding from PMID:22153508, which studied the polyhedral architecture of alphaB-crystallin. While primarily focused on CRYAB, interactions with CRYAA may have been detected in this study. "Protein binding" is uninformative and should be replaced with more specific terms for the alpha-crystallin hetero-oligomerization.

Reason: "Protein binding" is too vague. The context is alpha-crystallin oligomerization. A more specific term describing hetero-oligomerization or structural complex formation would be more informative, but GO:0042802 should not be used here because this row is not specific to CRYAA self-interaction.

Summary: IPI annotation for protein binding from PMID:23188086. This study investigated binding determinants of alphaB-crystallin, specifically recognition of the IxI motif. The IxI motif is important for sHSP oligomerization and substrate interaction. Interactions with CRYAA would be in the context of hetero-oligomerization. "Protein binding" is uninformative.

Reason: "Protein binding" is too vague. The context is sHSP subunit interactions via the IxI motif, relevant to hetero-oligomerization with CRYAB rather than identical CRYAA self-binding.

Summary: IPI annotation for protein binding from PMID:25416956 (proteome-scale interactome map). This is a high-throughput study. "Protein binding" from high-throughput interactome studies is uninformative and lacks the context of specific functional interactions.

Reason: "Protein binding" from high-throughput interactome mapping is too generic. CRYAA is known to form homo- and hetero-oligomers, but this broad interactome row is not specific enough to justify replacing the annotation with identical protein binding.

Summary: IPI annotation for protein binding from PMID:25910212 (macromolecular interaction perturbations in human genetic disorders). This is a high-throughput study examining how disease-associated mutations perturb protein interactions. "Protein binding" from such studies is uninformative.

Reason: "Protein binding" from a high-throughput interaction perturbation study does not tell us anything specific about CRYAA function. The actual molecular functions (chaperone holdase, structural lens protein, oligomerization) are better captured by other annotations.

Summary: IPI annotation for protein binding from PMID:29892012 (interactome perturbation framework for developmental disorders). This is a high-throughput study. "Protein binding" is uninformative.

Reason: "Protein binding" from a high-throughput interactome perturbation study is too generic to be informative about CRYAA function.

Summary: IPI annotation for protein binding from PMID:31515488 (disruption of protein interactions by genetic variants). This is a high-throughput study examining interaction perturbations. "Protein binding" is uninformative.

Reason: "Protein binding" from high-throughput variant effect mapping is too generic.

Summary: IPI annotation for protein binding from PMID:32296183 (reference map of human binary protein interactome). This is a high-throughput interactome mapping study. "Protein binding" is uninformative. GO:0042802 is already annotated from this same reference.

Reason: "Protein binding" is too vague. Although GO:0042802 "identical protein binding" is annotated separately from this reference, this broad interactome protein-binding row should not itself be converted to a self-interaction assertion.

Summary: IPI annotation for protein binding from PMID:32814053 (interactome mapping of neurodegenerative disease proteins). This is a high-throughput study. "Protein binding" is uninformative.

Reason: "Protein binding" from a high-throughput neurodegenerative disease interactome study is too generic. The finding that CRYAA interacts with disease-associated proteins could be relevant to its chaperone holdase function but "protein binding" does not capture this.

Summary: IPI annotation for protein binding from PMID:33961781 (dual proteome-scale networks for cell-specific interactome). This is a high-throughput study. "Protein binding" is uninformative.

Reason: "Protein binding" from a high-throughput proteome-scale interactome study is too generic.

Summary: IPI annotation for identical protein binding from PMID:12601044. This study demonstrated alphaA-crystallin self-interaction using a mammalian two-hybrid system, showing that the R116C cataract mutant had altered self-interaction. Self-interaction (homo-oligomerization) is a core property of CRYAA -- it forms homodimers and homotetramers as building blocks of larger homo-oligomers (UniProt). This annotation appropriately captures the self-interaction property.

Reason: CRYAA homo-oligomerization is a core structural property essential for both its lens structural role and chaperone function. The R116C cataract mutation alters this self- interaction (PMID:12601044), confirming functional importance.

