• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of nihpaAbout Author manuscriptsSubmit a manuscriptNIH Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Neuron. Author manuscript; available in PMC Mar 12, 2010.
Published in final edited form as:
PMCID: PMC2746109
NIHMSID: NIHMS99494

An Unbiased Expression Screen for Synaptogenic Proteins Identifies the LRRTM Protein Family as Synaptic Organizers

Summary

A delineation of the molecular basis of synapse development is crucial for understanding brain function. Co-cultures of neurons with transfected fibroblastoid cells have been used to demonstrate the synapse-promoting activity of candidate molecules. Here, we performed an unbiased expression screen for synaptogenic proteins in the co-culture assay using custom-made full-length cDNA libraries. Re-isolation of NGL-3/LRRC4B and neuroligin-2 accounts for a minority of positive clones, indicating that current understanding of mammalian synaptogenic proteins is far from complete. We identify LRRTM1 as a novel transmembrane protein capable of inducing presynaptic differentiation in contacting axons. All four LRRTM family members exhibit synaptogenic activity, LRRTMs localize to excitatory synapses, and artificially-induced clustering of LRRTMs mediates postsynaptic differentiation in dendrites. We generate LRRTM1 -/- mice and reveal altered distribution of the vesicular glutamate transporter VGLUT1, confirming an in vivo synaptic function. These results suggest a prevalence of LRR domain proteins in trans-synaptic signaling and provide a cellular basis for the recently reported linkage of LRRTM1 to handedness and schizophrenia.

Introduction

Chemical synapses represent the principal means of communication between neurons. While significant progress has been made in determining the developmental signals that guide axons to their targets, the molecular mechanisms that determine the final establishment of circuits remain incompletely understood. Synapses are asymmetric cellular junctions composed of a presynaptic vesicle release site, a synaptic cleft, and a specialized postsynaptic receptive apparatus. A major challenge in elucidating the mechanisms that govern synapse formation and maturation is the enormous variety of synapse types in the mammalian brain. It is likely that this diversity requires contribution from a large number of molecular signals. In addition to neuronal transmembrane proteins, secreted factors such as FGFs and neuronal pentraxins, and glial-derived factors such as thrombospondin and complement cascade proteins shape synapse development (Stevens et al., 2007; Waites et al., 2005).

Significant attention has focused on the role of synaptic cell adhesion molecules (CAMs) in synapse development (Ackley and Jin, 2004; Dalva et al., 2007; Ko and Kim, 2007; Yamagata et al., 2003). Given the potential for a trans-synaptic interaction that could bridge the presynaptic release apparatus with the postsynaptic density, Scheiffele and colleagues tested whether neuroligin expressed in transfected HEK cells could assemble presynaptic terminals when co-cultured with axons from pontine explants. Indeed, expression of neuroligin in HEK cells induced the clustering of synaptic vesicles in contacting axons (Scheiffele et al., 2000), via direct interaction with neurexins. In a parallel fashion, expression of β-neurexin in COS cells induces postsynaptic differentiation in contacting dendrites (Graf et al., 2004). These results show that a single molecular interaction can organize many aspects of presynaptic and postsynaptic assembly. Further studies indicate selective roles of different neuroligin and neurexin splice isoforms at excitatory versus inhibitory synapses (Chubykin et al., 2007; Craig and Kang, 2007). Moreover, analyses of knockout mice indicate an essential role for neurexins and neuroligins in synapse development (Missler et al., 2003; Varoqueaux et al., 2006). Neuroligin-1,2,3 triple knockout mice die at birth due to defects in excitatory and inhibitory transmission, although single neuroligin knockouts have only subtle phenotypes (Chubykin et al., 2007; Jamain et al., 2008; Varoqueaux et al., 2006).

The fibroblast-neuron co-culture assay has proven to be a useful tool in testing candidate proteins for a role in synaptic development (Biederer and Scheiffele, 2007; Craig et al., 2006). Co-culture assays have revealed synaptogenic activity for four families of neuronal CAMs, and in some cases their binding partners: neuroligins and partner neurexins, SynCAMs/Necls (Biederer et al., 2002), EphBs and partner ephrinBs (Aoto et al., 2007; Kayser et al., 2006), and netrin G ligands (NGLs/LRRC4s) (Kim et al., 2006). Knockout mice studies have validated roles for EphBs/ephrinBs (Aoto et al., 2007; Henderson et al., 2001; Henkemeyer et al., 2003; Kayser et al., 2006) as well as neurexins/neuroligins in synapse development. Evidence is accumulating that proteins testing positive for synaptogenic activity in this assay may not be essential for initiating the formation of or maintaining the integrity of synaptic junctions. Rather, these factors may serve a role in the maturation of the synapse, recruiting components necessary for synaptic function.

Here we used the fibroblast-neuron co-culture assay to search for novel synaptogenic proteins. We created and screened a set of full-length size-selected cDNA expression libraries from developing rat brain. Screening to date reveals a prevalence of leucine rich repeat (LRR) synaptogenic proteins arising from this unbiased approach. We identify the LRRTM (Leucine Rich Repeat Transmembrane Neuronal) protein family as able to instruct excitatory presynaptic differentiation and to mediate postsynaptic differentiation. Furthermore, we generate an LRRTM1 knockout mouse and reveal a modest synaptic phenotype, suggesting a synaptic basis for the recently reported linkage of LRRTM1 to handedness and schizophrenia (Francks et al., 2007). Parts of this work have previously been published in doctoral dissertations (Linhoff, 2008; Lauren, 2007).

Results

Expression Screening for Molecules Involved in Presynaptic Differentiation

To test the feasibility of converting the fibroblast-neuron co-culture assay into an expression screen, neuroligin-1 and neuroligin-2 cDNAs were diluted up to 1:1000 with either a soluble CFP expression plasmid or a pool of inactive cDNAs. Neuroligin-2 could be diluted 500-fold while still consistently yielding a positive synaptogenic signal on a single neuron-COS cell co-culture coverslip, while neuroligin-1 could only be diluted 100-fold (data not shown). Thus a pool size of ~250 clones would be likely to yield novel clones with synaptogenic activity of similar potency as neuroligins. This pool size of 250 is considerably smaller than typical pool sizes of 1,000 - 10,000 used for expression screens involving more sensitive detection technologies (e.g. Fournier et al., 2001). Furthermore, since we are expecting active proteins to be transmembrane or secreted, positive clones would have to be full-length to contain the signal peptide essential for proper trafficking.

Thus the first step in the screen was to generate an unamplified cDNA expression library with a high percentage of full-length clones (Figure 1A). We isolated mRNA from rat forebrain at P11, the peak of synaptogenesis (Micheva and Beaulieu, 1996). To ensure representation of full-length molecules, we used the biotinylated cap-trapper method that had been developed for large-scale sequencing projects (Figure S1A,B) (Carninci et al., 1996). The resultant cap-trapped cDNA was subjected to rigorous size fractionation and individual unamplified libraries were maintained. Mean insert size of the resultant libraries was confirmed by restriction digestion (Figure S1C), and sequencing of over 50 clones indicated that all contained the 5′ ATG. We were successful in creating high quality expression libraries for inserts up to 5 kb using pcDNA3.1 as the host expression vector (Figure 1A).

