LRFN2, also called SALM1, is a brain-enriched type-I single-pass synaptic membrane glycoprotein with an extracellular leucine-rich-repeat, immunoglobulin-like, and fibronectin type III region and a long cytoplasmic tail ending in an ESTV PDZ-binding motif. SALM1 occupies both postsynaptic and presynaptic membranes. At postsynapses, rodent SALM1 associates with PSD-95-family scaffolds and NMDA receptors and supports postsynaptic-density maturation, glutamate-receptor surface organization, synaptic transmission, and plasticity. At presynapses, mouse SALM1 binds CASK through its PDZ-binding tail and promotes F-actin- and PIP2-dependent cis clustering of neurexin, without evidence that SALM1 binds neurexin directly. SALM1 can also form cis complexes with itself and other SALMs, but it did not mediate SALM-SALM trans-cellular adhesion in heterologous cell-aggregation assays.
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0098839 postsynaptic density membrane | IBA GO_REF:0000033 | ACCEPT | Summary: The phylogenetic annotation places LRFN2/SALM1 at the postsynaptic density membrane, consistent with direct localization of the rat Lrfn2 ortholog and the conserved synaptic-membrane architecture of this subfamily. Reason: RGD:1311831 is rat Lrfn2/SALM1 rather than a different SALM paralog; the corresponding rat protein Q460M5 currently carries multiple experimental postsynaptic-density-membrane annotations from PMID:16495444. The paired PANTHER node is therefore supported by an exact orthologous seed, making this a sound transfer of a core localization. Propagation Review Root cause: NO FAILURE CORE Sources checked: PANTHER:PTN002911877 · PANTHER ancestral node PTN002911877 SUPPORTS TRANSFER RGD:1311831 · rat Lrfn2/SALM1 SUPPORTS TRANSFER |
| GO:0009986 cell surface | IBA GO_REF:0000033 | ACCEPT | Summary: The phylogenetic annotation places LRFN2 at the cell surface, matching its type-I membrane topology and experimental cell-surface localization of the mouse Lrfn2 ortholog. Reason: MGI:MGI:1917780 is mouse Lrfn2/SALM1, and the current mouse record carries a direct cell-surface annotation from PMID:16828986. The exact orthologous source and LRFN2-subfamily PANTHER inference support this core localization without relying on another SALM paralog. Propagation Review Root cause: NO FAILURE CORE Sources checked: MGI:MGI:1917780 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER PANTHER:PTN002911877 · PANTHER ancestral node PTN002911877 SUPPORTS TRANSFER |
| GO:0016020 membrane | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location vocabulary mapping assigns the broad membrane term to LRFN2. Reason: Human LRFN2 has a signal peptide and one transmembrane helix and is independently localized to the plasma membrane. The broad membrane term is therefore correct, although less informative than its synaptic membrane localizations. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0162 · membrane controlled vocabulary SUPPORTS TRANSFER |
| GO:0045202 synapse | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location vocabulary mapping places LRFN2 at the synapse. Reason: Synaptic localization is a conserved, core property of LRFN2/SALM1 and is supported by more specific postsynaptic-density, postsynaptic, and presynaptic localization evidence from exact rodent orthologs. This broad parent location is accurate. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0258 · synapse controlled vocabulary SUPPORTS TRANSFER |
| GO:0045211 postsynaptic membrane | IEA GO_REF:0000044 | ACCEPT | Summary: UniProt subcellular-location vocabulary mapping places LRFN2 at the postsynaptic membrane. Reason: The term is consistent with direct rodent SALM1 localization at postsynapses, the postsynaptic-density-membrane IBA, and LRFN2 interaction with postsynaptic PSD-95-family scaffolds and NMDA receptors. It does not exclude the separately demonstrated presynaptic pool. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB-SubCell:SL-0219 · postsynaptic membrane controlled vocabulary SUPPORTS TRANSFER |
| GO:0009986 cell surface | IEA GO_REF:0000107 | ACCEPT | Summary: Ensembl Compara transfers cell-surface localization from mouse Lrfn2/SALM1 to human LRFN2. Reason: Q80TG9 and ENSMUSP00000047573 represent the exact mouse Lrfn2 ortholog, whose current record carries direct cell-surface evidence from PMID:16828986. Human HPA plasma-membrane localization and conserved type-I topology independently support the transfer. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER ensembl:ENSMUSP00000047573 · mouse Lrfn2 protein SUPPORTS TRANSFER |
