Article Figures & Data
Figures
- Figure 1.
Generation and validation of AIDA-1 cKO mice. A, i, AIDA-1 domain schematic. SAM, Sterile α motif. A, ii, Targeting vector and cKO strategy: PTB domain flanked by loxP sites. AIDA-1 floxed mice (AIDA-1fl/fl) were crossed with CaMKIIαCre mice to generate forebrain-specific AIDA-1 cKO mice. B, PCRs using mouse-tail DNA were used to test for WT and floxed (fl) alleles and Cre recombinase. WT littermates expressing Cre recombinase (wt/wt; CaMKIIαCre) were used as controls for all experiments. Heterozygote mice were identified as AIDA-1fl/wt;CaMKIIαCre and cKO mice were identified as AIDA-1fl/fl;CaMKIIαCre. C, Western blot showing the progressive reduction of AIDA-1 expression in the hippocampus at different developmental stages from P7 to P35 and GAPDH (glyceraldehyde-3-phosphate dehydrogenase) control. D, Western blot showing expression of AIDA-1 splice variants (57 and 50 kDa) and GAPDH in hippocampus of AIDA-1 cKO mice at P35–P42. Graph shows quantification of Western blots [WT: [57 kDa], 100.0 ± 3.8%; [50 kDa], 100.0 ± 4.0%; heterozygote (HET): [57 kDa], 42.1 ± 4.0%; [50 kDa], 58.7 ± 5.8%; cKO: [57 kDa], 12.2 ± 3.1%; [50 kDa], 17.8 ± 5.8%; n = 4–9 mice], ***p < 0.0001. Summary data represent mean ± SEM.
- Figure 2.
Excitatory transmission at hippocampal Sch–CA1 synapses in AIDA-1 cKO mice. A, B, Loss of AIDA-1 did not significantly alter (A) input–output function or (B) paired-pulse ratio measured at 10, 30, 100, and 300 ms interstimulus intervals (ISI). C, Representative AMPAR-mEPSC traces (left) and summary plots (right) showing no change in AMPAR-mEPSC amplitude or frequency in WT and AIDA-1 cKO neurons. In all panels, summary data represent the mean ± SEM. Number of cells (c), slices (s), and animals (a) are indicated in parentheses.
- Figure 3.
Synaptic NMDAR transmission in AIDA-1 cKO mice. A, Representative averaged EPSCs (left) and summary data (right) showing that the NMDAR/AMPAR ratio was similar between WT and AIDA-1 cKO neurons (WT, 0.71 ± 0.05; cKO, 0.68 ± 0.07, p = 0.71). B, NMDAR-EPSCs recorded at different holding potentials in WT and AIDA-1 cKO neurons; representative averaged traces (left) and I–V plot (right). C, Normalized NMDAR-EPSCs (left) and summary data (right) showing faster NMDAR decay kinetics in AIDA-1 cKO compared with WT mice (WT, 81.2 ± 2.4 ms; cKO, 68.2 ± 1.5 ms). Summary data represent the mean ± SEM and the number of cells (c), and animals (a) are indicated in parentheses. ***p < 0.001. n.s., Not significant.
- Figure 4.
Altered GluN2A-mediated and GluN2B-mediated synaptic transmission and synaptic expression in AIDA-1 cKO mice. A, Averaged NMDAR-EPSCs (top) and summary data (bottom) showing less sensitivity to the selective GluN2B antagonist Ro25-6981 (500 nm) in AIDA-1 cKO mice compared with WT mice (WT, 62.2 ± 2.4% of baseline; cKO, 75.9 ± 2.3% of baseline). B, Averaged sample traces (top) and summary data (bottom) showing that NMDAR-EPSCs in AIDA-1 cKO mice are more sensitive to the selective GluN2A antagonist Zn2+ (200 nm) compared with WT mice (WT, 69.2 ± 2.5% of baseline; cKO, 48.7 ± 2.7% of baseline). In all panels, averaged sample traces were taken at times indicated by numbers on the summary plot. The number of cells (c), slices (s), and animals (a) are indicated in parentheses. C, Western blots and quantification showing unchanged expression of different synaptic proteins in whole hippocampal lysates from WT and AIDA-1 cKO mice (n = 4–10 mice; AIDA-1, 19.6 ± 4.5% of WT, n = 8 mice). D, Western blots and quantification (n = 3–10 mice) showing a decrease in GluN2B (GluN2B, 70.4 ± 12.3% of WT, n = 8), an increase in GluN2A (GluN2A, 153.0 ± 29.7% of WT, n = 8 mice) and decrease in PSD95 (PSD95, 83.19 ± 5.8% of WT, n = 4) protein abundance in hippocampal PSD fractions from AIDA-1 cKO mice. *p < 0.05, ***p < 0.001. Summary data represent the mean ± SEM. All experiments were performed using mice aged P35–P48.
