ERLN

Summary: Electronic mapping of the UniProt subcellular location (ER membrane, single-pass), which UniProt bases on the mouse ortholog Q3U0I6.

Reason: Correct and appropriately specific. Endoregulin is a single-TM peptide whose only known partner is the ER-resident SERCA pump; the mouse peptide colocalises with SERCA2b in reticulated ER membranes (PMID:27923914). The location is a prerequisite of its core function (inhibition of SERCA within the ER membrane). No human experimental localisation study was found, but the human peptide is functional in reconstituted membranes with SERCA and interacts with other ER-resident regulins in cells (PMID:34445594, PMID:36523160), and the TM helix and binding motif are conserved.

Summary: Curator transfer of the mouse Q3U0I6 IDA ER membrane annotation (PMID:27923914) to human endoregulin.

Reason: The donor annotation is sound (mouse ELN colocalises with SERCA2b in the ER of COS-7 cells, PMID:27923914) and the human and mouse peptides share the transmembrane helix and SERCA-binding motif. Human experimental localisation data are lacking, so this ISS is the appropriate evidence for human, and it underpins the core function.

Summary: Proposed new annotation: endoregulin binds SERCA and directly inhibits its Ca2+-ATPase activity.

Reason: Human GOA has no molecular function for ERLN, although this is the defining activity of the gene. The strongest human evidence is PMID:34445594, in which recombinant human ELN co-reconstituted with purified SERCA1a in proteoliposomes lowered the pump's maximal turnover rate. A purified two-component system shows that the peptide itself performs the inhibition, so the participation test for an enzyme-regulator function is met: ELN is the regulator, not a substrate or an indirect requirement. Mouse data corroborate it (PMID:27923914): the mouse peptide co-immunoprecipitates with SERCA2b, is competed off by PLN (same binding groove), and lowers the apparent Ca2+ affinity of SERCA3a. The two studies differ in mechanism (a Vmax effect on SERCA1a versus a KCa effect on SERCA3a) and in isoform, but both show inhibition, and a later live-cell assay found that ELN lowers ER Ca2+ stores (PMID:39921961, abstract only). Comparator check: PLN carries GO:0042030 (IDA and IBA), and SLN carries the more general GO:0004857 enzyme inhibitor activity (not an ancestor of GO:0042030 in current GO; both sit under GO:0140678 molecular function inhibitor activity). The regulins standing in the same role therefore carry this term, and its absence from ERLN reflects how recently the gene was characterised, not a convention. GO:0042030 is chosen over GO:0141110 transporter inhibitor activity for consistency with PLN and because the assay read out ATPase turnover.

Summary: Proposed new annotation: by binding SERCA, endoregulin reduces the pump's Ca2+ transport activity.

Reason: This is the process-level expression of the ATPase inhibitor activity, and endoregulin carries out the regulatory step itself by binding the pump. It is therefore a participant, not an upstream requirement. Mouse Q3U0I6 already carries this term (IDA, PMID:27923914, as acts_upstream_of_or_within), from oxalate-supported Ca2+ uptake assays. For human, the reconstitution assay with the purified human peptide (PMID:34445594) provides direct evidence. Comparator check: SLN carries this term by IDA and PLN carries its parent GO:1901894. Broader physiological processes (tissue Ca2+ signalling, epithelial function) are not proposed, because there are no loss-of-function data.