| GO:0050804 modulation of chemical synaptic transmission | IEA GO_REF:0000107 | ACCEPT | Summary: Ensembl Compara transfers modulation of chemical synaptic transmission from mouse Lrfn2/SALM1 to human LRFN2. Reason: The donor is the exact mouse ortholog and has multiple experimental annotations to this term from PMID:28604739, where Lrfn2 loss altered hippocampal AMPA/NMDA balance, synaptic maturation, and plasticity. The broad term captures a conserved core synaptic role without importing another SALM paralog's activity. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER ensembl:ENSMUSP00000047573 · mouse Lrfn2 protein SUPPORTS TRANSFER |
| GO:0098685 Schaffer collateral - CA1 synapse | IEA GO_REF:0000107 | KEEP AS NON CORE | Summary: Ensembl Compara transfers localization/activity at the Schaffer collateral-CA1 synapse from mouse Lrfn2/SALM1. Reason: The exact mouse ortholog was tested directly at hippocampal Schaffer collateral-CA1 synapses in PMID:28604739, which underlies multiple current SynGO experimental annotations. This anatomically specific transfer is compatible with conserved mammalian hippocampal expression and synaptic function, but it is an anatomically restricted instance rather than a core location for LRFN2 across its pre- and postsynaptic roles. Propagation Review Root cause: NO FAILURE NON CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER ensembl:ENSMUSP00000047573 · mouse Lrfn2 protein SUPPORTS TRANSFER |
| GO:0098793 presynapse | IEA GO_REF:0000107 | ACCEPT | Summary: Ensembl Compara transfers presynaptic localization of mouse Lrfn2/SALM1 to the human ortholog. Reason: Although SALM proteins are often discussed as postsynaptic organizers, this is not a paralog-derived inference; the exact mouse Lrfn2 ortholog has direct presynaptic localization evidence from PMID:31368584. That study also demonstrated a presynaptic SALM1 pool and a role in neurexin clustering, so the transfer is biologically supported. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER ensembl:ENSMUSP00000047573 · mouse Lrfn2 protein SUPPORTS TRANSFER |
| GO:0098794 postsynapse | IEA GO_REF:0000107 | ACCEPT | Summary: Ensembl Compara transfers postsynaptic localization of mouse Lrfn2/SALM1 to human LRFN2. Reason: The exact mouse ortholog has direct postsynaptic localization evidence from PMID:31368584, consistent with the independent rat postsynaptic-density-membrane evidence and conserved PSD-95 binding. The parallel presynaptic annotation reflects a separately observed pool rather than a contradiction. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER ensembl:ENSMUSP00000047573 · mouse Lrfn2 protein SUPPORTS TRANSFER |
| GO:0099175 regulation of postsynapse organization | IEA GO_REF:0000107 | ACCEPT | Summary: Ensembl Compara transfers regulation of postsynapse organization from mouse Lrfn2/SALM1 to human LRFN2. Reason: The exact mouse ortholog carries direct-assay and mutant-phenotype support from PMID:28604739; loss of Lrfn2 altered postsynaptic-density and spine maturation and postsynaptic receptor composition. This is a core, subfamily-specific SALM1 role and does not rely on presynaptic-induction phenotypes from SALM3 or SALM5. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER ensembl:ENSMUSP00000047573 · mouse Lrfn2 protein SUPPORTS TRANSFER |
| GO:0005886 plasma membrane | IDA GO_REF:0000052 | ACCEPT | Summary: Human Protein Atlas immunofluorescence directly places endogenous human LRFN2 at the plasma membrane. Reason: This direct human localization agrees with LRFN2's signal peptide, single transmembrane helix, extracellular adhesion domains, and independent cell-surface and synaptic-membrane annotations. Plasma membrane is a correct core location. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-8849882 | ACCEPT | Summary: The Reactome event places human SALM1/LRFN2 in an RTN3-associated membrane complex in brain, while the GO annotation records the broader plasma-membrane location. Reason: The event reports a relatively weak RTN3 interaction for SALM1, so it should not be used to infer the stronger SALM2/SALM3 association. However, plasma-membrane residence of human LRFN2 is independently supported by HPA immunofluorescence and type-I membrane topology, making the location itself correct. Supporting Evidence: Reactome:R-HSA-8849882 In the brain, reticulon 3 (RTN3) tightly associates with SALM2 and SALM3 to form a complex, and interacts relatively weakly with SALM1 and SALM4. