ZDHHC23

Summary: Golgi membrane is the best-supported site of ZDHHC23 activity.

Reason: The only direct localization data for ZDHHC23 (tagged protein in HEK293 cells) show predominant colocalization with the Golgi marker GM130 and partial overlap with the TGN marker TGN38; UniProt curators record Golgi membrane and TGN membrane on the basis of this paper. ZDHHC23 is a polytopic membrane protein, so Golgi membrane rather than Golgi lumen or Golgi apparatus is the appropriate term.

Summary: Bare protein binding from a high-throughput yeast two-hybrid map (partner GLIS1) carries no functional information.

Reason: Per the generic protein-binding policy, GO:0005515 is uninformative. The ZDHHC23-GLIS1 pair comes from the HuRI binary interactome screen; GLIS1 is a nuclear zinc-finger transcription factor with no established link to ZDHHC23 activity, and no more informative molecular function (e.g. a substrate- or adaptor-binding term) is supported by that paper. Removal does not imply the reported interaction is false.

Summary: ER as a site of activity is a family-level PAINT inference; the ZDHHC23-specific data place the enzyme predominantly at the Golgi/TGN.

Reason: This IBA derives from the deep DHHC node PTN000525256. The only experimental localization of ZDHHC23 (tagged protein in HEK293 cells) shows predominant Golgi (GM130) and partial TGN colocalization; the authors place dynamic BK channel palmitoylation at the TGN, and BK channels (not ZDHHC23) were the protein co-assayed with an ER marker. ZDHHC23 transits the ER as a newly synthesized membrane protein, but there is no evidence that it acts there, so ER is likely an over-propagation for this subfamily. Not REMOVE because ER residence has not been tested and excluded. ORGANELLES check (2026-11-06). Most human entries in this IBA's with/from list (ZDHHC1, 4, 6, 9, 11, 12, 14 and 22) have curated UniProt ER-membrane locations, most taken from the Ohno et al. 2006 survey of tagged human DHHC proteins (abstract-only in the cache). They include ZDHHC22, ZDHHC23's partner in BK-channel S-acylation, which UniProt records in both ER and Golgi membrane; current PANTHER places ZDHHC22 in a different family (PTHR12246), so it is not a sister of the target within this tree. ZDHHC23 itself is not among the seeds. The deep node carries both ER and Golgi assertions, and the only data on ZDHHC23 itself show Golgi/TGN; UniProt's ZDHHC23 entry records only Golgi and TGN membrane. UNDECIDED is the close alternative until ZDHHC23 is co-stained with an ER marker. On the is_active_in qualifier, no ZDHHC23 S-acyltransferase activity is shown in the ER. ZDHHC23 is a multi-pass membrane protein, so any ER pool would be ER membrane, but with no ER evidence on the protein no ER membrane row is added.

Summary: ZDHHC23 is a Golgi-resident integral membrane enzyme; the Golgi apparatus term is correct but less specific than the supported Golgi membrane location.

Reason: Overexpressed zDHHC23 colocalizes strongly with the cis/medial Golgi marker GM130 and more weakly with TGN38 in HEK293 cells, and the protein has six predicted transmembrane helices with a cytoplasmic DHHC domain, so it acts on the cytoplasmic face of Golgi membranes. The more specific term GO:0000139 Golgi membrane, which the gene already carries (IEA, GO_REF:0000044), is preferred. The IBA (PAINT node PTN000525256, a deep DHHC-family node) and the UniProt SubCell IEA rows agree and are refined the same way.

Proposed replacements: Golgi membrane

Summary: ZDHHC23 is a Golgi-resident integral membrane enzyme; the Golgi apparatus term is correct but less specific than the supported Golgi membrane location.

Reason: Overexpressed zDHHC23 colocalizes strongly with the cis/medial Golgi marker GM130 and more weakly with TGN38 in HEK293 cells, and the protein has six predicted transmembrane helices with a cytoplasmic DHHC domain, so it acts on the cytoplasmic face of Golgi membranes. The more specific term GO:0000139 Golgi membrane, which the gene already carries (IEA, GO_REF:0000044), is preferred. The IBA (PAINT node PTN000525256, a deep DHHC-family node) and the UniProt SubCell IEA rows agree and are refined the same way.

