LRFN4 is SALM3, one of five SALM/Lrfn synaptic adhesion molecules. The mapping is explicit in the SALM3 primary literature: “SALM3/Lrfn4” [PMID:26321637 “There are five known members in the family: SALM1/Lrfn2, SALM2/Lrfn1, SALM3/Lrfn4, SALM4/Lrfn3, and SALM5/Lrfn5.”]. This distinction is essential: SALM4 is LRFN3, not LRFN4. SALM1/LRFN2 and SALM2/LRFN1 glutamate-receptor results, and SALM4/LRFN3 cis-inhibitory results, are not evidence for LRFN4.
The primary SALM paper is PMID:16630835, SALM synaptic cell adhesion-like molecules regulate the differentiation of excitatory synapses. Its locally cached record is abstract-only and foregrounds SALM2, but it establishes the family-level PSD-95 interaction [PMID:16630835 “cell adhesion-like molecules termed SALM that interacts with the abundant postsynaptic density (PSD) protein PSD-95.”]. The reviewed human LRFN4/Q6PJG9 record attributes direct interaction with all four human PSD-95-family DLG scaffolds to this PMID [file:human/LRFN4/LRFN4-uniprot.txt “Interacts with DLG1, DLG2, DLG3 and DLG4.”]. The verified human accessions are DLG1/Q12959, DLG2/Q15700, DLG3/Q92796 and DLG4/P78352.
The same human record documents deletion of terminal LRFN4 residues 633–635 and the resulting “Missing: Loss of DLG1-, DLG3- and DLG4-binding.” [file:human/LRFN4/LRFN4-uniprot.txt]. This directly establishes a requirement for the terminal PDZ-binding tail in those three interactions. The record does not state whether this deletion also disrupts DLG2 binding, and the experiment does not establish the downstream consequence of tail loss in human neurons. Thus direct human IPI evidence supports PDZ domain binding, while the rat SALM3 study below remains bounded corroboration for neuronal PSD-95 association and postsynaptic clustering.
The strongest gene-specific mechanistic study identifies SALM3 as a postsynaptic ligand for all three LAR-family receptor protein tyrosine phosphatases. Binding requires the mini-exon B splice insert, and depletion of all three presynaptic LAR-RPTPs strongly suppresses SALM3-induced presynaptic differentiation [PMID:26321637 “In the present study, we demonstrated that SALM3 trans-synaptically interacts with all three types of presynaptic LAR-RPTPs in an alternative splicing-dependent manner.”; “In triple knockdown of LAR-RPTPs, SALM3-induced synapsin I clustering was decreased by 85.2% ± 3.0%”]. This is rodent evidence (mouse knockout and constructs; rat neuronal/synaptosomal assays), so transfer to human should be marked as orthology-based rather than described as direct human neuronal evidence.
Mouse Salm3 knockout provides in-vivo support for excitatory-synapse development: “Salm3−/− mice show a reduced excitatory synapse number but normal synaptic plasticity in the hippocampus” PMID:26321637. The paper further reports a cell-autonomous postsynaptic rescue and “substantial reductions in mEPSC frequency (~52.3%) and PSD density (~14.5%)” PMID:26321637. This supports a core role in organizing excitatory synaptic inputs, not a broad claim that SALM3 regulates all glutamate-receptor functions.
Earlier rat neuronal and biochemical work established SALM3 synaptogenic and scaffold-associated properties. SALM3 expression induces both excitatory and inhibitory presynaptic differentiation in contacting axons, the protein is enriched in synaptic fractions, and it forms a strong complex with PSD-95 [PMID:20410109 “We found that expression of the SALM family proteins SALM3 and SALM5 in nonneural and neural cells induces both excitatory and inhibitory presynaptic differentiation in contacting axons. SALM3 and SALM5 proteins are enriched in synaptic fractions, and form strong (SALM3) or weak (SALM5) complexes with postsynaptic density-95 (PSD-95)”]. Artificial aggregation of SALM3 on dendrites induces PSD-95 clustering [PMID:20410109 “Aggregation of SALM3, but not SALM5, on dendritic surfaces induces clustering of PSD-95.”]. SALM3 carries the C-terminal PDZ-binding motif characteristic of SALMs 1–3; the same study distinguishes SALM4/5, which lack this motif [PMID:20410109 “SALM4 and SALM5, unlike SALM1, SALM2, and SALM3, do not possess the C-terminal PDZ-binding motif”].
