Molecular mechanisms of vesicle formation and fusion

Molecular Mechanisms of Vesicle Transport/Fusion

One of the most active areas of research in cell biology these days is in trying to figure out the mechanisms involved in sorting and intracellular transport of proteins and vesicles. There is still a great deal to learn, but scientists have applied a number of approaches to identifying the cellular components that are involved in these processes. The book describes several of these on pp.733-743, and you should pay particular attention to them. One of the most productive approaches is through the study of secretion in the yeast Saccharomyces cerevisiae, or baker's yeast. This is a single celled eukaryote that is relatively easy to do genetics on. Scientists mutagenized yeast cells and isolated mutants that were defective in secretion of some substance (see Fig. 17-14 in Lodish et al.). Lots of different kinds of proteins might be affected by these mutations from those involved in processing newly synthesized proteins in the ER to those involved in fusion of secretory vesicles to the plasma membrane. These mutants, and the proteins that they encode, were named in sequence as SEC (secretory) mutants--Sec1, Sec2, etc. It's difficult to do genetics on multicellular organisms, so a biochemical approach has been taken in mammalian tissues to get at the same issues. In these cases, people try to reconstitute secretion, or some part of the process, in a test tube, or in cells that have been partially disrupted so that molecules can be added to them at will. The general strategy here is first to get something to happen, such as movement of vesicles from ER to Golgi, then to remove a part or fraction of the system so that it is disrupted. If you add back what you took away, and you thereby restore function, then you have an assay to figure out what the active components are. So the biochemists purify the various contents of the removed stuff, add back the purified substances, and try to see which ones are required for function. People using both methods have identified a large number of proteins that are likely to be involved in the transport pathways, and there is now a scramble afoot to figure out how they're involved, and in what order.
You might think that studies on yeast cells and studies on mammalian cells, which are so distantly removed in evolution, would yield very different results. It might seem likely that very different mechanisms would be at work in these two kinds of cells. In fact the similarities between them are quite striking and surprising, to the extent that some yeast proteins can substitute for mammalian proteins in reconstitution of the secretory pathway in vivo. This suggests that the pathway has been highly conserved over evolutionary time, since it first appeared.
Although the details differ somewhat from cell to cell and from compartment to compartment in the same cell, there are at least three principles emerging that seem to characterize all vesicle mediated transport within cells (Fig. 17-50).
First, formation of membrane vesicles from a larger membrane requires the assistance of a protein coat and "adapter proteins" that bind both to the coat proteins and to the "cargo proteins" in the membrane that is being pinched off to form a vesicle (Fig. 17-51).
Second, the process is facilitated by a number of proteins that interact with GTP or GDP--bind them, hydrolyze them, exchange them and so forth. In other words GTP seems to be one of the main sources of cellular energy that propels vesicle movement and fusion.
Third, fusion of a vesicle that has budded off from one compartment with the membrane of another compartment requires that there be complementary proteins in the two membranes--one associated with the vesicle membrane, and one associated with the target membrane.
There are at least two kinds of proteins that mediate budding of vesicles from large membrane. One is associated with vesicles that are going to be targeted from the Golgi apparatus or the plasma membrane to lysosomes--these involve a coat protein called clathrin (Fig. 17-50, 17-53). The second class seems mainly involved in passage between the ER and Golgi apparatus--these involve two kinds of coatamer proteins (COPs), one involved in anterograde (forward) transport and one involved in retrograde (backward) transport.
Individual clathrin molecules have a three legged structure (called a triskelion) that can self assemble into a kind of molecular geodesic dome. That is, if you purify clathrin and incubate it under the right conditions, then the clathrin will spontaneously assemble into an empty cage structure. In the cell of course when clathrin forms a cage it captures a sphere of membrane inside it. The clathrin molecules are attached to the membrane by a group of linker proteins that are called adaptors or assembly particles, which are made of a 4 molecule complex of proteins called adaptins. One kind of adaptor is supposed to be involved in coated vesicle formation from the TGN, the other in formation of coated vesicles at the plasma membrane. (Fig. 17-53) . The model is that the assembly particle attaches to membrane proteins with a particular binding sequence in their cytoplasmic domain and the clathrin binds to the assembly particle (Fig. 17-55). This forms first a coated pit, but once enough clathrin molecules have assembled, it was believed until recently that the free energy of their association, and the fact that they form an enclosed cage structure, is sufficient to drag along a piece of the membrane, forcing it to form a sphere encased by the clathrin. However, recent experiments suggest that a protein that binds and hydrolyzes GTP, called dynamin, is also essential for endocytosis of clathrin coated vesicles. (The experiment was to put a mutant dynamin gene into cells and show that it interfered with endocytosis, implying that normal dynamin is essential. This also means of course that formation of the coat isn't spontaneous because hydrolysis of GTP is required). Once formation of the coated vesicle occurs and the vesicle breaks free of the membrane that it was part of, the clathrin is rapidly lost. As you might expect, while clathrin assembles into the geodesic dome spontaneously, it does not dissociate easily. Indeed, other proteins and the energy derived from hydrolysis of ATP are necessary to disrupt the clathrin coat and form an uncoated vesicle. As we'll see in a few minutes, the "uncoated" vesicle is not merely naked membrane, but retains some proteins essential for it to fuse to another membrane.
