Association of SHANK Family with Neuropsychiatric Disorders: An Update on Genetic and Animal Model Discoveries

Abstract

The Shank family proteins are enriched at the postsynaptic density (PSD) of excitatory glutamatergic synapses. They serve as synaptic scaffolding proteins and appear to play a critical role in the formation, maintenance and functioning of synapse. Increasing evidence from genetic association and animal model studies indicates a connection of SHANK genes defects with the development of neuropsychiatric disorders. In this review, we first update the current understanding of the SHANK family genes and their encoded protein products. We then denote the literature relating their alterations to the risk of neuropsychiatric diseases. We further review evidence from animal models that provided molecular insights into the biological as well as pathogenic roles of Shank proteins in synapses, and the potential relationship to the development of abnormal neurobehavioral phenotypes.

Introduction

There are three SHANK genes in the human genome, which code a number of proteins particularly involved in synaptic scaffolding. These synaptic proteins can bind to a variety of other proteins enriched at the postsynaptic density (PSD) of excitatory synapses (Mossa et al. 2017; Monteiro and Feng 2017). Shank proteins are a core part of postsynaptic protein networks, thereby playing critical neurobiological roles in synapse formation, plasticity, signaling and transmission. In recent years, many genetic studies have extended strong evidence that alterations in SHANK genes are involved in the development of various psychiatric and neurological disorders, including autism spectrum disorders (ASD), Phelan–McDermid syndrome (PMS), schizophrenia (SCZ), bipolar disorders (BPD) and Alzheimer’s disease (AD; Drapeau et al. 2018; Zhang et al. 2017; Tatavarty et al. 2017; Kanani et al. 2018; Zhao et al. 2019; Alexandrov et al. 2017; MacGillavry et al. 2016). In this review, we first introduce the SHANK family genes, and their transcriptive and translational products. We then elaborate on the findings from genome-wide association studies (GWAS) that implicate their involvement in the development of a number of brain disorders. Finally, we denote findings from genetically modified animal models, which provide mechanistic insights into physiological and pathogenic roles of Shank proteins in structural and functional maintenance of synapses.

SHANK Family Genes and Their Coding Products

Three genes have been discovered to code Shank protein products, named as SHANK1 (also termed Synamon/SSTRIP/Spank-1), SHANK2 (also known as CortBP1/ProSAP1/Spank-3) and SHANK3 (also referred to as ProSAP2/Spank-2), respectively (Mossa et al. 2017). These genes code the corresponding final protein products, with multiple promoter activation, extensive alternative splicing of coding exons and intermediate isoform formation involved in the transcriptional and post-transcriptional processes (Mossa et al. 2017; Monteiro and Feng 2017) (Fig. 1).

Fig. 1.

SHANK family gene structure, protein domains and isoforms. Panels a, b and c show the chromosome locations, gene structures, domains and isoforms of SHANK1, SHANK2 and SHANK3, respectively. The protein-coding domains are marked and aligned to corresponding exons, including ANK, PDZ, PRO, SH3 and SAM. The positions of two promoters are indicated by black arrows. The exons in SHANK2 in red are alternatively spliced, with three identifiable promoters indicated by black arrows and two alternative stop codons (TAA) pointed by yellow arrows (b). The exons of SHANK3 in red are also alternatively spliced, with exon 11a being newly identified. The positions of six identified promoters are indicated by black arrows (c). ANK ankyrin repeat domain, PDZ postsynaptic density protein 95-discs large homologue 1-zonula occludens 1 domain, PRO proline-rich region, SAM sterile alpha motif, SH3 SRC homology 3 domain superfamily

SHANK1 is located on chromosome 19q13.3 in human, which spans ~ 55.1 kb, and contains five domains: ankyrin repeat domain (ANK, located at the N terminus), SRC homology 3 (SH3), postsynaptic density protein 95 (PSD95)-discs large homologue 1-zonula occludens 1 (PDZ), proline-rich domain (PRO) and sterile alpha motif (SAM, located at the C terminus) domains (Sheng and Kim 2000; Grabrucker et al. 2011) (Fig. 1a). It has 23 exons and 2 alternative promoters, which lead to the formation of 2 distinct transcript isoforms: Shank1A (the longest isoform, includes the above 5 domains) and Shank1B (contains PDZ, PRO and SAM domains). SHANK1 also contains alternative splicing sites that can generate different transcripts, for example, Shank1B (lacks the C-terminal SAM domain), Shank1C (lacks the N-terminal ankyrin repeat domain) and Shank1D (lacks ankyrin repeat domain, SH3 or SAM domains). Notably, Shank1 messenger appears to be expressed exclusively in the brain (Lim et al. 1999). In situ hybridization studies in rat show that the Shank1 mRNA is expressed predominately in the cerebral cortex, hippocampus and amygdala, moderately expressed in the hypothalamus and the substantia nigra, while low expression is present in the cerebellum, caudate nucleus, corpus callosum and subthalamic nucleus (Monteiro and Feng 2017; Lim et al. 1999).

SHANK2 is located on chromosome 11q13.3 in human (Fig. 1b). It spans ~ 621.8 kb and is the largest gene among the family. It contains 5 domains, with 25 exons, three alternative promoters and one stop codon. Its transcription results in three different protein isoforms, including Shank2E (containing all five protein domains), Shank2A (lacking the ankyrin repeat domain) and Shank2B (lacking the ankyrin repeat domain and the SH3 domain). Shank2 mRNA occurs in high levels in the brain and at lower levels in kidney and liver (Lim et al. 1999). Thus, Shank2 messenger is widely expressed in many regions, including the cortex, hippocampus, cerebellum, olfactory bulb and central gray (Monteiro and Feng 2017).

SHANK3 appears to be the best studied among the three family members, which is located on chromosome 22q13.3 in human (Fig. 1c). This gene spans ~ 55.1 kb in length and contains 24 exons, 6 alternative promoters and 1 alternative stop codon located in the exon 21b. Six distinct isoforms are produced following the transcriptional processing: Shank3a, Shank3b, Shank3c, Shank3d, Shank3e and Shank3f. Besides alternative promoter and mRNA splicing, the gene expression is regulated by epigenetic mechanisms such as DNA methylation and histone acetylation, resulting in differential tissue specific expression of (Monteiro and Feng 2017; Maunakea et al. 2020).

