Synaptic basal laminae are compositionally distinct from extrasynaptic basal laminae and are required for proper synapse formation. How the basal lamina is patterned with respect to the synapse and how it functions there remain unclear. Two C. elegans basement membrane proteins, NID-1 and CLE-1, homologues of nidogen and collagen XVIII respectively, are in the neural basal lamina and concentrated adjacent to presynaptic zones. Using a synaptobrevin (SNB)::GFP marker for the GABAergic type D motor neurons, we found that the presynaptic zones in
nid-1 null animals are smaller and more closely spaced than wild type, with an overall diffuse appearance along axons. These defects appear similar to, but slightly weaker than, those in
syd-2 mutants. Mutations in
syd-2, which encodes the C. elegans -Liprin, result in diffused and crowded synaptic GFP puncta1 . The vertebrate Leukocyte-common Antigen Related Receptor Protein Tyrosine Phosphatase (LAR-RPTP or LAR) has been shown to bind the laminin-nidogen complex, via its extracellular domain2 , and -liprin (LAR interacting protein) via its intracellular domain3. PTP-3, the C. elegans LAR homologue4 , has two isoforms that differ in the extracellular domain. Using an antibody common to both isoforms we found that endogenous PTP-3 is present along the nerve cord in a punctate pattern, indicating it is in a position to affect synaptogenesis. Thus we hypothesized PTP-3 could link
nid-1 and
syd-2 signaling and examined the role of
ptp-3 in synaptogenesis. There are currently two alleles of
ptp-3. One,
tm352, is a deletion that specifically affects the long isoform, including the domain known to bind laminin-nidogen in vertebrates. The second,
op147, is a Tc1 insertion in the first phosphatase domain.
tm352animals have subtle changes in SNB::GFP similar to, but weaker than, those seen in
nid-1animals.
op147animals appear to have normal SNB::GFP patterns.
nid-1;
tm352 double mutants appear more similar to
tm352single mutants, suggesting that
tm352 is epistatic to
nid-1. Interestingly,
nid-1;
op147double mutants appear more like
syd-2 animals than either single mutant. We also constructed
nid-1;
syd-2 double mutants and observed a
syd-2 like phenotype, indicating that
syd-2may be epistatic to
nid-1.
nid-1;
ptp-3(
tm352 or
op147);
syd-2 triple mutants are similar to
syd-2single mutants, indicating that
syd-2 is epistatic to both
nid-1 and
ptp-3. We can propose at least two possible models. First, SYD-2 may act through PTP-3 to pattern NID-1, which subsequently acts to somehow pattern the synapse. Alternatively, NID-1 may bind PTP-3 to define boundaries for SYD-2 within the synapse. We are investigating the localization of each of these molecules in the mutant backgrounds to differentiate these possibilities.