Identification of MAGUK scaffold proteins as intracellular binding partners of synaptic adhesion protein Slitrk2

Connor Loomisa,1, Aliyah Stephensa,1, Remi Janicota,b, Usman Baqaia,c, Laura Drebushenkoa,d, Jennifer Rounda

Abstarct

Synaptic adhesion proteins play a critical role in the formation and maintenance of synapses in the developing nervous system. Errors in synaptic adhesion constitute the molecular basis of many neuropsychiatric disorders, including schizophrenia, bipolar disorder, Tourette syndrome, and autism. Slit- and Trk-like proteins (Slitrks) are a family of leucine-rich repeat containing transmembrane proteins that promote synaptogenesis. These proteins localize to the postsynaptic density, where they induce synapse formation via trans-synaptic interactions with receptor protein tyrosine phosphatases. While trans-synaptic binding partners of Slitrks have been reported, little is known about the intracellular proteins that associate with Slitrks. Here we report an interaction between Slitrk2 and members of the PSD-95 subfamily of membrane associated guanylate kinases (MAGUKs). Coimmunoprecipitation from postnatal mouse brain indicates that PSD-93 and PSD-95 associate with Slitrk2 in vivo. Mapping analysis in yeast demonstrates that Slitrk2 interacts directly with PSD-95 via a non-canonical Src homology 3 (SH3) domain binding motif that associates with the SH3 domain of PSD-95. We also show that PSD95 induces robust clustering of Slitrk2 in 293T cells, and deletion of the SH3 domain in PSD-95 or the SH3 domain binding motif in Slitrk2 reduces this clustering. These data confirm PSD-95 as the first known intracellular binding partner of Slitrk2. Future studies will examine if Slitrk-MAGUK interactions mediate localization of Slitrks to synaptic sites and facilitate recruitment of additional intracellular signaling molecules involved in postsynaptic differentiation.

Keywords:
Slitrk
Leucine rich repeat
PSD-95
PDZ domain
Synaptic adhesion
Postsynaptic density

1. Introduction

Nearly every aspect of nervous system function relies on the proper establishment and maintenance of synaptic contacts between neurons. An array of transmembrane cell adhesion molecules (CAMs) serve as synaptic organizers, initiating contact between pre- and postsynaptic terminals and contributing to the specification of synapses as excitatory or inhibitory (Takahashi and Craig, 2013; Williams et al., 2010). These CAMs also organize the recruitment of relevant receptors and intracellular protein complexes required for synaptic differentiation and maturation (Jang et al., 2017; Sheng and Kim, 2011). Mutations in genes that code for synaptic CAMs are associated with many neuropsychiatric disorders, including autism, Obsessive compulsive disorder (OCD), and schizophrenia, demonstrating the importance of synaptic cell adhesion for proper nervous system function (Chen et al., 2014; Proenca et al., 2011; Südhof, 2008).
A diverse repertoire of postsynaptic CAMs participate in trans-synaptic interactions to promote synapse formation, including neuroligins, immunoglobulin (Ig)-domain containing proteins, and members of leucine-rich repeat (LRR) superfamily (de Wit and Ghosh, 2014; de Wit and Ghosh, 2016). Members of one such LRR superfamily, the Slit and Trk-like proteins (Slitrks), are enriched at the postsynaptic membrane, where they participate in trans-synaptic interactions with leukocyte common antigen-related receptor protein tyrosine phosphatases (LAR-
RPTPs) (Takahashi et al., 2012; Um et al., 2014; Yim et al., 2013). The six members of the Slitrk family are single-pass transmembrane proteins, each containing two leucine-rich repeat domains in the extracellular region (Aruga and Mikoshiba, 2003). When expressed in nonneuronal cells, Slitrks induce axonal contact and presynaptic differentiation in heterologous cell culture, and this ability of Slitrks to induce synaptogenesis requires interaction with presynaptic LAR-RPTPs (Linhoff et al., 2009; Takahashi et al., 2012; Yim et al., 2013). Modulation of Slitrk expression levels alters synaptic density in vitro and in vivo (Beaubien et al., 2016; Takahashi et al., 2012; Yim et al., 2013), supporting a key role for Slitrks in synapse formation, differentiation, and/or maintenance.
While trans-synaptic binding partners of Slitrks have been well characterized, little is known about the intracellular binding partners of Slitrks. Only one binding partner of Slitrk1, the adaptor protein 14-3-3, had been reported to date (Kajiwara et al., 2009). The cytoplasmic domains of Slitrks 2-5 contain putative phospho-tyrosine binding motifs but vary in size and amino acid composition, suggesting unique binding partners and divergent intracellular signaling mechanisms. The presence of a canonical class I PDZ binding motif at the carboxyl terminus of Slitrk2, combined with the prominent role of PDZ-containing MAGUK scaffold proteins in postsynaptic structure and function (Won et al., 2017; Zheng et al., 2011), prompted us to investigate a potential interaction between Slitrk2 and members of the MAGUK family.
Here we identify two members of the MAGUK scaffold family, PSD93 and PSD-95, as intracellular binding partners of Slitrk2. We show that the association of PSD-95 with Slitrk2 requires the SH3 domain of PSD-95, which binds a non-canonical SH3 domain binding motif near the carboxyl terminus of Slitrk2. We also demonstrate that PSD-95 induces robust clustering of Slitrk2 in non-neuronal cells, and these clustering data provide further support for a SH3-mediated interaction. These findings establish a molecular relationship between MAGUK family scaffolds and Slitrks and pave the way for investigations into the functional significance of this interaction.

