SynGAP is a synaptic Ras GTPase-activating protein (GAP) with four C-terminal splice variants: α1, α2, β, and γ. Although studies have implicated SYNGAP1 in several cognitive disorders, it is not clear which SynGAP isoforms contribute to disease. Here, we demonstrate that SynGAP isoforms exhibit unique spatiotemporal expression patterns and play distinct roles in neuronal and synaptic development in mouse neurons. SynGAP-α1, which undergoes liquid-liquid phase separation with PSD-95, is highly enriched in synapses and is required for LTP. In contrast, SynGAP-β, which does not bind PSD-95 PDZ domains, is less synaptically targeted and promotes dendritic arborization. A mutation in SynGAP-α1 that disrupts phase separation and synaptic targeting abolishes its ability to regulate plasticity and instead causes it to drive dendritic development like SynGAP-β. These results demonstrate that distinct intrinsic biochemical properties of SynGAP isoforms determine their function, and individual isoforms may differentially contribute to the pathogenesis of SYNGAP1-related cognitive disorders.
SynGAP is a GTPase-activating protein (GAP) that is highly enriched in dendritic spines of excitatory neurons (Chen et al., 1998; Kim et al., 1998). SynGAP is a Ras- and Rap- GTPase activating protein that facilitates the hydrolysis of small G protein-bound GTP (active) to GDP (inactive), thus negatively regulating the activity of small G proteins (Carlisle et al., 2008; Chen et al., 1998; Pena et al., 2008; Rumbaugh et al., 2006). SynGAP is encoded by the SYNGAP1 gene and is alternatively spliced to generate 4 distinct C-terminal isoforms: SynGAP-α1, SynGAP-α2, SynGAP-β, and SynGAP-γ (Li et al., 2001; McMahon et al., 2012). The C-terminal domain of SynGAP-α1 contains a class I PDZ ligand sequence (QTRV) which binds MAGUK family proteins such as PSD-95 (Chen et al., 1998; Kim et al., 1998); (Grant and O’Dell, 2001). Heterozygous deletion of Syngap1 in rodents causes severe deficits in long-term potentiation (LTP) at synapses of hippocampal CA1 pyramidal neurons that are innervated by Schaffer collaterals (SC), as well as severe working memory deficits (Kim et al., 2003; Komiyama et al., 2002; Rumbaugh et al., 2006).
In humans, loss-of-function variants in SYNGAP1 have been associated with Intellectual Disability (ID), epilepsy, Autism Spectrum Disorders (ASDs), and Neurodevelopmental Disability (NDD). While there are hundreds of genetic risk factors for these disorders, the significantly elevated frequency and 100% penetrance of loss-of-function variants in SYNGAP1 as well as the range of brain disorders associated with SYNGAP1 pathogenicity make it unique (Berryer et al., 2013; Carvill et al., 2013; Hamdan et al., 2011; Hamdan et al., 2009; Satterstrom et al., 2020).
Many loss-of-function variants of the SYNGAP1 gene have been causally associated with ID, epilepsy, ASD, and other NDDs. In a UK study of 931 children with ID, SYNGAP1 was the 4th most highly prevalent NDD-associated gene, and SYNGAP1 variants accounted for ~0.75% of all NDD cases (Fitzgerald et al., 2015). Patients with SYNGAP1 haploinsufficiency have high rates of comorbid epilepsy, seizures, and acquired microcephaly (Berryer et al., 2013; Carvill et al., 2013; Cook, 2011; Hamdan et al., 2011; Hamdan et al., 2009; Parker et al., 2015; Rauch et al., 2012; Tan et al., 2016; Fitzgerald et al., 2015; Vissers et al., 2010; Vlaskamp et al., 2019; Writzl and Knegt, 2013). Mental Retardation, Autosomal Dominant 5 (MRD5) (OMIM #612621) is caused by mutations in SYNGAP1. MRD5 is characterized by moderate-to-severe intellectual disability with delayed psychomotor development apparent in the first years of life (Holder et al., 2019). Nearly all reported cases of SYNGAP1-related ID and ASD are de novo mutations within/near exons or splice sites of SYNGAP1 (Vlaskamp et al., 2019).
Some key pathophysiological symptoms of ID and ASD observed in SYNGAP1 patients have been recapitulated in constitutive Syngap1 hetereozygous (Syngap1+/-) mice (Clement et al., 2012). Syngap1 heterozygous mice exhibit learning deficits, hyperactivity, and epileptic seizures (Clement et al., 2012; Guo et al., 2009). Additionally, several MRD5-associated SYNGAP1 missense mutations also cause SynGAP protein instability (Berryer et al., 2013). These data strongly suggest that SYNGAP1 haploinsufficiency is pathogenic in SYNGAP1-associated ID and ASD. Thus, several lines of evidence in mice and humans support that SynGAP is a critical regulator of synaptic plasticity, development, and behavior.
