A biophysical and molecular characterization of the interaction between the Alzheimer risk factor BIN1 and the neuronal scaffold protein p140Cap

Abstract

Bridging integrator 1 (BIN1) is a genetic risk factor for late-onset Alzheimer disease. BIN1's participation in endocytosis, membrane remodeling, and modulation of actin dynamics is well-characterized in non-neuronal cells. In neurons, BIN1 is enriched at presynaptic sites, where it facilitates excitatory neurotransmitter vesicle release. However, how BIN1 is involved in synaptic vesicle dynamics is not well understood. A C-terminal Src homology 3 (SH3) domain is invariant in all BIN1 isoforms and promotes protein-protein interactions with proteins harboring proline-rich motifs. While BIN1 interactions with dynamin, synaptojanin, RIN3, and tau have been identified and experimentally validated, the list of BIN1-interacting molecules is not exhaustive. Here, we report the neuronal scaffolding protein p140Cap, encoded by SRC kinase signaling inhibitor 1, as a BIN1 SH3 domain-interacting protein. We performed surface plasmon resonance to ascertain the affinity of BIN1-SH3 domain for p140Cap and identified a peptide containing three proline-rich motifs that exhibited biologically relevant affinity (K_D = 7.7 μM). Additional surface plasmon resonance experiments, coupled with alanine-scanning mutagenesis, revealed that two class II motifs, but not a class I motif, in p140Cap facilitated binding. Confocal microscopy and proximity ligation assays confirmed that BIN1 colocalizes with, and is within molecular distance of, p140Cap in cultured cells and in the mouse brain. Coimmunoprecipitation assays validated the interaction and glutathione S-transferase pulldown revealed that a rare BIN1 coding variant (rs138047593) significantly reduces p140Cap and tau binding, highlighting the impact of this mutant on interacting protein binding efficiency. The functional implications of BIN1:p140Cap interaction for neuronal functions warrant further investigation.

Keywords: Alzheimer disease; BIN1; SRCIN1; Src homology 3 domain; p140Cap; protein–protein interaction; rs138047593; surface plasmon resonance; tau.

Figure 1

SPR analysis of BIN1’s SH3 domain interactions with p140Cap and tau.A, schematic representation of the criteria used to select putative neuronal BIN1-SH3 domain–interacting proteins. B, schematic of the SPR experimental design using the Cytiva Biotin Capture chip. C, equilibrium analysis of the binding of 14-mer p140Cap_1033-1046 to biotinylated BIN1-SH3. D, schematic of full-length p140Cap and each 14-mer peptide assessed by SPR. Class I motifs are shown in bold and class II motifs are shown underlined. E-F, equilibrium analysis of the binding of peptides 14-mer p140Cap_498-512, and p140Cap_583-596, respectively, to biotinylated BIN1-SH3 domain. G, sequences of the two peptides from the proline-rich domain of tau assessed by SPR. H and I, equilibrium analysis of the binding of 6-mer tau_158-163 and 13-mer tau_154-166 peptides to biotinylated BIN1-SH3 domain, with class II motifs underlined. Shown are representative plots averaged from three technical replicates, and the data are represented as mean ± SD. BIN1, bridging integrator 1; SH3, Src homology 3; SPR, surface plasmon resonance.

Figure 2

SPR analysis of the BIN1–SH3 interaction with class II PRMs on p140Cap.A, schematic representation of the C-terminal proline-rich domain of p140Cap and the three canonical PRMs found within the 14-mer peptide sequence, where the class I motif is shown in bold and class II motifs are shown underlined. B, sequences of the 7-mer native and 14-mer alanine-substituted peptides synthesized for equilibrium assessment by SPR. C, schematic representation of the alanine-scanning mutagenesis experimental design and anticipated outcomes, where substitution of the basic residue alone or in combination with the corresponding proline residues is expected to result in decreased binding strength compared to the native 7-mer peptide. D, equilibrium analysis of the binding of the indicated peptides to biotinylated BIN1-SH3 domain; a-c, native class I, class IIa, and class IIb 7-mer p140Cap peptides, respectively; d-f, the corresponding 14-mer peptides containing alanine-substitution of the basic residue; g-i, the corresponding 14-mer peptides containing alanine-substitution of the basic residue and two corresponding proline residues. Shown are representative plots averaged from three technical replicates, and the data are represented as mean ± SD. E, summary of the mutagenesis/SPR results. BIN1, bridging integrator 1; SH3, Src homology 3; SPR, surface plasmon resonance.

