The neuronal protein Neurexin directly interacts with the Scribble-Pix complex to stimulate F-actin assembly for synaptic vesicle clustering

Abstract

Synaptic vesicles (SVs) form distinct pools at synaptic terminals, and this well-regulated separation is necessary for normal neurotransmission. However, how the SV cluster, in particular synaptic compartments, maintains normal neurotransmitter release remains a mystery. The presynaptic protein Neurexin (NRX) plays a significant role in synaptic architecture and function, and some evidence suggests that NRX is associated with neurological disorders, including autism spectrum disorders. However, the role of NRX in SV clustering is unclear. Here, using the neuromuscular junction at the 2-3 instar stages of Drosophila larvae as a model and biochemical imaging and electrophysiology techniques, we demonstrate that Drosophila NRX (DNRX) plays critical roles in regulating synaptic terminal clustering and release of SVs. We found that DNRX controls the terminal clustering and release of SVs by stimulating presynaptic F-actin. Furthermore, our results indicate that DNRX functions through the scaffold protein Scribble and the GEF protein DPix to activate the small GTPase Ras-related C3 Botulinum toxin substrate 1 (Rac1). We observed a direct interaction between the C-terminal PDZ-binding motif of DNRX and the PDZ domains of Scribble and that Scribble bridges DNRX to DPix, forming a DNRX-Scribble-DPix complex that activates Rac1 and subsequently stimulates presynaptic F-actin assembly and SV clustering. Taken together, our work provides important insights into the function of DNRX in regulating SV clustering, which could help inform further research into pathological neurexin-mediated mechanisms in neurological disorders such as autism.

Keywords: Drosophila; actin; cell adhesion; synapse; vesicles.

Figure 1.

DNRX is necessary for synaptic terminal aggregation and release of SVs. A–C″, synaptic boutons of wild-type (A–A″), dnrx mutant (B–B″), and pre-synaptic rescue (C–C″) at two instar larvae stage double stained for SYT (red) and HRP (green), which labels the pre-synaptic SVs and neuronal membrane, respectively. D–F″, synaptic boutons of wild-type (D–D″), dnrx mutant (E–E″), and pre-synaptic rescue (F–F″) at third instar larvae stage double stained for SYT (red) and HRP (green), which labels the pre-synaptic SVs and neuronal membrane, respectively. G–I″, synaptic boutons of wild-type (G–G″), dnrx mutant (H–H″), and pre-synaptic rescue (I–I″) at third instar larvae stage double stained for SYN (red) and HRP (green), which labels the pre-synaptic SVs and neuronal membrane, respectively. J, quantification of the SYT-diffused boutons ratio at two instar larvae stage in muscle 4 shows that dnrx mutants have disturbed the distribution of SVs and this phenotype can be rescued by pre-synaptic DNRX. K, quantification of the SYT-diffused boutons ratio at the third instar larvae stage in muscle 4 shows that dnrx mutants have disturbed the distribution of SVs and this phenotype can be rescued by pre-synaptic DNRX. L, quantification of the SYN-diffused boutons ratio at third instar larvae stage in muscle 4 shows that dnrx mutants have disturbed the distribution of SVs and this phenotype can be rescued by pre-synaptic DNRX. M–N″, confocal images of third instar larvae NMJ boutons double labeled with anti-BRP (red) and anti-GFP (green) in syt::gfp (M–M″), and dnrx²⁷³,syt::gfp mutants (N–N″) showing that loss of DNRX disrupts the distribution of terminal SVs. O–P, amplified confocal images of single third instar larvae NMJ bouton double labeled with anti-Brp (red) and anti-GFP (green) in syt::gfp (O), and dnrx²⁷³,syt::gfp (P) showing the dispersed distribution of SVs in dnrx mutants. Q, representative diagrammatic sketch of SVs in single bouton of wild-type and dnrx mutant. R–R″, STED images of third instar larvae single bouton of wild-type (R), dnrx mutant (R′), and pre-synaptic rescue (R″) labeled with SYT, showing the SVs were diffused in dnrx mutant and can be rescued by pre-synaptic DNRX. S, representative traces of spontaneous responses of indicated genotypes. T, quantification of mEJP frequency of the indicated genotypes, showing that the mEJP frequency was increased in dnrx mutant and the DNRX could fully rescue the defects at pre-synapse. Data are mean ± S.E. ***, p < 0.001; **, p < 0.01; and *, p < 0.05. ns, not significant. Two-tailed Student's t tests were used to compare genotypes. Scale bar, 5 (A–C″), 5 (D–F″), 2 (G–G″), 2 (H–H″), 2 (I–I″), 5 (M–N″), 2.5 (O–P), and 2.5 μm (R–R″).