2. Experimental procedures

2.1. DNA constructs

Full-length HA-tagged Slitrk constructs were a gift from Ann Marie Craig. PSD-95-pTagRFP was a gift from Johannes Hell (Addgene plasmid # 52671). Truncated versions of HA-Slitrk2 were constructed via PCR amplification using full-length HA-Slitrk2 as a template. HaSlitrk2ΔISQL encodes amino acids 1-842, HA- Slitrk2ΔISQL/DY encodes amino acids 1-833, and HA-Slitrk2ΔICD encodes amino acids 1-643. For yeast two-hybrid studies, the intracellular domains of Slitrks 2, 3, 4, 5, and 6 were sub-cloned into pGBKT7 (Takara Bio) to generate bait constructs. BD-Slitrk2 contains amino acids 697-846, while Slitrk2 BDSlitrk2ΔISXL contains amino acids 697-842. BD-Slitrk2ΔDY contains a Y-F substitution at position 833, BD Slitrk2PPPV contains a PPPV to AAAV substitution at position 740, and BD- Slitrk2ΔPNVP contains a PNVP to ANVA substitution at position 702. Full length and truncated versions of PSD-95 were sub-cloned into pGADT7 (Takara Bio) for yeast twohybrid studies and into pTag RFP-N for clustering studies. AD-PSD953PDZ encodes amino acids 1-430, AD-PSD-953PDZSH3 encodes amino acids 1-495, and AD-PSD-95SH3GK encodes amino acids 431-724.

2.2. Antibodies

Commercially validated rabbit polyclonal antibodies against Slitrk2 (PRS4457 and PRS4459) were obtained from Sigma-Aldrich. AntiHA.11 epitope tag antibody was obtained from BioLegend, and monoclonal myc antibody clone 9E.10 was obtained from BioRad. Mouse monoclonal antibodies against PSD-95, PSD-93, pan-MAGUK, SAP97, and SAP102 were obtained from the NIH NeuroMab Facility at UCDavis. Goat anti-rabbit Alexa 488 and goat anti-mouse Alexa 594 were obtained from Abcam. HRP-conjugated secondary antibodies for Western blotting were obtained from Thermo Scientific.

2.3. Preparation of crude synaptosomes

Enrichment of crude synaptic membranes was performed as described (Bermejo et al., 2014). Briefly, forebrain tissue was isolated from CD-1 mice at postnatal day 28 and homogenized in ice-cold 0.32 M sucrose/4 mM HEPES. The homogenate was centrifuged at 900 ×g for 10 min to remove the nuclear fraction (P1). The resulting supernatant was centrifuged at 10,000 ×g to obtain the crude synaptosome fraction (P2). The P2 pellet was lysed in 50 mM Tris pH 8.0, 150 mM NaCl, 2 mM EDTA, 1% sodium deoxycholate, 10 mM sodium fluoride, 1 mM PMSF, and 1× aprotinin/leupeptin to inhibit proteases. The lysate was then subjected to high-speed centrifugation, and the supernatant was used for coimmunoprecipitation.

2.4. Coimmunoprecipitation and Western blotting

Crude synaptosome lysates were pre-cleared with Protein A/G agarose beads and incubated with Slitrk2 antibody overnight at 4C with rotation. Lysates were then incubated with Protein A/G agarose beads for two hours at 4C to capture antibody complexes. Beads were then washed in lysis buffer and boiled in SDS-PAGE loading buffer containing 200 mM DTT. Denatured proteins were resolved using SDSPAGE electrophoresis and transferred to PVDF membranes. Membranes were blocked in 5% non-fat dry milk and incubated with appropriate primary antibodies overnight at 4C with rotation. Membranes were washed, incubated with HRP-conjugated secondary antibodies for one hour at room temperature, washed again, and exposed to ECL substrate (BioRad). Blots were imaged with a Syngene GeneGnome chemiluminescence imager.

