Session 4: Vesicle Trafficking

Transcript of Part 2: Genes and Proteins Required for Secretion

00:00:07.22	Hello.
00:00:08.22	My name is Randy Schekman.
00:00:10.00	I'm at the University of California, Berkeley, in the Department of Molecular and Cell Biology.
00:00:16.26	This is the second of three lectures on the theme of how cells export proteins.
00:00:23.13	In my first lecture, I described the history of the subject and how the pioneers of
00:00:29.05	cell biology were able to understand the structure of membranes, and how membranes within
00:00:35.13	eukaryotic cells relate to one another and convey protein molecules that are encapsulated for export
00:00:43.01	outside of the cell.
00:00:45.00	In this lecture, I'm going to describe the work that started in my laboratory in the mid-1970s
00:00:51.06	to try to understand the mechanism of this process in a simple eukaryotic organism,
00:00:59.00	baker's yeast.
00:01:00.00	So, let's drill down and see what yeast cells do to export protein molecules.
00:01:06.16	First, let's have a look at a... kind of a unique angle, that depicts the process of
00:01:13.26	membrane fusion, at the very end of the process of secretion and growth of the bud surrounding
00:01:23.14	a yeast cell.
00:01:24.14	So, this is an image taken through a plane of the plasma membrane bilayer.
00:01:30.26	The cell is frozen and then a... basically, a small hammer is used to crack the cell.
00:01:38.01	And a split is obtained right through the middle of the bilayer.
00:01:43.15	And this very special image shows a bulge.
00:01:47.11	You see this bulge on the outside of the yeast cell.
00:01:51.04	This is the bud portion of the cell that I talked about last time, where cells are growing
00:01:59.02	and exporting protein molecules.
00:02:01.16	And then, enriched within this bulge or the cell bud, you see small dimples, depressions,
00:02:09.22	little craters in the membrane, that likely depict the nascent membrane fusion events
00:02:19.11	where a secretory granule had just happened to fuse with the plasma membrane.
00:02:25.19	And what you can see is, in the depression, in the dimple, we are likely imaging the molecules
00:02:34.27	that have... were encapsulated within these vesicles, but which are now being shipped
00:02:39.08	outside of the cell.
00:02:40.13	So, it's a very special image that shows that these events of membrane fusion are
00:02:46.09	likely restricted to the bud portion of a dividing cell.
00:02:50.00	So, I told you at the end of my last lecture that this process, we believe, in yeast cells
00:02:58.04	would be responsible not only for secretion, but also that these little nascent fusion events
00:03:04.11	would, step-by-step, in an iterative process, allow the membrane that surrounds the bud
00:03:11.02	to grow, to enlarge.
00:03:12.25	And therefore, if this process of fusion were intimately linked to the growth of the cell,
00:03:19.18	it would be essential for the genes responsible for this pathway to be there for the cell to grow.
00:03:28.06	In other words, if these genes were crippled by mutation, by a chemical mutation,
00:03:32.20	the cell would die.
00:03:35.02	Well, in 1976 and 1977, I had the great fortune... good fortune to have a brilliant first-year
00:03:42.15	graduate student join my lab, by the name of Peter Novick.
00:03:46.02	Here is Peter in the laboratory, busy pipetting away.
00:03:51.03	Note, this is a typical image from the 1970s.
00:03:55.00	We all had long hair back then.
00:03:56.19	And I too had long hair.
00:03:58.02	Peter has now gone on to a very successful career of his own, continuing to study
00:04:05.07	this process in yeast.
00:04:06.19	He is now, as fate would have it, the George Palade Chair of Cell Biology
00:04:12.04	at the University of California at San Diego.
00:04:14.22	So, Peter and I devised a procedure to isolate temperature-sensitive lethal mutations of yeast,
00:04:23.26	specifically focusing on those that caused secreted proteins to accumulate
00:04:31.23	inside of the cell.
00:04:33.03	Now, we were doing this, of course, in the context of the University of California, Berkeley,
00:04:39.01	the birthplace of the Free Speech Movement and the home of student protests.
00:04:43.24	And so, of course, here we were, in the 1970s, proposing to kill living organisms.
