The ALS-Related SynGAP1 Pathogenic Variant Causes Dendritic Spine Loss: Potential Mechanisms of Early-Stage ALS Progression

Amyotrophic lateral sclerosis (ALS) is a rare and fatal neurodegenerative disease. The main clinical manifestations of ALS are progressive muscle weakness and atrophy. Patients usually die of respiratory failure because of respiratory muscle paralysis. The typical pathologic changes underlying the symptomatic stage are degeneration of upper and lower motor neurons (MNs; Brown and Al-Chalabi, 2017). Because the pathologic changes at this stage are difficult to reverse, it's crucial to identify pathologic changes occurring in the early stages of ALS to comprehensively understand ALS pathogenesis and possibly to develop early-stage intervention strategies.

Previous studies have shown that synaptic structural abnormalities and synaptic loss are tightly associated with early-stage ALS pathogenesis. (Sasaki and Maruyama, 1994). Further studies in animal models of ALS suggest that dendritic spine loss is an important pathologic phenotype that occurs before behavioral symptoms emerge (Fogarty et al., 2015, 2016). Because dendritic spines are the site of most excitatory input to motor neurons (Rekling et al., 2000), abnormalities in dendritic spines may reduce motor neuron excitability. Consistent with this, in vivo studies in mouse models of ALS found that motor neurons exhibited diminished excitability at the presymptomatic stage (Baczyk et al., 2020), corresponding to the structural abnormalities in glutamatergic synapses. Therefore, an in-depth investigation of the molecular events underlying the production of abnormal dendritic spines is of great importance for understanding the early stages of ALS pathogenesis.

Dendritic spine morphogenesis is regulated in part by synaptic Ras-GTPase activating protein 1 (SynGAP1). SynGAP1 is a switch for the small G-protein Ras, which blocks Ras signal transduction by activating the GTPase activity of Ras. SynGAP1 has four C-terminal isoforms—α1, α2, β, γ—which differ in biochemical characteristics and functions (see below; Gamache et al., 2020). Importantly, previous studies have found a link between SynGAP1 and ALS; the canonical ALS-associated protein fused in sarcoma (FUS) can bind to the 3'UTR of SynGAP1 mRNA and enhance the mRNA stability of the α2 isoform, which is required to maintain dendritic spine morphology (Yokoi et al., 2017). According to the above discussion, aberrant SynGAP1 expression profiles may be associated with dendritic spine abnormalities in ALS cases.

This reasoning led researchers to ask whether SynGAP1 itself harbored ALS-associated pathogenic mutations. If so, what pathologic phenotype does this mutation lead to in motor neurons? And what is the detailed mechanism? In a recent issue of The Journal of Neuroscience, Yokoi et al. (2022) identified an ALS-related SYNGAP1 variant. They investigated the molecular mechanism by which this variant causes neuronal dendritic spine loss.

To identify SynGAP1 mutations related to ALS, Yokoi et al. (2022) analyzed whole-exome sequencing data from a cohort of sporadic ALS patients in Japan. They found that seven patients had a variant in the 3'UTR of SYNGAP1 mRNA, rs149438267 (G→T). To investigate whether this variant contributes to ALS-related pathologic phenotypes in motor neurons, they used choline acetyltransferase–positive MNs derived from human-induced pluripotent stem cells (hiPSCs). The homozygous rs149438267 mutation was introduced into iPSCs using CRISPR–Cas9 technology, and then iPSCs were induced to differentiate into MNs. Compared with wild-type MNs, MNs carrying the homozygous rs149438267 variant (hereafter referred to as mutant MNs) had significantly fewer dendritic spines, which is a classic pathologic change in the presymptomatic stage of ALS.

Because the rs149438267 variant occurs in the 3'UTR region, which can interact with various RNA-binding proteins for post-transcriptional regulation, this variant may affect the expression profile of SynGAP1 mRNAs and finally cause neuronal pathologic phenotypes. So Yokoi et al. (2022) measured the expression levels of SynGAP1 in mutant MNs and wild-type MNs. Although the variant didn't affect the total amount of SynGAP1 protein in MNs, it increased the expression level of the SynGAP1-α1 protein. Moreover, overexpression of the SynGAP1-α1 resulted in decreased spine density in MNs.

Because there are no suitable antibodies to investigate other SynGAP1 isoforms with Western blots, Yokoi et al. (2022) analyzed changes in expression of other SynGAP1 isoforms using fragment analysis. This revealed abnormally low levels of SynGAP1-γ in mutant MNs expressing the rs149438267 variant. The researchers further explored the effect of reduced expression of SynGAP1-γ by restoring SynGAP1-γ expression levels in mutant MNs. This rescued spine loss, suggesting that SynGAP1-γ may have a role in promoting spine formation, and the reduction of SynGAP1-γ expression contributed to spine loss. In summary, the rs149438267 mutation may cause spine loss in motor neurons by upregulating SynGAP1-α1 expression and downregulating SynGAP1-γ expression.