2.5. Yeast two-hybrid analysis

Mapping experiments were performed using the Matchmaker Gold Yeast Two-Hybrid System (Takara Bio). Briefly, bait and prey plasmids were co-transformed into the Y2H Gold yeast strain using the Yeastmaker Yeast Transformation Kit (Takara Bio). Yeast were plated on SD -Leu/-Trp to select cells that received both plasmids. Colonies were then spot plated onto SD/-Ade/-His/-Leu/-Trp supplemented with 200 ng/ml Aureobasidin A and 40 μg/ml X-alpha-Gal to detect positive protein-protein interactions. Constructs co-transformed with empty bait or empty prey plasmids did not activate reporter genes.

2.6. Cell transfection

HEK 293T cells were grown in DMEM supplemented with 10% fetal bovine serum (Gibco) for a maximum of 12 passages. Cells at 80% confluency were transfected using Lipofectamine 2000 (Thermo Scientific). Cells were incubated for 24 h post transfection then fixed and immunostained as described below.below

2.7. Immunocytochemistry

HEK 293T cells were fixed with 4% paraformaldehyde in PBS and permeabilized with 0.5% Triton. Cells were blocked in 10% normal goat serum, then incubated overnight with appropriate primary antibodies in 5% normal goat serum. Cells were washed with PBS + 0.1% Tween-20 and incubated in Alexa-conjugated secondary antibodies for one hour at room temperature. Cells were washed again, and coverslips were mounted on glass slides using Fluoromount G (Electron Microscopy Sciences).

2.8. Imaging

Confocal imaging for images in Figs. 4 and 5 was performed on a Nikon C2 confocal microscope with a Plan Apo 60×/1.40 oil objective. Widefield imaging for quantification of cluster size was performed on a Nikon Eclipse 80i Fluorescence microscope using a 100×/1.30 oil objective. 293T clusters were identified and mean cluster area was calculated using the auto-detect ROI feature of NIS Elements AR software, with the experimenter blind to the condition during cluster identification. Images containing immunostaining with Alexa 594-conjugated secondary antibodies were pseudo-colored magenta. Background fluorescence was reduced in all microscopy images by specifying and subtracting a background ROI. Any adjustments to brightness were performed equally across all images in a data set.

2.9. Statistical analysis

Graphing and statistical analyses were performed using Graphpad Prism version 6. Specific statistical tests and sample sizes are indicated in the appropriate figure legends.

3. Results

3.1. Slitrk2 associates with MAGUK family scaffolds in postnatal mouse brain

To determine if members of the MAGUK scaffold family interact with Slitrk2, we generated crude synaptosome preparations from the forebrain of postnatal day 28 mice and employed a commercially available rabbit polyclonal antibody to immunoprecipitate Slitrk2. Inhouse validation of the Slitrk2 antibody confirmed that this antibody recognizes Slitrk2 and does not cross-react with other Slitrk family members (Supplementary Fig. 1). Immunoblotting demonstrated successful immunoprecipitation of Slitrk2, as evidenced by a prominent band at the predicted molecular weight (Fig. 1A). Immunoblotting with a pan-MAGUK antibody indicated coimmunoprecipitation of one or more MAGUK scaffold proteins with Slitrk2 (Fig. 1A). Subsequent immunoblotting with monoclonal antibodies specific to each MAGUK scaffold protein showed that PSD-93 and PSD-95 associate with Slitrk2, but SAP97 does not. A faint association of SAP102 with Slitrk2 was also observed. All four MAGUKs were detected in the P2 lysate, and no bands were observed in the control IP lacking the immunoprecipitating Slitrk2 antibody. Coimmunoprecipitation of Slitrk2 was also observed with immunoprecipitating antibodies against PSD-95 (Fig. 1B) and PSD93 (Fig. 1C). Expanded images of each Western blot are shown in Supplementary Fig. 2.

