Imbalanced post- and extrasynaptic SHANK2A functions during development affect social behavior in SHANK2-mediated neuropsychiatric disorders

Mutations in SHANK genes play an undisputed role in neuropsychiatric disorders. Until now, research has focused on the postsynaptic function of SHANKs, and prominent postsynaptic alterations in glutamatergic signal transmission have been reported in Shank KO mouse models. Recent studies have also suggested a possible presynaptic function of SHANK proteins, but these remain poorly defined. In this study, we examined how SHANK2 can mediate electrophysiological, molecular, and behavioral effects by conditionally overexpressing either wild-type SHANK2A or the extrasynaptic SHANK2A(R462X) variant. SHANK2A overexpression affected pre- and postsynaptic targets and revealed a reversible, development-dependent autism spectrum disorder-like behavior. SHANK2A also mediated redistribution of Ca2+-permeable AMPA receptors between apical and basal hippocampal CA1 dendrites, leading to impaired synaptic plasticity in the basal dendrites. Moreover, SHANK2A overexpression reduced social interaction and increased the excitatory noise in the olfactory cortex during odor processing. In contrast, overexpression of the extrasynaptic SHANK2A(R462X) variant did not impair hippocampal synaptic plasticity, but still altered the expression of presynaptic/axonal signaling proteins. We also observed an attention-deficit/hyperactivity-like behavior and improved social interaction along with enhanced signal-to-noise ratio in cortical odor processing. Our results suggest that the disruption of pre- and postsynaptic SHANK2 functions caused by SHANK2 mutations has a strong impact on social behavior. These findings indicate that pre- and postsynaptic SHANK2 actions cooperate for normal neuronal function, and that an imbalance between these functions may lead to different neuropsychiatric disorders.


SH-RX (P>90)
Fig. S1. Expression pattern of the SHANK2A(R462X) transgene as monitored by the coregulated nuclear-localized β-galactosidase. The nuclear β-galactosidase activity monitored by X-Gal staining is shown in a coronal brain section of an adult SH-RX mouse (C57Bl6/N; Tg (PtetO-nlacZ-SHANK2A(R462X) /Tg Camk2a-tTA ). The intensity of the X-Gal identifies the CA1 pyramidal, DG granular cells and cells in the piriform cortex as neuronal cell types with the strongest expression of Tg Camk2a-tTA driven PtetO-nlacZ-SHANK2A (R462X) . Lower expression was found in other forebrain regions including the amygdala and striatum. Scale bars are given in mm.  Immunoblot analysis of hippocampal extracts from 5 SH-WT and 5 control littermates and 6 SH-RX and 6 control littermates. B. Expression analysis of SH-WT P10-on hippocampi by nCounter revealed a significant down-regulation of Gria2, Gria3 and Grm1 indicated by asterisks (n=7 SH-WT and 8 control mice, 3 -5 months). C. Expression analysis of hippocampi by nCounter revealed a significant downregulation of the Gria2,3 and Grm5 mRNA for SH-WT Ad-off mice. For SH-RX Ad-off mice, Gria2 expression was downregulated (n = 7 SH-WT Ad-off and 8 Ctrl. AD-off mice at 5 -8 months of age; n = 7 SH-RX AD-off and Ctrl. AD-off mice at 5 -8 months of age). Unpaired two-tailed Student's t-test followed by the Benjamini-Hochberg test, *p ≤ 0.05. Error bars indicate the standard error of the mean (SEM).

Fig. S5. Peptide and protein quality control for the SWATH analysis.
Boxplots showing the coefficient of variation (CoV) at A. peptide level; B. protein level and C. protein level of the proteins quantified with one and with 2 or more peptides in each experimental condition, as a quality metric of the reproducibility of replicate measurements. The number (n) of peptides or proteins analyzed is given together with the median CoV (black line across the box). Replicas were for controls 14, for SH-WT 7 and for SH-RX 7 biologically independent samples (animals) of the synaptically enriched protein fraction. The box length indicates the interquartile range. e.g in (B) The control group showed 11% median CoV for the 2466 proteins quantified in 14 biological replicates.   Table S1: Alteration in synapse-enriched protein levels from the hippocampi of SH-WT and SH-RX mice. Mutations in genes that are highlighted in red are associated with neuropsychiatric or neurodegenerative disorders. Genes that are highlighted in green are also differentially expressed in Camk2a-tTA mice. Differently expressed genes related to postsynaptic functions are highlighted in yellow. Proteins that were found in SYNGO, and which were annotated in SYNGO are labeled by (++). Syngoportal.Org (Gene list using "ID convert tool" followed by "start gene set analysis" and finally by "annotations". P values ≤ 0.05 are highlighted in gray. Genes differentially expressed Genes associated with psychiatric disorders Genes differentially expressed in Camk2a-tTA mice (KT1) see Table S5 Postsynaptic localization Table S2: Differentially expressed hippocampal genes in SH-RX mice. Mutations in genes are as in Table S3 but now sorted according to FDR adjusted P values. Differentially expressed genes related to postsynaptic functions are highlighted in yellow (see Table S3). For color code in the first column, see Table S3. Genes differentially expressed in SH-WT and SH-RX (see Table S3).

