1. Volume 3, Issue 6, Pages 1996-2007 (June 2013)
Neurexin-1β Binding to Neuroligin-1 Triggers the Preferential Recruitment of PSD-95 versus Gephyrin through Tyrosine Phosphorylation of Neuroligin-1 
Grégory Giannone, Magali Mondin, Dolors Grillo-Bosch, Béatrice Tessier, Edouard Saint-Michel, Katalin Czöndör, Matthieu Sainlos, Daniel Choquet, Olivier Thoumine 
Cell Reports 
Volume 3, Issue 6, Pages (June 2013)
DOI: /j.celrep
Copyright © 2013 The Authors Terms and Conditions

2. Cell Reports 2013 3, 1996-2007DOI: (10.1016/j.celrep.2013.05.013)
Copyright © 2013 The Authors Terms and Conditions

3. Figure 1
Differential Formation of PSD-95 Clusters by Nrx1β-Occupied versus Unoccupied Nlg1
(A and B) Time-course images of PSD-95:GFP (green) in neurons coexpressing HA-Nlg1 and incubated with Nrx1β (A) or anti-HA (B), preclustered by Cy5-conjugated secondary antibodies (red).
(C) Kinetics of new PSD-95:GFP cluster formation upon addition of crosslinked Nrx1β or anti-HA, normalized by the number of pre-existing PSD-95:GFP clusters. Individual curves are shown (thin lines) together with the average of 5–8 cells (thick lines).
(D) Cumulative distributions of the normalized ratio between PSD-95:GFP and Cy5 enrichments.
(E) Mean ± SEM of the ratio between PSD-95:GFP and Cy5 enrichment levels for each treatment. Populations were analyzed by one-way ANOVA and compared by Dunn’s posttest (∗∗∗p < 0.001).
(F) Neurons expressing Nlg1WT were incubated with crosslinked Nrx1β (red) for 1 hr and then immunostained for endogenous PSD-95 (green).
(G) Recruitment kinetics of endogenous PSD-95 (dashed lines) and PSD-95:GFP (plain lines) at Nrx1β clusters. The enrichment factor is presented as the mean ± SEM per cell, and data points were compared with the control 10 min condition by one-way ANOVA, followed by nonparametric Dunnett’s test (∗p < 0.05; ∗∗p < 0.01).
See also Figures S1 and S2 and Movies S1 and S2.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

4. Figure 2
Turnover of Nlg1 and PSD-95 at Nrx1β- or Anti-HA-Induced Clusters
(A) PSD-95:GFP clusters induced by Cy5-labeled crosslinked Nrx1β for 1 hr.
(B) FRAP experiment on a PSD-95:GFP cluster (dashed circle).
(C) FRAP curves on Nlg1:GFP clusters crosslinked by Nrx1β (red) or anti-HA (black).
(D) FRAP curves on PSD-95:GFP. Control measurements were made on unbleached Nlg1:GFP or PSD-95:GFP clusters (gray curves). Data are presented as the mean ± SEM of clusters.
See also Table S1.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

5. Figure 3
Differential Mobility of Ligand-Occupied versus Unoccupied Nlg1 Molecules Revealed by Single-Nanoparticle Tracking
(A–C) Schematics and representative trajectories of Qdots coated with anti-HA or Nrx1β in neurons expressing Homer1c:GFP (gray pixels), and Nlg1WT or Nlg1ΔCter. Trajectories exhibit both freely diffusive (blue) and confined (red) episodes. Graphs show the corresponding diffusion coefficient (blue) and confinement index (red) over time. A decrease in diffusion coefficient is matched by an increase in confinement index.
(D) Distributions of diffusion coefficients on a logarithmic scale. The confined and freely diffusive events correspond respectively to the leftward and rightward populations fitted by Gaussian curves. Inset: zoom on the distribution of confined events.
(E) Percentage of confined events. The confined fractions obtained by double Gaussian fitting of the distributions were compared by unpaired Student’s t tests (∗∗p < 0.01; ∗∗∗∗p < ).
(F) MSD as a function of time at postsynapses.
(G and H) Neurons were cotransfected with Nlg1 plus either shRNA against PSD-95 or a plasmid containing both the shRNA sequence and PSD-95:GFP.
(G) Distributions of diffusion coefficients on a logarithmic scale.
(H) Percentage of confined events (∗p < 0.05; ∗∗∗p < ).
See also Figure S3.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