3.2. The SH3 domain of PSD-95 participates in a direct interaction with Slitrk2

A number of postsynaptic transmembrane proteins that contain a class I PDZ binding motif at their carboxy terminus are known to interact with the PDZ domains of PSD-95 (Irie et al., 1997; Kim et al., 2006; Kornau et al., 1995). Therefore, we predicted that PSD-95 binds to Slitrk2 via a putative class I PDZ binding motif (ISQL) located at the carboxyl terminus of Slitrk2. To test this hypothesis, we use the Matchmaker Yeast two-hybrid system (Takara Bio) to map the interaction between PSD-95 and Slitrk2. Specifically, we generated a bait construct consisting of the Gal4 binding domain fused to the intracellular domain of Slitrk2, as well as a prey construct consisting of the Gal4 activation domain fused to full-length PSD-95. Yeast cells cotransformed with the two full-length constructs grew on drop-out plates containing Aureobasidin A and turned blue in the presence of an X-gal derivative, confirming a positive interaction between Slitrk2 and PSD95 in yeast (Fig. 2B). Subsequent analysis of other Slitrk family members demonstrated that Slitrk5, which also contains a canonical class I PDZ binding motif at its C′ terminus, also interacts with PSD-95 in yeast (Supplementary Fig. 3).
To determine if the PDZ domains of PSD-95 mediate the interaction with Slitrk2, we generated truncated versions of PSD-95 and co-transformed these constructs with the full-length Slitrk2 intracellular domain (Fig. 2A). Surprisingly, we found that a truncated version of PSD95 containing only the PDZ domains did not interact with Slitrk2, while all versions of PSD-95 containing the SH3 domain displayed a positive interaction (Fig. 2B). Specifically, a version of PSD-95 containing only the SH3 and GK-like domain interacted strongly with Slitrk2, and a version of PSD-95 containing the PDZ and SH3 domains but lacking the GK-like domain exhibited a strong interaction with Slitrk2 (Fig. 2B).
To identify the SH3 domain binding site in Slitrk2, we introduced missense mutations into three potential SH3 domain binding sites in Slitrk2 (Fig. 3A) and tested each of these mutated constructs against full-length PSD-95 (Fig. 3B). We found that a DY motif located 15 amino acids from the carboxyl terminus of Slitrk2 is required for the interaction with PSD-95 in yeast. Interestingly, we also found that Immunoprecipitation of Slitrk2 from crude synaptosomal lysate resulted in coimmunoprecipitation of PSD-95 and PSD-93. A small amount of SAP102 also coimmunoprecipitated with Slitrk2. B–C. Immunoprecipitation of PSD-95 (B) and PSD-93 (C) from crude synaptosomal lysate resulted in coimmunoprecipitation of Slitrk2. No coimmunoprecipitation was observed in IgG controls. Images are representative of at least two independent immunoprecipitations from separate synaptosome preparations.
To further explore the relative importance of the SH3 and PDZ binding motifs in Slitrk2, we tested the mutated versions of Slitrk2 against a version of PSD-95 lacking the GK-like domain, which was shown to exhibit a strong interaction with full length Slitrk2 in Fig. 2.
Once again, mutation of the DY motif in Slitrk2 abolished the interaction (Fig. 3C). However, removal of the ISQL motif from the carboxyl terminus of Slitrk2 no longer abolished the interaction, suggesting that the ISQL motif may influence the strength of the interaction but is not required for PSD-95 association. Expression of all bait and prey proteins in yeast was verified by Western blotting and is shown in Supplementary Fig. 4.

3.3. PSD-95 induces clustering of Slitrk2 in mammalian cells

To determine if Slitrk2 and PSD-95 affect the subcellular localization of one another in mammalian cells, we co-expressed a HA-tagged version of Slitrk2 and a RFP-fused version of PSD-95 into HEK 293 T cells. Twenty four hours following transfection, we fixed the cells and immunostained for HA and PSD-95. When Slitrk2 and PSD-95 were expressed in the same cells, we observed robust clustering of Slitrk2 proteins (Fig. 4A), and these clusters colocalized heavily with clusters of PSD-95. Clustering was not prevalent in cells containing Slitrk2 alone, but significant cell surface expression of HA-Slitrk2 was observed (Fig. 4A).
To determine if the canonical class I PDZ binding motif at the carboxyl terminus of Slitrk2 is required for clustering, we co-expressed PSD-95 with a truncated version of HA-Slitrk2 that lacked this motif (Slitrk2-ΔISQL). Removal of the last four amino acids of Slitrk2 significantly reduced the mean Slitrk2 cluster area, but did not abolish clustering completely (Fig. 4B, C). Removal of the PDZ binding motif as well as the critical upstream DY motif from Slitrk2 (Slitrk2-ΔISQL/DY) reduced Slitrk2 clustering even further, with no significant difference in cluster size found between the ΔISQL/DY truncation and full-size Slitrk2 alone (Fig. 4B, C). When we co-expressed a version of Slitrk2 lacking the entire intracellular domain (Slitrk2-ΔICD), the average Slitrk2 cluster area was also reduced to baseline levels, similar to that observed for full-length Slitrk2 alone (Fig. 4B, C).
We observed consistent clustering of PSD-95, regardless of the Slitrk2 constructs being co-expressed (Fig. 4A, B), so we also quantified the effect of Slitrk2 on the mean area of PSD-95 clusters. We found no significant difference in the mean cluster area for PSD-95 in the presence or absence of full-length Slitrk2, suggesting that Slitrk2 has no discernable effect on the subcellular localization of PSD-95 (Fig. 4D).
To determine if the SH3 domain of PSD-95 is required for Slitrk2 clustering in HEK 293T cells, we repeated our clustering analysis with the truncated versions of PSD-95 described in Fig. 2. We found that fulllength PSD-95 and a version of PSD-95 containing the PDZ and SH3 domains clustered Slitrk2 to a similar extent (Fig. 5A, B, D), whereas further truncation of the SH3 domain from PSD-95 reduced Slitrk2 clustering (Fig. 5C, D). Removal of the SH3 and/or GK domain from PSD-95 did not alter the clustering of PSD-95 itself (Fig. 5E).

