Decrease of GSK-3β Activity in the Anterior Cingulate Cortex of Shank3b−/− Mice Contributes to Synaptic and Social Deficiency

Social deficiency is one of the core syndromes of autism spectrum disorders (ASD), for which the underlying developmental mechanism still remains elusive. Anterior cingulate cortex (ACC) plays a key role in integrating social information and regulating social behavior. Recent studies have indicated that synaptic dysfunction in ACC is essential for ASD social defects. In the present study, we investigated the development of synapses and the roles of glycogen synthase kinase 3β (GSK-3β), which mediates multiple synaptic signaling pathways in ACC by using Shank3b−/− mice (a widely used ASD mouse model). Our data revealed that Shank3b mutation abolished the social induced c-Fos expression in ACC. From 4 weeks post-birth, neurons in Shank3b−/− ACC exhibited an obvious decrease in spine density and stubby spines. The length and thickness of post-synaptic density (PSD), the expression of vesicular glutamate transporter 2 (vGlut2) and glutamate receptor 2 (GluR2), and the frequency of miniature excitatory post-synaptic currents (mEPSCs) were significantly reduced in Shank3b−/−ACC. Interestingly, the levels of phosphorylated GSK-3β (Ser9), which inhibits the activity of GSK-3β, decreased along the same time course as the levels of GluR2 increased in ACC during development. Shank3b mutation leads to a dramatic increase of pGSK-3β (Ser9), and decrease of pPSD95 (a substrate of GSK-3β) and GluR2. Local delivery of AAV expressing constitutively active GSK-3β restored the expression of GluR2, increased the spine density and the number of mature spines. More importantly, active GSK-3β significantly promoted the social activity of Shank3b−/− mice. These data, in together, indicate that decrease of GSK-3β activity in ACC may contribute to the synaptic and social defects of Shank3b−/− mice. Enhancing GSK-3β activity may be utilized to treat ASD in the future.


INTRODUCTION
Autism spectrum disorders (ASDs) are a group of neural developmental disorders characterized by repetitive stereotype behaviors and social defects (Takumi et al., 2019). Hundreds of related genetic mutations have been identified in human patients (Jacob et al., 2019). Among these candidate genes, Src-homology domain 3 (SH3) and multiple ankyrin repeat domains 3b (Shank3b) is one of the few genes which can cause the core syndrome of ASD at single mutation (Peca et al., 2011;Varghese et al., 2017). SHANK3 is a post-synaptic scaffold protein widely expressed by excitatory neurons, and interacts with multiple synaptic proteins (Monteiro and Feng, 2017). Even heterozygous Shank3b mice exhibit repetitive grooming and social defects (Dhamne et al., 2017;Qin et al., 2018). Previous studies have revealed that dysfunction of striatum glutamatergic transmission is essential for the repetitive grooming behavior of Shank3b −/− mice . The mechanism underlying the social deficiency of Shank3b −/− mice remains largely unknown.
Proper social behavior is essential for mammalian survival and reproduction. Among the brain regions involved in the regulation or generation of social behavior, anterior cingulate cortex (ACC) is regarded as the core region for integrating social information afferents from other brain regions such as the hippocampus, thalamus, and prefrontal cortex (PFC; Apps et al., 2016). Recently, our study demonstrated that the dysfunction of ACC, especially the dysfunction of excitatory synaptic transmission, accounted for the social deficiency in Shank3b −/− mice (Guo et al., 2019). Developmental synaptic deficiency has been proposed to be a key pathological change of ASDs (Lima Caldeira et al., 2019). How Shank3b mutation leads to synaptic defects in ACC is an interesting topic to be explored.
Glycogen synthase kinase 3β (GSK-3β) is a conserved serine/threonine kinase highly abundant in the brain. It is involved in multiple cellular process and signaling pathways, particularly Wnt signaling and mTOR signaling (Meffre et al., 2014;Hermida et al., 2017). In neurons, the function of GSK-3β is closely related to synaptic development and plasticity (Hur and Zhou, 2010). It can phosphorylate the N-methyl-D-aspartate (NMDA) receptor and post-synaptic density protein 95 (PSD95), thereby modulating the function of glutamic synapses (Peineau et al., 2007;Nelson et al., 2013). Considering that SHANK3B is located mainly in the post-synaptic components of excitatory synapses, we are curious as to whether GSK-3β are involved in the social deficiency of Shank3b −/− mouse.
In the present study, we investigated the synaptic defects in the ACC of Shank3b −/− mice during development. Our data demonstrated that the decrease of GSK-3β activity in ACC contributed to the synaptic and social deficiency of Shank3b −/− mice.