We recently discovered that SynGAP-α1 is rapidly dispersed from dendritic spines during LTP, which allows for concomitant spine enlargement and accumulation of synaptic AMPARs (Araki et al., 2015). SynGAP-α1 dispersion from the dendritic spines releases the inhibition of synaptic RAS activity which is required for the expression of LTP (Harvey et al., 2008; Murakoshi and Yasuda, 2012; Walkup et al., 2016; Zhu et al., 2002). Additionally, SynGAP is the third mostly highly expressed protein in the postsynaptic density (PSD) and can undergo multivalent interactions with PSD-95 via liquid-liquid phase separation (LLPS), a process of forming highly concentrated condensates with liquid-like properties, which may contribute to the formation of the PSD complex (Zeng et al., 2016). LLPS in cells is a phenomenon in which biochemical reactants are spatially clustered and concentrated in the absence of a surrounding membrane, allowing for organelle-like function without the physical and energetic barriers posed by lipid bilayers (Shin and Brangwynne, 2017). Although SynGAP is an ideal candidate to provide the structural basis of PSD (Zeng et al., 2018; Zeng et al., 2016), the phase separation of SynGAP was extensively characterized only with SynGAP-α1. The degree to which the other SynGAP isoforms undergo activity-dependent dispersion and LLPS remains largely unknown, as does the functional significance of these isoforms.
Although SYNGAP1 haploinsufficiency likely affects the expression of all SynGAP isoforms, only the α1 isoform has been rigorously characterized to date. Only a few functional studies of non-α1 SynGAP isoforms have been conducted to probe how these isoforms regulate synaptic physiology and disease pathogenesis (Li et al., 2001; McMahon et al., 2012). In these overexpression studies, the various SynGAP isoforms have been shown to have differing – and even opposing – effects on synaptic transmission (McMahon et al., 2012). However, as these were overexpression experiments, endogenous SynGAP was intact in this study, complicating interpretation of these results. It is currently unknown whether SYNGAP1-associated ID/ASD pathology is associated with select deficits of specific SynGAP isoforms that may underlie unique features of NDD.
Here, we report that SynGAP-α1 constitutes only 25–35% of total SynGAP protein in the brain, underscoring the importance of characterizing how the C-terminal SynGAP splice variants contribute to neuronal and synaptic development that are associated with the pathogenesis of SYNGAP1 haploinsufficiency. In developing neurons, the various SynGAP isoforms display differences in neuroanatomical and subcellular expression. We report that SynGAP-β is expressed earlier in development than the other SynGAP isoforms, and functions specifically to promote dendritic arbor development. In contrast, SynGAP-α1 reaches peak expression later in development, and regulates the processes underlying synapse strengthening, including AMPAR insertion and dendritic spine enlargement. Our findings describe unique roles for select SynGAP isoforms in mediating different facets of neuronal function. Furthermore, we identify isoform-specific differences in biochemical interactions between SynGAP and PSD-95, and show how these differences are related to the functional mode of each isoform, regulating either synaptic plasticity or dendritic structure. These results suggest that individual SynGAP isoforms mediate distinct, specialized regulation of neuronal and synaptic development and will inform potential therapeutic strategies for treating SYNGAP1-related disorders.
SynGAP isoforms have distinct and overlapping expression profiles during brain development
SYNGAP1 is alternatively spliced at several sites to include exons 18, 19, or 20 to generate four unique C-terminal isoforms: SynGAP-α1, SynGAP-α2, SynGAP-β, and SynGAP-γ (Figure 1A,B). SynGAP-α1 and SynGAP-α2 isoforms skip exon 19 and are produced by selective splicing of exon 20, whereby SynGAP-α1 contains a PDZ ligand (-QTRV) and SynGAP-α2 lacks this domain. The SynGAP-β isoform includes a frameshifting extension of exon 18 leading to early termination, which generates a SynGAP protein product with a partially truncated coiled-coil domain. The SynGAP-γ isoform includes exon 19, which contains a short coding sequence followed by a STOP codon (-LLIR*).
SynGAP isoforms are differentially expressed during brain development.
(A) Schematic of SYNGAP1 splicing at the C-terminus. SYNGAP1 is alternatively spliced within exons 18–20 to generate four unique C-terminal isoforms designated as α1, α2, β, and γ. (B) C-terminal amino-acid sequences of SynGAP isoforms encoding select protein domains. Coil-Coil domain (yellow) and PDZ ligand-binding domain (blue). Targeted epitopes of isoform-specific SynGAP antibodies (JH2469, JH7265, JH7206, and JH7366) are indicated as dotted lines. (C) Specificity of SynGAP isoform-specific antibodies. Immunoblots of SynGAP isoform expression in lysates prepared from HEK 293 T cells expressing individual GFP-tagged SynGAP isoforms and lysates prepared from brain tissue obtained from WT and Syngap1 +/- mice were shown. Quantification of relative SynGAP isoform levels with respect to total SynGAP expression measured from immunoblot were shown in Figure 1—figure supplement 1A. Two-way ANOVA followed by Tukey’s post hoc test (Isoform F(4,30)=1.900; p=0.13, Genotype F(1,30)=451.2; p