Early Postnatal Shank3 Downregulation in the Nucleus Accumbens Impairs Performance in Social Conditioning Paradigms in Male Mice

Abstract

Autism spectrum disorder (ASD) is a heterogeneous neurodevelopmental disorder characterized by reduced social interactions, as well as repetitive behaviors and restricted interests. Mutations in SHANK3, a scaffolding protein located postsynaptically at excitatory synapses, are associated with ASD, schizophrenia, and intellectual disability in humans. Similar autism-like phenotypes have been observed in Shank3-deficient rodent models. The mesolimbic dopamine pathway appears to be particularly sensitive to Shank3 disruptions. We have previously shown that Shank3 downregulation in the nucleus accumbens (NAc) (Shank3-NAcKD) during early postnatal development impaired social preference in the three-chamber test. Here, we aimed to assess whether this Shank3 downregulation would lead to deficits in social conditioning paradigms. Specifically, using the social instrumental task (SIT), we found that Shank3-NAcKD male mice performed fewer lever presses to gain access to social interaction with a nonfamiliar juvenile mouse. Moreover, these mice failed to develop a preference for the chamber associated with social stimuli in a conditioned place preference (CPP) task. Unsupervised analysis of locomotor motifs during CPP revealed distinct exploratory strategies, with an altered allocation of exploratory behaviors between the socially paired and unpaired chambers, suggesting a suboptimal direction of exploration towards relevant social-associated cues. Our current data expand on our previous research to understand the involvement of mesolimbic Shank3 expression in autism-like phenotypes. Additionally, our results underline that local Shank3 manipulation during early postnatal life leads to intricate social behavior deficits, highlighting the need for an in-depth dissection of behavioral phenotypes in rodent models of ASD.

Keywords: Shank3; autism spectrum disorder; behavioral classification; nucleus accumbens; social conditioned place preference; social instrumental task.

FIGURE 1

Postnatal downregulation of Shank3 in the NAc does not impair free social interaction. (a) Schema and representative image of injection site in the NAc with AAV‐scrShank3‐GFP or AAV‐shShank3‐luczsGreen in P6 mice. Mice were then tested in a free social interaction test during adulthood. Bar graphs for scr‐ (light blue, n = 7) and shShank3 (dark blue, n = 5) mice reporting the percentage of time. (b) Sniffing the sex‐matched juvenile conspecific (Mann–Whitney U test: p = 0.268). (c) Number of episodes sniffing the sex‐matched juvenile conspecific (unpaired t‐test: t ₍₁₀₎ = 0.832, p = 0.425). (d) Self‐grooming (unpaired t‐test: t ₍₁₀₎ = 0.909, p = 0.395), (e) rearing (unpaired t‐test: t ₍₁₀₎ = 0.378, p = 0.713), or (f) digging (unpaired t‐test: t ₍₁₀₎ = 1.626, p = 0.135) during the free social interaction test. All data are shown as the mean ± SE.

FIGURE 2

Deficits of Shank3 expression in the NAc result in impaired learning during a social instrumental task (SIT). (a) Top: schema of injection sites in the NAc with AAV‐scrShank3‐GFP or AAV‐shShank3‐luczsGreen in P6 mice. Mice were then tested in a social instrumental (SIT) task during adulthood. Bottom: schema of the SIT for one session. The mice can press the lever (I) to open a gridded auto‐guillotine door (II) and interact with a sex‐matched juvenile conspecific (III). The door stays open for 7 s before closing (IV). One session lasts 20 min, and the animals can press the lever at any time during the session. (b) Number of lever presses across days for scr‐ (light blue, n = 5) and shShank3 (dark blue, n = 8). Unlike nonlearners (NL), learner (L) mice exhibit an increase in lever presses across sessions. (c) Bar graphs reporting the mean number of lever presses per scr‐ (light blue, n = 5) and shShank3 (dark blue, n = 8) mice in the last 3 days of the task (day 23–25, Mann–Whitney U test: p = 0.020). (d) The proportions of L (learner, dark orange) and NL (non‐learners, light orange) mice are represented in pie charts. (e) Duration of door being open in seconds across days for scr‐ (light blue, n = 5) and shShank3 (dark blue, n = 8). (f) Bar graphs reporting the mean duration of door in an open state for scr‐ and shShank3 during the last 3 days of the task (day 23–25, Mann–Whitney U test: p = 0.065). (g) Bar graphs reporting the ratio of door open duration on the first three vs. the last 3 days of the task for scr‐ and shShank3 (Mann–Whitney U test: p = 0.045). (h) Time spent near the door zone when open across days for scr‐ (light blue, n = 5) and shShank3 (dark blue, n = 8). (i) Bar graphs reporting the mean duration of time spent near the door when open per scr‐ (light blue, n = 5) and shShank3 (dark blue, n = 8) mice during the last 3 days of the task (day 23–25, Mann–Whitney U test: p = 0.045). (j) Bar graphs reporting the ratio of time spent near the door when open on the first three vs. the last 3 days of the task for scr‐ and shShank3 (unpaired t‐test: t ₍₁₁₎ = 2.353, p = 0.038). Data are shown as the mean ± SE.

