Smaller dendritic spines, weaker synaptic transmission, but enhanced spatial learning in mice lacking Shank1

Abstract

Experience-dependent changes in the structure of dendritic spines may contribute to learning and memory. Encoded by three genes, the Shank family of postsynaptic scaffold proteins are abundant and enriched in the postsynaptic density (PSD) of central excitatory synapses. When expressed in cultured hippocampal neurons, Shank promotes the maturation and enlargement of dendritic spines. Recently, Shank3 has been genetically implicated in human autism, suggesting an important role for Shank proteins in normal cognitive development. Here, we report the phenotype of Shank1 knock-out mice. Shank1 mutants showed altered PSD protein composition; reduced size of dendritic spines; smaller, thinner PSDs; and weaker basal synaptic transmission. Standard measures of synaptic plasticity were normal. Behaviorally, they had increased anxiety-related behavior and impaired contextual fear memory. Remarkably, Shank1-deficient mice displayed enhanced performance in a spatial learning task; however, their long-term memory retention in this task was impaired. These results affirm the importance of Shank1 for synapse structure and function in vivo, and they highlight a differential role for Shank1 in specific cognitive processes, a feature that may be relevant to human autism spectrum disorders.

Figure 1.

Generation and characterization of Shank1 mutant mice. A, Domain structure of Shank1, showing ankyrin repeats (Ank), SH3 and PDZ domains, proline-rich region (Pro), and sterile α motif (SAM) domain. B, Schematic diagram of the Shank1 gene locus, the targeting vector, and the mutant allele after homologous recombination. The neomycin resistance cassette (NEO) replaces the exons coding the PDZ domain (white box) and their intervening introns. BamHI and EcoRV probes used for Southern analysis and PCR primers are shown. Restriction sites: B, BamHI; R, EcoRI; RV, EcoRV. C, Southern blot analysis of wild-type (+/+) and two independent clones of targeted ES cells (362 and 388). D, PCR genotype analysis of a representative litter from heterozygote intercross. The upper 517 bp band corresponds to amplified product of wild-type allele, and 282 bp band corresponds to the PCR product from the neo cassette. E, Immunoblot of forebrain membrane fractions from wild-type (+/+), heterozygous (+/−), and homozygous Shank1^−/− mice, probed with Shank1-specific antibody 1356 (left) or pan-Shank antibody 3856 (right). F, Cresyl violet staining of coronal sections of hippocampus from wild-type and Shank1^−/− mice.

Figure 2.

Altered composition of PSD fractions in Shank1 mutant mice. Analysis of Triton-extracted PSD fractions (PSDI) purified from forebrain of adult wild-type (+/+) and Shank1 knock-out (−/−) mice. A, SDS-PAGE and silver staining of PSD proteins revealed no major differences. B, Immunoblot analysis of PSD fractions from individual wild-type and Shank1^−/− mice for the indicated proteins. C, Quantitation of various proteins in PSD fractions based on results as shown in B, normalized to wild-type (100%; black histograms). Histograms show mean ± SEM (>6 mice from each genotype). There is a significant difference in total Shank (+/+, n = 6; −/−, n = 8; ***p < 0.0001), GKAP (+/+, −/−, n = 12 each; *p < 0.05), and Homer (+/+, n = 12; −/−, n = 14; **p < 0.002).

Figure 3.

Altered immunostaining of PSD proteins in Shank1-deficient neurons. A, B, Hippocampal neurons in dissociated culture (18–19 DIV) from wild-type (+/+) or Shank1^−/− mice were immunostained for the indicated proteins. A, Representative dendrites double labeled for PSD-95 (green) and with Shank1-specific antibody (1356) or pan-Shank antibody (3856) (red). Merged image is shown in color on the right. Scale bar, 5 μm. B, Dendrites immunostained for GKAP, PSD-95, Homer, or presynaptic marker Bassoon. C, Quantification of puncta density per 10 μm dendrite length for the indicated proteins (mean ± SEM). Shank1^−/− dendrites show a significant reduction in pan-Shank (n = 10 cells each; ***p < 0.001) and GKAP puncta density (+/+, n = 21 cells; −/−, n = 16 cells; **p < 0.01).

Figure 4.