- Figure 5.
shRNA-mediated knockdown of AIDA-1 alters synaptic NMDAR subunit composition in rat primary neurons. A, Left, Western blot showing reduction of AIDA-1 protein levels in cultured rat cortical neurons by two different AIDA-1-targeting shRNAs (shAIDA#1 and shAIDA#2) compared with a nontargeting control (shNT). Neurons were transduced with lentiviruses expressing shRNAs at 8 DIV for 7–12 d. Right, Quantification of shRNA-induced knockdown of AIDA-1 splice variants (57 and 50 kDa) and normalized to control (shNT): (shAIDA#1: [57 kDa], 12.2 ± 5.2%; [50 kDa], 29.7 ± 10.2%; shAIDA#2: [57 kDa], 28.1 ± 6.0%; [50 kDa], 27.5 ± 7.1%; n = 6–8 samples from 3 different cultures). Statistical significance was determined using one-way ANOVA and post hoc t test (2-tailed, unpaired) with Tukey's correction for multiple comparisons. B, Quantification for the mean intensity of GluN2B puncta (shAIDA#1, 81.6 ± 1.0% of shNT; shAIDA#2, 62.7 ± 1.0% of shNT), or GluN2A puncta (shAIDA#1, 156.4 ± 1.7% of shNT; shAIDA#2, 152.1 ± 1.5% of shNT) that colocalized with the synaptic marker Shank (n = 6–11 cells from 3 independent cultures). Neurons were transduced on DIV 7 and imaged on DIV 17–22. C, Representative images showing that AIDA-1 knockdown reduces synaptic levels of GluN2B and increases synaptic levels of GluN2A in DIV 21 rat primary hippocampal cultures. GluN2 subunit is red, shank is green, and MAP2 is blue in the merged image. D, Representative normalized NMDAR-EPSCs (top) and summary data (bottom) showing that shRNA-mediated knockdown of AIDA-1 accelerates the decay kinetics of NMDAR-EPSCs (shNT, 201.0 ± 9.4 ms; shAIDA#2, 163.5 ± 6.9 ms, n = 24 cells from 6 independent cultures). E, Averaged sample traces (top) and summary data (bottom) showing that shRNA-mediated knockdown of AIDA-1 decreased NMDAR-EPSC sensitivity to the GluN2B-specific antagonist Ro25-6981 (200 nm; shNT, 49.3 ± 3.2%; shAIDA#2, 68.5 ± 4.0%; n = 24 cells from 6 independent cultures). Neurons were transduced at DIV 7 and recorded at DIV 12–21. Summary graphs represent the mean ± SEM. **p < 0.01, ***p < 0.001.
- Figure 6.