Animals and Virus Injection
Shank3b −/− mice were a gift from Prof. Guoping Feng as described (Peca et al., 2011). The mice were maintained and housed in the animal facility of the Fourth Military Medical University at an ambient temperature of 25 • C, with a 12 h light and 12 h dark cycle and bred by mating heterozygous male with heterozygous female. The mice were given ad libitum access to food and water. All procedures conducted with mice were in compliance with the guidelines for experimental animal care and use by the Committee of the Animal Care and Use Committee of Fourth Military Medical University. The protocol of animal experiments was approved by the Committee of the Animal Care and Use Committee of Fourth Military Medical University.
Recombinant adeno-associated virus (rAAV) expressing constitutive active GSK-3β (converting Ser9 into Ala9, 1 × 10 13 PFU) driven by hSyn promoter was obtained from BrainVTA (Wuhan, China). The virus was injected into the bilateral ACC of Shank3b −/− mice (4 weeks old) at 0.2 mm right or left to Bregma, 0.75 mm anterior to Bregma with 400 nl per point and 1 mm in depth. AAV expressing luciferase was used as control.

Golgi Staining
Golgi staining was performed as previously described with minor modifications (Zhang et al., 2011). Animals were perfused with 0.9% saline solution or 0.01 M phosphate-buffered saline (PBS; pH 7.4). The brain tissue was removed and immersed in Golgi-cox solution (consisting 5% potassium chromate, 5% potassium dichromate, and 5% mercuric chloride) for further fixation, after which the tissue was maintained in the dark at room temperature for 2-3 days. Then, the brains were transferred to a fresh Golgi-Cox Solution for additional 14 days. After that, the brains were transferred to 25%-30% sucrose for 2 more days. Coronal sections (100-200 µm) were cut serially. For staining, brain sections were washed in deionized water for 1 min, placed in 50% NH 4 OH for 30 min, and subsequently in fixing solution (Kodak; Rochester, NY, USA) for an additional 30 min. The sections were then incubated in 5% sodium thiosulfate for 10 min. After rinsing with distilled water, dehydration with gradient ethanol was performed. The sections were finally mounted and observed under the bright field of confocal microscope FV1000, images were taken by z-stack scanning with the excitation wavelength of 405 nm, and then the virtual color was converted into red color.

Three-Chamber Test
The 3-Chamber apparatus was an opaque acrylic box with two pull-out doors and three chambers. Each chamber was identical in size (41 × 20 cm), with the dimensions of the entire box being 63 (length) 43 (width) × 23 cm (height). There was a 10-cm gap between adjacent chambers which could be opened or closed with the removable doors. Before tests, mice were individually habituated in the 3-Chamber apparatus for 10 min. After habituation, a C57 stimulus mouse of same age and same sex was placed in the inverted wired cylinder in the ''social chamber.'' The cylinder in the ''non-social chamber'' remained empty. The time the tested mice spent in the social vs. non-social chambers during the 10 min test period was measured. Only when all four paws entered the chamber, the mouse was considered to be within a specific chamber. The behaviors of each mouse were video-recorded during the entire test to assess the details of social behavior (rear, contact, sniff, grooming, stretch, withdrawal, and nose-to-nose). The chamber was cleaned by 75% ethanol between each test. The time and traveled distance were analyzed by using SMART3.0 software (Panlab Harvard Apparatus, Spain).

