LRFN5 is SALM5, the fifth synaptic adhesion-like molecule. It is not SALM3/LRFN4,
SALM4/LRFN3, SALM2/LRFN1, or SALM1/LRFN2. The downloaded human UniProt record is
Q96NI6 and contains no ALTERNATIVE PRODUCTS stanza, so I found no curated human
LRFN5 protein isoform that should be tracked separately. The important splice variants
in the synaptic literature belong instead to the binding partners PTPRF/LAR, PTPRD,
and PTPRS.
SALM5 lacks the C-terminal PDZ-binding domain found in SALM1-3
PMID:18227064.
Accordingly, the weak biochemical association with PSD-95 must not be described as a
canonical PDZ-tail interaction. SALM5 was weakly recovered with PSD-95, but aggregated
SALM5 did not cluster PSD-95 on dendrites
PMID:20410109.
The earliest direct adhesion study used rat brain and heterologous cells. SALM5 formed
homophilic rather than heterophilic trans-cellular associations
PMID:18227064.
This is genuine SALM5 biology but should not be conflated with the later heterophilic
SALM5-LAR-RPTP synaptic bridge.
Rat-neuron mixed-culture and knockdown experiments established that SALM5 induces both
excitatory and inhibitory presynaptic differentiation
PMID:20410109.
Loss of SALM5 decreased synapse abundance and function in both classes
PMID:20410109.
The later ligand study used mouse full-length SALM5, human and mouse LAR-RPTP constructs,
rat dissociated hippocampal neurons, and rat organotypic slices. Mouse SALM5 bound all
three LAR-family receptor phosphatases
PMID:27225731.
The LAR-binding-defective SALM5 S329/S360A mutant did not rescue AMPA-receptor-mediated
synaptic transmission after SALM5 knockdown
PMID:27225731.
Thus the best-supported model is a postsynaptic SALM5 dimer engaging presynaptic
LAR-RPTPs to organize presynaptic differentiation and maintain synaptic transmission.
An independent human extracellular-domain interactome corroborated binding of SALM5 to
PTPRF/LAR, PTPRD, and PTPRS
PMID:32822567.
This study supports the human interaction but does not by itself establish directionality,
cis/trans geometry, or synaptogenic function.
Two 2018 structural papers independently resolved a 2:2 SALM5-PTPδ assembly. One used
human SALM5 with mouse PTPδ and found that SALM5 dimerization is necessary for
synaptogenic activity, as summarized directly in the paper's abstract
PMID:29348429.
The other study used human SALM5 and human PTPδ fragments and observed a central SALM5
dimer bridging two PTPδ monomers
PMID:29348579.
A dimer-disrupting human SALM5 mutation retained PTPδ binding but abolished presynaptic
induction in a rat-neuron heterologous assay
PMID:29348579.
These structures use extracellular fragments (LRR-Ig for crystallography and, in some
functional experiments, an ectodomain displayed on HEK293 cells or beads), not intact
human neurons expressing endogenous full-length human LRFN5. They directly establish the
extracellular recognition mechanism but not an LRFN5 cytoplasmic signaling pathway.
The 2016 cell-aggregation study concluded that partner mini-exon B suppresses SALM5
binding
PMID:27225731.
In contrast, purified-protein SPR and structural studies found that PTPδ mini-exon B is
favored. In the human-human study, adding MeB increased SALM5 binding by about 20- to
27-fold
PMID:29348579.
The authors explicitly attributed the discrepancy as potentially assay-dependent. The
review should preserve this unresolved conflict rather than state one splice preference
as settled in vivo.
SALM5 also has a distinct, experimentally supported immune-regulatory role. Human and
mouse SALM5 bind HVEM/TNFRSF14, and the reciprocal human screen was selective for SALM5
among SALM family members
PMID:27152329.
The SALM5 LRR domain was sufficient for this interaction
PMID:27152329.
The functional inflammation experiments were primarily mouse, despite direct human
binding validation. SALM5-expressing cells suppressed macrophage IL-6 and TNF production
PMID:27152329.
This role is biologically credible and annotation-relevant, but it is a CNS immune-context
function rather than the central synaptic organizer mechanism.
Family-wide experiments found that each SALM enhanced neurite outgrowth in cultured
hippocampal neurons
PMID:18585462.
This supports a broader neuronal-development role but does not isolate an LRFN5-specific
mechanism as cleanly as the presynapse-induction and LAR-RPTP studies.
PubMed searches also recovered LRFN5 locus/copy-number associations with autism,
developmental delay, and schizophrenia. Those association studies were not added as core
functional references because they do not directly establish the molecular or cellular
function of LRFN5. Disease association should remain contextual and should not be used to
infer a specific GO activity.