The initial contact between axons and dendrites at early neuronal synapses is mediated by surface adhesion molecules and is thought to induce synaptic maturation through the recruitment of additional synaptic proteins. The initiation of synaptic maturation should be tightly regulated to ensure that synaptic maturation occurs selectively at subcellular sites of axo-dendritic adhesion. However, the underlying mechanism is poorly understood. Here, we report that the initial trans-synaptic adhesion mediated by presynaptic netrin-G1 and postsynaptic NGL-1 (netrin-G1 ligand-1) induces a cis interaction between netrin-G1 and the receptor protein tyrosine phosphatase LAR (leukocyte antigen-related), and that this promotes presynaptic differentiation. We propose that trans-synaptic adhesions at early neuronal synapses trigger recruitment of neighboring adhesion molecules in a cis manner in order to couple initial axo-dendritic adhesion with synaptic differentiation.

Synaptic cell adhesion molecules (CAMs) are thought to regulate various stages of synapse formation including the initial contact between axons and dendrites, the formation and stabilization of early synapses and their differentiation into mature synapses (Dalva et al., 2007; Han and Kim, 2008; Südhof, 2008; Brose, 2009; Shen and Scheiffele, 2010; Williams et al., 2010; Siddiqui and Craig, 2011; Yuzaki, 2011; Ko, 2012). Recently, a large number of synaptic adhesion molecules have been identified, examples of which include neuroligins, neurexins, SynCAMs, LRRTMs, NGLs, SALMs, LAR-PTPs, TrkC, Cblns-GluRδ, IL1RAPL1, IL1RAcP and Slitrks (Biederer et al., 2002; Lin et al., 2003; Kim et al., 2006; Ko et al., 2006; Wang et al., 2006; Südhof, 2008; de Wit et al., 2009; Ko et al., 2009; Linhoff et al., 2009; Woo et al., 2009b; Kwon et al., 2010; Matsuda et al., 2010; Siddiqui et al., 2010; Uemura et al., 2010; Takahashi et al., 2011; Valnegri et al., 2011; Yoshida et al., 2011; Takahashi et al., 2012; Yoshida et al., 2012; Yim et al., 2013). However, it is unclear how the trans-synaptic adhesion complexes are coupled to the recruitment of additional synaptic proteins for synaptic maturation.

In one candidate mechanism for synaptic maturation, the cytoplasmic regions of pre- and postsynaptic adhesion molecules interact with multi-domain scaffolding proteins, which further interact with and recruit other synaptic proteins (Han and Kim, 2008). In addition, synaptic adhesion molecules might interact with adjacent membrane proteins (e.g. receptors and adhesion molecules) in a cis manner, resulting in their recruitment and stabilization at early synapses. Perhaps a more important question would be whether the initial trans-synaptic interaction is coupled to the recruitment of adjacent membrane and cytoplasmic proteins in a regulated manner, allowing synaptic maturation to occur only at sites of early synaptic adhesion. Such regulated interactions could provide additional mechanisms of regulation and flexibility for synaptic assembly and disassembly.

The netrin-G proteins, netrin-G1 and netrin-G2 (also known as laminet-1 and laminet-2), were originally identified as novel glycosylphosphatidyl inositol (GPI)-anchored adhesion molecules that are expressed in distinct populations of neurons (Nakashiba et al., 2000; Nakashiba et al., 2002; Yin et al., 2002). Netrin-G1 and netrin-G2 are structurally similar to conventional netrins but do not interact with known netrin receptors, such as UNC-5H and DCC/UNC-40 (Nakashiba et al., 2000; Nakashiba et al., 2002), suggesting that they have novel ligands. Indeed, an early study identified NGL-1 as a specific ligand of netrin-G1 (Lin et al., 2003), whereas another study identified NGL-2 as a specific ligand of netrin-G2 (Kim et al., 2006). More recently, NGL-3, which is an orphan receptor, was found to interact with the LAR family of receptor protein tyrosine phosphatases (PTPs), which contains LAR, PTPδ and PTPσ (Woo et al., 2009b; Woo et al., 2009a; Kwon et al., 2010).

Functionally, netrin-G proteins were suggested to regulate neurite outgrowth and patterning of neuronal connections (Nakashiba et al., 2000; Nakashiba et al., 2002; Yin et al., 2002; Lin et al., 2003). More recently, however, a study using transgenic mice lacking netrin-G1 or netrin-G2 expression demonstrated that these molecules do not affect axonal path-finding but rather critically regulate the clustering of their postsynaptic partners (i.e. NGL-1 and NGL-2, see below) at specific segments of dendrites, which probably contributes to axonal cue-induced subdendritic differentiation (Nishimura-Akiyoshi et al., 2007).

NGL proteins and their ligands (netrin-G1, netrin-G2 and LAR-PTPs) have also been implicated in synapse formation. For example, NGL proteins expressed in heterologous cells induce presynaptic differentiation upon contacting axons of co-cultured neurons, suggesting that NGL-dependent trans-synaptic interactions regulate synapse development (Kim et al., 2006; Woo et al., 2009a; Kwon et al., 2010). In line with this, deletion of Ngl2 in mice or acute knockdown of NGL-2 by electroporation causes reductions in synaptic transmission and dendritic spine density at Schaffer collateral synapses in the CA1 region of the hippocampus in a synaptic-input-specific manner (DeNardo et al., 2012).

Notably, NGL-3 displays a presynapse-inducing activity much greater than those of NGL-1 and NGL-2 (Woo et al., 2009a). A possible explanation for this difference might be that, whereas NGL-3 and LAR directly stimulate synapse formation in a bidirectional manner, perhaps the major role for the interactions between NGL-1 and NGL-2 with netrin-G1 or netrin-G2 might be to determine the initial sites of axo-dendritic contact, and couple these events with subsequent maturation processes by recruiting, for instance, neighboring adhesion molecules with direct and stronger synaptogenic activities. In line with this, netrin-G1 and netrin-G2 are GPI-anchored proteins (Nakashiba et al., 2000; Nakashiba et al., 2002; Yin et al., 2002), and have been suggested to have co-receptors.

Here, we show that the binding of NGL-1 to presynaptic netrin-G1 induces a cis interaction between netrin-G1 and LAR, and that this promotes presynaptic differentiation. These results identify LAR as a co-receptor of netrin-G1, reveal a novel mode of synapse formation involving trans-induced cis interaction and suggest a mechanism for adhesion-site-specific synapse maturation.

NGLs and their ligands do not interact in cis under basal conditions

Presynaptic netrin-G1, netrin-G2 and LAR trans-synaptically interact with NGL-1, NGL-2 and NGL-3, respectively (Lin et al., 2003; Kim et al., 2006; Woo et al., 2009a). These interactions suggest that NGLs interact with each other on the postsynaptic surface in a cis manner, and similarly, that netrin-G1, netrin-G2 and LAR interact in cis with each other on the presynaptic surface.

We first explored this possibility in heterologous cells by doubly expressing netrin-G1 or netrin-G2 with LAR. Clustering of surface HA-netrin-G1 or HA-netrin-G2 by preclustered HA antibodies had no effect on Myc-LAR, which exhibited a diffuse distribution on the cell surface (Fig. 1A,B). Similarly, Myc-antibody-induced clustering of Myc-LAR had no effect on HA-netrin-G1 or HA-netrin-G2. By contrast, HA-neuroligin-1 and neurexin-1β-CFP were coclustered by neuroligin-1 clustering (Fig. 1A,B), consistent with their previously reported cis interaction (Taniguchi et al., 2007). In coimmunoprecipitation experiments, netrin-G1 or netrin-G2 and LAR did not form a complex in heterologous cells (Fig. 1C), further suggesting that they do not engage in a cis interaction. With respect to NGL, primary clustering of Myc-NGL-1 or Myc-NGL-2 by preclustered Myc antibodies did not induce coclustering of HA-NGL-3 (Fig. 1D). These results suggest that neither NGLs nor their ligands (netrin-G1, netrin-G2 and LAR) display cis interactions under basal conditions.

Fig. 1.

NGL proteins or their trans-synaptic ligands do not interact in cis under basal conditions. (A) Netrin-G1 and netrin-G2 do not interact in cis with LAR on the cell surface. HEK293T cells doubly expressing HA-netrin-G1 or HA-netrin-G2 and Myc-LAR, or HA-neuroligin-1 and neurexin-1β-CFP, were incubated with preclustered HA or Myc antibodies followed by immunostaining. (B) Quantification of the results in A. Clustering index indicates the average intensity ratio of secondarily clustered molecules in the region of primary clustering versus non-clustering. For example, the first bar refers to the average intensity ratio of Myc-LAR in the HA-netrin-G1 area versus non-HA area. Mean ± s.e.m., n = 15, ***P<0.001, ANOVA. (C) Netrin-G1 and netrin-G2 do not form coimmunoprecipitate complexes with LAR in heterologous cells. HEK293T cells doubly expressing Myc-netrin-G1 and LAR-FLAG-C1522S (phosphatase-dead), or Myc-netrin-G2 and LAR-FLAG-C1522S were immunoprecipitated by FLAG antibodies, and immunoblotted with Myc antibodies. (D) NGL-1 and NGL-2 do not interact in cis with NGL-3 on the cell surface. HEK293T cells doubly expressing Myc-NGL-1 and HA-NGL-3, or Myc-NGL-2 and HA-NGL-3, were incubated with preclustered Myc antibodies followed by immunostaining. Scale bars: 10 µm.

