Roger  Tsien (2011) - Engineering Molecules for Fun, Profit, and Clinical Relevance

I want to tell you today about some of my history and some of what we are most interested in now. I'm hoping I'm trying to pitch it to what young people typically most want to know about becoming a scientist, with some examples from my own career and some of what we are interested in now. If I have a theme it's that we are molecular engineers and that distinguishes us a little bit from the traditional sort of biological and physiological research which is based on trying to understand how do natural bio molecules and biological systems work. And the great advantage of that is that anything you discover that's true will last forever. And is something that we deeply understand about nature. Likewise in the chemical world the most prestigious chemical research, at least when I was a graduate student and post-doc, was the field of how to artificially re-synthesise natural bio molecules. So again people went into nature, found molecules that had complicated structures and proved whether or not that they could re-synthesise them. Typically by means other than the way the biological organism actually made them. But instead that didn't really grab me personally. I came to the preference to work on this question of can we design actually our own new molecules. And, of course we then have to build them, that will perform amusing and useful functions in biology. So this way we have enjoyment of working on biological systems which are still the most intricate and fascinating. And yet we have some creativity where we design molecules instead of simply trying to work out what nature has already figured out. And therefore it's a little bit like architecture or sculpture, and it has an ascetic element then it also has a less competitive element because no two sculptors will ever make the same sculpture. No two architects will ever design exactly the same building. Whereas you can have five groups all trying to understand a particular biological phenomenon and in principle there is one relatively correct answer. The first person who gets it wins the Nobel and the other four are consigned to the ash heap. That is I think too competitive for the likes of me. So there's the advantage when you create something you are not in such direct competition. This was even understood by Richard Feynman the famous physicist who said "what I cannot create I do not understand". That's the real test of when you have understood the molecules is when you can make something new out of them and because up to that point you are just rational. You think you understand and you say oh biology works this way but the real proof is that you can do something brand new with it. Not just explain after the fact. Traditionally it was assumed that only drug companies had the resources to, the multi- large groups necessary to do this sort of work, but I didn't care. I wanted to try this in my own lab. Now you might ask why? Well a lot of it in the end you have to figure out what you personally are like. And one of the features about me, and I've been criticised or commented in previous Lindau meetings that people didn't like me psychoanalysing myself but hey, you got to do this in your career if you want to be serious. You want to think what are you good at and what are you not good at. And in my case I happened to come from a whole family of engineers, my father, two brothers, two uncles, a bunch of cousins and yet I felt that biology had the most interesting questions. I didn't want to be an electrical or computer engineer. It also happens that I have been fascinated since childhood by pretty colours and therefore by the means that you make pretty colours. But one of the most important aspects is that I'm the youngest of three boys and it's characteristic that the youngest children have to find an ecological niche that's distinct from their parents and elder siblings. This is the subject of an entire book here called "Born to Rebel" by Frank Sulloway and any of you who are youngest children or any of you who are older children and want to understand your bratty little brothers and sisters, you might find this an interesting book. And it's not surprisingly written by a youngest child, Frank Sulloway admits that that's what he is. And his shining example is Darwin who is the youngest in his family. And Darwin therefore was not going to inherit the land in the British aristocracy. He was not going to be sent to the army. Traditionally the role of the third son was to be sent to the clergy, to become a clergyman and that's what he started his education. And then he found he really wasn't interested, or that interested in preaching. But much more interested in bugs and beetles and insects and animals and rocks and so on. And that set him on the path of natural history and that's partly because psychologically he needed to break away from the mould. And that's all what Frank Sulloway is writing about there. So I'm afraid a certain amount of that came to me that my elder brothers and father were regular engineers, or at least they started that way, so I had to do something