Original subtitles

Hello again, my name is Randy Schekman.

I am in the department of Molecular and Cell Biology

at the University of California, Berkeley.

In an earlier presentation I described a technique, genetics,

used to study the mechanism of protein secretion in a eukaryotic cell:

Baker's yeast, Saccharomyces cerevisiae.

I illustrated that this process uses fundamental principles that are conserved

in all cells that have a nucleus, eukaryotic cells.

Indeed, the process of neurotransmitter secretion in the brain

uses at its core the same machinery that is involved

in delivering a vesicle to the cell surface in yeast and fungi.

And this remarkable conservation of function

is illustrated by identifying genes that are shared

between diverse eukaryotic organisms

where it is possible to study their function in yeast, and then to

infer their function in more complicated mammalian systems.

Now having genes and having mutants is very useful and important, but by itself

it very rarely tells you how the molecules encoded

by the genes perform their function.

One would like to understand at a high level of resolution

how a membrane is shaped.

How protein molecules become inserted into membranes.

How the membranes are formed into a bud or a vesicle.

How that vesicle is delivered to a target membrane.

And then ultimately how the membranes merge by a process of fusion.

This fusion event is critical to understanding how

neurotransmitters are secreted in nerve cells, and how membranes enlarge in all cells.

To get at this level of sophistication it is necessary

to isolate the molecules, to isolate the proteins

and to understand how they interact with one another using functional techniques.

Biochemistry is one such functional technique

that allows one to understand at very high level of resolution

how these molecules work. If you can recapitulate a process

with purified molecules, you can have a much deeper understanding

of how these molecules most likely work inside of a cell.

So I'd like to sharpen the focus, instead of covering the broad contour

of the secretory pathway as I did in my last part.

I'd like to sharpen the focus on one particular station in the pathway.

One that has been successfully reconstituted

with isolated membranes and protein molecules.

Where one can now really begin to understand how this process works.

To illustrate this station that I am going to focus on,

let's consider what happens at an early stage in the secretory pathway.

And I've shown in the top panel an image not of a yeast cell, but of a mammalian cell.

A cultured mammalian cell growing in a plastic Petri dish in the laboratory.

An image taken by a wonderful electron microscopist,

my friend and colleague at the University of Geneva,

Lelio Orci, with whom we have collaborated for many years to study this process.

These events where membranes form into vesicles

were first appreciated by the brilliant and pioneering work of George Palade

and his colleagues in the 1950s at Rockefeller University.

Palade developed the tools of electron microscopy

to visualize these fragile membranes in cells that were professionally engaged

in the production of proteins for secretion

in the pancreas. Several tissues in the pancreas are responsible

for churning out large quantities of protein molecules

like insulin or zymogens that aid in protein digestion.

These molecules are manufactured, and cells of the pancreas devote

essentially all of their energy to this process.

So it becomes a very favorable organism, a system,

in which to investigate at least the morphology of this process.

And what Palade appreciated and what is revealed in Orci's diagram,

the electron micrograph, is as you see with the arrow on this diagram,

membrane buds form on the surface of the endoplasmic reticulum.

And that very event, the budding event, which I will describe in greater detail

is responsible for the first important separation

of proteins that have been inserted during their biogenesis

into the endoplasmic reticulum.

As Günter Blobel and his colleagues, Günter being a protégé of Palade's,

showed in pioneering work, proteins that enter the secretory pathway

do so being synthesized by ribosomes,

those electron dense dots on the surface of the ER membrane.

They are inserted into this membrane.

They either become embedded in the membrane

if they are integral membrane proteins,

or if they pass into the clear interior of that membrane,

they fold and become soluble proteins.

In this location there are literally, in mammalian cells, literally thousands of different molecules.

Some of them stay there to function in the ER.

They are channel proteins involved in new assembly events

or they are proteins that become part of the nuclear envelope, so they stay put.

However, a very large number of molecules do not remain.

They must be removed somehow from the ER membrane

to be targeted downstream in the secretory pathway.

They must leave the ER and go on

to this bus station, the Golgi apparatus, that I described in the last part.