Summary: IPI annotation for identical protein binding from PMID:19651604. This study demonstrated that alphaA-crystallin forms defined oligomers of 24 subunits as well as smaller oligomers and large clusters using biophysical methods and electron microscopy. The self-interaction (homo-oligomerization) is directly demonstrated.

Reason: CRYAA homo-oligomerization is directly demonstrated by biophysical analysis showing 24-subunit oligomers (PMID:19651604). This is a core property of the protein.

Summary: IPI annotation for identical protein binding from PMID:22085609. The F71L cataract-causing mutant study examined structural and functional properties including oligomerization. Self-interaction of CRYAA is documented.

Reason: Self-interaction (homo-oligomerization) is a core property of CRYAA confirmed by multiple studies. The F71L mutant study provides additional evidence.

Summary: IPI annotation for identical protein binding from PMID:25416956 (proteome-scale interactome map). CRYAA-CRYAA self-interaction is documented in this high-throughput study. UniProt lists 12 experiments supporting CRYAA self-interaction (IntAct). While from a high- throughput study, the self-interaction is well validated by targeted studies.

Reason: CRYAA homo-oligomerization is well documented. The high-throughput detection confirms targeted studies. UniProt lists 12 experiments for CRYAA-CRYAA interaction.

Summary: IPI annotation for identical protein binding from PMID:32296183 (reference map of human binary protein interactome). CRYAA self-interaction detected in this systematic study. Consistent with extensive prior evidence for homo-oligomerization.

Reason: CRYAA homo-oligomerization is a well-established core property, and this high-throughput study provides consistent confirmatory evidence.

Summary: IDA annotation for nucleoplasm based on curation of immunofluorescence data (GO_REF:0000052). PMID:19464326 showed that CRYAA (HSPB4) translocates to the nucleus during heat shock and resides in SC35 speckles, which are nuclear sub-structures within the nucleoplasm. The nucleoplasm localization is consistent with this stress-dependent translocation.

Reason: Nucleoplasm localization is supported by evidence of heat-shock-induced translocation to SC35 speckles (PMID:19464326). However, this is a stress-dependent secondary localization, not the primary constitutive location.

Summary: IDA annotation for cytosol based on curation of immunofluorescence data (GO_REF:0000052). CRYAA is a soluble cytoplasmic protein that exists free in the cytosol as oligomeric complexes. Multiple studies show uniform cytoplasmic distribution (PMID:29259299, PMID:19503744). Cytosol is a more specific sub-compartment of cytoplasm and is consistent with the known biology.

Reason: Cytosol is the primary constitutive location of CRYAA, where it exists as soluble oligomeric complexes. Consistent with multiple IDA studies showing cytoplasmic distribution.

Summary: IDA annotation for cytoplasm from PMID:29259299. This study used fluorescence microscopy of GFP-tagged wild-type alphaA-crystallin in human lens epithelial cells and showed that wild-type CRYAA was equally distributed in the cytoplasm. The mutant (p.116_118del) accumulated at the nuclear peripheral membrane.

Reason: Direct fluorescence microscopy evidence showing wild-type CRYAA is uniformly distributed in the cytoplasm of human lens epithelial cells (PMID:29259299).

Summary: IDA annotation for cytoplasm from PMID:26004348. This study on congenital cataract families included subcellular localization analysis. UniProt lists this as a supporting reference for cytoplasmic localization.

Reason: Cytoplasmic localization supported by localization data from PMID:26004348, consistent with the primary constitutive location of CRYAA.

Summary: IDA annotation for cytoplasm from PMID:19503744. This study on the R12C alphaA-crystallin mutation characterized subcellular localization. UniProt lists this as supporting evidence for cytoplasmic localization.

Reason: Cytoplasmic localization is documented as the primary location in this study of the R12C mutant, consistent with multiple other studies.

Summary: IDA annotation for cytoplasm from PMID:30340470. This study on a novel CRYAA mutation (congenital cataract and microphthalmia) included subcellular localization analysis. UniProt lists this as supporting evidence for cytoplasmic localization.

Reason: Cytoplasmic localization confirmed by this cataract mutation study, consistent with the known primary location of CRYAA.