Figure 1
Expression Screening Identifies LRRTM1 as a Synaptogenic Factor

Figure 1B illustrates the protocol that we used for the screen to identify cDNA pools that contained synaptogenic activity. Synaptogenic activity was assayed by the immunocytochemical detection of experimentally induced hemisynapses. The co-cultures were immunostained with synapsin I antibody to detect synaptic vesicle clustering at contact sites between COS cells and axons. Co-immunostaining for the postsynaptic markers PSD-95 family and gephyrin ensured that synapsin immunoreactivity associated with bona fide interneuronal synapses was not counted as a false positive. A cDNA pool was considered positive if it generated any COS cell with significant associated synapsin clusters unapposed to PSD-95 family or gephyrin. We screened positive cDNA pools using PCR to detect known synaptogenic factors. The first positive pool, PD026, was found in the 4-5 kb library, tested positive for neuroligin-2 using PCR, and was not fractionated to isolate the single responsible synaptogenic cDNA. Further screening within the 3-4 kb library resulted in another positive pool, PC064, which tested PCR negative for neuroligins. We subdivided the PC064 cDNA pool to isolate the synaptogenic clone. Generation of arrayed clones and DNA from a positive subpool aided in identifying the active clone. The isolated clone, PC064-89-29-40 (pool PC064, subpool 89, subpool 29, clone 40), was identified as leucine rich repeat transmembrane protein LRRTM1. The LRRTM protein family consists of four members each possessing 10 extracellular leucine rich repeats, a single transmembrane domain, and a ~70 amino acid cytoplasmic domain (Figure 1B and Figure S2) (Lauren et al., 2003).

Quantitation of the Synaptogenic Activity for LRRTM Family Members

We cloned each of the LRRTM family members from rat cDNA, fused CFP to the C-terminus, and tested each protein for synaptogenic activity in the co-culture assay. N-cadherin has been shown previously to not induce presynaptic clustering (Scheiffele et al., 2000); we confirmed this (see below) and used N-cadherin as a negative control. As an additional control we used the LRR protein, AMIGO, which has been shown to interact with the axons of hippocampal neurons and induce neurite outgrowth and fasciculation (Kuja-Panula et al., 2003).

To obtain a quantitative measure of each protein's ability to instruct presynaptic differentiation, we measured the amount of synapsin clustering associated with transfected COS cells and not associated with MAP2-positive dendrites to exclude interneuronal synapses. Robust synaptogenic activity was observed for LRRTM1-CFP, LRRTM2-CFP, LRRTM4-CFP, and Neuroligin-2-CFP (Figure 2A,B and Figure S3B). In some instances clustering of the CFP fusion protein could be observed apposed to synapsin puncta (Figure 2A,B insets). LRRTM3-CFP consistently yielded limited activity in the co-culture assay (Figure S3A). In contrast to the LRRTMs, synaptogenic activity was not observed for N-cadherin-CFP or the LRR-containing AMIGO-CFP (see Figure 2C for images of AMIGO-CFP). LRRTM1-CFP, LRRTM2-CFP, LRRTM4-CFP, and Neuroligin-2-CFP each induced a >14 fold increase in synapsin clustering when compared to results obtained from N-cadherin-CFP or AMIGO-CFP transfected cells (Figure 2D, n=20 transfected COS cells per construct; ANOVA p<0.0001; t-test versus N-cadherin-CFP and AMIGO-CFP p<0.005). In contrast, synapsin clustering associated with LRRTM3-CFP was 1.6 to 3.3 fold above that associated with N-cadherin-CFP or AMIGO-CFP (t-test versus N-cadherin and AMIGO p<0.05). In separate experiments to confirm the lack of presynaptic inducing activity of N-cadherin compared with no surface protein expression, we compared the percentage of cells exhibiting any clusters of synapsin not associated with PSD-95 or gephyrin for COS cells expressing N-cadherin-CFP (1.6 ± 0.3 %) or expressing CFP (1.8 ± 0.5 %; p>0.1).

Figure 2
Quantitation of the Synaptogenic Activity of LRRTM Family Members

Axonal contact area determined from dephospho-tau immunoreactivity was also higher for LRRTM2-CFP, LRRTM4-CFP, and neuroligin-2-CFP than for N-cadherin-CFP or AMIGO-CFP (Figure S3C; ANOVA p<0.0001), suggesting an effect on axon adhesion as well as presynaptic differentiation. The less than two-fold increase in axon contact area for LRRTM2-CFP or LRRTM4-CFP was not sufficient to explain the >14-fold increase in synapsin clustering, indicating a direct presynaptic effect. Thus, measures of synapsin area normalized per axon contact area also revealed significant induction of presynaptic differentiation by LRRTMs (Figure 2E). Although the most synaptogenic factors tended to have increased axon contact area, the same phenomenon did not hold true for LRRTM1-CFP. This result is likely due to the reduced surface expression of LRRTM1 relative to the other family members. We observed that LRRTM1 accumulates extensively in the endoplasmic reticulum of transfected non-neuronal cells (data not shown), and this result has been observed previously (Francks et al., 2007).

Since synapsin is a vesicle associated protein, we also tested for other markers of presynaptic differentiation. LRRTMs induced clustering of the active zone cytomatrix protein, bassoon (Figure S4A-D), as well as the integral synaptic vesicle membrane protein, synaptophysin (Figure S4E-H).

LRRTMs Instruct Excitatory Presynaptic Differentiation

We tested whether the synaptogenic activity was restricted to excitatory or inhibitory presynaptic differentiation. The CFP fusion proteins were scored blind for significant clustering of the presynaptic vesicular neurotransmitter transporters, VGLUT1 and VGAT, to identify excitatory or inhibitory presynaptic differentiation respectively. Significant clustering of VGLUT1 unapposed to PSD-95 family was observed for LRRTM1-CFP, LRRTM2-CFP, LRRTM3-CFP, LRRTM4-CFP, and neuroligin-2-CFP (Figure 2F-J; ANOVA p<0.0001; t-test versus N-cadherin-CFP and AMIGO-CFP p<0.0005). In contrast, only LRRTM2-CFP and neuroligin-2 showed robust clustering of VGAT unapposed to gephyrin (Figure S3D-F; ANOVA p<0.0001).

LRRTM Instructs the Development of Functional Glutamate Release Sites

Since LRRTM1 and LRRTM2 demonstrated the most potent synaptogenic activity of the four family members, we focused further efforts on these two family members. To test whether the induced presynaptic clusters were functional for release, we used whole cell patch voltage clamp to record currents from HEK293T cells co-transfected with LRRTM2-CFP and NMDA receptor subunits YFP-NR1 and NR2A and co-cultured with hippocampal neurons. A similar assay was used to show that neuroligin-1 co-expressed with NMDA receptors could induce excitatory postsynaptic current (EPSC)-like events in HEK cells co-cultured with cerebellar granule neurons (Fu et al., 2003). Here, bursts of spontaneous activity much like NMDA receptor currents typical of cultured neurons were recorded from the HEK293T cells expressing LRRTM2-CFP and NMDA receptors (Figure 3A). These bursts of high amplitude events were abolished by tetrodotoxin (TTX) and reduced to miniature EPSC-like events, indicating that the bursts arise from action potential dependent neuronal network activity which drives glutamate release onto the transfected cell. In contrast, HEK293T cells co-transfected with YFP-NR1, NR2A, and N-cadherin-CFP as a negative control exhibited only rare spontaneous events and lacked the bursting activity observed for LRRTM2-CFP expressing cells (Figure 3B). These rare synaptic-like events may correspond to fusion of immature synaptic vesicles or may arise from orphan release sites of axons (Krueger et al., 2003). The frequency and amplitude of events onto N-cadherin expressing cells was not significantly altered by TTX. Amplitude of events onto LRRTM2-CFP expressing cells was reduced 4.6-fold by TTX (n = 5 cells with frequency >0.1 Hz), indicating the formation of multiple functional release sites. Quantitative analysis confirmed a greater than 20-fold overall increase in frequency of spontaneous EPSC-like events and 6.6-fold increase in frequency of miniature EPSC-like events in HEK293T cells expressing LRRTM2-CFP compared with N-cadherin-CFP expressing cells (Figure 3C; n ≥ 10, p<0.005 t-test). The amplitude of spontaneous but not miniature events was also significantly increased for LRRTM2-CFP versus N-cadherin-CFP expressing cells (Figure 3D; p<0.05 t-test comparison of mean cell values). Mean amplitude of miniature events was 89 and 99 pA for N-cadherin or LRRTM2-CFP expressing cells, respectively, suggesting that sufficient levels of diffuse surface NMDA receptors were obtained in the HEK cells to provide a sensitive assay of release. All synaptic-like events were abolished by NMDA receptor antagonist APV (data not shown). Thus the presynaptic specializations induced by LRRTM2 expression are functional with respect to evoked and spontaneous glutamatergic synaptic vesicle exocytosis.