Resident-Juvenile-Intruder Home-Cage Test
Social interaction was examined as described with minor modifications (Felix-Ortiz and Tye, 2014). Briefly, a male adult Shank3b −/− mouse was allowed to explore freely for 1 min (habituation) in his home cage. Another novel juvenile (3-4 weeks old) male C57BL/6 mouse was introduced to the cage and allowed to explore freely for 3 min (test session). Juvenile mice were used to avoid mutual aggression. All behaviors were video recorded and analyzed by a researcher blind to the testing condition. The time and frequency of direct contact (pushing the snout or head underneath the juvenile's body and crawling over or under the juvenile's body) were measured.

Electron Microscopic Study
Animals were perfusion fixed with a mixture of 4% paraformaldehyde containing 1% glutaraldehyde. Tissue sections of 50 µm were prepared with a vibratome and further fixed with 1% osmium tetroxide, dehydrated with graded ethanol, replaced with propylene oxide, and flat-embedded in Epon 812. The sections were trimmed under a stereomicroscope and mounted onto blank resin stubs for ultrathin sectioning. Ultrathin sections (70-90 nm) were prepared on an LKB Nova Ultratome (Bromma). After being counterstained with uranyl acetate and lead citrate, the sections were examined under a JEM-1230 electron microscope (JEM, Tokyo).

Morphological Analysis
All images of Golgi staining and immunofluorescent staining were taken by Olympus FV1000. For sholl analysis, in cases where individual neurons could not be identified, branches of non-target neurons were manually erased so that the target neuron with full processes was clearly left in the field. The branch intersections and length of dendrites were determined by serial circles surrounding cell body every 5 µm until reaching the end of the longest dendrite using ImageJ. For spine analysis, IMARIS 7.5 filament tracer was used to reconstruct each dendrite. Using customized algorithms, spines were classified as filopodia, stubby, long-thin, or mushroom (Berry and Nedivi, 2017). For the fluorescent intensity, images were taken under the same setting (at least five images from each animal) and analyzed by ImageJ as described (Bhat et al., 2017).

Data Analysis
Western-blotting images were analyzed by ImageJ. For immunohistochemistry and Western-blots, at least three biological repeats were performed for each experiment. For behavior study, at least 15 mice were included in each group. The data were presented as means ± standard error of mean (SEM). The normality was assessed by the Shapiro-Wilk test. Data were analyzed by one-way analysis of variance (ANOVA) except for Sholl analysis which was analyzed by two-way ANOVA, followed by Dunnett post hoc using SPSS l6.0 (Chicago, IL, USA) or by unpaired, two-tailed Student's t-test. P-values less than 0.05 were considered as statistically significant.