Summary: IDA annotation for structural molecule activity from PMID:16303126. This study showed that alpha-crystallins (alphaA- and alphaB-) suppress aggregation of the cataract-causing T5P gammaC-crystallin mutant both in vitro and in transfected cells. The study demonstrated a dual role: increasing solubility and reducing aggregate size. However, "structural molecule activity" is not the best descriptor for this chaperone function. The paper is really about chaperone holdase activity, not structural function per se. The more specific term GO:0005212 "structural constituent of eye lens" better captures the structural role.

Reason: The evidence in PMID:16303126 is about chaperone holdase activity (suppressing aggregation of T5P gammaC-crystallin), not structural molecule activity per se. An in-situ holdase activity term would better capture the molecular function demonstrated, but none exists: GO:0140309 is carrier-specific (escorting between cellular components, per go-ontology#30552) and GO:0044183 is foldase-specific. For the true structural role, GO:0005212 is already annotated.

Proposed replacements: holdase chaperone activity

Summary: IDA annotation for protein-containing complex from PMID:16303126. CRYAA forms large oligomeric complexes of approximately 540 kDa (PMID:8943244). UniProt documents that CRYAA forms heteropolymers (3 CRYAA : 1 CRYAB), homodimers, homotetramers, and larger homo-oligomers. It is also part of a complex with BFSP1 and BFSP2 for lens intermediate filament formation. The protein-containing complex annotation is correct but very generic.

Reason: CRYAA forms large homo- and hetero-oligomeric complexes as a core feature of its biology. The annotation is correct, though generic. The oligomeric complex is essential for both structural and chaperone functions.

Summary: IPI annotation for identical protein binding from PMID:16303126. This study used both in vitro sedimentation assays and cell transfection to demonstrate that alpha-crystallins form complexes. CRYAA homo-oligomerization is well established and essential for function.

Reason: CRYAA homo-oligomerization is a core property demonstrated in this study and many others. The self-interaction is essential for forming the large oligomeric complexes required for both structural and chaperone functions.

Summary: IPI annotation for protein binding from PMID:12235146. This study investigated the role of C-terminal extensions of alphaA- and alphaB-crystallins by domain swapping, demonstrating that the C-terminal extension plays a crucial role in structure and chaperone activity. The protein-protein interactions are alphaA-alphaB hetero-oligomerization. "Protein binding" is too generic.

Reason: "Protein binding" is uninformative. The interactions documented are alpha-crystallin subunit interactions (hetero-oligomerization) relevant to chaperone function. An in-situ holdase activity term would better capture the molecular function demonstrated by the chaperone activity assays used, but none exists: GO:0140309 is carrier-specific (per go-ontology#30552) and GO:0044183 is foldase-specific. GO:0042802 "identical protein binding" already captures the hetero-oligomerization aspect.

Proposed replacements: holdase chaperone activity

Summary: IMP annotation for protein stabilization from PMID:12235146. This study showed that domain-swapped chimeras of alphaA and alphaB crystallins have dramatically altered chaperone-like activity -- the chimeric alphaB with CRYAA C-terminal extension showed enhanced chaperone activity while the chimeric alphaA with CRYAB C-terminal extension almost lost activity. This demonstrates that CRYAA contributes to protein stabilization (preventing aggregation of destabilized proteins). GO:0050821 "protein stabilization" is defined as "any process involved in maintaining the structure and integrity of a protein and preventing it from degradation or aggregation." This is an excellent fit for CRYAA holdase activity.

Reason: GO:0050821 "protein stabilization" accurately describes the core holdase function of CRYAA -- preventing aggregation and maintaining protein integrity. The evidence from PMID:12235146 demonstrates that the C-terminal extension of CRYAA is crucial for this function. This is one of the most appropriate BP terms for the CRYAA holdase chaperone activity.

Summary: IDA annotation for nucleus from PMID:19464326. The cached abstract excerpt available here is about HSPB7, so explicit text support is taken from the UniProt CRYAA subcellular-location statement, which records heat-shock-induced nuclear translocation to SC35 splicing speckles. While not the primary constitutive localization, the stress-induced nuclear translocation is experimentally documented in the UniProt record.