Proposed replacements: Golgi membrane

Summary: Palmitoylation-dependent targeting of substrates to membranes is a family-level role; for ZDHHC23 it is reflected in KCNMA1 delivery to the plasma membrane.

Reason: The IBA comes from the deep DHHC node PTN000525256 (donors yeast Erf2-type and rice DHHC proteins). ZDHHC23 adds the lipid to the substrate, which is the signal that promotes membrane delivery, so it does do part of the work. For ZDHHC23 the evidence specifically concerns exit of BK channel alpha subunits from the TGN to the plasma membrane, captured by GO:0072659; this broader family-level term is kept but is not the central description. Mouse genetic data show the trafficking effect is conditional (lost when the beta1 subunit is present), which also argues against treating it as core.

Summary: InterPro-based palmitoyltransferase activity is correct but less specific than the demonstrated protein-cysteine S-palmitoyltransferase activity.

Reason: The DHHC domain (residues 259-309) carries an intact DHHC motif (D286-H287-H288-C289), with C289 as the S-palmitoyl cysteine intermediate, and ZDHHC23 S-acylates cysteines in the BK channel S0-S1 loop. Replace with GO:0019706, which the gene already carries with experimental and IBA support.

Proposed replacements: protein-cysteine S-palmitoyltransferase activity

Summary: ZDHHC23 palmitoylates the BK channel S0-S1 loop; knockdown and overexpression in HEK293 cells and Zdhhc23 deletion in mouse smooth muscle reduce substrate S-acylation.

Reason: ZDHHC23 catalyses the step itself, so the process term is appropriate. siRNA knockdown in an imaging screen reduced S0-S1 loop palmitoylation-dependent surface delivery, and independent genetic deletion of mouse Zdhhc23 (same PANTHER subfamily SF475) reduced endogenous BK alpha-subunit S-acylation.

Summary: ZDHHC23 palmitoylates cysteines of the BK channel S0-S1 loop.

Reason: The BK channel study identifies ZDHHC23 (with ZDHHC22) as the acyltransferase for the S0-S1 loop cysteines; this is the core molecular function. Mouse Zdhhc23 knockout data corroborate it for the endogenous substrate.

Summary: Protein-cysteine S-palmitoyltransferase activity is inherited from the DHHC ancestor and ZDHHC23 retains the catalytic DHHC motif.

Reason: PAINT placed this function at PTN000525256, covering the DHHC family; ZDHHC23 retains the catalytic DHHC cysteine and has its own experimental support (it appears among the donors, which is expected rather than circular).

Summary: The Rhea/EC mapping (RHEA:36683, EC 2.3.1.225) matches the curated catalytic activity of ZDHHC23.

Reason: UniProt records the protein-cysteine palmitoylation reaction with experimental evidence from PMID:22399288; the IEA mapping to GO:0019706 is correct.

Summary: ZDHHC23-dependent palmitoylation promotes delivery of BK channel alpha subunits to the plasma membrane.

Reason: In HEK293 cells, overexpression of zDHHC23 increased surface expression of the S0-S1 reporter and knockdown reduced it. This is the substrate-specific outcome of the enzyme's catalysis. The IBA comes from PTN008583751, a ZDHHC23-specific node seeded by human ZDHHC23 and mouse Zdhhc23, so the propagation is confined to the orthologous clade. Caveat: in mouse vascular smooth muscle, where beta1 subunits are present, Zdhhc23 deletion reduced BK current without changing surface expression, so the trafficking role is context dependent.

Summary: ZDHHC23-dependent palmitoylation promotes delivery of BK channel alpha subunits to the plasma membrane.

Reason: In HEK293 cells, overexpression of zDHHC23 increased surface expression of the S0-S1 reporter and knockdown reduced it. This is the substrate-specific outcome of the enzyme's catalysis. The IBA comes from PTN008583751, a ZDHHC23-specific node seeded by human ZDHHC23 and mouse Zdhhc23, so the propagation is confined to the orthologous clade. Caveat: in mouse vascular smooth muscle, where beta1 subunits are present, Zdhhc23 deletion reduced BK current without changing surface expression, so the trafficking role is context dependent.