00:04:50.22	And naturally, this engendered a certain amount of criticism, indeed, even protest
00:04:55.23	against our work.
00:04:56.23	Here you see one such protest, "End the torture in the labs", "Yeast have feelings too".
00:05:03.23	We had to work hard to convince the experimental subjects committee at Berkeley that
00:05:10.02	yeast are not sentient beings, they approved our work, and we have since killed trillions of yeast cells
00:05:15.06	with no evidence of any torture in the labs.
00:05:19.10	Well, in 1978, after Peter obtained the first such temperature-sensitive lethal mutation
00:05:29.03	that caused secretory proteins to accumulate inside the cell, just by chance George Palade,
00:05:35.08	who I described in detail in my last lecture visited, UC Berkeley for two honorific lectures.
00:05:41.24	And I had the pleasure of telling him about our effort.
00:05:44.27	But, more importantly, Peter joined a group of other graduate students to organize a dinner
00:05:51.00	that evening.
00:05:52.00	It was May of 1978.
00:05:53.23	And at the dinner, he was able to tell Dr. Palade about his work and the new evidence
00:05:59.10	that he had of a mutation that blocked secretion.
00:06:01.28	And Palade naturally was quite interested and suggested that Peter have a closer look
00:06:08.07	by thin-section electron microscopy, the technique that I told you about last time, that Palade
00:06:13.25	used to such great effect in understanding the organization of eukaryotic cells.
00:06:19.17	We very quickly processed this first secretion mutant for electron microscopy.
00:06:24.22	And one of the great memories of my career came when, in the summer of 1978, Peter,
00:06:30.28	in the basement of our biochemistry building, called me excitedly down to the electron microscope
00:06:37.26	to examine images such as you see here.
00:06:41.21	In contrast to the image that I showed at the end of my last lecture, where one sees
00:06:47.14	just a few small vesicles in the bud portion of a dividing cell, in this mutant,
00:06:54.17	which we called sec-1, short for secretion-defective mutant number 1, the cell continues to make
00:07:01.16	mature secretory vesicles.
00:07:04.03	But, instead of a few in a bud portion of the cell, the cell now fills its entire cytoplasmic volume
00:07:12.23	with thousands of such vesicles.
00:07:15.24	They have nowhere to go because this gene, the sec-1 gene, encodes a protein that is
00:07:24.14	required for the granule to be attached... to become attached to the plasma membrane
00:07:32.07	of the cell.
00:07:33.07	And in the absence of that gene, in the absence of that functional protein, the vesicle has
00:07:37.27	nowhere to go, so it continues to be made, to fill the entire cytoplasmic volume.
00:07:43.20	We now know that this gene is evolutionarily conserved.
00:07:48.03	It is present in all eukaryotic cells, wherever a vesicle has to dock and fuse with a target membrane.
00:07:55.10	In fact, we even know, in the brain, in the nerve terminal, as I described in my last
00:07:59.02	lecture, that the synaptic vesicles responsible for fusion and secretion of neurotransmitters
00:08:06.22	rely crucially on a neuronal equivalent of the sec-1 gene product to organize the
00:08:14.18	fusion of a synaptic vesicle with the presynaptic membrane.
00:08:17.17	So, a billion years of evolution, and this pathway, which was... which evolved in microorganisms,
00:08:26.07	has been passed on over the eons, to be used, in molecular terms, in very similar ways by...
00:08:33.21	by human beings.
00:08:36.20	Well, this was of course a very exciting event, to see this first secretion mutant.
00:08:42.00	We knew, on the basis of how we had discovered this mutation, that there must be many more
00:08:46.21	genes to be found.
00:08:48.10	And so Peter devised another really simple and elegant way to isolate more such mutations.
00:08:55.13	He observed in the light microscope that this mutant, sec-1, when the cells are warmed to
00:09:02.09	37 degrees centigrade, continue to make these vesicles
00:09:05.07	-- indeed, they continue to make all of their macromolecules --
00:09:08.07	but they don't get any bigger.
00:09:09.16	They don't enlarge.
00:09:10.25	They must be making more mass, probably squeezing out water.
00:09:15.20	And then they probably become more dense, more... more compact, more material
00:09:23.25	within a confined volume.
