Loss of predominant Shank3 isoforms results in hippocampus-dependent impairments in behavior and synaptic transmission

Abstract

The Shank3 gene encodes a scaffolding protein that anchors multiple elements of the postsynaptic density at the synapse. Previous attempts to delete the Shank3 gene have not resulted in a complete loss of the predominant naturally occurring Shank3 isoforms. We have now characterized a homozygous Shank3 mutation in mice that deletes exon 21, including the Homer binding domain. In the homozygous state, deletion of exon 21 results in loss of the major naturally occurring Shank3 protein bands detected by C-terminal and N-terminal antibodies, allowing us to more definitively examine the role of Shank3 in synaptic function and behavior. This loss of Shank3 leads to an increased localization of mGluR5 to both synaptosome and postsynaptic density-enriched fractions in the hippocampus. These mice exhibit a decrease in NMDA/AMPA excitatory postsynaptic current ratio in area CA1 of the hippocampus, reduced long-term potentiation in area CA1, and deficits in hippocampus-dependent spatial learning and memory. In addition, these mice also exhibit motor-coordination deficits, hypersensitivity to heat, novelty avoidance, altered locomotor response to novelty, and minimal social abnormalities. These data suggest that Shank3 isoforms are required for normal synaptic transmission/plasticity in the hippocampus, as well as hippocampus-dependent spatial learning and memory.

Figure 1.

Loss of major naturally occurring Shank3 proteins in Shank3^ΔC/ΔC mice. A, Quantification and representative Western blots of hippocampal lysates showing loss of the major isoforms of Shank3 with C′ (P. Worley, JH3025), SH3 (Abcam), and N′ (P. Worley) antibodies. B, Quantification and representative Western blots of lower-molecular-weight bands that appear or increase in intensity in the Shank3^ΔC/ΔC hippocampal lysates using SH3 domain and N′ antibodies. C, Larger representative Western blots showing the comparison of bands detected by the three Shank3 antibodies. D, E, Quantification and representative blots of whole hippocampus lysates with antibodies against PSD proteins and receptors that interact directly or indirectly with Shank3. In A, D, and E, data are normalized to β-actin control and then to average of WT levels. In B, data are normalized to β-actin control and then to average of Shank3^ΔC/ΔC. Error bars represent SEM, **p < 0.01, ***p < 0.001.

Figure 2.

Increased mGluR5 protein levels in hippocampal synaptosome and PSDII fractions of Shank3^ΔC/ΔC mice. A, Quantification of PSD proteins in synaptosome fractions of the hippocampus shows a complete loss of the major isoforms of Shank3 using the C terminus, N terminus, and SH3 domain antibodies against Shank3 in Shank3^ΔC/ΔC mice (*p < 0.0001) as well as an increase in mGluR5 (*p < 0.0001) compared with WT (WT, n = 8; ΔC/ΔC, n = 6). Representative blots are shown inset for proteins showing significant differences. B, Quantification of the <100 kDa bands that appear or increase in the hippocampal synaptosomes of Shank3^ΔC/ΔC mice. C, Quantification of PSD proteins in PSDII fractions of the hippocampus shows a complete loss of Shank3 using the C-terminal antibody (*p < 0.05) as well as the N-terminal antibody (*p < 0.001) of Shank3 in Shank3^ΔC/ΔC mice as well as a robust increase in mGluR5 (*p < 0.01) compared with WT (for each group, n = 3 sets of hippocampi pooled from 2 mice each). D, Quantification and representative Western blot of the <100 kDa bands that appear or increase in the hippocampal PSDII fraction of Shank3^ΔC/ΔC mice. For A–D, data were normalized to β-actin levels and then to the average of WT (A, C) or ΔC/ΔC (B, D). Data shown as average ± SEM. Representative blots are shown inset for proteins showing significant differences.

Figure 3.