6. Figure 4
Acute Binding of Soluble Nrx1β Quickly Freezes the Lateral Motion of Nlg1 Molecules
(A and B) Schematics and representative trajectories of anti-HA Qdots (red curves) on neurons expressing Homer1c-GFP (gray pixels), acquired before (A) or after 10 min incubation with soluble Nrx1β (B).
(C and D) Corresponding diffusion coefficients versus time, superimposed for several Qdots. Note the almost zero free diffusion upon Nrx1β addition.
(E and F) Distributions of diffusion coefficients before (black) and after (red) addition of Nrx1β multimers, for Nlg1WT and Nlg1ΔCter.
(G) Ratio of the fractions of confined events, measured after and before addition of Nrx1β or vehicle. Confined fractions determined by double Gaussian fitting were compared before and after treatment by paired Student’s t test (∗∗∗p < ; ∗∗∗∗p < 0.001).
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

7. Figure 5
Nlg1 Tyrosine Phosphorylation Controls Direct Binding to Gephyrin
(A and B) Neurons (6–7 DIV) cotransfected with Nlg1WT and PSD-95:GFP were incubated with crosslinked Nrx1β-Fc for 30 min in the presence of 5 mM orthovanadate, 50 μM genistein, or control solution.
(A) Representative images.
(B) Cumulative distributions of PSD-95:GFP enrichment factor. Populations were analyzed by nonparametric one-way ANOVA and compared by Dunn’s posttest.
(C) Sequence alignment of the cytoplasmic portion of rat Nlg1, Nlg2, and Nlg3. The conserved gephyrin and PSD-95 PDZ domain-binding motifs are indicated by boxes, and the unique Y782 is shown in red.
(D) Extracts of COS cells nontransfected (−) or transfected with Nlg1WT or Nlg1YA were immunoprecipitated with anti-HA and immunoblotted for anti-HA or anti-pY.
(E) Extracts of neurons (8 DIV) were immunoprecipitated with antibodies to endogenous Nlg1 and immunoblotted with anti-Nlg1 or anti-pY.
(F) Brain extracts from WT or Nlg1 KO mice were immunoprecipitated and immunoblotted with anti-Nlg1. FT, flow-through; SM, starting material.
(G) Sequences of the biotinylated Nlg1 peptides encompassing the gephyrin-binding motif, with Y782 and modifications highlighted. λ represents Norleucine.
(H) Nlg1 peptides immobilized on streptavidin-coated beads were used to pull down recombinant gephyrin in vitro. Samples were run on polyacrylamide gels and Coomassie stained. The three bands correspond to 6His:gephyrin (82 kDa), BSA (69 kDa) to reduce nonspecific binding, and streptavidin (13 kDa) to control sample loading.
See also Table S2.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

8. Figure 6
Recruitment of Gephyrin by Nlg1 Tyrosine Point Mutants in Cluster and Synapse Induction Assays
(A) Neurons (8 DIV) cotransfected with Venus:gephyrin plus Nlg1WT, Nlg1YA, or Nlg1YF were incubated with crosslinked Nrx1β-Fc.
(B) Cumulative distributions of the Venus:gephyrin enrichment factor at Nrx1β-Fc clusters. Populations were analyzed by nonparametric one-way ANOVA and compared by Dunn’s posttest (∗∗∗p < 0.001).
(C) Neurons (9–10 DIV) transfected with GFP alone or cotransfected with GFP plus Nlg1WT or Nlg1 point mutants were immunostained for endogenous synaptotagmin and gephyrin.
(D) Numbers of synaptotagmin or gephyrin puncta per dendrite area (mean ± SEM). Lower bars represent gephyrin puncta opposed to synaptotagmin puncta, and upper bars correspond to extrasynaptic gephyrin puncta. Populations were analyzed by parametric one-way ANOVA and Bonferroni’s posttest (∗∗∗p < 0.001; ns, not significant).
See also Figures S4 and S5.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

9. Figure 7
Recruitment of PSD-95 by Nlg1 Tyrosine Point Mutants in Cluster and Synapse Induction Assays
(A and C) Neurons (6–7 DIV) cotransfected with PSD-95:GFP plus Nlg1WT or Nlg1 point mutants were incubated with crosslinked Nrx1β-Fc or anti-HA.
(B and D) Cumulative distributions of the PSD-95:GFP enrichment factor. Populations were analyzed by nonparametric one-way ANOVA and compared by Dunn’s posttest (∗∗∗p < 0.001; ns, not significant).
(E) Neurons (8–10 DIV) transfected with GFP alone or cotransfected with GFP plus Nlg1WT or Nlg1 point mutants were immunostained for endogenous synaptotagmin and PSD-95.
(F) Numbers of PSD-95 puncta per dendrite area (mean ± SEM). Lower bars represent PSD-95 puncta opposed to synaptotagmin puncta, and upper bars correspond to extrasynaptic PSD-95 puncta. Populations were analyzed by parametric one-way ANOVA and Bonferroni’s posttest (∗∗∗p < 0.001).
See also Figure S6.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