And that event, that sorting event, as you'll see, is achieved

by a very special machinery that coats the surface of

the ER membrane and literally separates molecules according to their final destination.

Those that are to be removed are attracted to this coat complex

that I'll describe and are sifted

into a bud, which then separates from the ER membrane by a process

of fission in which the bud is clipped by fission from its donor.

And these vesicles once separated, shown here in a cluster of vesicles,

congregate in a structure that then

gives rise to the first station of the Golgi apparatus,

the cis compartment, which you see enlarged in this electron micrograph.

So the process of budding, targeting, and fusion is reproduced

early in the secretory pathway

just as it is later in the pathway when vesicles must bud from the Golgi apparatus

and be targeted to the cell surface.

It is this series of events that I am going to focus on

because it has been possible in several laboratories,

to study this in great detail.

Now I am going to highlight some of the parts of the pathway that

I will illustrate in greater detail in a moment.

But let's start first with this cartoon, which describes

the forward, or anterograde, event

in which membranes and soluble proteins destined for transport out of the ER

become packaged into special carriers that I will refer to as COPII vesicles.

You see this on the top limb of this cartoon.

So literally hundreds, possibly thousands, of different protein molecules

must be recognized by a machinery that surrounds this vesicle

and pinches it from the donor membrane,

delivering it via a structure shown in the middle of this diagram

ultimately to the first station of the Golgi apparatus, the cis compartment.

Now some of the protein molecules that are transported are going to leave the cell

and be secreted, and so they will flow via process of maturation

through several sequential stations in the Golgi apparatus.

But some proteins that are delivered by a COPII vesicle must be reused.

They are, for instance, membrane proteins that serve

to address the vesicle to its target, so called SNARE proteins.

Most elegantly functionally dissected by Jim Rothman

and his colleagues now at Columbia University.

These SNARE molecules permit a vesicle to seek out

and engage a proper target and to form a productive union

that permits membrane fusion.

But the SNARES, having done their thing, must be returned back

back to the site of the endoplasmic reticulum,

so that they can perform their function all over again.

It's very much like an escalator.

An escalator flows in one direction, and then the stairs,

or the elements of the escalator

are returned, recycled, and this goes around and around

whether or not people are being moved on the escalator.

Much in the same way, membrane material flows bi-directionally between these first

two stations of the secretory pathway.

SNARE proteins and other molecules then must somehow be removed.

They must be retrieved,

and that retrieval event is achieved by another vesicle that flows in the other direction,

the so called retrograde direction, and invokes yet another

complex of cytoplasmic proteins, called COPI.

So this, the logic is, a bi-directional flow:

COPII vesicles moving things forward, ultimately allowing

most proteins to leave the Golgi apparatus,

and COPI vesicles moving those proteins back

that must be recycled, reused, delivered back to the ER.

If this process is interrupted by a drug or by mutations in essential sec genes

as I described in my last presentation,

then the process comes to a screeching halt.

A block at this limb or at this limb very quickly arrests traffic and causes the elaborate

and exaggerated endoplasmic reticulum membranes

that I showed near the end of my last presentation.

So now with this in mind, I'd like to describe in a little bit of detail

what we know about how the COPII proteins work.

And in the first approach, I'll tell you how it was that we were able,

and many other investigators have been able,

to understand this process using a simple biochemical technique.

So what one would like is to be able to reproduce the formation of this transport

vesicle with isolated ER membranes and soluble cytoplasmic proteins

and then to use this vesicle budding reaction in the test tube

to isolate functional forms of those molecules that are

necessary to pinch this membrane from the ER.

And to do this in my laboratory, two terrific graduate students

David Baker and Michael Rexach collaborated in the mid-1980's.

At the same time, Susan Ferro-Novick, a former graduate student of mine conducted

elegant and very complementary experiments at her laboratory at Yale University.

And also at the same time, Bill Balch at the Scripps Clinic at San Diego in La Jolla

performed similar biochemical analysis using membranes

and soluble proteins from mammalian cells.

Here's the assay that Michael Rexach developed to study

the formation of a transport vesicle in vitro.