Gene
For the color codes in the first column …… see Table S3.

The open field test
The mouse was placed in the corner of a white acrylic open-field box (40 × 40 × 40 cm) and allowed to explore the arena freely for 10 min while its path was monitored and tracked by a video camera, placed 1 m above the center of the arena. The automatic detection of the mouse's traveled distance and the time spent in the central zone (15 cm apart from the walls) was recorded.

The dark-light box test
The

The neophobia test
Each subject was placed in an arena with an unfamiliar drink (100 μl sweetened condensed milk) in the center. The mouse was allowed to roam the arena for 10 min and the latency, as well as the number of contacts with the drink, was manually assessed 17 .

The burrowing test
The burrowing test is based on the mouse's behavior towards the displacement of items from the tube within its home cage 18 . The tube was filled with 200 g of food pellets covered with 60 g of bedding. The test was performed at 5 p.m. and the pellets remaining in the tube were weighed after 2 h. Then the tube was placed in the cage again. After 12 h, the weight of the remaining pellets in the tube was finally assessed.

The puzzle box test
The puzzle box test was slightly modified from the one described in 20 . The puzzle box consisted of two compartments (a brightly-lit start zone and a smaller covered goal zone) separated by a barrier that had a narrow underpass (about 4 cm wide). Each mouse was introduced into the start zone and the task was to enter the goal zone where it could find some bedding from its home cage. Mice underwent a total of 11 trials over 4 consecutive days, with three trials per day on the first three days, and two trials on the last day. On day 1, during trial 1, the underpass was unblocked and the barrier had an open door above the underpass. In trial 2 and trial 3, the barrier had no doorway and the animals had to enter the goal zone via a small underpass. On day 2, trial 4 was identical to trial 2 and trial 3. In trial 5 and trial 6, however, the underpass was filled with sawdust and the animals had to dig through the sawdust. On day 3, the animals had to repeat trial 6 first as well as in trial 7. In trial 8 and trial 9, the animals were presented with the underpass being blocked by a cardboard plug that the mice had to pull out with their teeth and paws to enter the goal zone. Trial 10 on day 4 was again a repetition of trial 9. At the end of the test, in trial 11, the task was to repeat trial 1, like on the first day. After each trial, mice were left for 1 min inside the goal zone.

The fear conditioning test
The fear conditioning test evaluates natural fear learning as described before 21 . For the acquisition session, at first, each mouse had to spend 180 s as habituation in the new arena. Then, an auditory tone was presented for 30 s at a level of 90 dB and frequency of 5,000 Hz with a rise time of 50 ms. A mild foot shock (0.5 mA) was administered during the last 2 s of the tone presentation and co-terminated with the tone. After the shock presentation, an inter-trial interval of 90 s preceded the second and third identical trials. Following the third shock presentation, the mouse remained in the arena for an additional 90 s. On the following day, the contextual testing was conducted similar to the acquisition session including lighting and odor, but without the tone and the foot shock. The experiments lasted for 300 s. On the third day, the cued memory was tested by placing each mouse in a new chamber with different odors, allowing it to habituate for 180 s. The same tone cue as in the acquisition session was then activated for 30 s. Then, an inter-trial interval of 90 s proceeded the second and third trials. The third tone was activated for 300 s until the end of the experiment. The video freeze software was used to record and measure freezing time and numbers.

Amphetamine injection in the open field
Baseline activity in the open field was measured by placing the mouse in the corner of a wooden arena measuring 40 x 40 x 40 cm and allowing it to explore freely for 60 min, while its path was monitored and tracked by a video camera placed 1 m above the center of the arena. Automatic detection of the distance traveled by the mouse as well as its speed was recorded with the SYGNIS tracker software (Sygnis AG).
After 1 h, mice were removed from the open field, i.p. injected with 5 mg/kg amphetamine and immediately returned to the open field for another 60 min.