rgn

Summary: Trans-Golgi network residence is propagated from TGN38/TGOLN2, but Rgn lacks the membrane anchor and sorting signal that define that localization, and shows no detectable sequence conservation with the seeds.

Reason: PANTHER:PTN000575073 is the single annotated node of PTHR23211 and its IBD for trans-Golgi network is seeded by vertebrate TGN38/TGOLN2 proteins, which are type I membrane proteins whose trans-Golgi residence and recycling depend on a single transmembrane segment and the cytoplasmic YQRL tyrosine sorting motif. Although M9PFV8 is classified in the same PANTHER subfamily (PTHR23211:SF0), it has no transmembrane segment and no C-terminal tyrosine motif (the UniProt record carries no transmembrane feature; our hydropathy scan found none after the signal peptide), and its domain architecture (signal peptide, C-type lectin-like domain, charged disordered tail) is unlike TGN38. Local alignment of Rgn to the TGOLN2 and TGN38 seeds is confined to the compositionally biased disordered region and is not distinguishable from composition-matched shuffled controls, whereas Rgn's lectin-like domain aligns clearly to PTHR22803 C-type lectins. The target therefore lacks the determinants of the propagated compartment, and no experiment places Rgn in the trans-Golgi network. Transit of a signal-peptide protein through the Golgi would not justify an is_active_in assertion.

Summary: TGN-derived transport-vesicle residence is propagated from the TGN38 recycling cycle, which requires a transmembrane cargo-receptor architecture that Rgn lacks.

Reason: The same node (PANTHER:PTN000575073) propagates trans-Golgi network transport vesicle localization seeded by mouse TGN38 (MGI:105079). That localization reflects TGN38 cycling between the trans-Golgi network and the plasma membrane as an integral membrane protein, driven by its cytoplasmic tyrosine motif. Rgn has no transmembrane segment and no tyrosine sorting motif, its architecture is a signal peptide plus C-type lectin-like domain plus charged disordered tail, and no sequence conservation with the seeds was detected above shuffled controls. No experiment places Rgn in TGN-derived vesicles.

Summary: The C-type lectin-like domain is real, but it lacks the canonical calcium/sugar-binding motifs and no ligand has been measured, so carbohydrate binding is a fold-level over-inference.

Reason: The rule transfers carbohydrate binding from the presence of a C-type lectin-like domain. C-type lectin-like domains are a fold, and many bind proteins, lipids or nothing carbohydrate-like. In canonical Ca2+-dependent CTLDs, sugar binding uses the EPN (mannose-type) or QPD (galactose-type) motif and the WND motif at calcium site 2; the Rgn CTLD has none of these, and the UniProt PROSITE-ProRule annotation itself cautions that conserved residues are missing. No ligand assay exists for Rgn. The original description of Rgn raises Wingless glycoprotein binding only as a speculation.

Summary: The ISS donor, galactose-specific lectin-galC1 (DL1), belongs to a different PANTHER family (PTHR22803) and retains sugar-binding motifs that Rgn lacks.

Reason: The donor FBgn0016675 (lectin-galC1, UniProt Q9VIX1) is a small secreted galactose-specific C-type lectin classified in PTHR22803, whereas Rgn is classified in PTHR23211. The similarity between the two is limited to the C-type lectin-like domain (31.5% identity over ~110 residues). That relationship supports a shared fold, not a shared ligand, and Rgn lacks the QPD/EPN and WND motifs used for Ca2+-dependent sugar binding. PMID:16475980 characterises DL1, not Rgn. Same action as the IEA row for this term.

Summary: rgn is induced specifically at the blastema during imaginal-disc regeneration, which matches the genetic requirement shown in the same study.

Reason: In situ hybridisation shows rgn is off in undamaged prothoracic leg discs and is activated at the blastema after either fragmentation or ubiquitous wg expression. Taken alone, expression would be weak evidence of participation, but here it agrees with the IMP row from the same paper. Same action as the IMP row.

Summary: Reducing rgn dosage delays blastema formation in imaginal-disc regeneration, and rgn deficiency in gut progenitors impairs epithelial repair.

Reason: Two independent P-element alleles in CG6014 (rgn1, rgn2) fail to complement. Heterozygous animals delay blastema formation by about one day under both wg-induced and fragmentation-induced regeneration, and they reduce regeneration-induced transdetermination. Precise excisions of rgn1 suppress its semi-lethality and revert the transdetermination phenotype. An independent 2026 study (abstract only) reports that rgn deficiency in intestinal progenitors disrupts Notch signalling, depletes intestinal stem cells and causes intestinal atrophy during epithelial repair. GO:0042246 is as specific as the evidence allows, because GO has no imaginal-disc regeneration term and the molecular mechanism is unknown. The process term is broad and may partly reflect a regulatory role (the authors speculate that Rgn attenuates Wg signalling), so regulation terms are not proposed.