 I would suggest that we get started right away for your sake. First things first, welcome everybody. I'm Eric Ljondal and I'm professor of biophysics in the department here, sitting out at SILIFE lab when I'm not travelling, which I tend to do a lot. But the next four or five weeks you're all mine, and that means that I'm all yours. So until what is at the end of April, you are my focus. First things first, I sent out an email yesterday, a couple of them over there, how many got them? What I'm going to tell you about the next hour might very well be basic concepts and possibly a repetition for you. I will even repeat things in this course. So what you're going to find me doing is I will kind of layer things. So today we're going to talk about very simple chemistry even. We might start to touch a little bit about physics. Tomorrow I will then kind of swap sites. And then we're going to talk much more about physics, free energies, and then after a while doing physics, we're going to come back to chemistry. And then we're going to start to, can we now interpret this chemistry with the physics we learned? And once we've done that, we might very well head back to physics again. Okay, based on that, can we take the physics one step further up so that superficially it might even sound like I'm repeating myself in the course. But the point is that every time we do this, we're moving one step up the ladder so that it's important that you know the basic stuff, because the next time we talk about amino acids, I'm going to assume that you know all the basic stuff about the architecture so that we can look at things more from principles. We're going to speak a little bit about basic properties of proteins, architectures, and then what keeps proteins together. And then I'm just going to introduce entropy a little bit. Because again, this is a great fun mix of physics and chemistry. So those, many of you might have taken this course. Do you know what this mantra is called? A structure to function. Good! Awesome! So this is something called the central dogma of molecular biology. And there's only one thing that you should understand from this course. This is what it is. Sequence leads to structure, leads to function. Everything since Watson and Crick in molecular biology builds on this. So sequence, those are your amino acids. A particular sequence will mean that you get a particular structure, for instance, of a protein. And the particular structure of a protein, this is a fairly boring protein, but it might, for instance, mean that it forms an ion channel that actually conducts ions in your nervous system. If you're changing the structure, you're changing the function, and the way you change the structure is by changing the sequence. There is also a feedback loop in this that has to do with evolution, of course, that if the function starts to affect the organism in positive or negative ways, that might mean that the random sequence alteration might change. But again, the central dogma of molecular biology. Sequence to structure is what you talked a lot about in bioinformatics. We're going to spend this course talking a little bit more about how we go from structure to function and what the relation here is. The amino acids, on the other hand, I do hope you've heard about. This might sound like a very corny physical introduction to amino acids, and it's kind of intentional. So amino acids are small, simple molecules, but the important part of this is that they're building blocks. So you hopefully know that there is a C alpha, there is an amino group, and there is a carboxyl group, and then you have almost always a hydrogen and then some sort of R group, which is variant. And they have a couple of individual properties, which again, this sounds like I'm repeating myself, but this is important. They are zwitterionic, and the zwitterionic property of amino acids means that even though a small amino acid like alanine, it's going to be neutral in solution, but it's neutral because it effectively has one positive charge in the amino group and a negative charge in the carboxyl group. And again, that is way more important than you think. We're going to come back to that when we talk about membranes. They are also all chiral, that I hope you know. Only the L amino acids are coded for in genes. First, are you all on top of what the chirality means? What does it mean then? Exactly. And so the point is that if you have a central atom, then with four different groups attached, and it is important that they are different, there is no way I can take this one on the left and rotate it to coincide with the one on the right, so that there is a sort of, well, they're mirror copies of each other, and that means that most chemical properties are identical, but their physical properties are not quite identical. One of them you call L, and the other one is called D, and that has to do with the order of these groups. Only the L ones are coded for in our genes. Why? So why is the enzyme shaped in that way? We don't know. It's random. We have no idea why everything is, we do not know why things ended up as L amino acids from the start. So you can think of this as a spontaneous symmetry break, right? The chemistry is in principle the same when you're looking at individual amino acids, but when the first organisms at some point started to have L amino acids, then as you say what eventually happened, we ended up having enzymes which are very large, complicated molecules that stitched these together. Suddenly those enzymes would only recognize the L amino acids, so this becomes a self-propelling force in nature, because we are L amino acids, and all our bodies code for L amino acids, we couldn't really introduce a D amino acid. But you could imagine a mirror life where everything was D amino acids, all the enzymes, all the genes of all the organisms. So I'm not going to say that it's a freak of nature. In the lab we can of course create D amino acids. Could you imagine why that could be useful? Yes? Sorry, say that again. But then I know already when I create it. So there is one very cool concept nowadays that we're going to come back to. When we're building modern drugs, we frequently try one of the coolest concepts in the pharmaceutical industry, it's something called biologicals. So biological drug, most drugs like Losec, Pylosec or anything, these are small organic compounds, they're not proteins. But on the other hand, proteins are really cool, really large molecules that you can get to do anything. So you could of course create drugs that are proteins. What would happen when you take such a pill? Your stomach would digest the drug, and then it no longer works. But your stomach is full of enzymes and it's breaking this down with enzymes. So what if we create a drug based on D amino acids? Then it would not be broken down by your stomach. So sometimes you can actually use these things that things are, normally it's bad not to be compatible, right? But in some cases it's actually an advantage not to be compatible with the body. And we're going to come back to this. But the reason why I spend this time, this is the first example how a physical principle, chirality, can actually be related to a biological function. Which is kind of the concept we're going to come back to again and again and again in this course. You have hopefully also studied a bunch of these amino acids, you can represent these in a bunch of ways. How many of you know all these amino acids by heart? Good, then we have something in common. I know them, but I would know them. But I would have to think hard about something. I know what a triptophan looks like. I know what a pheonanone looks like. If you start to ask me, let's see, isoleucine versus leucine versus valanine, I would have to think a couple of seconds before I count the number of CH2 groups. But the point is that it's not the number of CH2 groups that matters. Same thing here, you need to understand the principles. And the principle of these, you can of course represent this in lots of different ways. You can draw them in lots of different ways. And if you have a photographic memory and would like to remember the exact structure of each and every one of them, go right ahead. If 20 is not enough, there are a bunch of non-natural amino acids too, so that you can learn as many as you want. But what I would like you to know are the principles of the amino acids. You need to learn how to classify them. So there are some amino acids that are small and hydrophobic. Isoleucine, isoleucine valine. And that I would know instantly within one second. If I see it in a protein structure, it's a valine. That's going to be a hydrophobic site. On the other hand, there are amino acids that are polar, but not really charged. Serine, threonine, aspergine, glutamine. Well, that depends on a bit on the pH. They are likely water soluble. If you find one of those in the site, it's going to be a site that likes to bind molecules that are not hydrophobic. And then there are things that are electrically charged. Arginine, histidine, depending on pH, lysine, aspartic and glutamic acid. And if you start defining one of those in a membrane where things are normally extremely hydrophobic, it's likely super important. And again, aspartic acid versus glutamic acid, knowing that there is an extra CH2 group in the glutamic acid, well, that's a nice bonus, but it's not that CH2 group that's important. But you can't mistake aspartic acid for valine. And then similarly, there are a bunch of special cases like cysteine, glycine and proline. And you need to know why these are special. You need to know because when you find these in proteins, it's frequently super important. We're just about to submit a paper when one of my students has found some very important mutations related to proline in the iron channels. I'm going to come back to this in a few weeks. But that also means that there is not just one way to skin this cat. You can classify them based on whether they are small or large. That will depend on how much volume you have in the binding site. You can classify them based on whether they are hydrophobic or hydrophilic. You can classify them based on whether they are charged or polar. You can classify them based on whether they are chiral or not. It actually turns out one amino acid is not chiral. Do you remember which one? Sorry? Glycine. Why is glycine not chiral? It has two hydrogens, right? The side chain for glycine is a hydrogen. When you have two hydrogens, the requirement for being chiral is that you need to have four different groups. Because there are two hydrogens that are identical, you can rotate it to burn. Proline, if you want to be really corny here, proline is technically not an amino acid but an amino acid. I will not ask you about that. If I ask you about proline, you're more than welcome to write it, but it's not knowledge I expect you to know. But sit down and look at this. Spend an hour. You have no idea how many hours and months and years I spent just looking at amino acid and understanding them. It's not just a list of CH2 groups. You need to understand the different ways to classify these. And a lot of research, even in this department, is still based on understanding the differences between amino acids. What happens in a membrane protein if you're replacing a methionine with a cysteine or with a histidine? This is important for biology. Much more so than you think. And that brings you to the genetic code that I hope you've seen. This is another thing that I saw. I would actually cry if you learned this by heart. But for the reason is that you only have a finite amount of time for knowledge. Learning this by heart is a complete waste of time. There are some things that are good to know. Stop codons, for instance, recognizing that some of those are stop. And maybe a start codon. But again, knowing all the varieties that create Phelanalanine versus Proline versus Lucine versus Cytus and Lucine, forget about it. Who discovered this? Watson and Crick. But there is something stranger. There is a redundancy. So with four different bases, there are a total of 64 different combinations. Three of these actually turns out to be a signal for stop. But there are only 20 amino acids. So some amino acids are coded for by more than one codon. Why? Can you imagine anything that you could do with this? What it leads to? Yep. It could. That's useful evolutionary. But that's not really what I'm looking for. Primarily, what does it lead to? If you're looking at it. Which amino acids are going to be more common, do you think? So why are you... I'm not saying whether it's right or wrong. But why are you saying those amino acids? Why are you saying that it would be common? Exactly. So the number of codons... And this actually works. You can study this in nature. The relative occurrence of all amino acids in all organisms is roughly proportional to the number of codons coding for it. And that's why arginine with 1, 2, 3, 4, 5, 6 is going to be much more common than some of those like tyrosine with only two or so. And it's... I love to ask this question. That makes sense because you see all these complicated reasonings about based on hydrophobicity and membranes and everything. But this is just statistics. This is just counting. Why nature did it this way? We have no idea. But these are the building blocks nature needed and therefore it was created in these fractions. Amino acids also undergo this something called polymerization. I'm well aware that this is repetition for some of you but this is important. Amino acids are stable as individual molecules but under some conditions you can get two amino acids to bind together and the spice of NOH and NH that then becomes an H2O. Then we create a peptide bond between them. This was discovered a long, long, long time ago over 100 years ago by Emil Fischer that amino acids will create long sequences. And this is typically done with enzymes. If you want to go into all the details physics and chemistry here that we're going to do later this turns out to be a very special bond. Most of these other bonds you can rotate around but this is a very stiff bond. It's effectively almost a double bond between... well it is a double bond between the carbon and nitrogen here. So you typically can't rotate wrong around this bond. But we're going to come back and tell you what it is and what's appropriate. And the special part of proteins. The proteins are polymers. And you've all seen tons of polymers. This is polymer, plastic. Proteins are so-called hetero polymers and that they are polymers but in contrast to all the simple polymers they're not assembled by identical groups but they're assembled by different groups. The reason why this is