Fig. 1.

NGL proteins or their trans-synaptic ligands do not interact in cis under basal conditions. (A) Netrin-G1 and netrin-G2 do not interact in cis with LAR on the cell surface. HEK293T cells doubly expressing HA-netrin-G1 or HA-netrin-G2 and Myc-LAR, or HA-neuroligin-1 and neurexin-1β-CFP, were incubated with preclustered HA or Myc antibodies followed by immunostaining. (B) Quantification of the results in A. Clustering index indicates the average intensity ratio of secondarily clustered molecules in the region of primary clustering versus non-clustering. For example, the first bar refers to the average intensity ratio of Myc-LAR in the HA-netrin-G1 area versus non-HA area. Mean ± s.e.m., n = 15, ***P<0.001, ANOVA. (C) Netrin-G1 and netrin-G2 do not form coimmunoprecipitate complexes with LAR in heterologous cells. HEK293T cells doubly expressing Myc-netrin-G1 and LAR-FLAG-C1522S (phosphatase-dead), or Myc-netrin-G2 and LAR-FLAG-C1522S were immunoprecipitated by FLAG antibodies, and immunoblotted with Myc antibodies. (D) NGL-1 and NGL-2 do not interact in cis with NGL-3 on the cell surface. HEK293T cells doubly expressing Myc-NGL-1 and HA-NGL-3, or Myc-NGL-2 and HA-NGL-3, were incubated with preclustered Myc antibodies followed by immunostaining. Scale bars: 10 µm.

NGL-1 binding to netrin-G1 causes netrin-G1 to interact in cis with LAR in heterologous cells

We next tested whether ligand binding could induce cis interactions between NGLs or their ligands. NGL-1 and netrin-G1 were singly expressed in two different groups of HEK293T cells; when these cells were mixed, the expressed proteins concentrated at the cell–cell interface, indicative of their trans-cellular adhesion (Fig. 2A,B). Importantly, when HEK293T cells doubly expressing netrin-G1 and LAR were mixed with NGL-1-expressing cells, NGL-1, netrin-G1 and LAR all concentrated at the cell–cell interface (Fig. 2A,B), suggesting that the trans interaction between NGL-1 and netrin-G1 caused netrin-G1 to interact in cis with LAR. By contrast, NGL-2 failed to induce coclustering of netrin-G2 with LAR (Fig. 2A,B). In control experiments, NGL-1 did not induce coclustering of netrin-G2 and LAR, and similarly, NGL-2 failed to induce coclustering of netrin-G1 and LAR. In addition, NGL-1 did not cocluster with LAR in the absence netrin-G1 coexpression. Here, we used a phosphatase-dead form of LAR (LAR-FLAG-C1522S) in order not to make transfected cells unhealthy, which can occur through tyrosine dephosphorylation of target proteins. This, however, suggests that the NGL-1-induced cis interaction between netrin-G1 and LAR does not require the phosphatase activity of LAR.

Fig. 2.

NGL-1 binding to netrin-G1 induces netrin-G1 to interact in cis with LAR. (A) Surface NGL-1 in one cell induces trans-cellular coclustering of netrin-G1 and LAR in another cell at the cell–cell interface, forming a zipper-like, protein-enriched line. A group of HEK293T cells expressing NGL-1-EGFP were co-cultured with another group of cells doubly expressing Myc-netrin-G1 and LAR-FLAG-C1522S for 24 hours, followed by triple staining of the proteins. Additional combinations of transfection include NGL-2-EGFP with Myc-netrin-G2 and LAR-FLAG-C1522S; NGL-1-EGFP with Myc-netrin-G2 and LAR-FLAG-C1522S; NGL-2-EGFP with Myc-netrin-G1 and LAR-FLAG-C1522S; NGL-1-EGFP with Myc (Myc only) and LAR-FLAG-C1522S. (B) Quantification of the results in (A). Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. Mean ± s.e.m., n = 15, ***P<0.001, ANOVA. (C) The binding of NGL-1 to netrin-G1 causes NGL-1 to coprecipitate with LAR. Soluble NGL-1-Fc or NGL-2-Fc proteins were incubated with HEK293T cells doubly expressing Myc-netrin-G1 or Myc-netrin-G2 and LAR-FLAG-C1522S for 2 hours, followed by precipitation of NGL-1-Fc and immunoblotting of netrin-G1 or netrin-G2 and LAR. (D) A control experiment showing that NGL-1 does not crossreact with LAR, whereas it binds and clusters netrin-G1. HEK293T cells expressing Myc-LAR or Myc-netrin-G1 (positive control) were incubated with preclustered soluble NGL-1-Fc for 2 hours, followed by immunostaining. (E) The interaction between NGL-3 and LAR does not cause NGL-3 to interact in cis with NGL-1 or NGL-12, or LAR to interact in cis with netrin-G1 or netrin-G2. Preclustered soluble NGL-3-Fc proteins were incubated with HEK293T cells expressing LAR and netrin-G1 or netrin-G2. Alternatively, preclustered soluble LAR-Fc proteins were incubated with cells expressing NGL-3 and NGL-1 or NGL-2. Scale bars: 10 µm.

Fig. 2.

NGL-1 binding to netrin-G1 induces netrin-G1 to interact in cis with LAR. (A) Surface NGL-1 in one cell induces trans-cellular coclustering of netrin-G1 and LAR in another cell at the cell–cell interface, forming a zipper-like, protein-enriched line. A group of HEK293T cells expressing NGL-1-EGFP were co-cultured with another group of cells doubly expressing Myc-netrin-G1 and LAR-FLAG-C1522S for 24 hours, followed by triple staining of the proteins. Additional combinations of transfection include NGL-2-EGFP with Myc-netrin-G2 and LAR-FLAG-C1522S; NGL-1-EGFP with Myc-netrin-G2 and LAR-FLAG-C1522S; NGL-2-EGFP with Myc-netrin-G1 and LAR-FLAG-C1522S; NGL-1-EGFP with Myc (Myc only) and LAR-FLAG-C1522S. (B) Quantification of the results in (A). Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. Mean ± s.e.m., n = 15, ***P<0.001, ANOVA. (C) The binding of NGL-1 to netrin-G1 causes NGL-1 to coprecipitate with LAR. Soluble NGL-1-Fc or NGL-2-Fc proteins were incubated with HEK293T cells doubly expressing Myc-netrin-G1 or Myc-netrin-G2 and LAR-FLAG-C1522S for 2 hours, followed by precipitation of NGL-1-Fc and immunoblotting of netrin-G1 or netrin-G2 and LAR. (D) A control experiment showing that NGL-1 does not crossreact with LAR, whereas it binds and clusters netrin-G1. HEK293T cells expressing Myc-LAR or Myc-netrin-G1 (positive control) were incubated with preclustered soluble NGL-1-Fc for 2 hours, followed by immunostaining. (E) The interaction between NGL-3 and LAR does not cause NGL-3 to interact in cis with NGL-1 or NGL-12, or LAR to interact in cis with netrin-G1 or netrin-G2. Preclustered soluble NGL-3-Fc proteins were incubated with HEK293T cells expressing LAR and netrin-G1 or netrin-G2. Alternatively, preclustered soluble LAR-Fc proteins were incubated with cells expressing NGL-3 and NGL-1 or NGL-2. Scale bars: 10 µm.

Biochemically, soluble NGL-1-Fc fusion proteins incubated with HEK293T cells doubly expressing netrin-G1 and LAR caused coprecipitation of NGL-1-Fc with LAR only in the presence of netrin-G1 (Fig. 2C). By contrast, NGL-2-Fc failed to coprecipitate with LAR even in the presence of netrin-G2 (Fig. 2C). In a control experiment, NGL-1-Fc bound to netrin-G1 but not to LAR (Fig. 2D). These results suggest that netrin-G1 binds its ligand (NGL-1), and thereafter interacts with LAR in a cis manner. We next tested whether the binding of NGL-3 to LAR could cause LAR to cocluster with netrin-G1 or netrin-G2 in the reverse orientation. In HEK293T cells expressing LAR and netrin-G1 or netrin-G2, NGL-3-Fc successfully clustered LAR on the cell surface but failed to induce coclustering of LAR with netrin-G1 or netrin-G2 (Fig. 2E). In addition, soluble LAR-Fc successfully clustered NGL-3 on the cell surface but failed to induce coclustering of NGL-3 with NGL-1 or NGL-2 (Fig. 2E). These results suggest that NGL-1, but not NGL-2 or NGL-3, is capable of inducing its specific receptor (netrin-G1) to interact in cis with LAR.