Shank3^ΔC/ΔC mice exhibit impaired spatial learning. A–D, Training days for the Morris water maze task. For each day of training, data were averaged across four daily trials. A, Latency to reach hidden platform on successive water maze days. Shank3^ΔC/ΔC mice take longer to reach the submerged platform. B, Swim speed on successive water maze training days. The average swim speed was unchanged in Shank3^ΔC/ΔC mice. C, Distance traveled before reaching the hidden platform on successive water maze training days. Shank3^ΔC/ΔC mice travel a more circuitous route (longer distance) before reaching the submerged platform. D, Percentage time spent in thigmotaxis on successive water maze training days. E, Time spent in target quadrant and other quadrants during probe trial in which target platform is removed. Shank3^ΔC/ΔC mice spend more time in the target quadrant versus other quadrants but less time in target quadrant compared with littermate controls. F, Number of target location crossings and corresponding phantom platform location crossings in other quadrants during the probe trial. Shank3^ΔC/ΔC mice fail to show a preference for the target platform location. G–J, Training trials for the Morris water maze reversal task. No differences were observed during training for the Morris water maze reversal learning task in latency to platform (G), mean swim speed (H), distance traveled (I), or thigmotaxis (J). K, On the probe trial for the Morris water maze reversal task, Shank3^ΔC/ΔC mice failed to show preference for the target quadrant and spent equal time in all four quadrants. L, On the probe trial for the Morris water maze reversal task, Shank3^ΔC/ΔC mice did not show preference for the target platform location (n = 18 in all panels, data depicted as average ± SEM, *p < 0.05).

Figure 4.

Shank3^ΔC/ΔC mice exhibit impairments in other behavioral tasks. A, Latency to fall from or to go one full revolution on the rotarod task. Shank3^ΔC/ΔC mice exhibit motor coordination impairments in eight trials of rotarod test conducted over 2 d (n = 19). Legend in A applies to C and I. B, Latency to lick hindpaw on the hotplate task. Shank3^ΔC/ΔC mice show hypersensitivity to heat on a hotplate (n = 18). Legend in B applies to D–H and J–L. C, Width of nest built as a function of time in a nest-building task. Shank3^ΔC/ΔC mice exhibit impairments in nest-building behavior over a 90 min period (n = 19). D, Number of marbles buried during a 30 min marble-burying task. Shank3^ΔC/ΔC mice show impaired marble burying behavior (n = 19). E, Time spent in dark and light chambers during dark/light task. Shank3^ΔC/ΔC mice spend more time in the dark than littermate controls (n = 19). F, Latency to enter the light chamber in the dark/light task. Shank3^ΔC/ΔC mice exhibit dramatically increased latency to enter the light side (n = 19). G, Fraction of time in the open arms versus time in other arms in the elevated plus maze task. Shank3^ΔC/ΔC mice spend the same time in open versus closed arms when compared with littermate controls (n = 18). H, Ratio of time spent in the center to time spent in the periphery in an open-field task. Shank3^ΔC/ΔC mice behave the same as littermate controls (n = 19). I, Locomotor activity as measured by number of photobeam breaks during successive 5 min intervals over a 2 h period. Shank3^ΔC/ΔC mice exhibit normal locomotor habituation over the full 2 h period (n = 19). J, Number of photobeam breaks during the initial 5 min of the locomotor task shown in I. Shank3^ΔC/ΔC mice show decreased activity, initially suggesting abnormal locomotor response to novelty (n = 19). K, Total distance traveled during the 10 min open-field task. Shank3^ΔC/ΔC mice have decreased locomotor activity in the open field (n = 19). L, Number of photobeam breaks during the 10 min dark/light task. Shank3^ΔC/ΔC mice have decreased locomotor activity in dark/light (n = 19). Error bars represent SEM, *p < 0.05.

Figure 5.

Shank3^ΔC/ΔC mice exhibit minimal social interaction deficits and normal startle reactivity and PPI. A, Time spent in chambers with empty cages. For the first trial of three-chambered social interaction test, Shank3^ΔC/ΔC mice were allowed to explore a three-chambered apparatus and showed no initial preference for either end of the box (n = 17). Legend in A applies to B–D and F. B, In the second trial when given a choice between social or inanimate target, both WT and Shank3^ΔC/ΔC mice show a preference for a caged social target versus inanimate object. However, Shank3^ΔC/ΔC mice avoided the inanimate object and spend less time sniffing it than the WT group (n = 17). C, In the third trial, when given a choice between novel social target versus a familiar social target, Shank3^ΔC/ΔC mice failed to show a preference for the novel social target, unlike their WT littermate pairs (n = 17). D, Shank3^ΔC/ΔC mice show normal social interaction with a juvenile conspecific mouse and, when presented with the same mouse 3 d later, exhibit normal social memory (n = 18). Shank3^ΔC/ΔC mice exhibit normal response to startle (E) and show no deficits in PPI (E; n = 18). Shank3^ΔC/ΔC mice show no change in total time spent in repetitive grooming behavior (G) or in time spent grooming per bout (H; n = 19). However, when tested at an older age, Shank3^ΔC/ΔC mice show a significant increase in overall time spent grooming (I) and time spent grooming per bout (J; n = 16). Error bars represent SEM, *p < 0.05.

