Tlr12 (mouse, UniProt Q6QNU9, MGI:3045221) — curation notes

These notes rest on the UniProt record, the GOA rows and the cached publications;
the Falcon deep-research report arrived after the review was written and is
cross-checked at the end of this file.

Naming: this entry is the receptor the literature calls "TLR11"

Both UniProt entries of the pair carry the caution "There is some confusion
regarding the nomenclature of this gene. In the literature, Tlr12 is frequently
referred to as Tlr11 and vice-versa"
(file:mouse/Tlr12/Tlr12-uniprot.txt). The direction can be fixed from sequence
data in the cached full texts:

• Q6QNU9 (MGI Tlr12, this gene) starts MGRYWLLPGL ..., RefSeq NP_991392,
  signal peptide 1..21.
• Q6R5P0 (MGI Tlr11) starts MPRMERHQFC ..., RefSeq NP_991388, signal 1..30.

Cloning primers published for the receptors called TLR11 encode the Q6QNU9
N-terminus: PMID:21097503 and
PMID:24078692 (ATG GGC CGC TAC TGG = M-G-R-Y-W),
while the primer for the receptor called TLR12 encodes the Q6R5P0 N-terminus
PMID:24078692. Koblansky's "TLR12" open reading
frame is PMID:23246311,
again the other entry. MGI records the widely used knockout allele as
Tlr12 (MGI:3045751) with the note that every reference identifies it as a
Tlr11 allele while all associated sequence information indicates an allele of
Tlr12 — consistent with the same inversion. UniProt likewise cites PMID:15001781
under this entry for FUNCTION, TISSUE SPECIFICITY and DISRUPTION PHENOTYPE.

So the literature on "TLR11" — uropathogenic bacteria (Zhang 2004), Toxoplasma
profilin (Yarovinsky 2005, Pifer 2011), flagellin (Mathur 2013) — is about this
gene product.

Established biology of Q6QNU9

• Ligand-restricted innate receptor, not responsive to the standard TLR agonist
  panel: PMID:15001781
  specifically to uropathogenic bacteria; knockouts are
  PMID:15001781.
• The bacterial ligand was later identified as flagellin:
  PMID:23101627
  and PMID:23101627.
• It is also the receptor for apicomplexan profilin:
  PMID:15860593;
  loss impairs the anti-Toxoplasma response
  PMID:19050265.
• It is an intracellular receptor that depends on UNC93B1:
  PMID:21097503
  intracellular receptor residing in the endoplasmic reticulum, and
  PMID:23290966.
• Of the pair, this entry is the MyD88-recruiting member:
  PMID:24078692.

Annotation decisions (summary)

• GO:0002224 (IBA) accepted; the numbered children GO:0034170/GO:0034174 are named
  after the literature receptors and cannot be assigned until GO and MGI settle the
  inversion (raised as a question, identically to the paralogue's review).
• GO:0038023 (IBA) modified to GO:0038187 pattern recognition receptor activity:
  two chemically defined PAMPs (flagellin, parasite profilin) are bound directly.
• GO:0005886 plasma membrane (IBA) modified to endosome — this is an
  ER/endolysosomal, UNC93B1-dependent receptor; the donor set of the PAINT node is
  surface TLRs.
• GO:0042832 defense response to protozoan (IMP, PMID:19050265, allele MGI:3045751)
  accepted: allele and paper both belong to this gene on the evidence above.
• NEW GO:0009617 response to bacterium proposed, the counterpart of the row sitting
  on the paralogue: the knockout of this allele is susceptible to uropathogenic
  bacteria and the receptor binds bacterial flagellin directly.
• Generic IEA rows (GO:0007165, GO:0016020, GO:0051707) kept as non-core.

Deep-research cross-check (2026-09-30)

Compared Tlr12-deep-research-falcon.md against the finished review, and also read the
paralogue's report, because both reports are written around the literature names and are
therefore inverted relative to the database symbols. The Tlr12 report is about
literature "TLR12" = Q6R5P0 = MGI Tlr11, and the Tlr11 report is about literature
"TLR11" = Q6QNU9 = this entry. Material was mapped accordingly before comparison.

