Part 3: The Ultimate Fate of the Vesicle

If you haven't yet, you'll probably wish to peruse Part 1, Part 1.5, and Part 2.

I have two things to say in advance:
1. I'll know when I've grown up and become a real science blogger, because I'll be able to write all this in a single post, rather than have to break it up into trois.
2. Now's the part where you're going to hate me. Because I am now going to argue that I have explained to you enough about the difference between endocytosis and kiss-and-run that I can show you that the Nature news piece written about the article I've been jonesing over... is wrong.

Why is that? Let's review the difference between endocytosis & kiss-and-run: in endocytosis, the vesicle membrane joins with the plasma membrane, and then stays as part of the plasma membrane until it later gets pulled back out of it. In kiss-and-run, the vesicle transiently bonds with the plasma membrane, before pulling back, potentially before it's even had the chance to release all its transmitter.

Now let's see what the Nature news piece has to say about what this article has demonstrated:

Following fusion, the vesicle membrane is recycled back inside the cell to form a new vesicle that is refilled with neurotransmitter ready for the next round of communication. The initial fate of the vesicle once it has released its contents is a matter of some contention. One theory — termed 'kiss-and-run' — is that the vesicle membrane remains as a separate entity from the main plasma membrane, like a drop of oil on water, to allow efficient recycling of the membrane-bound molecules and proteins used by the vesicles for identification, docking and fusion.

By labelling a vesicle-membrane protein called synaptotagmin with a fluorescent tag and using STED microscopy, Jahn and colleagues could follow the fusion of individual vesicles at the plasma membrane (Fig. 2 on page 937). Their study provides some of the most compelling evidence to date that at least some membrane constituents remain grouped together after vesicles fuse with the plasma membrane (rather than diffusing freely within the membrane like a drop of water on water), which is consistent with the kiss-and-run theory.

(emphasis mine)

Now as far as I can tell, they're not talking about kiss-and-run at all. They're talking about the fate of the vesicle during endocytosis. Now, I realize that it's quite possible that there is not a single person reading this who cares about this. But I will point out in my defense that I spent two hours last nite poring over the actual paper trying to figure out what it had to do with kiss-and-run, convinced I wasn't understanding it correctly, before I realized something: for the first time in my life, I was confident that I was right and Nature was wrong.

It makes a guy paranoid, I tell ya.

Anyways, onto the paper. So this lab had this revolutionary imagine technique, and decided to apply it to vesicle recycling. To track the vesicles, they developed antibody tag for the vesicle protein synaptotagmin. It's function doesn't really matter in this context, what matters is that it's integral to the vesicle--and only found in the plasma membrane due to vesicle fusion--so antibodies bound to it will show the locations of vesicles in the cell, and where vesicles have fused with the plasma membrane. They practiced their technique on the labelled preparation to show the difference in resolution attainable:

d, Comparison of confocal (left) and STED (right) counterpart images of a labelled preparation reveals a marked increase in resolution by STED. Scale bar, 500 nm.

They then performed a series of experiments to prove that their technique for separately visualizing surface and internalized pools of vesicles would viably distinguish between the two. Basically, they first labelled the cells under conditions preventing endocytosis, and visuallized it. Since there was no access to the inside of the cell for the antibodies, only surface antibodies would be visualized. They then added attempted labelling with primary antibodies under conditions allowing endocytosis; this was followed by labelling the primary antibodies on the surface with non-fluorescent antibodies before permeabilizing the cells and adding secondary fluorescent antibodies. Thus, in this preparation the surface populations of synaptotagmin would be blocked from fluorescent labelling, while only the internal populations of vesicles that had undergone endocytosis would be fluorescently labelled.

Once they demonstrated that this was a viable approach, they then visualized this with the STED technique:


g, STED image of surface-stained synapses (conditions as in a[low temperature & no calcium - JME]). Scale bar, 1 microm. h, STED image of internalized vesicles (conditions as in c[permitting active endocytosis; surface staining blocked; permeabilized after fixation - JME]). i, Quantification of the brightness of single dots (in arbitrary units), compared with the dot brightness derived from single primary antibodies adsorbed to glass (see Methods). The graphs represent averages of 3–5 independent experiments (plusminus s.e.m.). Note that the bin size is 100 units for the single antibody graph and 300 units for the other images.

So what does this show us? First off, note that the synaptotagmin is sorted into discrete dots. Combined with the fact that their graph shows us that each dot--in both the surface and internalized images--seems to represent multiple antibodies, this shows us that synaptotagmin seems to maintain aggregated into discrete pools after fusing with the plasma membrane. Also, there are more dots in the internalized image--demonstrating the high rate at which endocytosis occurs--but the surface dots are generally brighter, since the surface is exposed to the antibodies for a longer period of time during its preparation, thus increasing the efficiency of labelling.

Next, they strongly stimulated the cells in order to promote large-scale rounds of exocytosis and endocytosis while their antibodies were present. They then analyzed the comparative brightness of the stimulated and unstimulated surface images, and followed by analyzing the dot diameters from a number of preparations:


a, Typical STED image of a heavily stimulated, surface-stained, non-permeabilized preparation. Compare with Fig. 2g. Scale bar, 1 microm. b, Comparison of dot brightness of the surface pool in stimulated and unstimulated preparations (mean plusminus s.e.m. from 4–5 independent experiments). c, Quantification of dot FWHM for different preparations (see Methods). We analysed the dots from 3–5 different experiments. To ensure that dot FWHM measurements were not affected by dots that were closely together, a stringent chi2 cutoff of less than 0.01 (difference between the fit and the data) was placed on the lorentzian fit to the dots.

So what does this show us? Even after heavy, extended endocytosis, synaptotagmin is still localized in discrete batches on the plasma membrane. These are clusters of approximately the same brightness during both conditions of heavy endocytosis and during conditions in which endocytosis is predicted: thus, synaptotagmin is equally clustered in both circumstances. Finally, the fact that dot size maps very consistently across preparation implies that each dot represents the staining of a single vesicle.

So this shows us that when vesicles fuse with the membrane during endocytosis, not only does the synaptotagmin from each vesicle not diffuse across the membrane, but it seems to remain aggregated in amounts corresponding to the original vesicle. This strongly implies that when the vesicle fuses with the membrane, it somehow manages to maintain a distinct entity separate from its surroundings. How is this accomplished? Well, never let it be said that Neuroscience is not a discipline with plenty of questions remaining for future generations of researchers.
--
Willig KI, Rizzoli SO, Westphal V, Jahn R, Hell SW (2006) STED microscopy reveals that synaptotagmin remains clustered after synaptic vesicle exocytosis. Nature 440:935-939.