Membrane Proteins

The membrane proteins are the most important components in all the various types of membranes in and around cells. The two components of a membrane are the lipids and the proteins. The proteins essentially do all of the difficult and most important tasks. Originally we thought that proteins in the membrane are simply floating around in a sea of lipids that didn’t do very much: this is a Singer and Nicolson models. Now it’s actually known that there are positive functions for some of the lipids: they’re involved in signaling and so on, but for the purposes of this talk I’m going to focus on the proteins which actually carry out many-many of the individual tasks, hundreds of thousands of different tasks in the cell membrane.

Each different membrane protein has a different function and carries out different tasks in the membrane: transporting small molecules in and out of the membrane, transporting big molecules like proteins in and out of the membrane, signaling from the inside to the outside of the cell or from the outside to the inside of the cell… So they have a very important role, and understanding how proteins are built, their structure and how they function in doing their different tasks is the key to understanding how membranes work and how cells work.

When I started working in the area I used to be a structural biologist interested in protein structures in generally. About 1970 it was clear that we didn’t know very much about membrane proteins, and so that was when my own interest in this area started. We began to think about this, and theories like Gorter and Grendal, Singer and Nicolson models emerged that were all rather generic. The idea is to focus on specific membrane protein (and there are thousands of them), find out how that particular membrane protein works and then move to the next one, and then you end up having a general picture of how all membrane proteins are built, what is their structure.

And obviously the function of a membrane protein is the key thing in understanding how it interacts with the outside world: generally structural biologists (and I am a structural biologist) believe that if you just see the structure of something you can look at it and you can often guess, it’s sort of obvious how it works. In much the same way you see a car with four round wheels runnung along, and you kind of see a large part of how it works just by looking at it. So the general philosophy of the structural biologists is to find out the structure so you’ll be able to either deduce or guess or have one or two hypotheses that you can test. Therefore the structure, we think, is the key to understanding membrane proteins and those are the key to understanding membranes themselves.

So when I started nothing was really known about membrane proteins, but there were non membrane protein structures like enzymes or hemoglobin, the oxygen carrier in the blood. And the two features that are the building blocks of proteins in general were first worked out by Linus Pauling who had a Nobel Prize for this and other work to do with the structure of molecules. In the 1950s Pauling and Corey came up with a model building exercise that allowed them to speculate, and at that stage it wasn’t proved, that that all proteins would have secondary structure elements. It means that the polypeptide chain, the chain of amino acids that make up the proteins, would be arranged in three dimensions in a particular way. And there are two hypotheses that turned out to be correct: there is what they call the alpha helix, which is a helical arrangement of the polycontinuous polypeptide chain in which there were 3.6 amino acid residues per turn of the helix, and the other structure was beta structure where the strands were fully stretched out, not in helix, and then they would go along and the next strand would come back either in an anti parallel or a parallel beta structure. Soon after this the structure of myoglobin was solved experimentally on using X-ray diffraction and it was shown to be composed entirely of alpha-helical stretches, exactly the Pauling alpha helix with a heme group, which is the red oxygen binding pigment in the center of it.

A little bit later other structures came along (lysozyme, chymotrypsin) and now there are a hundred thousands sets of coordinates deposited in the protein data bank with tens of thousands of different protein structures, and they all have alpha helixes and beta sheets in them. And that’s all of proteins. Membrane proteins in the 70’s weren’t known, so many people were wondering how it would be.

In a membrane you have this lipid bilayer, a hydrophobic barrier between the inside and outside the cell. You have to get protein molecules either on the lipid or particularly extending through the lipids, transmembrane proteins.

And so one idea was that you cannot have the parts of the protein that need to interact with water interacting with the lipid bilayer. That means that in a protein molecule you have a polypeptide chain, and there are components of the polypeptide that must be solvated by water. So you have a peptide bond, you have an NH-group and you have a carbonyl group so you have the peptide bond with NH—CO and these are the two moieties that are involved in Pauling’s alpha helix and beta sheet. If you have a beta sheet, all the hydrogen bonds are satisfied inside the beta sheet, and if you have an alpha helix, all these hydrogen bonds between the NH and the CO are in the alpha helix, so it was a reasonable hypothesis. I certainly had a review in 1980 in a meeting in the Alps where we said: ‘All membrane proteins will be either composed of a bundle of alpha helix stretches of the polypeptide crossing the lipid bilayer one time, two times, three times, n times, or there would be beta structure’. And the only way to remove all the hydrogen bonding is to have a beta barrel and there were beta barrels, for example, chymotrypsin as soluble enzyme that was composed of two six stranded beta barrels. So the hypothesis was that you would have either membrane proteins made up of alpha helixes or membrane proteins made out of beta barrels.