important is we're going to come back and talk about this later on in the course and understand why does nature need to polymerize proteins? Why couldn't we build proteins from amino acids at single building blocks? And as you might guess already it's going to be all the details of this hetero polymerization that creates specific properties of specific proteins. The person to discover this was Frederick Sanger in 1952 and let's see I think I should have more paper. I will try to hand out a couple. I'm not going to ask you to I'm not going to ask you exam questions based on specific details about these papers. So why on earth do I hand out the papers? It's up to you. If you're reading these papers I find, particularly reading classical papers I find to be a really good way to get an idea how did people think when they first discovered this. So the point is not that you should know these papers by heart but it's a great way to learn about how they discovered this. So this is Fred Sanger's very first paper from 1952 when they sequenced insulin which was a tour de force at the time. It's less than 100 residues. You got the Nobel Prize for this a few years later. This is the basis of all high throughput sequencing and everything new that we can sequence entire human genomes. But at the time it was not even obvious that a protein is uniquely determined by a unique amino acid sequence. All these things that you find I would actually argue that most of the things are completely for granted and obvious. The reason why they're so obvious now is because somebody got a Nobel Prize for it 50 years ago. This was not obvious in the 1950s. And it was so not easy to find a way to determine the sequence of these molecules. So as I mentioned that these peptide bonds have very special properties. There is going to be an electron resonance on this so that the bond we create here is going to be a bond that you typically cannot rotate around, so called double bonds. In almost all cases for proteins if you have to make a bet they're going to be trans bonds. So that is the oxygen and the hydrogen here are in different sites. And that simply has to do with steric hindrance. It's very hard to have cis bonds. When you have a proline after this because a proline does not have this hydrogen for prolins it's actually quite common that they are cis bonds but it's kind of 50-50 for prolins. Completely horrible. You can't predict it. And there are even a ton of PDV structures where this is simply wrong. And I know because a French colleague of ours one of the proteins we're working on we realized just weeks ago that there is a structure in the protein data bank and the proline peptide bond has been assigned incorrectly. It should be translated as this. So again this might also seem very corny more theoretical chemistry and everything but understanding things like peptide bonds is going to be super important for functioning proteins. And it comes back to the sequence the fact that it's a proline leads to structure the fact that it's a trans or cis bond leads to function how it's going to create a binding sign for all that. There are lots of proteins in the world. How do you define whether things is a protein or just two, three amino acids? Titan is a gigantic protein that we're going to come back to in your muscles. 30,000 amino acids. There are some of these that are only 100. So when are things a protein? So that's right. So if you engineer a beautiful protein in the lab and it folds beautiful into a protein but it doesn't do anything I seem almost unfair against it just because it could be a really beautiful structure or if you have a mutation so that for whatever reason hemoglobin doesn't bind oxygen anymore is it no longer a protein? So people have thought about this for years. Physicists they frequently like to say anything that's small enough to have some sort of structure almost an alpha helix they like to call it a protein. A chemist, biochemist in particular will buff at that and say pff, that's a polypeptide. Protein has to do something. I would say that the definition somehow you need to have a buried part of it. There has to be some amino acids that are not really exposed to water. So that the smallest proteins we're going to find are going to be something like 25, 30 amino acids. And again, then I'm slightly more in the physicist side. There are lots of people are going to say that those are just fancy polypeptides. But you can imagine the number of ways you could stitch together titanium. Molecular weight 3 million. 30,000 amino acids. Each and every one of them matters. It doesn't mean that you can't have mutations but we can't randomly start to change things there then it will likely no longer work. So I'm still, even after half a career in this I'm still in shock and awe of how nature can do this. And it almost never goes wrong. If you think that computers are good or anything, computers break all the time. These processes they virtually never go wrong. There are no machines that come anywhere near what you have in the body. We're going to spend, not today but in a couple of lectures we're going to systematically go through proteins and talk about different classes and what they do. There are a bunch of fibrous proteins which are likely the ones that you don't think of as proteins. Lots of things in hair, skin and all these structures is actually proteins too. They're looking fairly boring because they're insoluble, strong. There can be lots of hydrogen bonds but they're kind of scaffolds, building blocks. We are going to spend most of the course talking about globular or water soluble proteins. Globular is pretty much just a keyword that they're roughly spherical. They can have a huge amount of different complex folds and it's a beautiful target to understand protein folding with. The peptide chain here always interacts with itself just, again, create some sort of fold and that's going to be a recurring topic for the course. Why do things fold? When do they fold? When is the fold stable and when is the fold not stable? And how unique are folds? And then it's our... 10 years ago this was our pet research area. Today it's probably one of the hottest areas. Membrane proteins. When I took a course roughly in these areas in biophysical chemistry and learned in the fall of 1992, I think it was, or 1993, Professor of Biophysical Chemistry. We didn't have a coursebook at the time so we had a beautiful long compendium that he had handwritten and then distributed. And then at some point he asked us to take out and strike over this sentence where it said that there are roughly 110 protein structures in the protein data bank because this has increased tremendously. There were actually 300 at the time and one membrane protein. Or two. Do you know how many protein structures we know today? No, not structures. So you're a bit high. Roughly a bit over 100,000. I think there are roughly 130,000 structures in the PDB but some of them are DNA nucleic acids. There are now over 1,000 membrane protein structures now. You have no idea how spoiled you are. Because again, trying to understand membrane proteins without having the structure is futile. So that suddenly we actually, we have the code, we can decipher the code, we can see what these structures look like. And it's, again, today we take it for granted that you have structures of all these channels. When I started as a PhD student, there was still not a single structure or nothing, always based on guesswork of what the structures might look like. And now we know how they work in mechanical detail. We're going to spend a lot of time, well, at some point these proteins assemble, right? And we are, I will not repeat this now, but we're going to spend a lot of time talking about what is where in these proteins and why do they get their function. In some cases, like hemoglobin, actually, you might end up having some sort of other group. It's called a prosthetic group, the heme group, that helps the protein achieve a function in this particular case, bind oxygen. So that effectively becomes part of the fold. And I think, yes, I do have a couple of slides to show you an even more complicated example. This is titanium, that super long protein I showed you. So the way titanium creates function is that, let's do it top down. So in the diaphragm, diaphragm, your muscles that you use to breathe, if you look at this in the microscope, you will eventually find these muscle fibers and muscle bundles. And in the muscle bundles, you're going to find a small fiber. They're created of a couple of hundred or a few thousand muscle fibers. If you get even stronger microscopes, you will be able to get the fibrils, which is just the keyword for a small part of the fiber. And in this fibril, you're going to start finding something called units called sarcomeres that are repeats. Again, I don't expect you to. The point is not for you to know these details. And if you start drilling further down into the sarcomeres, you will eventually find some small strange chains in the microscope. And if you were to get an electron microscope, you would eventually be able to see that those chains actually consist of proteins. So each small ball here is effectively a fairly large protein, not just fairly large. It's this three million molecular weight protein with 30,000 amino acids. And if you start drilling down in all the detail, it turns out that some of these domains actually consist of beta sheets. I can talk about sheets. And what this structure does is that you can effectively restretch it. So if you apply a force there and a force there, you can take this protein and pull it apart a bit. And when you release the force, it's going to pull in again. So this is kind of the spring that you will later on use with ATP and everything to regulate how we apply force in the muscles. So this force, ultimately, when you drill down far enough, is going to be related to forces on individual proteins. And again, this is not going to be a course of physiology, so we're going to not talk about musculature. But you see that it's hierarchical. And at this point, what determines this is going to be individual amino acids that determine whether the protein works or not. Sequence leads to structure, leads to function. Central dogma again. Proteins, we're going to look at them. One thing that you will realize in this course I'm going to try to show you some movies of proteins based on simulations. When I was roughly your age, this was almost unheard of. We couldn't simulate far enough. But today, we can actually see proteins in microscopic detail and understand quite a bit of them. And we now know that if you look at single domain proteins, they're very hard and that it's easy to crystallize. So if you just overexpress enough of them, you can grow a small crystal with this protein. And when you grow a small crystal of this protein, it turns out that we can frequently determine the structures with extra crystallography. For larger proteins, this is much more complicated that the individual parts or the domains of the protein might be fairly rigid. But think about your muscles, right? You can't really crystallize your muscles. They're large, floppy and everything. So to understand the structure of single parts of proteins we're fairly good at. Understanding this type of structure, we are really bad at. And this easily fools you. Because if you start looking at the scientific literature, there are all these success stories about in nature of science. And we understand so much about structure. And you can easily come away with the idea that all important science has already been done. There's nothing left for you to do. But the point is that there are no scientific papers about the failures. We don't publish papers on things that don't work yet. So the scientific literature always has and always will look finished. Because again, we only publish the things that are finished. In the next 20 years, I think that by far the hardest part is not just looking at larger proteins, but even looking at proteins that, for instance, take shape. An iron cell will have to open or close. At some point, you're going to have organelles, skin and everything in your body. That is also structure. But it's structure that goes way beyond the level of a single protein. And there are virtually no studies of this. Because until very recently, we haven't had the tools to do it. This is, again, a pet research area of mine. We're going to come back to it. Historically, the way to determine protein structure is x-ray crystallography. I'm actually heading there tomorrow. Magslab, brand new synchrotron down in the south of Sweden. And this is the strongest synchrotron in the world. So what you do is that you take a micro crystal, a very, very small piece of a protein. A protein that you even see the crystal here. So you have a crystal that might be a couple of micrograms or something fitted on the very edge of a needle. And then you take this protein and they radiate it with an x-ray beam while you're turning it around. And what's then going to happen is that based on the exact location of all the atoms in the crystal. Again, we're going to come back to this. You will end up with a pattern that in some directions you're going to have constructive interference that you amplify the light. And then you will get small dot there. And in other directions, again, depending on exactly where your atoms are placed, they're going to cancel each other. And then you don't get the dot. So effectively, we get a dot pattern. And from these dot patterns, you would then need to... Well, we see the dot pattern here even. From that pattern, you would then need to deduce that that's what it looks like. How do you do that? In principle, yes. So these dot patterns has to do with the reciprocals space. So that's effectively of an image of the protein in Fourier space. But there's one small problem here. So if you take... How many of you are familiar with Fourier transforms? I'm so not going to go through Fourier transforms. Actually, no, I can. If you look... If you listen to audio, an audio signal to you, it's music, right? It's a sequence of events. For instance, when you hear me speak. But you could also decompose my speech into frequencies. So how much is it at 100 hertz? How much is it at 200 hertz? How much at 400 hertz? So instead of looking at things as a function of time, you can compose this signal as a function of sine waves at different frequencies. So you can choose to represent either a signal of time or as a function of frequency. And this is the representation of a protein in direct space, that is, x, y, and z coordinates. And this is effectively a representation of the protein in... Well, it is a representation