rosetteless (Salpingoeca rosetta, F2U5Y1 / PTSG_03555) — notes

Automated deep research was not available in this session (no provider API
keys configured), so there is no -deep-research-*.md file. These notes were
compiled manually from the cached full text of the primary papers.

Identity

• UniProt F2U5Y1 (TrEMBL, unreviewed), ORF PTSG_03555, EMBL EGD82922.1.
• UniProt's submitted name "Lung surfactant protein A" comes from the genome
  project's EMBL record (ECO:0000313|EMBL:EGD82922.1). It is a
  similarity-based label; nothing links this protein functionally to animal
  surfactant protein A (SFTPA1/2). The name used in the literature is
  Rosetteless (Rtls).
• The locus-to-gene mapping: the mapped mutation disrupts "a splice donor in
  the gene EGD82922" PMID:25299189.
• Domain architecture: signal peptide, two C-type lectin-like domains
  (Pfam PF00059 x2; SMART CLECT x2), Ser/Thr-rich mucin-like stretches, two
  internal repeats PMID:25299189.
• Carbohydrate binding has not been tested PMID:25299189.

Genetics

• Forward genetic screen (Levin et al. 2014). The rtls^l1 allele is a T-to-C
  change in the splice donor of intron 7, causing mis-splicing and ~25%
  protein levels. Mutants make normal chains and grow normally but never form
  rosettes PMID:25299189.
• No complementation was possible in 2014 PMID:25299189.
• Blocking antibody against Rtls inhibits rosette development in wild type
  without affecting growth PMID:25299189.
• Independent CRISPR allele (rtls^PTS1, premature termination) reproduces
  the phenotype PMID:32496191.
• Knocked out again with a selection-based pipeline (2025; abstract only in cache) PMID:41037400.

Localisation

• Secreted; in rosettes it forms a thick layer at the basal poles, filling
  the rosette interior PMID:25299189.
• In single cells and chains, weak membrane-associated patches and material
  deposited on the substrate PMID:25299189.
• The CRISPR mutant lacks secreted Rtls at the basal end PMID:32496191.
• Treated as an ECM component by later work PMID:30556809.

Relationship to jumble and couscous

• In the glycosyltransferase mutants, Rtls still reaches the basal pole but
  is not upregulated or secreted after rosette induction PMID:30556809.
  Rtls is itself predicted to be heavily glycosylated PMID:30556809.

GO modelling issues

• GOA has no annotations at all for F2U5Y1 (checked 2026-09-30).
• No GO term exists for clonal rosette/colony development in a unicellular
  organism. GO:0007275 multicellular organism development is not
  taxon-restricted away from choanoflagellates, but GO places Dictyostelium
  sorocarp development (GO:0030587) outside it (under GO:0099120
  socially cooperative development and GO:0048856 anatomical structure
  development), so using GO:0007275 for a colony would be a modelling
  decision, not an obvious fit. Proposed an NTR instead.
• MF: the protein's role looks structural (an extracellular layer joining
  cells), but no biochemical activity has been measured. GO:0005201
  extracellular matrix structural constituent is used in core_functions as
  an inference, flagged in knowledge_gaps; it is not asserted as an
  annotation.