Impaired NMDAR-dependent plasticity at Sch–CA1 synapses in AIDA-1 cKO mice. A, Average traces (left) and summary plot (right) showing that LTP induced by HFS was impaired in AIDA-1 cKO compared with WT mice (last 10 min of HFS-LTP: WT, 173.5 ± 11.2% of baseline; cKO, 128 ± 4.8% of baseline). B, Similar results were observed by inducing LTP with TBS (last 10 min of TBS-LTP: WT, 153.7 ± 6.2% of baseline; cKO, 128.8 ± 7.2% of baseline). C, Average traces (left) and summary plot (right) showing that NMDAR-LTD induced by low-frequency stimulation (LFS) was absent in AIDA-1 cKO compared with WT mice (LFS; last 10 min: WT, 74.8 ± 3.4% of baseline; cKO, 97.2 ± 3.1% of baseline). D, mGluR-LTD induced by application of group I mGluR agonist (S)-3,5-dihydroxyphenylglycine (DHPG) is normal in AIDA-1 cKO mice (mGluR-LTD; last 10 min: WT, 83.6 ± 5.9% of baseline; cKO, 75.7 ± 6.6% of baseline, p = 0.753). In all panels, averaged sample traces were taken at times indicated by numbers on the summary plot. Summary data represent the mean ± SEM and the number of slices (s) and animals (a) are indicated in parentheses; ***p < 0.001, **p = 0.012. Experiments were performed using mice at P35–P48.
- Figure 7.
AIDA-1 preferentially associates with GluN2B-containing NMDARs. A, Coimmunoprecipitations from mouse whole hippocampal lysates prepared at P42 show that AIDA-1 interacts with GluN1, with PSD95, and preferentially with GluN2B over GluN2A. The extent of GluN2B or GluN2A binding to AIDA-1 was determined by normalizing the immunoprecipitation (IP) signal by the input (n = 5 independent coimmunoprecipitations; input, 15 μg; GluN2B, 69.5 ± 3.3%; GluN2A, 10.5 ± 1.2%). B, Primary rat hippocampal neurons at DIV 21 were immunostained for the dendritic marker MAP2, synaptic marker Shank, AIDA-1, and GluN2A or GluN2B. C, The extent of colocalization of AIDA-1 with GluN2A or GluN2B was quantified by Mander's coefficients, and shows that AIDA-1 colocalizes preferentially with GluN2B (n = 6–8 cells from 4 independent cultures; GluN2A-AIDA, 10.5 ± 1.2%; GluN2B-AIDA, 50.6 ± 1.8%). D, Coexpression of GluN2B (eGFP-GluN2B) or GluN2A (eGFP-GluN2A) with AIDA-1 (57 kDa isoform; mKATE2, a red fluorescent protein) in rat primary hippocampal neurons at DIV 14 for 3 d shows extensive colocalization of AIDA-1 and GluN2B in dendrites, and much less with GluN2A. E, Representative Western blot showing AIDA-1 isoform and GluN2B do not coimmunoprecipitate when coexpressed in HEK293FT cells. Data presented as mean ± SEM. ***p < 0.001.
- Figure 8.
AIDA-1 associates with the GluN2B transport machinery and regulates GluN2B export from the ER. A, Coimmunoprecipitations from whole mouse hippocampus showing that AIDA-1 associates with components of the GluN2B transport machinery KIF17, CASK, and SAP97, but not KIF1b. B, mKATE2-tagged AIDA-1 was coexpressed with either YFP-tagged KIF17 or CASK in HEK cells. Coimmunoprecipitations show that AIDA-1 can interact with both KIF17 and CASK. C, Loss of AIDA-1 does not alter the total levels of KIF17 and CASK in hippocampal lysates (n = 4 mice). Summary graphs represent the mean ± SEM. D, Western blots for GluN2B, calnexin (ER marker), and 58K Golgi protein (Golgi marker) showing different biochemical fractions (top) from hippocampal tissue of WT and AIDA-1 cKO mice. H represents whole lysate. Below, A plot of the relative abundance of GluN2B in consecutive iodixanol gradient fractions normalized to H (bottom left) and quantification of GluN2B in pooled ER fractions (bottom right). Bar graph shows quantification (n = 3 mice) of total GluN2B in ER fractions (area under the curve; cKO, 132.1 ± 1.5% of WT, p = 0.024). Results show GluN2B accumulates in the ER fractions in AIDA-1 cKO mice. Summary graphs represent the mean ± SEM. *p < 0.05.