The full length structure of SHANK3 gene contains ANK, SH3, PDZ, PRO and SAM in human. However, in the mouse genomic DNA, there include another protein domain DUF535, with its function remained unclear. Shank3 mRNA expression appears to be high in the brain, heart and spleen. In the rat brain, Shank3 messenger is enriched in the hippocampal formation, thalamus, striatum and cerebellar granule cells (Lim et al. 1999; Boeckers et al. 1999). Overall, the Shank proteins are best known as synaptic scaffolding apparatus involving in the targeting, anchoring and dynamically regulating neurotransmitter receptors and signaling molecules at postsynaptic sites (MacGillavry et al. 2016). These functions appear to be executed via their domain-specific interaction with other proteins (Fig. 2). For instance, the amino-terminal ankyrin repeat domain interacts with the PSD protein SHARPIN and probably binds to the cytoskeleton through an interaction with spectrin alpha chain, non-erythrocytic 1 (SPTAN1, also known as α-fodrin) (Lim et al. 2001; Böckers et al. 2001). The N-terminal region preceding the ankyrin repeat domain may interact intra-molecularly with the six ankyrin repeats, and could restrict the access to binding partners such as SHARPIN and SPTAN1 (Mameza et al. 2013) (Fig. 2). The class I PDZ domain of SHANK3 can partnership with the SAP90/PSD95 associated protein 1 (SAPAP1, also known as GKAP1) and the glutamate receptor 1 (GluR1; also known as GRIA1) subunit of AMPA receptors (AMPARs) (Monteiro and Feng 2017; Naisbitt et al. 1999) (Fig. 2). The latter is important for dendritic spine formation and synaptic transmission. The PRO domain binds to the Homer and cortactin 10 proteins, which are important for cytoskeleton regulation, synaptic transmission and plasticity (Naisbitt et al. 1999; Hayashi et al. 2009) (Fig. 2). Moreover, the carboxy-terminal SAM domain of Shank proteins is known to self-multimerize, which is required for the localization of the proteins to the PSD (Naisbitt et al. 1999; Boeckers et al. 2005). Besides the above, Shank proteins could cofunction with many other partners to modulate neuronal and synaptic activity (Monteiro and Feng 2017; Guilmatre et al. 2014).

Fig. 2.

Schematic drawing illustrating the potential protein–protein interaction network of the SHANK at the postsynaptic density of glutamatergic synapses. NMDA-R N-methyl-d-aspartate receptor, BDNF brain-derived neurotrophic factor, AMPA alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid, DLGAP1, 2, 3 discs large associated protein 1, 2, 3, SLC6A3 solute carrier family 6 (neurotransmitter transporter, dopamine), member 3

Association of SHANK Gene Variations with Neuropsychiatric Disorders

While the SHANK genes share a very high degree of sequence and domain homology, different proteins and isoforms are differentially expressed in the central nervous system according to developmental stages, cell types and brain regions. Therefore, the neurobiological roles of Shank proteins in the brain are fairly complicated, and could be more or less differentially related to neuronal/synaptic activities and functions. Since the identification of the first mutation in SHANK genes, there has been a spectacular increase in our knowledge on the role of this family in the development of neuropsychiatric disorders. During the past several years, distinct SHANK gene variations are found to be significantly associated with the risks of different disorders (Durand et al. 2007; Toro et al. 2010; Sudhof 2008).

Autism Spectrum Disorder (ASD)

Genetic defects of SHANKs have been identified in patients with ASD in multiple reports. In 2007, Durand et al. first described a specific mutation, SHANK3 interstitial or terminal deletion or frameshift, in ASD patients (Durand et al. 2007). Accumulating genetic association studies indicate that SHANK3 variation is estimated to contribute to ~ 0.5% of ASD, which represents one of the most replicated and well-characterized genetic defects in ASD (Jaramillo et al. 2020; Bey et al. 2018). Different forms or variations of genetic defects are found in patients with ASD, including: (1) terminal microdeletion of SHANK3 in 22q13.3 or ring chromosome of 22 (Drapeau et al. 2018; Jeffries et al. 2005; Okamoto et al. 2007; Bonaglia et al. 2011; Sarasua et al. 2014); (2) point mutation (Moessner et al. 2007; Gauthier et al. 2009; Boccuto et al. 2013; Soorya et al. 2013; Cochoy et al. 2015); (3) small intragenic deletion (Marshall et al. 2008; Chen et al. 2016); (4) insertion mutation (Durand et al. 2007; Jaramillo et al. 2017, 2020; Kolevzon et al. 2011; Bonaglia et al. 2011); (5) chromosome translocations (Moessner et al. 2007; Bonaglia et al. 2001); (6) microduplication (Chen et al. 2016) and (7) frameshift variant (Kanani et al. 2018) (Table 1). Microdeletion of SHANK3 in Phelan–McDermid Syndrome (PMS) is the most common molecular defect in all types of mutations of this gene in ASD. In addition to SHANK3, mutations in SHANK1 and SHANK2 have been found in ASD patients. In 2010, Berkel et al. reported patients with ASD and mental retardation with de novo copy number variations (Berkel et al. 2010). Evidence of SHANK2 mutations causing ASD have been further identified independently in different ASD patients cohorts (Pinto et al. 2010; Wischmeijer et al. 2011; Leblond et al. 2012; Chilian et al. 2013). Mechanistically, Zaslavsky et al. showed that the SHANK2 mutations in association with ASD can lead to hyperconnectivity of human neurons (Zaslavsky et al. 2019). In comparison, there are much less reports about the involvement of SHANK1 in ASD. In 2012, Sat et al. reported an inherited deletion of 63.8 kb, encompassing SHANK1 and CLEC11A genes, segregating in six autistic individuals, four males and two females, from the same family (Sato et al. 2012). Notably, the four males were clinically menefested as high-functioning autism, while the two females carrying the same deletion only displayed anxiety and shyness. In addition, Wang et al. recently identified de novo genetic mutations of SHANK1 gene among a Chinese ASD cohort of 1543 ASD probands (Wang et al. 2016a, b). It should be noted that a recent meta-analysis study on SHANKs mutations in patients of ASD suggest that no SHANK1 duplications occur in patients with ASD (Leblond et al. 2014). Studies also show that SNP–SNP interactions in SHANK family may confer ASD risk in the Northeast Han Chinese population (Qiu et al. 2018; Bai et al. 2018), whereas there exists no correlation between SHANK1 SNPs and ASD (Qiu et al. 2019). In sum, SHANK2 and SHANK3 have genome-wide significance for their association with ASD. On the other hand, given the rare frequency of SHANK1 deleterious mutations, the clinical relevance of SHANK1 to the development of ASD remains to be ascertained.

Table 1.

