LRIT1 review notes

Identity and architecture

Human LRIT1 (Q9P2V4) is a 623-aa type-I single-pass membrane glycoprotein. The
reviewed UniProt record maps a cleaved signal peptide, an extracellular
LRR–Ig-like–fibronectin-type-III region, one transmembrane helix, and a short
cytoplasmic tail. UniProt curates no alternative protein products. LRIT1 belongs
to the gene-specific PANTHER subfamily PTHR45842:SF9 together with mouse Q8K099
and rat Q9JMH2; the broad PTHR45842 parent also contains functionally distinct
LRFN/SALM proteins and is not a safe functional-transfer unit.

The original rat Pal paper cloned the human homolog but functionally studied rat
Pal. Its abstract describes a "putative type I transmembrane protein" and reports
that rat "Pal immunoreactivity was distributed diffusely on the disk membrane in
the lamellar regions." It also states only that "The human homolog of Pal was
mapped to chromosome 10q23.2–23.3 using fluorescence in situ hybridization."
PMID:10777785. The same paper tested expressed rat Pal in HeLa cells using
GRP78 as an ER marker: "To confirm the subcellular localization of Pal, double
staining of Pal and GRP78 was performed." PMID:10777785. These species and
assay boundaries are essential: ER and outer-segment evidence is rat-derived,
not direct endogenous-human evidence.

Photoreceptor synapse evidence

Two adjacent 2018 mouse studies establish complementary LRIT1 biology.

• One study reports that Lrit1 "regulates the synaptic connection between cone
  photoreceptors" and that loss causes "an impairment of signal transmission"
  PMID:29590622. It also reports that "Lrit1 interacts with Frmpd2, a
  photoreceptor scaffold protein" PMID:29590622. The local cache is
  abstract-only, so the detailed experimental annotations defer to the curator
  and are not over-interpreted.
• The second study provides full-text native-retina and reconstituted interaction
  evidence: "Together, these findings establish LRIT1 as a binding partner of
  mGluR6." It localizes the protein to the retina synapse and concludes that it
  is "prominently present in the cone synaptic cleft." PMID:29590623. Mouse
  knockout changes cone-bipolar response sensitivity, background adaptation,
  and temporally demanding visual performance.

The two knockout studies differ in structural emphasis. PMID:29590622 reports
aberrant cone-pedicle morphology and selective connection defects, whereas
PMID:29590623 reports altered gain and adaptation without grossly altered
photoreceptor-synapse architecture. This is not treated as a contradiction in
the shared molecular function: both support LRIT1 at the cone synapse and an
LRIT1–mGluR6 axis, while the detailed structural consequence may depend on the
allele, assay, or analysis.

Curation boundaries

• The human core is an explicit exact-ortholog transfer from mouse Q8K099; direct
  human retinal physiology has not been demonstrated.
• G protein-coupled glutamate receptor binding is supported by mGluR6
  co-immunoprecipitation. LRIT1 is not itself a glutamate receptor, kinase, or
  signaling enzyme.
• Regulation of synapse assembly is used for the LRIT1–mGluR6/FRMPD2 organizer
  role and cone-connectivity phenotype. Synaptic membrane adhesion is not asserted
  because direct membrane-attachment evidence is absent, and no permanently stable
  ternary LRIT1–mGluR6–FRMPD2 complex is implied.
• Modulation of chemical synaptic transmission is direction-neutral because
  Lrit1 loss raises dark-adapted cone-synapse sensitivity but impairs background
  adaptation; a single positive or negative direction would collapse distinct
  illumination states.
• Dendrite is marked over-annotated/source-weak because Lrit1 transcript in
  bipolar cells and protein puncta apposed to dendritic tips do not establish
  LRIT1 protein within dendrites.
• ER membrane is retained as a real but non-core rat/heterologous context. The
  relationship among ER, rat outer-segment disk, and mouse synaptic pools is an
  open trafficking/localization question.
• No LRIT3, ELFN1, NYX, or LRFN/SALM function is transferred solely from shared
  LRR–Ig-like architecture.