Figure 3
LRRTM Instructs Assembly of Functional Vesicle Release Sites

The LRR Domain of LRRTM is Necessary and Sufficient for Synaptogenic Activity

To determine if the LRR domain was necessary for synaptogenic activity, a deletion mutant lacking the LRR domain, ΔLRR-LRRTM2-CFP, was tested in the co-culture assay. The deletion mutant was expressed on the cell surface, but presynaptic differentiation was not observed in axons that contacted the surface of transfected COS cells (Figure S5). To determine whether the LRR domain of LRRTM2 is sufficient to instruct presynaptic differentiation, a fusion protein of the LRR domain of LRRTM2 with a myc-epitope tagged placental alkaline phosphatase (AP), LRRTM2 LRR-AP, was linked via biotinylated anti-myc antibody onto Neutravidin beads. Contact of these LRR-AP coated beads with isolated axons of hippocampal neurons induced clustering of synapsin or VGLUT1 at contact sites (Figure 4A,B, right panels). Beads coated with the control AP protein did not display synaptogenic activity (Figure 4A,B, left panels). Random counts revealed synapsin clustering at 34.1 % of LRR-AP bead contacts and only 2.1 % of control AP bead contacts (n = 50 fields). The mean synapsin intensity under all LRR-AP beads was 6.3-fold higher than that under control AP beads (p<0.0001). Clustering of the active zone marker, bassoon, could also be detected at LRR-AP bead-axon contact sites. In fact, there appeared to be separation between the bassoon labeled active zone and the VGLUT1 positive vesicle pool, with the active zone most closely apposed to the bead surface (Figure 4B, right panel, inset). The LRR-AP beads did not appear to cluster VGAT at contact sites with GAD positive axons (data not shown). These results show that the LRR domain of LRRTM2 is necessary and sufficient to induce excitatory presynaptic differentiation without any contribution from other factors.

Figure 4
The LRR Domain of LRRTM Instructs Presynaptic Differentiation

Localization of Recombinant LRRTMs to Excitatory Synapses Independently of the PDZ Domain Binding Site

To begin to assess subcellular localization, we created extracellular YFP fusions of LRRTM1 and LRRTM2 and transfected hippocampal neurons in culture. YFP-LRRTM-1 or -2 expressed at low level in hippocampal neurons traffic to dendrites and assume a punctate, synaptic-like pattern (Figure 5A-C). YFP-LRRTM-1 or -2 colocalized with PSD-95 family proteins opposite VGLUT1 puncta. YFP-LRRTM-1 and -2 appeared to be exclusively co-localized with the excitatory postsynaptic scaffolding proteins of the PSD-95 family but not with the inhibitory postsynaptic scaffolding protein, gephyrin. Quantitation of thresholded puncta revealed 72.6% overlap of YFP-LRRTM2 puncta with PSD-95 family, compared with 4.5% overlap of mirror images as control, and 8.0% overlap with gephyrin. These results suggest that the function of LRRTMs may be restricted to glutamatergic and not GABAergic hippocampal synapses, although more definitive tests and more diverse contexts are required.

Figure 5
Recombinant LRRTMs Localize to Excitatory Postsynaptic Sites Independently of PDZ Domain Binding

The C-terminus of all four LRRTM members ends in a pattern of residues –E-C-E-V that resembles the X-S/T-X-V Class I PDZ domain ligand pattern (Sheng and Sala, 2001). Furthermore, a glutamate at the -3 position is preferred for PDZ domains 1 and 2 of PSD-95, for which several natural ligands terminate in –E-S/T-D-V (Sheng and Sala, 2001). Although a cysteine has not been reported in the -2 position for PDZ domain ligands, the similarity prompted us to test whether LRRTMs might bind PSD-95. We tested whether YFP-LRRTM2 could co-localize with PSD-95 in COS cells. YFP-LRRTM2 frequently forms small clusters in transfected COS cells. When PSD-95-mRFP was co-transfected with YFP-LRRTM2, the two proteins co-localized precisely in larger discrete aggregates (Figure 5D). Similar co-clustering has been observed for co-expressed PSD-95 and other ligands and is thought to reflect the formation of large complexes via multimerization of PSD-95 and binding to ligand. Deletion of the C-terminal E-C-E-V residues of YFP-LRRTM2 disrupted the co-localization with PSD-95 (Figure 5E). We also confirmed the interaction by co-immunoprecipitation of myc-tagged PSD-95 with YFP-LRRTM2 co-expressed in COS cells (Figure 5F). Again deletion of the C-terminal E-C-E-V residues of YFP-LRRTM2 resulted in loss of the interaction. Thus LRRTM2 can interact with PSD-95 via its C-terminal E-C-E-V. However, we were unable to co-immunoprecipitate LRRMT2 and PSD-95 family proteins from rodent brain (data not shown). The major in vivo interacting partner for LRRTM2 may be a different PDZ domain protein, or interaction with the PSD-95 family may be regulated in vivo, perhaps via modification of the -2 cysteine.

To determine whether the PDZ domain binding site is essential for postsynaptic clustering of LRRTMs, we again assessed localization of YFP-LRRTM2 constructs in cultured neurons. Deletion of the C-terminal E-C-E-V did not abolish clustering of YFP-LRRTM2ΔPDZ at glutamate postsynaptic sites colocalizing with PSD-95 (Figure 5G,H). However, deletion of the C-terminal 55 residues leaving only 17 residues after the transmembrane domain did abolish clustering of YFP-LRRTM2ΔC-term (Figure 5I). Deletion of this C-terminal domain also abolished polar distribution of the LRRTM2 protein, resulting in detection of the YFP-LRRTM2ΔC-term on the surface of axons as well as dendrites (Figure S6). Thus the C-terminus of LRRTM2 but not the specific PDZ domain binding site is required for postsynaptic clustering and for polarized distribution.

Overexpression of LRRTM in Neurons Increases Clustering of Presynaptic Antigens Via the Extracellular Domain

We wondered whether high level expression of YFP-LRRTM2 in neurons might increase presynaptic clustering onto the expressing neurons, similar to effects in the COS cell co-cultures. Indeed, high level expression of YFP-LRRTM2 in neurons resulted in greatly enhanced clustering of bassoon and synapsin onto the expressing neurons in comparison with non-transfected neighbor neurons (Figure S6). Enhanced presynaptic clustering was also induced by YFP-LRRTM2ΔC-term and thus mediated by the extracellular domain of LRRTM2 (Figure S6). Although YFP-LRRTM2ΔC-term reached the surface of axons as well as dendrites, the enhanced presynaptic clustering was largely limited onto the somatodendritic domain, perhaps via other mechanisms promoting axon-dendrite contact over axon-axon contact.

Postsynaptic Differentiation Mediated by LRRTMs

We next determined whether LRRTMs may be able to mediate postsynaptic as well as induce presynaptic differentiation. We tested whether artificial clustering of LRRTMs at non-synaptic sites on dendrites might result in co-clustering of postsynaptic proteins. Hippocampal neurons were transfected to express low levels of YFP-LRRTM1 or YFP-LRRTM2, and then cultured in the presence of beads coated with anti-GFP antibodies. The antibody coated beads induced robust clustering of YFP-LRRTMs at non-synaptic sites (sites lacking VGLUT1 or VGAT) on the dendrites. Such artificial clustering of YFP-LRRTMs resulted in co-clustering of the glutamatergic postsynaptic proteins NMDA receptor essential subunit NR1 (Figure 6A,B), PSD-95 family (Figures 6C and S7A), and SynGAP (Figure S7C). In contrast, artificial clustering of YFP-LRRTMs did not result in co-clustering of the inhibitory postsynaptic protein, gephyrin (Figure 6D and S7B). NR1 immunofluorescence at bead-induced YFP-LRRTM2 clusters showed a 3.2-fold enrichment compared with the rest of the dendrite, and PSD-95 a 2.9-fold enrichment (p<0.005, n = 6-21 cells), whereas gephyrin immunofluorescence was not enriched. These results suggest that LRRTMs may be involved in bidirectional signaling to recruit both postsynaptic and presynaptic components.

