dc.description.abstracteng | Synapses are essential structures for inter-cellular communication in the central nervous system
between neuronal cells. They form highly-specialized compartments that convert electrical to chemical
signals, i.e. neurotransmitter release by synaptic vesicle fusion. A highly dynamic network of
interacting proteins facilitates the cycle of synaptic vesicle recruitment, docking, priming and Ca2+-
triggered exocytosis. Subsequently, fused vesicles are retrieved by endocytosis and are prepared for
another round of the cycle. Synaptosomes are pinched-off synaptic nerve terminals that can be
prepared from neuronal tissue. Synaptosomes are model systems for synapse function, because they
are physiologically active and can be stimulated to release neurotransmitters.
The goal of the present study was to identify and quantify the protein interaction dynamics present in
the synapse. This was attempted by quantitative chemical cross-linking mass spectrometry (XL-MS) of
synaptosomes in resting and excited state. XL-MS is an established method in structural biology that
provides low resolution structural information about protein conformations and interactions.
However, analysis of more complex samples like synaptosomes is challenging. The database search
space exponentially increases with the number of theoretically cross-linkable peptides derived from a
protein sequence database, which impairs sensitivity. Furthermore, synaptosome preparations
frequently contain co-migrating myelin fragments and intra- and extrasynaptosomal mitochondria.
Therefore, proteomic analyses of synaptosomes are dominated by proteins of mitochondrial and
myelinic origin. This thesis has established a biochemical workflow to deplete contaminant proteins
originating from myelin fragments and intrasynaptosomal mitochondria, thereby enabling a stronger
focus on synaptic proteins. In addition, a peptide-focused database search approach for XL-MS was
developed, which first identifies peptides that participate in a cross-linking reaction followed by
providing these cross-linking candidates for database search. In contrast to considering all theoretically
cross-linkable peptides derived from a proteome, the search space is much smaller resulting in a higher
sensitivity. The novel approach was validated on purified cross-linked complexes of known structure
and on in vivo cross-linked bacteria.
Combinging the peptide-focused database search approach with the improved protocol for
synaptosome purification resulted in the creation of a cross-linking protein interaction network of
resting and excited synaptosomes. Numerous known and novel protein interactions were identified
involving, e.g. ion transporting ATPases, synapsins, 14-3-3 scaffold proteins, G-proteins and
Stxbp1/Munc18-1. Furthermore, quantitative XL-MS allowed the quantification of significant changes
in protein conformations and interactions upon stimulation of synaptosomes. Significantly changed
cross-linked residues were observed in Ca2+- and Ca2+/calmodulin-binding proteins, e.g.
synaptotagmin, Anxa6, alpha spectrin, and Camkv. Ion channels like PMCA, Na+/K+ ATPase, SERCA, I3PR
and VDAC exhibited significantly changed cross-linked residues under excited conditions. The implied
conformational changes agreed with the respective ion channel function. Moreover, previously
unknown conformational changes were observed in this thesis, e.g. a major domain movement in I3PR
that might turn the channel inactive, a possible monomerization of Cend1, and an enhanced
interaction between CamkII and neutral ceramidase.
A quantitative XL-MS analysis of changing protein interactions in complex samples like stimulated
synaptosomes was not attempted before. This thesis therefore analyzed the most complex and
transiently changing system by quantitative XL-MS, to date. | de |