Developmental trajectory of synaptic vesicle recycling in wild type and neurodevelopmental disorder mqodels

Abstract

The presynapse performs an essential role in brain communication via the activity-dependent release of neurotransmitters. This is made possible by exocytosis of synaptic vesicles (SV) that package the neurotransmitters, and these SVs are recycled via endocytosis. However, the sequence of events through which a presynapse acquires functionality is relatively poorly understood, which is surprising since mutations in genes essential for its operation are heavily implicated in neurodevelopmental disorders (NDDs). We addressed this gap in knowledge by determining the developmental trajectory of SV recycling pathways in wild type (WT) murine hippocampal neurons from primary cultures. Using the genetically-encoded calcium indicator (GECI) GCaMP6f, we revealed that the majority of nerve terminals displayed activity-dependent calcium influx from 3 days in vitro (DIV). Synaptophysin-pHluorin (sypHy) imaging indicated that SVs are recycled via evoked exocytosis and endocytosis from the first week in vitro as well, although the number of responsive nerve terminals continued to increase until the second week in vitro. However, the most intriguing discovery was that activity-dependent bulk endocytosis (ADBE) was only observed from DIV 14 onwards. The delayed acquisition of ADBE was reported through a tetramethylrhodamine (TMR)-dextran uptake assay and was corroborated by examination of horse radish peroxidase (HRP) uptake via electron microscopy. This series of optical assays suggests that ADBE is acquired as an extra step of synaptic development. After determining the developmental trajectory of SV recycling in WT neurons, I then investigated whether this was altered in NDD models. First, I tested SV recycling in Fmr1 knockout (KO) neurons which model Fragile X Syndrome, one of the most common monogenic NDDs that shows high comorbidity to autism spectrum disorder (ASD). Fmr1 KO neurons displayed reduced ADBE at DIV 14 compared to neurons from their littermate controls, as reported previously. Intriguingly, however, this deficit recovered at a later timepoint (DIV 21), indicating that the phenotype was not because of a mechanistic failure of ADBE due to the absence of FMRP. The mechanism behind this delay and the compensation will be the next key question to address. Through GST pull-down assay and mass spectrometry, I have identified Dynamin-1, a GTPase that constricts and fissions SVs during endocytosis, as a potential novel interacting partner of FMRP, opening up a new avenue for studying SV recycling in Fmr1 KO neurons. SV recycling was also investigated in Dual Specificity Tyrosine Phosphorylation-Regulated Kinase 1A (Dyrk1A) heterozygous (HET) neurons, which model DYRK1A syndrome, another monogenic NDD that is highly associated with ASD, intellectual disability and epilepsy. DYRK1A is located in the Down syndrome critical region on human chromosome (HSA) 21, meaning that it impacts neurodevelopment in a dosage-dependent manner. Similar to Fmr1 KO neurons, Dyrk1a HET neurons displayed reduced ADBE in comparison to WT neurons in both TMR-dextran and HRP uptake assay. However, in addition to depressed ADBE, they also showed reduced evoked exocytosis as well as SV pool size. This reduction in SV exocytosis was not due to altered activity-dependent calcium influx, since Dyrk1a HET neurons did not show any significant difference with WT neurons in calcium imaging using GCaMP6f. This finding highlights the presynaptic downscaling in Dyrk1a HET neurons downstream of activity-dependent calcium influx. The mechanism behind such phenotype could be further investigated based on the proteomics and phospho-proteomics data which highlighted the disrupted phosphorylation of SV recycling proteins. To summarise, it appears that Dyrk1a HET and Fmr1 KO neurons may share depression of ADBE as a potential convergent phenotype but through separate pathology.