Connexin-32 supplies the membrane subunits of connexons and the conduit of gap-junction channels. Human Cx32 expression experiments support intercellular conductance and coupling, while assembly measurements distinguish hemichannel precursors from completed channels. These are structural and channel activities of one functional unit (PMID:10234007(https://pubmed.ncbi.nlm.nih.gov/10234007/), PMID:14627639(https://pubmed.ncbi.nlm.nih.gov/14627639/)).
Disease variants can affect different steps. Trafficking defects and altered gating should be separated from reduced protein abundance. Increased S85C hemichannel opening is a variant result, not an assertion that normal Cx32 remains constitutively open (PMID:12460545(https://pubmed.ncbi.nlm.nih.gov/12460545/), PMID:11891346(https://pubmed.ncbi.nlm.nih.gov/11891346/)).
Human mutations and peripheral-nerve expression establish an important disease context. The original experiment underlying the existing nervous-system-development assertion remains unresolved because the full paper could not be recovered; mutation association alone does not settle that process claim (PMID:8266101(https://pubmed.ncbi.nlm.nih.gov/8266101/)).
The three inherited PAINT assertions were read against the actual tree, ancestral placements and selected alignment windows. They follow the AN0/PTN008334405 placement and are supported by independent human channel evidence. Donor count and target self-evidence were not treated as defects. The exact lineage and literal source data are retained in the evidence archive.
The positive HuRI interactions remain recorded, even where generic binding is functionally uninformative. Three assertions retain unresolved partner-product suffix differences. The GJB2 interaction does not by itself establish a mixed-subunit connexon; heterotypic channel compatibility is a different observation (PMID:32296183(https://pubmed.ncbi.nlm.nih.gov/32296183/)).
The canonical human protein has 283 residues. Mutant constructs, partner isoforms and historical translation bands are not additional natural GJB1 products. The original cloning and topology studies include rat junction material, and the older structural model uses truncated Cx43 density; none is presented as a direct full-length human atomic structure (PMID:2875078(https://pubmed.ncbi.nlm.nih.gov/2875078/), PMID:2460334(https://pubmed.ncbi.nlm.nih.gov/2460334/), PMID:15383278(https://pubmed.ncbi.nlm.nih.gov/15383278/)).
The source guide distinguishes original readings, cached availability, donor contexts and remaining uncertainty. It also identifies the metadata limits in the clinical bibliography.
All 44 generic HuRI protein-binding assertions are removed for lack of a specific functional interpretation. Their positive pairs and literal source identifiers remain preserved. The three partner-product discrepancies remain unresolved and are now represented as suspected issues requesting curator checking; no replacement accession or false interaction is asserted. The distinction is between the usefulness of the generic function term and the identity of the experimental product.
The nervous-system-development assertion remains undecided. The source assessment now acknowledges the relevant regeneration result while keeping normal developmental participation unresolved. The accepted channel, transport, communication and structural assembly core is unchanged.
The redundant replacement attached to the broad cell-communication term and the canonical-only alternative-products block were removed. The canonical 283-residue sequence and its verification remain in the evidence archive and the earlier notes. All original annotation fields, source bytes and four exact excerpts remain unchanged. The source guide records the current scope; the evidence archive distinguishes initial preparation outcomes from the follow-up.