Compromised Response of ACC Neurons to Social Stimulus in Shank3b −/− Mice
We first analyzed the response of ACC neurons to social stimulation. The expression of c-Fos, a quick-response gene widely used to reflect neuronal activation, was examined at 1 h after the 3-chamber test. The testing mice were placed into the ''social chamber'' for 10 min, and the control mice placed in the ''non-social chamber'' for 10 min. The average social time was approximately 270 s. Very few or no c-Fos-positive neurons could be observed in the ACC of both WT and Shank3b −/− mice under control conditions ( Figure 1A, left panels; Figure 1B). A lot of c-Fos-positive neurons were found in the ACC of social stimulated WT mice, while the number of c-Fos-positive neurons in the ACC of social stimulated Shank3b −/− mice remained at a similar level as that in WT control ( Figure 1A, right panels; Figure 1B). Both in WT and Shank3b −/− mice, there was no c-Fos expression in striatum, the key region involved in the grooming phenotype of Shank3b −/− mice (Supplementary Figure S1). These data indicated that ACC, but not striatum, responds to social stimulation, which can be compromised by Shank3b mutation.
Abnormal Synaptic Development in the ACC of Shank3b −/− Mice Because SHANK3B is mainly expressed in synaptic sites of excitatory neurons, we focused on the synaptic development of pyramidal neurons in ACC. Golgi staining was conducted at 2, 3, and 4 weeks post-birth. A gradual morphological differentiation was observed. Typical spines along dendrites were not detected until 4 weeks post-birth (Supplementary Figure S2). Therefore, in the present study, we mainly analyzed phenotype at 4 weeks post-birth. Sholl analysis showed that there was no difference in the total dendritic length between Shank3b −/− ACC neurons and WT ACC neurons. However, significantly less basal dendrite branches and more proximal apical branches were found in Shank3b −/− ACC neurons (Figure 2A), indicating that Shank3b may influence the dendrite development of ACC pyramidal neurons.
We next evaluated the spine development in the ACC neurons of Shank3b −/− mice with focus on basal dendrites. Quantification showed that the density of spines along basal dendrites was remarkably reduced in Shank3b −/− neurons as compared with that of WT neurons (Figure 2B, left and middle panels). Further analysis of the sub-types of spines showed that there was no change of immature filodopia and thin spines. However, the number of stubby spines decreased significantly ( Figure 2B, right panel), indicating that the maturation of excitatory synapses may be affected by Shank3b mutation. In line with this result, immunohistochemistry and Westernblotting showed that the expression of vesicular glutamate transporter 2 (vGlut2) was significantly reduced in the ACC of Shank3b −/− mice ( Figure 2C). Electron microscopic study further revealed that the average length and thickness of PSD were significantly reduced in the ACC of Shank3b −/− mice ( Figure 2D). Further, patch-clamp recording showed that the frequency of mEPSC was significantly reduced in the ACC of Shank3b −/− mice ( Figure 2E). These data indicate that Shank3b mutation may impair the formation and function of excitatory synapses in ACC.
Decrease of GSK-3β Activity and GluR2 Expression in the ACC of Shank3b −/− Mice To explore the possible underpin mechanisms, we focused on GSK-3β signaling, a key kinase involved in the dendrites growth and polarization of neurons. GSK-3β interacts with multiple signaling pathways, such as Wnt/β-catenin signaling, mTOR signaling and MAPK signaling (Hur and Zhou, 2010;  Hermida et al., 2017), and phosphorylates many cytoskeleton proteins (Buttrick and Wakefield, 2008), thereby regulating dendrite development. We first investigated the expression of phosphorylated GSK-3β (p-GSK-3β, Ser 9), an inactive form of GSK-3β, and glutamate receptor 2 (GluR2), a major AMPA receptor which is correlated to synaptic localization of SHANK3 (Grimes and Jope, 2001;Ha et al., 2018). During postnatal development of ACC, the expression of GluR2 increased from 3 weeks, and remained at a high level from then on ( Figure 3A). Interestingly, the expression of p-GSK-3β (Ser 9) in ACC dramatically decreased from 3 weeks, and stayed at a low level from then on ( Figure 3B). These data revealed an inverse expression trend of GSK-3β activity and GluR2 protein during development. We then analyzed the expression of GSK-3β and GluR2 in the ACC of Shank3b −/− mice at 4 weeks ( Figure 3C). The expression of GluR2 decreased significantly in Shank3b −/− ACC. The levels of p-GSK-3β (Ser 9) were significantly increased in the ACC of Shank3b −/− mice, while the levels of total GSK-3β kept unchanged ( Figure 3C). Since p-GSK-3β (Ser 9) is an inhibitory form of GSK-3β, we next examined the phosphorylation of PSD95, a substrate of GSK-3β. The results showed that p-PSD95 was significantly reduced in Shank3b −/− ACC ( Figure 3C). Further, we assessed the expression of Akt and p70-S6, two key proteins which modulate GSK-3β activity. The phosphorylation of Akt was increased in the ACC of Shank3b −/− mice while the expression of total Akt and p70-S6 remained unchanged (Figure 3C, Supplementary  Figure S3A). The levels of p-GSK-3β in mPFC and striatum of Shank3b −/− mice remained at similar levels as those in WT mice (Supplementary Figure S3B). These data indicated that GSK-3β activity is lowered in Shank3b deficient neurons.