in the Fourier space or the reciprocals space. And that's essentially the frequency. So if this... Assuming that there's two angstroms between all these alpha carbons, that's a frequency, right? So how many signals do we have spaced roughly two angstrom apart? So this somehow describes the periodic patterns in something. And I'm not sure about you, but to me it's not entirely easy to go back. It even turns out that it's impossible to go back because if you take this protein and turns it into a signal that is a function of all the periodicity in it, it's actually a complex number. And again, we're not going to go through complex number in mathematics. You don't need to worry about it. Can you see any complex numbers here? So a complex number is both... something about both an amplitude and a phase. And now we're talking about deep matter. You lose the phase here. You only see... You don't see the direction here. You only see roughly how strong it is. So there is actually not enough information here to go back. So we can't do it. So somehow you need to create a model of this and then optimize the model to fit this as good as possible. Which is something we do with computers nowadays, but this was not always done with computers. Do you know what this is? It's a famous plot. Sorry? Yes. Or, well, yes and no. So this is Rosalind Franklin's first DNA scattering pattern. It's not the structure, right? You can't see the DNA there. But what she saw here, and again, I guess to somebody who has worked in the fields, this would not be obvious to me, but to them it was obvious when you saw this pattern, it means that it corresponded that you had some sort of helical structure in this. This is not obvious, but if you spend 10 years in the field, I guess it is obvious at some point. But again, you still don't get the structure. You just see that there has to be something spiral shaped here. And there has to be something that has to do... If you measure the distance between these dots, you can somehow guess roughly how far they should be spaced apart. And again, I'm not sure about you, but I so could not guess that. And that is the problem that you have to take a good model and see if your model fits this. You can't just take this, first they didn't have computers at the time, you can't just take pushes in the computer and ask to get the model. So that's why these samples were even lying around in offices for quite a while, until two junksters pretty much stole her results. So if you take these pictures and start to measure it, you can actually measure these layers. And the layer lines here, you can use this to derive a whole lot of what the structures looked like. The DNA double helix. That was pretty much Watson... They bet with... Francis Crick pretty much sat in his office and calculated because he was the physicist. And then he had Jim Watson sitting and building the molecular models. And they built model after model after model, until they found a model that, if the model is incompatible with the data, you need to throw it out and start over. Or if you find a model that seems to work, continue working on it. If you find boring with this, again, we only show you the successes. And science is not success. Science is mostly trial and error. So can you... There is another potential model that would also be there. If you take the DNA but turn it, do it on the other side, that the backbone of DNA, think of that as a spiral stairs. You have the backbone in the middle. And then you have all the DNA bases pointing out. Is that... It's going to be a structure that has roughly the right helicity here. It's going to have the roughly the right layering. And this was even proposed by Linus Pauling in nature. No, sorry, PNAS. My bad. Plan of the nucleic acid structure showing several nucleotide. He got a Nobel Prize, not for this particular paper. No, he has a Chemistry Prize and a Peace Prize. Is this bad? I think we have to convince it is. In hindsight, it is bad. But the point is, the reason why I'm showing you this, I think this is awesome. Because what did Linus do? One of the core concepts of this course. He built a model. So he tried to... This data is too complex. They didn't have computers. So can you sit down with paper and pen and think? This was an awesome model. There was only one problem is that there were other models that were better. And the hallmark of a good scientist is not that you make mistakes, that you don't make mistakes. The first one to say, when Watson and Crick's model was published, the first one to say that that is the right one. That was Linus Pauling. My models are wrong. And I don't think you pull the paper. Because this was an interesting model. It was roughly compatible with the data. It's just that their model, everybody was in shock and all about how good their model was. You make mistakes, but learn from your mistakes. And even if it turns out that you were wrong but somebody else was right, accept that. And that brings me to the second paper today. That I think you... And this paper you should read. We have forced you to read. This is how a paper should be. It's two pages. So I don't even need to staple it. This is Watson and Crick's paper on the structure of the salt that talks to your ribonucleic acid. It's kind of a boring title. By all points, you will never read a paper that's right. It is actually pretty good, too. I wish more papers were like this. Read this because we're going to talk about that a little bit tomorrow. There's a lot to learn about how modern science works and biophysics and everything. There is a PDF of the paper of Mondo, too. Yes? It wasn't bad. Not the information at the time, but there are two things with Watson and Crick's model that turned out to be stellar. I'm not going to tell you because they are present in this paper. So we will talk about that tomorrow. But there are two things that means that if you're a good scientist and life scientist, that means that this model is simply compatible with life. It's much better and more compatible with the data. So in the interest of encouraging all of you to read the paper, I'm not going to tell you now, but I will tell you tomorrow morning. This is a structure, just like proteins have structures. And we're going to come back and talk much more about nucleic acids later on. So I will not spend too much time on it now. And I think Samuel Messon does, too. So you have a backbone, which is pretty much a scaffold, which is called of phosphates and sugars. And then you have four different building blocks, adenine, thymine, guanine, and cytosine. And depending on the order in which you assemble this, we're going to create different contents here. When the reason why this paper is so corny is that when this paper was published, nobody was really sure what this molecule was useful for. Which, again, we're talking about something it lived 60 years ago. We did not know what DNA in the first place was useful for. And this gives you a bit of perspective on how fast research has moved. And everything that holds DNA strands together and everything, I will come back. Let's talk about that tomorrow. It's much more fun than when you know a little bit more. So overall, you will normally see DNAs paired in this double helix, where the bases belong together. That, of course, creates all these patterns that for reproducibility. You can take a DNA strand, pull it apart, and create two strands that are identical copies of the first one. You can also take a very much related molecule, RNA, that just has to do with whether you have an oxygen or not. RNA ends up being a completely different type of molecule. You have the same type of pairing between bases, but based on slightly different properties in the backbone in particular, RNA creates structures that are not double helical. Actually, if you're interested in structural biology, DNA is very boring, because DNA is just going to be that double helix. Depending on the amount of water and hydration, you can get that to form slightly different helices, but it's always a boring double helix. RNA is really cool. Again, when I was in your age, we couldn't determine the RNA structure. But in ribosomes and everything, there are a huge number of molecules where we have RNA structure in it, too. Sadly, I think the book doesn't talk about RNA structure, because it was too new for the book, too. This is a tRNA. But the first thing that we were able to discover, let me go back to proteins. The way we determine protein is actually that, just as a DNA structure, we can use X-ray crystallography to determine structures of proteins. That is another result that seems to be obvious today. Per roots and all, the first structure that they determined took 22 years. So you're around 20 or so now. Starting a project now and you will finish it when you're 45. Then you believe in your hypothesis, right? And again, in hindsight, it's obvious because they did succeed. But when they started, nobody else had done it. It was not clear, can you even determine the structure of a protein? Is there a unique structure of a protein in the first place? We also know how intimately connected the structure is to life and everything. That particular structure we're going to come back to, hemoglobin, one of the hallmarks. Today we know much more advanced structures, ion channels, doing various things. These were the first membrane protein ever determined. There is a pattern in all these structures that you likely see that there are lots of these alpha helices, and I showed you some beta sheets that are all of you familiar with sheets and helices? We will come back to them a couple of times. It's not because I don't think you know what a helix and sheet is, but it has to do with this reposition. Trust me, you do not know what a helix and sheet is. You have seen them, but fundamentally understanding what they are, that takes time and you need to look at them again and again and again. So what does a protein do to different things? What do glycine do to different things? Why are some things helical? Why are there no beta sheets and membrane proteins? We can understand all these things from physics, and that's why I'm going to keep going back to them. Let's see. I might take that here. So talking about the proteins here, when do you think that people determine alpha helices and beta sheets? When or rather how did we find out about helices and beta sheets? Yes, it's perfectly fine to guess. So one way of thinking about it is based on those first structure. It's the collection of coordinates. And by the time you have several thousand atoms, it's going to be very complicated to see this. So classifying things into sheets and helices, it's a great way to look at this here, that you can probably even start to see that there is some sort of overall shape here, right? If I just showed you a cloud of atoms, you wouldn't even see how they are connected. So classifying things into these building blocks allows us to take a step back, and I don't care if it's a valine or isoleus in there. It's a larger level building block. So now we're talking to take this step back, looking at things at a fundamental higher level. So once you have an x-ray structure, it's obviously much easier to classify things into larger building blocks, right? That would be a great hypothesis and it's wrong. Alpha helices were postulated before the first structure was determined. And the reason why I need to tell this by the break is that there is a certain person here that I kind of kind of disturb it, and that person was Linus Pauling. And this is not Linus Pauling papers, but this is a paper by Dave Eisenberg. It's a review article in PNAS a few years ago. Review articles I can also recommend because they're usually short. And good review articles are not written by a researcher wanting to pitch all their own results, and that happens. But review articles actually are written with the intention of teaching you something. Read them and learn something about not but history. So this is Dave Eisenberg who wrote a very short four-page paper in PNAS about the discovery of alpha helix and beta sheets. And this is also the hallmark of a really good model, right? Because as wrong as Pauling was for the DNA structure, he was so right with helices. So he sat down and just looked and looked and looked and looked at amino acids. And looking enough at amino acids, what are the regular patterns and shapes in which we could build amino acids? And then it turns out there aren't a whole lot of regular shapes. But it could come up with that we can likely construct them either helically or somewhat more linearly. And if the first structure he would call alpha and the second one beta. And the really cool thing is of course if you have a model and you postulate it and a few years later it turns out that they show up in those secretary structures. Then it's a really cool research result. Which he did not get the Nobel Prize for, but he got again a few others. Don't feel too sorry for him. So if you look at these structures and again we're going to come back to membrane proteins here. So this is an example of aquaporin which is a water structure. And in the membrane proteins we're going to have lots of helices that help create in this case a particular channel on the inside here. And Peter and Rod shared the 2003 Nobel Prize for chemistry for the discoveries of channels. Both the cases A channel I showed on the other slide and this aquaporin structure. There are others examples here and this I guess if you search Google enough you can find old slides that wrote 10 years ago. So I actually in my defense I had these slides before these people got the Nobel Prize. RNA polymerase that is the structure that starts to read your genetic information and extract information in DNA to the messenger RNA. And I was a postdoc at Stanford not in Roger's lab but his colleague Mike's. And Roger got a very well deserved Nobel Prize for determining these structures. You pretty much get the Nobel Prize for one paper. There is once you take that information you bring it on to this protein factory we're going to come back to this too. That's the ribosome which is a gigantic structure of roughly 50-60 chains of proteins and lots of RNA. And the ribosome will decode your RNA and create the protein. And then that's going to be a nascent chain that shows up. And that's when proteins are going to fold. Tom Steitz, Peter Moore and Vinky Ramakrishna got the Nobel Prize for determining the structure of the ribosome a few years ago. There are more membrane proteins here deep protein coupled receptors. When I was your age everybody said these are going to be so important but sadly there people had already spent two decades trying to crystallize