The brain is a structure that has fascinated generations of scientists due to its complexity and its relevance in a plethora of biological processes. It is made up of a few hundred nerve cells (so-called neurons) in the worm C. elegans (Albertson and Thomson, 1976; White et al., 1986) and can contain up to several hundred billions of nerve cells like in African elephants (Herculano-Houzel et al., 2014). These highly specialized cells communicate with each other at structures termed synapses, where signals are transferred in a highly regulated way, and a majority of synapses communicate via chemicals called neurotransmitters (NTs). The typical chemical synapse can be subdivided into three functionally and morphologically distinct parts: (I) the signal-emitting presynaptic part including the active zone (AZ), where NTs are released from vesicular structures (synaptic vesicles, SVs) within the neuron either following an electrical stimulus (leading to an action potential, AP; AP-evoked release) or spontaneously (i.e. without a corresponding stimulus); (II) the synaptic cleft through which the NT molecules have to diffuse in order to be sensed by (III) the postsynaptic part, which harbors membrane receptors that specifically detect certain kinds of neurotransmitters and allow the signal to propagate electrically or via intracellular signaling cascades. This process of neurotransmission is the physiological basis of all behavior and movement found in animals. In the last few decades, intense research has elucidated many details of neurotransmission and their implications for neurological diseases. However, a considerable amount of questions is still open and unanswered. To contribute to an answer to some of these questions, the work conducted in the framework of this cumulative thesis comprises three interconnected subprojects.
The first subproject aimed to determine how synapses (at the neuromuscular junction formed by nerves terminating on muscles, NMJ) in the invertebrate model organism Drosophila melanogaster (commonly known as the fruit fly) engage isoforms of the protein Unc13 to regulate the coupling distance between SVs and the source of calcium influx, which triggers their release. Building a partial computational model of the synapse in conjunction with experimental data of my colleagues, I could substantiate the claim that two Drosophila isoforms of the protein localize at different distances from the calcium source and differentially ensure regulated release of SVs (Publication I: Böhme et al., Nature Neuroscience 2016). Using an experimental approach to visualize the release of single SVs at individual synapses and correlating it with protein levels at those synapses, I further contributed to prove that the Unc13A isoform constitutes the main molecular correlate of SV release sites (Publication II: Reddy-Alla et al., Neuron 2017). Parts of these results were then reviewed in Publication III: Böhme et al., FEBS Letters 2018, where I used the model built for Publication I and showed the influence of different coupling distances on synaptic short-term plasticity (which describes how the synapse responds to quickly succeeding stimuli). We further investigated the role of Unc13A in synaptic signaling maintenance on different timescales, where I showed the generally geometric and plastic patterning of Unc13A at single AZs using a computational approach and STED images acquired by a colleague (Publication IV: Böhme et al., Nature Communications 2019). Using findings from these publications, we then built an advanced stochastic model of synaptic SV release, specifically taking into account the broad distribution of coupling distances found at the AZ. To make our simulations agree with experimental data on short-term plasticity, we found it necessary to include a calcium dependent mechanism that rapidly regulates the number of readily releasable vesicles. These results are shown in Publication V: Kobbersmed et al., eLife 2020.
The goal of the second subproject was to investigate different forms of the cellular membrane signaling lipid diacylglycerol (DAG) and their interaction with a functionally downstream signaling molecule (protein kinase C, PKC), as it is not clear whether or how subtle differences in lipid structure influence kinetics and signaling properties. To this end, our collaborators generated DAGs exhibiting a chemical ‘cage’ that keeps them from translocating over cellular membranes. These caged compounds were then individually applied to the outside of cells and their biological function restored by acutely removing the cage through a UV laser flash. An intracellular, fluorescently tagged DAG binding protein (a domain of PKC) could then be used to indicate lipid dynamics and protein recruitment in the cell membrane over time. However, due to the temporal convolution of concurrent dynamic processes like trans-bilayer movement, sensor binding/unbinding and metabolism, the exact quantification of kinetic properties required a computational assay that we developed. Using this in silico model of lipid dynamics and signaling in a cellular environment, we could show vastly different kinetic properties and lipid-protein interactions only depending on relatively small structural differences. Our computer-aided quantification provided evidence of differential effector protein binding and lipid availability in different parts of the cell. The work and results are shown in Publication VI: Schuhmacher et al., PNAS 2020.
In the third subproject, we set out to determine how spontaneous and AP-evoked SV exocytosis are regulated and whether they are functionally overlapping or separate. Using a genetically encoded calcium indicator, which we expressed in the postsynapse of the Drosophila NMJ, we correlated the activity of individual AZs in both release modes with levels of presynaptic proteins, expanding on our findings from Publication II: Reddy-Alla et al., Neuron 2017. Furthermore, we pharmacologically investigated the involvement of different presynaptic voltage gated channels in spontaneous release, and determined the degree of overlap in postsynaptic sensing. This work showed that many presynaptic proteins (e.g. Unc13A) predict both SV release modes, while some show differential influence. We further showed that postsynaptic receptors generally detect NT released via both modes, and that presynaptic voltage-gated Ca2+ channels (VGCCs) are involved in the generation of spontaneous SV release. Lastly, we observed that both release modes draw on the same SV pool. The work is shown in Manuscript in preparation: Grasskamp et al.