FIGURE 3

Shank3 expression deficits in the NAc lead to impaired association in a conditioned place preference (CPP) task. (a) Top: schema of injection sites in the NAc with AAV‐scrShank3‐GFP or AAV‐shShank3‐luczsGreen in P6 mice. Mice were then tested in a conditioned place preference (CPP) task during adulthood. Bottom: schema illustrating the CPP procedure and apparatus. The protocol included a 15‐min pre‐test, a conditioning phase (30 min per day over four consecutive days), and a 15‐min post‐test (cf. Materials and Methods for more details). (b) Representative apparatus occupancy per scr‐ (light blue) and shShank3 (dark blue) mice during pre‐TEST and post‐TEST. (c) Left: preference index calculated at pre‐ and post‐TEST for scrShank3 mice (n = 15, Wilcoxon signed‐rank test: p = 0.026). Right: percentage of time passed in the chambers for scrShank3 mice (n = 15, two‐way ANOVA: phase × time in chamber interaction: F _{(1, 56)} = 12.69, p = 0.0008; phase main effect: F _{(1, 56)} = 0.1531, p = 0.697; time in chamber main effect: F _{(1, 56)} = 4.977, p = 0.0297, followed by Bonferroni's multiple comparisons test). (d) Left: preference index calculated at pre‐ and post‐TEST for shShank3 mice (n = 12, Wilcoxon signed‐rank test: p = 0.42). Right: percentage of time passed in the chambers for shShank3 mice (n = 12, two‐way ANOVA: phase × time in chamber interaction: F _{(1, 44)} = 0.605, p = 0.441; phase main effect: F _{(1, 44)} = 0.983, p = 0.327; time in chamber main effect: F _{(1, 44)} = 0.038, p = 0.847, followed by Bonferroni's multiple comparisons test). (e) Left: distance moved measured during pre‐ and post‐TEST for scr‐ (light blue, n = 15) and shShank3 (dark blue, n = 12) mice (two‐way ANOVA: phase × group interaction: F _{(1, 46)} = 0.1.786, p = 0.188; phase main effect: F _{(1, 46)} = 0.937, p = 0.338; group main effect: F _{(1, 46)} = 2.351, p = 0.1321, followed by Bonferroni's multiple comparisons test). Right: Correlations between distance moved and preference index in the post‐TEST for scr‐ (light blue, n = 15) and shShank3 (dark blue, n = 12) mice. Pearson correlation coefficients and relative p values are reported per both groups. (f) Left: Schematic illustration of the video analysis. Pose estimation of 10 body parts with DeepLabCut and consecutive segmentation of the mice behavior in motifs with VAME. Right: Dendrogram reporting the clustering and behavioral annotations of the 41 identified motifs. (g) Motif usage in the empty versus social‐associated chambers is shown for scr‐ (left, n = 15) and shShank3 (right, n = 12) mice during the post‐TEST phase. Each point represents the mean motif usage per group, with the corresponding SEM. The difference between mean usage in the two conditions was calculated for each motif, and a paired t‐test was performed to assess whether the differences significantly deviated from the diagonal. Only motifs with significant differences in usage are highlighted, with colors corresponding to the behavioral annotations described in panel (f). All the data are shown as the mean ± SE.