00:09:25.13	Indeed, he did the following experiment, which Illustrated that property of these cells.
00:09:32.05	He took a mixture of 99% wild type yeast cells and 1% sec-1 mutant cells, and he mixed these two.
00:09:42.28	And he incubated the mixture at 37 degrees centigrade.
00:09:45.05	And he applied the sample to the top of a tube that forms a gradient,
00:09:50.09	that would allow cells to be separated according to their buoyant density.
00:09:54.01	And he found conditions that allowed all of the normal, secretion-normal cells, to be...
00:10:01.04	to be retained at the top of this tube, at a lowboy density, whereas all of the
00:10:06.24	sec-1 mutant cells, having a higher buoyant density, sedimented to form a pellet at the bottom
00:10:11.10	of the tube.
00:10:12.10	So, he could effect a complete separation of these two populations of cells.
00:10:17.02	This is a biochemist's idea of how to select mutants.
00:10:20.00	And so what we did was we took a yeast culture, we exposed it to chemical mutagen, grew the
00:10:25.26	cells for a while, incubated the cells at a... at a restrictive growth temperature,
00:10:30.21	repeated this centrifugation step, punctured a hole in the bottom of the tube, and collected
00:10:36.01	the densest 1% of cells, plated them out on petri plates, and looked among those cells
00:10:41.15	for those that were temperature-sensitive in their ability to form colonies.
00:10:46.05	And then, among those, Peter found several hundred more mutations that defined 23 different genes,
00:10:55.00	each of which is required to produce a protein that is uniquely important
00:11:03.11	at a different point in the process of protein secretion.
00:11:07.00	He found 10 genes, mutations in 10 genes, that looked just like sec-1 in accumulating
00:11:13.14	small vesicles.
00:11:14.24	But these are different genes, which means that they are... there are at least 10 different proteins.
00:11:19.09	We now know there many more, but at least 10 different proteins that are required to
00:11:22.20	take those mature vesicles and deliver them into a bud to fuse with the plasma membrane.
00:11:28.16	He found two genes that appeared very different in the electron microscope.
00:11:33.03	Here, for example, is one such mutant.
00:11:36.04	This one is called sec-7.
00:11:37.08	And in this cell, when it's warmed to a high temperature, the cell accumulates a structure,
00:11:44.05	rarely seen... rarely if ever seen in a normal yeast cell, that looks just like the
00:11:49.03	stack of membranes that I showed you in my last lecture, discovered in nerve cells
00:11:54.15	in the 19th century, called the Golgi apparatus.
00:11:57.12	Sure enough, this structure can be seen, highly exaggerated in this mutant, because in this
00:12:04.04	mutant proteins are manufactured, delivered to the Golgi apparatus, but, because of a
00:12:11.07	defect in the sec-7 gene, they can't leave the Golgi apparatus.
00:12:15.08	So, the Golgi continues to build up to form this enormous exaggerated structure in the cell.
00:12:22.07	Now, when these cells are cooled back to room temperature, the mutant sec-7 protein is
00:12:31.15	restored to activity and the Golgi decomposes, and the proteins that have accumulated
00:12:38.22	within this structure now can be secreted to the cell surface in vesicles.
00:12:43.11	Another structure that was seen in nine mutations in the original collection caused a defect
00:12:50.20	in traffic of proteins out of the endoplasmic reticulum.
00:12:55.04	So, in this case, these mutations caused proteins to build up in the first station in this pathway.
00:13:02.01	And as a result, this organelle becomes much more elaborate, highly involved in the cell.
00:13:07.27	The nucleus, the membrane that surrounds the nucleus, the nuclear envelope, is distorted,
00:13:13.12	much wider than it is in a normal yeast cell, all because these proteins build up in the structure.
00:13:19.06	And as before, when the cells are cooled back to room temperature, the mutant protein refold
00:13:25.19	and proteins can the ER and progress as normally through the secretory pathway.
00:13:31.00	Now, this... after we had collected these mutants and published the work, we recognized
00:13:36.26	that although we had many genes there was one target that we were particularly interested in
00:13:40.27	that seemed to evade our ability to ob... to define mutations.