Figure 6
Clustering of LRRTMs Mediates Postsynaptic Differentiation

Localization of Endogenous LRRTM2 to Excitatory Synapses In Vivo

In order to assess the endogenous localization of LRRTM2, we generated rabbit polyclonal antibodies against peptide sequences in the C-terminal cytoplasmic tail of LRRTM2 (Figures 7A and S8). All four LRRTM mRNAs could be detected in cultured hippocampal neurons by RT-PCR (data not shown). However, with the LRRTM2 antibody only a weak signal was detected by Western blot and no endogenous immunofluorescence was detected in cultured neurons (data not shown), even though these antibodies could detect LRRTM2 immunofluorescence in YFP-LRRTM2 transfected cells (Figure S8D-G) and in brain sections (Figure 7C-F). Thus expression level of LRRTMs is low in cultured neurons.

Figure 7
Localization of LRRTM2 to Excitatory Synapses In Vivo

To address the localization of LRRTM2 in the brain, we first performed synaptic fractionation from adult rat brain. Biochemical fractionation of whole brain homogenate shows a strong enrichment of LRRTM2 in the PSD fraction, the detergent-resistant material following extraction of the synaptic plasma membrane fraction (Figure 7B). This PSD fraction is also highly enriched in PSD-95. We then examined localization of endogenous LRRTM2 in tissue sections using confocal microscopy. LRRTM2 is widely distributed in neuropil regions in discrete puncta throughout the brain. Consistent with the in situ hybridization data (Lauren et al., 2003), LRRTM2 protein was detected in cortex, thalamus, striatum, olfactory bulb, cerebellum, and all hippocampal subfields (Figure 7C,D and data not shown). Also consistent with LRRTM2 mRNA distribution, LRRTM2 protein is more abundant in deep than in superficial layers of neocortex. The protein further exhibits laminar-specific concentrations independent of mRNA distribution. In the CA1 region of hippocampus, LRRTM2 is more concentrated in stratum lacunosum moleculare than in radiatum or oriens (Figure 7C), and in CA3 LRRTM2 is most concentrated in stratum lucidum (Figure 7E). High expression of LRRTM2 mRNA by cerebellar granule cells corresponds to abundance of the protein at cerebellar glomeruli (Figure 7F). At these large discrete cerebellar glomerular rosettes and CA3 mossy fiber synapses, LRRTM2 immunofluorescence overlapped significantly with the general synaptic marker bassoon and the excitatory synaptic marker VGLUT1. Thus both biochemical and immunofluorescence data supports localization of LRRTM2 to glutamatergic postsynaptic sites in adult rodent brain.

Altered Distribution of VGLUT1 in LRRTM1 -/- Mice

As shown above, all four LRRTMs have synaptogenic activity. LRRTM1 and LRRTM2 are most potent at inducing presynaptic differentiation, target to glutamate postsynaptic sites, and clustering of either on dendrites mediates postsynaptic differentiation. Expression patterns of the four LRRTMs overlap extensively in vivo (Lauren et al., 2003). Thus it may require knockout of multiple family members to reveal a strong phenotype. We started an in vivo analysis of function here by generating genetically targeted mice lacking the family member isolated from the screen, LRRTM1 (Figure 8A-C). LRRTM1 -/- mice survived in the expected Mendelian ratios, were fertile, and displayed no overt phenotype. Brain morphology appeared grossly normal as assessed by cresyl violet stain (excepting that one of twelve LRRTM1 -/- mice showed anomalous ventroculomegaly, data not shown). Cellular organization and synaptic distribution assessed by DAPI stain and bassoon immunofluorescence were indistinguishable from wild type. Figure 8D shows images from the hippocampal formation which along with thalamus normally expresses the highest levels of LRRTM1 (Lauren et al., 2003).

Figure 8
Normal Brain Morphology but Altered VGLUT1 Immunofluorescence in LRRTM1 -/- Mice

A quantitative immunofluorescence analysis revealed a synaptic defect in specific regions of the hippocampal formation in LRRTM1 -/- young adult mice (Figure 8E-G). The measures revealed a selective increase in the size of VGLUT1 puncta in CA1 stratum radiatum and stratum oriens, but no change in CA1 stratum lacunosum moleculare or CA3 stratum lucidum. Puncta density and intensity did not differ (Figure S9). The finding that immunofluorescence for the active zone protein bassoon was not significantly different, in puncta size, density, or intensity, in these same double labeled samples serves as an internal procedural control and indicates a selective change in the glutamatergic vesicle localized VGLUT1. The increased size of fluorescence puncta for VGLUT1 may indicate a dispersal of synaptic vesicles. A similar increase in size of fluorescence puncta for synaptophysin in neurons lacking β-catenin corresponded to a dispersal of synaptic vesicles and reduction in reserved pool vesicles (Bamji et al., 2003). LRRTM2 is highly co-expressed in CA1 with LRRTM1, and both LRRTM2 and LRRTM4 are co-expressed in CA3 (Lauren et al., 2003). Specifically the region of CA1 with the highest level of LRRTM2 immunoreactivity, lacunosum moleculare, did not show any alteration in VGLUT1 distribution in the LRRTM1 -/- mice, synaptic defects were only seen in radiatum and oriens which have relatively lower levels of LRRTM2 (See Figure 7C for LRRTM2 immunoreactivity in CA1 regions).

Prevalence of Novel and LRR Proteins from the Unbiased Screen

While these results validate LRRTMs as novel synaptogenic proteins, we sought to determine the extent to which known synaptogenic proteins plus LRRTMs explain the spectrum of brain synaptogenic activity. A total of 521 pools of cDNA expression plasmids, representing 123,050 independent clones, were tested in the co-culture assay. Eight pools scored positive for ability to cluster synapsin in contacting axons, in the absence of postsynaptic markers (Table 1). Four pools contained known synaptogenic proteins, neuroligin-2 (Figure S10; Scheiffele et al., 2000) or NGL-3/LRRC4B (Figure S11; Kim et al., 2006). Thus, two of the four previously identified synaptogenic families (neuroligins, EphBs, SynCAMs, and NGLs) were re-isolated in this screen. The additional positive pools do not contain known synaptogenic proteins or LRRTMs as assessed by PCR, and presumably include novel clones. A majority of the synaptogenic proteins scoring positive in this unbiased screen were previously unidentified using other approaches. Of the three synaptogenic protein families identified here, all contain C-terminal PDZ domain binding sequences conserved among all family members, and two contain extracellular leucine-rich repeats (Table 1).

Table 1
Results from COS Cell and Neuron Co-Culture Screen of Expression Libraries

We next tested whether the results of this unbiased screen could be used to further predict additional synaptogenic proteins. New candidates were chosen based on primary sequence by the presence of predicted extracellular LRR domains, a single transmembrane domain, and a C-terminal sequence bearing a consensus PDZ domain binding site and/or highly conserved among family members and species. One of the six additional candidates chosen, Slitrk2, tested positive in the co-culture assay for ability to induce clusters of synapsin in contacting hippocampal axons (Figure S12).

Discussion

We performed here a function-based discovery screen in a mammalian system for proteins that can instruct presynaptic differentiation. In addition to isolating positive pools containing neuroligin-2 and NGL-3 cDNAs, we identified a novel synaptogenic cDNA, LRRTM1. We show that LRRTM1 instructs excitatory presynaptic differentiation in contacting axons, mediates excitatory postsynaptic differentiation in dendrites, localizes to excitatory postsynaptic sites, and is required for normal VGLUT1 distribution in vivo. These results provide a cellular basis for the linkage of LRRTM1 to handedness and schizophrenia (Francks et al., 2007). Results from this unbiased screen suggest that only about half of mammalian synaptogenic proteins have been found by other approaches, and reveal a prevalence of LRR proteins as synaptic organizers. Thus the first major contribution of this work is to develop a function-based unbiased screen to identify novel mammalian synaptic proteins. The second major contribution is to characterize the synapse organizing activity of LRRTM proteins.