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-8849891 | ACCEPT | Summary: The Reactome event describes SALM1/LRFN2 binding to PSD-95-family postsynaptic scaffolds, consistent with synaptic plasma-membrane residence. Reason: SALM1 is explicitly included among SALMs1-3 that bind DLG-family PDZ scaffolds and are enriched in synaptic fractions. This supports the broad plasma-membrane location without importing the presynaptic-differentiation activity that the same record restricts to SALM3 and SALM5. Supporting Evidence: Reactome:R-HSA-8849891 SALMs 1-3 interact with the PDZ domain containing proteins PSD 95 (DLG4) and synapse associated protein 97 (SAP97 or DLG1) and SAP102 (DLG3) |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-8849900 | ACCEPT | Summary: The Reactome event models cis homo- and heteromeric complexes among SALM1-3 in the membrane, supporting plasma-membrane localization of SALM1/LRFN2. Reason: SALM1 is directly included in the cis complexes, and its established single-pass surface topology supports the location. The cis association does not establish homophilic adhesion across cell junctions and is not treated as such here. Supporting Evidence: Reactome:R-HSA-8849900 SALM1, SALM2, and SALM3 form homo- and heteromeric complexes in a cis manner. |
| GO:0005886 plasma membrane | TAS Reactome:R-HSA-8849906 | ACCEPT | Summary: The Reactome event specifically describes SALM1/LRFN2 association with NMDA receptors at early excitatory synapses, consistent with plasma- membrane residence. Reason: This event concerns SALM1 rather than another SALM paralog and describes both direct extracellular association with GRIN1 and indirect recruitment through PSD-95. These interactions occur in a synaptic membrane context, so the broader plasma-membrane annotation is sound. Supporting Evidence: Reactome:R-HSA-8849906 SALM1 can directly interact with the extracellular domain of the NR1 subunit of NMDA receptor or indirectly by binding to PSD-95, which may recruit NMDA receptor via the NR2 subunits of NMDA receptors |
| GO:0030165 PDZ domain binding | IPI PMID:34251337 Sorting nexin-27 regulates AMPA receptor trafficking through... | NEW | Summary: The human LRFN2/SALM1 C-terminal PDZ-binding motif directly binds the rat SNX27 PDZ domain; ortholog studies also support binding to PSD-95-family scaffolds and CASK. Reason: GO:0030165 exactly captures the direct interaction measured between a human LRFN2 C-terminal peptide and the recombinant rat SNX27 PDZ domain by ITC. Exact rodent-ortholog studies independently support binding of the conserved terminal motif to PSD-95-family PDZ domains and CASK. This does not place SNX27 at the postsynaptic density or assert membership in a stable obligate complex. Supporting Evidence: PMID:34251337 This established that the isolated recombinant PDZ domain of SNX27 directly bound to a synthetic peptide corresponding to the LRFN2 PDZ binding motif, S-S-E-W-V-M-E−3-S-T-V−0 with a high micromolar affinity (Kd = 1.6 µM) (Figure 2D). PMID:16828986 C-termini of Lrfn1, Lrfn2 and Lrfn4 were bound by PDZ domains of postsynaptic protein PSD95, re-distributing PSD95 to cell periphery where the Lrfn proteins were detected. PMID:31368584 Together, these data identify SALM1 as a novel binding partner of the CASK/Mint1/Lin7b presynaptic organizer complex, interacting directly with CASK, via its PDZ binding domain. |
| GO:0035255 ionotropic glutamate receptor binding | ISO PMID:16495444 A novel family of adhesion-like molecules that interacts wit... | NEW | Summary: Rat SALM1/Lrfn2 associates with NMDA-receptor subunits, including an extracellular or transmembrane interaction with the NR1 subunit. Reason: GO:0035255 captures the demonstrated NMDA-receptor binding more precisely than the parent glutamate-receptor-binding term. Q460M5 is the exact rat Lrfn2 ortholog, and the primary study separately supports association with endogenous brain NMDA receptors and NR1 interaction in a heterologous assay. Transfer to human is orthology-based rather than direct experimentation in human neurons. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q460M5 · rat Lrfn2/SALM1 SUPPORTS TRANSFER Supporting Evidence: PMID:16495444 SALM1 also interacts with the NMDA receptor NR1 subunit through its extracellular or TM1 domains. |