Types of genetic defects in shank3 of ASD patients

Genetic defect			SHANK3 isoforms affected	Other genes disrupted	Diagnosis	Reference
Microdeletion of SHANK3 in 22q13.3 or ring chromosome of 22			All isoforms disrupted	More than 30 other genes	PMS	Bonaglia et al. (2011), Jeffries et al. (2005). Okamoto et al. (2007) and Drapeau et al. (2018)
Frameshift variant	Varianted copy number
	c.1231delC, p.Arg411Val		–	–	ASD	Kanani et al. (2018)
Point mutation	Rearrangement	Exon/intron/domain
	E809X, Q1243X, G1271Afs * 15, L1142Vfs * 153	Exon 21/Pro-Rich-Domain	SHANK3	–	ASD	Gouderet al. (2019)
	c.1527 G>A	Exon 12/SH3	SHANK3a	De novo 17q12 microduplication	ASD/ID	Soorya et al. (2013)
	c.5008 A>T	Exon 22/SAM	SHANK3a	–	ASD/ID	Cochoy et al. (2015)
	c.4402 G>T	Exon 22	–	–	ASD	Boccuto et al. (2013)
	c.1304 + 48 C>T	Intron 10	–	c.421 C > G	ASD	Boccuto et al. (2013)
	P.Q321R	Exon 8/ANK	SHANKa-b	–	ASD	Moessner et al. (2007)
	c. 2265 C + 1 del G	Exon 19	SHANKa-e	–	ASD	Gauthier et al. (2009)
Deletion	Deletion length
	106 kb		–	ACR, RABL2B, LOC150417, RPL23AP82, LOC105373100	ASD	Chen et al. (2016)
	147 kb		–	Microduplication at 19q13.4	ASD	Chen et al. (2016)
	277 kb		–	MAPK8IP2, ARSA, 1.4 Mb gain at 22q13.33	ASD	Moessner et al. (2007)
	4.36 Mb		–	–	ASD	Moessner et al. (2007)
	142 kb		–	ACR, RABL2B	ASD	Durand et al. (2007)
Insertion mutation	Exon/domain
	Exon 21		–	–	ASD	Durand et al. (2007)
	Exon 11		–	–	ASD	Kolevzon et al. (2011)
	Exon 13		–	–	ASD	Mei et al. (2016)
Chromosome translocations	Chromosome
	t(14;22)(q32.33;q13.31) 3.2 Mb deletion		–	–	ASD	Moessner et al. (2007)
	t(12;22)(q24.1;q13.3)		–	–	ASD/PMS	Bonaglia et al. (2001)
Duplication	Duplication length
	1.8 Mb		–	More than 60 other genes	ASD/ADHD	Chen et al. (2016)

ASD autism spectrum disorder, PMS Phelan–McDermid Syndrome, ID intellectual disability, ADHD attention deficit hyperactivity disorder

Phelan–McDermid Syndrome (PMS)

PMS, also called as 22q13.3 deletion syndrome, is a rare and complex neurodevelopmental disorder characterized clinically by global developmental delay, mild dysmorphic features, motor deficits, variable degrees of intellectual disability (ID), and absent or delayed speech (Drapeau et al. 2018). The genetic changes in PMS involved the deletion of the distal long arm of chromosome 22 that could be resulted from pure terminal deletion, ring chromosomes, unbalanced translocations, mosaicisms of these anomalies and SHANK3 mutations (Bonaglia et al. 2011; Luciani et al. 2003; Phelan and McDermid 2012). The first case was a child reported in 1985, until 1994 Nesslinger et al. showed additional cases and proposed that the symptoms described in patients with chromosome 22q13.3 deletions might be a common phenotype (Watt et al. 1985; Nesslinger et al. 1994). In 2001, a PMS patient’ karyotype showed a de novo balanced translocation of t(12;22)(q24.1;q13.3) disrupting SHANK3 in exon 21, providing evidence that SHANK3 is the detrimented gene associated with the neurobehavioral deficits in PMS patients (Bonaglia et al. 2001). Indeed, small deletions of SHANK3 gene were described in many more cases, such as a short simple repeat located between exons 8 and 9, and the deletion breakpoint located in intron 8 (Toro et al. 2010; Bonaglia et al. 2006). In 2016, Zwanenburg et al. reported four PMS patients with a small deletion of 182 to 224 kb including three common genes SHANK3, ACR, and RABL2B (Zwanenburg et al. 2016). The PMS patient carrying an approximately 100 kb deletion were clinically presented as mild intellectual disability, autistic symptoms and speech delay (Britt-Marie et al. 2002). The patient carrying a terminal deletion of the last 130 kb of chromosome 22q also showed autistic features, speech delay and mild intellectual disability (Flint et al. 1995; Wong et al. 1997). Overall, it appears that SHANK3 mutation or disruption is not only a highly expressed single-gene risk factor for autism, but also a genetic causal factor for PMS (Bonaglia et al. 2011). Notably, more than 80% of patients with PMS are considered to meet the clinical criteria for ASD (De Rubeis et al. 2018). It should be further mentioned that individuals with PMS are at high risk of developing a number of neuropsychiatric conditions in adolescence or early adulthood, including catatonia, bipolar disorder and lasting regression of skills (Kohlenberg et al. 2020; Kolevzon et al. 2019).

In addition to the above studies demonstrated the link of SHANK3 alterations to PMS, a few reports showed changes in SHANK2 gene in individuals with developmental deficits. Thus, a de novo 3.5 Mb deletion of SHANK2 from 11q13.2 to 11q13.4 was related to global developmental delay in a case study (Marcou et al. 2017). Furthermore, Fischetto et al. identified a paternal microduplication (400 kb) of SHANK2 in a boy with mild development delay (DD) by single nucleotide polymorphism (SNP) array analysis (Fischetto et al. 2017).

Schizophrenia (SCZ)

There are many studies reported alterations in the SHANK genes in patients with SCZ. In 2007, Failla et al. reported a patient with SCZ exhibited a 22q13.3-qter duplication encompassing SHANK3 (Failla et al. 2007). The patient developed initial behavioral abnormalities (agitation, social closer and passivity) at the age of 13, followed by auditory hallucinations, loss of self-control, dysfunction of self-awareness with temporary amnesia, disorientation, aggressiveness and sleep disturbances. Her symptoms of apathy, poor personal hygiene and adaptive difficulties met the diagnostic criteria for borderline intellectual dysfunction and disorganized SCZ. In 2010, Gauthier et al. reported two de novo mutations (R1117X and R536W) in 2 families by sequencing SHANK3 in 185 SCZ patients, with 1 mutation being found in three brothers diagnosed with SCZ (Gauthier et al. 2010). In 2015, Zhang et al. reported that circulating microRNA-7 (miR-7) level is significantly increased in SCZ patients, which included 50 patients and 50 healthy controls (Zhang et al. 2017). They also showed that the levels of Shank3 mRNA and protein are reduced by overexpression of miR-7, and increased by miR-7 knockdown. Further morede Sena Cortabitarte et al. investigated the relationship between SHANK3 and SCZ, indicating that the p.G1011V variant would be a promising candidate variant for follow-up studies in larger samples and functional investigation (Ana et al. 2017). The role of SHANK2 in SCZ was first reported by Peykov et al. (2015); they sequenced SHANK2 in a cohort of 481 SCZ patients and observed a significant increase in the number of rare missense variants in these individuals as compared with controls. They also detected four non-synonymous variants in the SCZ cohort (i.e., S610Y, R958S, P1119T and A1731S) relevant to functional impairments to various degrees (Peykov et al. 2015). Specially, A1731S overexpression reduced the number of SHANK2-Basoon-positive synapses. Then, in 2016, whole-genome sequencing in numerous families showed that at least three members diagnosed as schizophrenia were caused by mutations of SHANK2 (Homann et al. 2016). Additionally, Lenertz et al. analyzed 199 SCZ patients and 206 healthy controls, and found that the T allele of variant rs3810280 located in SHANK1 promoter led to serious damage to auditory working memory, which was replicated in the other patients at-risk for SCZ (Lennertz et al. 2012). Moreover, Fromer et al. reported a de novo SHANK1 frameshift mutation in a SCZ patient (Fromer et al. 2014). Taken together, emerging evidence points to a potential association between the SHANK gene family and the risk of SCZ, while future investigations are needed to include more patients.