When this project was initiated, there was considerable uncertainty as to whether neuroligin was unique as a synaptogenic factor or whether it was one of several such proteins. While an unbiased search for such activity was likely to provide the most convincing resolution of this issue, it remained unclear if a broad synaptogenesis screen was feasible. After screening over 105 cDNAs, we can conclude that multiple synaptogenic protein families exist, and that our knowledge of synaptogenic factors does not yet include a full catalogue of relevant cell surface signals. The unbiased screen as performed here isolated only the most potent synaptogenic factors. We identified neuroligin-3 by PCR in a pool that was scored as negative for synaptogenic activity in the screen. Yet neuroligin-3 is able to induce presynaptic differentiation in co-culture (Chih et al., 2004). Thus smaller pool sizes may be needed to identify synaptogenic factors less potent than neuroligin-2, NGL-3 and LRRTM1.

The LRRTM protein family was initially identified while looking for proteins sharing sequence similarity to the Slit family of axon guidance molecules (Lauren et al., 2003). The LRRTM protein family is restricted to vertebrates and consists of four members in mammals. All four family members are expressed in mouse brain from P0 or earlier through adult (Lauren et al., 2003). Each LRRTM family member displays the ability to instruct excitatory presynaptic differentiation.

LRRTM3 displayed the weakest activity among the LRRTMs despite good apparent cell surface localization. Since LRRTM3 expression is low in hippocampus and higher in cortex (Lauren et al., 2003), we also tested LRRTM3-CFP and LRRTM2-CFP in a co-culture assay using neocortical neurons. LRRTM3 was much less effective as a synaptogenic factor using both neuronal types (data not shown). Additionally, LRRTM3-CFP was often concentrated in the COS cells along the length of contacting hippocampal axons in a non-punctate distribution (data not shown), suggesting the presence of a binding partner along the length of axons and not specifically at presynaptic sites. These observations suggest that the function of LRRTM3 may be significantly different from that of the other family members. LRRTM3 was recently identified in an siRNA screen to promote processing of amyloid precursor protein by BACE1 to increase Aβ secretion, a function not shared by the other family members (Majercak et al., 2006).

LRRTM2 localized specifically to glutamatergic and not GABAergic synapses in hippocampal culture (Figure 5C), and the LRRTM2 LRR domain on beads could instruct glutamatergic but not GABAergic presynaptic differentiation (Figure 4). Thus it is not clear why LRRTM2 on COS cells was able to instruct GABAergic as well as glutamatergic presynaptic differentiation (Figure S3D,F). We suspect that the local concentration of LRRTM2 available for presentation to axons may reach higher levels on COS cells than on beads or at postsynaptic sites, allowing for low affinity interactions with GABAergic axons. Similarly, neuroligin-2 instructs both excitatory and inhibitory presynaptic differentiation in the co-culture assay, yet neuroligin-2 is exclusively localized to inhibitory synapses under normal conditions (Chih et al., 2005; Graf et al., 2004; Varoqueaux et al., 2004). The distribution of neuroligin-2 can be shifted to excitatory synapses upon overexpression of PSD-95, leading to the idea that the ratio of postsynaptic organizing molecules may control the ratio of excitatory to inhibitory inputs (Levinson et al., 2005). Perhaps like neuroligins, LRRTM2 may function at both synapse types under some conditions.

Our immunohistochemical localization of LRRTM2 in brain was consistent with the in situ hybridization pattern (Lauren et al., 2003), widespread expression particularly prominent in cerebellar granule cells, the deeper layers of neocortex, and all major cell layers of the hippocampus. The laminar distribution of LRRTM2 within the hippocampus (Figure 7C) suggests a selective targeting of the protein. NGL-1, an LRR domain containing synaptic adhesion protein related to NGL-3 isolated in our screen, is also selectively concentrated in CA1 dendrites in stratum lacunosum moleculare opposite temporoammonic inputs from the entorhinal cortex (Nishimura-Akiyoshi et al., 2007). NGL-1 redistributed equally to radiatum and oriens upon targeted deletion of its axonal ligand netrin-G1. The laminar-selective distribution of LRRTM2 may also be related to expression of an axonal ligand by specific inputs. The LRRTM2 LRR-AP protein did not bind cells expressing LRRTM2-CFP under conditions where positive control neurexin-AP bound cells expressing neuroligin-2-CFP (data not shown), suggesting the absence of LRRTM homophilic binding.

The combined results of our screen suggest that LRR proteins may be prevalent as synaptic cell adhesion and organizing proteins, with 2 of the 3 proteins isolated here containing LRR domains. Screening of an additional six LRR protein candidates chosen only by primary sequence further identified Slitrk2 as synaptogenic. The Slitrk family was recently identified based on homology to Slit secreted axon guidance molecules and Trk neurotrophin receptors (Aruga and Mikoshiba, 2003). Little is reported about cellular function other than ability to alter neurite outgrowth, but mutations in Slitrk1 are associated with Tourette's Syndrome (Abelson et al., 2005), and targeted deletion of Slitrk1 in mice results in behavioral abnormalities (Katayama et al., 2008). LRRTMs and Slitrks were not included in a set of 160 candidates tested recently by RNAi for a role in synapse development; LRRC4C and LRRN6A were tested but found to have no effect in this study which identified a role for cadherins and semaphorins in synapse development (Paradis et al., 2007). Mammalian LRR transmembrane proteins previously shown to have a synaptic function are NGLs and SALMs (Ko and Kim, 2007). NGL-1 is reported to be important for axon outgrowth (Lin et al., 2003), and subsequently NGLs were found to bind PSD-95 (Kim et al., 2006). Like LRRTMs, NGL-2 can promote excitatory presynaptic differentiation using both bead and cellular co-culture assays (Kim et al., 2006). NGL-1 and NGL-2 bind netrin-G1 and netrin-G2, respectively. However, direct aggregation of netrin-G2 on axons does not mediate presynaptic differentiation, nor does netrin-G2 induce postsynaptic differentiation in co-culture (Kim et al., 2006). Thus, like LRRTMs, the axonal binding partner that mediates the synaptogenic activity of NGLs remains to be fully identified. SALMs also bind PSD-95, and artificial clustering of SALM2 on dendrites directly mediates postsynaptic differentiation, but SALMs do not instruct presynaptic differentiation in co-culture assays (Ko et al., 2006; Wang et al., 2006). A recent expression screen in Drosophila also identified several other LRR proteins as important in synaptic target selection (Kurusu et al., 2008). Determining how these multiple synaptic LRR proteins function cooperatively, and identifying their binding partners, whether unique or common, is emerging as a major challenge.

A specific component of this challenge will be to define the precise role of LRRTM1-4 at synapses. The ability of a protein to induce presynaptic specializations in the co-culture assay does not necessarily imply that its endogenous function is in synapse initiation. Given the highly interconnected network of proteins at presynaptic specializations in axons, any protein that tightly binds to a transmembrane component of such complexes could potentially instruct presynaptic differentiation. Therefore, the co-culture screen could yield proteins that are involved in later stages of synaptic maturation, perhaps in recruiting specific components that regulate selective synaptic parameters. In some respects, neuroligins fit this description. While neuroligins are potent synaptogenic factors in the co-culture assay, and neuroligin-1,2,3 triple knockout mice die due to defects in synaptic function, loss of all neuroligins does not result in an obvious effect on overall synapse morphology (Varoqueaux et al., 2006). Furthermore, like the phenotype of LRRTM1 -/- mice reported here, the phenotype of individual neuroligin knockouts is subtle. The only defect reported for the neuroligin-1 null mouse is a reduced NMDA to AMPA ratio at CA1 synapses in acutely prepared slices (Chubykin et al., 2007; Varoqueaux et al., 2006). Similarly, triple mutation of EphB1 plus EphB2 plus EphB3 is required to observe any reduction in spine density in CA1 or neocortex, since single or double EphB mutants exhibit no spine phenotype (Henkemeyer et al., 2003; Kayser et al., 2006).