A very simple principle was employed.

In a gently prepared extract, a lysate of yeast cells,

indeed in a gently prepared lysate of mammalian cells,

the endoplasmic reticulum does not fragment.

It remains as large envelopes, big tubular networks,

that are so large that they will centrifuge

and sediment to form a pellet at the bottom of the tube with just a brief spin.

It is possible to take this tube with large ER membranes

and put it in a microcentrifuge

and turn the centrifuge on and off,

and immediately all of the ER membrane pellets to the bottom of the tube.

However, if these membranes are incubated with the right conditions

of proteins and ATP nucleotide

small vesicles bud from the ER in vitro,

and these vesicles are so small that they cannot be centrifuged in a microcentrifuge.

They are on the order of 70-80 nanometers in diameter

and in order to pellet them out of suspension

it is necessary to spin them at a very high centrifugal force in an ultracentrifuge.

So just to summarize then, it is possible to develop a functional vesicle budding assay

relying on the fact that vesicles are so small

that they can be separated with 100% efficiency

from ER membranes by a brief spin in a low speed centrifuge.

Now with this functional assay it is possible

to collect the vesicles separate from the ER membranes

and then to dissolve these vesicles or ER membranes

in a denaturing detergent, sodium dodecyl sulfate.

Separate the protein molecules from this denaturing solution on a polyacrylamide gel

and then to use various techniques, either radioactive proteins

or antibody molecules to diagnose the efficiency of packaging or capture

of a membrane protein into vesicles from the donor ER membrane.

I won't show you this kind of experiment, instead I would like to focus just,

for simplicity, on what these vesicles look like.

So using this technique we were able to purify the proteins.

I've already indicated that it is a complex. I'll tell you more about this

complex in a moment.

We were able to purify the proteins,

and with pure proteins and nucleotide, ATP and GTP,

we could reproduce the budding event with isolated membranes.

And then Charles Barlowe, a wonderful postdoc in the lab,

did the, a kind of magic experiment, where

he mixed all of the proteins that we had purified

with membranes and in collaboration with Lelio Orci,

we visualized what these vesicles looked like by using the techniques of electron microscopy

that I described in my last presentation.

And here is a broad view of the vesicles that are produced in this biochemical reaction.

And of course we were very pleased to see how homogeneous the vesicles were,

but we were quite surprised to see that all of the membranes

formed in this condition are studded with an electron dense,

but kind of fuzzy, coat material.

Indeed, we could show using antibody molecules that the coat material

consists of the sec proteins that we knew were required to form the vesicle by

pinching from a donor membrane.

That is to say that this material, under these conditions of incubation,

persist on the surface of the vesicle as though it's

been responsible for the formation of the vesicle.

In higher magnification it becomes clear just how thick and fluffy this coat material is.

Now the impression that one has of this coat

is that it is rather an amorphous structure,

but nothing could be farther from the truth.

Because in other images, where the vesicles are stained

and a surface contour is examined,

it is clear that there is structure associated with the coat molecules as they

have been retained on these vesicles. Some kind of regular repeat structure.

Within the last year, it has become apparent that this regular repeat structure

has a detailed mechanism of assembly,

most recently illustrated by Bill Balch and his colleagues

studying this process in mammalian cells where they found that the outer surface

of the coat forms a very regular array.

And this can happen even without membranes.

So this structure almost certainly polymerizes with some regular

geometry and the details of that have yet to be discovered.

Now to orient this process in the pathway

it was necessary to see exactly which membrane

is involved in forming a COPII vesicle.

In a homogenate of yeast cells, or of mammalian cells

there are many different membranes that co-mingle in the lysate.

And it's very difficult looking at these specimens to be sure that the COPII coat

actually is pinching a membrane from the endoplasmic reticulum.

So to address this issue a terrific postdoc in my lab, Sebastian Bednarek,

isolated yeast nuclei as a source of endoplasmic reticulum,

a source that could be readily separated

from all the other membranes in the cell.

And he examined the ability of isolated yeast nuclei to serve as a substrate

for budding and COPII vesicle formation.