different. The one thing they all hated was chemistry that's maybe why I ended up doing chemistry. And furthermore I found that at least when I was starting inter-disciplinary barriers prevented most biologists from understanding chemistry. They knew about molecules to the extent of being able to build a linear alphabetic string, you know twenty amino acids, make up of proteins and four make up DNA and one extra makes up RNA. And other than those twenty-five letters most people in biology couldn't do chemistry that was deviated from those linear building blocks. And that meant big opportunities and few competitors and that's another theme I don't particularly like competition. And so I'm always trying to get away from it and find some area that's relatively under-populated where a little bit of work can produce a rather large bang, we hope. And so I'm going to cut past a lot of the early stuff we did for lack of time and go onto the beginning of our work on GFP, Green Flouescent Protein, which I have to mention because of course that's what took me to Stockholm and eventually to here. And this is a video of the jelly fish and I'm going to try to reproduce for you now, I think this is the only video I know, despite searching the internet, this is the only video of the actual glow of the jelly fish. And I did it after the Nobel because I found that I needed it for talks like this. Now I had to go all the way up to Alaska because the jelly fish is no longer available in Friday Harbour where Professor Shimomura actually discovered it, because of pollution and so on. And run off or maybe global warming or a mixture of causes it's basically extinct at Friday Harbour. So I had to go up to Alaska and let's try to run this movie. This is a beaker with a jelly fish in it, poor little jelly fish is trapped in there and you actually can see it by this glow which is we are shining a UV light on it. And this is the glow of the GFP in the ring around the periphery of the jelly fish. In a moment we are going to, after a little bit of hemming and hawing we have to turn out the lights so then we can poke the jelly fish to make it glow by bioluminescence. But of course in the dark it's a little hard to poke it so we had to practice making sure that the stick was in right place. You can hear a bit of recorded sound, you can have the sound up a little bit, and jelly fish is slightly rotating here. Get the stick in place, turn off the lights. So a real demonstration such as it is on film, one of the rare chances I get to do science...ha-ha. Okay, so that's the initial phenomenon not awfully impressive and out of that was built the three shares of the Nobel Prize. Plus a small revolution in biology because the protein that changes the colour of the flash from blue to green is the green fluorescent protein. And that protein was cloned by Douglas Prasher and then shown to work by Marty Chalfie in other organisms. And here for example you can make a mouse that's glowing fluorescently throughout its body. Because by putting the DNA for the green fluorescent protein into the mouse now the mouse and all this progeny once we've incorporated it in the germ line now the mouse is permanently green. Just for those of you who are interested in greenness of chemistry this is encoded by DNA, not toxic, synthesisable by any organism that can tolerate air. And it does so in 230 sequential compensations roughly. One cyclization, one or two autoxidations goes on 100% yield in aqueous solution and the only by-product is hydrogen peroxide. And so that's pretty good for so-called green chemistry. We changed its colours and fixed up the brightness of the green from the jelly fish, made a cyan, a blue version, eventually a yellow version. All of that with the help of the crystal structure which I guess I didn't have in here but it's familiar to most of you as an eleven stranded beta barrel almost perfectly cylindrical. We've used it now to do all sorts of biochemistry inside cells. But we want to not just tag proteins that's sort of 90% of what most people want to do. But we particularly from my past interest wanted to see what the inter-cellular biochemistry was doing. And so we devised ways of making sensors artificially in which say we would tie a cyan fluorescent protein and a yellow fluorescent protein together with a protease sensitive linker. And then we would capitalise on the quantum mechanical phenomenon of fluorescent resonance energy transfer or FRET. Where if the CFP and YF P are in proximity they talk to each other. The cyan protein when excited transfers its energy to the yellow florescent protein. But if we break that linkage in the two halves of the molecule drift apart then we disrupt energy transfer and we go back to omitting cyan. This can be used to measure calcium if you put a calcium sensitive linker there you can look at cyclic AMP by grafting this onto the cyclic AMP dependent protein kinase. Ed Fischer is here and he shared the Nobel for some of these important discoveries about protein kinases. And in general phosphorylation we can also make sense of it this way by using phosphor amino acid binding proteins together with