Rescue of Synaptic Development and Social Behavior by ACC Expression of Constitutive Active GSK-3β
To investigate the effects of enhancing GSK-3β activity on synapse development, we injected AAV-expressing constitutive active GSK-3β (aGSK-3β) into the ACC of Shank3b −/− mice.
Our preliminary experiments showed that it was very hard to precisely inject the virus into ACC at 3 weeks or earlier, possibly due to the small size of ACC (data not shown). We performed virus injection at 4 weeks post-birth and the area of virus diffusion was determined by mixing Hoeschst33342 with virus aliquots (Figure 4A). Three weeks later, significant up-regulation of total GSK-3β was confirmed by Western-blotting ( Figure 4B). Although the total length of dendrites was not significantly affected by aGSK-3β, more dendrite intersections were found in Shank3b −/− ACC treated with aGSK-3β than control ( Figure 4C). Further, Sholl analysis showed that branches along both the basal and apical dendrites were significantly increased in aGSK-3β-treated neurons ( Figure 4C). More importantly, not only was the spine density increased (Figure 4D), but the mature spines (stubby and mushroom spines) were also remarkably increased (Supplementary Figure S4). In addition, the expression of GluR2 was significantly up-regulated in the Shank3b −/− ACC treated with aGSK-3β ( Figure 4E). These data indicated that aGSK-3β could rescue the synapse formation in Shank3b deficient ACC. We next tested whether over-expressing aGSK-3β in ACC could affect the social behavior of Shank3b −/− mice. In the 3-chamber test which measures social approach and preference for social novelty, control Shank3b −/− mice spent similar time exploring the cylinder containing social mice as exploring the empty cylinder ( Figure 5A). Shank3b −/− mice treated with aGSK-3β spent obviously more time in the box containing social mice ( Figure 5A). In the resident-juvenile-intruder home-cage test which measures social interaction, Shank3b −/− mice treated with aGSK-3β showed significantly more direct contact with the juvenile intruder, while Shank3b −/− mice treated with control virus showed no obvious interests to the juvenile intruder ( Figure 5B). These data suggested that over-expressing aGSK-3β in ACC can rescue the social deficiency of Shank3b −/− mice.