NGL-1 binding to netrin-G1 on the axonal surface induces coclustering of netrin-G1 with LAR

As NGL-1 and netrin-G1 are predominantly detected in dendrites and axons, respectively (Kim et al., 2006; Nishimura-Akiyoshi et al., 2007), we tested whether the NGL-1-binding-induced coclustering of netrin-G1 and LAR occurs in axons. When cultured hippocampal neurons transfected with netrin-G1 [8–9 days in vitro (DIV)] were incubated with preclustered NGL-1-Fc proteins, they specifically bound to netrin-G1 clusters on MAP2-negative axons (Fig. 3A). By contrast, control Fc proteins failed to cause netrin-G1 clustering on axons. We then doubly expressed netrin-G1 and LAR in hippocampal neurons and incubated these neurons with NGL-1-Fc. Following NGL-1-Fc treatment, we observed coclustering of netrin-G1 with LAR at sites of NGL-1 binding (Fig. 3B,C). In control neurons treated with Fc alone, LAR did not cocluster with netrin-G1 in axons and mainly appeared in the soma and dendrites. It seems that LAR proteins in axons are present in amounts smaller than those in the cell body and proximal dendrites and/or are diffusely distributed in axons, making it difficult to detect unless they are clustered.

Fig. 3.

NGL-1 binding to netrin-G1 on the axonal surface causes coclustering of netrin-G1 with LAR. (A) NGL-1-Fc binds to netrin-G1 clusters on MAP2-negative axons. Cultured rat hippocampal neurons at 8–9 DIV were transfected with Myc-netrin-G1 for 3 days, and incubated with preclustered NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and MAP2. (B) NGL-1-Fc binding to netrin-G1 induces coclustering of netrin-G1 with LAR. Cultured rat hippocampal neurons at 8–9 DIV were cotransfected with Myc-netrin-G1 and LAR-FLAG-C1522S for 3 days, and incubated with preclustered NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and FLAG. (C) Quantification of the results in B. Integrated intensities of LAR clusters were normalized to those of netrin-G1. Mean ± s.e.m., n = 15, ***P<0.001, ANOVA. (D) NGL-1-Fc binding to endogenous netrin-G1 induces coclustering of netrin-G1 with endogenous LAR. Cultured rat hippocampal neurons at 11–12 DIV were incubated with preclustered NGL-1-Fc for 2 hours, followed by staining for Fc, netrin-G1 and LAR. Quantification of the results could not be performed because endogenous netrin-G1 signals in Fc alone panels were almost undetectable unless they are clustered by NGL-1-Fc. Scale bars: 10 µm.

Fig. 3.

NGL-1 binding to netrin-G1 on the axonal surface causes coclustering of netrin-G1 with LAR. (A) NGL-1-Fc binds to netrin-G1 clusters on MAP2-negative axons. Cultured rat hippocampal neurons at 8–9 DIV were transfected with Myc-netrin-G1 for 3 days, and incubated with preclustered NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and MAP2. (B) NGL-1-Fc binding to netrin-G1 induces coclustering of netrin-G1 with LAR. Cultured rat hippocampal neurons at 8–9 DIV were cotransfected with Myc-netrin-G1 and LAR-FLAG-C1522S for 3 days, and incubated with preclustered NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and FLAG. (C) Quantification of the results in B. Integrated intensities of LAR clusters were normalized to those of netrin-G1. Mean ± s.e.m., n = 15, ***P<0.001, ANOVA. (D) NGL-1-Fc binding to endogenous netrin-G1 induces coclustering of netrin-G1 with endogenous LAR. Cultured rat hippocampal neurons at 11–12 DIV were incubated with preclustered NGL-1-Fc for 2 hours, followed by staining for Fc, netrin-G1 and LAR. Quantification of the results could not be performed because endogenous netrin-G1 signals in Fc alone panels were almost undetectable unless they are clustered by NGL-1-Fc. Scale bars: 10 µm.

We next tested whether NGL-1 can induce coclustering of endogenous netrin-G1 with LAR. Incubation of cultured neurons with NGL-1-Fc induced coclustering of endogenous netrin-G1 with endogenous LAR at sites of NGL-1-Fc binding (Fig. 3D). The specificity of netrin-G1 and LAR antibodies could be supported by the strong and punctate signals of endogenous netrin-G1 and LAR induced by the incubation of cultured neurons with their specific ligands NGL-1 and NGL-3, respectively (supplementary material Fig. S1A,B). These results suggest that the binding of NGL-1 to netrin-G1 on axonal surfaces causes netrin-G1 to interact in cis with LAR.

LAR is required for NGL-1-dependent presynaptic differentiation

NGL proteins have the capacity to induce presynaptic differentiation in contacting axons (Kim et al., 2006; Woo et al., 2009b). Incubation of Myc-netrin-G1-expressing neurons with NGL-1-Fc caused coclustering of Myc-netrin-G1 with synapsin I (a presynaptic protein; Fig. 4A,B) and VGlut1 (an excitatory presynaptic vesicle protein; supplementary material Fig. S2A,B). In addition, NGL-1-Fc induced uptake of synaptotagmin I luminal domain antibodies at sites of NGL-1-Fc and Myc-netrin-G1 coclustering (supplementary material Fig. S2C,D), suggesting that NGL-1-Fc can induce recycling of presynaptic vesicles. In control experiments, Fc alone did not cause synapsin I coclustering at sites of netrin-G1 clusters (Fig. 4,A,B).

Fig. 4.

LAR is required for NGL-1-dependent presynaptic differentiation. (A) NGL-1-Fc binding induces coclustering of netrin-G1 with synapsin I. Cultured hippocampal neurons transfected with Myc-netrin-G1 (DIV 8–11 or 9–12) were incubated with NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and synapsin I. (B) Quantification of the results in A. Integrated intensities of synapsin I clusters were normalized to those of netrin-G1. Mean ± s.e.m., n = 15, ***P<0.001, Unpaired t-test. (C) Knockdown of LAR expression by shRNA (sh-LAR) in heterologous cells. Controls were an empty vector (sh-vec), a mismatch control sh-LAR (sh-LAR*) and knockdown-resistant LAR rescue constructs (LAR-FLAG res; wild-type and phosphatase-dead C1522S). Intensities of LAR were normalized to those in the control lane (sh-vec). n = 6, **P<0.01, ANOVA. (D) Reduced NGL-1-Fc-induced coclustering of netrin-G1 with synapsin I in LAR-knockdown neurons and rescue of this effect by knockdown-resistant LAR (LAR-FLAGres and LAR-FLAG-C1552Sres). Cultured hippocampal neurons transfected with Myc-netrin-G1 and sh-LAR, sh-vec or sh-LAR*, or sh-LAR and LAR-FLAGres or LAR-FLAG-C1522Sres (DIV 8–11 or 9–12) were incubated with NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and synapsin I. (E) Quantification of the results in D. n = 15, **P<0.01, ANOVA. Scale bars: 10 µm.

Fig. 4.

LAR is required for NGL-1-dependent presynaptic differentiation. (A) NGL-1-Fc binding induces coclustering of netrin-G1 with synapsin I. Cultured hippocampal neurons transfected with Myc-netrin-G1 (DIV 8–11 or 9–12) were incubated with NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and synapsin I. (B) Quantification of the results in A. Integrated intensities of synapsin I clusters were normalized to those of netrin-G1. Mean ± s.e.m., n = 15, ***P<0.001, Unpaired t-test. (C) Knockdown of LAR expression by shRNA (sh-LAR) in heterologous cells. Controls were an empty vector (sh-vec), a mismatch control sh-LAR (sh-LAR*) and knockdown-resistant LAR rescue constructs (LAR-FLAG res; wild-type and phosphatase-dead C1522S). Intensities of LAR were normalized to those in the control lane (sh-vec). n = 6, **P<0.01, ANOVA. (D) Reduced NGL-1-Fc-induced coclustering of netrin-G1 with synapsin I in LAR-knockdown neurons and rescue of this effect by knockdown-resistant LAR (LAR-FLAGres and LAR-FLAG-C1552Sres). Cultured hippocampal neurons transfected with Myc-netrin-G1 and sh-LAR, sh-vec or sh-LAR*, or sh-LAR and LAR-FLAGres or LAR-FLAG-C1522Sres (DIV 8–11 or 9–12) were incubated with NGL-1-Fc for 2 hours, followed by staining for Fc, Myc and synapsin I. (E) Quantification of the results in D. n = 15, **P<0.01, ANOVA. Scale bars: 10 µm.

To test the possibility that LAR mediates NGL-1-induced synapsin I clustering, we performed knockdown experiments in cultured neurons using a previously reported shRNA for LAR (sh-LAR) and a mismatch control (sh-LAR*) (Fig. 4C) (Mander et al., 2005). In neurons transfected with sh-LAR (DIV 8�–11 or 9–12), NGL-1-Fc induced lower levels of synapsin I coclustering compared with those in control neurons expressing an empty shRNA vector (sh-vec) or sh-LAR*, as indicated by the intensity ratios of synapsin I and netrin-G1. In rescue experiments, neurons cotransfected with knockdown-resistant LAR-FLAG (LAR-FLAGres) (Fig. 4C) and sh-LAR showed synapsin I coclustering comparable to that in control neurons transfected with sh-vec (Fig. 4D,E). Interestingly, neurons cotransfected with phosphatase-dead LAR-FLAG (LAR-FLAG-C1522Sres) and sh-LAR showed synapsin I coclustering comparable to that observed in neurons transfected with wild-type LAR-FLAGres (Fig. 4D,E). These results suggest that LAR is required for NGL-1-dependent presynaptic differentiation, and that the phosphatase activity of LAR is not required for this induction.