Bipolar Disorder (BPD)

A substantial number of studies support that SHANK3 mutations are related to the risk of BPD. In 1996, Sovner et al. first reported a patient diagnosed with atypical BPD, whose genetic phenotype was a 22q13 deletion syndrome involving SHANK3. Following this report, a series of case studies have shown that patients with the PMS often present a high incidence of BPD. In 2012, Verhoeven et al. described two patients with PMS and BPD in the same family (Verhoeven et al. 2012). The younger brother displayed severe intellectual disability (ID), hyperactivity together with temper tantrums, repetitive behaviors, lack of language communication, loss of interest in daily activities and increase of his social withdrawal. The older brother presented with a mild ID, recurrent major depressive episodes and unstable pattern of mood. In 2012, Denayer et al. reported four patients with PMS and BPD (Denayer et al. 2012). These patients displayed ASD, severe to profound ID, disruptive behaviors, anxiety and self-absorbing behaviors. In 2012, Vucurovic et al. also found seven patients with PMS that displayed psychiatric features of BPD, with this mental status related to SHANK3 complex multiple deletions (Vucurovic et al. 2012). In other clinical and preclinical studies, SHANK3 changes could be correlated with manic or depressive episodes of BPD (Crisafulli et al. 2013; Ortiz et al. 2015). In 2014, Noor et al. identified a 276 kb duplication encompassing several genes in a BPD case, including seven exons of SHANK2 (Noor et al. 2014). More recently, four point mutations in SHANK2 (c.3979G>A; c.2900A>G; c.4461C>T; c.4926G>A) have been identified in BPD patients (Yang and Jiang 2018). To date, there is no report regarding SHANK1 mutation screening in patients with BPD. However, there is one study reporting that psychostimulant-induced hyperactivity is reduced rather than enhanced in shank1 knockout mice. This finding was considered to speak against the observed behavioral phenotype of this gene in BPD (Sungur et al. 2018). In brief summary, mutations in SHANK3 and SHANK2 are associated with the risk of BPD based on genetic studies, whereas whether this situation exists for shank1 remains unclear.

Alzheimer’s Disease (AD)

AD is a common age-related neurodegenerative disease causing dementia in the elderly. The pathological hallmark of AD includes extracellular deposition of β-amyloid (Aβ) and intraneuronal accumulation of phosphorylated tau (pTau) (Sheng et al. 2012). A pathogenic role of Aβ products, especially the soluble Aβ oligomers, has been proposed for AD. Thus, it is suggested that the Aβ oligomers impair synaptic transmission and cause synaptic dysfunction and degeneration (Haass and Selkoe 2007; Crews and Masliah 2010). Studies have been carried out to understand the interplay between the Shank family and AD pathogenesis. Roselli et al. reported that Aβ can disrupt two scaffold proteins, Homer1b and Shank1 (Roselli et al. 2009). Specifically, Aβ treatment decreased levels of these proteins, and induced ultrastructural changes in the postsynaptic sites. Pham et al. observed a reduction in the levels of Shank1 and Shank3 in the brain of APP transgenic mice, with a decrease of Shank1 levels also found in the frontal cortex of patients with AD (Pham et al. 2010). Gong et al. reported a significant increase in the levels of Shank2 in synaptosomes isolated from the middle frontal gyrus in patients with AD as compared with controls (Gong et al. 2009). This group also showed a decrease in the level of Shank3 and an increase of Shank2 protein in AD relative to control, while no significant difference in Shank1 were found between the AD and control groups. In 2012, a study reported a reduced level of Aβ (479 mg/L) in the cerebrospinal fluid in a bipolar disorder patient with dementia carrying a SHANK3 deletion mutation (Vucurovic et al. 2012). Lately, Zhao et al. found that microRNA-34a could mediate a down-regulation of Shank3 in sporadic AD cases (Vucurovic et al. 2012). Together, there is no report thus far indicating distinct SHANKs changes in AD. However, experimental studies point to some connections between Shank protein dysregulation and the molecular neuropathology of AD.

Temporal Lobe Epilepsy (TLE)

To date, there are a few studies on SHANK genes involved in the pathogenesis of temporal lobe epilepsy. In 2013, a study by Han et al. demonstrated that Shank3 overexpression could induce seizures in vivo (Han et al. 2013). To explore the underlying mechanism, Wang et al. carried out a kinome-wide siRNA screen and identified multiple kinases that potentially regulate Shank3 protein stability and serve as druggable target (Wang et al. 2019a, b, c). Specifically, ERK2 was found to bind, phosphorylate and promote the poly-ubiquitination-dependent degradation of Shank3. In 2016, Zhang et al. investigated Shank3 expression of in patients with intractable TLE and in pilocarpine-induced rat model of epilepsy, Shank3 protein levels and immunolabeling were increased in the neocortex of TLE patients and epileptic animals relative to controls (Zhang et al. 2016). Additionally, our group found that following seizure-induction, Shank1/2/3 mRNA was downregulated at acute stage, upregulated at latent stage, and returned to the basal level at chronic stage (Fu et al. 2020). We have also characterized an involvement of DNA methylation in the regulation of shank3 in epilepsy (Fu et al. 2020). Further, inhibition of Shank3 expression could significantly mitigate seizure severity at the acute stage of pilocarpine induced-epilepsy (own unpublished observation). Studies are currently carried out to explore the transcriptional mechanism relative to epileptogenesis in animal models and in clinical setting.

SHANK Gene Modified Animal Models of Neuropsychiatric Disorders

As elaborated above, SHANK gene alterations appear to be widely linked to the risks of multiple neuropsychiatric disorders according to GWAS data from human populations. The genetic variation profiles and the clinical manifestations appear to be also remarkably heterogeneous among the affected families and individuals. Clearly, the SHANK gene family plays pleiotropic neurobiological roles in the nervous system, which warrant thorough mechanistic investigations. In this regard, a number of animal models carrying specific Shank mutations have been established, with researches on these models extended valuable insights into the biological functions of this family as well as mechanistic underpinning of related brain disorders. In this section, we further denote major findings from studies on animals carrying Shank mutations, especially in the context of molecular and functional deficits related to neurobehavioral abnormalities.