Figure 2.

DNRX regulates SV clustering and release dependent on presynaptic F-actin. A–C‴, confocal images of third instar larvae NMJ type Ib boutons at muscles 12/13 triple labeled with Texas Red phalloidin (red), anti-DLG (green), and anti-HRP (blue) in wild-type (A), dnrx mutants (B), ok6>dnrx rescue (C). The phalloidin signal contained in white circles correspond to HRP, largely reflecting the presynaptic F-actin. D and E, summary graph of relative fluorescence intensity of F-actin correspond to HRP (D) and DLG (E) showing that the F-actin fluorescence intensities were significantly reduced in dnrx mutant and can be restored to normal levels. F–G‴, representative confocal images of third instar larvae NMJ labeled with anti-HRP (blue), anti-SYT (red), and anti-GFP (green) in wild-type and dnrx mutant, respectively. The light green spot represents the pre-synaptic F-actin decreased in the dnrx mutant. H–L″, synaptic boutons of wild-type (H–H″), dnrx mutant (I–I″), cortactin mutant (J–J″), and pre-synaptic overexpress Cortactin (K–K″), and the active form of DPak (L–L″) in the dnrx mutant double stained for SYT (red) and HRP (green), which labels pre-synaptic SVs and neuronal membrane, respectively. M, quantification of SYT-diffused bouton ratio shows presynaptic branched F-actin can fully rescue the diffused distribution of SVs in dnrx mutant. N, quantification of mEJP frequency of the indicated genotypes, showing that mEJP frequency was increased in dnrx mutant and the Cortactin, the active form of DPak could rescue the defects at pre-synapse. O, representative traces of spontaneous responses of the indicated genotypes. Data are mean ± S.E. ***, p < 0.001; **, p < 0.01; and *, p < 0.05. ns, not significant. Two-tailed Student's t tests were used to compare genotypes. Scale bar, 5 (A–C‴), 5 (F–G‴), and 5 μm (H–L″).

Figure 3.

DNRX is co-localized with Scribble in the nervous system. A–A″, staining for scribble::gfp knock-in fly embryo with anti-DNRX (blue) and GFP is the spontaneous green fluorescence, showing that DNRX and Scribble are co-localized in the central nervous system. B–B″, third instar larvae brains of scribble::gfp staining with anti-DNRX (red) and GFP is the spontaneous green fluorescence, showing that DNRX and Scribble are co-localized in the central nervous system especially concentrated at mushroom body. C–C‴, confocal images of the third instar larvae neuromuscular junction at muscles 6/7 staining with anti-DNRX (blue), anti-DLG (red), and GFP is the spontaneous green fluorescence in scribble::gfp integrated with the DNRX overexpression fly, showing that DNRX and Scribble were co-localized in neuromuscular junction. D–D″, scribble::gfp knock-in fly adult brain staining with anti-DNRX (red) and GFP are the spontaneous green fluorescence, showing that DNRX and Scribble are co-localized in the central nervous system. E–E‴, epithelial cell staining with anti-DNRX, anti-DLG (red), and GFP is the spontaneous green fluorescence, showing that DNRX and Scribble are not co-localized at epithelial cells. Scale bar, 50 (A–A″), 50 (B–B″), 20 (C–C‴), 100 (D–D″), and 20 μm (E–E‴).

Figure 4.