In addition to this clear evidence for redundancy among members of each family of synaptogenic proteins, many synaptic phenotypes tend to be apparent exclusively in vivo. For example, dissociated cortical cultures from the neuroligin triple knockouts show normal excitatory transmission and synapse density and composition (Varoqueaux et al., 2006). These observations, coupled with the apparent low level of expression of LRRTMs in hippocampal neurons in dissociated culture suggest that the function of these synaptic organizing molecules may be assessed accurately only by in vivo studies. The altered distribution of VGLUT1 but not bassoon in specific hippocampal subfields with in vivo deletion of LRRTM1 as reported here supports a role in selective recruitment of synaptic components. Further ultrastructural, electrophysiological and behavioral analyses will be required to fully characterize the effect of deletion of LRRTM1.

Recently LRRTM1 was shown to be linked via paternal transmission with handedness and schizophrenia (Francks et al., 2007). This finding is intriguing given our results indicating a synaptic function for LRRTM1. Schizophrenia is likely to be etiologically complex with many genetic, epigenetic, and environmental influences. Nonetheless, the identification of genetic alterations associated with complex disorders even in rare cases can lead to the development of animal models, such as the neuroligin-3 R451C knock-in, neuroligin-4 knockout, and neuroligin-2 transgenic mouse models of autism (Hines et al., 2008; Jamain et al., 2008; Tabuchi et al., 2007). Thus it is possible that our molecular and cellular level studies on LRRTM1 may lead to new directions of research on schizophrenia.

Experimental Procedures

More detailed experimental procedures are described in the Supplemental Data.

Generation of the Full-Length, Size-Selected Expression Libraries

Rat brains were dissected at P11, and the cerebellum and brainstem were removed. Total RNA was prepared using the method of Carninci and Hayashizaki (1999), and mRNA was further purified using the Ambion Poly(A)Purist MAG kit. For purification of full-length cDNA, the biotinylated cap-trapper protocol was implemented (Carninci et al., 1996). The first-strand synthesis products are oxidized and 5′-cap and 3′-hydroxyl groups are biotinylated. Limited RNAse I digestion cleaves unhybridized RNA, thus eliminating the 5′ biotin tag on first-strand products that are not full length. The remaining biotinylated mRNA:cDNA hybrids are purified using streptavidin magnetic beads, and single-stranded full-length cDNA is released following sodium hydroxide hydrolysis of the mRNA strand. For second-strand synthesis we developed a modification of the single-stranded linker ligation method (SSLLM) to attach a linker with a BamH I site to the cDNA (Shibata et al., 2001). Full-length cDNA was digested and size-fractionated using a 1% low-melt agarose gel before ligation into the pcDNA3 vector.

Cell Culture

Cultures of hippocampal neurons were prepared from E18 rat embryos according to previously described protocols (Goslin et al., 1998; Kaech and Banker, 2006). COS-7 and HEK293T cells were cultured in DMEM-H supplemented with 10% fetal bovine serum. Co-cultures of COS or HEK293T cells and neurons were performed essentially as described (Graf et al., 2004). For the expression screen, COS cells were transfected in 12-well plates with 500 ng of plasmid from each expression pool. Transfected COS cells were harvested by trypsinization after 7 or 24 hr, seeded onto neuron coverslips pre-grown for 8-9 DIV, and fixed for screening 20 hr later.

Immunocytochemisty and Culture Imaging

For screening, co-cultures were fixed for 12 min with 4% paraformaldehyde and 4% sucrose in PBS (pH 7.4) followed by permeabilization with PBST (PBS + 0.1% Triton X-100). An initial blocking step was performed with PBS-BSA/NGS (3% BSA, 5% normal goat serum) for 30 min at 37°C. Co-cultures were incubated overnight in anti-synapsin I (1:2000; Millipore), anti-PSD-95 family (IgG2a; 1:500; clone 6G6-1C9; Affinity Bioreagents; recognizes PSD-95, PSD-93, SAP102 and SAP97), and anti-gephyrin (IgG1; 1:1000; mAb7a; Synaptic Systems) in PBS-BSA/NGS. After washing with PBS, co-cultures were incubated in Alexa-488 conjugated goat anti-rabbit (1:500; Molecular Probes) and Texas Red conjugated goat anti-mouse (1:500; Invitrogen) was used to detect gephyrin and PSD-95 in the same channel. After washing in PBS, the coverslips were incubated with 200 ng/ml DAPI in PBS for 20 min, washed, and mounted in elvanol (Tris-HCl, glycerol, and polyvinyl alcohol, with 2% 1,4-diazabicyclo[2,2,2]octane). Primary antibodies used for other experiments were: rabbit anti-VGAT (1:2000; Synaptic Systems), anti-SynGAP (1:2000; Affinity BioReagents), mouse anti-Tau-1, clone PC1C6 (IgG2A; 1:2000; Millipore; recognizes dephosphorylated tau), anti-bassoon (IgG2A; 1:1000; Stressgen), anti-synaptophysin (IgG1; 1:1000; BD Biosciences), anti-NMDAR1, clone 54.1 (IgG2a; 1:1000; BD Biosciences), guinea pig anti-VGLUT1 (1:4000; Millipore), and chicken anti-MAP2 (IgY; 1:8000; Abcam).

Images were acquired on a Zeiss Axioplan2 microscope with a 63X 1.4 numerical aperture oil objective and Photometrics Sensys cooled CCD camera using Metamorph imaging software (Molecular Devices) and customized filter sets. Controls lacking specific antibodies confirmed no detectable bleed-through between channels AMCA, CFP, YFP or Alexa-488 (imaged through a YFP filter set), Alexa-568, and Alexa 647. For quantitation, sets of cells were co-cultured and stained simultaneously and imaged with identical settings. All imaging and analysis were done blind.

Electrophysiology

Whole cell patch clamp recordings were made at room temperature in Mg2+ free extracellular solution (168 mM NaCl, 2.4 mM KCl, 10 mM HEPES, 10 mM D-glucose, 1.3 mM CaCl2, 20 μM glycine, pH 7.4, osmolality adjusted with sucrose). HEK293T cells were voltage clamped at -60 mV, spontaneous currents recorded, and then mEPSC-like events recorded in the presence of 1 μM TTX. All activity was blocked with 100 μM APV. NMDA receptor-mediated spontaneous events and mEPSC-like events were both measured with the Axograph software using an optimally scaled sliding template (Clements and Bekkers, 1997) and criteria of 3 times the SD. The measures likely underestimate frequency for the LRRTM2 group without TTX.

Bead Clustering Experiments

For clustering of YFP-LRRTMs, we used a protocol that had been previously used to cluster YFP-neuroligins (Graf et al., 2004). For the presynaptic induction assay, 1 μm NeutrAvidin labeled Fluospheres were mixed by constant inversion at room temperature in HBS + BSA (10 mM HEPES, pH 7.3, 145 mM NaCl, 5 mM KCl, 2 mM MgCl2, 2 mM CaCl2, 100 μg/ml BSA). The beads were incubated with biotinylated anti-myc tag (IgG1; Millipore 16-170; 17 μg per 10 μl beads), washed, then incubated with 60 μg of either the soluble control AP protein or the LRRTM2 LRR-AP protein for 2 hr in HBS + BSA. Coated beads were resuspended in conditioned media and applied to 8 DIV hippocampal neurons. Neurons were cultured in the presence of coated beads for 20-24 hr, and fixed the following day for immunocytochemistry.

Generation and Analysis of LRRTM1 -/- Mice

The entire mouse Lrrtm1 open reading frame (ORF) is encoded by a single exon. Thus, to generate LRRTM1 knock out mice this exon was chosen for gene targeting. The plasmid used in the construction of the targeting construct was obtained from Dr. Peter Mombaerts (Rodriguez et al., 1999). The ES cell clones were analyzed by Southern blotting for the evidence of homologous recombination (Figure 8).