The other kind of coat is one formed with coatomer proteins (coatomer means coat-forming), which are abbreviated COPs, since there are two kinds, they're called COP1 and COP2. These vesicles seem primarily to be involved in the default pathway. There are at least 6 COPs, identified as a,b, b ',g,d,e,z. The genes for several of these have been purified and sequenced. Beta and Zeta re similar to the assembly particles of the clathrin-adaptor complex. This suggests that they serve similar functions in coatomer coated vesicles as the related proteins do in coated vesicles, but these functions remain to be identified with certainty. For the coat to bind to membranes, they must interact with ADP-ribosylation Factor (ARF), a GTP binding protein whose function is uncertain, but which probably provides the energy and location necessary to facilitate coatomer formation on the membrane of the Trans Golgi Network (Figs. 17-58 and next figure for model).
Once the vesicle pinches off the initial membrane, it sheds its coat, travels, usually in a directed way, through the cell to its "target" membrane, and then fuses with the new membrane. This fusion process is a complex one about which a considerable amount has recently been learned. Using both genetic approaches in yeast and biochemical approaches in mammalian cells, it has been possible to identify a number of the proteins that are involved in vesicle fusion with the target membrane. As should be obvious if you think about it, it is energetically unfavorable for two phospholipid membranes to approach close enough to each other to touch, because they must squeeze out all the water molecules bound to the phospholipid head groups and protein molecules that would keep them apart. Thus fusion must require some input of energy. In addition, there must be some kind of specific interaction between the vesicle and its target membrane. That is, if a vesicles buds from the ER, there must be some way to ensure that it will fuse with the cis Golgi network and only with the CGN, not plasma membrane, TGN, lysosomes, etc. Similarly vesicles coming from the TGN should only recognize the plasma membrane and no other cellular structure.
How is this specificity accomplished? One current hypothesis is that there are complementary proteins associated with the vesicle membrane and the target membrane, that interact with each other and stabilize contact between the two membranes long enough for them to be fused. The binding and unbinding of these two proteins is regulated by several other proteins, especially one kind that hydrolyzes GTP and thereby provides the cellular energy necessary to promote fusion. This hypothesis is called the SNARE hypothesis by James Rothman and colleagues who developed it.
I should say that the terminology here is somewhat confusing. Rothman et al. call the protein in the vesicle membrane that is required the v-SNARE and protein in the target membrane t-SNARE. Other proteins are called NSF, SNAP, and the group of GTP-binding proteins is called Rabs. It gets worse, as I'm about to show you. What I want you to get from this information is not the specific names of the players, but the general features that seem to characterize these membrane fusion processes.
The next figure shows you a current model derived from studies on several different kinds of membranes about how the fusion complex is organized. As you see, it is believed that about 7 different proteins are required. One is the v-SNARE class, an integral membrane protein of the vesicle. Another is the t-SNARE, an integral membrane protein of the target. These differ for every fusion complex--show on the overhead. In addition there are several soluble proteins that bind to these membrane associated proteins and presumable help to hold them together. Some of these soluble proteins seem to be the same for all the fusion pathways (e.g. Sec 17 & 18), some are specific for only one. Demonstrate. Finally all the fusion pathways are stimulated by a small GTP binding protein of the Rab family (over 30 different Rab proteins have been identified) that is embedded by a lipid tail in the vesicle membrane for the duration of the fusion event, but that is recycled back to the origin and reused in formation and fusion of new vesicles (Fig. 17-59).
The currently accepted model for this process is shown on the next overhead. The idea is that the Rab starts out anchored in the membrane of the nascent (define) vesicle. It's thought that the Rab may assist in the assembly of the docking/fusion complex on the vesicle membrane, and that budding of the membrane vesicle cannot occur until this happens. The vesicle then moves to contact and begin to fuse with the target membrane. Here specific recognition requires the right combination of v-SNARE and t-SNARE. That is, if a vesicles contains one kind of v-SNARE, then it will only bind to one kind of t-SNARE which will be present in only one membrane of the cell. Once fusion of the vesicle is initiated, Rab interacts with another protein in the target membrane called a GTPase Activating Protein. This turns on an enzymatic activity in the Rab, which catalyzes breakdown of GTP to GDP. This in turn is thought to trigger a conformational change in the Rab which allows it to be plucked out of the membrane by GDI, a protein that binds Rab and prevents loss of GDP. This complex somehow finds its way back to the initial membrane where two proteins, GDI displacement factor (GDF) and Guanine nucleotide exchange factor (GEF) promote unbinding of Rab from GDI and a swap of GTP for GDP. Binding of GTP again alters the conformation of Rab so that it can reinsert into the membrane and begin the cycle all over again.
I'm showing you the Rab cycle not so you can memorize all these acronyms, but because this cycle is similar to many others that are important in membrane associated events in the cell, and also very similar to the Ran transport cycle for nuclear proteins. The basic features are that the functional protein alternates between an active and inactive state, and that the switch between them is the GTP--GDP conversion. Second, other proteins help recycle this functional protein back to the active state by catalyzing an exchange of GDP and GTP. Thus, GTP hydrolysis must provide the energy source for the vesicle formation and fusion process, but we don't begin to understand how. This kind of scheme will show up again later when we discuss signal transduction at membranes.
To reiterate the general scheme is that vesicle formation is driven by formation of a coat of Clathrin or coatomer proteins, it is driven by GTP hydrolysis and a group of GTP binding, hydrolyzing and exchange proteins that are involved somehow in both vesicle formation and vesicle fusion (in this case the Rab proteins and all the ones that interact with them), and the fusion requires specific interaction between proteins in the vesicle membrane and target membrane, which are sometimes called v-SNARES and t-SNAREs.