Smaller dendritic spines and thinner PSDs in Shank1 mutant mice. A, DiI-labeled dendrites from CA1 pyramidal neurons of adult wild-type (+/+) and Shank1^−/− mice (three representative segments shown). Scale bar, 5 μm. B, Quantification of dendritic spine density in wild-type versus Shank1^−/− neurons. Data are presented as scattergrams (with mean ± SEM superimposed), each point corresponds to the mean spine density for a single neuron (+/+, n = 33 neurons from 4 animals; −/−, n = 27 neurons from 4 animals). Shank1^−/− mice show a slight reduction in spine density (p < 0.05; t test). A total of 9210 wild-type spines and 7797 mutant spines were measured. C, D, Cumulative frequency plots for spine length (C) and spine head width (D) from wild-type (black line) and Shank1^−/− mice (gray line). Scattergrams, with mean ± SEM, are shown at the right. Shank1^−/− mice showed a small but highly significant shift toward smaller spines (length, p < 0.0005, K–S; width, p < 0.001, K–S). E, Representative electron micrographs of hippocampal CA1 striatum radiatum synapses from wild-type and Shank1^−/− mice. The PSD is visible as an electron-dense layer adjacent to the postsynaptic membrane. Note thinner PSD in Shank1^−/− synapses. Scale bar, 100 nm. F, G, Cumulative frequency distribution of PSD length (F) and thickness (G) from wild-type (black line) and Shank1^−/− mice (gray line). Sixty randomly selected synapses from each of four wild-type and six Shank1^−/− mice were measured by blind observers.

Figure 5.

Decreased synaptic strength in Shank1 mutant mice. A, Left, Sample traces (average of 10 consecutive responses) represent the responses evoked with seven different stimulus intensities from wild-type (+/+) or Shank1^−/− hippocampal slices. The same sample traces are shown at two different scales. Stimulus artifacts were truncated. Right, Summary graph of the input–output relationships of field EPSPs of wild-type mice (+/+; n = 15 slices from 9 mice) and Shank1^−/− mice (−/−; n = 12 slices from 7 mice). Symbols indicate the mean ± SEM. The input–output relationship of Shank1^−/− is significantly weaker than that of wild-type mice (*p < 0.05; Student's t test). B, Left, Top and Middle, Two consecutive sample mEPSC traces from wild-type and Shank1^−/− mice. Left, Bottom, Averaged mEPSC from wild-type mice (left) and Shank1^−/− mice (center) (average of 300 traces) and superimposed traces (sup; right). Note that the time course of the events is the same. Right, Summary graphs of the amplitude and frequency of mEPSCs in wild-type (n = 12 cells/6 mice) and Shank1^−/− (n = 13 cells/6 mice) animals. There was no significant difference in mEPSC amplitude (NS, not significant). The frequency of mEPSCs in Shank1^−/− mutants is significantly less than that of wild-type mice (*p < 0.02; Student's t test).

Figure 6.

Synaptic plasticity is unchanged in Shank1 mutant mice. A, Top, Sample traces of field EPSPs of wild-type (+/+) and Shank1^−/− mice (−/−) recorded at the times indicated in summary graph. Below, Summary graph of the averaged time course of LTP (+/+, n = 12 slices/10 mice; −/−, n = 12 slices/12 mice). Initial EPSP slopes were measured, and the values were normalized to the averaged slope value measured during the baseline period (−30 to 0 min). Tetanic stimulation (100 Hz, 1 s) was applied at 0 min. B, Sample traces and summary graph of the averaged time course of LTD (+/+, n = 9 slices of 9 mice; −/−, 7 slices of 7 mice). Low-frequency stimulation (1 Hz, 15 min) was applied at 0 min. C, Sample traces and summary graph of the averaged time course of late-phase LTP (+/+, n = 9 slices of 9 mice; −/−, n = 8 slices of 8 mice). Four trains of tetanic stimulation (100 Hz, 1 s) were applied from 0 min at 5 min intervals. There was no statistically significant difference between wild-type and Shank1^−/− mice.

Figure 7.