00:13:45.27	And that was in the machinery that Palade and Blobel, as I described in my last lecture,
00:13:55.02	showed to be required for the very first step, where secretory proteins move through a channel
00:13:59.26	in the ER membrane.
00:14:01.11	We had no such mutations and we wondered why.
00:14:04.01	And then another... as chance would have it, another brilliant graduate student,
00:14:08.26	by the name of Ray Deshaies, joined the group.
00:14:11.02	Here is Ray at a special celebration of the lab.
00:14:14.19	I'll tell you about Ray's work.
00:14:16.08	His wife, Linda Silveira, also a graduate student in the lab, worked on another project.
00:14:21.23	And Linda Hicke, another graduate student in the lab, whose work I'll describe later
00:14:26.04	in this half-hour section.
00:14:28.03	But let me tell you about Ray's idea and how he was able to conceive of a very special,
00:14:34.12	simple genetic means to identify the channel protein in the ER responsible for protein
00:14:42.13	translocation from the cytoplasm.
00:14:45.20	Here's the idea for Ray's selection, really a very simple one.
00:14:49.22	We know that most cytoplasmic proteins that act, for instance, as enzymes, remain in the cytoplasm,
00:14:57.16	where they have access to their chemical substrates.
00:15:01.15	For example, the last step in the biosynthesis of histidine is achieved by a cytoplasmic
00:15:08.24	enzyme called histidinol dehydrogenase, which takes histidinol and converts it to histidine,
00:15:15.11	which is an amino acid that's used in protein biosynthesis.
00:15:18.07	Now, we also know, from experiments that we did in yeast and which John Beckwith and
00:15:25.10	Tom Silhavy did in E coli, that if you take the gene that encodes a cytoplasmic protein,
00:15:32.18	an enzyme, and you fuse to the 5' end of that gene the sequence for a signal peptide,
00:15:43.18	a signal that would be responsible for secretion, as I described in my previous lecture,
00:15:49.19	you create a hybrid protein that now, in yeast, can be ina... inappropriately translocated
00:15:59.08	into the lumen of the ER.
00:16:01.28	In this case, by attaching a signal peptide to the N-terminus of histidinol dehydrogenase,
00:16:08.22	this enzyme is now sequestered in the ER, where it will no longer have access to
00:16:15.19	its hydrophilic substrate, histidinol.
00:16:18.21	And as a result, if this cell can't make histidine, it can't grow, unless histidine is provided
00:16:27.14	in the growth medium.
00:16:29.28	If you grow the cells on a minimal medium without histidinol, in this case... without...
00:16:36.20	without histidine, they simply won't grow.
00:16:39.01	Well, that is a perfect setup for a genetic selection, because it allows you to expose
00:16:44.24	this cell to a chemical mutagen and to look for mutations that may, for example,
00:16:52.25	cripple the machinery through which this fusion protein would be transported.
00:16:59.03	And possibly allow enough of the fusion protein to remain behind in the cytoplasm to catalyze
00:17:06.23	the conversion of histidinol to histidine, and thus allow the cell to grow.
00:17:12.07	So, to repeat, the cell contains a fusion protein that misappropriates histidinol dehydrogenase
00:17:20.16	to the lumen of the ER.
00:17:22.20	You expose that cell to a chemical that causes mutations.
00:17:26.13	And you look for cells that can now grow in the absence of histidine, in the presence
00:17:31.20	of histidinol, by virtue of the fact that the mut... that the mutation in the cell
00:17:37.16	has crippled the machinery.
00:17:39.04	Now, these mutations, of course, would kill the cell, because, as I already illustrated,
00:17:46.04	mutations that block secretion kill the cell.
00:17:48.18	So, we looked for temperature-sensitive mutations that allow the mutant protein to misbehave
00:17:57.13	just enough to leave some histidinol dehydrogenase in the cytoplasm, but mutations that
00:18:04.27	when warmed to a fully restrictive temperature, 37 degrees, kill the mutant protein completely
00:18:11.27	and prevent the cell from growing under any circumstances.
00:18:16.04	That was the crucial experiment that Deshaies did that allowed us to discover several genes,
00:18:24.04	the first of which, called sec-61, encodes a membrane protein that threads through
00:18:31.15	the ER bilayer ten times.
