Synaptic vesicle recycling in preclinical models of intellectual disability, autism spectrum disorder and epilepsy

Abstract

The development of the central nervous system is dysregulated in neurodevelopmental disorders such as intellectual disability, autism spectrum disorder, and epilepsy. These three disorders have different clinical features, yet there is high comorbidity between them. They can be difficult to study due to their highly complex aetiologies, however there are various monogenic diseases that can cause all of them, including SYNGAP1 haploinsufficiency where the synaptic guanosine triphosphatase (GTPase)-activating protein (SYNGAP) protein levels are highly reduced; Fragile X syndrome where the fragile X mental retardation protein (FMRP) is no longer translated; and DNM1 epileptic encephalopathy where mutations in the Dynamin1 gene alter the protein function. These monogenic conditions are synaptopathies as the proteins affected play important roles in synapse stability and neurotransmission.

Because of the high comorbidity between these disorders, it is hypothesised that there may be a common mechanism underlying them. We hypothesise that a deficit in presynaptic vesicle recycling may be part of a common mechanism underlying intellectual disability, autism spectrum disorder, and epilepsy especially in SYNGAP1 haploinsufficiency, Fragile X syndrome, and DNM1 epileptic encephalopathy.

Using various fluorescent presynaptic activity reporters including synaptic pHluorins, tetramethylrhodamine dextran and calcium dyes to compare presynaptic activity in in vitro models of these monogenic conditions, we found differences in synaptic vesicle (SV) endocytosis in the genetically altered conditions compared to wildtype controls.

We observed various SV endocytosis defects in clathrin-mediated endocytosis (CME) or activity-dependent bulk endocytosis (ADBE) in our models. We observed enhanced CME in SynGAP1 KO mouse hippocampal neurons. This enhanced SV endocytosis was accompanied by decreased SV cargo on the plasma membrane. Rat SynGAP1 KO hippocampal neurons did not display enhanced SV endocytosis, nor did neurons with the GTPase-activating (GAP) domain of SynGAP deleted. This was perhaps due to the altered time course of development between these rodent species.

In mouse and rat models of Fragile X syndrome, CME was not altered compared to wildtype controls. However, in a rat model, we observed fewer nerve terminals undergoing ADBE which is the dominant SV endocytosis mode during elevated neuronal activity. De novo epileptic encephalopathy-associated mutations in DNM1 had differential effects on SV recycling through both CME and ADBE. Mouse hippocampal neurons overexpressing Dyn1R237W, Dyn1I289F and Dyn1H396D all showed less CME compared to overexpression of Dyn1WT. Moreover, fewer nerve terminals overexpressing Dyn1H396D were found to undergo ADBE. We also found that a large-conductance potassium (BK) channel opener can accelerate clathrin-mediated endocytosis and thus may be able to rescue the impaired SV endocytosis caused by these mutants.

Although there is not yet a common underlying pathway at the presynaptic level between these conditions, SV recycling dysfunction is present across all of these models. Furthermore, we propose an axis of pathophysiology model where optimal SV endocytosis is required for optimised neural performance. We propose that either decreased or increased SV endocytosis can lead to the synaptic dysfunction observed in these models.