And here is an image from a publication of his. To orient you,

this is the interior of the nucleus, the inner nuclear

membrane that surrounds chromatin.

A space between the two membranes, just the same as the space

in the lumen of the endoplasmic reticulum.

And now the outer nuclear membrane, the membrane that

is in direct contact with the cytoplasm.

Isolated yeast nuclei, incubated with pure COPII proteins

and nucleotide, generate COPII buds and vesicles exclusively from that location.

So we can be confident that the COPII coat does its thing

by pinching membranes from the endoplasmic reticulum.

Now I am going to summarize a great deal of biochemical work by

a large number of talented students and postdocs in my lab and Susan Ferro's lab,

and Bill Balch's lab, in the form of a cartoon

that was assembled by a graduate student in the lab, David Madden,

that illustrates the steps in the pathway of COPII coat assembly.

But before I do that I need to introduce my key collaborator,

and that is Lelio Orci, who is not only a brilliant electron microscopist,

but a very talented artist, and he is shown here in his garden in Geneva,

harvesting what he claimed were

two heads of lettuce, but they look suspiciously like COPII vesicles to me.

So we are grateful to Lelio for his collaboration.

Now to the illustration that I am going to use to highlight

the stations in the assembly of the COPII coat.

Here are the actors. The process, as you'll see, is initiated

by a small GTP binding protein called Sar1.

It is very similar to other small GTP binding proteins

involved at many important events, in very many important events in the cell.

Sar1, as you'll see, become activated by acquiring GTP,

and when it is activated, it extends an N-terminal, amphipathic helix

that allows the Sar1 molecule to become embedded in the ER membrane.

In that location, it acquires in turn two proteins complexes.

One, called Sec23/24, isolated by a wonderful graduate student in my lab, Linda Hicke.

This, as you'll see, is the core of the COPII coat

that allows cargo molecules to be distinguished

from molecules that remain in the ER membrane.

And then an outer layer, a scaffold complex, as Bill Balch showed,

a structure forming scaffold complex consisting of two other sec proteins,

Sec13 and Sec31.

Sar1 is directed to the ER membrane by touching

another important protein, called Sec12,

an integral membrane protein that exposes to the cytoplasm a catalytic domain

that allows Sar1 to discharge GDP and acquire GTP.

This is a so-called guanine nucleotide exchange factor

that operates exclusively on Sar1. And since Sec12 is an integral protein

and is itself a permanent resident of the ER, it almost never leaves.

This orients the activation of Sar1 and allows it to bind exclusively to the ER.

Now in the animation that you'll see in a moment,

the assembly of the coat is linked to the sorting of molecules

indicated by two categories: green molecules are membrane proteins

that are on the go-they are to leave the ER,

and red molecules are stopped.

They remain in the ER and are somehow ignored by the coat.

So now let's have a look, not necessarily in real time, at how this process works.

So Sar1, in the cytoplasm, interacts with Sec12,

acquires GTP, diffuses in the plane of the membrane, acquires Sec23/24 molecules,

which then collide and sample different membrane proteins.

Molecules that are destined for transport are picked up,

and then molecules are clustered together by this

scaffold complex, which assembles on the membrane,

sculpts the vesicle from the membrane,

and then on hydrolysis of GTP by Sar1,

the subunits of the coat are shed from the membrane surface,

discharged back into the cytoplasm, to be re-used for new vesicle budding events.

Leaving behind a naked vesicle that exposes

membrane proteins including SNARE proteins

that allow this vesicle to dock and fuse with a target membrane.

Well, this serves then as a highlight of a key step in the secretory pathway,

where we stared with genetics to understand the genes that are required,

and then used biochemistry to delve into the function of these protein molecules.

As should be apparent, there are many other stations in the pathway,

traffic from the Golgi to the cell surface, where a similar

interplay between genetics and biochemistry has

and will continue to illustrate important mechanistic details.

In my next chapter, I am going to describe

how the process of membrane transport can go awry

in a human in the form of diseases that specifically effect

the machinery that I've described. And you'll see the rather surprising,

and in some cases profound effect, of mutations that affect human health.