Salvador Luria (1981) - Membranes and Transmembrane Channels

Ladies and Gentlemen, students and colleagues. It’s very pleasant to be a scientist speaking in a theatre because they give you the flowers before the performance, whereas an actress has to wait until after the performance. Probably the idea is that if they waited until after the performance we wouldn’t deserve the flowers. I think it’s going to be a difficult task. We followed 2 beautiful talks. A little bit like 2 marvellous symphonies which are so beautiful that one forgets that they continue, that the last movement continues 15 or 20 minutes longer than one expected. But I hope I still have enough of the time allotted to me for my own talk. But if the boat is to leave and I continue to talk and the boat is leaving, please feel free to leave and go for the promenade on the 'Bodensee'. Another apology I must make is that I cannot or I do not wish, like many of my colleagues, to talk about the fruit of their meditations and discoveries of many years. Because I feel that as long as I can talk about something that I am actually doing now and which I do not yet understand very well, I keep young. And this is one of the problems that one has in life is to keep as young as possible, or at least to forget that one is getting old. If I wanted to give to this lecture a very impressive title, I would say I want to talk about the molecular biology of transmembrane channels. But it’s really much less impressive than that. The question is that, as most of you know, biological membranes are rather strange things. On the one hand they have a stupid framework consisting of phospholipid molecules which are very impermeable to almost everything except water. And which are really relatively indifferent. Only the specialists really concern themselves about the internal changes with temperature or with other conditions. But in order for substances to go into cell or from one part of the cell into another, if the portions, if the parts of the cells are subdivided by membrane, there has to be some mechanism. And these mechanisms are channels. And these channels are proteins or group of proteins about which the task or people, the very few people fortunately still in the world - nothing like DNA or recombinant DNA. But there are still and already a substantial number of people beginning to understand that there is such a thing as interest in the way things go in and out of membrane. Of course neurophysiologists being the first ones who had to worry about these problems. But what one would like to know about these channels is, of course, first of all the specificity of the conductance for various substances. The rates at which various substances go through and what determines those rates. The dependence of the conductance for various substances on external stimuli, either chemical or electrical. And finally how they are formed and organised within the membrane. And to make clear about the various types of channels that we have, I would like to show you in the first slide a few types of membranes. This being the cytoplasmic membrane, let’s say of something, possibly of a bacterium or of a cell or of nerve fibres. Some are intelligent channels, for example the ones that regulate the transport of substances across the membrane. Either purely because of differences in the concentrations of specific substances, or because of actual coupling with energy from ATP or in other ways, as we’ll see later. Then we have not only the intelligent but the neurotic channels, that is those which are gated by the electrical potential. That is the membrane potential as it determines the permeability of channels for sodium or for potassium and so on. Then we have somewhat less intelligent channels which do not discriminate very much. And an example of that are the gap junctions that form between cells, for example in the very early embryo of most animals. And which can be formed, open or closed, either functionally or also developmentally, so that cells at a certain point stop being tightly coupled by gap junctions. And then finally we have another class which I would call the most stupid ones, but not quite stupid enough for my own taste. And an example of that are the so-called porins or matrix proteins or the outer membrane of bacteria. Bacteria of course like mitochondria have 2 membranes, an inner cytoplasmic membrane with all of the intelligent things. And an outer membrane which, not having any energy available, has to provide holes or channels through which substances up to about the molecular weight of 1,000 can go through and reach the active region. Now all of these membranes, all of this type of membranes on this slide have one unpleasant feature. In the course of the growth or the formation of the cell, the proteins of these channels have to go and insert themselves into the membranes. And therefore in general the biochemist who has tried to work with them finds that they are more or less distasteful, because they generally are poorly soluble in water. And when it comes to things which are poorly soluble in water biochemists prefers to turn the other way and look into the swimming pool. Now for example in the first class we can put the acetylcholine receptor – which is practically the only receptor for a neurotransmitter that has been purified and isolated in a satisfactory way. And that took approximately I’d say 10 to 15 years. In the case of the outer membrane pore of bacteria the matrix protein has been purified in a couple of laboratories after 5 or 6 years of work. So when many years ago I became interested in the idea of how things may go through membrane and so on, I thought, in fact in became interested because I became interested in a class of proteins which are really very remarkable. They are called colicins of the E group or the E1 group. And these colicins are antibiotics that attach