4. Discussion

Here we report the first known intracellular binding partner of synaptic adhesion protein Slitrk2. Specifically, we show evidence of an in vivo interaction between Slitrk2 and two members of the MAGUK scaffold family. MAGUKs are well established as key molecular organizers at the post synaptic density (Won et al., 2017; Zheng et al., 2011), and these data add Slitrk2 to an expanding list of postsynaptic receptors and cell adhesion molecules that associate with these essential intracellular scaffolds.
Immunoprecipitation from crude synaptosome preparations reveals that Slitrk2 binds to PSD-93 and PSD-95 in vivo. Association of PSD-93 and PSD-95 with the same postsynaptic transmembrane receptor is not unprecedented, with both scaffolds capable of forming molecular complexes with Neuroligins and the inward rectifier K+ channel Kir2.3 (Inanobe et al., 2002; Irie et al., 1997). It remains to be determined if coimmunoprecipitation of both scaffolds with Slitrk2 represents a single molecular complex containing all three proteins, or distinct pools of Slitrk2/PSD-93 and Slitrk2/PSD-95 interactions. A significant number of excitatory synapses contain both PSD-93 and PSD-95 (Sans et al., 2000; Sun and Turrigiano, 2011), and PSD-93 and PSD-95 are capable of forming ligand-induced hetero-oligomers (Zeng et al., 2018), raising the possibility that both scaffold proteins are binding simultaneously to the same Slitrk molecule in vivo.
Our mapping analysis disproved our initial hypothesis that the interaction between Slitrk2 and PSD-95 would be solely PDZ-mediated. Instead, we found that the SH3 domain of PSD-95 is required for association with Slitrk2, and the SH3 binding site is likely a DY motif located 15 amino acids from the carboxyl terminus of Slitrk2. This DY motif deviates from the traditional PXXP motif that characterizes many SH3 binding sites and has been verified as a non-canonical SH3 binding site in the context of other protein interactions (Saksela and Permi, 2012). SH3-mediated binding has also been reported for other membrane receptors that associate with PSD-95. Kainate receptors bind to PSD-95 via an SH3-mediated interaction (Garcia et al., 1998), and the SH3 domain of PSD-95 has been shown to bind NMDA receptors in a sub-type specific manner (Cousins et al., 2009; Cousins and Stephenson, 2012). The finding that Slitrk5 also exhibits an interaction with PSD-95 is consistent with a two-part binding mechanism, since Slitrk5 is the only other Slitrk family member to possess both a canonical class I PDZ binding motif at its C′-terminus and an upstream DY motif.
Subsequent mapping analysis in yeast demonstrated that removal of the PDZ binding motif (ΔISQL) from the carboxyl terminus of Slitrk2 abolished the interaction with full-length PSD-95. This finding suggests that the PDZ-binding motif may stabilize the SH3 mediated interaction or may be essential for the proper three-dimensional structure of the carboxyl terminal region of Slitrk2. The fact that Slitrk2-ΔISQL displays a weak interaction with a version of PSD-95 lacking the GK domain might be explained by differences in the expression levels of the two versions of PSD-95. An alternative explanation is that the presence of the GK domain has an inhibitory effect on the association between Slitrk2 and PSD-95, and removal of the GK domain from PSD-95 strengthens the interaction with Slitrk2. In support of this explanation, others have shown evidence of an intramolecular association between the SH3 and GK domains of PSD-95 that seems to have an inhibitory effect on binding of PSD-95 to other proteins (McGee and Bredt, 1999).
When co-expressed in HEK 293T cells, PSD-95 induced robust clustering of Slitrk2. PSD-95 has been shown to induce clustering of other postsynaptic membrane proteins as well, including Neuroligins, the inward rectifier potassium channel Kir2.3, and Semaphorin 4B (Burkhardt et al., 2005; Graf et al., 2004; Horio et al., 1997; Levinson et al., 2005; Prange et al., 2004). The observation that removal of the C′-terminal PDZ binding motif from Slitrk2 reduced but did not abolish clustering entirely is in agreement with our mapping analysis, which showed that the PDZ domains of PSD-95 are dispensable for the interaction. Removal of both the PDZ binding motif and SH3 binding site from Slitrk2 reduced clustering to levels indistinguishable from that of Slitrk2 alone, confirming that the interaction is primarily SH3 mediated. Further support for an SH3-mediated interaction comes from the observation that removal of the SH3 domain from PSD-95 reduced Slitrk2 clustering. Removal of the entire intracellular domain of Slitrk2 (ΔICD) also reduced clustering to baseline levels, but did not appear to impact Slitrk2 trafficking to the cell surface, since immunostaining for both full-length Slitrk2 and Slitrk2-ΔICD displayed a peripheral staining pattern characteristic of cell surface proteins. This observation argues against a role for PSD-95 in trafficking Slitrk2 from intracellular compartments to the cell membrane, but further investigation into the membrane trafficking of Slitrk2 is needed.