| GO:0097115 neurexin clustering involved in presynaptic membrane assembly | ISO PMID:31368584 SALM1 controls synapse development by promoting F-actin/PIP2... | NEW | Summary: Presynaptic mouse SALM1 promotes F-actin/PIP2-dependent cis clustering of neurexin during presynaptic membrane assembly. Reason: GO:0097115 exactly captures the measured F-actin/PIP2-dependent neurexin cis clustering at the presynaptic membrane. The exact mouse Lrfn2 ortholog localizes directly to presynapses, and depletion/rescue experiments support a causal role. The evidence does not establish direct SALM1-neurexin binding or SALM1-mediated trans homophilic adhesion. Propagation Review Root cause: NO FAILURE CORE Sources checked: UniProtKB:Q80TG9 · mouse Lrfn2/SALM1 SUPPORTS TRANSFER Supporting Evidence: PMID:31368584 Together, our data suggest that SALM1 organizes synapse development by promoting F‐actin/PIP2‐dependent cis‐oligomerization of Neurexin at the presynapse. |
Loading supporting content…
Download this section (compressed HTML)Q: Which PSD-95-family, CASK-family, sorting, and glutamate-receptor partners bind endogenous human LRFN2 at pre- versus postsynaptic membranes, and which interactions require the terminal ESTV motif?
Q: What molecular bridge connects presynaptic SALM1 and CASK to F-actin/PIP2-dependent neurexin cis clustering without direct SALM1-neurexin binding?
Q: Which developmental stage, cell type, or circuit variable explains the divergent Lrfn2 knockout effects on hippocampal plasticity and on excitatory versus inhibitory transmission?
Q: Do SALM1-containing cis homo- or heteromeric assemblies regulate surface trafficking or partner avidity while remaining distinct from the trans-cellular adhesion mechanisms of SALM4 and SALM5?
Q: Which membrane side and extracellular partner account for the LRFN2-dependent cone-to-bipolar-cell transmission phenotype, and is this retinal role conserved in humans?
Experiment: Generate isogenic human induced-pluripotent-stem-cell-derived excitatory neurons with endogenous LRFN2 knockout, ESTV deletion, and repaired controls. Use compartment-resolved proximity labeling and quantitative interactomics together with AMPA/NMDA receptor surface tracking, synaptic nanostructure imaging, and paired-neuron electrophysiology to test localization-specific rescue.
Hypothesis: Human LRFN2 uses its terminal ESTV motif to recruit distinct postsynaptic PSD-95/SNX27 and presynaptic CASK modules that jointly support synapse organization and transmission.
Type: Endogenous human-neuron structure-function analysis
Experiment: Rescue LRFN2-depleted neurons with ESTV-tail, polybasic-region, and combined mutants expressed at matched endogenous abundance. Quantify CASK binding, membrane PIP2, cortical F-actin, single-molecule neurexin cluster size, synaptic-vesicle clustering, and neurotransmitter release; separately test purified components for direct interactions.
Hypothesis: The SALM1 PDZ-binding tail and juxtamembrane polybasic region provide separable inputs that couple CASK and PIP2/F-actin to presynaptic neurexin clustering.
Type: Presynaptic mechanism dissection
Experiment: Introduce domain-selective LRFN2 substitutions that preserve surface delivery, compare them with ESTV deletion in rodent and human neurons, and measure direct NR1 binding, PSD-95/SNX27 occupancy, receptor diffusion and surface retention, postsynaptic-density morphology, and AMPA/NMDA current ratios.
Hypothesis: Postsynaptic SALM1 regulates NMDA- and AMPA-receptor organization through separable extracellular receptor association and intracellular PDZ-scaffold recruitment.
Type: Postsynaptic receptor-interface epistasis
Experiment: Perform matched conditional deletion and knock-in rescue in hippocampal excitatory neurons, hippocampal inhibitory inputs, cortical neurons, and retinal cone circuits. Use the same age windows and assays for receptor composition, synapse morphology, basal transmission, plasticity, and retinal output.
Hypothesis: The divergent mouse phenotypes reflect circuit- and developmental-stage-specific requirements rather than opposite biochemical functions of Lrfn2.
