Forty GOA rows were reviewed using gene-specific primary evidence. PMID:10859168(https://pubmed.ncbi.nlm.nih.gov/10859168/) establishes the unfused-myoblast phenotype and membrane localization; PMID:19270174(https://pubmed.ncbi.nlm.nih.gov/19270174/) identifies Sns as the ligand of Kirre/Roughest; PMID:18971929(https://pubmed.ncbi.nlm.nih.gov/18971929/) and PMID:19515699(https://pubmed.ncbi.nlm.nih.gov/19515699/) establish the nephrocyte diaphragm function. PMID:24485456(https://pubmed.ncbi.nlm.nih.gov/24485456/) directly measures Drosophila adhesion heterocomplexes despite its worm-focused title. Binding to Mec2 and Unc-5 is kept noncore; Roughest binding is specified as cell adhesion molecule binding. Abstract-only evidence leaves homophilic adhesion and the specific actin-focus result unresolved, without rejecting the experiments.
The sequence analysis finds that selected Q0E9F2 isoform B contains all 1479 residues of RefSeqA identically, plus 63 residues around the membrane-entry region. The insert has its own predicted transmembrane segment. This is a concrete topology concern, not absence of a separate assay for every isoform. Gene-level annotations are reviewed as gene-level evidence; the two broad developmental ProtNLM claims remain UNC for this exact longer sequence pending topology/functional assessment. Membrane localization remains supported and is LSP. These findings do not imply wrong predictor input or prove the longer isoform inactive.
The selected frozen UniProt record is retained in sns-uniprot-source.json, with the complete original prediction metadata in sns-predictions-source.json. It maps the exact accession to this FlyBase gene; current sequence identity does not establish the historical predictor input. Gene-level experiments support conserved functions unless a relevant isoform difference is established. Falcon research was requested with perplexity-lite fallback alongside publication caching. Primary publications and sequence observations, rather than AI summaries or ARBA assertions, support the curated conclusions.
The wrapper timed out and its fallback reported insufficient quota, but the original Falcon job later produced a complete report. The report and primary sources were inspected; no duplicate provider submission was made. The report correctly summarizes gene-level adhesion/fusion and nephrocyte functions, and cautions that foundational constructs are not isoform-B-specific. The exact hydrophobic insertion remains established by the local analysis; report statements about the standard single-pass model do not resolve its topology.
ProtNLM2 files were not consulted. All 40 GOA rows were re-examined against the cached primary literature
and the PANTHER/PAINT placement.
Identity. Q0E9F2 is the unreviewed TrEMBL entry for FlyBase isoform B (CG33141-RB, 1542 aa). The
gene has no Swiss-Prot entry; the other UniProt products are A1Z7J1 (isoform A, 1479 aa), A0A0B4LEH1
(isoform C, 1492 aa) and Q9V4Y0 (SNS, 1482 aa), all in PTHR11640:SF136. Per the existing
sequence analysis, isoform B contains all 1479 residues of isoform A plus
a 63-residue insertion (1060–1122) with an extra predicted TM segment (1089–1112). All Ig/FN3 domains and
the shared cytoplasmic tail are retained. Q0E9F2 is the entry that carries every gene-level GOA row, so it is
the de facto annotated product of the gene and not a fragment. Gene-level experiments are reviewed here.
Evolutionary placement. PANTHER PTHR11640:SF136 (NEPHRIN) holds Sns, its paralog Hbs, human/mouse/rat
nephrin (NPHS1) and C. elegans SYG-2. The partners Kirre/Duf and Rst are in SF31 (Kirre/Neph1-type).
The IBA rows come from PTN002267852, with sns itself among the donors. That is expected and not circular.
Family review interpro/panther/PTHR11640/PTHR11640-review.yaml also places Q0E9F2 in SF136, so the placement
is the same. One minor wording point: its SF136 description reuses the family-level heterogeneity text
("extends beyond nephrin to kin-of-irre proteins"). That describes the whole PTHR11640 family and not SF136,
because Kirre/Rst sit in SF31.
Changes from the previous review
- GO:0001751 compound eye photoreceptor cell differentiation (NAS, PMID:20599904): KEEP_AS_NON_CORE → MODIFY
to GO:0001745 compound eye morphogenesis. The paper is about ommatidial separation by preferential adhesion,
not photoreceptor differentiation.
- Two bare GO:0005515 rows (Mec-2 Y2H, Unc-5 interactome): KEEP_AS_NON_CORE → REMOVE under the protein-binding
policy. The interactions themselves are not disputed. The Rst rows remain MODIFY → cell adhesion molecule binding.
- GO:0016020 membrane: ACCEPT → MODIFY to plasma membrane, which the gene already carries.
- GO:0009653 / GO:0030154 / GO:0160108 (ARBA): KEEP_AS_NON_CORE → TOO_GENERAL. OLS ancestry confirms each is
entailed by retained specific terms (compound eye/chaeta morphogenesis, myoblast fusion, garland nephrocyte
differentiation, larval visceral muscle development).
- GO:0016202 regulation of striated muscle tissue development and GO:1901739 regulation of myoblast fusion
(ComplexPortal NAS): MODIFY → GO:0007520 myoblast fusion. Sns does part of the work of fusion (it is the FCM
ligand of Duf/Rst) and has no shown regulatory role.
- GO:0030036 actin cytoskeleton organization (IMP, PMID:18003739): UNDECIDED → KEEP_AS_NON_CORE. Full text
was not retrievable (PMC publisher restriction). The abstract says known fusion components regulate the
actin foci, and the reconstitution work cited in the Falcon report supports an Sns–WASP/SCAR link. The
Arp2/3 machinery carries out the polymerization, so the row is not core.
- GO:0007156 homophilic cell-cell adhesion (IGI with kirre): stays UNDECIDED. The IGI partner and the known
Sns–Kirre trans binding suggest heterophilic adhesion between homotypic cells, but the full text is not
available, so the curator is not overruled.
- Rows citing PMID:18971929, PMID:38878170 and PMID:15955848 now quote those papers directly.
- The description was rewritten around nephrin-subfamily placement. Status set to COMPLETE.
Relative for next review. Human NPHS1 (nephrin, O60500) is in the same PANTHER subfamily SF136 and is
named as the Sns ortholog (PMID:19515699 title). It is the best-characterised member, and the repo has no
human NPHS1 review. The paralog hbs (same SF136) already has a complete review.
Modules. modules/dmel_irm_myoblast_adhesion.yaml (kirre_sns and rst_sns complexes) includes Q0E9F2.