Figure 3

Colocalization of neuronal BIN1 and p140Cap in cultured cells.A, HeLa (top) and N2a (bottom) cells were cotransfected with plasmids encoding neuronal BIN1 TID (magenta) and p140Cap-myc/His (green) and visualized by immunostaining. The colocalization between BIN1 and p140Cap (white) is represented in the merged images. 100X scale represents 50 μm, 100X 2.8 scale represents 20 μm. B, COS-7 cells were cotransfected with neuronal EGFP-BIN1 (pseudocolored magenta) and p140Cap-myc/His (green) and visualized by EGFP fluorescence and antibody staining, respectively, along with F-actin staining by Alexa Fluor 647–labeled phalloidin (cyan). Deconvolved confocal image stacks were used to calculate the degree of overlap in voxels between pairs of channels as Pearson’s colocalization coefficients (n = 20 cells from two transfections; results are plotted as mean ± SD). The bottom gray-scale panels depict the colocalization maps (bottom). One-way ANOVA revealed a significant difference between the three groups [F(2,57) = 467.8, p < 0.0001]. C, transfected cells were immunostained and analyzed as described above to ascertain the overlap between BIN1 (magenta) p140Cap (green) and α-tubulin (cyan). Deconvolved confocal image stacks were used to calculate the degree of overlap in voxels between pairs of channels as Pearson’s colocalization coefficients (n = 20 cells from two transfections; results are plotted as mean ± SD). One-way ANOVA indicated a significant difference between the three groups [F(2,57) = 254.3, p < 0.0001]. Post hoc Tukey’s analysis was performed. ∗∗∗∗p < 0.0001. BIN1, bridging integrator 1; SH3, Src homology 3.

Figure 4

PLA analysis of neuronal BIN1 and p140Cap in the mouse brain.A, representative images of PLA in the cortex (CTX) and hippocampal CA1, CA3, and dentate gyrus (DG) regions of the mouse brain. Images were acquired using a 20X objective (the scale bar represents 200 μm). The boxed regions are shown as a higher magnification image acquired using a 60X objective (the scale bar represents 50 μm). The PLA signal intensities from biological replicates of BIN1-cKO (n = 4), nTg (n = 3), and BIN1-OE (n = 4) animals were quantified, and the results are plotted as mean ± SD (right). The statistical significance was determined using ANOVA: CTX [F(2,8) = 54.39; p < 0.0001; CA1 [F(2, 8) = 79.76, p < 0.0001]; CA3 [F(2, 8) = 109.2, p < 0.0001]; DG [F(2, 8) = 46.00, p < 0.0001]. Post hoc Tukey’s analysis was performed. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001; and ∗∗∗∗p < 0.0001. BIN1, bridging integrator 1; BIN1-OE, BIN1 overexpression; cKO, conditional KO; nTg, nontransgenic; PLA, proximity ligation assay; SH3, Src homology 3.

Figure 5

Quantitative analysis of Bin1 and p140Cap expression in the brains of mice expressing varying levels of BIN1.A, immunoblot analysis of forebrain [hippocampus (HPC) and cortex (CTX)] or midbrain protein extracts from nTg, BIN1-cKO, and BIN1-OE animals (n = 3 mice per genotype). The blots were probed with antibodies against BIN1, p140Cap, PSD95, CaMKII, and synaptophysin. All bands observed in the blots represent distinct BIN1 isoforms [neuronal and ubiquitous isoforms (25, 52)], while p140Cap consistently appears as a doublet. ImageStudio 6.0 was used to quantify BIN1 and p140Cap signal intensities, and the results (normalized to nTg) are plotted as mean ± SD. One-way ANOVA found significant differences for BIN1 levels in HPC + CTX [F(2, 6) = 18.78, p = 0.0026] and midbrain [F(2, 6) = 9.345, p = 0.0144]; post hoc Tukey’s analysis was performed. ∗p < 0.05 and ∗∗p < 0.01. B, representative immunofluorescence images of mouse brains immunostained with antibodies against polyclonal BIN1 antibody (magenta) and mAb p140Cap (green). Images were acquired using a 20X objective from the cortex (CTX) and hippocampal CA1, CA3, and dentate gyrus (DG) regions (the scale bar represents 200 μm). The boxed regions are shown as a higher magnification image, acquired using a 60X objective (the scale bar represents 50 μm). BIN1 and p140Cap signal intensities were quantified from BIN1-cKO (n = 4 animals), nTg (n = 3 animals), and BIN1-OE (n = 4 animals), and the results are potted as mean ± SD (right). Two-way ANOVA revealed statistical differences in BIN1 [F(3,31) = 22.31, p < 0.0001] and p140Cap levels [F(2, 31) = 11.76, p < 0.001]. Post hoc Tukey’s analysis found significant differences in BIN1 levels in all regions (∗∗∗∗p < 0.0001) and p140Cap levels in CA3 (∗∗p < 0.01). BIN1, bridging integrator 1; BIN1-OE, BIN1 overexpression; cKO, conditional KO; nTg, nontransgenic; SH3, Src homology 3.