Another notable observation from our 293T experiment is that clustering of PSD-95 was largely unaffected by the presence or absence of Slitrk2. This finding suggests that the clustering relationship between PSD-95 and Slitrk2 is hierarchical, with PSD-95 driving subcellular changes in Slitrk2 localization. This observation is consistent with prior evidence that PSD-95 clusters at synapses before other postsynaptic proteins (Rao et al., 1998) and argues against a role for Slitrk2 in recruiting PSD-95 to the postsynaptic density. Other molecules and posttranslational modifications, including palmitoylation, Ephrin-B3, and alpha-actinin, have been shown to mediate synaptic localization of PSD-95 (El-Husseini et al., 2000; Hruska et al., 2015; Matt et al., 2018).
As this paper was under review, another paper was published confirming a molecular interaction between Slitrk2 and members of the MAGUK scaffold family (Han et al., 2019). Specifically, they report that Slitrk2 associates with PSD-95, PSD-93 and SAP-102 in postnatal mouse brain, which corroborates our coimmunoprecipitation data in Fig. 1. Han et al. demonstrate that the class I PDZ binding motif (ISQL) in Slitrk2 is important for the interaction with PSD-95 in 293T cells, which matches our finding that removal of the ISQL motif from Slitrk2 compromises PSD-95 binding in yeast. Coincidentally, Han et al. also performed clustering analysis in non-neuronal cells and found that co-expression of Slitrk2 and PSD-95 produced colocalized clusters in the cytoplasm, and removal of the ISQL motif in Slitrk2 appears to reduce this colocalization. Our work complements Han et al.’s clustering data by quantifying the degree of clustering with both wild-type and truncated proteins and confirming that removal of the ISQL motif in Slitrk2 moderately reduces mean cluster size. Overall, the degree of clustering observed in our experiments is noticeably higher, possibly due to the presence of an RFP tag in our PSD-95 expression construct.
Our work is unique from Han et al. (2019) in that we offer evidence of a second, SH3-mediated mode of interaction between Slitrk2 and PSD-95. By performing mapping analysis with a version of PSD-95 containing only the SH3 and GK domains, we show that the PDZ domains of PSD-95 are actually dispensable for the association. The presence of a second, SH3-mediated point of contact between Slitrk2 and PSD-95 is further supported by our clustering experiments, which show that removal of a non-canonical SH3 binding site (DY) from Slitrk2 or removal of the SH3 domain from PSD-95 further reduces clustering, compared to the PDZ-associated truncations alone. Han et al.’s work is unique in demonstrating functional relevance for the Slitrk2-PSD-95 interaction, showing that removal of Slitrk2′s C′-terminal PDZ-binding motif reduces excitatory synapse density and synaptic transmission in hippocampal neurons.
Given the mounting evidence of an interaction between Slitrk2 and PSD-95 and its potential influence on synaptic density, it will be of great interest to determine if PSD-95 directs the subcellular localization of Slitrk2 in neurons. PSD-95 has been shown to localize other cell adhesion molecules to synaptic sites with synaptic sub-type specificity. For example, knockdown of PSD-95 is sufficient to shift Neuroligin1 and Neuroligin3 from excitatory to inhibitory synapses (Gerrow et al., 2006; Levinson et al., 2005), whereas overexpression of PSD-95 shifts localization of Neuroligin2 from inhibitory to excitatory synapses (Graf et al., 2004; Levinson et al., 2005). However, there is also evidence that neuroligins localize to synaptic sites independent of PSD-95 binding (Dresbach et al., 2004), so it will be of great interest to determine if PSD-95 dictates localization of Slitrk2 to synaptic sites, or if it is vital for other aspects of Slitrk function, such as the recruitment of intracellular signaling molecules required for postsynaptic differentiation.