Domains mediating the cis interaction between netrin-G1 and LAR

Netrin-G1 contains one VI domain (also known as the laminin globular domain) and three V domains (V1–V3; laminin epidermal growth factor-like or LE-like repeats). The extracellular region of LAR contains three immunoglobulin domains (Ig) and eight fibronectin III (FNIII) repeats. To identify the domains responsible for mediating the cis interaction between netrin-G1 and LAR, we generated deletion variants of these proteins (Fig. 5A), and performed binding assays in which HEK293T cells expressing NGL-1 were mixed with HEK293T cells doubly expressing netrin-G1 and LAR.

Fig. 5.

Domains that mediate the cis interaction between netrin-G1 and LAR. (A) Deletion variants of netrin-G1 and NGL-1. (B) Netrin-G1 lacking the domain V region (ΔV) fails to interact in cis with LAR upon NGL-1 binding. A group of HEK293T cells expressing NGL-1-EGFP were mixed with another group of HEK293 cells doubly expressing Myc-netrin-G1 (wild-type or ΔV) and LAR for 1 day, followed by triple staining. (C) Quantification of the results in (B). Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. Mean ± s.e.m., n = 15 **P<0.01, Unpaired t-test. (D) Myc-netrin-G1 ΔV displays normal binding to NGL-1-Fc. HEK293T cells expressing Myc-netrin-G1 (wild type or ΔV) were incubated with preclustered NGL-1-Fc proteins. (E) LAR mutants lacking the FN or Ig domains fail to interact in cis with NGL-bound netrin-G1. A group of HEK293T cells expressing NGL-1-EGFP and another doubly expressing Myc-netrin-G1 and LAR (wild type and ΔFN/ΔIg mutant in pDisplay vector) were mixed cultured for 1 day, followed by triple staining. (F) Quantification of the results in E. Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. n = 15, ***P<0.001, ANOVA. Scale bars: 10 µm.

Fig. 5.

Domains that mediate the cis interaction between netrin-G1 and LAR. (A) Deletion variants of netrin-G1 and NGL-1. (B) Netrin-G1 lacking the domain V region (ΔV) fails to interact in cis with LAR upon NGL-1 binding. A group of HEK293T cells expressing NGL-1-EGFP were mixed with another group of HEK293 cells doubly expressing Myc-netrin-G1 (wild-type or ΔV) and LAR for 1 day, followed by triple staining. (C) Quantification of the results in (B). Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. Mean ± s.e.m., n = 15 **P<0.01, Unpaired t-test. (D) Myc-netrin-G1 ΔV displays normal binding to NGL-1-Fc. HEK293T cells expressing Myc-netrin-G1 (wild type or ΔV) were incubated with preclustered NGL-1-Fc proteins. (E) LAR mutants lacking the FN or Ig domains fail to interact in cis with NGL-bound netrin-G1. A group of HEK293T cells expressing NGL-1-EGFP and another doubly expressing Myc-netrin-G1 and LAR (wild type and ΔFN/ΔIg mutant in pDisplay vector) were mixed cultured for 1 day, followed by triple staining. (F) Quantification of the results in E. Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. n = 15, ***P<0.001, ANOVA. Scale bars: 10 µm.

Unlike wild-type netrin-G1, a mutant netrin-G1 lacking the domain V region (Myc-netrin-G1 ΔV) failed to cocluster with LAR upon NGL-1 binding (Fig. 5B,C). Myc-netrin-G1 ΔV displayed normal binding to NGL-1 (Fig. 5B), as also confirmed by NGL-1-Fc binding (Fig. 5D), consistent with the reported involvement of the N-terminal VI (lamin globular) domain in NGL-1 binding (Seiradake et al., 2011). These results suggest that the domain V region of netrin-G1 is important for the cis interaction with LAR. The LAR deletion variants lacking the eight FN domains (LAR-ΔFN) and the three Ig domains (LAR-ΔIg) both failed to interact with NGL-1-bound netrin-G1 (Fig. 5E,F), suggesting that the FN and Ig domains in LAR are both important for the interaction with NGL-1-bound netrin-G1.

Alternative splicing in netrin-G1 regulates NGL-1 and LAR binding

Netrin-G1 displays extensive alternative splicing, mainly in the domain V region (Nakashiba et al., 2000; Nakashiba et al., 2002; Yin et al., 2002; Aoki-Suzuki et al., 2005; Meerabux et al., 2005; Eastwood and Harrison, 2008). In the fetal human brain, netrin-G1c and netrin-G1d are most abundant (Meerabux et al., 2005) (Fig. 6A). We used netrin-G1c in the above-described experiments. Netrin-G1m is the longest isoform and contains all three V domains and a short unknown domain (Ukd). We first used soluble NGL-1-Fc binding assays to test whether these splice variants display different levels of NGL-1 binding. Netrin-G1m on the cell surface showed reduced levels of NGL-1 binding relative to that in netrin-G1c and netrin-G1d (Fig. 6B,C), suggesting that alternative splicing of netrin-G1 regulates NGL-1 binding. Netrin-G1m showed reduced cis interactions with LAR upon NGL-1 binding in a mixed HEK cell assay (Fig. 6D,E). The extents of the reductions were similar to those for NGL-1 binding to netrin-G1, suggesting that this reduction in cis interaction is largely attributable to reduced NGL-1 binding. We also tested whether neuronally expressed netrin-G1m display altered cis interactions with LAR upon NGL-1 binding. Netrin-G1m displayed reduced cis interactions with LAR upon NGL-1-Fc binding, compared with netrin-G1c and netrin-G1d (Fig. 6F,G). Together, these results suggest that alternative splicing in netrin-G1 regulates its trans interaction with NGL-1, and these changes might subsequently affect the cis interaction between netrin-G1 and LAR.

Fig. 6.

Alternative splicing in netrin-G1 regulates its interaction with NGL-1 and LAR. (A) Splice variants of netrin-G1 used in this study. (B) Netrin-G1m show reduced levels of NGL-1 binding. HEK293T cells transfected with netrin-G1 splice variants were incubated with NGL-1-Fc (2 hours, non-permeabilized live cells). (C) Quantification of the results in (B). Mean ± s.e.m., n = 15, *P<0.01, ANOVA. (D) Netrin-G1m show reduced levels of cis interaction with LAR in heterologous cells. One group of HEK293T cells expressing NGL-1-EGFP and another group doubly expressing Myc-netrin-G1 splice variants and LAR-C1522S were mixed cultured for 1 day, followed by triple staining. (E) Quantification of the results in D. Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. n = 15, **P<0.01, ANOVA. (F) Netrin-G1m show reduced levels of cis interaction with LAR in neurons. Cultured hippocampal neurons doubly expressing Myc-netrin-G1 splice variants and LAR-FLAG-C1522S (DIV 8–11 or 9–12) were incubated with NGL-1-Fc for 2 hours, followed by triple staining. (G) Quantification of the results in F. n = 15, **P<0.01, ANOVA. Scale bars: 10 µm.

Fig. 6.

Alternative splicing in netrin-G1 regulates its interaction with NGL-1 and LAR. (A) Splice variants of netrin-G1 used in this study. (B) Netrin-G1m show reduced levels of NGL-1 binding. HEK293T cells transfected with netrin-G1 splice variants were incubated with NGL-1-Fc (2 hours, non-permeabilized live cells). (C) Quantification of the results in (B). Mean ± s.e.m., n = 15, *P<0.01, ANOVA. (D) Netrin-G1m show reduced levels of cis interaction with LAR in heterologous cells. One group of HEK293T cells expressing NGL-1-EGFP and another group doubly expressing Myc-netrin-G1 splice variants and LAR-C1522S were mixed cultured for 1 day, followed by triple staining. (E) Quantification of the results in D. Clustering index indicates the average intensity ratio of LAR in the area of cell–cell interface versus non-interface area. n = 15, **P<0.01, ANOVA. (F) Netrin-G1m show reduced levels of cis interaction with LAR in neurons. Cultured hippocampal neurons doubly expressing Myc-netrin-G1 splice variants and LAR-FLAG-C1522S (DIV 8–11 or 9–12) were incubated with NGL-1-Fc for 2 hours, followed by triple staining. (G) Quantification of the results in F. n = 15, **P<0.01, ANOVA. Scale bars: 10 µm.

NGL-1 binding-defective netrin-G1 fails to interact in cis with LAR

To obtain further evidence supporting the notion that NGL-1 binding to netrin-G1 drives the interaction of netrin-G1 with LAR, we generated a netrin-G1 point mutant (Y86T in the domain VI or laminin globular domain) that we predicted would lose NGL-1 binding based on the crystal structure of netrin-G1 (Seiradake et al., 2011). This mutant indeed failed to interact with soluble NGL-1-Fc (Fig. 7A). In cultured neurons expressing netrin-G1-Y86T, NGL-1-Fc did not induce coclustering of netrin-G1-Y86T with LAR (Fig. 7B,C). In this experiment, we used both wild-type neurons and netrin-G1-deficient neurons obtained from previously reported knockout mice (Nishimura-Akiyoshi et al., 2007) to minimize the effect of endogenous netrin-G1 proteins. The results were qualitatively similar in the two cell types (wild type and knockout), suggesting that the binding of NGL-1 to netrin-G1 drives the cis interaction of netrin-G1 with LAR.