00:18:34.11	And which we now know is the channel in the ER, not only in yeast but in all eukaryotes,
00:18:41.01	through which polypeptides progress.
00:18:43.17	Well, let me summarize not only that but all of the work that I've described until now in yeast,
00:18:49.28	in the form of a simple cartoon that depicts the stages in the pathway
00:18:55.17	through which secretory proteins progress.
00:18:58.04	This is very much like what Dr. Palade was able to demonstrate in the mid-1970s,
00:19:05.01	but with the added bonus that each of these stages illustrated in this cartoon now can...
00:19:11.20	is populated with genes that are required at each of these steps along the pathway.
00:19:19.22	This pathway is evolutionarily conserved.
00:19:22.00	All of the genes that I've described are found in human cells.
00:19:25.25	And as a result of this discovery, it became feasible to use yeast cells as a
00:19:33.23	production vehicle for the synthesis and secretion of clinically useful quantities of human proteins.
00:19:41.11	And so, in the early 1980s, biotech companies were able to harness yeast cells by introducing genes,
00:19:48.12	such as the gene for human insulin or the gene for the hepatitis B virus surface protein,
00:19:55.12	and use yeast cells, then, to produce quantities.
00:20:00.01	For instance, one third of the world's supply of recombinant human insulin is made by secretion
00:20:04.24	in yeast.
00:20:05.28	Or, all of the hepatitis B vaccine that's made in the world, that's used for vaccination purposes,
00:20:12.22	is made by producing vesicles in yeast cells that house the hepatitis B virus protein,
00:20:19.22	which can then stimulate the immune system.
00:20:22.10	So, this was a practical benefit of the basic science that we and others in my laboratory performed.
00:20:29.12	But we were interested in understanding mechanism.
00:20:33.04	And though we had the tools available to define the genes, by itself the existence of
00:20:40.26	these genes, in the early 1980s, didn't tell us what we really wanted to know, which was,
00:20:45.22	how does this process work?
00:20:47.17	What do these genes encode?
00:20:49.04	What do the molecules, the protein molecules encoded by these genes do to manufacture vesicles
00:20:55.01	that allow cells to grow and secrete proteins?
00:20:59.05	And for this, I'm going to introduce two new observations that allowed us to make progress.
00:21:04.14	The first was a closer look at this first step in the pathway, performed by
00:21:10.17	a wonderful postdoctoral fellow in the lab at the time by the name of Chris Kaiser, who had a close...
00:21:15.08	a very much closer look by microscopy and genetics at the genes required to convey proteins
00:21:20.26	between the ER and Golgi apparatus.
00:21:24.10	Here's a summary of what Chris Kaiser discovered.
00:21:28.07	He found that among the set of genes required for the movement of proteins between these
00:21:34.06	two organelles there are two subsets which show extensive genetic interactions,
00:21:42.10	among each group separately.
00:21:44.25	And which, in the first instance, are required
00:21:48.11	to produce vesicles by budding from the endoplasmic reticulum.
00:21:53.18	And in the second instance, to take these vesicles and to deliver them,
00:21:58.22	by membrane fusion, to the Golgi apparatus.
00:22:02.11	Now, we cloned and sequenced these genes.
00:22:06.00	And very interestingly, two of the genes required for the fusion of these vesicles at the Golgi apparatus
00:22:13.13	turn out to be the yeast equivalents of two proteins that James Rothman and his
00:22:20.11	colleagues had purified from mammalian cells that seemed to be responsible for the fusion
00:22:26.20	of vesicles in the mammalian Golgi apparatus.
00:22:30.25	He purified two proteins, called NSF and alpha-SNAP, which turned out to be the human or mammalian
00:22:40.08	equivalents of the yeast genes sec-18 and sec-17.
00:22:44.28	So that, by the end of the 1980s, we were able to appreciate not only the evolutionary conservation
00:22:50.08	of this pathway but, at the detailed molecular, mechanistic level, genes in yeast
00:22:57.04	had the equivalent function to proteins obtained from mammalian cells.
00:23:02.21	Now, in the final part of this lecture, I want to take a step back to tell you about
00:23:09.19	a historical precedent for how you can use this kind of genetics to bootstrap an understanding
00:23:17.01	of biochemical molecular mechanism.