Agreement (from the Tlr11 report, which is this gene): pattern-recognition receptor with
two protein ligands, Toxoplasma profilin and bacterial flagellin from uropathogenic
E. coli and Salmonella; intracellular endosomal/endolysosomal rather than surface
receptor; UNC93B1-dependent trafficking; MyD88-dependent signalling with this receptor as
the MyD88-recruiting member of the pair; IL-12 as the principal output; protection against
uropathogenic bacteria and against Salmonella entry into Peyer's patches; epithelial
expression in intestine, lung and skin; no functional human counterpart.

Additions taken up, each verified in a primary paper:

• A fourth, independent identity check on the naming inversion, plus direct
  two-ligand binding data. The report cites a 2016 paper that the review had not used;
  fetched as publications/PMID_26859749.md. It is unambiguously about this entry:
  PMID:26859749,
  matching Q6QNU9's UniProt features (SIGNAL 1..21, TOPO_DOM 22..709) and not Q6R5P0's
  1..30 signal. It shows one receptor binding both PAMPs through different ectodomain
  regions: PMID:26859749
  Added as supported_by on the GO:0038023 -> GO:0038187 MODIFY row.
• Compartment nuance for the location call. Flagellin recognition is placed in an
  acidic endolysosomal compartment
  PMID:26859749,
  which supports the proposed GO:0005768 endosome; but profilin binding needs the uncleaved
  receptor and is seen at neutral pH, so that event may occur before cleavage, possibly at
  the surface or in early endosomes. Recorded in the GO:0005886 MODIFY reason; the proposed
  replacement term is unchanged, and the existing suggested question about compartments
  already covers it.
• The pair is not rodent-restricted. Verified with
  PMID:37874499
  and PMID:23246311.
  description corrected; human absence unchanged.
• The transcriptional route is not simply NF-kappa-B.
  PMID:24078692
  and PMID:24078692.
  The description, the GO:0002224 summary and the core-function description no longer
  assert NF-kappa-B activation; a suggested question records the cell-type dependence.

Not taken up: the microneme protein MIC3 as a second Toxoplasma ligand (single 2023
report, no IL-12 induction, candidate only) and expression in testicular germ cells
(descriptive immunohistochemistry with no function), neither sufficient for a GO
annotation.

Actions: unchanged for every existing annotation and for the proposed GO:0009617. Changes
were confined to description, two reason/summary fields, three new supported_by
quotes, two new references and one new suggested question. The Falcon file itself is not
cited as annotation support, because its gene-name mapping is inverted; the validator
warning about unused deep research is accepted for that reason.

2026-10-12 Ortholog/family audit (no change)

GOA refreshed via just fetch-gene mouse Tlr12: unmodified, all 7 GOA rows already covered; the 8th review row (GO:0009617 response to bacterium) is a proposed NEW annotation not in GOA, as expected. Status already COMPLETE.

Ortholog consistency: no human counterpart (TLR12 absent in human); no non-mouse TLR12 reviews exist, so nothing to reconcile cross-species.

Family consistency (vs Tlr11, Tlr13, human TLR3/7/8/9): consistent. GO:0002224 IBA ACCEPT, GO:0038023 IBA -> MODIFY to GO:0038187, GO:0038187 core MF, and the IEA generics (GO:0007165, GO:0016020, GO:0051707) KEEP_AS_NON_CORE all match Tlr11/Tlr13 and the family convention. Plasma-membrane IBA MODIFY -> GO:0005768 endosome matches Tlr11 (Tlr13 uses GO:0010008 endosome membrane because it has an experimental endosome-membrane row).

Heterodimer: consistent with Tlr11. Both assert no complex CC term and describe the obligate heterodimer in prose/core_functions, this entry supplying the MyD88-recruiting function PMID:24078692. Nomenclature inversion (literature TLR11 = this entry Q6QNU9) handled identically: the UPEC/flagellin function (GO:0009617 NEW, GO:0038187 flagellin support) is placed here and left UNDECIDED on Tlr11, consistent with the allele evidence (MGI:3045751 = Tlr12) and UniProt's citation of PMID:15001781 under Q6QNU9. No defects; outcome no_change.