Hypoactivity, rotarod performance, and increased anxiety-related behaviors in Shank1 mutant mice. A, Horizontal activity in open-field test is shown as number of beam breaks (mean ± SEM). Black line, Wild type (n = 24); gray line, Shank1^−/− (n = 24). Shank1^−/− mice show significantly reduced activity (genotype effect, F_(1,46) = 21.60; p < 0.0001, repeated measures two-way ANOVA). B, Total distance traveled and movement time during 30 min open-field session is significantly reduced in Shank mutants (****p < 0.0001; t test). C, Mice (n = 15 per genotype) were tested for their ability to stay on the accelerating Rotarod (measured as latency in seconds to falling off). Three trials were performed on two consecutive days, with each trial terminated after 300 s. Repeated measures ANOVA revealed a significant genotype effect (F_(1,28) = 21.19; p < 0.0001). D, Shank1^−/− mutants make significantly fewer light–dark transitions and have longer latencies to enter the light compartment in the light–dark transition assay (***p = 0.001; *p = 0.05; n = 7 per genotype).

Figure 8.

Impaired contextual fear memory in Shank1 mutant mice. A, Wild-type mice (black line; n = 20) and Shank1^−/− mice (gray line; n = 18) showed similar freezing responses during the conditioning phase. The horizontal bar denotes exposure to conditioned stimulus (tone), and the vertical arrow notes timing of unconditioned stimulus (footshock). B, Decreased freezing in Shank1^−/− mice at 1 and 24 h after fear conditioning when exposed to the same context (**p < 0.01; ***p < 0.0001). C, Cued fear memory was tested in an altered context 48 h after conditioning. The mean percentage of time spent freezing before tone and during presentation of the tone is shown. There is no significant difference in freezing in response to the conditioned stimulus.

Figure 9.

Enhanced acquisition and impaired retention of spatial memory by Shank1 mutants in the eight-arm radial maze task. Two of eight arms were baited to test simultaneously reference and working memory. A, Total number of reference memory errors during acquisition training (+/+, black line, n = 15; −/−, gray line, n = 14). Mice received a total of 84 trials; data are presented in blocks of four trials. B, Total number of working memory errors (revisiting errors) across training. Shank1^−/− mice make significantly fewer reference memory errors (genotype effect, F_(1,27) = 10.98; p < 0.003, repeated measures two-way ANOVA) and working memory errors (genotype effect, F_(1,27) = 10.41; p < 0.004) than wild-type mice. C, Shank1^−/− mice correctly select a baited arm with their first arm selection more frequently than wild type (genotype effect, F_(1,27) = 49.72; p < 0.0001). Data are shown in blocks of eight trials. D, Latency (time in seconds) to complete the trials in the eight-arm radial maze task is presented as means of four trials. E, Impaired spatial memory retention in Shank1^−/− mice. Mice were retested in the radial maze task 28 d after completion of initial training (white bars); data are the mean of four trials. “End-training” (black bars) shows the performance during the last four-trial block of the acquisition phase shown in A. There is no significant difference in the performance of wild-type mice, but mutants perform significantly worse after 28 d without exposure to the maze. *p < 0.02; NS, p > 0.05, t test. F, Reversal training shows enhanced behavioral plasticity in Shank1^−/− mutants. Position of baits was changed, and mice were retrained to learn new positions of baits. There was no difference in initial reversal trials (trials 1–16, genotype effect, F_(1,27) = 0.30; p = 0.59), but Shank1^−/− mice made fewer reference memory errors compared with wild type over last 20 trials (genotype effect, F_(1,27) = 7.81; p < 0.01). G, Graph comparing standard (black) and intensive training protocols (gray) in the eight-arm radial maze task. For the intensive training, wild-type and Shank1^−/− mice received a total of 60 trials over 22 training days. H, Despite intensive training, wild-type mice (n = 17) still make more reference memory errors than Shank1^−/− mice (n = 18) (genotype effect, F_(1,33) = 10.86; p < 0.003, repeated measures two-way ANOVA). Data are presented in blocks of four trials. I, Defect in memory retention in Shank1^−/− mutants after intensive training (+/+, −/−, n = 9 mice each). When tested 28 d after completion of intensive training, mutant mice make more reference memory errors compared with last four-trial acquisition block, whereas wild-type performance shows no significant deterioration. “End-training” represents last four-trial block shown in H. **p < 0.005; NS, p > 0.05, t test.