00:23:19.28	So, now we're going to take a step back, 20 years, to a crucial landmark experiment conducted
00:23:27.20	by two investigators at Caltech in 1965.
00:23:32.23	Here they are.
00:23:33.23	These are Bob Edgar, a bacterial virus geneticist,
00:23:40.15	and his young protege, a new assistant professor at Caltech, by the name of William Wood.
00:23:46.26	Edgar was a classic geneticist.
00:23:49.27	He used the bacteriophage T4 to understand the genes that are required for the production
00:23:57.01	of infectious virion particles.
00:23:59.08	He discovered that mutations in these genes blocked the production of virus particles
00:24:04.16	and caused infected E coli cells to accumulate incomplete virions.
00:24:11.02	But he had no idea what these gene products may do to promote the assembly of the virus.
00:24:17.14	Bill Wood had come as a trained biochemist... trained in the laboratory of Paul Berg
00:24:23.26	at Stanford University, so he was well versed in biochemical analysis.
00:24:28.19	And he saw the great advantage of what Edgar had done to conceive of a strategy,
00:24:33.25	that I'll show you now, that allowed this team to develop a cell-free reaction that reproduced the production of
00:24:42.15	virion particles in the test tube, employing the genes that Edgar had discovered by
00:24:48.03	his genetic approach.
00:24:49.23	Here's the basic experiment.
00:24:53.02	One starts with bacteria that are infected with two different virus mutants,
00:25:01.08	each by themselves incapable of making infectious virions.
00:25:06.02	Now, if these cells had been infected with the two viruses together, genetic complementation
00:25:13.09	inside of the infected cell would have allowed one defective genome to complement the other
00:25:19.21	to produce infectious viruses.
00:25:22.16	But if the mutant viruses were used to infect two different populations of bacteria,
00:25:29.08	nothing would happen.
00:25:30.12	What Wood properly recognized was that if these two different populations were broken
00:25:38.13	and the cytoplasm fractions from each were mixed in the test tube there may be
00:25:47.11	a form of biochemical complementation that would allow each virion to provide the missing component
00:25:56.10	for a completion of virus assembly in the test tube.
00:26:00.04	Indeed, that is what's happened.
00:26:02.02	And the data shown on the right shows a beautiful result, where, at the outset of the experiment,
00:26:09.06	very few if any infectious virus particles are detected.
00:26:12.28	But, as the two mutant samples are incubated in the test tube together, three logs of infectious
00:26:21.07	virion particles are produced during a 30 or 40-minute incubation.
00:26:26.17	This is a result that warms the heart of any biochemist.
00:26:30.19	And provided a historical precedent that we in my laboratory could use to try to
00:26:38.21	identify biochemical entities associated with defective gene products.
00:26:43.08	We struggled for some years to achieve such a reaction.
00:26:46.26	But eventually another brilliant graduate student by the name of David Baker,
00:26:51.02	who has gone on to his very successful own career, joined the lab, and within a very short period of time
00:26:58.08	David had devised a very simple way of breaking open yeast cells that would
00:27:04.11	allow communication, vesicular communication, between the ER and Golgi apparatus to be reproduced
00:27:12.11	in the test tube.
00:27:14.10	His work was followed by the efforts of two graduate students, whose work I'd like to
00:27:20.04	tell you about now.
00:27:21.19	The first was Linda Hicke, whose picture I showed you a few minutes ago.
00:27:25.20	Linda was a graduate student in the lab working on a gene that we now know to be required
00:27:31.03	to form vesicles that bud from the ER.
00:27:34.15	And she used the system that David Baker had developed to do a variation on the Wood and Edgar
00:27:42.07	experiment that I just told you about.
00:27:44.10	And I want to show you the data for that, because it's an experiment that
00:27:48.01	still warms my heart to this day.
00:27:50.12	She took the following combinations of extracts of cells.
00:27:55.18	The dark column shows a sample that was incubated with membranes that by themselves showed
00:28:05.17	no transport as defined by the Baker assay, but could be restored to activity when the sample
00:28:12.28	was incubated at a permissive temperature for the cell-free reaction, in this case,
00:28:18.25	15 degrees.