10. Figure S1
Anti-HA and Nrx1β Induce Similar Crosslinking of Nlg1, but Nrx1β Triggers a More Prominent Recruitment of PSD-95, Related to Figure 1
(A) Untransfected neurons at 8 DIV were incubated with cross-linked Nrx1β for 30 min, then immunostained for endogenous Nlg1. The enrichment of endogenous Nlg1 at Cy5 clusters was 1.20 ± 0.04 (9 cells/239 clusters).
(B–G) 6-7 DIV hippoccampal neurons transfected with Nlg1:GFP were treated for 30 min with Nrx1β (B) or anti-HA (C) cross-linked with Cy5-conjugated secondary antibodies. Nlg1:GFP shows a fairly diffuse membrane distribution at this developmental stage (left panels). After 30 min treatment, both cross-linked Nrx1β and anti-HA induced a prominent and fast formation of Nlg1:GFP aggregates (middle panels), which co-localized almost perfectly with Cy5-labeled surface clusters (right panels).
(D) Density of Nrx1β or anti-HA clusters induced on 8 DIV neurons expressing Nlg1WT. The densities were similar in both conditions (unpaired t test, p > 0.05, n = cells out of 4-5 experiments).
(E) Number of Nlg1:GFP clusters induced by cross-linked Nrx1β or anti-HA for 30 min, normalized by the number of pre-existing Nlg1:GFP clusters (p > 0.05 by unpaired t test).
(F) Mean size of Nrx1β or anti-HA clusters. The sizes were similar in both conditions (unpaired t test, p > 0.05).
(G) The mean enrichment factor of Nlg1:GFP at contact sites compared to the surrounding Nlg1:GFP was higher for cross-linked Nrx1β (ratio 5.3 ± 0.3; 4 cells, 48 clusters) than for cross-linked anti-HA (2.8 ± 0.1, 5 cells, 66 clusters). However, the ratio between Nlg1:GFP and Cy5-labeled secondary antibodies enrichment for either Nrx1β (0.74 ± 0.03; 4 cells; 48 clusters) or anti-HA antibodies (0.72 ± 0.04; 5 cells, 66 clusters), were comparable (unpaired t test, p > 0.05). Thus, the Cy5 signal was used thereafter to normalize the amount of Nlg1 contained at contact sites in both conditions.
(H) Nrx1β-Fc before or after reduction using DTT was separated by SDS-PAGE under non-reducing conditions, and revealed by Western blot using an anti-human Fc antibody. Before reduction, both monomeric and multimeric forms of Nrx1β-Fc were present. After DTT treatment, all multimeric Nrx1β-Fc was reduced to monomeric Nrx1β-Fc. To check if monomeric Nrx1β-Fc could re-multimerize over time, we let the purified protein solution up to 24 hr in culture medium before running the gel. Note that the monomeric form stays predominant.
(I) 7 DIV neurons co-transfected with Nlg1 and PSD-95:GFP (green) were immuno-stained for synapsin (red) after Nrx1β cross-linking for 1 hr. Contrary to pre-existing PSD-95 clusters, newly formed PSD-95 clusters did not co-localize with synapsin (arrows).
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

11. Figure S2
Quantification of Enrichment Factors and Influence of Endogenous Nlg1, Related to Figure 1
(A–C) 6-7 DIV hippoccampal neurons co-transfected with PSD-95:GFP and Nlg1WT (A, B) or Nlg1ΔCter (C) were treated for 30 min with Nrx1β (A,C) or anti-HA (B) cross-linked with Cy5-conjugated secondary antibodies. PSD-95:GFP aggregates (left), Cy5 clusters as raw and segmented images (middle), and mask superimposed to the PSD-95:GFP image (right). Cy5-labeled Nrx1β clusters which reflect Nlg1 aggregation, are co-localized with intense PSD-95:GFP clusters, whereas Cy5-labeled anti-HA clusters recruit less PSD-95:GFP. Cy5-labeled Nrx1β clusters recruit very faint PSD-95:GFP fluorescence when cells express Nlg1ΔCter.
(D–F) 8 DIV primary neurons from Nlg1 KO mice were co-transfected with Nlg1WT and PSD-95:GFP and incubated with either Nrx1β-Fc (D) or anti-HA (E) cross-linked with Cy5 labeled secondary antibodies.
(F) Cumulative distributions of the PSD-95:GFP enrichment factor at Nrx1β-Fc (red) or anti-HA (black) clusters. Populations were compared using a non-parametric Mann-Whitney t test (p < , anti-HA (6 cells, 309 clusters) versus Nrx1β (7/349)).
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