protein kinase substrates. I'll show you just one example that's all I have time for involving a calcium change. And this is looking at the early development of a zebra fish embryo where red means high calcium and the red pre-stages and predicts every cleavage furrow. So just before the cleavage furrow happens you see this wave of high red that is the high calcium, which then triggers actomyosin like contractions that then pinch the daughter cells apart by tightening the belt and constricting them. And this goes on and on over minutes past fertilization, this of course time-lapse and after a while the divisions become essentially a synchronous and so on. The first one was the most spectacular. Because it was a nice big single equatorial belt. This is the work of my former post-doc Atsushi Miyawaki. So this is just an example of the fun types of things you can learn biologically, but I picked it of course because it's a pretty movie. I told you that I loved pretty colours and it was always very frustrating to me that as a kid my favourite colour was bright red and that the green fluorescent protein refused to give us any reproducible bright red colours. It got us cyan and blues and yellowish yucky green. But never a good honest red. And so it was a revelation when a Russian group showed that corals which are actually distant biological relatives jelly fish, very distant. But they of course have many colours but what nobody had guessed before then was that some of those colours are exactly homologous to the jelly fish GFP. And in particular they picked out this protein from discosoma that they called DsRed, for its bright red colour. But they didn't know what the structure of the chromophor was or how this protein became red. And this was something we managed to contribute by doing mass spec on protein, on hydrolysis and so on. Here we've actually drawn for you the structure, the core structure of the DsRed chromophor consisting of a hydroxybenzylidine imidazolinone coupled to an acylimine. That's quite a mouth full. But here is it drawn out using the actual bacteria that expressed DsRed, using them as the ink. So we have some dextrous people in the lab who can use sterile toothpicks and pick up bacteria, draw them on the petri-dish in the scale of about 1 angstrom = 1cm on the plate. And so you both get to see the beautiful red colour of the real protein together with the structure of the chromophor. And it turned out to be not so difficult to change the colour of this DsRed to a wide variety of colours that pretty much filled out the rest of the spectrum. You see this is what we managed to get out of the jelly fish. And with the coral proteins we were able to diversify them into this whole family and the only question left was what to call these guys. There were so many of them all at once that we had to come up with a more interesting nomenclature. And after a lot of argument in the lab we came up with these notoriously fruity names where the name of the fruit gives you a little bit of a mnemonic for the colour. Like Honeydew, banana orange, tomato, tangerine, strawberry, cherry etc. And again these have proven fairly popular but of course the field keeps moving on, this was 2004 and there are many, many other flavours now available. But just to give you one example of the use of these different colours. Here I've taken an example from the monitoring of the cell cycle. Using the cell cycle dependent synthesis and breakdown of certain proteins and you'll hear more about the mechanism of breakdown from Prof. Hershko and Prof. Ciechanover later this morning so it's actually appropriate. These are cells which are filled with a pair of proteins and we arranged so that one protein that's basically greenish, it's actually yellowish green is made during mitosis. It's made during the G2 phase of the cell cycle, G2S. And then it's broken down during interface. There are known trigger signals that tell a protein make me now and break me down at another stage. And that's how we control the cell cycle as Tim Hunt got his prize for. And then we've another set of proteins that do exactly the opposite, majoring inter-phase and broken down during the active dividing part of the cell cycle. And we use those signals to control the red protein, which actually is M cherry. And so during the cell cycle when the cell is actively dividing it would be green because it's making green and destroying red. And then when it's in inter-phase it will be red and not green. It's just like stop lights. Atsushi Miyawaki could have picked it the other way round it's totally arbitrary which ones you make and which ones you break down but it's easier for all of us who are used to traffic lights to make green being go and red being stop. And so if this movie will play, you will see one week in the life of these two cells, as they divide and every time they divide they cycle between green and red, in this fun time-lapse way. And so this again indulges my love of pretty colours. And eventually most of them sort of get a little bit arrested by density dependent mechanisms and that's why they are mostly red near the end. Now this is just a stunt