References

Aruga, J., Mikoshiba, K., 2003. Identification and characterization of Slitrk, a novel neuronal transmembrane protein family controlling neurite outgrowth. Mol. Cell.Neurosci. 24 (1), 117–129.
Beaubien, F., Raja, R., Kennedy, T.E., Fournier, A.E., Cloutier, J.F., 2016. Slitrk1 is localized to excitatory synapses and promotes their development. Sci. Rep. 6, 27343.
Bermejo, M.K., Milenkovic, M., Salahpour, A., Ramsey, A.J., 2014. Preparation of synaptic plasma membrane and postsynaptic density proteins using a discontinuous sucrose gradient. Journal of visualized experiments: JoVE 91.
Burkhardt, C., Müller, M., Badde, A., Garner, C.C., Gundelfinger, E.D., Püschel, A.W.,2005. Semaphorin 4B interacts with the post-synaptic density protein PSD-95/SAP90 and is recruited to synapses through a C-terminal PDZ-binding motif. FEBS Lett. 579 (17), 3821–3828.
Chen, J., Yu, S., Fu, Y., Li, X., 2014. Synaptic proteins and receptors defects in autism spectrum disorders. Front. Cell. Neurosci. 8, 276.
Cousins, S.L., Stephenson, F.A., 2012. Identification of NMDA receptor subtype-specific binding sites that mediate direct interactions with the scaffold protein, PSD-95. J.Biol. Chem. jbc–M111.
Cousins, S.L., Kenny, A.V., Stephenson, F.A., 2009. Delineation of additional PSD-95 binding domains within NMDA receptor NR2 subunits reveals differences between NR2A/PSD-95 and NR2B/PSD-95 association. Neuroscience 158 (1), 89–95.
Dresbach, T., Neeb, A., Meyer, G., Gundelfinger, E.D., Brose, N., 2004. Synaptic targeting of neuroligin is independent of neurexin and SAP90/PSD95 binding. Mol. Cell.Neurosci. 27 (3), 227–235.
El-Husseini, A.E., Craven, S.E., Chetkovich, D.M., Firestein, B.L., Schnell, E., Aoki, C., Bredt, D.S., 2000. Dual palmitoylation of PSD-95 mediates its vesiculotubular sorting, postsynaptic targeting, and ion channel clustering. J. Cell Biol. 148 (1), 159–172.
Garcia, E.P., Mehta, S., Blair, L.A., Wells, D.G., Shang, J., Fukushima, T., … Marshall, J., 1998. SAP90 binds and clusters kainate receptors causing incomplete desensitization.Neuron 21 (4), 727–739.
Gerrow, K., Romorini, S., Nabi, S.M., Colicos, M.A., Sala, C., El-Husseini, A., 2006. A preformed complex of postsynaptic proteins is involved in excitatory synapse development. Neuron 49 (4), 547–562.
Graf, E.R., Zhang, X., Jin, S.X., Linhoff, M.W., Craig, A.M., 2004. Neurexins LOXO-195 induce differentiation of GABA and glutamate postsynaptic specializations via neuroligins. Cell 119 (7), 1013–1026.
Han, K.A., Kim, J., Kim, H., Kim, D., Lim, D., Ko, J., Um, J.W., 2019. Slitrk2 controls excitatory synapse development via PDZ-mediated protein interactions. Sci. Rep. 9, 17094.
Horio, Y., Hibino, H., Inanobe, A., Yamada, M., Ishii, M., Tada, Y., … Kurachi, Y., 1997. Clustering and enhanced activity of an inwardly rectifying potassium channel, Kir4. 1, by an anchoring protein, PSD-95/SAP90. Journal of Biological Chemistry 272 (20), 12885–12888.
Hruska, M., Henderson, N.T., Xia, N.L., Le Marchand, S.J., Dalva, M.B., 2015. Anchoring and synaptic stability of PSD-95 is driven by ephrin-B3. Nat. Neurosci. 18 (11), 1594.
Inanobe, A., Fujita, A., Ito, M., Tomoike, H., Inageda, K., Kurachi, Y., 2002. Inward rectifier K+ channel Kir2. 3 is localized at the postsynaptic membrane of excitatory synapses. Am. J. Phys. Cell Phys. 282 (6), C1396–C1403.
Irie, M., Hata, Y., Takeuchi, M., Ichtchenko, K., Toyoda, A., Hirao, K., … Südhof, T.C., 1997. Binding of neuroligins to PSD-95. Science 277 (5331), 1511–1515.
Jang, S., Lee, H., Kim, E., 2017. Synaptic adhesion molecules and excitatory synaptic transmission. Curr. Opin. Neurobiol. 45, 45–50.
Kajiwara, Y., Buxbaum, J.D., Grice, D.E., 2009. SLITRK1 binds 14-3-3 and regulates neurite outgrowth in a phosphorylation-dependent manner. Biol. Psychiatry 66 (10), 918–925.
Kim, S., Burette, A., Chung, H.S., Kwon, S.K., Woo, J., Lee, H.W., Kim, E., 2006. NGL family PSD-95–interacting adhesion molecules regulate excitatory synapse formation.Nat. Neurosci. 9 (10), 1294.