Figure 6.

Synaptic plasticity at hippocampal CA3–CA1 synapses is altered in Shank3^ΔC/ΔC mice. A, LTP is decreased in Shank3^ΔC/ΔC mice (n = 6) compared with WT controls (n = 8). Arrow indicates onset of 100 Hz train for 1 s. Inset, Average of 15 consecutive traces immediately before (black) and 60 min after (gray) 100 Hz tetanus. Calibration: 0.3 mV (WT) or 0.55 mV (Shank3^ΔC/ΔC), 5 ms. Legends in A also apply to B and C. B, mGluR–LTD from 6- to 8-week-old mice is not significantly affected by exon 21 deletion (WT, n = 9; Shank3^ΔC/ΔC, n = 7). Bar indicates 5 min bath application of DHPG. Inset, Average of 15 consecutive traces immediately before DHPG wash-in (black) and 60 min after the start of DHPG washout (gray). Calibration: 0.3 mV, 5 ms. C, There no significant difference in mGluR–LTD from 3- to 4-week-old Shank3^ΔC/ΔC mice (n = 11) compared with WT (n = 9). Bar indicates 10 min bath application of DHPG. Inset, Average of 15 consecutive traces immediately before DHPG wash-in (black) and 60 min after the start of DHPG washout (gray). Calibration: 0.3 mV (WT) or 0.22 mV (Shank3^ΔC/ΔC), 5 ms. Error bars represent SEM, *p < 0.05.

Figure 7.

Synaptic transmission is altered at hippocampal CA3–CA1 synapses in Shank3^ΔC/ΔC mice. A, The NMDA/AMPA ratio is decreased in Shank3^ΔC/ΔC mice (n = 25) compared with WT (n = 22). Fifteen consecutive traces (gray) and average trace (black) at −70 mV (bottom) and at +40 mV (top) from WT mice (left) and Shank3^ΔC/ΔC mice (right). Legend in A applies to C and D. Cumulative frequency of mEPSC amplitude (B) and mean mEPSC amplitude (C) were unchanged, but mEPSC frequency (D) was significantly decreased in Shank3^ΔC/ΔC compared with WT (WT, n = 15; Shank3^ΔC/ΔC, n = 22). Inset, One minute raw traces from a WT CA1 neuron (black) and a Shank3^ΔC/ΔC CA1 neuron (gray). Calibration: 10 pA, 2.5 s. Legend in B also applies to E and F. E, PPR is not different between WT and Shank3^ΔC/ΔC mice at interstimulus intervals of 30–500 ms. n = 10 for each genotype. F, The relationship of stimulus intensity to fEPSP slope is decreased in Shank3^ΔC/ΔC mice. Inset, Relationship of fiber volley to fEPSP slope is similar between WT and Shank3^ΔC/ΔC mice. n = 10 for each genotype. Error bars represent SEM, *p < 0.05.

Figure 8.

No morphological deficits were observed in the CA1 hippocampal neurons in Shank3^ΔC/ΔC mice. A, No differences between genotypes were observed in quantitative assessment of branching via Sholl analysis. Legend in A also applies to C. B, Representative examples of WT and Shank3^ΔC/ΔC spine density at 90 μm from the soma at 100× magnification. Scale bar, 5 μm in WT (also applies to Shank3^ΔC/ΔC). C, No differences between genotypes were observed in spine density in the apical dendrites of CA1 hippocampus pyramidal neurons. n = 20 neurons from 5 mice for each genotype. Error bars represent SEM, *p < 0.05.