Fig. 7.

NGL-1 binding-defective netrin-G1 fails to interact in cis with LAR. (A) Netrin-G1-Y86T fails to interact with NGL-1. HEK293T cells expressing Myc-netrin-G1 (WT and Y86T) were incubated with NGL-1-Fc, followed by immunostaining. (B) Netrin-G1-Y86T fails to cocluster with LAR upon NGL-1 binding. Cultured hippocampal neurons expressing Myc-netrin-G1 and LAR-FLAG-C1522S or neurons from netrin-G1-deficient mouse brain coexpressing Myc-netrin-G1 (wild type or Y86T) and LAR-FLAG-C1522S were incubated with NGL-1-Fc for 2 hours followed by triple staining. (C) Quantification of the results in B. Mean ± s.e.m., n = 15, **P<0.01, ANOVA. Scale bars: 10 µm.

Fig. 7.

NGL-1 binding-defective netrin-G1 fails to interact in cis with LAR. (A) Netrin-G1-Y86T fails to interact with NGL-1. HEK293T cells expressing Myc-netrin-G1 (WT and Y86T) were incubated with NGL-1-Fc, followed by immunostaining. (B) Netrin-G1-Y86T fails to cocluster with LAR upon NGL-1 binding. Cultured hippocampal neurons expressing Myc-netrin-G1 and LAR-FLAG-C1522S or neurons from netrin-G1-deficient mouse brain coexpressing Myc-netrin-G1 (wild type or Y86T) and LAR-FLAG-C1522S were incubated with NGL-1-Fc for 2 hours followed by triple staining. (C) Quantification of the results in B. Mean ± s.e.m., n = 15, **P<0.01, ANOVA. Scale bars: 10 µm.

In the present study, we found that the trans-synaptic adhesion between postsynaptic NGL-1 and presynaptic netrin-G1 triggers a cis interaction of netrin-G1 with LAR on the presynaptic surface. This induced cis interaction is important for presynaptic differentiation, suggesting that this is a case in which a postsynaptic adhesion molecule plays an instructive role in driving contact-dependent and regulated presynaptic differentiation.

The suggested cis interaction between netrin-G1 and LAR is supported by similarities in their expression patterns. For instance, mRNA encoding netrin-G1 is first detected in the hippocampus from postnatal day 3 (P3), becomes abundant in the CA1 region until P7 and is then abundant in the dentate gyrus (DG) after P14 (Eastwood and Harrison, 2008). In the cortex, netrin-G1 mRNA is detected on the piriform cortex on P0 and is highly expressed in layers II and III of the entorhinal cortex throughout the adult stage (Nakashiba et al., 2002; Nishimura-Akiyoshi et al., 2007). Similarly, mRNA encoding LAR is detected in the hippocampus on P0, is abundant in the CA1 and CA3 regions at P7, and mainly exists in the DG after P14. In cortical layers, LAR mRNA begins to be detected from P0 and becomes specific to layer II and layer III from P14 (Honkaniemi et al., 1998; Kwon et al., 2010).

This study extends the known roles of synaptic organizers. Pre- and postsynaptic organizers are currently thought to couple initial axo-dendritic synaptic adhesions with the recruitment of cytosolic proteins at the pre- and postsynaptic sides, thereby promoting synaptic maturation. Here, we suggest a novel mechanism that neighboring adhesion molecules could also be recruited by synaptic organizers. In the present study, we identify LAR as a novel co-receptor for netrin-G1. Netrin-G1 is a GPI-anchored protein that lacks the cytoplasmic region, making it unlikely to interact with cytoplasmic proteins during synaptic development. Thus, netrin-G1 has been expected to interact with a co-receptor that contains a cytoplasmic region.

LAR contains two cytoplasmic tyrosine phosphatase domains. The membrane proximal phosphatase domain (D1) is catalytically active, whereas the membrane distal domain (D2) is inactive but can interact with various cytosolic proteins (Nam et al., 1999; Blanchetot et al., 2002). An important binding partner of the D2 domain is the multi-domain adaptor liprin-α (Pulido et al., 1995), which is further connected to key presynaptic proteins, including RIM, ELKS/ERC and CASK (Schoch et al., 2002; Ko et al., 2003; Olsen et al., 2005). In addition, studies using C. elegans, Drosophila and mammalian cells have indicated that LAR and liprin-α crucially regulate pre- and postsynaptic development (Wyszynski et al., 2002; Dunah et al., 2005; Patel et al., 2006; Spangler and Hoogenraad, 2007; Jin and Garner, 2008).

Our results suggest that the tyrosine phosphatase activity of LAR is not required for NGL-1-dependent coclustering of netrin-G1 with synapsin I. Although further details remain to be explored, it should be noted that LAR directly and trans-synaptically interacts with NGL-3, and that NGL-3 induces presynaptic differentiation more strongly than NGL-1 in mixed culture assays (Woo et al., 2009b). In addition, LAR directly and trans-synaptically interacts with IL1RAcP, a postsynaptic adhesion molecule, in addition to NGL-3 (Yoshida et al., 2012). Whether LAR-dependent presynaptic differentiation requires phosphatase activity thus remains to be studied under the context of both cis and trans interactions.

Our study reveals that the netrin-G1–NGL-1 complex has a bidirectional function in synaptic development. Netrin-G1 and netrin-G2 are expressed in distinct subsets of neurons and distribute mainly to axons (Nakashiba et al., 2000; Nakashiba et al., 2002; Yin et al., 2002), suggesting that netrin-G proteins differentially regulate axonal functions in a pathway-specific manner. Consistent with this notion, netrin-G1 and netrin-G2 are required for surface clustering of NGL-1 and NGL-2, respectively, in subdendritic segments of target neurons (Nishimura-Akiyoshi et al., 2007). These previous results suggest that this is a case of an axonal factor (netrin-G1 or netrin-G2) playing an instructive role in driving postsynaptic differentiation. However, it was unclear whether NGL-1 and NGL-2 play any instructive roles in presynaptic differentiation. Here, we demonstrate that NGL-1 regulates presynaptic differentiation through netrin-G1 and LAR, which suggests a novel role for a postsynaptic factor (NGL-1) in driving presynaptic differentiation, and is reminiscent of the reported enhancement of presynaptic release by postsynaptic neuroligin-1 and PSD-95 through trans-synaptic adhesion (Futai et al., 2007).

An important aspect of the cis interaction between netrin-G1 and LAR is that it is ‘induced’ by NGL-1 binding. Reported cis interactions between neuronal adhesion molecules have been shown to regulate neurite outgrowth and synaptic development, such as those of SynCAM–SynCAM for synaptic development (Fogel et al., 2011). Examples include the interactions of FLRT–FGF for neurite outgrowth (Wheldon et al., 2010), ephirinA5–Eph3 for growth cone navigation and axon targeting (Marquardt et al., 2005; Carvalho et al., 2006) and NB-3–L1(CHL1) for apical dendrite orientation (Ye et al., 2008). However, none of these cis interactions are known to be induced by trans adhesions.

The NGL-1-induced netrin-G1-LAR cis interaction might involve a conformational change. In the immune system, T cells recognize specific antigens presented on antigen-presenting cells (APCs) and form immunological synapses. Certain T cell receptors (TCRs) undergo conformational changes upon binding of a major histocompatibility complex/antigen peptide (MHCp) ligand, leading to the exposure of a specific epitope on TCRs (Risueño et al., 2005). Similarly, the ligand binding of integrins causes a conformational change in which a bent and low-affinity conformation is converted, by separation of the α- and β-subunits, into an extended state that has a high affinity for extracellular matrixes (Carman and Springer, 2003).

The regulated and stepwise cis interactions of NGL-1, netrin-G1 and LAR might have several advantages. First, this would ensure that LAR, a receptor tyrosine phosphatase that could critically regulate local phospho-protein environments, is clustered only at the sites of axo-dendritic adhesion. Stepwise clustering might create additional sites of regulation for synaptic development. In addition, it could facilitate synaptic disassembly. For example, removal of trans-synaptic ligands by proteolytic cleavage, as recently reported for neuroligin-1 (Peixoto et al., 2012; Suzuki et al., 2012) and NGL-3 (Lee et al., 2013), might rapidly reduce the affinity of cis interactions and trigger the disassembly of adhesion protein complexes.

Netrin-G1 has been associated with Rett syndrome, autism spectrum disorders, schizophrenia, and bipolar disorder (Borg et al., 2005; Eastwood and Harrison, 2008; O'Roak et al., 2012). Many aspects of these disorders involve abnormalities in brain development. Therefore, the critical role of netrin-G1 in mediating NGL-1-dependent LAR clustering and presynaptic differentiation, as identified in this study, might contribute to the development of these disorders.