DNRX binds with the PDZ domains of Scribble via the C-terminal PDZ-binding motif. A, schematic representation of the protein structure of DNRX and Scribble. The polypeptide structure of the GST-fused DNRX C-terminal section with and without the C-terminal PDZ-binding motif and the PDZ domains of Scribble with HA and HIS tags. B, schematic graph of full-length HA-tagged Scribble of Drosophila and HIS-tagged α-Neurexin of mouse constructs. C, immunoprecipitated (IP) from whole Drosophila adult brain extracts with GFP nano-antibody fused beads in scribble::gfp and wild-type strains, respectively. Anti-GFP antibody precipitated the DNRX from the lysates, whereas the control IgG and wild-type sample did not. Suggesting DNRX physically interacts with Scribble in vivo. Input is 1%. D, Western blots of anti-His immunoprecipitates using protein lysate of HEK293 cells that can express the proteins of HA-Scribble and HIS-α-Neurexin, showing that HA-Scribble was coimmunoprecipitated with HIS-α-Neurexin in mouse. E, Coomassie Brilliant Blue for the purified DNRX C-terminal protein and deleted final 7-amino acid protein. F, Western blot validation for the purified 4 PDZ domain polypeptides of Scribble. G and H, GST pulldown with the DNRX C-terminal protein and deleted a very C-terminal 7-amino acid protein, respectively. Incubating with 4 PDZ domains of Scribble. Western blot results showed DNRX directly binds with all the PDZ domains of Scribble.

Figure 5.

DNRX interacts with Scribble and sustains the terminal SV aggregation. A–B‴, representative images of larvae brains of the indicated genotypes were stained with DNRX (red), HRP (green), and DAPI (blue), showing the level of DNRX was decreased in scribble mutant. C–D″, representative images of larvae NMJ of the indicated genotypes were stained with DNRX (green) and HRP (red), showing the level of DNRX was decreased in the scribble mutant. E, quantification of the fluorescence intensity for DNRX both in brain and NMJ of the indicated genotypes. F and G, Western blot analysis of protein lysates prepared from heads using anti-DNRX antibody showing that the level of DNRX was reduced in scribble mutant both in larvae (F) and adult (G) stages. H, quantitative analysis of data for Western blot analysis of the indicated genotypes for adult flies. I, Western blot analysis of protein lysates prepared from heads using anti-GFP antibody showing that the level of Scribble was reduced in the dnrx mutant in larvae stage. J, quantitative analysis for Western blots of the indicated genotypes. K–L″, representative images of larvae brain of the indicated genotypes stained with DNRX (red) and GFP (green), showing the level of Scribble was decreased in dnrx mutant. M–N‴, representative images of larvae NMJ of the indicated genotypes were stained with HRP (blue), DLG (red), and GFP (green), showing the level of Scribble was decreased in dnrx mutant. O, quantification of the fluorescence intensity for Scribble::GFP of the indicated genotypes. P–R″, synaptic boutons of wild-type (P–P″), dnrx mutant (Q–Q″), and pre-synaptic overexpress Scribble in dnrx mutant (R–R″) double stained for SYT (red) and HRP (green), which labels the pre-synaptic SVs and neuronal membrane, respectively. S, quantification of the ratio of SYT-diffused boutons in muscle 4 shows that dnrx mutants have disturbed the terminal cluster of SVs and this phenotype can be rescued by pre-synaptic Scribble. Data are mean ± S.E. ***, p < 0.001; **, p < 0.01; and *, p < 0.05. Two-tailed Student's t tests were used to compare genotypes. Scale bar, 50 (A–B‴), 5 (C–D″), 50 (K–L″), 20 (M–N‴), and 5 μm (P–R″).

Figure 6.