them. So the sad part is that we would never be able to get structure of deep protein coupled receptors until we got structures in particular Brian Kubilke and Ray Stevens actually. So Brian got the Nobel Prize but not Ray Stevens and today actually that number is old. There are probably 50-plus structures now. These are this is nirvana for the pharmaceutical industry almost all drugs that you make today trying to target either deep protein coupled receptors or membrane-incentral receptors. So you can kind of start to see a pattern here. If you are aiming for to get a Nobel Prize in particular in chemistry, going into structural biology is a pretty good bet and historically that has always meant X-ray crystallography until a few years ago. So that there was an ugly duckling method that you could the problem with X-ray crystallography is you need to start by crystallizing the protein. And if you can't crystallize the protein you could never get a structure and then you just put the sample in the fridge and forget about it and then you waste to that PhD students' career. I'm not kidding here. There are careers that are broken by X-ray crystallography. And the problem with this ugly duckling method is you could try to use microscopy and image individual proteins. The only problem is that the resolution was not good enough and everything. Until roughly five years ago when there was a leap in technology with the new generation of detectors. Cryo-electron microscopy. And suddenly we jumped overnight from roughly five angstrom resolution which means that we can't identify side-chains to two to three angstrom resolution which means we see side-chains. And this changed everything. Thank you, Ramakrishnan, who got the Nobel Prize for their X-ray crystallography studies on ribosomes. They no longer do X-ray crystallography. They only do cryo-electron microscopy. We have out at Psylight Lab, and we're going to show you that later in the course, we actually got a set of these high-end microscopes and we're going to get a second one here from now on. Everybody in Sweden is jumping on this too. They're really sexy pieces of equipment. And there was another I have this paper in Nature, Well Worth Reading, another one of these review papers to pay, well, four pages. This is what's happening in Structure of Biology right now. And this we brought up already last year and what happened last fall. Richard Henderson got the Nobel Prize in Chemistry for this. Together with Joachim Frank and Jacques de Boucher. We're talking about cryo-EM in the course too. The difference between cryo-EM Cryo-EM is much, much, much easier than X-ray crystallography. All that stuff about Fourier transform and that's complicated. What we do with cryo-EM is that we just have a really good microscope instead. But we're going to need to be able to first we need to freeze your samples into a thin slice and we'll talk about that later on. And then we can't use a normal light microscope because what happens is the microscope eventually. If you start to look at very small things. Once you get significantly below the wavelength of light, you can't image things with the wavelength that's larger than the object. You might have heard about this super-resolution microscopy a Nobel Prize a few years ago. We will talk about that later. But what you can do is that you don't have to image with light. You can image with electrons. And in quantum chemistry you might know that everything is a particle depending on well they're always a wave in the particle but depending on the energy it has you're going to have either more like wave-like properties or particle-like properties. And electrons if we accelerate them to a few hundred thousand volts, they're going to be waves that have wavelengths roughly of one angstrom or even 0.1 angstrom. So then and one angstrom is good because that's roughly where protein structure is. So we can actually use an electron microscope the only problem is that it's pretty hard to detect these electrons and it's very noisy so all the technology development actually has to do with these detectors. And with this works it's great because you're going to get pictures of proteins. Note the plural form there because it's not just one protein you're going to get a picture of ten proteins hundred proteins, a thousand proteins ten thousand proteins, a hundred you're going to get a picture of maybe 1.5 million copies of your proteins from different directions. So think of X-rays, take an entire human and throw the human in an X-ray machine and take random X-rays from different directions and then try to determine how all the joints in your skeleton works from that. And a human is very simple compared to protein. So the problem here is that you get up with a gigantic problem with tons of data and then you need to get a structure from that and that's what we do in computers nowadays. We're going to talk a little bit about that later on in the course too. 20 minutes past 10 now. I have roughly 30 slides remaining after the break. So I think this is a great place to take a break. Normally I will always be around at break. Today I might have to sneak away and just take with all of the models on whether they need my assistance for any of the students that they're accepting this year. So I would suggest that we take 30 minutes and be back here at 10 to 11 and I will try to be back already in 20 minutes in case you have some questions for me. All right. I would suggest that I get started again. So one thing, I think I fixed this web access for two people during the break and in both of them the case was that you were already, if you are on the list there then you're registered. If you're not on that list then you're not registered and then it's very important that you contact me. But for both of these they were registered with their Stockholm University ID, SUID, email address. I don't care what email address you use. The only thing I care about is that I can reach you. So in this particular case I could register them manually and add them as students with another email address. So that if you still don't get emails from me tonight or if you particularly if you didn't get emails yesterday talk to me and I can add you manually or forward your SUID because in theory there could be some important things coming on that address. But I will make sure that you get that way. Structural biology as well as we talked about the reason why I bring up these things is that it's not noble prizes per se are completely unimportant for science. It's a fun bit of PR for science and everything. The good thing as a student to study is that they describe very much not necessarily what's happening in things right now but cool things that have happened the last 10 years and it's described where the entire field is moving and a lot of these things are not written in textbooks. Most textbooks for instance are going to say that actually I think most textbook structured biology would ignore cryoEM because it was such a niche technique that was not useful for anything. And of course today in science it's the opposite. We have some of the leading groups in the world give up an extra crystallography instead because cryoEM is so much more efficient for them. So the reason why I mention these noble prizes it describes what's happening in science rights at least the last 10 years. We are going to do a study visit later on in the course completely voluntary but I'm going to take you around at SciLifeLab and show you a couple of these techniques both cryoEM and other things happening in our labs. We might show you some of the computational stuff but it's probably not that interesting to just look at the computers. The x-ray part is a bit more boring because the only thing you will see in the lab is the over expression and purification of the proteins. The actual structure determination here for that you need a synchrotron. And if you think that occasionally we and others say that these cryoelectron microscopes are expensive because there might be 5 or 10 million euros. If you are comparing that with the synchrotron suddenly you realize how insanely cheap the cryoelectron microscopes because here we talk about billions of euros. Then you have to share it and go to a special facility. But to give you an idea of what's happened several years ago there were already in particular for membrane proteins people are really interested in cryoelectron microscopy. Can you imagine any reason for that? This is a hard question. So what is that we needed to determine a x-ray crystallography structure of something? We need a crystal, yeah. And then you need something to crystallize. Yes, it's more complicated and that's a membrane protein. Membranes, we'll come back to the internals of membranes but membranes are pretty much oil. How many oil crystals have you seen? You can actually force things to crystallize but it's hard. They absolutely don't want to crystallize. They hate to crystallize. And if we want to crystallize then we're going to need to violate them so much so that their structure, as you say, they will start to change. There are some tricks around this and people have been able to determine structures. Again, Bob McKinnon's and Peter Ager's structures were extra crystals. But becoming a crystal that very much goes against everything a membrane protein wants to be. And you've now combined this with the fact that these are some of the most important proteins for the pharmaceutical industry. Those G-protein coupled receptors, these are entirely rumors but I think that you're talking about billions of dollars, of course, of determining that first structure. And there are even pharmaceutical companies that paid part of this to have a one-year lead time where only they would have a particular structure. So, large pharmaceutical companies have their own teams that try to determine the structures of G-protein coupled receptors. Because the amount of money to be able to design a drug you need the structure of a protein. And the amount of money you will make from these drugs is insane. We're going to touch on that in the last part of the course but we don't have billions of dollars. The problem though is that cryo-electromyroscopy is potentially really tempting because you do not need a crystal. We would take images directly of proteins. But before roughly 2011, and that's the advances that the paper I passed around showed, the resolution you would have getting to something like if you nanometer resolution, that's easy almost anybody, you can do that in a low-end microscope. But at 5 nanometer resolution you're going to see something like this, you're only going to see the shape. Now we know what the structure is, so I placed the structure in here but the only thing you would see would be the grey shape. And I guess based on that grey shape you could not say the location of the helices is high chance, you're right. So cryean people are even joking about calling this for blobology. Now because at this level it is blobs and you can probably start to see, we can laugh at it that I do too, right. But it's knowing that there is a transmembrane domain here and then there's some sort of linker domain here. This is super important but of course you can design drugs just by seeing that grey shape. And then what happened is that we got to something like a few angstroms resolution, maybe five angstrom, and they were somewhere here. This is amazing because now you can actually start to see these are actually alpha helices but again that is the only thing you will get from the microscope. Can you place the side chains? In this case you probably can start to do it, but that is a really good structure. But this is the hard part and this is where you're going to need a computer or something to help you model things because I certainly can't say how the hydrogen bonds here are placed or how a single chain is rotated. But somewhere a few years ago we started to get this point where you can actually identify individual amino acids. Not all of them, but you might see a helix here and then you have something sticking out of the helix. It's a fairly large blob. And then you're going to need to start looking at your amino acids very carefully. And imagine that you something have a helix here something that's well, blob-like. And then you have something that goes like that again. So that is some sort of large side chain and then we know that we're going to have a helix and then some sort of large side chain again. And then you need to start looking at your sequences. And suddenly you find that you have a tryptophan and then lots of amino acids and then you have a tryptophan again. Then you have identified it, right? And then, of course, you can have computers help you do this. But then you know, oh, and you're going to need to look at the entire sequence and try to place this sequence in these electrons. So that's something that takes months to do manually here. But then we can start to build an atomic model into this electron density. And that is how you go from this picture to that picture. So this is now an atomic model that people have built into this. But this is a model. Just as Linus Pauling and everybody sat down and created DNA or protein models, today students and professors for that matter we're creating atomic models and making sure that they are compatible with the cryo-electron density structure. But then you can actually get a structure of a membrane protein that you cannot crystallize. And the reason why this is undergoing an explosion right now is that a few years ago this is Yifan Cheng who managed to determine the first structure of the TRPV1 receptors. And there were some of the first membrane proteins we determined with these new techniques. The reason why there is such an explosion now is that when these new techniques have appeared all the X-ray crystallographers that did what I told you before the break have this sample that doesn't work so you throw it in the fridge and save it for a rainy day. This is the rainy day. So now what we're having everywhere all across the world, people are pulling out these samples, they're just pulling them on a grid and then they're shooting electrons on it and if they're lucky it's a structure. So there are probably two structures per week now in high impact journals and that will have to slow down but it's crazy right now. But it's very fun to be in the middle of revolutions. This TRPV1 is actually a pain receptor and heat receptor too and it's activated by capsaicin in chili peppers for instance a very small molecule that binds to this and when it binds to this receptor your nerve system senses heat or pain. But this comes back to the diversity the reason why this particular channel can sense heat