Identical conditions were used for immunostaining as well as capture and analysis of confocal images for each condition. Each experiment was performed blind and in duplicate. Mice were perfused at 4-5 weeks with 2% paraformaldehyde in PBS, brains post-fixed and cryoprotected in 30% (w/v) sucrose in PBS. Cryostat sections (20 μm) were incubated in blocking solution containing 0.1 M PBS, 5% normal goat serum, 5% BSA, and 0.25% Triton X-100, then overnight at 4°C with anti-Bassoon (1:1000, Stressgen) and anti-VGLUT1 (1:300; NeuroMab N28/9) in blocking solution, then with Alexa dye conjugated secondary antibodies. Confocal images (0.37 μm optical section at 5 μm below the tissue surface) were captured sequentially on a Fluoview FV500 confocal system from five separate fields per anatomical region per animal. Three wildtype and two knockout animals were used for analysis. Images were analyzed using MetaMorph. A single threshold was set for each staining condition to capture clusters that were clearly distinguishable and to minimize merged clusters. The number, size, and intensity of the puncta were measured. Data from wild type and knockout images were compared using Student's t-tests.

Supplementary Material

Acknowledgments

We thank Xiling Zhou and Huaiyang Wu for consistent excellence in preparation of hippocampal neuron cultures. Fernanda Laezza, Vanessa Santiago, and Ann Chow provided assistance with preparation of RNA, cDNA pools, and PLAP fusion proteins. Katherine Walzak and Susan Hand provided assistance with co-culture assays and immunofluorescence for revision experiments. We thank Dr. Robert Holt and team at the Michael Smith Genome Sciences Centre for arraying the cDNA subpool and preparing DNA in 384-well format. We also thank Drs. Joshua Sanes for input in the early stages of this work and Tabrez Siddiqui for helpful comments on the manuscript. This work was supported by CIHR MOP-84241, NIH MH070860, and CRC and MSFHR salary awards to A.M.C., NIH NS03220 and NS39962 to S.M.S., CIHR and MSFHR fellowships to F.A.D. and JSPS fellowship to H.T.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • Abelson JF, Kwan KY, O'Roak BJ, Baek DY, Stillman AA, Morgan TM, Mathews CA, Pauls DL, Rasin MR, Gunel M, et al. Sequence variants in SLITRK1 are associated with Tourette's syndrome. Science. 2005;310:317–320. [PubMed]
  • Ackley BD, Jin Y. Genetic analysis of synaptic target recognition and assembly. Trends Neurosci. 2004;27:540–547. [PubMed]
  • Aoto J, Ting P, Maghsoodi B, Xu N, Henkemeyer M, Chen L. Postsynaptic ephrinB3 promotes shaft glutamatergic synapse formation. J Neurosci. 2007;27:7508–7519. [PubMed]
  • Aruga J, Mikoshiba K. Identification and characterization of Slitrk, a novel neuronal transmembrane protein family controlling neurite outgrowth. Mol Cell Neurosci. 2003;24:117–129. [PubMed]
  • Bamji SX, Shimazu K, Kimes N, Huelsken J, Birchmeier W, Lu B, Reichardt LF. Role of beta-catenin in synaptic vesicle localization and presynaptic assembly. Neuron. 2003;40:719–731. [PMC free article] [PubMed]
  • Biederer T, Sara Y, Mozhayeva M, Atasoy D, Liu X, Kavalali ET, Sudhof TC. SynCAM, a synaptic adhesion molecule that drives synapse assembly. Science. 2002;297:1525–1531. [PubMed]
  • Biederer T, Scheiffele P. Mixed-culture assays for analyzing neuronal synapse formation. Nat Protoc. 2007;2:670–676. [PubMed]
  • Carninci P, Hayashizaki Y. High-efficiency full-length cDNA cloning. Methods Enzymol. 1999;303:19–44. [PubMed]
  • Carninci P, Kvam C, Kitamura A, Ohsumi T, Okazaki Y, Itoh M, Kamiya M, Shibata K, Sasaki N, Izawa M, et al. High-efficiency full-length cDNA cloning by biotinylated CAP trapper. Genomics. 1996;37:327–336. [PubMed]
  • Chih B, Afridi SK, Clark L, Scheiffele P. Disorder-associated mutations lead to functional inactivation of neuroligins. Hum Mol Genet. 2004;13:1471–1477. [PubMed]
  • Chih B, Engelman H, Scheiffele P. Control of excitatory and inhibitory synapse formation by neuroligins. Science. 2005;307:1324–1328. [PubMed]
  • Chubykin AA, Atasoy D, Etherton MR, Brose N, Kavalali ET, Gibson JR, Sudhof TC. Activity-dependent validation of excitatory versus inhibitory synapses by neuroligin-1 versus neuroligin-2. Neuron. 2007;54:919–931. [PMC free article] [PubMed]
  • Clements JD, Bekkers JM. Detection of spontaneous synaptic events with an optimally scaled template. Biophys J. 1997;73:220–229. [PMC free article] [PubMed]
  • Craig AM, Graf ER, Linhoff MW. How to build a central synapse: clues from cell culture. Trends Neurosci. 2006;29:8–20. [PMC free article] [PubMed]
  • Craig AM, Kang Y. Neurexin-neuroligin signaling in synapse development. Curr Opin Neurobiol. 2007;17:43–52. [PMC free article] [PubMed]
  • Dalva MB, McClelland AC, Kayser MS. Cell adhesion molecules: signalling functions at the synapse. Nat Rev Neurosci. 2007;8:206–220. [PubMed]
  • Fournier AE, GrandPre T, Strittmatter SM. Identification of a receptor mediating Nogo-66 inhibition of axonal regeneration. Nature. 2001;409:341–346. [PubMed]
  • Francks C, Maegawa S, Lauren J, Abrahams BS, Velayos-Baeza A, Medland SE, Colella S, Groszer M, McAuley EZ, Caffrey TM, et al. LRRTM1 on chromosome 2p12 is a maternally suppressed gene that is associated paternally with handedness and schizophrenia. Mol Psychiatry. 2007;12:1129–1139. 1057. [PMC free article] [PubMed]
  • Fu Z, Washbourne P, Ortinski P, Vicini S. Functional excitatory synapses in HEK293 cells expressing neuroligin and glutamate receptors. J Neurophysiol. 2003;90:3950–3957. [PubMed]
  • Goslin K, Asmussen H, Banker G. Rat hippocampal neurons in low-density culture. In: Banker G, Goslin K, editors. Culturing Nerve Cells. Cambridge: MIT Press; 1998. pp. 339–370.
  • Graf ER, Zhang X, Jin SX, Linhoff MW, Craig AM. Neurexins induce differentiation of GABA and glutamate postsynaptic specializations via neuroligins. Cell. 2004;119:1013–1026. [PMC free article] [PubMed]
  • Henderson JT, Georgiou J, Jia Z, Robertson J, Elowe S, Roder JC, Pawson T. The receptor tyrosine kinase EphB2 regulates NMDA-dependent synaptic function. Neuron. 2001;32:1041–1056. [PubMed]
  • Henkemeyer M, Itkis OS, Ngo M, Hickmott PW, Ethell IM. Multiple EphB receptor tyrosine kinases shape dendritic spines in the hippocampus. J Cell Biol. 2003;163:1313–1326. [PMC free article] [PubMed]
  • Hines RM, Wu L, Hines DJ, Steenland H, Mansour S, Dahlhaus R, Singaraja RR, Cao X, Sammler E, Hormuzdi SG, et al. Synaptic imbalance, stereotypies, and impaired social interactions in mice with altered neuroligin 2 expression. J Neurosci. 2008;28:6055–6067. [PubMed]
  • Jamain S, Radyushkin K, Hammerschmidt K, Granon S, Boretius S, Varoqueaux F, Ramanantsoa N, Gallego J, Ronnenberg A, Winter D, et al. Reduced social interaction and ultrasonic communication in a mouse model of monogenic heritable autism. Proc Natl Acad Sci U S A. 2008;105:1710–1715. [PMC free article] [PubMed]