Animal Models with Shank1 Mutations

There have been studies using mice in which exons 14-15 of Shank1 are deleted (Shank1 Δex14–15), resulting in the destruction of the PDZ region (Hung et al. 2008; Silverman et al. 2011; Wohr et al. 2011).These mice display mild anxiety-related behavior during the light–dark test, reduced exploratory locomotion in a novel open field and reduced ultrasonic vocalizations (USV). The animals also show reduced motor coordination and contextual fear conditioning, but present enhanced spatial learning or working memory. Levels of the postsynaptic proteins guanylate kinase-associated protein (GKAP) and Homer are reduced in the brain, along with reduced spine density, smaller and thinner PSD in CA1 region of the hippocampus. Moreover, studies have demonstrated early communication deficits and repetitive behaviors in the Shank1 knockout mouse model of ASD (Sungur et al. 2014, 2016). Notably, in Shank1 mutant mice with a targeted replacement of exons 14 and 15, the animals exhibit normal sociability, but insufficient object exploration and deficit in object recognition (Sungur et al. 2017). The findings from this mouse line also suggest that Shank1 is involved in neocortical-dependent associative learning (Collins and Galvez 2018) (Table 2).

Table 2.

Molecular and behavioral phenotypes of shanks mutant animals

Gene	Animal species	Exons targeted	Disorders associated	Behavioural features				References
				Repetitive behaviours	Social behaviors	Learning and Memory	Other behaviors
SHANK1	Mouse	Exon 7 (Δex7)	ASD	Increased self-grooming in females, increased locomotor activity and behaviours, reduced digging	Impaired social interaction	Impaired novel objective recognition memory	Increased anxiety-like behaviors	Schmeisser et al. (2012)
	Mouse	Exons 14-15 (Δex14-15)	ASD	Increased repetitive self-grooming, reduced motor coordination	Deficit in social communicative behavior	Impaired fear conditioning, enhanced spatial learning, impaired retention of spatial memory	Increased anxiety-like behaviors in the light–dark test, reduced exploratory locomotion	Hung et al. (2008), Silverman et al. (2011) and Wohr et al. (2011)
	Mouse	Exons 14-15 (Δex14-15)	ASD	Not mentioned	Delay in early development of ultrasonic communication, deficit in social communication	Not mentioned	Not mentioned	Sungur et al. (2016)
	Mouse	Exons 14-15 (Δex14-15)	BPD	Reduced psychostimulant-induced hyperactivity	Not mentioned	Not mentioned	Not mentioned	Sungur et al. (2018)
	Mouse	Exons 14-15 (Δex14-15)	ASD	Not mentioned	No significant changes in social behaviors	Insufficient object exploration, deficit in object recognition	Not mentioned	Sungur et al. (2017)
	Mouse	Shank1 KO	–	Not mentioned	Not mentioned	Decline of neocortical dependent associative learning and memory consolidation	Not mentioned	Collins and Galvez (2018)
	Mouse	Shank1 KO	ASD	Not mentioned	Deficit in social communication	Not mentioned	Not mentioned	Sungur et al. (2016)
SHANK2	Mouse	Exons 6-7 (Δex6-7)	ASD	Enhance jumping, decreased digging behaviours, increased grooming in a novel object recognition arena	Reduced interaction, normal levels of social novelty recognition and olfactory	Impaired spatial learning and memory, normal levels of novel object recognition memory	Impaired nesting behavior, hyperactivity, increased anxiety-like behaviour	Won et al. (2012) and Lim et al. (2017)
	Mouse	Exon 7 (Δex7)	ASD	Not mentioned	Social deficits	Deficits in spatial learning and memory	Not mentioned	Lim et al. (2017)
	Mouse	L7-cre; Exons 6-7 (Δex6-7)	ASD	No evidence of hyperactivity, task-specific repetitive behaviour	ASD-like social impairments	Impairments in motor learning	Intact baseline motor performance	Peter et al. (2016)
	Mouse	Exons 4-7 (Δex4-7)	ASD	No lesions	Normal initiation of social interaction in identical test, perturbed recognition of social novelty	Not mentioned	No anxiety-like behaviour	Peca et al. (2011)
	Mouse	Exon 24 (Δe24)	BPD	Hyperactivity	Deficits in social behaviors	Impaired in spatial learning, deficits in cognitive behaviors	Enhanced reward-seeking behavior, mania-like behaviors, circadian rhythms change	Pappas et al. (2017)
	Mouse	Pv-Cre; Exons 6-7 (Δex6-7)	ASD Seizure	Moderate hyperactivity, enhanced self-grooming in novel environment	Normal levels of social interaction	Normal learning and memory	Normal anxiety-like behavior	Lee et al. (2018)
	Mouse	Viaat-Cre; Exons 6-7 (Δex6-7)	ASD	Hyperactivity	Social interaction deficits, mild social communication deficits	Not mentioned	Anxiety-like behaviors	Kim et al. (2018)
	Mouse	CaMKII-Cre; Exons 6-7 (Δex6-7)	ASD	Repetitive self-grooming, mild hyperactivity	Social communication deficits	Not mentioned	Normal anxiety-like behaviors	Kim et al. (2018)
	Mouse	Exons 6-7 (Δex6-7)	ASD	Not mentioned	Not mentioned	Not mentioned	Impaired basal tactile perception and acute pain response, decreased sensitivity to chronic pain and inflammatory pain	Ko et al. (2016)
	Mouse	Exons 4-9 (Δex4-9B)	ASD	Excessive self-grooming, inflexible behavior	Reduced social sniffing, reduced reciprocal social interactions	Impaired novel object recognition	–	Bozdagi et al. (2010) and Yang et al. (2012)
SHANK3	Mouse	Exons 4-9 (Δex4-9J)	ASD	Excessive self-grooming, increased head pokes	Reduced interest, decreased bidirectional social interactions	Impaired short and long memory	–	Wang et al. (2011)
	Mouse	Exons 4-9 (Δex4-9)	ASD/PMS	Increased repetitive grooming,	Abnormal social interaction	Impaired novel and spatial object recognition learning and memory	Normal locomotor activity and locomotor habituation	Jaramillo et al. (2016)
	Mouse	Exon 9 (Δex9)	–	Increased self-grooming	Normal	Impaired spatial memory	Increased rearing in a novel environment, absence of anxiety-like behavior	Lee et al. (2015)
	Mouse	Exon 11 (Δex11)	ASD	Increased self-grooming	Impaired social interaction	Impaired spatial memory	No anxiety-like behaviour, exhibit restricted interest and an avoidance phenotype when exposed to inanimate objects, exhibit only minor deficits in motor coordination, increased aggression and hyposensitivity to pain	Schmeisser et al. (2012) and Vicidomini et al. (2017)
	Mouse	Exons 12-13 (SHANKE13)	ASD	Increased repetitive grooming	Social interaction deficits	Impaired spatial learning	Reduced rearing activity	Jaramillo et al. (2017)
	Mouse	Exons 13-16 (Δex13-16)	ASD/PMS	Increased self-grooming	Normal social interaction	Deficits in discrimination learning	Normal locomotor activity	Copping et al. (2017)
	Mouse	Exons 13-16 (Δex13-16)	ASD	Excessive self-injurious grooming, causing skin lesion	Dysfunctional social interaction behavior, decreased reciprocal interaction in interacting dyadic test, decreased anogenital sniffing, decreased frequency of nose-to-nose interaction	Not mentioned	Increased anxiety-like behavior, reduced exploratory locomotion	Peca et al. (2011), Guo et al. (2019) and Wang et al. (2017)
	Mouse	Glutamatergic Shank3 (exons 14–16) Δex14–16	–	Enhanced repetitive self-grooming, normal digging	Enhanced direct social interaction, normal social approach and social novelty recognition	Not mentioned	Normal locomotor activity and enhanced anxiolytic-like behavior	Yoo et al. (2019a, b)
	Mouse	GABAergic Shank3 (exons 14–16) Δex14–16	ASD	Enhanced self-grooming and suppressed digging and enhanced rearing	Enhanced direct social interaction and suppressed social communication	Not mentioned	Show hypoactivity in a novel environment, increased anxiety-like behaviors in the light–dark test, differential influences on other types of anxiety-like behaviors	Yoo et al. (2018)
	Mouse	Global Shank3 Exons 14-16 Δex14–16	ASD	Enhanced self-grooming but digging and climbing behavior was suppressed	Enhanced direct social interaction and suppressed social communication	Not mentioned	Show hypoactivity in a novel environment, increased anxiety-like behaviors in the light–dark test, differential influences on other types of anxiety-like behaviors	Yoo et al. (2018)
		Exon 21 (Δex21)	ASD	Increased grooming in older mice	Minimal social interaction, normal communication	Impaired in spatial learning, impaired spatial memory during the probe trial	Impaired motor coordination, hypersensitivity to heat, avoidance toward inanimate objects, absence of anxiety-like behavior, aberrant locomotor response to novelty	Kouser et al. (2013), Qin et al. (2018) and Ma et al. (2018)
		Exon 21 (lnsG3680)	ASD	Robust repetitive/compulsive grooming behavior, skin lesion between 4 and 6 months of age	Social interaction deficits	Impaired motor learning, normal basic working memory	Increased anxiety-like behavior, reduced locomotion, impaired coordination capability, reduced PPI	Zhou et al. (2016)
		Exon 21 (R1117X)	SCZ	Allogrooming/barbering, skin lesion between 4 and 6 months of age	Social interaction deficits, social dominance behavior	Impaired motor learning, normal basic working memory	Increased anxiety-like behavior, reduced locomotion, impaired coordination capability, impaired PPI	Zhou et al. (2016)
		Exons 4-22 (Δex4-22)	ASD	Excessive self-grooming causing skin lesion	Normal	Impaired spatial memory and learning	Reduced locomotion, reduced startle reactivity	Wang et al. (2016a, b)
	Mouse	Exons 4-22 (Δex4-22)	PMS	Excessive Self-grooming	Normal social interaction	Contextual learning impaired, working memory not altered	Developmental delays, response delays to auditory startle, reduced spontaneous locomotion, impaired motor coordination and balance sensory abilities (sound discrimination, strong visual and olfactory function), increased freezing under fear condition, increased novelty-induced anxiety	Drapeau et al. (2018)
	Mouse	Exons 4-22 (Δex4-22)	PMS	Excessive self-grooming	Normal social interaction	Contextual learning impaired, working memory not altered	Developmental delays, response delays to auditory startle, reduced spontaneous locomotion, impaired motor coordination and balance sensory abilities (sound discrimination, strong visual and olfactory function), increased freezing under fear condition, increased novelty-induced anxiety	Drapeau et al. (2018)
	Mouse	Shank3B KO	ASD	Exhibited repetitive grooming	Deficits in reciprocal social interactions	Impaired learning and memory	Reduced open field activity, deficits in sensory responses, show anxiety-related behaviors	Dhamne et al. (2017)
	Mouse	Shank3B KO	ASD	Not mentioned	Not mentioned	Show learning deficits	Enhanced pitch discrimination	Rendall et al. (2019)
	Mouse	Shank3B KO	ASD	Increased repetitive grooming	Normal social communication, impaired social interactions	Not mentioned	No anxiety-like behaviors	Balaan et al. (2019)
	Mouse	Exon 8 (Q321R)	ASD	Increased repetitive grooming	Normal social interaction and social communication	Normal learning and memory	Normal locomotor activity, increased anxiety-like behavior, decreased seizure susceptibility	Yoo et al. (2019a, b)
	Rat	Exon 6 (Δex6)	PMS	Not mentioned	Not mentioned	Deficits in long-term social recognition memory	No anxiety-like behaviors, deficits in attention	Harony-Nicolas et al. (2017)
	Rat	Exon 6 (Δex6)	PMS	Not mentioned	Not mentioned	Not mentioned	Exhibit degraded cortical responses to sound	Engineer et al. (2018)
	Rat	Shank3 knock-out	ASD/PMS	Not mentioned	Behavioral responses to pro-social 50-kHz USV are reduced, deficits in a number of key juvenile social play interactions	Not mentioned	Norman motor activity in a novel open field arena, reduced ultrasonic vocalizations	Berg et al. (2018)
	Macaques	Exon 21 (lnsG3680)	ASD	Increased repetitive behaviours	Social interaction deficits	Impaired spatial learning	Motor deficits	Zhou et al. (2019)