is based on the specific amino acid composition of the channel. And now we talk about quite a few amino acids of course because these are large structures and we don't even know this yet right that here we only talked about it from a structural point of view but throughout this course and in particular not tomorrow because tomorrow we're going to do more physics but we're going to come back to this and try to understand what these things do and what happens when we insert different amino acids and that's also why when we do that and when that is more boring physics free energy and everything remember what we spoke about today the reason we want to do that is because we want to understand the biology. Everything in this course we want to understand the biology. Physics for physics says it's great fun if you're a physicist but we're not pure physicists but to understand the biology we will have to drill down into the physics even if some of there might be more equations than some of you like. So if you look at this particular molecule how floppy or not is this? So this is a great thing by not having a white protein. I can draw. We can rotate around that bond we can actually rotate around that bond and to tell the truth we can rotate pretty much every single bond here apart from the ones that are yellow well we can rotate there too but that's not important. So there are lots of things that can rotate here and depending on the rotations of this bond even this small chain with just 5 amino acids can adopt a whole lot of different conformations. So for instance if you start rotating a bond here in the middle we change the entire left part of the molecule relative to the right part. The hydrogen is not going to be that important. And it turns out that these bonds some of the bonds here in the amino acids high chain are more important than others. Can you imagine which ones? So there are kind of 3 classes of bonds and this is not about their names so if you rotate that bond around the hydrogen absolutely nothing happens and I would actually say that even if it's a CH3 group it's purely local and it doesn't really influence anything else it's just a bit of noise we don't care. If you start to rotate say that bond it's going to rotate that entire group right? And that will still influence this the group is going to be placed in different ways but it's mostly local. It might if we rotate again this is not entirely linear so if we rotate that bond I might flop around the the tyrosine side chain here a bit but it's still fairly local it's not going to change the entire protein. On the other hand the bonds that goes in this main chain here if you start rotating those we're going to change the entire geometry of the chain. So these are the bonds in the main chain here are going to be the ones that determine the overall conformations your protein can adapt and once you have folded things into a protein there might we already hinted that there is a unique structure of the protein so once they're folded in a protein they will adapt one specific conformation but what conformation do they adopt and why and how? This was a very fundamental problem roughly around the 1950s because again we knew that proteins were poly polymers and polypeptide chains and what determines? There's a much more fundamental question than you think forget about specific details Is this determined by biology or physics? You could of course imagine the body having some very advanced enzymes or something that assemble your proteins in a specific way that's our most things when we build a house we just don't throw bricks in a wall and hope that this house erects spontaneously then we need to invest energy and to build a house in a specific way we want it then you could imagine that the body had to do that using ATP or whatever I don't know and the other alternative is I know this has to be based on the laws of physics and part of the reason why we have this course is that it is mostly not entirely but mostly based on the laws of physics but that is so an obvious result so let's think about this and let's go back to mathematics these torsions have names and you're going to hear this so many times in each amino acid there are two torsions the one before the alpha carbon is called Phi and the one after the alpha carbon is called Psi they're just names there are going to be more torsion names but if you look at those for one amino acid an obvious question is to ask ourselves is how many different conformations are there for an entire protein based on those two bonds so there are two premino acids and this comes back to this model thing at some point we need to start to assume and I will violently assume things I might be of one order of magnitude I can be of by two orders of magnitude but we're going to see in the end it doesn't really matter so if we take one bond and we have to measure the angle and there are 360 degrees to a turn and again if it varies by one degree that's probably not that important and at some point you can decide well how much does it have to vary to be important and let's just pick 10 degrees it's a completely arbitrary number but it's not 180 degrees and it's not one degree so if each of these torsions is sampled in 10 degree units every single torsion can be in 36 different states 10, 20, 30 all the way up to 360 degrees so that for each residue the phi one can be in 36 different ones and the psi one can also be in 36 different ones so that's 36 squared possibilities 36 to the power of 2 for one amino acid and remember that list of proteins I showed you the smallest list of proteins in that list was 104 residues the smallest one and if we again assume that each residues roughly independent of the others well there are 100 different residues so we should raise that to the power of 100 so multiply that number 100 times so that's roughly 36 to the power of 200 or 10 to the power of 308 that's a cool number because if you try to do that in python do that calculation in python you will see what happens so this number is actually too large you can represent that double precision on a computer so most low end calculators at least and probably even in python it's going to say that it's an error you can't calculate that so it's a pretty large number this is probably about the same it's probably more than the number of atoms in the universe and our idea or hypothesis right now it's only going to be one of those states that is the true native structure it's pretty insane how nature finds that from that go down to one if you look at the previous slide this might look like a simple problem it's about as far away from as simple as you got and it's insane that we know that this is governed by the laws of physics but it's going to be completely impossible to find that that's my well let's come back to that keep that thought for a second because this far we've said that proteins are unique right that they do one thing hemoglobin does one thing so whether it's one or ten it's not that whole collection of things but it's a very small fraction of those things that are native if you look more specifically about these hiding bonds we're going to come back to this too but there are these cis trans-isomerations which is just a fancy way of saying that they can rotate for the phion Psi bonds they are free to rotate so it's very easy for them to rotate so they will rotate even when you fold the protein there are other bonds like the peptide groups and the peptide group remember that I said it was double here right and that means that it's typically going to be stuck either in trans or in cis and it will normally never change normally a secure there it requires much more energy so it actually turns out that the number I showed you was an underestimate because I didn't even account for the peptide bonds here and we didn't account for the side chains and this is why I could just assume roughly ten degrees you could of course argue that it was an did I properly account for one state and everything but I underestimated it and yes we could argue is it ten to the power of 300 or ten to the power of 400 the point is it doesn't matter it's an astronomically large number and I think I think the Richard Feynman who at one point noted that based on what the world is doing right now and the modern finance and everything and how those numbers are doing those numbers are typically larger than astronomical numbers so when it comes to really large numbers there are a lot of natural numbers but the point is that there are some degrees of freedom here that can change under special reasons cis and trans related to this proline and there are others that will change naturally in proteins time whether the things involved prolias are cis or trans it depends a bit and it depends in the worst possible way that you can actually change it with some phosphatases and everything there are a bunch of molecules that can change a proline here to be whether trans before the proline or let's see yes either trans or cis before the proline and that will even create different functions for it so all the single most important degree of freedom are going to be these torsion degrees of freedom and now we start to simplify things we start to model you might not be aware of it remember that structure of hemoglobin your 10,000 atoms you could of course decide to treat this by thinking about the XYZ coordinates of each atom and that's certainly a viable way of doing it it's correct but you can have tons of very long lists of XYZ coordinates and you just look at them they're just going to be numbers so here we rather choose to represent protein structure by looking at the internal coordinates so what are the bond lengths what are the angles that's roughly with a hydrogen it's going to be one angstrom the angle here is roughly 120 degrees right the bonds they're just going to vibrate a little bit you can see it with spectroscopy but the bond is for a hydrogen the carbon might be between 0.99 and 1.1 angstrom I don't care about it in the terms of a protein it's roughly fixed same thing the angles here they might vary a few degrees but they're not really going to change in a protein but these torsion degrees of freedoms they are the one that actually will change and they will change even as a function of temperature and they're going to determine everything with proteins so in the terms of the degrees of freedom for a protein the torsions is where we should look we're going to come back to that I haven't forgotten your questions and it turns out that there are a bunch of ways we can measure how these spin to move in the lab and that's important because modulus are really good and useful but the problem is you need to test your model against reality so the easiest way is to do some sort of IR spectrum and the vibrational frequencies are in the order of multiple terahertz and these vibrations are usually based on the hydrogens vibrating relative to the oxygen you see beautiful peaks in a water spectrum actually the problem is that these peaks are so broad so that anything else you're going to need to detect has to be on different frequencies from water because water will drown all other signals so let's get back to these torsions actually I, let's see I thought there was a movie sorry the movie is going to be on the next one this is an example of a completely straight horizontal protein and there in this particular amino acid which is an alanine see it's three groups there you have the phi torsion just before the C-alpha and the psi torsion just after the C-alpha I so wish there was a good mnemonic rule for that but I don't know one this is one thing that you have to learn by heart you need to be able to place the phi and psi torsions in a protein because they are different the way you define this is that if you look at the chain here we have our central C-alpha the carbons before the C-alpha is the nitrogen of the previous amino acid and then we have the carbon of the previous amino acid and actually even the C-alpha of the previous amino acid and if you take these one, two, three, four four atoms always define a torsion here so you have one plane of these three atoms and another plane of these three atoms so you will be able to rotate around that bond and finally we can define an angle between that plane and the plane by the three atoms afterwards so the phi one here will be carbon and nitrogen of the previous amino acid and the alpha carbon and carbon of the current amino acid while the psi torsion is nitrogen of the previous one and then C-alpha, C of the present one and nitrogen of the next this is so not obvious you are not going to see why you need to look in the book go online, in worst case I can bring a molecular building no, I'm not going to bring it tomorrow because I need to fly out tomorrow I would bring a molecular building block toolkit later on this becomes much more natural if you sit and look at it in a molecular viewer and twist and turn it you will learn this by heart but it will require a little bit of work I thought that I had a movie this is going to be in a few later slides so in contrast to other angles dihedral angles are really defined by the angles of two planes so if we have four general atoms for a general definition I don't care what the atoms are whether they are alpha, carbons or oxygens and then we need some ways to identify them and usually when we identify we start in this is at i so it's ijkl so ijk defines one plane, the blue here and jkl defines another plane so the angle between these two planes define a dihedral angle and then of course if you have two planes should you measure the angle between from blue to red here or from blue to red in the other direction there are two ways of doing it and when scientists are involved you can imagine what that leads to there are two definitions so it turns out people in polymer physics do one thing and people in biochemistry do another thing don't worry about the definition for now I'm not going to ask you about it but it's also through this measure we can measure things like cis and trans isomerization cis versus trans along that peptide bond is exactly the same things it's a torsion angle around the bond in the middle that is right there roughly 0 degrees or 180 degrees so it comes back to torsions or dihedrals they're essentially equivalent you can use any word you want for it they determine everything in not just proteins but all complex molecules actually what torsions are allowed? I said that we could imagine placing each torsion in roughly a 10 degree unit I'm sorry that was a horrible lie rather than no it wasn't a horrible lie it was a very simple model but it's going to turn out that my simple model was a bit too simple so if we take a molecule