  • Kaech S, Banker G. Culturing hippocampal neurons. Nat Protoc. 2006;1:2406–2415. [PubMed]
  • Katayama K, Yamada K, Ornthanalai VG, Inoue T, Ota M, Murphy NP, Aruga J. Slitrk1-deficient mice display elevated anxiety-like behavior and noradrenergic abnormalities. Mol Psychiatry 2008 [PubMed]
  • Kayser MS, McClelland AC, Hughes EG, Dalva MB. Intracellular and trans-synaptic regulation of glutamatergic synaptogenesis by EphB receptors. J Neurosci. 2006;26:12152–12164. [PubMed]
  • Kim S, Burette A, Chung HS, Kwon SK, Woo J, Lee HW, Kim K, Kim H, Weinberg RJ, Kim E. NGL family PSD-95-interacting adhesion molecules regulate excitatory synapse formation. Nat Neurosci. 2006;9:1294–1301. [PubMed]
  • Ko J, Kim E. Leucine-rich repeat proteins of synapses. J Neurosci Res. 2007;85:2824–2832. [PubMed]
  • Ko J, Kim S, Chung HS, Kim K, Han K, Kim H, Jun H, Kaang BK, Kim E. SALM synaptic cell adhesion-like molecules regulate the differentiation of excitatory synapses. Neuron. 2006;50:233–245. [PubMed]
  • Krueger SR, Kolar A, Fitzsimonds RM. The presynaptic release apparatus is functional in the absence of dendritic contact and highly mobile within isolated axons. Neuron. 2003;40:945–957. [PubMed]
  • Kuja-Panula J, Kiiltomaki M, Yamashiro T, Rouhiainen A, Rauvala H. AMIGO, a transmembrane protein implicated in axon tract development, defines a novel protein family with leucine-rich repeats. J Cell Biol. 2003;160:963–973. [PMC free article] [PubMed]
  • Kurusu M, Cording A, Taniguchi M, Menon K, Suzuki E, Zinn K. A screen of cell-surface molecules identifies leucine-rich repeat proteins as key mediators of synaptic target selection. Neuron. 2008;59:972–985. [PMC free article] [PubMed]
  • Lauren J. Characterization of LRRTM and NGR Gene Families: Expression and Functions. Helsinki University Printing House; Helsinki: 2007.
  • Lauren J, Airaksinen MS, Saarma M, Timmusk T. A novel gene family encoding leucine-rich repeat transmembrane proteins differentially expressed in the nervous system. Genomics. 2003;81:411–421. [PubMed]
  • Levinson JN, Chery N, Huang K, Wong TP, Gerrow K, Kang R, Prange O, Wang YT, El-Husseini A. Neuroligins mediate excitatory and inhibitory synapse formation: involvement of PSD-95 and neurexin-1beta in neuroligin-induced synaptic specificity. J Biol Chem. 2005;280:17312–17319. [PubMed]
  • Lin JC, Ho WH, Gurney A, Rosenthal A. The netrin-G1 ligand NGL-1 promotes the outgrowth of thalamocortical axons. Nat Neurosci. 2003;6:1270–1276. [PubMed]
  • Linhoff MW. Expression Screening for Proteins Involved in CNS Presynaptic Differentiation and Initial Characterization of the Synaptogenic LRRTM Protein Family. Washington Unversity in St. Louis; St. Louis: 2008.
  • Majercak J, Ray WJ, Espeseth A, Simon A, Shi XP, Wolffe C, Getty K, Marine S, Stec E, Ferrer M, et al. LRRTM3 promotes processing of amyloid-precursor protein by BACE1 and is a positional candidate gene for late-onset Alzheimer's disease. Proc Natl Acad Sci U S A. 2006;103:17967–17972. [PMC free article] [PubMed]
  • Micheva KD, Beaulieu C. Quantitative aspects of synaptogenesis in the rat barrel field cortex with special reference to GABA circuitry. J Comp Neurol. 1996;373:340–354. [PubMed]
  • Missler M, Zhang W, Rohlmann A, Kattenstroth G, Hammer RE, Gottmann K, Sudhof TC. Alpha-neurexins couple Ca2+ channels to synaptic vesicle exocytosis. Nature. 2003;424:939–948. [PubMed]
  • Nishimura-Akiyoshi S, Niimi K, Nakashiba T, Itohara S. Axonal netrin-Gs transneuronally determine lamina-specific subdendritic segments. Proc Natl Acad Sci U S A. 2007;104:14801–14806. [PMC free article] [PubMed]
  • Paradis S, Harrar DB, Lin Y, Koon AC, Hauser JL, Griffith EC, Zhu L, Brass LF, Chen C, Greenberg ME. An RNAi-based approach identifies molecules required for glutamatergic and GABAergic synapse development. Neuron. 2007;53:217–232. [PMC free article] [PubMed]
  • Rodriguez I, Feinstein P, Mombaerts P. Variable patterns of axonal projections of sensory neurons in the mouse vomeronasal system. Cell. 1999;97:199–208. [PubMed]
  • Scheiffele P, Fan J, Choih J, Fetter R, Serafini T. Neuroligin expressed in nonneuronal cells triggers presynaptic development in contacting axons. Cell. 2000;101:657–669. [PubMed]
  • Sheng M, Sala C. PDZ domains and the organization of supramolecular complexes. Annu Rev Neurosci. 2001;24:1–29. [PubMed]
  • Shibata Y, Carninci P, Watahiki A, Shiraki T, Konno H, Muramatsu M, Hayashizaki Y. Cloning full-length, cap-trapper-selected cDNAs by using the single-strand linker ligation method. Biotechniques. 2001;30:1250–1254. [PubMed]
  • Stevens B, Allen NJ, Vazquez LE, Howell GR, Christopherson KS, Nouri N, Micheva KD, Mehalow AK, Huberman AD, Stafford B, et al. The classical complement cascade mediates CNS synapse elimination. Cell. 2007;131:1164–1178. [PubMed]
  • Tabuchi K, Blundell J, Etherton MR, Hammer RE, Liu X, Powell CM, Sudhof TC. A neuroligin-3 mutation implicated in autism increases inhibitory synaptic transmission in mice. Science. 2007;318:71–76. [PMC free article] [PubMed]
  • Varoqueaux F, Aramuni G, Rawson RL, Mohrmann R, Missler M, Gottmann K, Zhang W, Sudhof TC, Brose N. Neuroligins determine synapse maturation and function. Neuron. 2006;51:741–754. [PubMed]
  • Varoqueaux F, Jamain S, Brose N. Neuroligin 2 is exclusively localized to inhibitory synapses. Eur J Cell Biol. 2004;83:449–456. [PubMed]
  • Waites CL, Craig AM, Garner CC. Mechanisms of vertebrate synaptogenesis. Annu Rev Neurosci. 2005;28:251–274. [PubMed]
  • Wang CY, Chang K, Petralia RS, Wang YX, Seabold GK, Wenthold RJ. A novel family of adhesion-like molecules that interacts with the NMDA receptor. J Neurosci. 2006;26:2174–2183. [PubMed]
  • Yamagata M, Sanes JR, Weiner JA. Synaptic adhesion molecules. Curr Opin Cell Biol. 2003;15:621–632. [PubMed]
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

  • Gene
    Gene
    Gene links
  • Gene (nucleotide)
    Gene (nucleotide)
    Records in Gene identified from shared sequence links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene
    HomoloGene links
  • MedGen
    MedGen
    Related information in MedGen
  • Nucleotide
    Nucleotide
    Published Nucleotide sequences
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • Protein
    Protein
    Published protein sequences
  • PubMed
    PubMed
    PubMed citations for these articles
  • Substance
    Substance
    PubChem Substance links
  • Taxonomy
    Taxonomy
    Related taxonomy entry
  • Taxonomy Tree
    Taxonomy Tree

Recent Activity

    Your browsing activity is empty.

    Activity recording is turned off.

    Turn recording back on

    See more...