NEW YORK (GenomeWeb News) – Combining mass spectrometry, immunoblotting, and electron microscopy, a team led by researchers at the University of Göttingen has built a 3D model of a mouse synaptic bouton.
Detailed in a study published last week in Science, the model offers both insights into the mechanisms of synaptic vesicle trafficking and cellular trafficking more generally, said Sven Truckenbrodt, author on the paper and a graduate student in the lab of Göttingen researcher Silvio Rizzoli, who led the project.
Synaptic boutons, or synaptosomes, are the location of synaptic vesicle recycling, the process by which these vesicles fuse with the cell membrane to release neurotransmitters and are then taken back into the cell and refilled with neurotrasmitters.
Because this process takes place entirely within the synaptic bouton, it can be studied in isolation, and as such, Truckenbrodt told ProteoMonitor, it provides a useful model for the study of cellular trafficking.
The synaptic bouton also has the advantage, from an organelle proteomics perspective, of being reasonably easy to purify. Sufficiently purifying target components has traditionally been one of the key challenges to organelle proteomics. Indeed, this issue has led researchers to develop new methods relying not on biochemical purification but rather by fractionating cells by density and comparing protein levels across these different fractions.
Because the synaptic bouton is a widely studied system, however, good enrichment protocols were already in place, "so we had the advantage that we could just use these protocols that work quite well," Truckenbrodt noted.
The researchers began their analysis by using quantitative immunoblotting to measure levels of 62 synaptic proteins of particular interest. They then determined the amount of protein per microgram of synaptosome by using fluorescence microscopy to measure the number of particles in the overall synaptosome preparation and then electron microscopy and immunostaining of known synaptic markers to determine the portion of that preparation that was, in fact, synaptosome.
The immunoblot analysis detected roughly 40 percent of the synaptosomes' protein content. Seeking to further validate and expand upon this finding, the Göttingen researchers then turned to mass spec, analyzing the preparations using a Thermo Fisher Scientific LTQ Orbitrap Velos. This analysis identified roughly 1,100 additional synaptosome proteins, bringing the total portion of proteins identified between the two analyses to approximately 88 percent of the total protein weight of the synaptosome preparation.
As Truckenbrodt noted, the analysis still leaves as an open question the composition of the remaining 12 percent.
"We basically can't say much about these proteins that couldn't be assigned," he said. "We can't say how diverse this [population] is, how many different types of proteins were missing."
However, he said, between the two analyses, the researchers were able to identify essentially all the proteins known to be key to synaptic vesicle trafficking.
"We identified basically all the proteins that we knew from previous studies to be very important," he said, adding that while their immunoanalysis had difficulty with certain quickly degrading proteins, the mass spec work compensated for this issue.
These protein measurements led to a pair of what Truckenbrodt characterized as "very unexpected" findings.
One of these findings was that proteins in the same pathway and same steps of a particular pathway appeared to be present in very similar numbers, a curious discovery, he said, given the disparate natures of these proteins.
"We had no idea how well the numbers would correlate in these trafficking pathways," he said. "And we actually didn't expect to find such good correlation, because all of these proteins are transported to the synapse on different carriers and all of them have vastly different lifetimes. We still have no idea how the cell regulates these copy numbers to match so well."
Also surprising, he said, was the discovery that while proteins involved in delivery of synaptic vesicles to the cell membrane for release of their neurotransmitters are quite abundant – with some present at levels in the range of 26,000 copies per synaptosome – the proteins involved in the endocytotic step, wherein the synaptic vesicles are taken from the plasma membrane back into the cell, are considerably less abundant, typically in the range of 1,000 to 2,000 copies.
This indicates that the synaptic bouton is able to simultaneously endocytose around 10 percent of its synaptic vesicles, Truckenbrodt said, noting that "it was very surprising to find these two arms of the trafficking pathway to be so differently provided for."
Using stimulated emission depletion microscopy, the researchers obtained localization data within the synaptic bouton for 60 proteins involved in synaptic vesicle trafficking, building a 3D model of the space comprising around 300,000 proteins.
This model, the authors noted, suggests one possible rationale for the disparity of the level of proteins involved in delivery versus endocytosis of synaptic vesicles – specifically that the crowded nature of the active zone where vesicles are delivered to the membrane could require high protein copy numbers to deal with constraints on organelle and protein diffusion and allow for rapid neurotransmitter release.
Endocytosis, by contrast, can take place more slowly, they wrote, allowing that process "to proceed with proportionally lower numbers of cofactor proteins." Keeping copy numbers of these proteins low could also be a means of preventing the active zone from becoming even more crowded, they added.
The researchers limited their modeling to proteins quantified both by immunoblotting and mass spec, Truckenbrodt said, so as "to be very sure that the [copy] numbers were correct."
The availability of good affinity reagents was also a limitation, he added. "We would have done more, but at some point you run out of decent antibodies."
"If new proteins turn out to be important and new antibodies [to them] become available, then we can use that to expand our model even further," he said.