ASD autism spectrum disorder, PMS Phelan–McDermid Syndrome, SCZ schizophrenia, PPI pre-pulse inhibition, BPD bipolar disorder

Animal Models with Shank2 Mutations

Studies have been carried on a few Shank2 mutant mouse lines (Table 2). One group of investigators developed Shank2 exons 6 and 7 deletion mice (Shank2 Δex6-7) (Won et al. 2012; Lim et al. 2017). The animals exhibit increased self-grooming, increased anxiety-like behavior and locomotor activity, reduced social interaction by USV, impaired spatial learning and memory. Another group generated Shank2 mutant mice with the exon 7 deleted (Shank2 Δex7), the mutant mice display social deficits and impaired spatial learning and memory (Lim et al. 2017). Moreover, Lim et al. demonstrated that enhancing inhibitory synaptic function can reverse spatial memory deficits in Shank2 mutant mice (Lim et al. 2017). In 2010, Berkel et al. generated shank2 mutant mice with a 120 kb deletion selectively disrupting exon 7, and found that the mutation did not result in autism-like behavior nor intellectual disability (Berkel et al. 2010). In addition, Shank2 Δe24−/− mice are anhedonic, display significantly increased locomotor activity, abnormal reward-seeking behavior, have perturbations in circadian rhythms, and show deficits in social and cognitive behaviors. The major behavioral features in these mice are considered to be reminiscent of bipolar disorder (Pappas et al. 2017).