Kornau, H.C., Schenker, L.T., Kennedy, M.B., Seeburg, P.H., 1995. Domain interaction between NMDA receptor subunits and the postsynaptic density protein PSD-95.Science 269 (5231), 1737–1740.
Levinson, J.N., Chéry, N., Huang, K., Wong, T.P., Gerrow, K., Kang, R., … El-Husseini, A., 2005. Neuroligins mediate excitatory and inhibitory synapse Formation: involvement of psd-95 and neurexin-1β in neuroligin-induced synaptic specificity. Journal of Biological Chemistry 280 (17), 17312–17319.
Linhoff, M.W., Laurén, J., Cassidy, R.M., Dobie, F.A., Takahashi, H., Nygaard, H.B., … Craig, A.M., 2009. An unbiased expression screen for synaptogenic proteins identifies the LRRTM protein family as synaptic organizers. Neuron 61 (5), 734–749.
Matt, L., Kim, K., Hergarden, A.C., Patriarchi, T., Malik, Z.A., Park, D.K., … Zhang, Y., 2018. α-Actinin anchors PSD-95 at postsynaptic sites. Neuron 97 (5), 1094–1109.
McGee, A.W., Bredt, D.S., 1999. Identification of an intramolecular interaction between the SH3 and guanylate kinase domains of PSD-95. J. Biol. Chem. 274 (25), 17431–17436.
Prange, O., Wong, T.P., Gerrow, K., Wang, Y.T., El-Husseini, A., 2004. A balance between excitatory and inhibitory synapses is controlled by PSD-95 and neuroligin. Proc. Natl.Acad. Sci. 101 (38), 13915–13920.
Proenca, C.C., Gao, K.P., Shmelkov, S.V., Rafii, S., Lee, F.S., 2011. Slitrks as emerging candidate genes involved in neuropsychiatric disorders. Trends Neurosci. 34 (3), 143–153.
Rao, A., Kim, E., Sheng, M., Craig, A.M., 1998. Heterogeneity in the molecular composition of excitatory postsynaptic sites during development of hippocampal neurons in culture. J. Neurosci. 18 (4), 1217–1229.
Saksela, K., Permi, P., 2012. SH3 domain ligand binding: What’s the consensus and where’s the specificity? FEBS Lett. 586 (17), 2609–2614.
Sans, N., Petralia, R.S., Wang, Y.X., Blahos, J., Hell, J.W., Wenthold, R.J., 2000. A developmental change in NMDA receptor-associated proteins at hippocampal synapses. J. Neurosci. 20 (3), 1260–1271.
Sheng, M., Kim, E., 2011. The postsynaptic organization of synapses. Cold Spring Harb. Perspect. Biol. a005678.
Südhof, T.C., 2008. Neuroligins and neurexins link synaptic function to cognitive disease.Nature 455 (7215), 903.
Sun, Q., Turrigiano, G.G., 2011. PSD-95 and PSD-93 play critical but distinct roles in synaptic scaling up and down. J. Neurosci. 31 (18), 6800–6808.
Takahashi, H., Craig, A.M., 2013. Protein tyrosine phosphatases PTPδ, PTPσ, and LAR:presynaptic hubs for synapse organization. Trends Neurosci. 36 (9), 522–534.
Takahashi, H., Katayama, K.I., Sohya, K., Miyamoto, H., Prasad, T., Matsumoto, Y., … Craig, A.M., 2012. Selective control of inhibitory synapse development by Slitrk3PTPδ trans-synaptic interaction. Nature neuroscience 15 (3), 389.
Um, J.W., Kim, K.H., Park, B.S., Choi, Y., Kim, D., Kim, C.Y., … Choii, G., 2014. Structural basis for LAR-RPTP/Slitrk complex-mediated synaptic adhesion. Nature communications 5, 5423.
Williams, M.E., de Wit, J., Ghosh, A., 2010. Molecular mechanisms of synaptic specificity in developing neural circuits. Neuron 68 (1), 9–18.
de Wit, J., Ghosh, A., 2014. Control of neural circuit formation by leucine-rich repeat proteins. Trends Neurosci. 37 (10), 539–550.
de Wit, J., Ghosh, A., 2016. Specification of synaptic connectivity by cell surface interactions. Nat. Rev. Neurosci. 17 (1), 4.
Won, S., Levy, J.M., Nicoll, R.A., Roche, K.W., 2017. MAGUKs: multifaceted synaptic organizers. Curr. Opin. Neurobiol. 43, 94–101.
Yim, Y.S., Kwon, Y., Nam, J., Yoon, H.I., Lee, K., Kim, D.G., … Ko, J., 2013. Slitrks control excitatory and inhibitory synapse formation with LAR receptor protein tyrosine phosphatases. Proceedings of the National Academy of Sciences 110 (10), 4057–4062.
Zeng, M., Ye, F., Xu, J., Zhang, M., 2018. PDZ ligand binding-induced conformational coupling of the PDZ–SH3–GK tandems in PSD-95 family MAGUKs. J. Mol. Biol. 430 (1), 69–86.
Zheng, C.Y., Seabold, G.K., Horak, M., Petralia, R.S., 2011. MAGUKs, synaptic development, and synaptic plasticity. Neuroscientist 17 (5), 493–512.