themselves to certain bacteria and kill them. And as I will show you in this very brief talk, they kill them by doing exactly what a biochemist would like to do: by inserting themselves in the phospholipid bilayer of the cytoplasmic membrane. And they are creating a hole whose conductance is non-specific, relatively non-specific, and allows substances to pass, which have molecular weights up to about 800 or 900. And so what I am going to do now is to tell you a little bit the history of the battle that myself and my students have carried out against these proteins for a number of years. The next slide shows the effect of some of the proteins of this group. If you mix bacteria with sufficient amount of one of these proteins, you'll find that a whole series of physiological effects occur: The synthesis of protein RNA, the next should be DNA, and glycogens are arrested. Active transports, most of them, are actually blocked. Bacterial motility is blocked. And ATP levels decrease. And the remarkable point that I would like you to keep in mind is that all of these events are the results of a single molecule of colicin, attaching itself to a single bacterium. The kinetics is strictly first order. And as we know now, one molecule is sufficient to produce these effects. And we thought about strange possibilities of how to explain this several years ago, although now we have an explanation which is extremely simple and somewhat less interesting. Now whenever you see that one molecule of a protein interacting with a bacterium can do all of these things, you immediately think that it must act at the level of the energy metabolism. Because that’s the one place where all of these phenomena are going to be bound. Now the next slide shows a very simple way. A scheme of the energy metabolism of a bacterium like Escherichia coli growing with glucose as the only or main carbon source. And of course out of glucose can get phosphate or pyruvate, which is used in bacteria to bring in more glucose and mostly used to produce ATP. ATP is used for all of the syntheses. On the other hand we have electron transport with all of its intermediates. And the electron transport as well as the ATP with the mediation of a calcium magnesium activated ATPase. Both serve to generate what used to be called energised stratum membrane, and which we now call the proton motive force which has been shown now directly to energise bacterial motility and the active transport of a variety of substances. I hope everybody in this audience will forgive me when I say 'energise directly', because I do not want to enter into the question of whether the proton motive force is used directly at the level of certain proteins to transport substances. Or there might be some intermediates which Peter Mitchell would like not to hear about. And I agree with him. Now given this complexity it was quite clear that in some way the colicin had to affect somewhere this general complex. Either at the level of the ATP itself, therefore reducing this. Or at the level of the proton motive force itself. Or at the coupling of the proton motive force with these functions. And we solved this problem by making a second slide. The next slide. I give you about 10 seconds time to see how it differs from the other. (laughter) This is a mutant of course. I have to have bacterial mutants, otherwise I would not be true to my past. And this is a mutant that is blocked in the ITPase, which is the only ATPase functional in the bacterial cell. In such a mutant the proton motive force can be generated only by electron transport, and not from ATP. And therefore the effects of colicin on this mutant should reveal whether we are damaging here or here. And the next slide shows what happens when colicin acts on the ATPase-defective strain of E. coli. And here it turns out of course that respiration continues, which is fortunate, otherwise the experiments could not be done. But the synthesis of micro molecules, RNA, DNA and proteins and glycogen continue. And the ATP levels increase so that - what is happening here? We have created bacteria which are still able to make protein, to make RNA and DNA. They have a higher level of ATP now because the ATP is not wasted in charging the proton motive force, this squiggle. But the motility is still abolished and the accumulation of potassium, rubidium, galactosides, amino acids is still abolished. But now we notice that pre-accumulated substances come out of the cell. So we conclude that the effect of the colicin on the energy metabolism must be on the site of the proton motive force. And that probably some leakage from the cells is being generated by the single molecule protein. Now the next slide simply tells you a little bit more graphically what the proton motive force must be. It can be generated from ATP through the ATPase or from electron transport. It activates motility in the bacterial cells. It activates active transport. And it is involved in chemotaxis, but that’s not to be discussed here. Now what is the proton motive force? Since I had to learn it a few years ago from reading Mitchell and now students fortunately can study it in text books which are beginning to incorporate something about Mitchell’s theory. The next slide. I put something very elementary. The proton motive force is simply an energy charge across the membrane which consists of 2 components: a membrane potential and a difference in pH, that is immense potential for protons. And the 2 together amount to a force, to a potential of 200 millivolts. Now all of these, according to Mitchell’s theory, are generated because the respiratory chain or ATP through the ATPase, separately and more or less independently, can expel protons from the cell, creating a proton gradient. And the proton gradient across the membrane generates these 2 elements. And it's a membrane potential which is of course in existence already in