We first demonstrate a positive role for actin in clustering SVs at synapses. transport packets; large dense-core vesicles made up of active zone components, and synaptic vesicle transfer packets, pleiomorphic vesicles made up of synaptic vesicle-associated proteins as well as proteins critical for exocytosis and endocytosis (McAllister, 2007). It is still unclear what signals mediate the accumulation of these transport packets at developing synapses. Recent work has resurrected the concept that artificial cell contact is sufficient to induce the clustering of synaptic vesicles (SVs) and that MK-4256 actin polymerization is required in this process. Indeed, vesicles accumulate at points of contact between axons and beads coated with poly-d-lysine, poly-l-lysine, or growth factors (Burry and Hayes, 1986;Burry et al., 1986;Lee and Peng, 2006;Lucido et al., 2009). Actin polymerization at contact points precedes vesicle clustering, and treatment with latrunculin A abolishes the accumulation of vesicles at discrete sites along the axon (Kuromi and Kidokoro, 1998;Zhang and Benson, 2001;Lee and Peng, 2006;Lucido et al., 2009). This suggests, but does BMP2 not directly demonstrate, a role for actin in SV clustering. Indeed, although gross actin depolymerization disrupts vesicle clustering, this can be attributed to a variety of factors including the weakening of strong cellcell adhesion. It also remains unclear what signals enhance the preferential polymerization of actin at sites of cellcell contact. The cadherin/-catenin MK-4256 adhesion complex has been shown to play an important role in clustering SVs at synapses (Iwai et al., 2002;Togashi et al., 2002;Bamji et al., 2003,2006;Lee et al., 2008). Perturbation of intercellular cadherin interactions (Togashi et al., 2002) or ablation of -catenin (Bamji et al., 2003) dramatically impairs the accumulation of SVs at contact sites. We have recently exhibited that cadherin/-catenin complexes localize SVs to contact sites by recruiting the PDZ protein, scribble, to developing synapses, and that the mislocalization of SVs is usually phenocopied in scribble knockdown cells (Sun et al., 2009). In this study, we identify a molecular pathway through which cellcell contact can translate to the recruitment and localization of SVs to incipient synapses. We first demonstrate a positive role for actin in clustering SVs at synapses. We next demonstrate that -pix, a Rac/Cdc42-specific guanine nucleotide exchange factor (GEF), can enhance actin polymerization at synapses and can recruit SVs to synapses. Knockdown of -pix results in the mislocalization of SVs MK-4256 along the axon, which can be rescued by enhancing actin polymerization through cortactin overexpression. This directly implicates actin in mediating the effects of -pix on SV clustering. Finally, we show that -pix forms a complex with cadherin, -catenin, and scribble at synapses, and that scribble is important for the localization of -pix at synapses. Together, our data suggest that cadherin/-catenin/scribble complexes recruit -pix to sites of cellcell contact, and that this enhances the local polymerization of actin, which can trap SVs as they translocate along the axon. == Materials and Methods == == == == Recombinant DNAs and siRNAs == To suppress expression MK-4256 of endogenous scribble, two short hairpin RNAs (shRNAs) corresponding to mouse scribble (GenBank accession no.NM_134089) nucleotides 33963416 (shRNA1) and 12801300 (shRNA2) were transiently transfected into mouse hippocampal neurons (Sun et al., 2009). An shRNA specifically against human scribble (GenBank accession no.NM_015356) nucleotides 18421862 was used as a control (shRNA control) (Dow et al., 2007;Sun et al., 2009). To suppress expression of endogenous -pix, three interfering RNA oligonucleotides against rat -pix (Invitrogen) were transiently transfected into rat hippocampal neurons. Sequences of siRNA-1 and siRNA-2 correspond to rat -pix (GenBank accession no.NM_053740) nucleotides 2145 and 17791803, respectively. A mixture of siRNA-1 and siRNA-2 (siRNA-M) was further used as a control for off-target effects. siRNA-negative control duplexes (siRNA-C) were purchased from Invitrogen. To knockdown expression of N-cadherin, we transfected neurons with a previously published N-cadherin siRNA (N-cad siRNA) (Aiga et al., 2011). As controls, cells were transfected with scrambled N-cadherin siRNA (N-cad siRNA-C), or cotransfected with N-cad siRNA plus siRNA-resistant N-cadherin-CFP (N-cad-CFP) (Aiga et al., 2011). GFP-actin, GFP–pix, HA–pix, dominant-negative -pix (DN–pix), HA-tagged cortactin (Cort-HA), synaptophysin-RFP (Syn-RFP), synaptophysin-GFP (Syn-GFP), and the calponin homology domain name of utrophin tagged with RFP (UtrCH-RFP).