like this and twist and turn it, it's going to turn out remember we also have a side chain here so if I start turning this randomly at some point those atoms might bump into other atoms and if atoms are bumping into each other the energy is going to be so high that the molecule doesn't want to be there and you can represent this in a very simple plot if you have two variables, phi and psi let's put one of them on the x-axis phi and this goes from minus 180 degrees to plus 180 degrees and let's place the other one on the y-axis minus 180 degrees to plus 180 degrees for the psi angle and if we don't know anything else you have the benefit of knowing protein structures so we can just go and look up things in the protein data bank even the first hemoglobin structure and then look at residue one for the first residue, what is the phi angle and what is the psi angle say that it's an alanine and let's assume that the phi was minus 90 and psi is 0, then I put the dot there black dot and then we can take the second residue in that protein and then we keep putting dots so why do we put dots? well, we collect data and we want to learn something and when you collect enough data it's going to turn out that there are going to be parts of these two-dimensional plots that is full of black dots and there are other parts that are completely empty and the parts that are full of dots are the parts where it's easy for these amino acids to be while the parts that are completely empty are the parts where they would collide they would clash into other things and it turns out that very large parts is disallowed so this is specifically called a ramachandran diagram fancy name for something that's actually pretty simple the ramachandran diagram just describes the distributions of phi and psi torsions in a protein or you can do this any way you want you could say what is the average ramachandran diagram for hemoglobin or you could say what is the average ramachandran diagram for say old proleins if you want to find out this is versus trans so this is apparently a well populated region and that is another well populated region and here is another region but here we don't have anything and if we group this for amino acids it's turned out that what I showed you on the previous slide that's kind of what we have for general amino acids on average lots of things there, lots of things there and then a little scattered things glycine on the other tends to have things in more places why do you think glycine has more areas where it can be it's small and likely more floppy it doesn't have this bulky side so that glycine is happier to be placed in lots of other ways proline appears to be the opposite why? so proline has a ring in it so that the phi angle that would break the ring so the proline has to be at the phi value of roughly minus 90 it can't change from that so proline can pretty much only be along that line already here can you start to think about some things what would happen with proline versus glycine when you see them in proteins which one of these do you think would be more common in say turns because it's small and floppy right if you take a proline and put it in some other large elements say like an alpha helix what do you think might happen it would break it right because it's it can't adapt and then residues before proline is similar the proline will spill over there too it can't really place in other ways this again has a load rather than worrying about 10 atoms and the XYZ coordinates of every atom we've now taken this down to two numbers per residue it's your first model and this model the point is not that the Ramesh Handen diagram is something complicated to learn it helps you because instead of worrying about 10 atoms multiplied by XYZ that's 30 degrees of freedom for amino acid now you have 2 degrees of freedom per amino acid everything else is just noise and by now you know the secondary so it actually turns out that the region down here is alpha helices and the region up there is beta sheets and that's another type of helix which is actually not an alpha helix so already at this stage we touched upon this Christian Anfinsen who was a Danish biochemist he had a very early postulate which at the time was extreme he was the first one to argue that the way proteins fold is based entirely on physics so if you take a protein and pull it out, destroy the structure somehow it will reform again I think I'm going to have a couple of slides talking about that tomorrow the book will talk about it too could you imagine any way to test this but how do you know if it has folded exactly right you need to find a protein that has some well defined function it can't be too complicated and you need to be able to measure whether the protein works or it has been destroyed and then you need to show a cell, take it out to a test tube first you show that the protein works and then you could destroy it in some neat way so that disrupt the structure of something and then you show if you now restore the conditions if this is based entirely on the laws of physics the protein should somehow be able to refold I can't determine what the structure is but I don't care because if I can show that the function is reinstated then whatever the structure because it's not the structure itself we're interested in in this case will it find its structure and that's what Christian Amfelsen showed and what do you think he got for that a Nobel Prize and he in particular predicted that the structure the confirmation that the protein is going to adopt corresponds to something called the global lowest value of the free energy and we're going to talk more about free energy tomorrow than you might have done in the rest of your life but this is another core concept in this course you have to understand what free energy is because it determines everything what happens in the lab but then van der forder Cyrus Leventhal came up with a an interesting based on things you already know and he did that ball of the envelope calculation because it was a physicist that you're saying that this can fold in one or a few seconds in a test tube we're starting from something like 10 to the power of 300 conformations and this is going to happen in one second there is a bit of a problem here because there is no way this can happen in nature and let's just assume that the numbers I had they were horribly off so let's pick something simpler instead of saying 36 squared states per amino acid let's say that each amino acid can be just alpha helix or beta sheet it's definitely not overestimating it for 100 residue we're still talking about something 2 to the power of 100 and that would take years for chemistry to test all those conformations so this is something that's famous called Leventhal's paradox so what is a paradox? I want to rescue Cyrus here too because it's not that Cyrus was stupid and argued that this was wrong and that's a paradox is a seemingly contradictory thing right? we know that it falls in physics because Christian Amphis's experiments was right and yet somehow it finds the best conformation but it's impossible in nature to test all these conformations which is fun because that is seemingly a contradiction so we're going to need to find in this course and we will solve that, we will eventually get away around Leventhal's paradox but this requires pretty advanced physics and that comes back to the central dogma that we spoke about we know that amino acid sequence leads to the 3D folded structure that leads to the folded structure and we can of course somehow say that there is something amazing going on here that finds back what Amphis and really predicted was this step given one amino acid sequence you will form into a unique structure it might be slightly more than one structure but it's unique somehow, it's not completely random and in hemoglobin partly because I showed you the structures and partly because you need to trust me for now a hemoglobin always looks roughly the same way so Amphis and really claimed that amino acids will self-organized into advanced structure and at some point and that's what we're going to come back to later on once you have this structure it's going to bind or interact with other proteins and that's what's going to cause specific function and this is also very much based on physics there are a couple of things that I don't think the book talks about this but that we so we take this 3D folded structure and destroy it like Christian Amphis and well that we denaturate the protein and then we're going to come back and talk about that in the course the opposite of that when I restore the state then we re-naturate it so this naturally has to do with the native state of the protein this is generally true but not for everything to every rule in biology there is an exception and today we know that there are proteins there are things that happen in the body that is not described by this very simple rules so some proteins are for instance modified post-translation where you add or remove special chemical groups there might be some co-factors that are required for binding or at least for the function the heme group in hemoglobin as an example of that so if you denaturate it it's not obvious that these groups will re-bind but for any small protein this is true still 70 years later which is a pretty amazing result so there is a Ben Robson this is all the cartoon that the great protein see his amazing split-second leap from fully extended to tightly coiled and then there is a quote there that I don't know how he does it so that cyrus live in thought but what you know is that this has to be based on those torches there are no other really important degrees of freedom in the protein so we're going to need repeatedly in the course we're going to drill down back to the torches and understand how does nature sample these torches and can we somehow find a way to guide nature so that we stabilize a specific state of the protein the reason why these proteins are important has to do with energies and we're going to be talking a lot about energies in this course too so why do we need to talk about energies well this is where physics enter so in the lab you're probably used to talking about reactions that happen versus reactions that don't happen unfortunately real life is not that easy particularly if you're looking at molecules we have to start talking about probabilities but there are some reactions that are likely to happen and there are some reactions that are less likely to happen a likely event is if I drop this pen it's going to fall to the floor and it works a less likely reaction is for the pen to spontaneously go up to my hand and it doesn't happen but that is not saying that it's impossible it is just a couple of hundred orders of magnitude less likely and the way we're going to talk about this in physics with these bonds the barrier it takes to rotate around this bond is a couple of kilocalories per mole and we're going to talk about units tomorrow too so if there is a barrier for something will it happen well in the naive way of saying that that means going uphill and we don't want to go uphill but on the other hand we're not stuck we are living at room temperature so there is some thermal motion from the temperature so that's some barriers you can get over so for instance if I want to go out and have lunch if I try to go straight through that door it's likely not going to work if I open the door well technically there is a one centimeter threshold there but I will of course just walk over it so to me the threshold is not a significant barrier on the other hand if you are an ant that threshold will require quite a bit of energy to get over it so that the size of these barriers related to whether it requires a lot of energy or not and whether the temperature decides that some doors we will not be able to go through because they are effectively locked and other doors we will just slide right over and it turns out that the reason why these torsions are so important is that they happen to have energy barriers that are of the ballpark we can actually get over them with thermal motion well we don't go over them all the time so if you actually change you are going to be in a specific torsion so they can't change but they don't change all the time and that in turn will we talked about when you have the sequence of amino acid on a very local scale the location of the torsions here and the relative interactions are going to determine whether you fold things into an alpha helix or beta sheets and these are going to be very strong interactions why this happens is a very good question and this is likely partly based on evolution so why do we even build life with amino acids it's a very good question that I don't have an answer to but the reason why this has proceeded through evolution is like they are good building blocks because they are small and they are stable and they achieve all the things nature needs but they are not infinitely stable which is also important but based on a handful of these called bricks and mortar or something we can then assemble these into slightly larger building blocks which you then would call tertiary structure something so that the primary structure would be the amino acid sequence the secondary alpha helix and beta sheets maybe turns tertiary structure will start to describe how have we placed different alpha helixes relative to each other and in this case there would be one subunit of hemoglobin and at some point you might have multiple subunits sticking together to a full hemoglobin molecule with four subunits and if you have a ribosome there are going to be 58 protein subunits in it again when I was a student we didn't really have a whole lot of those structures a part of hemoglobin in your world you are going to see plenty of structures in side chains because we know more now but under all this structure is not really nature this is just our model of nature because what this is this is just a collection of lots and lots and lots of atoms with XYZ coordinates some interactions between the atoms are stronger because there might be a bond between two carbons here some of them are much weaker that there is a weak van der Waals interactions between two molecules and in nature there is not really any fundamental difference but for me to try to understand that on that level is completely hopeless but thinking about this hierarchically is a model that I use to make it easier to understand proteins so let's look a little bit about these models we are going to come back to them and I realized I skipped a little bit far ahead by handing out these alpha helices were these structures one of the structures that Linus postulated and I even remember doing a lab where