Type: Cross-circuit conditional-genetics comparison
Experiment: Map endogenous cis interfaces and stoichiometry by cross-linking mass spectrometry and single-molecule imaging, create interface-specific mutants that retain folding and surface expression, and compare effects on scaffold/receptor recruitment, synapse organization, and quantitative cell-cell adhesion assays.
Hypothesis: SALM1 cis oligomerization controls membrane organization but does not confer SALM1-mediated trans homophilic adhesion.
Type: Cis-assembly interface analysis
What is not known — curated, literature-grounded statements of the open unknowns (the inverse of core functions).
Gap: The endogenous scaffold, glutamate-receptor, and trafficking partners occupied by LRFN2 at human pre- and postsynaptic membranes, and the requirement for its terminal ESTV motif in human neurons, remain unknown.
OPEN BIOLOGY RESIDUAL_SUBGAP
What is known: Human LRFN2 is directly localized to the plasma membrane, and a human LRFN2 tail peptide directly binds SNX27. PSD-95-family, CASK, NMDA-receptor, and neuronal functional evidence otherwise comes principally from exact rodent orthologs or heterologous systems.
Significance: Establishing the endogenous human partner hierarchy is necessary to connect the conserved binding activities to physiological human synapse organization.
What would resolve it: Use endogenous compartment-resolved tagging in human induced-pluripotent-stem- cell-derived neurons, compare wild type with ESTV-deleted LRFN2, and combine quantitative interactomics, receptor tracking, synaptic imaging, and functional rescue.
Provenance (the field's own admissions):
Gap: How presynaptic SALM1 couples CASK and its juxtamembrane basic region to F-actin/PIP2-dependent neurexin cis clustering is unresolved, including whether the relevant interactions are simultaneous, transient, or indirect.
OPEN BIOLOGY RESIDUAL_SUBGAP
What is known: Mouse SALM1 binds CASK through its PDZ-binding tail and is required for neurexin clustering and presynaptic organization, but the study does not demonstrate direct SALM1-neurexin binding or a stable obligate SALM1-CASK-Mint1-Lin7b- neurexin complex.
Significance: Resolving this linkage would define SALM1's presynaptic mechanism without conflating cis regulation of neurexin with a trans-synaptic ligand interaction.
What would resolve it: Combine endogenous CASK and SALM1 perturbations with ESTV-tail and polybasic- region rescue mutants, PIP2 manipulation, single-molecule neurexin clustering, and direct binding assays using purified cytoplasmic components.
Provenance (the field's own admissions):
Gap: The direction and cellular basis of Lrfn2-dependent synaptic plasticity are not settled, and the relative contributions of excitatory versus inhibitory synapses differ between mouse models and brain regions.
OPEN BIOLOGY RESIDUAL_SUBGAP
What is known: One knockout study reported immature hippocampal synapses and enhanced LTP, whereas an independent line showed suppressed NMDAR-dependent plasticity and region-specific inhibitory or excitatory transmission defects. These outcomes should not be collapsed into a single signed regulatory effect.
Significance: A mechanistically resolved phenotype is needed before assigning a more specific positive or negative regulation term for plasticity or transmitter-specific synapse development.
What would resolve it: Compare conditional, cell-type-specific Lrfn2 loss and matched rescue alleles across hippocampal and cortical circuits using standardized development, receptor-composition, morphology, and electrophysiology assays.
Provenance (the field's own admissions):
Gap: The physiological role of SALM1-containing cis homo- and heteromeric SALM assemblies is unknown, as is whether they regulate trafficking, membrane avidity, or synaptic partner selection.
OPEN BIOLOGY RESIDUAL_SUBGAP
What is known: SALM1-3 co-immunoprecipitate and form same-cell complexes, but SALM1-3 did not form trans-cellular SALM associations in the tested assay. The evidence therefore does not support homophilic trans adhesion or a defined stable complex for human LRFN2.
Significance: Defining the function of cis assembly would explain a conserved interaction while preventing inappropriate transfer of SALM4/5 trans-association mechanisms.
What would resolve it: Determine endogenous SALM stoichiometry and cis interfaces in neurons, then test interface-specific mutants that preserve folding and surface delivery for effects on receptor/scaffold binding, trafficking, and synaptic organization.
Provenance (the field's own admissions):
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)