here, again it makes a pretty movie the real biological application of the cell cycle in large masses of cancer cells inside the body where you don't have the ability to watch them one by one in time-lapse. But here the biochemistry colour read out is tremendously helpful. And I take one last example from fluorescent proteins again of a visually spectacular sort. Many of you will know of the beautiful work that Jeff Lichtman who published a way they combinatorially turn on and off red, yellow and blue proteins by sort of rolling the dice on each one of them. It's like pulling a molecular slot machine and depending on how the tumblers settle you can come out with either all of them switched on. In which case you get a sort of white, or you could have predominantly green or predominantly one green and two red makes orange and so on. So there's this whole combinatorial gamut of colours that each neuron picks its own colour by pulling the slot machine independently of its neighbours. And with that you get these beautiful sorts of maps like here. Where all the cells in the dentigerus come out with different colours and the usefulness is that this is like the colour coding on wires that electronic engineers use. They use coloured wires to help them trace from here to there, you don't have to follow the detailed pathway if they match in this beautiful colour pattern. You can track them over long distances. That's just one little example. So that's perhaps my most successful, influential type of work scientifically. But I want to admit that just because you are a Nobel Prize winner doesn't mean you have some humiliating failures and I want to confess. For example one of our big failures which started with a good idea and this is an email I wrote to a man called Peter Hegemann all the way back in 1999, saying, Professor Hegemann I've been interested for some time in potential methods by which mammalian neurons might be transfected with genes whose product could permit light-triggering of depolarization and action potentials. And this really was not such a brilliant idea. We have long been interested in ways of stimulating cells with light not just reading out their biochemical activity. We had already done so with small molecules. Now with the power of GFP we are getting much better at reading say calcium signals or all sorts of biochemical signals in the cell and it became obvious let's try to do that genetically as well. And I looked around in the literature what is the simplest biological system where you can shine light and turn on a cell and that when you read the literature was probably green algae based on the work do Professor Hegemann who was a plant biologist. And that's why I wrote to him hoping that he could give me some of his genes that might act for us the way that Doug Prasher's gift of the jelly fish GFP had made our work possible now. I'm not a good gene cloner from scratch. We go and find interesting genes that other people may not have recognised the potential usefulness off and we asked them and Professor Hegemann was happy to give us some of the genes for the absence of vulvox in Chlamydomonas. Unfortunately what actually happened is that Rene Meijer a student in my lab spent two or three years trying to get those genes to work and the worse thing is that for one week it seemed to be working really well. And that later proved to be an artefact that we didn't know it at the time. I kept this going for a year or two trying to reproduce that result that eventually proved to be a waste of time. And so then when we finally gave up on these two proteins we went into the genome or Rene went into the genome and fished out two more plausible candidates. Better candidates than these ones and he started with ChR1 and he spent the year trying to make it work and he failed. And by that point he failed on three proteins - that's chlamyopsin, vovoxopsin, and ChR1, and he said Prof. Tsien, I need a new project. Because I got to get a PhD and so we started something completely different, we left ChR2 in the freezer and that turned out to be the right one. That other people have now taken in all over the place in neurobiology and I suspect in a few years we will get those people appropriate Nobel Prize including Hegemann and his collaborators. Rene did get his PhD by the way but he is discouraged by his little foray into biology he went back to being an optical instrumentation engineer. And he actually seems to be quite happy and not bitter at me for having bobbled his attempt that otherwise might have gotten him famous. So just because you think that, I know, I'm up here I must be imminent, no we make mistakes and foul up projects too. So I want to end with our most recent efforts which are to start apply molecular imagining to patients. This has to do partly with my wish in my declining phase to do something more clinically relevant than the work we've done up to now who's main clinical relevance was to provide tools for high through-put drugs screening, and that's okay but it's pretty indirect. And we didn't have much involvement. Also you may have noticed that what we are developing from the pure