Many molecular changes related to synapses have been reported in Shank2 mutant mice. InΔex6-7 mice, the protein level of triheteromeric subunit GluN1 of the N-methyl-d-aspartate (NMDA) glutamate receptor is increased in the hippocampus, while the level of GluA1 subunit of amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptors (AMPARs) is reduced (Won et al. 2012). In shank2 Δex7 mice, synaptic proteins of the glutamate receptor sbunits, GluN1, GluN2A and GluA2, in striatum, and GluN1 and GluN2B in hippocampus, are increased (Schmeisser et al. 2012). The Shank2 Δex6-7 mice display impaired basal tactile perception, acute pain response as well as decreased sensitivity to chronic pain and inflammatory pain (Ko et al. 2016). Reduced spine density in hippocampal CA1 is also observed in Shank2Δex7 mice. The structure of PSD is not apparently changed in the latter two Shank2mutant mouse lines.

A series of studies have demonstrated that specific cell type deletions of Shank2 in mice can result in diverse synaptic and behavioral abnormalities. Male mice lacking Shank2 in excitatory neurons (CaMKII-Cre;Shank2^fl/fl) exhibit social interaction deficits and mild social communication deficits, hyperactivity, and anxiety-like behaviors. Notably, male mice lacking Shank2 in GABAergic inhibitory neurons (Viaat-Cre;shank2^fl/fl) also show social repetitive self-grooming, communication deficits, and mild hyperactivity. These behavioral changes were associated with distinct changes of hippocampal and striatal neural circuits in the two mouse lines (Kim et al. 2018). Mice with Shank2 deletion in parvalbumin neurons exhibit moderate hyperactivity, enhanced self-grooming and suppressed seizure susceptibility, whereas the social behaviors, learning and memory and even anxiety level in the animals appear normal (Seungjoon et al. 2018). In addition, mice with Shank2-deficient in cerebellar Purkinje cells display abnormal task-specific repetitive behavior, impairments in motor learning and ASD-like social impairment (Peter et al. 2016).

A recent study found a novel Shank2 transcriptional variant—exon 4′, located between exons 4 and 5 (Lee et al. 2020). Exon 4′ appears to be a modifying domain participating in genetic compensation for ASD patients, and may bring out phenotypic heterogeneity in Shank2 Δex6-7 mice model (Lee et al. 2020). Wegener et al. also demonstrated genetic interactions in the Shank2 mutant mice, by comparing Shank2 Δex7 mice with Shank2 Δex6-7 mice (Wegener et al. 2018).

Animal Models with Shank3 Mutations and Knockout (KO)

Multiple lines of Shank3 mutant mice and, more recently, rats and macaques that carry global, conditional and point mutations or knockout (KO) in shank3, have been generated (Berg et al. 2018; Harony-Nicolas et al. 2017; Zhou et al. 2019) (Table 2). Molecular and behavioral characterizations have provided much information about normal and disease-related functions of Shank3. Overall, these animals display diverse synaptic, neuronal, brain connectivity and behavioral abnormalities. Studies on these animals have extended substantial understanding into how Shank3 mutations lead to various phenotypic abnormalities (Drapeau et al. 2018; Schmeisser et al. 2012; Peca et al. 2011; Bozdagi et al. 2010; Wang et al. 2011, 2017; Kouser et al. 2013; Lee et al. 2015; Jaramillo et al. 2015; Wang et al. 2016a, b; Copping et al. 2017; Guo et al. 2019; Yoo et al. 2018, 2019a, b; Dhamne et al. 2017; Kabitzke et al. 2018; Vicidomini et al. 2016; Wang et al. 2019a, b, c; Qin et al. 2018; Wang et al. 2019a, b, c).

The Shank3 mutation mouse models include those with deletions of exons 4-9 (Shank3Δex4-9, Shank3Δex4-9B and Shank3Δex4-9J), exons 4-7 (Shank3 Δex4-7), exons 12-13 (Δex12-13), exons 13-16 (Shank3Δex13-16), exons 14-16 (Shank3Δex14-16), exon 6 (Shank3 Δex6), exon 8 (Shank3 Δex8), exon 9 (Shank3 Δex9), exon 11 (Shank3 Δex11), exon 21 (Shank3 Δex21), exons 4-22 (Shank3 Δex4-22) and so on (Drapeau et al. 2018; Schmeisser et al. 2012; Berg et al. 2018; Peca et al. 2011; Bozdagi et al. 2010; Wang et al. 2011, 2017; Kouser et al. 2013; Lee et al. 2015; Jaramillo et al. 2015; Wang et al. 2016a, b; Copping et al. 2017; Guo et al. 2019; Yoo et al. 2018, 2019a, b; Dhamne et al. 2017; Kabitzke et al. 2018; Vicidomini et al. 2016; Engineer et al. 2018). Repetitive behaviors and increased self-grooming are observed in all Shank3 mutant lines, most of which also show social interaction deficits. Anxiety-like behaviors and abnormal locomotor activity, and learning and memory deficits, are commonly found in these Shank3 mutant animals. In behavioral phenotyping experiments, self-injurious grooming causing skin lesions is observed in some Shank3 hile normal social behaviors are found in other Shank3 mutant lines (Δex4-7, Δex13-16, Δex9, glutamatergic exons 14-16, Δex21, Δex4-22 and Shank3BKO) (Drapeau et al. 2018; Peca et al. 2011; Kouser et al. 2013; Lee et al. 2015; Wang et al. 2016a, b; Copping et al. 2017; Yoo et al. 2018, 2019a, b; Chantell et al. 2018; Yoo et al. 2019a, b). Normal learning and memory are observed in the Shank3 (Δex8) mice (Yoo et al. 2019a, b), while no anxiety-like behaviors are found in the Shank3 (Δex4-7, Δex9, Δex21 and Shank33B KO) mouse (Peca et al. 2011; Kouser et al. 2013; Lee et al. 2015; Chantell et al. 2018) and Δex6 rat (Berg et al. 2018). The locomotor activity appears normal in the Shank3 (Δex4-9, Δex8, Δex13-16 and glutamatergic exons 14-16) mice (Jaramillo et al. 2011; Copping et al. 2017; Yoo et al. 2018, 2019a, b) and in the Shank3 KO rat (Berg et al. 2018).

The Shank3 Δex4-22 mice present normal levels of social interest, but they persist in unsuccessful efforts to engage the social partner and exhibit non-social behaviors that include repetitive self-grooming behavior during social encounter event (Wang et al. 2016a, b). Although the Shank3 Δex9 mice do not show autistic-like behavior, they show mildly impaired spatial memory and increased rearing in a novel environment (Lee et al. 2015). In 2013, Han et al. reported that Shank3 transgenic mice with an N-terminal EGFP-tag into this gene exhibited manic-like behaviors, which appears to recapitulate the phenotype of human SHANK3 duplication (Han et al. 2013). In addition, Zhou et al. generated two Shank3 mutant mouse lines (lnsG3680 and R1117X mutation) (Yang et al. 2016). The mice with the ASD-linked lnsG3680 mutation exhibit stronger repetitive/compulsive grooming behavior. The mice with the SCZ-linked R1117X mutation manifest stronger allogrooming and social dominance-like behavior.