Reason: Nuclear localization is stress-induced (heat shock) rather than constitutive. CRYAA resides in SC35 speckles under stress conditions. This is a secondary localization.

Summary: IDA annotation for cytoplasm from PMID:19464326. This study used confocal microscopy to characterize subcellular localization of HSPB family members and showed CRYAA (HSPB4) in the cytoplasm under basal conditions.

Reason: Cytoplasmic localization is directly demonstrated by confocal microscopy in PMID:19464326, consistent with the primary constitutive location of CRYAA.

Summary: IPI annotation for protein binding from PMID:14752512. This study demonstrated by GST pulldown and coimmunoprecipitation that alphaA-crystallin binds to pro-apoptotic Bax and Bcl-X(S) proteins both in vitro and in vivo, preventing their translocation to mitochondria during staurosporine-induced apoptosis. The R116C cataract mutant showed much weaker affinity. "Protein binding" is uninformative for this specific anti-apoptotic binding activity.

Reason: "Protein binding" is too generic. The specific interaction is binding to pro-apoptotic Bax and Bcl-X(S) to sequester them and prevent apoptosis. This is better captured by the negative regulation of apoptotic process annotation (GO:0043066) already present. No single MF term perfectly captures anti-apoptotic binding, but the IPI evidence supports the GO:0043066 BP annotation.

Summary: NOT annotation for protein refolding (ISS from bovine ortholog via GO_REF:0000024). This is a critical annotation that explicitly states CRYAA does NOT perform protein refolding. CRYAA is a holdase chaperone that prevents aggregation of denatured proteins but does not actively refold them. This is consistent with all experimental evidence: PMID:8943244 describes the chaperone-like activity as "the ability to suppress nonspecific aggregation" with no mention of refolding capability. The negated GOA assertion is therefore consistent with CRYAA being a holdase without foldase activity. This NOT annotation should be retained because it records the biologically important distinction between CRYAA holdase activity and active protein refolding.

Reason: The negated annotation correctly states that CRYAA does not perform active protein refolding. Retaining this NOT assertion is useful because it prevents propagation of a foldase interpretation and because it rules out GO:0044183 "protein folding chaperone" as the replacement for the obsolete GO:0051082.

Summary: GO:0051082 "unfolded protein binding" is now formally obsolete (go-ontology#30962). This IPI annotation is based on PMID:8943244, which demonstrated that recombinant human alphaA-crystallin (CRYAA) suppresses heat-induced aggregation of aldose reductase (UniProtKB:P07320, the WITH/FROM interactor) and singlet-oxygen-induced aggregation of gamma-crystallin in stoichiometric amounts. The paper explicitly defines this as "chaperone-like activity" meaning the ability to suppress nonspecific aggregation of destabilized proteins. The C-terminal truncation mutant (R157STOP) showed markedly reduced chaperone-like activity despite preserved secondary structure, confirming that the C-terminal region is essential for the holdase function. The experimental evidence clearly supports that CRYAA binds partially denatured proteins and prevents their aggregation, which is holdase (not foldase) activity. No adequate MF replacement term yet exists: GO:0140309 is carrier-specific (escorting between cellular components, per go-ontology#30552) and GO:0044183 is foldase-specific, so the row is modified to the proposed holdase chaperone activity term (NTR).

Reason: GO:0051082 is now formally obsolete. The experimental evidence from PMID:8943244 clearly demonstrates holdase chaperone activity -- CRYAA suppresses aggregation of heat-denatured aldose reductase and singlet-oxygen-damaged gamma-crystallin. This is not "unfolded protein binding" in a passive sense; it is active suppression of aggregation. The IPI evidence (with aldose reductase UniProtKB:P07320 as interactor) is strong. Neither obsoletion-replacement candidate fits: GO:0044183 requires assisting folding, which CRYAA does not do, and GO:0140309 requires escorting the client to a destination, which CRYAA does not do. Replace with the proposed holdase chaperone activity term (NTR). GO:0050821 "protein stabilization" is also relevant as a BP term.