There are several unanswered questions. Here, we mainly investigated the NGL-1-dependent cis interaction of netrin-G1 with LAR. The LAR family contains two additional members, PTPδ and PTPσ, which are known to regulate presynaptic differentiation by trans-synaptically interacting with diverse postsynaptic adhesion molecules including NGL-3, TrkC, Slitrks, IL1RAPL1 and IL1RAcP (Kwon et al., 2010; Takahashi et al., 2011; Valnegri et al., 2011; Yoshida et al., 2011; Takahashi et al., 2012; Yoshida et al., 2012; Yim et al., 2013). Therefore, PTPδ and PTPσ could also regulate NGL-1-dependent presynaptic differentiation in a cis manner.

Another question is whether trans-induced cis interactions would also be observed in other synaptic adhesion molecules such as neurexins. Although this possibility remains to be explored, it is conceivable that the mechanisms that we proposed in this study cooperate with neurexins for presynaptic differentiation. Known interactions of presynaptic proteins predict that LARs and neurexins cooperate, as we suggested previously (Woo et al., 2009a). Specifically, LAR directly interacts with liprin-α (Pulido et al., 1995) and neurexins indirectly interact with liprin-α through CASK (Hata et al., 1996; Olsen et al., 2005). Given that liprin-α indirectly associates with synaptic vesicles through presynaptic proteins including RIM and ELKS/ERC (Schoch et al., 2002; Ko et al., 2003), LAR and neurexins might converge onto liprin-α for cooperative presynaptic differentiation.

In conclusion, our study proposes a novel principle for synapse formation: trans-synaptic adhesions induce cis interactions with neighboring adhesion molecules to promote synaptic development. In addition, our study suggests LAR as a co-receptor of netrin-G1 and proposes a novel role for synaptic organizers in recruiting neighboring membrane proteins.

DNA constructs and antibodies

A Myc epitope was inserted into the site between amino acid (aa) residues 43 and 44 in human netrin-G1c (BC030220), and between residues 32 and 33 in mouse netrin-G2a (AB052336) in pEGFP-N1. For rat netrin-G1d and mouse netrin-G1m (Meerabux et al., 2005), constructs were subcloned into pEGFP-N1 with N-terminal Myc tagging (EGFP not fused with the protein). Full-length human NGL-1 (NM020929, aa 1–641) and mouse NGL-2 (DQ177325, aa 1–652) were subcloned into pEGFP-N1. For Fc tagging, full-length ectodomains of NGL-1 (aa 1–484) and NGL-2 (aa 1–483) were subcloned into modified pEGFP-N1, where EGFP was replaced with human Fc. Human LAR (Y008115, aa 1–1881) C1522S (phosphatase dead) was subcloned into pEGFP-N1 with C-terminal FLAG tagging. Deletion variants of LAR in pDisplay (LAR-ΔIg and LAR-ΔFn) have been described (Kwon et al., 2010). sh-RNA resistant LAR rescue constructs (LAR-FLAGres) were obtained by introducing silent mutations using QuikChange kit; nucleotides 5461–5479 were replaced by 5′-GGCCTGCACAGCTATACAG-3′. For deletion of V domains in netrin-G1 (ΔV), aa 297–793 were deleted. For LAR deletion constructs, full-length LAR-Ecto (aa 17–1163), LAR-Ig1–3 (aa 35–295) and LAR-FN1–8 (aa 309–1078) were subcloned into pDisplay. Polyclonal EGFP antibodies (#1997) were generated against H6-EGFP (aa 1–240) in guinea pigs. Synapsin I (Millipore), HA (Santa Cruz Biotechnology), FLAG (Sigma), Myc (Santa Cruz Biotechnology), VGlut1 (SYSY), synaptotagmin I luminal domain (SYSY) and LAR (NeuroMab) antibodies were purchased. Netrin-G1 antibody has been described (Nakashiba et al., 2002) we performed preclearing by repeatedly incubating netrin-G1 KO brain slices with the antibodies to increase the specificity.

HEK293T cell adhesion assays

HEK293T cells were cultured on 60 mm dishes and incubated for 24 hours after DNA transfection. One dish was transfected with NGLs, and the other dish was doubly transfected with netrin-G1 or netrin-G2 and LAR. HEK293T cells from the two dishes were dissociated, transferred to 18 mm cover glasses and incubated for 24 hours.

Neuron transfection and immunocytochemistry

Primary hippocampal neurons were prepared from embryonic day 18 rats or postnatal day 0–1 knockout mice. Neurons were transfected using a calcium-phosphate-based mammalian transfection kit (Clontech). Neurons were transfected with DNA constructs at 8–9 DIV and maintained until 11–12 DIV. Soluble preclustered NGL-1-Fc proteins were incubated with neurons at 11–12 DIV for 2 hours. Neurons were fixed with 4% paraformaldehyde, 4% sucrose, permeabilized with 0.2% Triton X-100 in phosphate-buffered saline, and incubated with primary antibodies, followed by secondary Cy3-, Cy5- or FITC-conjugated antibodies (Jackson ImmunoResearch). For netrin-G1 and LAR endostaining, neurons were fixed and permeabilized with methanol for 10 minutes at −20°C. Synaptotagmin I luminal domain uptake assay was performed as previously described (Kraszewski et al., 1995). Briefly, cultured neurons were incubated with synaptotagmin I luminal domain antibodies (1∶10) in Krebs-Ringer HEPES solution for 1 minute.

Image acquisition and quantification

Z-stacked images were randomly captured by confocal microscopy (LSM510, Zeiss) and were analyzed using MetaMorph image analysis software (Universal Imaging). For quantification of preclustered antibody and cell–cell interface signals on HEK cell surfaces, average intensities of the clusters from ∼15 cells were analyzed. For quantification of synapsin I clusters induced by netrin-G1 clustering, captured neuronal images were thresholded, and the integrated intensities of synapsin I clusters were normalized to those of netrin-G1. Values displayed indicate mean ± s.e.m. and statistical significance was determined by one-way ANOVA (Tukey's test).

Author contributions

Y.S. and E.K. designed the experiments, analyzed results and wrote the manuscript; Y.S., H.J and P.P. performed experiments; E.K. and S.I. jointly supervised Y.S.

Funding

This study was supported by the Institute for Basic Science (IBS), and, in part, by the Funding Program for World-Leading Innovative R&D on Science and Technology from the Japanese Society for the Promotion of Science and an Intramural Research Grant from the RIKEN Brain Science Institute.