And we now have the structure of thousands of membrane proteins and almost all of them consist of a transmembrane alpha helical bundle or a beta barrel. Occasionally there will be a beta barrel with one alpha helix going through the middle or there will be a beta barrel with helixes on the outside. Generally speaking, the overall idea that they’re either transmembrane helixes of transmembrane beta barrel holds. Most of the membrane proteins in the outer membranes of bacteria, mitochondria and chloroplasts have this beta barrel structure and most of the membrane proteins in the inner membranes, which includes nuclear membrane, endoplasmic reticulum, ribosomes, are composed of alpha helical bundle proteins.

So there is a general understanding now of what to expect in different parts of the membranes in the different organelles and it fits with earlier fundamental theories of Pauling helixes and beta sheets, how they all fit into the membrane. But with time a lot more information has come about individual membrane proteins, usually from one person or one group of scientists focusing on specific membrane proteins. When I was a young scientist we began by saying: ‘The most interesting membrane proteins might be for example in humans: the ion channels’. You have muscles and nerves, and they conduct a nerve impulse from one end to the other and the way they do this was worked out by Hodgkin and Huxley when they analyzed sodium and potassium ion fluxes in and out of the membranes. They didn’t know what it was but they said: when a nerve impulse passes either in nerve or muscle, first channels open that are permeable to sodium ions and they flow from the inside of the cell to the outside of the cell. The cell voltage drops from 60 or 70 millivolts negative inside to zero and then that opens other channels and you get a nerve impulse propagating. The recovery happens when the sodium channels close and then potassium channels open: they flow back and the membrane goes back to resting.

These were very important membrane proteins and they still are and people still work on them, but in the 70s they were too difficult to work on. So we looked around for simpler membrane proteins that were more stable, easier to work on. The one that I focused on for about 10 or 15 years was bacteriorhodopsin which turned out to be a membrane protein with seven transmembrane helixes: exactly the kind of alpha helical bundle that would be consistent with the Pauling alpha helix. We found out its structure first at low resolution in 1975 and then at high resolution in 1990. Another membrane protein that had a lot of impact in the early years was from the work of Hartmut Michel who is a research scientist at the Max Planck Institute of Biochemistry in Martinsried in Germany and he focused on the reaction centers which is the site at which light energy after being captured in chloroplasts is funneled down and converted into an ion gradient across the membrane. He crystallized this complex of four proteins in 1983 and the crystals were beautifully behaved, beautifully organized, And he and Hans Deisenhofer solved the structure published in 1985 and so on. And this also consisted of a bundle of I think it was 11 transmembrane helixes, so the L and the M subunits each had five helixes surrounding chlorophyll at the reaction centers involved in light capture and then one other transmembrane helix. So bacteriorhodopsin had seven transmembrane helixes, the reaction centers had eleven, and so that set the scene throughout the mid 80s.

Soon after that Michael Garavito who had been working with Jürg Rosenbush in Switzerland they determined the very first structure of a bacterial outer membrane protein and that turned out to consist of 16 strands of beta barrel. So by 1990-1991 we had two or three examples of membrane proteins that had either transmembrane alpha helical bundles or beta barrels thus giving in great detail a picture of the structure, the general title which had been hypothesized before. Throughout the 90s and 2000 and so on the methods improved, some structures were determined by electron cryomicroscopy which was our original area of interest, others were determined by making 3D crystals and doing X-ray crystallographic analyses of the structure. There were other powerful methods that were added in, for example, lipidic cubic phase was a particular trick for making membrane proteins crystallized in three dimensions and now many of these structures are done by that method. Most recently the electron cryomicroscopy method has been improved technically by better microscopes and detectors, and now you can determine structures of membrane proteins and other proteins without ever making a crystal, without doing electron crystallography, without doing x-ray crystallography.

And so now there are increasing numbers of quite important membrane protein structures being determined by these newer methods. Now we have the protein data bank which is the depository for all the protein structures, nucleic acid structures, ribonucleoprotein structures, all the different proteins and BAR structures in biology: there are now a hundred thousands of these deposited. It would take you a long time just to look at them all. Of these one thousand, probably a few thousand are membrane protein structures, so still we may have knowledge of the three-dimensional structures of perhaps half the proteins in the world, either as a 3D image of the structure itself or some related similar homologous structure.

In summary, having started with nothing we now have a knowledge of thousands of different membrane protein structures: beta barrel structures, alpha helical structures, and then there are a few that are hybrids of the two.

There are also just the occasional special membrane proteins, for example, aquaporin which has a function of allowing water molecules to pass through all cells but particularly in the red blood cells, controls the size and osmotic pressure. It has a very special arrangement of the polypeptide where one of the strands of the protein goes in and turns around halfway through the membrane and comes back, and a similar one from the other side. So there are perhaps a very small number of ultra special membrane proteins that do not fit into these categories and that’s because they have a particular function. I would say, now we have a really excellent idea about membrane protein structure and this knowledge permeates the thinking of all the people in biology who are studying different aspects of biology.