I was going to have you do this lab but in the late 1990s mid-1990s in London we sit down and did it for computers that were very slow at the time so I took that Ramos Chandon diagram and the reason we changed the angle by 10 degrees at a time and then you had a gigantic spreadsheet because the computers couldn't do it at the time so I had 36 by 36 squares and I set an entire afternoon and if this particular orientation of phi and psi would make atoms clash I would put a cross in a bird and in the next one they did not clash you would put a zero there and then you would effectively draw out your Ramos Chandon diagram and see what regions are allowed and what regions are not allowed and then it turns out that for alpha helices in particular there are a couple of regions that are not only allowed but in these regions they can also form hydrogen bonds so that the oxygen here will form a hydrogen bond to the hydrogen slightly further up in the helix and that's going to stabilize the structure even more and that's of course the reason why these things tend to be locked in there is a an advanced structure there are only going to be two alpha helices that I will ask you to know about but in principle depending on how hard you twist the spiral or not you can make it either slightly wider or slightly thinner like any spiral you have and there is a broad nomenclature that you have NM helices where the N here describes that you make from I to I plus M so 5 would be that you make hydrogen bond to rescue 1 to rescue 6 and then we had M atoms when you take one turn in this entire helix how many atoms do you have in the backbone and again it's just a model to try to classify things and if you do this there is going to be a bunch of different helices and the good thing for you is that they don't really exist most of them and one more the most common one is that you make a hydrogen bond to an atom to rescue 4 steps out and it turns out to be roughly 13 atoms per turn that's a 413 helix that you have never heard about but you have heard about the alpha helix and that is the alpha helix you need to know that an alpha helix forms hydrogen bonds to rescue 4 units further out in the helix because this is what locks it in because otherwise can't it be roughly 4.5 or 3.5 why is it exactly 4 the reason why it is exactly 4 is because of those hydrogen bonds if it was not exactly 4 you would start to have an offset that becomes larger and larger and larger and larger as you go in the helix so it is exactly 4 and if it was not for that hydrogen bond it would be extremely floppy so what happens if you now make this helix try to turn it tighter what's going to happen is that you just you add strain and strain and strain it's still 4 until I have so much strain that it's going to be easier for them to jump and form a hydrogen bond to something that's only 3 residues and that's what we call the 3.10 helix and that is the other helix that you will actually see in proteins so that's a helix that's one tighter it's not as good but you will see it in some ion channels and we'll actually see it in some structures later on all the other helixes are pretty much irrelevant but just for completion's sake if you take a helix, the same alpha helix and try to wind it to be less wound eventually it's going to move from 4 to 5 instead and that's actually called the pi helix I have never seen the 2.7 helix in my life so I'm not really going to ask you about that on the exam but just in case you see it in the book so alpha helix is by far the most important one and the 3.10 helix is a tighter wound version of the alpha helix oh, we have some examples of it the reason why the alpha helix is nicest is that it's not really strained the amino acids are happy it might look that there is a lot of space here on the inside, it's not really because all these atoms have a finite size radius so that is pretty much it's adequately packed and everything atoms are neither they're neither pushing into each other and there isn't a vacuum either so the 3.10 helix is relaxed the 3.10 helix you can probably start to see here that it's going to start to clash here on the inside it's a much, much more tighter wound and the one thing with an alpha helix you will gradually move around and put the cytosines all over the place with the 3.10 helix all the cytosines are going to be right on top of each other and that is the one thing reason why it's actually important just a few cases of biology and the reason why the pi helix is not this nice is that suddenly this hole on the inside starts becoming too large so you essentially have a bit of vacuum here in the middle and that is very bad nature abhors vacuum so that if you have to take a guess and a helix it's almost 9.9% of all helices and proteins are alpha helices and the last 0.1% are going to be 3.10 helices everything else pretty much does that yep we are I think I'm going to talk about molecular interactions what happens with interactions if you forget about electrostatics for a while all atoms are involved in something called van der Waals or Lenard-Jones interactions that you've probably heard about and they look something like this so if you are at very short distances you're going to have electronic repulsion that's powder repulsion and at very large distances all atoms attract each other noble gases even so you're going to get the shape that looks something like this so that there is a distance here that is the best distance to be at and if you start moving atoms too close to each other they're going to repel that's bad if you start to move them too far away from each other they would like to attract but again liking to attract means that it's not as good to be here as there so that the alpha helix would be there and the pi helix would be there so that the 3-10 helix is strained because they would like to be further apart the pi helix is strained because they would like to be closer so if you take a guess here now I only spoke about the alpha helix here because the beta sheet is to tell the truth it's pretty boring, it's just something that's straight which one of those is more stable why? because those side chains are so there is definitely some stabilization in this helix and for that reason all proteins are alpha helical do you see beta sheets in proteins? so it's not quite such an easy story because yeah well it depends on the class of protein beta sheets are certainly not rare let's have a look at them the structures so the things that are local structure just ordered for amino acids and I think this partly starts to come into Leventhal if there was too much diversity here it would be too difficult to search it seems like nature is already following the same strategy as we did organized things into building blocks don't try to move every atom but have some sort of hierarchical structure and there are pretty much just helices and sheets and maybe turns so for all of these building blocks hydrogen bonds tend to be paired there should be good local packing and the side chains shouldn't clash into each other that's certainly true for the alpha helix and the alpha helix achieves this by interacting with itself with the local structure so how does the beta sheet achieve it? exactly so this is also, it's a local or secondary structure because it's about how individual amino acids are oriented and again it's much more local than say the tertiary structure or quaternary structure in the entire ribosome but it's still not, it's more global than alpha helices because you could imagine having one sheet here and then lots of amino acids here and then you're coming back here are these individual strands they're stabilized by interacting and forming hydrogen bonds between multiple strands and here we see the same hemorrhageal things of course we can choose to model this with every single atom here and trying to track where the hydrogen bonds are but to me this is much simpler because here I see it schematically and I don't care about the individual just as I know that an alpha helix is stabilized by hydrogen bonds inside the helix the second I know that I don't have to draw them anymore because it just confuses and clutters the image and that's why we never draw those hydrogen bonds but I draw the helices as the spiral forms as that same thing with the beta sheets I know that there will be hydrogen bonds there but I don't care about exactly how they're oriented so it's easier to remove them yep sorry why that strand doesn't form alpha helices that's a really good question let's talk about it either we talk about it later today or tomorrow we will go through it in detail it's a super good question why does an amino acid form an alpha helix and when does it form a beta sheet so what determinants do you think the laws of it's physics but right now based on what I told you we don't know yet so that because based on Christian Anfinsen's results it can't be biology especially enzymes forming them but we don't really know yet why if they prefer alpha helix or beta sheets so we really need to come back to that so one thing that can happen with beta sheets already at this stage that depending on what order you put them in you can either go up down up down so that they are anti-parallel or you can go up here and then we have a large chain that goes out to the whiteboard here and then it comes back here and then we go up again out again maybe we have a helix out here and then we come into the third here so in this case all of them go from N to C terminus, N to C terminus, N to C terminus so here they are all in the same direction but here they are in opposite directions both of them will occur but there's slightly different properties or of course how different these are they depend on the scale you're interested in right now we're looking at individual amino acids and the individual sheet and then this matters if you're looking at the gigantic ribosome you might actually be there is some beta sheet there I don't really care whether it's parallel or anti-parallel so all depends on the scale which you're interested in looking at beta sheets have some very very nice properties compared to alpha helices so first they are long extended chains so it's a fairly simple structure as I mentioned the hydrogen bonds are formed between chains never in an individual chain this leads to this pleated sheets which is this it's basically down up down up down up down up that's the pleatedness and it actually turns out that it's going to be just so slightly twisted so already now I will spill the beans beta sheets have one nice properties that every second amino acid is going to point down and every single amino acid is going to point up down up down up and this means that you can create something what if you pick all the amino acids are hydrophobic and even amino acids are hydrophilic you're going to create a sheet where you have one hydrophilic side and one hydrophobic side and that is virtually impossible to do with an alpha helix so just from a functional point of view there are likely going to be things that we can build or nature can build with beta sheets that would be difficult to build with alpha helices but that will of course require us to have certain amino acids that prefer to be in beta sheets and why they prefer to be in beta sheets we're going to need to come back to so sorry this was the slide that I jumped a bit Pauling, Corey and Branson there were a series of eight beautiful papers I'm so not going to ask you to read all of them I distributed this Dave Eisenberg review these were the papers where they talked about this in theory before any of these patterns had been shown in structures and then went through all the things I talked about what are the possible helical twists you can have what are the possible, how could you place amino acids stable right next to each other they identified all the hydrogen bond all the patterns and everything in 1951 it's even before the structure of DNA and it took roughly 10 years before we then got the first actual structure of proteins that confirmed all of it I think it's a marvelous example of good theoretical research theoretical research does not have to mean theoretical physics because Pauling is very much a chemist and in theory all of you could do this with paper and pen we're not talking about quantum mechanics it's about down and using your brain and thinking and of course doing this after the structure is known after the structure known as no big deal but doing it 10 years before that's a mastermind so you can even place these different amino acids in the Rammage-Handrun diagram so this is an alpha helix 310 helices would be down here 270, you see 270 helices don't exist you could in theory imagine taking a helix and rather than having it right-handed make it left-handed then it would be out there it doesn't really happen either but friend of Ford would now start to say something so what do we know about the laws of physics are the laws of physics left-handed or right-handed if you take a mirror image of a molecule how does that change its energy the laws of physics are symmetric right so why on earth does it come that the alpha helix is common why is the alpha helix right-handed while there is no left-handed alpha helix all alpha helices are right-handed that's because your amino acids are left-handed actually left and right doesn't these definitions has well a right-handed helix has to do with the order that's you're twisting it up the left-handedness of them these are different lefts and right the right-handedness of the helix is the pattern with the helix moves the left-handedness of the amino acid has to do with how you define your groups but the point is it's because the amino acids are chiral you break the symmetry on the amino acid level and because the amino acids are chiral they will have a built-in preference for one direction again, if all amino acids were D amino acids the Ramesh-Handen diagram would be mirrored because we would instead prefer the left-handed alpha helix the reason why we only see right-handed alpha helix is because it's built from chiral amino acids that has to do with these kind of the rule book of nature you can't as you mentioned then, this means that the enzymes because these enzyme proteins will be built from these alpha helixes so that every enzyme in nature will also have a handedness built in and that enzyme will then only recognize L amino acids so that where the symmetry break originally came from, I have no idea but that is why everything in nature has a handedness you could imagine a parallel word that was the mirror of ours and as long as everything was a mirror, it would work alpha helixes have 3.6 residues per turn too could be worth knowing there is another reason why