Aoki-Suzuki
M.
,
Yamada
K.
,
Meerabux
J.
,
Iwayama-Shigeno
Y.
,
Ohba
H.
,
Iwamoto
K.
,
Takao
H.
,
Toyota
T.
,
Suto
Y.
,
Nakatani
N.
 et al. (
2005
).
A family-based association study and gene expression analyses of netrin-G1 and -G2 genes in schizophrenia.
Biol. Psychiatry
57
,
382
393
.
Biederer
T.
,
Sara
Y.
,
Mozhayeva
M.
,
Atasoy
D.
,
Liu
X.
,
Kavalali
E. T.
,
Südhof
T. C.
(
2002
).
SynCAM, a synaptic adhesion molecule that drives synapse assembly.
Science
297
,
1525
1531
.
Blanchetot
C.
,
Tertoolen
L. G.
,
Overvoorde
J.
,
den Hertog
J.
(
2002
).
Intra- and intermolecular interactions between intracellular domains of receptor protein-tyrosine phosphatases.
J. Biol. Chem.
277
,
47263
47269
.
Borg
I.
,
Freude
K.
,
Kübart
S.
,
Hoffmann
K.
,
Menzel
C.
,
Laccone
F.
,
Firth
H.
,
Ferguson-Smith
M. A.
,
Tommerup
N.
,
Ropers
H. H.
 et al. (
2005
).
Disruption of Netrin G1 by a balanced chromosome translocation in a girl with Rett syndrome.
Eur. J. Hum. Genet.
13
,
921
927
.
Brose
N.
(
2009
).
Synaptogenic proteins and synaptic organizers: “many hands make light work”
Neuron
61
,
650
652
.
Carman
C. V.
,
Springer
T. A.
(
2003
).
Integrin avidity regulation: are changes in affinity and conformation underemphasized?
Curr. Opin. Cell Biol.
15
,
547
556
.
Carvalho
R. F.
,
Beutler
M.
,
Marler
K. J.
,
Knöll
B.
,
Becker-Barroso
E.
,
Heintzmann
R.
,
Ng
T.
,
Drescher
U.
(
2006
).
Silencing of EphA3 through a cis interaction with ephrinA5.
Nat. Neurosci.
9
,
322
330
.
Dalva
M. B.
,
McClelland
A. C.
,
Kayser
M. S.
(
2007
).
Cell adhesion molecules: signalling functions at the synapse.
Nat. Rev. Neurosci.
8
,
206
220
.
de Wit
J.
,
Sylwestrak
E.
,
O'Sullivan
M. L.
,
Otto
S.
,
Tiglio
K.
,
Savas
J. N.
,
Yates
J. R.
 3rd
,
Comoletti
D.
,
Taylor
P.
,
Ghosh
A.
(
2009
).
LRRTM2 interacts with Neurexin1 and regulates excitatory synapse formation.
Neuron
64
,
799
806
.
DeNardo
L. A.
,
de Wit
J.
,
Otto-Hitt
S.
,
Ghosh
A.
(
2012
).
NGL-2 regulates input-specific synapse development in CA1 pyramidal neurons.
Neuron
76
,
762
775
.
Dunah
A. W.
,
Hueske
E.
,
Wyszynski
M.
,
Hoogenraad
C. C.
,
Jaworski
J.
,
Pak
D. T.
,
Simonetta
A.
,
Liu
G.
,
Sheng
M.
(
2005
).
LAR receptor protein tyrosine phosphatases in the development and maintenance of excitatory synapses.
Nat. Neurosci.
8
,
458
467
.
Eastwood
S. L.
,
Harrison
P. J.
(
2008
).
Decreased mRNA expression of netrin-G1 and netrin-G2 in the temporal lobe in schizophrenia and bipolar disorder.
Neuropsychopharmacology
33
,
933
945
.
Fogel
A. I.
,
Stagi
M.
,
Perez de Arce
K.
,
Biederer
T.
(
2011
).
Lateral assembly of the immunoglobulin protein SynCAM 1 controls its adhesive function and instructs synapse formation.
EMBO J.
30
,
4728
4738
.
Futai
K.
,
Kim
M. J.
,
Hashikawa
T.
,
Scheiffele
P.
,
Sheng
M.
,
Hayashi
Y.
(
2007
).
Retrograde modulation of presynaptic release probability through signaling mediated by PSD-95-neuroligin.
Nat. Neurosci.
10
,
186
195
.
Han
K.
,
Kim
E.
(
2008
).
Synaptic adhesion molecules and PSD-95.
Prog. Neurobiol.
84
,
263
283
.
Hata
Y.
,
Butz
S.
,
Südhof
T. C.
(
1996
).
CASK: a novel dlg/PSD95 homolog with an N-terminal calmodulin-dependent protein kinase domain identified by interaction with neurexins.
J. Neurosci.
16
,
2488
2494
.
Honkaniemi
J.
,
Zhang
J. S.
,
Yang
T.
,
Zhang
C.
,
Tisi
M. A.
,
Longo
F. M.
(
1998
).
LAR tyrosine phosphatase receptor: proximal membrane alternative splicing is coordinated with regional expression and intraneuronal localization.
Brain Res. Mol. Brain Res.
60
,
1
12
.
Jin
Y.
,
Garner
C. C.
(
2008
).
Molecular mechanisms of presynaptic differentiation.
Annu. Rev. Cell Dev. Biol.
24
,
237
262
.
Kim
S.
,
Burette
A.
,
Chung
H. S.
,
Kwon
S. K.
,
Woo
J.
,
Lee
H. W.
,
Kim
K.
,
Kim
H.
,
Weinberg
R. J.
,
Kim
E.
(
2006
).
NGL family PSD-95-interacting adhesion molecules regulate excitatory synapse formation.
Nat. Neurosci.
9
,
1294
1301
.
Ko
J.
(
2012
).
The leucine-rich repeat superfamily of synaptic adhesion molecules: LRRTMs and Slitrks.
Mol. Cells
34
,
335
340
.
Ko
J.
,
Na
M.
,
Kim
S.
,
Lee
J. R.
,
Kim
E.
(
2003
).
Interaction of the ERC family of RIM-binding proteins with the liprin-alpha family of multidomain proteins.
J. Biol. Chem.
278
,
42377
42385
.
Ko
J.
,
Kim
S.
,
Chung
H. S.
,
Kim
K.
,
Han
K.
,
Kim
H.
,
Jun
H.
,
Kaang
B. K.
,
Kim
E.
(
2006
).
SALM synaptic cell adhesion-like molecules regulate the differentiation of excitatory synapses.
Neuron
50
,
233
245
.
Ko
J.
,
Fuccillo
M. V.
,
Malenka
R. C.
,
Südhof
T. C.
(
2009
).
LRRTM2 functions as a neurexin ligand in promoting excitatory synapse formation.
Neuron
64
,
791
798
.
Kraszewski
K.
,
Mundigl
O.
,
Daniell
L.
,
Verderio
C.
,
Matteoli
M.
,
De Camilli
P.
(
1995
).
Synaptic vesicle dynamics in living cultured hippocampal neurons visualized with CY3-conjugated antibodies directed against the lumenal domain of synaptotagmin.
J. Neurosci.
15
,
4328
4342
.
Kwon
S. K.
,
Woo
J.
,
Kim
S. Y.
,
Kim
H.
,
Kim
E.
(
2010
).
Trans-synaptic adhesions between netrin-G ligand-3 (NGL-3) and receptor tyrosine phosphatases LAR, protein-tyrosine phosphatase delta (PTPdelta), and PTPsigma via specific domains regulate excitatory synapse formation.
J. Biol. Chem.
285
,
13966
13978
.
Lee
H.
,
Lee
E. J.
,
Song
Y.
,
Kim
E.
(
2013
).
LTD-inducing stimuli promote cleavage of the synaptic adhesion molecule NGL-3 through NMDA receptors, matrix metalloproteinases, and presenilin/g-secretase.
Philos. Trans. R. Soc. B.
(in press).
Lin
J. C.
,
Ho
W. H.
,
Gurney
A.
,
Rosenthal
A.
(
2003
).
The netrin-G1 ligand NGL-1 promotes the outgrowth of thalamocortical axons.
Nat. Neurosci.
6
,
1270
1276
.
Linhoff
M. W.
,
Laurén
J.
,
Cassidy
R. M.
,
Dobie
F. A.
,
Takahashi
H.
,
Nygaard
H. B.
,
Airaksinen
M. S.
,
Strittmatter
S. M.
,
Craig
A. M.
(
2009
).
An unbiased expression screen for synaptogenic proteins identifies the LRRTM protein family as synaptic organizers.
Neuron
61
,
734
749
.
Mander
A.
,
Hodgkinson
C. P.
,
Sale
G. J.
(
2005
).
Knock-down of LAR protein tyrosine phosphatase induces insulin resistance.
FEBS Lett.
579
,
3024
3028
.
Marquardt
T.
,
Shirasaki
R.
,
Ghosh
S.
,
Andrews
S. E.
,
Carter
N.
,
Hunter
T.
,
Pfaff
S. L.
(
2005
).
Coexpressed EphA receptors and ephrin-A ligands mediate opposing actions on growth cone navigation from distinct membrane domains.
Cell
121
,
127
139
.
Matsuda
K.
,
Miura
E.
,
Miyazaki
T.
,
Kakegawa
W.
,
Emi
K.
,
Narumi
S.
,
Fukazawa
Y.
,
Ito-Ishida
A.
,
Kondo
T.
,
Shigemoto
R.
 et al. (
2010
).
Cbln1 is a ligand for an orphan glutamate receptor delta2, a bidirectional synapse organizer.
Science
328
,
363
368
.
Meerabux
J. M.
,
Ohba
H.
,
Fukasawa
M.
,
Suto
Y.
,
Aoki-Suzuki
M.
,
Nakashiba
T.
,
Nishimura
S.
,
Itohara
S.
,
Yoshikawa
T.
(
2005
).
Human netrin-G1 isoforms show evidence of differential expression.
Genomics
86
,
112
116
.
Nakashiba
T.
,
Ikeda
T.
,
Nishimura
S.
,
Tashiro
K.
,
Honjo
T.
,
Culotti
J. G.
,
Itohara
S.
(
2000
).
Netrin-G1: a novel glycosyl phosphatidylinositol-linked mammalian netrin that is functionally divergent from classical netrins.
J. Neurosci.
20
,
6540
6550
.
Nakashiba
T.
,
Nishimura
S.
,
Ikeda
T.
,
Itohara
S.
(
2002
).
Complementary expression and neurite outgrowth activity of netrin-G subfamily members.