Finally, Yokoi et al. (2022) sought to elucidate the mechanism by which the rs149438267 variant alters the relative abundance of SynGAP1 C-terminal isoforms in MNs. RNA pull-down assays and liquid chromatography with tandem mass spectrometry analysis revealed increased binding ability of FUS and heterogeneous nuclear ribonucleoprotein K (HNRNPK) to the 3'UTR of SynGAP1 rs149438267 variant mRNA. Like FUS, HNRNPK can regulate mRNA stability (Thiele et al., 2004; Natarajan et al., 2022). However, in the hiPSC-derived MNs, Yokoi et al. (2022) found that only HNRNPK significantly affected the expression of each isoform of SynGAP1. The binding motifs of HNRNPK in the 3'UTR of SynGAP1 mRNA partially overlapped with the exon splicing enhancer sequence. Therefore, excessive recruitment of HNRNPK may interfere with the alternative splicing of SynGAP1 mRNA. For direct evidence, Yokoi et al. (2022) treated mutant MNs with antisense oligonucleotides targeting the HNRNPK binding motif and found that this restored normal SynGAP1-γ expression levels and rescued spine loss.

Overall, the main finding of Yokoi et al. (2022) was that the rs149438267 variant caused abnormal changes in the expression level of SynGAP1-α1 and SynGAP1-γ, which ultimately resulted in dendritic spine loss in MNs. As mentioned previously, SynGAP1 can generate four major C-terminal isoforms—α1, α2, β, and γ—through the splicing of exons 18, 19, and 20 at the 3′ terminal of mRNA. Most previous studies have focused on the α1 isoform. SynGAP1-α1 can undergo liquid–liquid phase separation (LLPS) with the scaffold protein PSD-95 via multivalent interactions and form the dynamic liquid-like postsynaptic dense (PSD). As a result, the SynGAP1-α1 is highly enriched in the PSD. Activation of NMDA receptors at excitatory postsynaptic membranes leads to phosphorylation of SynGAP1-α1 by CaMKII. This weakens the binding between SynGAP1-α1 and PSD-95, causing SynGAP1-α1 to disperse into the aqueous cytoplasmic phase (Zeng et al., 2016). This relieves the inhibition of Ras by SynGAP1-α1, leading to activation of the Ras pathway, which triggers two key events in the dendritic spine, (1) the rapid transport of AMPA receptors (AMPARs) to the spinous membrane and (2) actin polymerization leading to dendrite enlargement. Together, these events mediate to long-term synaptic potentiation (LTP; Araki et al., 2015). In contrast, α2, β, and γ isoforms have weak LLPS ability, their abundance in the PSD region is lower than that of SynGAP1-α1, and they have no significant role in regulating LTP (Araki et al., 2020). Given these findings and those of other studies showing that overexpression of SynGAP1-α1 can attenuate AMPAR-mediated miniature EPSCs (McMahon et al., 2012), one can reasonably speculate that excessive SynGAP1-α1 can interfere with the phase-transition behavior of the proteins in the PSD region on postsynaptic membrane excitation, affecting the activity of the Ras pathway, altering the electrophysiological characteristics of glutamatergic synapses, disrupting the normal excitability of motor neurons, and ultimately leading to neuronal damage.

The main pathologic change in MNs in the study by Yokoi et al. (2022) was neuronal spine loss. The morphogenesis of dendritic spines is significantly controlled by the GTPase signaling pathway, and enhanced Ras signaling can increase spine density (Gärtner et al., 2005). According to the discussion in the previous paragraph, abnormally high expression of SynGAP1-α1 may disrupt Ras signaling, which may relate to the loss of dendritic spines observed by Yokoi et al. (2022). However, although the effect of SynGAP1-γ on Ras activity was weak, it had a relatively large effect on the GTPase activity of Rap, another small G-protein (Araki et al., 2020). Intriguingly, activation of Rap GTPase activity (i.e., inhibiting Rap signaling) can increase spine density (Pak et al., 2001), so the potential interaction between SynGAP1-γ and Rap may be related to the promoting effect of SynGAP1 on spine growth.

In summary, Yokoi et al. (2022) identified an ALS-related SynGAP1 pathogenic variant and further revealed the detailed mechanism by which this variant leads to dendritic spine loss. Because the loss of dendritic spines is a canonical presymptomatic neuropathological change of ALS, this study may enrich our understanding of early-stage ALS pathogenesis. Simultaneously, this study refined our understanding of the biological functions of SynGAP1 C-terminal isoforms, as well as the regulatory mechanisms of dendritic spine morphogenesis.