Proposed replacements: holdase chaperone activity

Summary: IMP annotation for visual perception from PMID:9467006. This study identified the R116C missense mutation in CRYAA as causing autosomal dominant congenital cataract (ADCC) in a family. The logic is that mutation in CRYAA causes cataract, which impairs visual perception. However, CRYAA does not directly participate in phototransduction or visual signaling. It maintains lens transparency, which is a prerequisite for vision but not part of the perception pathway itself. The IMP evidence connects CRYAA to visual impairment through lens opacity, not through a direct role in visual perception.

Reason: While CRYAA mutations cause cataracts that impair vision (PMID:9467006), CRYAA does not participate in the visual perception signaling pathway. It maintains lens structural integrity. GO:0002088 "lens development in camera-type eye" and GO:0005212 "structural constituent of eye lens" better capture the role. Marking as over-annotated because the connection to visual perception is indirect.

Summary: IMP annotation for negative regulation of apoptotic process from PMID:14512969. This study identified the R49C mutation in CRYAA and showed that the R49C mutant protein failed to protect lens epithelial cells from staurosporine-induced apoptotic cell death, whereas wild-type CRYAA did protect. The IMP logic is that loss-of-function mutation leads to failure to suppress apoptosis, implicating wild-type CRYAA in negative regulation of apoptosis. This is well-supported anti-apoptotic activity, consistent with the IBA and IDA annotations to the same term.

Reason: Anti-apoptotic function is well demonstrated by mutant phenotype: the R49C mutant fails to protect from staurosporine-induced apoptosis (PMID:14512969). This is a secondary function compared to the core structural and chaperone holdase roles.

Summary: GO:0051082 "unfolded protein binding" is now formally obsolete (go-ontology#30962). This IMP annotation is based on PMID:18407550, which identified the R116H (c.346G>A) mutation in CRYAA as the cause of autosomal dominant congenital cataract in a large Chinese family. The mutant phenotype evidence is that the R116H mutant protein showed loss of chaperone activity in the DTT-induced insulin aggregation assay, increased hydrophobicity, and increased binding affinity to lysozyme. The logic is: mutation in CRYAA causes loss of chaperone (holdase) activity, leading to cataract, therefore wild-type CRYAA enables this chaperone function. This is valid IMP evidence for chaperone holdase activity. However, the term "unfolded protein binding" does not accurately capture the functional consequence measured, which is suppression of protein aggregation (holdase activity). No adequate replacement exists yet -- GO:0140309 is carrier-specific and GO:0044183 is foldase-specific -- so the row is modified to the proposed holdase chaperone activity term (NTR).

Reason: GO:0051082 is now formally obsolete. The IMP evidence from PMID:18407550 is based on the R116H cataract-causing mutation showing loss of chaperone activity in DTT-induced insulin aggregation assay. This demonstrates the functional importance of CRYAA holdase activity but the term "unfolded protein binding" mischaracterizes the function. Neither obsoletion-replacement candidate is correct: GO:0140309 is carrier-specific (escorting between cellular components, per go-ontology#30552) and GO:0044183 requires assisting folding. Replace with the proposed holdase chaperone activity term (NTR).

Proposed replacements: holdase chaperone activity

Summary: IDA annotation for cytoplasm from PMID:14752512. This study demonstrated that alphaA- and alphaB-crystallins reside in the cytosol and prevent the translocation of pro-apoptotic Bax and Bcl-X(S) from cytosol to mitochondria during staurosporine-induced apoptosis. The abstract states that "alpha-crystallins prevent the translocation of Bax and Bcl-X(S) from cytosol into mitochondria during staurosporine-induced apoptosis" (PMID:14752512), implying CRYAA co-localizes with these targets in the cytoplasm. This is consistent with multiple other IDA studies confirming cytoplasmic localization (PMID:29259299, PMID:19464326, PMID:19503744, PMID:26004348, PMID:30340470).

Reason: Cytoplasmic localization is the primary constitutive location of CRYAA. PMID:14752512 demonstrates CRYAA in the cytosol where it interacts with and sequesters Bax and Bcl-X(S). This is consistent with all other localization studies and represents a core annotation.