DISCUSSION
Recently, we reported that ACC is a key region for the social deficit of Shank3b −/− mice. In the present study, we further analyzed the effects of Shank3b mutation on the morphological development of ACC neurons with focus on synaptogenesis. Our data revealed a defect of excitatory synaptic development and a decrease of GSK-3β activity in the ACC of Shank3b −/− mice. Interestingly, the expression of constitutive active GSK-3β in ACC rescued both the synaptic and social deficiency caused by Shank3b mutation. Our data further supported a crucial role of ACC in social function. Considering the extensive connection of ACC, it will be interesting to investigate how the interaction between ACC and its connecting regions, such as PFC, affects social behavior.
Of the current hypothesis about the development of ASD, ''developmental synaptopathie'' has drawn more and more attention in recent years (Ebrahimi-Fakhari and Sahin, 2015), which implies that the disruption of the synaptic excitation and inhibition (E/I) is crucial for the progress of ASD (Howell and Smith, 2019). Shank3 is one of the few genes whose single mutation is sufficient to induce typical autism syndrome. The dysfunction of excitatory synapses in the striatum of Shank3 −/− mice has been attributed to the reduction of NMDA/AMPA ratio which in turn results in the silence of excitatory synapses (Jaramillo et al., 2016). Our data showed that in the ACC of Shank3 −/− mice, the density of spines, the length and thickness of PSD, and the expression of GluR2 were significantly reduced. These data demonstrated the abnormal development of excitatory synapses in the ACC of Shank3b −/− mice. The reduction of vGluT2 is interesting. Our electrophysiological data are in line with this observation. It is possible that post-synaptic SHANK3B may affect synaptic transmission through neurexinneuroligin-mediated transsynaptic signaling (Arons et al., 2012). Considering that previous studies have reported presynaptic localization of Shank3 in developing neurons (Halbedl et al., 2016;Wu et al., 2017) and that our data were obtained at 4w post-birth, it may be possible that SHANK3 may play a role in the presynaptic function during development. Because SHANK3B could also be expressed by GABAergic neurons, the possibility that abnormal inhibition may affect the morphology of neurons and regulate the structural and functional plasticity of glutamatergic synapses is interesting, and deserves to be further investigated.
The existence of six protein-protein interaction domains in SHANK3B imparts it with the ability to interact with or affect diverse synaptic proteins. Dozens of Shank3-interacting proteins have been identified, such as Contactin, Homer, and SAPAP (Monteiro and Feng, 2017). Recently, SHANK3B has been reported to directly interact with β-catenin (Qin et al., 2018), a Wnt signaling protein which can modulate synapse formation and stability. As an upstream molecule of β-catenin in Wnt signaling, GSK-3β is mainly distributed in post-synaptic components. Our data showed a significant enhancement of p-GSK-3β (S9) in the ACC of Shank3b −/− mice, suggesting that GSK-3β activity was inhibited in Shank3b −/− mice, which is consistent with previous reports that Wnt/βcatenin is activated in Shank3b deficient PFC (Qin et al., 2018). The inverse expression trend of GSK-3β and GluR2, and the rescue effects of aGSK-3β on GluR2 indicated a relationship between the expression of GluR2 and GSK-3β activity. The involvement of GSK-3β in the synapse dysfunction of other ASD models (e.g., FMR1 −/− mice) has been documented (Mines et al., 2010;Caracci et al., 2016). Knocking out GSK-3β in neurons resulted in fewer spines (Kondratiuk et al., 2017), similar to what we have observed. Whether SHANK3B FIGURE 5 | Effects of over-expressing aGSK-3β in Shank3b −/− ACC on social behavior. (A) Three-chamber tests of control Shank3b −/− mice and Shank3b −/− mice treated with aGSK-3β at 3 weeks after virus injection. Notice that aGSK-3β treatment significantly rescued the social preference of Shank3b −/− mice. (B) Resident-juvenile-intruder home-cage test of control Shank3b −/− mice and Shank3b −/− mice treated with aGSK-3β. Notice that aGSK-3β treatment significantly enhanced social contact of Shank3b −/− mice with juvenile intruder. Values represent mean ± SEM. * * P < 0.01. * * * P < 0.001. N = 15 mice per group (A,B). One-way ANOVA (A). Student's t-test (B).
could directly interact with GSK-3β is worthy of further exploration.
Importantly, our data showed that expressing constitutive active GSK-3β in Shank3b −/− ACC could rescue both synaptic development and social function. The rescue of GluR2 expression may be achieved by its downstream β-catenin signaling. So far, there still lacks effective treatment for the social deficiency in ASD patients. Considering too much activation of GSK-3β by constitutive a GSK-3β may lead to inhibitory effects, pharmacologically enhancing GSK-3β activity might be an alternative for the improving social behavior in ASD patients.
Considering the multiple signaling GSK-3β involved and the wide distribution of GSK-3β, it is possible that this synaptic function of SHANK3B may also be important for the pathology of many other neuropsychiatric and neurodevelopmental disorders, such as Alzheimer's disease and Schizophrenia.

DATA AVAILABILITY STATEMENT
The datasets generated for this study are available on request to the corresponding author.

ETHICS STATEMENT
The animal study was reviewed and approved by the Committee of the Animal Care and Use Committee of Fourth Military Medical University. Written informed consent was obtained from the owners for the participation of their animals in this study.

AUTHOR CONTRIBUTIONS
MW, XL and YH conducted most of the experiments, collected and analyzed the data. HZ and YZ contributed to morphological analysis. JK contributed to electron microscropic study. FW contributed to electrophysiology. JC contributed to Golgi staining. XL contributed to behavior analysis. SW and YW conceived the study, provided financial support, analyzed the data, and prepared the manuscript.

ACKNOWLEDGMENTS
We thank Drs. Li Zhang, Lirong Liang, and Chuchu Qi for their technical help, and Prof. Wenting Wang for helpful discussion.