FBXW7-AS1 / B0L3A2 (DEspR, "DEAR") — curation notes

Session 2026-10-08 (MICROPROTEINS Tier 4). Contested-peptide case: UniProt
B0L3A2 (DESPR_HUMAN, PE1, 85 aa, single TM) sits on the locus HGNC calls
FBXW7-AS1, "RNA, long non-coding", antisense to FBXW7 exon 5.

Locus and existence

• The ORF is interrupted at codon 14 by a TGA in the reference genome. UniProt
  records RNA editing as the resolution: "The nonsense codon (UGA) at position 14
  is modified to a sense codon (UGI), which is the equivalent of UGG. ADAR is
  responsible of FBXW7-AS1 expression and RNA-editing" (UniProt RNA EDITING, cited
  to PMID:33853558). The originating laboratory reports ADAR1 CRISPR knockout
  abolishing DEspR protein [PMID:33853558, "we show that ADAR1-dependent DEspR
  expression, via CRISPR/cas9-knockout studies"].
• Protein-level evidence is antibody-based: anti-DEspR mAb pull-down from isolated
  cell membranes gives a 17.5 kDa glycosylated band that shrinks after PNGase F,
  and this is explicitly not corroborated by mass spectrometry — "We note that
  although DEspR was detected consistently in denatured conditions via Western blot
  analysis of membrane protein pull-down products using 5g12e8 and 6g8g7 anti-DEspR
  mAbs, MS did not detect DEspR-specific peptides" PMID:27301377. The authors
  attribute non-detection to the known difficulty of small glycosylated single-TM
  proteins.
• Cell-surface presentation is reproducible across cell types and laboratories by
  flow cytometry: "To determine the cell surface expression of DEspR on circulating
  immune cells, we used flow cytometric analysis" [PMID:34789851 — Maine Medical
  Center, with the originating lab as co-authors], and "Blocking anti-DEspR murine
  precursor and humanized antibodies bind cell-surface DEspR, internalize, and
  co-translocate with gal1/gal3 nuclear shuttling proteins to the nucleus"
  PMID:33853558.

So: the locus designation (lncRNA) and the protein entry (PE1) are in open
conflict, and the protein-existence evidence is antibody-only. Plasma-membrane
localisation is nevertheless the best-supported single claim about this entry.

The receptor claim, and why it is contested

• Origin: a rat brain cDNA isolated by antisense-oligonucleotide screening under
  "molecular recognition theory", reported as a dual ET-1/angiotensin II receptor
  with "a single predicted transmembrane region with distinct ET-1 and AngII
  putative binding domains" and "ET-1- and AngII-induced coupling to a Ca2+
  mobilizing transduction system" PMID:9508787.
• The angiotensin half did not hold outside Dahl S rats: "mouse Dear does not bind
  ANG II, similar" to Dahl R rat Dear PMID:16293765, and UniProt carries a CAUTION
  to that effect. The ligand pair was reassigned to ET-1 plus the VEGF-A signal
  peptide (VEGFsp).
• Human evidence (PMID:24465725) is: anti-DEspR mAb staining of DEspR+ Cos1
  transfectants displaced by ligand — "7c5b2-immunostaining of DEspR+Cos1 cell
  transfectants was effectively blocked by the antigenic peptide and by ET1" — plus
  competition FACS with VEGFsp and ET1, and ligand-specific phosphoproteomics:
  "DEspR-mediated signaling activated phosphoproteins implicated in angiogenesis
  with some overlap with VEGF-VEGFR2 signaling pathways (FAK, PKCa, ERK1/2), but
  also distinct from VEGF-VEGFR2". The authors' own summary is deliberately modest:
  "Collectively, these findings support the functionality of VEGFsp- and ET1-binding
  to DEspR."
• The endothelin field has not taken the receptor up. Watts: "The pharmacology and
  physiology of DEAR has not been developed" PMID:19907001. Thorin & Clozel: "All
  these data come from one group of scientists, and the physiology and pharmacology
  of DEAR has not been studied in depth" PMID:21081213. The Davenport et al.
  IUPHAR review states "Current evidence only supports the existence of two
  subtypes, ETA and ETB, according to NC-IUPHAR nomenclature" and "No further
  Family A GPCRs have been identified that might bind ET peptides"
  PMID:26956245 — note this is a statement about Family A GPCRs, so it is not
  a direct refutation of an 85-aa single-pass protein, but it shows the field does
  not count DEspR among ET receptors.

Per CLAUDE.md ("do not overrule curators from incomplete evidence") and the Tier 4
addendum on contested peptides, the experimental rows are not removed. They are
instead corrected where the GO term definition does not fit the protein, which
is an ontology question rather than a re-reading of the experiments:

• GO:0004962 endothelin receptor activity is defined as "Combining with endothelin
  and transmitting the signal across the membrane by activating an associated
  G-protein; promotes the exchange of GDP for GTP on the alpha subunit of a
  heterotrimeric G-protein complex" (QuickGO). No G-protein coupling has ever been
  shown for this 85-aa single-pass protein — the 1998 rat work showed Ca2+
  mobilisation, the 2014 human work phosphoproteomics. The ET-1-binding,
  signal-transmitting claim is better expressed by
  GO:0004888 transmembrane signaling receptor activity plus
  GO:0017046 peptide hormone binding.
• GO:0086100 endothelin receptor signaling pathway is likewise "A G
  protein-coupled receptor signaling pathway initiated by endothelin binding" →
  GO:0007166 cell surface receptor signaling pathway.
• GO:0038085 vascular endothelial growth factor binding is "Binding to a vascular
  endothelial growth factor". The reported ligand is the VEGF-A signal peptide,
  and UniProt states the protein "Does not bind the VEGFA mature protein". So the
  term overstates the claim; GO:0005048 signal sequence receptor activity (with
  GO:0042277 peptide binding as the generic fallback) matches what was assayed.
• GO:0038084 vascular endothelial growth factor signaling pathway requires VEGF
  binding its receptor; the paper's own phosphoproteomics is explicitly "distinct
  from VEGF-VEGFR2". → GO:0007166.

Protein-binding row

The GO:0005515 IPI row cites PMID:17446437 with EDN1 (P05305) and VEGFA (P15692)
as partners. That paper is a human SNP-haplotype association study — "Here we
report the association of human ATP1A1 (P<0.000005) and Dear (P<0.03)" with
hypertension — and the cached record is abstract-only. Bare protein binding carries
no functional information (CLAUDE.md), and the two ligand interactions it points at
are already covered by the ligand-binding rows, so the row is removed; removal does
not assert the interactions are false.

No new annotations proposed

Mouse knockout lethality with "impaired angiogenesis, dysregulated neuroepithelial
development" PMID:16293765 is a necessity phenotype in another species, with the
human protein's own existence contested; it does not pass CLAUDE.md's participation
test for a NEW angiogenesis process term, and no ISS row exists to review. Nothing
is proposed as NEW.

Core function

One activity is recorded: a cell-surface, ET-1/VEGFsp-binding signalling receptor at
the plasma membrane, with the directionality and mechanism flagged as unresolved.
protein binding is not used.