Determinants of neuropeptide secretion: From nuclear signal to dense-core vesicle fusion sites

Neuropeptides and neuromodulators are by far the largest group of chemical signals in the brain. In comparison to neurotransmitter release from SVs, their secretory pathway is poorly understood, and many fundamental aspects are still unknown. Such aspects include a general understanding of the DCV population in neurons, their fusion efficiency, and the relation between stimulation frequency and DCV fusion. In particular, the organizing principles for neuropeptide release from DCVs remain elusive. Furthermore, the molecular mechanisms that promote efficient DCV fusion during long-term plasticity are still unknown. Therefore, in this thesis, we aimed to study the determinants underlying neuromodulators exocytosis in mammalian CNS neurons. In chapter 2, using a combination of confocal microscopy and super-resolution microscopy, we characterized the distribution of DCVs in mammalian neurons. We observed that the number of DCVs per μm neurite was consistent between different culture systems but slightly increased in labeled axons in ex vivo sections of the mouse cortex. Ca2+ influx triggered the fusion of 1–6% of the total number of DCV, with approximately 80% of DCV fusion events occurring at the axon. Our findings in chapter 2 led to the question of how different patterns of stimulation affect the properties of single DCV fusion. Chapter 3, starting from pioneering works in the PNS (Andersson et al., 1982; Lundberg et al., 1986), compared different stimulation patterns, ranging from a single 4 Hz train to a repetitive 50 Hz burst, TBS stimulation, and short 200 Hz trains. These data indicate that DCV fusion is highly dependent on the frequency of AP delivery, providing the first characterization of the release properties of single DCV in dissociated CNS neurons. In chapter 4, based on previous studies in D. melanogaster (Shakiryanova et al., 2007; Wong et al., 2009) and C. elegans (Hoover et al., 2014), we investigated whether CaMK2 promotes neuromodulator exocytosis, in particular BDNF. We confirmed that in the absence of CaMK2, BDNF secretion is reduced. However, the trafficking, accumulation, or fusion efficiency of DCVs was not altered in the absence of both αCaMK2 and β. Instead, the reduced secretion of neuromodulators was entirely due to a decreased expression of DCV cargo. Reduced neuromodulator expression was restored by active βCaMK2 or by promoting CREB phosphorylation. Thus, CaMK2 is a crucial component of a feedback loop that upregulates the expression of DCV-resident neuromodulators dependent on their secretion. One of the main aspects of neuronal communication is the presence of specialized active zones to ensure fast synaptic transmission (Sabatini and Regehr, 1996), yet the determinants of DCV fusion sites are still elusive. In chapter 5, we investigate these determinants. Starting from evidence from vacuolar fusion in the yeast S. cerevisiae (Alpadi et al., 2013; Peters et al., 2004), we show that dynamin proteins are significant regulators of DCV exocytosis. Via acute pharmacological inhibition and genetic perturbation of dynamin, we demonstrate that dynamin is crucial for DCV exocytosis during prolonged stimulations, in addition to the well-established function in SV endocytosis. Moreover, in the absence of dynamin, a significant proportion of syntaxin1 was present in inaccessible clusters that could be restored by αSNAP overexpression. In addition to its fission role in endocytosis, we propose that dynamin promotes the recovery of DCV fusion sites downstream of αSNAP mediated SNARE disassembly.