00:28:20.01	This dark line is of cytosol taken from a sec23 mutant, a mutant that would be defective
00:28:28.20	if the cell had been incubated at 37 degrees, but would be nearly normal if the cell
00:28:33.15	was incubated at a low temperature.
00:28:35.17	Indeed, this cell-free reaction was quite active at the temperature 15 degrees.
00:28:42.05	She also prepared cytosolic proteins from a cell that has a wild type copy of sec23.
00:28:49.17	And similarly, this mixture of wild-type cytosol and membranes produced transport activity
00:28:56.11	as measured in the Baker assay.
00:28:58.27	Now, crucially, as independent samples were incubated at slightly higher temperatures,
00:29:07.13	from 25 to 30 degrees, which we found to be a restrictive temperature for our
00:29:14.15	biochemical assay, the activity for transport with cytosol carrying the wild-type SEC23 protein
00:29:23.14	was more or less intact.
00:29:25.11	But importantly, the activity associated with cytosol carrying the mutant SEC23 protein
00:29:32.22	was down quite markedly.
00:29:35.12	And clearly, then, a temperature-sensitive traffic reaction indicated that this assay,
00:29:43.07	the Baker assay, faithfully reproduced the pathway that we had deduced on the basis of
00:29:51.21	genetic analysis, confirming that this was a functional assay that would allow us
00:29:58.04	to purify these proteins.
00:29:59.20	Now, this observation was further simplified by a very important contribution of
00:30:06.01	another terrific graduate student by the name of Michael Rexach.
00:30:09.06	Rexach observed that, in the course of the cell-free reaction that Baker had devised,
00:30:16.20	membranes in the lysate, specifically ER membranes in this gentle lysate, remained very large,
00:30:25.04	and could be sedimented to the bottom of a centrifuge tube with a very low-speed centrifugation.
00:30:31.21	Her further observed that if these large membranes were incubated with wild-type cytosolic proteins,
00:30:40.02	during the course of an incubation at 30 degrees, small vesicles formed that could not be sedimented
00:30:48.19	to form a pellet at the bottom of a tube.
00:30:51.21	And instead would have to be sedimented only after a very high-speed centrifugation.
00:30:57.21	So, a simple differential centrifugation, of the sort that I described at
00:31:03.22	the outset of my last lecture, was sufficient to separate vesicles that bud from the ER in vitro
00:31:12.00	from the ER membranes.
00:31:14.06	Further, Rexach showed that the mutants, such as sec23, are defective in the production
00:31:21.24	of these small vesicles.
00:31:23.15	Well, this then allowed us to begin to fractionate all of the proteins that we knew to be
00:31:32.22	required for vesicle budding, those genes that Chris Kaiser had described that are responsible
00:31:38.15	for vesicle budding in vivo turn out also to be required for vesicle budding in vitro.
00:31:44.02	And as a result, we were able to discover that these genes have a unique function to
00:31:51.08	assemble on the surface of the ER membrane, to form a bud that pinches from the ER membrane.
00:32:00.09	In order to visualize this process, we developed a collaboration with
00:32:03.19	one of the great morphologists in the world, today,
00:32:06.13	a man by the name of Lelio Orci, shown here in his office in Geneva,
00:32:11.09	with whom I had the pleasure of collaborating for over 20 years on experiments of the following sort.
00:32:18.06	We purified all of the proteins necessary to bud vesicles from the ER, the genes that
00:32:24.12	we had already described by genetic analysis and their protein products, and we took these
00:32:29.25	purified proteins and we mixed them with ER membranes, we sedimented away the ER membranes
00:32:35.00	at low speed, and obtained a high speed pellet fraction.
00:32:38.10	And with the guidance of Lelio Orci, we were able to visualize the vesicles that formed
00:32:44.07	in the test tube, and were amazed to see that a uniform population of about 80-nanometer vesicles
00:32:53.26	were produced in the test tube, each of which is coated by a... what appeared,
00:33:02.08	at least initially, to be a fuzzy electron-dense coat, consisting of the proteins that we added
00:33:09.12	to perform the budding reaction.
