GOA changes since the original review (234 rows in total):
- 26 new rows were seeded as PENDING. 11 are donor-split duplicates of reviewed rows: the same
term and evidence with a different WITH/FROM (MGI:105384 mouse vs UniProtKB:P11142 human), or
a different qualifier. The other 15 are new terms or new evidence. All 26 are resolved: 16
ACCEPT, 10 KEEP_AS_NON_CORE.
- GO:0051082 unfolded protein binding: the IEA, ISO and IDA (PMID:11133993) rows are retired.
QuickGO confirms the term is obsolete ("this binding term should be replaced by an activity
term such as protein folding chaperone (GO:0044183) or unfolded protein holdase activity
(GO:0140309)"). GOA now carries GO:0044183 by IDA from the same paper, plus IEA (ARBA) and
ISO (mouse). The earlier reviews had already recommended MODIFY to GO:0044183, and the new
rows are ACCEPTed PMID:11133993. Handled the same way as the
St13 and Uggt1 re-reviews. No proposed_new_terms entry is needed, because the core activity
has a GO term (GO:0044183).
- 24 rows are retired in total. Their reviews are kept, and a retirement sentence was added to
each.
Action changes (from -> to):
- GO:0005515 protein binding (8 IPI rows, MARK_AS_OVER_ANNOTATED is no longer allowed):
- PMID:15708368 (SGTA) -> MODIFY to GO:0051087 protein-folding chaperone binding
PMID:15708368
- PMID:9528774 (Hsp40/Hop/Hip; retired) -> MODIFY to GO:0051087 PMID:9528774
- PMID:23159318, 18307834, 19457116, 17877381, 10198213 and 12514190 -> REMOVE. The abstracts
do not support a more specific MF, and removal does not mean the interactions are false.
PMID:12514190's abstract does not mention Hsc70 at all.
- GO:0046777 protein autophosphorylation (IDA PMID:8420978, retired): REMOVE ->
MARK_AS_OVER_ANNOTATED. The old review wrongly said the paper has no autophosphorylation
data. In fact PMID:8420978, but at <1% stoichiometry and uncoupled from ATPase activity.
- GO:1904593 prostaglandin binding (IPI PMID:21445266): REMOVE -> MARK_AS_OVER_ANNOTATED. The
full text does identify Hspa8 as a 15d-PGJ2 target, but the "binding" is covalent Michael
addition to free thiols PMID:21445266.
- GO:0001664 GPCR binding (ISO from human P11142): REMOVE -> KEEP_AS_NON_CORE. The human source
is a real IPI with GPR37/Pael-R PMID:12150907, and rat has its own direct evidence for GPCR binding
PMID:10866672. propagation_review updated.
- GO:0048018 receptor ligand activity (ISO from human): REMOVE -> MARK_AS_OVER_ANNOTATED. The old
propagation comment called the human IDA a likely misannotation, which is unfounded
PMID:17785435. It is a single, cell-type-specific study, so it is
marked over-annotated for rat rather than removed.
- GO:0005886 plasma membrane: IEA MARK_AS_OVER_ANNOTATED -> KEEP_AS_NON_CORE, and the retired IBA
row the same, for consistency with the new ISS row. The old rationale ("not the core site")
supports non-core status, not over-annotation. UniProt lists Cell membrane, and rat hsc73
assembles with A1 receptors at the cell membrane (PMID:10866672).
- IEP rows for processes other than "response to X" -> MARK_AS_OVER_ANNOTATED, because a change
in expression is not participation (same reasoning as the St13 re-review):
- kidney development, cerebellum development, forebrain development, skeletal muscle tissue
development, estrous cycle and G1/S transition.
- For skeletal muscle development, the abstract itself says HSC73 is not developmentally
regulated PMID:12909603.
- forebrain development (PMID:9878698) comes from an adult-brain localization study.
- GO:1990832 slow axonal transport and GO:0008088 axo-dendritic transport (IEP PMID:11933046):
KEEP_AS_NON_CORE -> MARK_AS_OVER_ANNOTATED. Hsc73 is cargo here [PMID:11933046 "labeled in
"slow component b" of axonal transport along with the molecular chaperone Hsc73 and actin"].
- GO:0006606 protein import into nucleus (IMP PMID:8892974): KEEP_AS_NON_CORE ->
MARK_AS_OVER_ANNOTATED. hsc70 is the import substrate in this study PMID:8892974.
- GO:0046034 ATP metabolic process (ISO): KEEP_AS_NON_CORE -> MARK_AS_OVER_ANNOTATED. Using ATP
to power the chaperone cycle is already captured by ATP hydrolysis activity.
propagation_review added.
- Legacy rows: about 200 rows had boilerplate reasons and no supported_by.
Verbatim support was added to every ACCEPT and KEEP_AS_NON_CORE row, from UniProt P63018
CC-lines, the falcon deep research, or cached abstracts. 18 donor-side papers were cited and
added to references; PMID:37767704 was newly cached.
- core_functions: GO:0061738 late endosomal microautophagy was added to the CMA/cargo-selector
function, keeping the eMI ACCEPT rows consistent with the core
PMID:21238931.
- description: the closing sentence about "annotated processes" was replaced with plain biology.
Open questions:
- GO:0005102 signaling receptor binding (IPI PMID:11133993) is kept as non-core. The transferrin
receptor is bound as a chaperone client, so a curator may prefer a different framing.
- GO:0035651 AP-3 adaptor complex binding (mouse IDA PMID:19010779): the abstract does not name
Hsc70. This is deferred to the curator's full-text reading.
- GO:0019899 enzyme binding (IPI PMID:21958194, alpha-enolase) and GO:0031686 A1 adenosine
receptor binding are both kept as non-core. Neither is clearly a client versus regulatory
interaction.
Review follow-up 2026-10-17: In response to the PR #4627 review, the GO:1904593 prostaglandin
binding summary no longer says Hspa8 is modified "via Cys603". PMID:21445266 identified Hspa8 by
2D gel and MS (Table 2, spots 1-3), but the Cys603 site came only from a sequence-homology search
against IkB-kinase beta Cys179 (5/16 identity), and Hspa8 is not in the Figure 10 pull-down set.
The summary now says this, and the homology sentence is quoted. The action (MARK_AS_OVER_ANNOTATED)
is unchanged. The GO:0007519 skeletal muscle tissue development IEP reason now explains why
MARK_AS_OVER_ANNOTATED was kept rather than REMOVE: the abstract reports HSC73 rising to adult
levels by postnatal day 3, and the paper is abstract-only. The three open questions above, plus the
VMAT2 inhibition observation (PMID:19457116; candidate GO:0141110 transporter inhibitor activity,
not asserted), are now suggested_questions in the review YAML.