the cell. But it is added on to the membrane potential. And in addition there is a gradience of protons which tends therefore to bring protons back into the cells. And it generates 100 millivolt energy charge at pH6 and that of course is dependent on the external pH. Now the question was at this point, can one measure this? Can one find out how one molecule of colicin affects this proton motive force? If it is true, our hypothesis, that this is where it affects it. And the next slide shows how methods developed in any number of laboratories but not in our own, although we used them passively, is that you can measure separately the delta side, that is the membrane potential in a bacterial cell, in which we cannot introduce a microelectrode because the smallest microelectrode is bigger than the biggest cell of E. coli. Therefore we had to do it chemically. It can be done by taking lipophilic cations, radioactively labelled, and then find that these cations are going to partition themselves in and out of the cells depending on the membrane potential. Because of course they can pass through the lipid bilayer. And if the membrane potential is high or negative inside, then they will accumulate here. And they come out when the membrane potential goes down. As far as measuring the pH inside the bacterium, we cannot put a pH-electrode - they don't make them this small. But we can use a radioactive weak acid for which the cell doesn't either transport or a metabolism. And we use for example for E. coli butyric acid, which has been used by many other people. And butyric acid will partition itself according to the difference in pH. The higher the pH here, the lower the concentration of hydrogen ion. Therefore less of the substance will be in this form inside, because in order to come through it has to be protonated. So these are just the tricks that we learned from the membrane physical chemists. And applying these tricks - the next slide - shows that when you add one of these lipophilic cations to a bacterial cell, it accumulates very rapidly. You add the colicin and it goes out. And the membrane potential decreases from about 110 millivolt to less than 10 millivolt in a matter of minutes. So it’s clear that one result, one consequence of the colicin molecule, is to abolish the membrane potential. The next slide is a little bit more puzzling, but we know the answer. This is actually printed directly from the computer. And the computer has been sort of unkind enough not to leave out the set of points that don’t fit the curve. But you have to live with the computer and you have to be honest. In fact, my secretary who was preparing the slides suggested that I have another side made without these 3 points, but I resisted the temptation. But anyway the curve is good enough to show that colicin added to the cell has practically no effect on the delta pH. But that other substances, which are known to be proton conductors, caused a decrease of the difference in concentration of ions. And this turns out to be a very simple reason. As I’ll tell you in a moment, before ending, is that the cells do in fact lose the membrane potential and they do in fact become also permeable to protons, so that protons do pass through. And the reason we don’t see it is because these cells, the rate of penetration of protons through the channels produced by the colicin is much slower than the rate of production and expulsion of protons through the electron transport system that the bacteria have. So that the bacteria are pumping out protons about 20 times faster than the protons can leak inside the cells. And so that the system is working nicely. Now at this point we have to ask the question. This is all very nice. We know that the colicin abolishes the membrane potential. But the question is what does it do. Does it act on the cytoplasmic membrane? Does it act on the outside? Does it disrupt something that we don’t know? Or something of that kind. At this point what we did, obviously, is we went to a simple system. And the simple system is to use liposomes. The next slide please. Liposomes, vesicles of phospholipid which are produced by a variety of ways, sonication, drying, sonication and so on. And they form vesicles that have a single bilayer of phospholipid around them. And these were produced using phospholipid directly from the cell, from the bacteria Escherichia coli, and in approximately the same proportions in which they are present in the cell, about phosphatidylethanolamine, phosphatidylglycerol and cardiolipin in appropriate proportions. The interesting thing is the reason, why are they so uniform? The reason they are all so uniform is because my colleague Doctor Kayalar, who did this work with the liposomes, was asked at the meeting whether his liposomes were really all nice and uniform and large. And he didn’t have a slide. So when he came back he made many slides. And we chose the one that shows them all approximately like the same and all very large. If you looked at some other slides, you would find that there are some somewhat smaller. But these are approximately an average of about 600, 700 millimicron in diameter. And the smallest one was probably about 400. So they are really nice. There is a reason for telling you that. And the reason will come when I talk with the last slide. But anyway these are nice liposomes. If you make liposomes in a solution that contains one or more radioactive substances, and then you pass these liposomes through a column or you dialyse them through a medium, you can get rid of the external radioactive substance and you have liposome loaded with radioactive substances suspended in a medium without. So by measuring the radioactivity outside the liposome it’s very easy to see if anything comes out. It’s a little bit more complicated but we have done it to see if anything goes in. And the next slide now shows what happens if