00:33:11.11	Now, we now know, and I'll summarize in the next two slides, that the proteins that
00:33:16.25	do this assemble stepwise to produce a bud, to pinch the bud to form a vesicle,
00:33:23.06	and to capture membrane and secreted proteins
00:33:27.11	that are designed to be conveyed from the ER to the Golgi apparatus.
00:33:32.07	So, let me show you a summary slide and then a higher-resolution image of how these proteins work
00:33:37.23	in my last two slides of this section.
00:33:41.10	First, this is a cartoon, summarizing a great deal of work over a period of many years,
00:33:46.20	that describes how this process works.
00:33:48.17	And let me just summarize it for you in just a little detail.
00:33:52.18	We know that this process of budding begins with a small GTP-binding protein called Sar1
00:34:02.20	that acquires GTP by interacting with a membrane protein called Sec12, where it lodges
00:34:10.09	into the ER membrane to begin to deform that membrane to form a bud.
00:34:16.04	We then know that two proteins, in the form of a heterodimer of two Sec gene products,
00:34:22.17	Sec23 and Sec24, assemble onto the dimple that's formed by Sar1, and begin to
00:34:31.18	sample different membrane proteins for capture into a nascent bud.
00:34:37.01	They recognize... specifically, the Sec24 molecule recognizes sequences on membrane proteins
00:34:45.06	that are signals for traffic out of the ER.
00:34:49.12	And complexes are formed that then are clustered together through the intervention of
00:34:56.25	the outer layer of this coat, the Sec31/Sec13 complex, which literally envelops the membrane,
00:35:05.22	in the form of a scaffold, to sculpt the bud and to pinch it, collecting, as it does so,
00:35:13.02	not only the inner layer of the coat but also cargo molecules that are designed to be transported,
00:35:19.05	while excluding proteins in the ER membrane that are designed to remain behind and
00:35:24.12	not to be transported.
00:35:25.18	Now, we now know... through the work of other laboratories that have taken these proteins
00:35:30.18	and developed atomic-resolution crystal structures, we know a great deal about how the molecules
00:35:38.19	of this COPII coat cooperate to form this bud.
00:35:43.15	And let me summarize that in my last slide for this part.
00:35:47.05	This is a lower-resolution image.
00:35:49.17	We now have much higher atomic-resolution images.
00:35:53.02	The coat consists of two layers.
00:35:55.23	There's an inner layer of proteins, consisting of the GTP binding protein Sar1 and its partners
00:36:02.24	Sec23 and 24, responsible for tagging cargo molecules designed to be transported.
00:36:10.06	And then, an outer layer, this outer layer of two other Sec, proteins Sec13 and 31,
00:36:17.04	that form a scaffold.
00:36:18.19	Indeed, this scaffold, as was first described by William Balch and his colleagues in La Jolla,
00:36:24.06	this scaffold has the unusual ability to self-assemble into a regular polyhedron,
00:36:32.03	a cube octahedral structure with squares and triangular facets that forms the kind of exoskeleton
00:36:41.10	that surrounds the membrane, capturing cargo molecules for budding from the ER membrane.
00:36:48.08	Well, we know a lot about these, not only in yeast but also in mammalian cells.
00:36:53.11	We even know that some human genetic diseases are the result of mutations in different subunits
00:36:59.08	of this COPII coat, once again confirming the evolutionary conservation of this pathway.
00:37:06.06	And emphasizing something that I feel very strongly with... about, and which I'll
00:37:09.23	leave you with for this part.
00:37:12.02	Which is that the pursuit of science for its own curiosity-driven thirst for understanding,
00:37:21.19	inevitably, when one discovers things of a fundamental nature such as I've described...
00:37:28.01	inevitably has practical application, in this case, in the biotechnology industry,
00:37:34.25	and even in understanding at a fundamental level how human diseases may evolve.
00:37:39.12	I'll leave you with that for this part of the lecture series.
00:37:44.06	And we'll start in a few minutes with my third lecture, which will describe a process
00:37:50.14	that probably doesn't happen in yeast, but which also involves the capture, in this case,
00:37:56.21	of RNA molecules into vesicles that may be transported within the human body and promote not only
00:38:04.05	normal development but also may be subverted in human disease.
00:38:08.15	Thank you.