biologist point of view it can be dismissed as just techniques development. Oh, the Tsien Lab they just make techniques. Now from a chemistry point of view I think it's pretty interesting engineering. But from the biologist point of view and also prestigious journals point of view. But the great thing is that when you start to work on something clinically relevant the very same thought process suddenly becomes translational research. Bench to bedside and this is what the granting agencies think is the greatest stuff. And it's absolutely no different except that you put a veneer of clinical relevance on it. But you know it's not just a veneer, my father, my nephew, my PhD supervisor, many colleagues for example died of cancer. My mother had a stroke actually recently. I was diagnosed with cancer and fortunately I'm okay. And so all of this concentrates one's mind. The problem with fluorescent proteins is what makes them so powerful in biology is that they are codable by DNA. Which means that we can directly tap in all the power of molecular biology and make them sensitive to you know the fundamentals of life processes. Unfortunately the one place that fluorescent proteins become the gene transfer or genetics a handicap is when you want to work in real sick people. Now, it's one thing to work on stem cells that are derived from humans or human cells that are taken out of a person. But when you want to go back into a live person it is neither technically feasible nor ethically allowable right now at least to put say a jelly fish gene into you if you were the patient. Because you know now we have to ask what good is it to you personally. It may be very good for us as researchers, us, but is it good for you. And further more humans are two thick and opaque for fluorescents even though fluorescents has obviously been pretty good to me. So we need other techniques like magnetic resonance imaging, MRI is the subject of its own set of Nobel Prizes. It's great except that when we want to use molecular contrast it's not very sensitive, it needs a lot of the contrast agent to show up, those typically are Gd chelates. So what do we do? Well we study proteases in cancer because they are biologically important, they are actually responsible and necessary for metastasis, which is when cancer cells leave the primary tumour and go to distance sites. That's what kills cancer patients. Not the primary tumour genesis though of course that's where everything starts. But what kills people is this migration to distant organs. And that requires that the cancer cells acquire the ability to cut their way through normal tissue including the basement endothelial membrane that lines blood vessels, including normal extra cell matrix. And they use enzymes such as matrix metalloproteinases 2 and 9 to do so and fairly universally, perhaps you could escape that requirement but it's nearly universal. So we try to sense the activity of those proteases that enable metastasis because of its biological relevance. And also because it's amenable to the mechanism I'm about to describe where if you have a highly positively charged set of amino acids, a row of arginines for example they love to stick to the outside of cells by electro-static interaction. And then go in by processes that are still somewhat poorly understood. But this process as I just described is electro-static but very indiscriminate because essentially all cells have negative charges or negatively charged coating around their surface. So what we contributed beyond this known phenomenon is the idea of masking the positive charges with a non-stick backing paper made out of equal set of negative charges. And the polycadine and polyamine, the arginines and glutamates neutralise each other and keep each other out of the way until we cut the linker, which was shown here in green and we used the MMPs to cut the linker. So the tumour cells but not normal cells have the ability to tear off the backing paper and restore the stickiness, unmask the stickiness of the polycadine which then sticks to the tissue. So an example of what it looks like is here in a tumour that has been labelled with GFP, now this is an example of the usefulness of GFP, but this is something can only do in a mouse not in a human being. Human tumours never come glowing green, not in a patient at least. But we can inject our synthetic molecule which has a deep red dye on it and because of this mechanism I've just described it labels the cancer and makes it glow red. Because only the cancer can cut this linker right in the middle between the glycine and leucine unmasked the negative, throw away the negative charges; unmask the positive charges which then cause adhesion to the cell. As an important control if we remake the molecule with d-amino acids here. So we haven't changed the molecular weight. We haven't changed the hydrophobicity but this otherwise matching molecule is not an enzyme substrate then the tumour does not light up. Then we said we wanted to be able to work in magnetic resonance imaging and this is the evidence that we can. Here is a mouse with a tumour before we injected it and in this initial controlled