alpha helixes are very stable and that has to do with electrostatics and now we're becoming a bit physical here remember that I said about alpha helixes being charges things are a bit more complicated if you go into real molecular detail because you have atoms, for individual atoms we use the thinking of them as either they are an atom or they are an ion, ions have charges plus or minus unit charges or two unit charges and normal atoms are neutral but when you move things into molecules say water, even a simple water molecule with oxygen there and two hydrogens you will actually make the hydrogen slightly positive while the oxygen is slightly negative and you can see this this is why I can take for instance a comb and charge it with that electrically and then you can defect a beam of water because that you effectively have a dipole in the water so that's pointing from minus to plus the water molecule itself is neutral but it has more plus here and more minus here and we tend to talk about things as partial charges so water would have roughly minus 8 there and roughly plus 0.4 and this will vary but very rough so on the molecular level charges are not just unit charges for an entire molecule it has to be a unit charge but the individual atoms will have more the reason why you get this is that the electrons the oxygen pull the electrons from the hydrogen so that there are more electrons here and this will be the case in amino acids too so some of these love electrons the nitrogen loves electrons the oxygen loves electrons so the oxygen will steal electrons from this carbon the nitrogen will steal electrons from that hydrogen so if you start looking at the details here we can have some atoms here that are slightly or even fairly positively charged and other atoms that are quite negatively charged and this is particularly true over this peptide bond so if you go from oxygen to carbon so we have one arrow pointing that way minus to plus and then we have a second arrow here and also pointing that way so that actually turns out there is quite a strong very strong dipole pointing along each peptide bond and what happens then if you line all these up in the large helix it turns out that all these peptide bonds are going to point in the same direction so along the entire helix we're going to have dozens or hundreds of these peptide bonds all pointing in the same direction that will effectively mean it would be the same as if you had a partial negative charge in one end of the CO terminus and a positive charge in the NH terminal so the helix itself would be neutral but it would be spread so you have more positive charges here and more negative charges in that end of the helix and at this point I can understand if you're getting tired and he's deviating into physics you take this course to understand biochemistry you didn't want to study charges in physics the reason why this is important is if you want the Nobel Prize these simple ion channels are guided by physics it's an ion channel so what's the big deal it turns out that this is a potassium ion channel and we're going to come back to the ions there are two important ions in your membrane potassium and sodium K and Na and both of them have one unit charge so how difficult can it be to make this particular channel should only conduct potassium and it should not connect any sodium but that can be very different just create a hole that lets through potassium but not sodium how much do you know about ions this is chemistry from your upper secondary studies let me draw a potassium ion maybe something like that and then let me draw a sodium ion please create a hole that lets through this one but not that one you're allowed to make mistakes roughly one times out of 10 billion and this all your signaling in your nervous system depends on this and you're only allowed to use building blocks you're allowed to use four or eight alpha helices that's it this starts to be pretty difficult right so the way again you don't get a Nobel Prize just for determining a structure you get a Nobel Prize for explaining what this means so in this particular structure there are actually four helices so these are all alpha helices but occasionally we draw things in different ways to emphasize different parts and in this particular case there are one, two, three, four helices pointing in almost like guns to a binding site in here and you can even see in the structure a useful bit of ion density here and it turns out that all these helices are constructed so that they are in the way that it would have a slightly negative charge here and a slightly positive charge here so there's a binding site here where there's negative, negative, negative, navigative so what would you like to bind there? something positive right so it's a water cavity sorry so what happens here and this is beautiful what happens is that an ion is not doesn't just exist in isolation so what happens when you put an ion in water so if you take this ion and put it in water what parts of the water of the ion, oxidant and they will not only be turned towards it it's going to be so hard that the ion effectively binds the water so it's going to create what you call a hydration shell of water around the ion that it carries with itself stripping the hydration shell from the water requires a bit of energy because you would need to replace it with something else that's negative so which one of these, this is a bit of physics but which one of these do you think binds its water's hardest so if electrostatics gets closer it becomes stronger so while this is a smaller ion it's going to bind its hydration water harder so it's going to be easier to strip the hydration water from that one than from that one so the way these channels work and again, my point is just illustrating why electrostatics and physics is important we're going to come back and talk about channels later this would be a large blob here and you would need to go through the filter here you can't do that with your water so you have two choices one, ion coming along here but you can't strip the water so if the water was stuck so that that sodium atom is going to be stuck there, it can't get further when you have a potassium ion on the other hand, the potassium ion comes in here and it can actually strip its water and replace the water by being coordinated by these four helices and whilst we've stripped the water off the larger potassium ion, the ion comes through here and this is of course based on nature has an advantage of 1.3 billion years of trial and error but this is the reason why these, again even physical interactions such as lictostatics and dipoles, the one thing the way they found this out in the structure is that they noticed the strong helix dipoles here so by understanding the physical properties of that, they could decipher the mechanism of the channel can you repeat by the potassium pudding it wasn't too big no, so the potassium is the one that does go through yes, but the point, because it's larger you can imagine having all the charge in the center here and that means that the hydration water is going to be slightly further away from the charge so that means this is a larger ion but it's going to be easier to strip its hydration water and that is the key because that once you add all the hydration water around this that one is larger and we can't strip this water around because it binds it harder so if you just look at the ion it appears to be a paradox that how on earth could you get that ion through by the point that we can strip the hydration water and potassium without the hydration water is smaller than sodium with hydration water and that is why we're going to start talking more about physics, that's why we need to understand interactions, we're going to need to understand electrostatics, we're going to need to understand Father-Wall's interactions because that this is not just fancy physics or understanding why proteins, well this is going to determine why proteins work in the first place so what do you think might happen if you have mutations in these helices it will likely start to influence how good these channels are at conducting things and it turns out that there are a ton of diseases related to the mutations and helices in exactly an ion channels and then you're not going to conduct ions in the same way and that sounds like a minor problem and the only problem is that your body depends on conducting trillions of ions per day so suddenly your nerve system is no longer going to work or your heartbeat is not going to work which is kind of bad so the point is that they stabilize it by helices sorry, by electrostatics and I think I already mentioned the last part here about the special properties of water and that is something that we're going to come back to sorry, can I take three more minutes, I have a few more slides I mentioned that electrostatics was important the reason why electrostatics was important and why we're going to come back to it is that it's a strong it's a super strong interaction if I take two charges here and separate them by one angstrom that is a bit extreme but you're talking about 300 kilocalories per mole it's an insanely strong interaction so that electrostatics tends to determine lots of things if you have large charges in proteins they're likely not random and if you can compare that to a torsion which is like 100 times smaller you need to have a rough 150 or 400 I don't care about but you need to have a gut feeling for these energies because they're later on going to determine things that happen versus things that don't happen and they also, they are very long range because charges decay as one over our charge interactions and because of the electrostatics is so strong that's also why we get all these hydrogen bonds when an oxygen in one water interacts with the hydrogen in another water you can break this bond whereas this is almost a covalent bond it's super strong and in ice all of them are formed perfectly we're going to have these hydrogen bonds in DNA too that's what keeps all these strands together and again that's why the reason why DNA is always a sub-helix it's not just a minor electrostatic interactions they are so strong that it's astronomically bad to pull them apart and that is what keeps DNA folded up we're going to talk more about the hydrogen bonds more but I'm going to show you one example here we'll ask you a question so this is a small computer simulation of liquid water so what is the difference between liquid water and ice well it turns out that you might imagine that you might have ice being perfectly solid and all the hydrogen bonds formed that's actually quite true in ice at zero degrees Kelvin you had an average of two hydrogen bonds per water but it's not at all true that when we melt ice that we break these hydrogen bonds we go from say 2 to 1.7 hydrogen bonds per water so the hydrogen bonds are still there even in liquid water but there is something that happens when ice melts into water and that's going to be ultimately very related to things like protein stabilization and folding and we're going to talk a lot about that tomorrow but that's when we're going to need to start talking about these free energy concepts so that on average the hydrogen bonds are formed and I already mentioned that they exist everywhere even in the individual base pairs that's the reason why adenine pairs with titanium two hydrogen bonds while guanidine pairs with cytosine one, two, three hydrogen bonds and what I'm so not going to go through now and now but I will at least tempt your appetite a bit protein folding is very much going to be about hydrogen bonds are things paired with water or when do hydrogen bonds form or don't form internally and everything so hydrogen bonds is the first part of solving these riddles when things form or not and this is where we're going to approach this later so now we started talking about biological point of view tomorrow I'm going to go into the physics because we need to understand the things that I mentioned about probabilities what things happen and what things don't happen and once we've a little bit more physics we can move back to the biology and revisit these things but rather than today I was hand waving but when you see this on either Friday or next week then we will be able to approach the same things again but with physical rigor and then we're going to use equations to actually predict exactly why things happen and why they don't happen and that brings me to the summary slide there are what I spoke about today corresponds to chapter one and two of protein physics for your sake read them even if you think it's repetition and I'm going to spend the first 40 minutes or so it depends on how talkative you are and talking about this the worst thing that can happen here is that I don't finish all the slides in a lecture I will still try to finish exactly run noon and then I'll just push those slides to the next lecture never ever feel that you're interrupting me we're doing this for your sake not mine and if you want to go you're more than welcome to read ahead and what I'm going to do tomorrow at the end of each lecture I have a slides of study questions you can ask me absolutely anything but these are the questions I'm not going to answer them tomorrow you're going to answer them tomorrow so I will start throwing these questions out if all of you participate that's awesome if there are only a handful of you participating I will deliberately start going through the classroom because the more nervous you are at answering these questions the more important it is that you do it because this is going to be about learning and I don't expect that you will know all these by heart tomorrow some of them are easier than others I don't think we spoke about this this might be something that you need to look up how much ATP does your body use in a day look it up as an exercise you want it's more than you think good tomorrow we're going to spend talking about physics but we'll leave that for tomorrow and then I have more handouts for you and everything so you see you in the same place here tomorrow morning how many of you don't have computer accounts? okay those of you who just spent the last four months dwelling around in the basement and hacking in those computers you probably don't need to go there again but if any of you are unsure whether your computer accounts work or if you don't have a computer account set up Bjorn Dari and Eric will meet us in the computer room and then we can head over there and both make sure you get in if you need an entry card and also make sure that you have an account set up that works for you but I would suggest we do that right away because it's only going to take ten minutes