Figure 6

BIN1-SH3 domain-dependent redistribution of p140Cap in HeLa cells.A, schematic representation of EGFP-tagged BIN1 constructs cotransfected with p140Cap-myc/His/His. B, cartoons depicting the anticipated experimental outcomes. C, HeLa cells cotransfected with p140Cap-myc/His and EGFP-tagged BIN1 WT (WT-BIN1), and BIN1 lacking the N-terminal BAR domain (ΔBAR) or the C-terminal SH3 domain (ΔSH3). Cells were immunostained to visualize p140Cap (green) and EGFP-BIN1 fluorescence (pseudocolored in magenta). 100X images scale represents 50 μm, and inset image scale represents 5 μm. The train of yellow arrows in the inset images indicates likely p140Cap localization along the cytoskeleton. Blue arrows indicate overlap in p140Cap and BIN1 localization. BAR, Bin1-Amphiphysin-Rvs; BIN1, bridging integrator 1; EGFP, enhanced green fluorescence protein; SH3, Src homology 3.

Figure 7

GST pull-down analysis of BIN1-SH3 domain interaction with p140Cap and tau.A, GST pull-down assay of BIN1-SH3 binding to p140Cap and tau from mouse brain lysates. Mouse brains were homogenized in Hepes buffer, and 0.5 mg aliquots of 1% Triton X-100 soluble proteins were incubated with 25 μg of GST or GST-BSH3 coupled to glutathione-Sepharose. Bound proteins were analyzed by immunoblotting with antibodies against total tau and p140Cap. B, quantification of the efficiency of GST pulldowns. Immunoblots were analyzed on the LI-COR system, and the signal intensities from two independent experiments were quantified using ImageStudio 6.0. The efficiency of GST pulldown was calculated by comparing signal intensities of p140Cap or tau in the pulldowns with the corresponding signal in the input lane. The results are plotted as mean ± SD, and statistical significance was determined using an unpaired two-tailed t test; ∗∗p = 0.0035. C, Richardson diagram of the rat BIN1-SH3 domain crystal structure (PDB: 1BB9). Acidic residues mutated to their nonacidic counterparts in constructs used in GST pull-down experiments are represented by sticks; blue = MutB, green = MutA, magenta = MutK (K42R GWAS SNP). D, table outlining the mutations introduced into each construct. E, GST pulldowns from mouse brain homogenates using WT, BIN1-SH3, MutK, MutA, and MutB were conducted in triplicate as described above. F, bar chart from three technical replicates displaying the efficiency of GST pulldown of tau and p140Cap from each construct used. The pull-down efficiency was calculated as described above, and the results are plotted as mean ± SD. Statistical significance was analyzed by one-way ANOVA: p140Cap pulldown [F(3, 8) = 20.88, p = 0.0004]; tau pulldown [F(3, 8) = 104.0, p = 0.0001]. Post hoc Dunnett's analysis was performed. ∗∗p < 0.01; ∗∗∗p < 0.001; and ∗∗∗∗p < 0.0001. G, schematic representation of TurboID constructs cotransfected with p140Cap-myc/His. H, coimmunoprecipitation analysis of BIN1 and p140Cap. Subconfluent cultures of N2a-p140Cap cells grown on 60 mm dishes were transfected with plasmids encoding V5-tagged BIN1 TID or V5-tagged Cyto TID. The cells were lysed in 400 μl of co-IP buffer supplemented with 1% IGEPAL and incubated with 3 μl of mouse anti-V5 mAb. The resulting immunoprecipitates and input lysates, corresponding to 2.5% of the volume used for immunoprecipitation, were probed with antibodies against BIN1 and p140Cap. BIN1, bridging integrator 1; GST, glutathione S-transferase; GWAS, genome-wide association studies; SH3, Src homology 3.