12. Figure S3
Increased Confinement of Nlg1 at Synapses Depends on Interaction with PSD-95, Related to Figure 3
Distributions of diffusion coefficients on a logarithmic scale, for synaptic (red), extra-synaptic (blue), and total (black) trajectories.
(A–C) Both unoccupied Nlg1 (QDHA) (A), occupied Nlg1 (QDNrx1β) (B), and unoccupied Nlg1SWAP (QDHA) (C), displayed more confined events for the synaptic compared to extra-synaptic location.
(D) However, no differences were found between synaptic and extra-synaptic locations for unoccupied Nlg1ΔCter (QDHA). Those results support the view that the increased confinement of Nlg1 at synapses depends on an interaction between Nlg1 and PSD-95.
(E) Distributions of diffusion coefficients on a logarithmic scale for all conditions. The distributions were fitted by two Gaussian curves (dotted lines) corresponding to confined (left) and free diffusive (right) events.
(F) Cumulative distributions of the confinement probability for all conditions. Ligand occupied Nlg1 (QDNrx1β) (red) were more often confined than unoccupied Nlg1 (QDHA) (black). However, no differences were found between unoccupied Nlg1 and unoccupied Nlg1SWAP (QDHA) (blue). Unoccupied Nlg1ΔCter (QDHA) (green) were mostly free diffusive.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

13. Figure S4
Density of Anti-HA and Nrx1β Clusters on Neurons Expressing Nlg1 Tyrosine Point Mutants, Related to Figure 6
(A and B) 8 DIV hippoccampal neurons transfected with either Nlg1WT, Nlg1YA or Nlg1YF, were treated for 30 min with Nrx1β ligands (A) or anti-HA antibodies (B), cross-linked with Cy5-conjugated secondary antibodies. The number of clusters per unit area of dendrite was quantified. Conditions were compared by parametric one-way ANOVA.
(A) Density of Nrx1β clusters. No significant differences between Nlg1 mutants, n = cells out of 4-5 experiments.
(B) Density of anti-HA clusters. No significant differences between Nlg1 mutants, n = 9-23 cells out of 2-5 experiments.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

14. Figure S5
Endogenous Nlg1 Colocalizes Strongly with PSD-95 and to a Lesser Extent with Gephyrin Puncta, Related to Figure 6
(A and B) Untransfected neurons were immunostained at DIV 10 for endogenous Nlg1 and PSD-95 (A), or endogenous Nlg1 and gephyrin (B). Solid arrowheads indicate strong colocalization of endogenous Nlg1 with PSD-95 or gephyrin puncta, while open arrowheads indicate no recruitment.
(C) Cumulative distributions of the enrichment factor of endogenous Nlg1 at PSD-95 or gephyrin puncta. 10 cells were analyzed for each condition. Statistics are as follows: 877 PSD-95 clusters, median Nlg1 enrichment = 1.37; and 574 gephyrin clusters, median Nlg1 enrichment = Populations were compared by unpaired Student t test (∗∗∗p < 0.005). The no enrichment condition corresponds to an enrichment factor equal to 1.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions

15. Figure S6
Model of Excitatory and Inhibitory Scaffold Assembly Based on Nrx-Induced Changes in the Phosphotyrosine Level of Nlgs, Related to Figure 7 and Discussion
Schematic diagram summarizing the effects of Nlg1 aggregation versus Nrx1β binding on PSD-95 and gephyrin recruitment, depending on the tyrosine phosphorylation state of Nlg1. In terms of Nlg1 phosphorylation and scaffold recruitment, the left and right panels are likely to correspond to the situations of inhibitory and excitatory synapses, respectively.
Cell Reports 2013 3, DOI: ( /j.celrep )
Copyright © 2013 The Authors Terms and Conditions