you take liposomes that have been made with radioactive rubidium and radioactive sucrose, both of them in the same solution, and you add the colicin. You see that rubidium comes out almost completely, because when you disrupt the liposome very little remains. I don't know how to point here, I'm not tall enough. Sucrose comes out at first more slowly and then even more slowly for quite a while. Now the exit is not simply disruption or a formation of a rapid channel. If there is a control you take the same kind of things but you add here valinomycin. Valinomycin makes an immediate channel for rubidium or for potassium, and they immediately equilibrate. Whereas if you add colicin K, you get again the same kind of slow exit. And what is most interesting, if we think in terms of the natural channels, is that for each substance we study we find that the channel produced by colicin has a specific rate of exit. Which means that this channel produced by a single molecule of colicin has a conductivity which is different for rubidium. It turns out that it’s about ½ if you measure it for sulphate ion. It’s similar for hydrogen ions, and similar to that for sulphate ion. About ¼ for sucrose and much shorter, much lower but still definite for substances like glucose-6-phosphate. And we can measure up to a polymer, a sugar polymer of approximately molecular weight of 750, and it comes out very slowly. Whereas inulin, a molecule of about 4,000 molecular weight, does not come out at all. And I may add that in the case of several other substances that we have studied, they seem to fall into, what I will say approximately, certain classes of conductances. Now the question is, is this really due to one molecule of colicin? Well, that was done - unfortunately I don't have a slide. But it was done in the case of rubidium by taking this initial part of this curve for various concentrations of rubidium and measuring very accurately. And Doctor Celik Kayalar, who is now at Berkeley, has found that the dependence of this initial slope on the concentration of colicin is exactly let’s say between 0.9 and 1.2. Definitely first order reaction. So a single molecule forms the channel. Now the question is, how does it form the channel? How does it go in? And here we have a puzzle which I think will be of interest to other people in membrane biology – if there are some present besides the neurophysiologists, who deal with much higher forms. Some time ago a group under Albert Einstein, Doctor Finkelstein and his co-workers, got some colicin K from us. And they placed it in a system in which you have an artificial membrane, one of the Montal-type, in which 2 monolayers of phospholipid are brought up against one another. And these are connected let’s say in 2 cuvettes into which you can put electrodes. And you can connect them through a voltage CRM system. And you can find out whether the channel, what is the conductance for sodium or potassium ions of the channels, if any. And whether the channel is open or closed immediately when you add the colicin or whether they require an electrical gating. The next slide which is the last shows these 2 systems. Here are the liposomes in water. And here is the type of artificial membrane produced by having 2 monolayers of phospholipid trapped in this little hole. And you have a system here which you can change the polarity and you have a voltage CRM that you can measure recurringly. Now the remarkable thing is that on this system they found that the conductors for potassium and sodium through this thing, produced by adding the colicin on one side or on the other, is definitely gated. That is in order for the sodium or potassium to start going through, you need to apply a potential difference of the order between 10 and 20 millivolt – about 20 millivolt you have open channels - and they close back and forth. So we said we have here the sodium channel, the potassium channel, marvellous. Except that when we went to do the same experiment with liposomes, it turns out that the channels produced by colicin are not gated at all. How do we know? Well any of you who are membranologist would know. All you have to do is to have different potassium concentration, add some valinomycin, and you change the membrane potential. And we changed the membrane potential between minus 100 inside to plus 100 inside in steps. And at each step the additional colicin made no difference at all. The additional colicin continued to make the same kind of channel conductances as it did at any of the other voltages. Now here we have a peculiar situation in membranology. This is not the first case. For example the outer membrane of the bacterial cell, the matrix protein, the channel, has been purified, as I said, by 2 laboratories. The laboratories of Nakae and Nikaido in Japan and in Berkeley find that when they put in artificial membranes that the channel's conductance is completely non-gated, independent of electrical potential. Whereas Schindler and Rosenbusch in Basel find that the same protein put in the same membrane, but in this form rather than in liposome, in this kind of set up, you find the gated situation. And this is also true for the outer membrane channel protein from mitochondria, although that has been done in a very partial way. So here we have a situation in which the simplest arrangement gives us a very simple answer. A more complicated arrangement gives an answer which in my opinion is probably completely wrong. And I mean, apart from the fact that in a more strictly professional audience I would say be careful before you use that kind of membrane. This reinforced my own conviction ever since I started doing experiments in the middle of the 1930s. If you can do something in a simple way without any instrument, do it that way. And you are likely to be more right than people who have big powerful apparatuses. Thank you very much. Applause.