Summary: IDA annotation for negative regulation of intracellular transport from PMID:14752512. This study demonstrated using GST pulldown assays and coimmunoprecipitation that alphaA- and alphaB-crystallins bind to pro-apoptotic Bax and Bcl-X(S) both in vitro and in vivo, and "prevent the translocation of Bax and Bcl-X(S) from cytosol into mitochondria during staurosporine-induced apoptosis" (PMID:14752512). This is a specific form of negative regulation of intracellular transport -- CRYAA sequesters these proteins in the cytosol, preventing their mitochondrial translocation. The cataract-causing R116C mutant shows "much weaker affinity to Bax and Bcl-X(S)" (PMID:14752512), further supporting that wild-type CRYAA actively prevents this transport. This is a secondary, non-core function tied to the anti-apoptotic role rather than the primary structural or chaperone holdase functions.

Reason: The annotation is well supported by direct experimental evidence from PMID:14752512 showing that CRYAA prevents translocation of Bax and Bcl-X(S) from cytosol to mitochondria. However, this is a secondary function tied to the anti-apoptotic role, not the core structural or chaperone holdase activity of CRYAA. Marked as non-core for consistency with the anti-apoptotic annotations (GO:0043066).

Summary: IDA annotation for negative regulation of apoptotic process from PMID:14752512. This study directly demonstrated that "Human alphaA- and alphaB-crystallins prevent staurosporine-induced apoptosis through interactions with members of the Bcl-2 family" (PMID:14752512). Using GST pulldown and coimmunoprecipitation, the authors showed that alpha-crystallins bind to pro-apoptotic Bax and Bcl-X(S) both in vitro and in vivo, sequestering them in the cytosol and preventing their translocation to mitochondria. As a result, "alpha-crystallins preserve the integrity of mitochondria, restrict release of cytochrome c, repress activation of caspase-3 and block degradation of PARP" (PMID:14752512). The anti-apoptotic function was confirmed in human lens epithelial cells, ARPE-19 cells, and H9c2 cells under staurosporine, etoposide, or sorbitol treatment. The R116C cataract mutant showed much weaker affinity to Bax and Bcl-X(S), consistent with loss of anti-apoptotic activity contributing to cataract pathology. This is a well- documented secondary function of CRYAA beyond its primary structural and chaperone roles.

Reason: Anti-apoptotic activity of CRYAA is directly demonstrated by IDA evidence in PMID:14752512 using multiple experimental approaches (GST pulldown, coimmunoprecipitation, functional apoptosis assays in multiple cell types). However, this is a secondary function compared to the core structural lens protein and chaperone holdase roles. Marked as non-core for consistency with the IBA and IMP annotations for the same GO term.

Summary: IMP annotation for visual perception from PMID:14512969. This study identified the R49C missense mutation in CRYAA as causing autosomal dominant "nuclear" cataract in a four- generation family. The mutant protein "was abnormally localized to the nucleus and failed to protect from staurosporine-induced apoptotic cell death" (PMID:14512969). The logic connecting CRYAA to visual perception is that the R49C mutation causes cataract (lens opacity), which impairs vision. However, CRYAA does not directly participate in the visual perception signaling pathway (phototransduction). It maintains lens structural integrity and transparency, which is a prerequisite for light transmission but not part of the perception process itself. GO:0002088 "lens development in camera-type eye" and GO:0005212 "structural constituent of eye lens" more accurately capture the actual role. This is the same over-annotation issue identified for the IEA (GO_REF:0000043) and IMP (PMID:9467006) visual perception annotations already reviewed.

Reason: While the R49C CRYAA mutation causes autosomal dominant cataract that impairs vision (PMID:14512969), CRYAA does not participate in the visual perception signaling pathway. It maintains lens transparency through its structural and chaperone roles. The connection to visual perception is indirect -- through lens opacity rather than involvement in phototransduction. GO:0002088 "lens development in camera-type eye" and GO:0005212 "structural constituent of eye lens" better capture the actual function. Consistent with the assessment of the other visual perception annotations for CRYAA.