Mech. Dev.
111
,
47
60
.
Nam
H. J.
,
Poy
F.
,
Krueger
N. X.
,
Saito
H.
,
Frederick
C. A.
(
1999
).
Crystal structure of the tandem phosphatase domains of RPTP LAR.
Cell
97
,
449
457
.
Nishimura-Akiyoshi
S.
,
Niimi
K.
,
Nakashiba
T.
,
Itohara
S.
(
2007
).
Axonal netrin-Gs transneuronally determine lamina-specific subdendritic segments.
Proc. Natl. Acad. Sci. USA
104
,
14801
14806
.
O'Roak
B. J.
,
Vives
L.
,
Girirajan
S.
,
Karakoc
E.
,
Krumm
N.
,
Coe
B. P.
,
Levy
R.
,
Ko
A.
,
Lee
C.
,
Smith
J. D.
 et al. (
2012
).
Sporadic autism exomes reveal a highly interconnected protein network of de novo mutations.
Nature
485
,
246
250
.
Olsen
O.
,
Moore
K. A.
,
Fukata
M.
,
Kazuta
T.
,
Trinidad
J. C.
,
Kauer
F. W.
,
Streuli
M.
,
Misawa
H.
,
Burlingame
A. L.
,
Nicoll
R. A.
 et al. (
2005
).
Neurotransmitter release regulated by a MALS-liprin-alpha presynaptic complex.
J. Cell Biol.
170
,
1127
1134
.
Patel
M. R.
,
Lehrman
E. K.
,
Poon
V. Y.
,
Crump
J. G.
,
Zhen
M.
,
Bargmann
C. I.
,
Shen
K.
(
2006
).
Hierarchical assembly of presynaptic components in defined C. elegans synapses.
Nat. Neurosci.
9
,
1488
1498
.
Peixoto
R. T.
,
Kunz
P. A.
,
Kwon
H.
,
Mabb
A. M.
,
Sabatini
B. L.
,
Philpot
B. D.
,
Ehlers
M. D.
(
2012
).
Transsynaptic signaling by activity-dependent cleavage of neuroligin-1.
Neuron
76
,
396
409
.
Pulido
R.
,
Serra-Pagès
C.
,
Tang
M.
,
Streuli
M.
(
1995
).
The LAR/PTP delta/PTP sigma subfamily of transmembrane protein-tyrosine-phosphatases: multiple human LAR, PTP delta, and PTP sigma isoforms are expressed in a tissue-specific manner and associate with the LAR-interacting protein LIP.1.
Proc. Natl. Acad. Sci. USA
92
,
11686
11690
.
Risueño
R. M.
,
Gil
D.
,
Fernández
E.
,
Sánchez-Madrid
F.
,
Alarcón
B.
(
2005
).
Ligand-induced conformational change in the T-cell receptor associated with productive immune synapses.
Blood
106
,
601
608
.
Schoch
S.
,
Castillo
P. E.
,
Jo
T.
,
Mukherjee
K.
,
Geppert
M.
,
Wang
Y.
,
Schmitz
F.
,
Malenka
R. C.
,
Südhof
T. C.
(
2002
).
RIM1alpha forms a protein scaffold for regulating neurotransmitter release at the active zone.
Nature
415
,
321
326
.
Seiradake
E.
,
Coles
C. H.
,
Perestenko
P. V.
,
Harlos
K.
,
McIlhinney
R. A.
,
Aricescu
A. R.
,
Jones
E. Y.
(
2011
).
Structural basis for cell surface patterning through NetrinG-NGL interactions.
EMBO J.
30
,
4479
4488
.
Shen
K.
,
Scheiffele
P.
(
2010
).
Genetics and cell biology of building specific synaptic connectivity.
Annu. Rev. Neurosci.
33
,
473
507
.
Siddiqui
T. J.
,
Craig
A. M.
(
2011
).
Synaptic organizing complexes.
Curr. Opin. Neurobiol.
21
,
132
143
.
Siddiqui
T. J.
,
Pancaroglu
R.
,
Kang
Y.
,
Rooyakkers
A.
,
Craig
A. M.
(
2010
).
LRRTMs and neuroligins bind neurexins with a differential code to cooperate in glutamate synapse development.
J. Neurosci.
30
,
7495
7506
.
Spangler
S. A.
,
Hoogenraad
C. C.
(
2007
).
Liprin-alpha proteins: scaffold molecules for synapse maturation.
Biochem. Soc. Trans.
35
,
1278
1282
.
Südhof
T. C.
(
2008
).
Neuroligins and neurexins link synaptic function to cognitive disease.
Nature
455
,
903
911
.
Suzuki
K.
,
Hayashi
Y.
,
Nakahara
S.
,
Kumazaki
H.
,
Prox
J.
,
Horiuchi
K.
,
Zeng
M.
,
Tanimura
S.
,
Nishiyama
Y.
,
Osawa
S.
 et al. (
2012
).
Activity-dependent proteolytic cleavage of neuroligin-1.
Neuron
76
,
410
422
.
Takahashi
H.
,
Arstikaitis
P.
,
Prasad
T.
,
Bartlett
T. E.
,
Wang
Y. T.
,
Murphy
T. H.
,
Craig
A. M.
(
2011
).
Postsynaptic TrkC and presynaptic PTPσ function as a bidirectional excitatory synaptic organizing complex.
Neuron
69
,
287
303
.
Takahashi
H.
,
Katayama
K.
,
Sohya
K.
,
Miyamoto
H.
,
Prasad
T.
,
Matsumoto
Y.
,
Ota
M.
,
Yasuda
H.
,
Tsumoto
T.
,
Aruga
J.
 et al. (
2012
).
Selective control of inhibitory synapse development by Slitrk3-PTPdelta trans-synaptic interaction.
Nat. Neurosci
15
,
389
398, S381-S382
.
Taniguchi
H.
,
Gollan
L.
,
Scholl
F. G.
,
Mahadomrongkul
V.
,
Dobler
E.
,
Limthong
N.
,
Peck
M.
,
Aoki
C.
,
Scheiffele
P.
(
2007
).
Silencing of neuroligin function by postsynaptic neurexins.
J. Neurosci.
27
,
2815
2824
.
Uemura
T.
,
Lee
S. J.
,
Yasumura
M.
,
Takeuchi
T.
,
Yoshida
T.
,
Ra
M.
,
Taguchi
R.
,
Sakimura
K.
,
Mishina
M.
(
2010
).
Trans-synaptic interaction of GluRdelta2 and Neurexin through Cbln1 mediates synapse formation in the cerebellum.
Cell
141
,
1068
1079
.
Valnegri
P.
,
Montrasio
C.
,
Brambilla
D.
,
Ko
J.
,
Passafaro
M.
,
Sala
C.
(
2011
).
The X-linked intellectual disability protein IL1RAPL1 regulates excitatory synapse formation by binding PTPδ and RhoGAP2.
Hum. Mol. Genet.
20
,
4797
4809
.
Wang
C. Y.
,
Chang
K.
,
Petralia
R. S.
,
Wang
Y. X.
,
Seabold
G. K.
,
Wenthold
R. J.
(
2006
).
A novel family of adhesion-like molecules that interacts with the NMDA receptor.
J. Neurosci.
26
,
2174
2183
.
Wheldon
L. M.
,
Haines
B. P.
,
Rajappa
R.
,
Mason
I.
,
Rigby
P. W.
,
Heath
J. K.
(
2010
).
Critical role of FLRT1 phosphorylation in the interdependent regulation of FLRT1 function and FGF receptor signalling.
PLoS ONE
5
,
e10264
.
Williams
M. E.
,
de Wit
J.
,
Ghosh
A.
(
2010
).
Molecular mechanisms of synaptic specificity in developing neural circuits.
Neuron
68
,
9
18
.
Woo
J.
,
Kwon
S. K.
,
Kim
E.
(
2009a
).
The NGL family of leucine-rich repeat-containing synaptic adhesion molecules.
Mol. Cell. Neurosci.
42
,
1
10
.
Woo
J.
,
Kwon
S. K.
,
Choi
S.
,
Kim
S.
,
Lee
J. R.
,
Dunah
A. W.
,
Sheng
M.
,
Kim
E.
(
2009b
).
Trans-synaptic adhesion between NGL-3 and LAR regulates the formation of excitatory synapses.
Nat. Neurosci.
12
,
428
437
.
Wyszynski
M.
,
Kim
E.
,
Dunah
A. W.
,
Passafaro
M.
,
Valtschanoff
J. G.
,
Serra-Pagès
C.
,
Streuli
M.
,
Weinberg
R. J.
,
Sheng
M.
(
2002
).
Interaction between GRIP and liprin-alpha/SYD2 is required for AMPA receptor targeting.
Neuron
34
,
39
52
.
Ye
H.
,
Tan
Y. L.
,
Ponniah
S.
,
Takeda
Y.
,
Wang
S. Q.
,
Schachner
M.
,
Watanabe
K.
,
Pallen
C. J.
,
Xiao
Z. C.
(
2008
).
Neural recognition molecules CHL1 and NB-3 regulate apical dendrite orientation in the neocortex via PTP alpha.
EMBO J.
27
,
188
200
.
Yim
Y. S.
,
Kwon
Y.
,
Nam
J.
,
Yoon
H. I.
,
Lee
K.
,
Kim
D. G.
,
Kim
E.
,
Kim
C. H.
,
Ko
J.
(
2013
).
Slitrks control excitatory and inhibitory synapse formation with LAR receptor protein tyrosine phosphatases.
Proc. Natl. Acad. Sci. USA
110
,
4057
4062
.
Yin
Y.
,
Miner
J. H.
,
Sanes
J. R.
(
2002
).
Laminets: laminin- and netrin-related genes expressed in distinct neuronal subsets.
Mol. Cell. Neurosci.
19
,
344
358
.
Yoshida
T.
,
Yasumura
M.
,
Uemura
T.
,
Lee
S. J.
,
Ra
M.
,
Taguchi
R.
,
Iwakura
Y.
,
Mishina
M.
(
2011
).
IL-1 receptor accessory protein-like 1 associated with mental retardation and autism mediates synapse formation by trans-synaptic interaction with protein tyrosine phosphatase δ.
J. Neurosci.
31
,
13485
13499
.
Yoshida
T.
,
Shiroshima
T.
,
Lee
S. J.
,
Yasumura
M.
,
Uemura
T.
,
Chen
X.
,
Iwakura
Y.
,
Mishina
M.
(
2012
).
Interleukin-1 receptor accessory protein organizes neuronal synaptogenesis as a cell adhesion molecule.
J. Neurosci.
32
,
2588
2600
.
Yuzaki
M.
(
2011
).
Cbln1 and its family proteins in synapse formation and maintenance.
Curr. Opin. Neurobiol.
21
,
215
220
.

Supplementary information