Salvador Luria (1981)

Membranes and Transmembrane Channels

Salvador Luria (1981)

Membranes and Transmembrane Channels

Comment

It’s a pity that Salvador Luria lectured only once in Lindau. He must have been a brilliant teacher with a talent to inspire young researchers. His first graduate student at Indiana University where he taught between 1943 and 1950 was James Watson, and the Center of Cancer Research at MIT, which he was asked to set up in 1972 and directed until his retirement in 1985, generated four Nobel laureates, namely David Baltimore, Susuma Tonegawa, Philipp Sharp and Robert Horvitz. “Salva was a visionary who protected his young faculty from unnecessary interruptions, thus allowing their research programs to flourish in an ideal scientific environment. He was also a role model for how a scientist could shape and lead a community“, his early MIT recruit Philipp Sharp later recalled Luria’s excellence.[1]

In cooperation with Max Delbrück and Alfred D. Hershey with whom he shared the Nobel Prize in Physiology or Medicine 1969, Luria “set the solid foundations on which modern molecular biology rests“, according to the Karolinska Institute[2]. Working with bacteriophages - viruses that infect bacteria – Delbrück and Luria had demonstrated in a famous experiment that “resistant bacteria arise by mutations of sensitive cells independently of the action of virus“[3], i.e. that bacteria are affected by natural selection. This knowledge laid the foundation for both explaining the development of antibiotic resistance and for discovering restriction enzymes, the major tools of genetic engineering. When molecular biology began to flourish at the end of the 1950s, genetic analyses bored Luria however. Rather than being involved “in putting together little pieces of a large puzzle whose overall features were already evident", he wanted to dig in “unplowed fields“. He decided to work on cell membranes.[4] He investigated the effect of certain bacteriocins and discovered that these colicins exert their lethal effect by opening ion channels in the cell membranes of bacteria. These findings took already center stage in his Nobel lecture and still are in the focus of his interest in this lecture in Lindau.

Luria distinguishes between intelligent, neurotic, less intelligent and stupid membrane channels, and complains that they all share one unpleasant feature: “The proteins of these channels have to insert themselves into the membranes, and the biochemist who works with them finds them more or less distasteful, because they generally are poorly soluble in water”. In this regard, colicins are a remarkable class of proteins, he says: “They attach themselves to certain bacteria and kill them – by exactly what a biochemist would like to do, by inserting themselves in the phospholipid bilayer and creating a hole whose conductance is relatively non-specific and allows substances to pass with a molecular weight of up to 800 or 900”. Amazingly, one single molecule of colicin can trigger in one bacterium a whole series of physiological effects from arresting the synthesis of macromolecules to blocking active transport mechanisms and decreasing ATP levels. Referring to Peter Mitchell’s (Nobel Prize in Chemistry 1978) chemiosmotic theory, Luria suggests that colicin influences the proton motive force across the bacterial membrane. But can one measure how one molecule of colicin affects the proton motive force? What does colicin do to abolish the membrane potential? And what are the best experimental systems to explore that? These are the questions Luria discusses in the second half of his talk, which he concludes with an advice from his personal experience: “If you can do something in a simple way without any instruments, do it that way and you are likely to be more right than people who have a big powerful apparatus.”

Joachim Pietzsch

[1] Cf. The Salvador Luria Papers (National Library of Medicine: Profiles in Science. http://profiles.nlm.nih.gov/ps/retrieve/Narrative/QL/p-nid/164)

[2] http://www.nobelprize.org/nobel_prizes/medicine/laureates/1969/press.html

[3] Luria, SE and Delbruck M. Mutations of Bacteria from Virus Sensitivity to Virus Resistance. Genetics 28, (November 1943): 491-511, 510.

[4] Cf. The Salvador Luria Papers

Cite


Specify width: px

Share

COPYRIGHT

Cite


Specify width: px

Share

COPYRIGHT


Related Content