A human ectodomain interactome independently supports binding of human SALM3 to all three human LAR-RPTPs by surface plasmon resonance [PMID:32822567 “With the exception of PTPRF-SALM4, we observed binding of all LAR-PTPRs to all SALMs”]. The study also cautions that the assay reveals a spectrum of binding responses rather than proving identical physiological functions for every SALM–PTPR pair.
Mouse SALM3 structural work shows that its isolated LRR domain forms a stable antiparallel dimer; SEC–MALLS of LRR and LRR–Ig fragments supported dimeric species, and “the LRR domain is sufficient alone for stable dimer formation” PMID:32665594. The crystallized entity associated with PDB 6TL8 is the mouse SALM3 LRR fragment, not full-length human LRFN4.
For the ligand complex, SAXS of mouse SALM3 LRR–Ig with the PTPσ Ig1–3 fragment supports a 2:2 assembly, but the authors explicitly caution that the low-resolution model is “likely representing an average snapshot of a number of conformations present in solution” PMID:32665594. Thus the evidence supports stoichiometry and a flexible extracellular binding arrangement; it does not define a single rigid atomic structure of a full-length transmembrane complex.
The modeled interfaces have functional support. In heterologous synapse-formation assays, “the interface mutants targeting the SALM3-PTPσ interaction abolished synapsin I clustering,” and a dimer-interface mutant also abolished synaptogenic activity PMID:32665594. These experiments used mouse SALM3 constructs, HEK293T cells and cultured hippocampal neurons. They strengthen the conserved SALM3/LRFN4 mechanism but remain rodent/partial-construct evidence rather than direct full-length human neuronal validation.
In rat brain, SALMs 1–3 strongly co-immunoprecipitate with each other; in heterologous cells, all five SALMs can form homo- and heteromeric complexes [PMID:18227064 “In brain, we found that SALMs 1-3 strongly co-immunoprecipitated with each other, whereas SALMs 4 and 5 did not”]. The same paper found no trans associations for SALMs 1–3 [PMID:18227064 “Both SALMs 4 and 5 formed homophilic, but not heterophilic associations, whereas no trans associations were formed by the other SALMs.”]. Thus SALM3 cis complexes should not be converted into SALM3 homophilic trans-cellular adhesion.
Family-wide assays report that overexpression of each SALM enhanced neurite outgrowth and that C-terminal PDZ-binding domains of SALMs 1–3 are needed for most measured outgrowth effects [PMID:18585462 “Over-expression of each SALM resulted in enhanced neurite outgrowth, but with different phenotypes.”; “the C-terminal PDZ binding domains of SALMs 1-3 are required for most aspects of neurite outgrowth.”]. These are cultured-neuron overexpression results and do not establish a uniquely SALM3-specific in-vivo developmental process.
Reticulon 3 co-immunoprecipitated with SALMs 1–4 in brain, suggesting a possible family-level trafficking interaction [PMID:19681166 “A 19-kDa band, identified as reticulon 3C, bound to all four SALMs”]. This is supporting context, not a demonstrated core molecular function of LRFN4.
LRFN4 also has experimentally studied functions outside neurons. In human THP-1 cells and primary monocytes, expression rises with macrophage differentiation, and LRFN4 signaling regulates transendothelial migration and cell elongation through actin reorganization [PMID:21704618 “We also found that expression of LRFN4 in the monocytic cell line THP-1 and in primary monocytes was upregulated following macrophage differentiation.”; “LRFN4 signaling regulated both the transendothelial migration of THP-1 cells and the elongation of THP-1 cells via actin cytoskeleton reorganization.”].
A follow-up human-cell study places 14-3-3 and NCK1 in the LRFN4 signaling complex and Rac1 downstream of LRFN4-mediated elongation [PMID:22677168 “We found that 14-3-3 and NCK proteins complexed with LRFN4, and they were involved in LRFN4-mediated cell elongation.”; “Finally, we demonstrated that a Rac1 small GTPase was involved in LRFN4-mediated cell elongation.”]. This establishes a context-specific monocytic signaling role; it should not be silently folded into the rodent synaptic mechanism.
The BioPlex-style and HuRI binary-interactome annotations are high-throughput interaction calls. The cached narrative texts describe proteome-scale maps but do not expose the LRFN4-specific pairs in the article body; the exact pairs likely reside in supplementary datasets. They are reasonable interaction records but do not define a coherent core molecular function. In particular, the HuRI paper states that “the cellular function of most individual PPIs remains to be elucidated” PMID:32296183.