DNRX interacts with DPix through Scribble and regulates the activity of Rac1, and maintains the terminal SV aggregation and release. A, Western blots of anti-GFP immunoprecipitates using protein lysates of adult Drosophila heads showing immunoprecipitated proteins for DNRX and DPix in vivo. Input is 1%. B, Western blots of anti-GFP immunoprecipitates using protein lysate of adult Drosophila heads showing loss of Scribble, DNRX could not immunoprecipitate with DPix in vivo. Input is 1%. C, schematic representation of DNRX interacts with DPix through Scribble and activates Rac1. D, Western blots showing the level of Rac-GTP and total Rac1 in adult heads of wild-type, dnrx mutant. E, summary graph showing that the activated Rac1 was decreased after knocking out DNRX. F, Western blots showing the level of Rac-GTP and total Rac1 in adult heads of elav dcr/+, dnrx, and scribble knocking down flies. G, summary graph showing that the activated Rac1 was decreased after knocking down the expression of DNRX and Scribble using the RNAi method. H, summary graph showing that the mean fluorescence intensity of DPix-GFP was decreased both in NMJ and VNC in dnrx mutant when compared with wild-type controls. I and J″, representative images of larvae VNC of the indicated genotypes were stained with DNRX (red) and GFP (green), showing DNRX regulates the amount of DPix. K and L‴, representative images of larvae NMJ type Ib boutons of the indicated genotypes were stained with HRP (blue), DLG (red), and GFP (green), showing DNRX regulates the amount of DPix. M–O″, confocal images of third instar larvae NMJ type Ib boutons of the indicated genotypes double labeled with anti-SYT (red) and anti-HRP (green), which labels the pre-synaptic SVs and neuronal membrane, respectively. P, quantification of the ratio of the SYT-diffused boutons in muscle 4 shows that in dnrx mutants the distribution of SVs has been disturbed and this phenotype can be rescued by pre-synaptic DPix. Q, representative traces of spontaneous responses of the indicated genotypes. R, quantification of mEJP frequency of indicated genotypes, showing that mEJP frequency was increased in the dnrx mutant and the Scribble and DPix could rescue the defects at pre-synapse. Data are mean ± S.E., ***, p < 0.001; **, p < 0.01; and *, p < 0.05. ns, not significant. Two-tailed Student's t tests were used to compare genotypes. Scale bars, 50 (I–J″), 5 (K-L‴), and 5 μm (M–O″).

Figure 7.

The PDZ-binding motif is essential for the effect of DNRX on F-actin assembly and SV release. A–D‴, representative images of third instar larvae NMJ type Ib boutons at muscles 12/13 triple labeled with Texas Red phalloidin (red), anti-DLG (green), and anti-HRP (blue) in wild-type (A), dnrx mutants (B), dnrx rescue (C), and dnrx^ΔPDZ rescue (D). The phalloidin signal contained in white circles corresponds to HRP, and largely reflects the presynaptic F-actin. E and F, summary graph of the relative fluorescence intensity of F-actin corresponding to HRP (E) and DLG (F) showing that F-actin fluorescence intensities were significantly reduced in dnrx mutant and could not be rescued by inducing DNRX^ΔPDZ at presynapse. G, representative traces of spontaneous responses of the indicated genotypes. H, quantification of mEJP frequency of the indicated genotypes, showing that mEJP frequency was increased in the dnrx mutant and that DNRX^ΔPDZ could not rescue the defects. I, Western blots showing the level of Rac-GTP and total Rac1 in adult heads of wild-type, dnrx mutant, dnrx rescue, and dnrx^ΔPDZ rescue. J, summary graph showing that the activated Rac1 was decreased after knocking out DNRX and this defect can be rescued by inducing the full-length of DNRX at presynapse, however, driving DNRX^ΔPDZ at presynapse in the dnrx mutant had no rescue effect. Data are mean ± S.E. ***, p < 0.001; **, p < 0.01; and *, p < 0.05. ns, not significant. Two-tailed Student's t tests were used to compare genotypes. Scale bar, 5 μm (A–D‴).

Figure 8.

The model shows that DNRX regulates SV aggregation and release via presynaptic F-actin by modulating Scribble–DPix. Presynaptic DNRX interacts with Scribble–DPix to regulate the activity of Rac1, the activated Rac1 (Rac1-GTP) then affects actin polymerization to finally modulate the SVs localization and release.