state the tumour is not any different in MRI brightness from the rest of the tumour. But after we've injected the correct molecule which in this case is a little more complicated but has six of these peptides hanging off it now the tumour lights up. And this is something we hope can be done on human beings, you could inject the same molecule and it could be even done as part of your annual check-up. And we would see by anything glowing whether you had a cancer anywhere in your body because MRI can scan throughout the body. This is the control where the material we injected has got the d-amino acids and so it doesn't accumulate. And it's important to detect cancer early because if we could detect cancer at this sort of stage then when the tumour is localised we have a good chance of survival. Whereas if we wait till out here it costs a huge amount more to the health care system like a hundred times more but the prognosis is terrible and the cancer is generally caught too late. Unfortunately industry prefers to work out here because a) they can make a lot more money and b) it's a lot easier to enrol the patients when they are on the verge of death than when they are, most of them are still relatively healthy when you are trying to do the screening. So let me end up with one last movie or so which is showing you this in action and here I want to introduce the idea. We almost have a separate type of peptide that can stain peripheral nerve which is what we are most afraid of cutting during surgical operations. And here's the movie, in the beginning state you could hardly see where the tumour was. In the fluorescent mode here when we turn looking at fluorescent you can see where the tumour is but you can't operate. The best is to combine the two, this is white light alone can you see where the tumour is? No? Fluorescent you can but you can't operate but we use the computer to combine the views fifteen times a second. So that it looks as though the tumour has GFP but it's really not GFP it's our synthetic molecule. And now we've revealed, here's a big massive tumour remember because it was glowing green in fake colour, that means it's got the protease active. And here's the nerve that's been exposed but you might be ready to try to save the nerve and cut out the tumour by cutting along this line. But that would be a mistake. Because in a moment now when we check with the nerve peptide, where the nerve runs this is the missing branch, the hidden branch diverted by the tumour. This is the branch of the nerve that you couldn't have seen by regular white light. And in our overlay mode we colour it light blue just because it's another colour that's not in the body and now the job of the surgeon is made much more easy. Cut out the green stuff, leave the light blue stuff and the light blue is the nerves that you don't want to injure and there they are. So by mechanisms like this we can actually improve the survival after surgery in various tumour models and here's one where we happened to be able to raise the survival rate about a factor of five. Of course much more work needs to be done and my last slide is some lessons for young scientists. Try to find important problems that will either influence a lot of other research, answer important questions or maybe give you clinical relevance. But they have to be ones that give you some internal sensual pleasure or at least on the other side make you more comfortable with your neurosis. Including for example getting away from my older relatives. Accept that your batting average will be low but hopefully not zero. It's better to have a lot of ideas and you saw that in some cases we struck out like in the Channelrhodopsin, in other cases we were lucky. You have to learn to make lemonade from lemons, nature often gives you strange directions and learn to exploit them. Not necessarily stick forever with your original plan. I will mention by the way something completely non-scientific, but that if you want to say alive and healthy for longer enough to outlive your competitors at least scientifically I do suggest that you get regular exercise. And there's now scientific evidence that having exercise actually promotes the growth of stem cells in the brain and injects the pulses of growth hormones and so on. It's good for you, particularly especially in a meeting like this remember prizes are ultimately a matter of luck, you try lots of things not necessarily your most brilliant ideas, maybe the ones that are most successful do not be motivated or impressed by prizes because you cannot predict them and you cannot run your life by them. And just because so many people want to have our signatures or photographs and think that we are some sort of idol I hope to have shown you that we have a field to play just like everybody else. Of course it's absolutely crucial you have to find the right collaborators both junior and senior and you exploit them but in a kind way so that both of you benefit. And only that way can we do the best science. And so here with one of our final pictures painted with fluorescent bacteria are a lot of the people who were responsible for this work. Thank you.