A comprehensive characterization of brain and retina synaptic vesicle proteome by quantitative mass spectrometry.
von Sunit Mandad
Datum der mündl. Prüfung:2015-06-09
Betreuer:Prof. Dr. Henning Urlaub
Gutachter:Prof. Dr. Reinhard Jahn
Gutachter:Prof. Dr. Tobias Moser
Gutachter:Prof. Dr. Silvio Rizzoli
EnglischSynaptic vesicles (SVs) are essential organelles for the transfer of information between a pre-synaptic nerve terminal and a post-synaptic target. SVs release their neurotransmitter content into the synaptic cleft in a tightly regulated process, requiring the orchestrated interaction of a number of proteins involved in exocytosis, endocytosis and vesicular recycling. While the molecular anatomy of brain SV is known, the molecular profile of SVs of sensory systems, such as the retina, is not well understood. The major reason is the unavailability of reliable and efficient isolation protocol for SVs with low amounts of starting material. In this study, establishment of a modified isolation protocol resulted in successfully purifying highly pure SVs from retina that formed the basis for a reliable determination of its absolute proteome quantification and molecular composition. Remarkably, this protocol allowed recovery of microgram quantities of highly pure and functional SVs from as low as eight bovine retinas. Maximal vesicle recovery was achieved by the introduction of a harsh homogenization step (powdering the tissue in liquid nitrogen using mortar and pestle and subsequent homogenization using Ultra-turrax) followed by subcellular fractionation. Notably, after fractionation by differential and rate-zonal centrifugation, performing an immunoprecipitation with a monoclonal antibody against synatophysin greatly improved the purity and yield of SVs from retina. The purity of the preparation was ensured by western blotting, electron microscopy and mass spectrometry. The data derived from the iBAQ-MS based absolute quantification of proteins in purified bovine brain and retina SVs from frozen starting materials showed that the copies of SV-integral proteins such as synaptotagmin, SCAMP5, VGlut1 and synaptophysin were similar, if not identical, in bovine brain and retina SVs. Interestingly, however, the copies of v-SNARE protein VAMP2 and tetraspanin protein synaptogyrin-1 were drastically reduced; ~ 6 and 2 fold, respectively, in retina as compared to brain. On the other hand, surprisingly, a three to four fold increase in the copy number of the membrane glycoprotein SV2 were quantified in retina as compared to brain. In addition to differences observed in SV integral proteins, intriguing observations also surfaced when SV associated proteins were quantified. Three copies of the ribbon synapse specific protein syntaxin-3 were found associated to the purified retina SVs, however its brain-specific isoform syntaxin1A/B was totally absent. In addition, the brain-specific Rab3a and synapsin were not quantifiable in our pure retina SVs, suggesting the preparation of SVs to be of majorly ribbon in origin. These striking differences in the retina SV proteome to that of brain highlights the specialized functionality of retina synapses. Although the molecular profile of synaptic proteins in transverse sections of retina is reported in literature, this is the first study where retina and brain SV proteome have been compared in terms of absolute copies. The findings put forward in this study provides a basis for detailed functional analysis in the future. In parallel, highly pure rat brain synaptosomes were quantified by iBAQ mass spectrometry in an aim to characterize the average number of presynaptic proteins. As a further validation of the obtained results, a battery of quantitative western blots was run for a selected group of proteins. These results, in combination with additional biophysical and biochemical data, were used to build a three-dimensional model of an average synaptic nerve terminal. As an independent project, the temporal turnover of synaptic proteins in nerve terminals of the brain and retina were analyzed using a modified version of the SILAC mice approach. To our knowledge this is the first study where a lysine6 diet has been used for labelling proteins over various timeframes to determine protein turnover in vivo. The mice were fed with lysine6 diet for 5, 14 and 21 days and its incorporation in the proteins of brain and retina were analysed by quantitative mass spectrometry. The data shows that the turnover of synaptic proteins in retina is faster than in brain. Strikingly, the turnover of proteins that are involved in similar SV-recycling pathways, correlated well with their respective copy numbers. Future work should be focussed on elucidating the physiological meaning of these interesting observations.
Keywords: iBAQ-MS; SVs