Findings from the animal models support that Shank3 mutations can affect the structure and function of excitatory synapses. The loss of Shank3 leads to an increase in metabotropic glutamate receptor 5 (mGluR5) in both synaptosome and postsynaptic density-enriched fractions in the hippocampus (Kouser et al. 2013). Co-localization studies show that mGluR5 and PSD-95 are increased in the striatum in Shank3Δex4-22 mice (Wang et al. 2016a, b). GluA2, GluN2A, and GluN2B, SAPAP3/GKAP3, Homer1, and PSD-93 are reduced in Shank3Δex13-16 mice (Peca et al. 2011). Wang et al. reported that Shank3 mutation preferentially affects glutamatergic synaptic transmission in the striatopallidal medium spiny neurons (MSNs), while enhancing this transmission via a Gq-coupled human M3 muscarinic receptor (hM3Dq) can rescue the repetitive grooming behavior (Wang et al. 2017). In 2017, Jaramillo et al. generated Shank3 mutant mice with a transcriptional stop cassette prior to exon 13 (Shank3E13). These mice showed impaired hippocampal LTP and striatal glutamatergic synaptic transmission, displayed increased repetitive grooming and decreased rearing, spatial memory deficits and impaired social interaction (Jaramillo et al. 2017). In addition, conditional knockout of Shank3 in the anterior cingulate cortex (ACC) resulted in excitatory synaptic dysfunction. The animals exhibited apparent social interaction deficits, which can be reversed by restoration of Shank3 expression in the ACC, or systemic administration with an α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor-positive modulator (Guo et al. 2019).

Lately, whole genomic and neuronal subtype specific Shank3 deletion mouse models are generated (Drapeau et al. 2018; Yoo et al. 2018, 2019a, b). The synaptic and behavioral deficits seen in global Shank3^Δ14–16, GABAergic Shank3^Δ14–16 and glutamatergic Shank3^Δ14–16 deletion mice are characterized in detail. For instance, neuronal excitability is abnormally increased in layer 2/3 pyramidal neurons in the medial prefrontal cortex (mPFC) in mouse with a glutamatergic Shank3 and global Shank3 deletion models. Direct social interaction and repetitive self-grooming are abnormally increased in the above three lines. One study also examined the changes of adult genetic reversal to rescue phenotypes in a Shank3 mutant mice, indicating that re-expression of Shank3 in adult mice leads to improvement in synaptic structure and function in the striatum, along with rescue of behavioral abnormalities including social interaction deficit and repetitive grooming, whereas anxiety and motor coordination deficit could not be recovered (Mei et al. 2016). Another study showed that early restoration of Shank3 expression in the knockout mice could ameliorate the ASD-like behavioral phenotypes (Jaramillo et al. 2020).

In the past, Shank genes editing has been mainly focused on rodents, but recently Zhou et al. created genetically engineered non-human primate model for the first time. Primate models of psychiatric disorders are expected to better recapitulate the behavioral phenotypes of autism spectrum and other disorders in humans, and may be more suited for the evaluation of therapeutics. The cynomolgus macaques with Shank3 mutation showed increased repetitive behaviour, motor deficits and sleep disturbances as well as social and learning impairments, resembling the clinical features of human autism (Zhou et al. 2019).

We attempt to summarize the molecular and behavioral phenotypes of shanks mutant animals, with the major features from different models included in Table 2. Overall, all Shank3 mutant animals manifest repetitive and increased self-grooming behaviors, most of them showed abnormal social communication and interaction, impaired learning and memory, reduced locomotion activity and increased anxiety-like behavior. However, it should be noted that different or different extents of behavioral deficits can exist in Shank3 mutant lines, even in the same Shank3 mutants as shown in different studies. Therefore, much additional work is needed to understand the role of Shank proteins in context of the involvement of neural pathways and circuits.

Conclusions and Future Perspectives

SHANKs genes can generate multiple protein isoforms that are distinctively expressed in the brain dependent on developmental stages, cell types and anatomical locations. Growing evidence from genome association studies strongly suggests a connection between SHANK gene alterations and some neuropsychiatric disorders, especially ASD and SCZ, with mutations in SHANK2 and SHANK3 may be a potential monogenic cause for ASD. More researches on SHANK association with TLE are expected to uncover some of the neurobiological role of the proteins in modulating excitatory synaptic circuitries involved in epileptogenesis. In this regard, it is worth noting that most neuropsychiatric disorders are caused by the absence or haploinsufficiency of SHANK genes, whereas SHANK overexpression is observed in temporal lobe epilepsy.

Animals carrying different shank mutations show distinct and shared molecular and behavioral characteristics. All mutants appear to alter excitatory synapses and their neurotransmission. Various mutations of the same gene may cause synaptic and circuitry defects resulting in different neurobehavioral phenotypes and clinical features. Again, more studies are needed to understand underlying neurocircuitry changes. Albeit numerous questions remain unresolved, studies on SHANK family in human and animal models have provided novel molecular bases for neuropsychiatric disorders.

Abbreviations

PSD

Postsynaptic density

ASD

Autism spectrum disorders

PMS

Phelan–McDermid Syndrome

AD

Alzheimer’s disease

SCZ

Schizophrenia

BPD

Bipolar disorders

ID

Intellectual disability

DD

Developmental delay

GWAS

Genome-Wide Association Studies

SNP

Single nucleotide polymorphism

NMDA

N-Methyl-d-aspartate

ANK

Ankyrin repeat domain

PDZ

Postsynaptic density protein 95-discs large homologue 1-zonula occludens 1 domain

PRO

Proline-rich region

SAM

Sterile alpha motif

SH3

SRC homology 3 domain superfamily

BDNF

Brain-derived neurotrophic factor

AMPA

Alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid

DLGAP1, 2, 3

Discs large associated protein 1, 2, 3

SLC6A3

Solute carrier family 6 (neurotransmitter transporter, dopamine), member 3

Author Contributions

L-W and D-L drafted the manuscript; W-BX and B-XZ draw the figures; Z-HL designed the outline and revised the manuscript; X-XY proof-edited and finalized the manuscript; BX provided financial support. All of the authors have approved of the publication of the manuscript.

Funding

This work has been supported through funding from the National Natural Science Foundation, China (Grant Number 81671299), The Hunan Natural Science Foundation (Grant Number 2018JJ2648).

Data Availability

All data and material files are available upon request.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interests.

Ethical Approval

All procedures were followed in accordance with the Ethical Standards of the Responsible Committee on Animal and Human Experimentation (The Ethics Review Committee of Xiangya Hospital, Numbers 201603296 and 201603297) and with the Helsinki Declaration of 1964 and later versions.