Video 1-1: Protein sequencing and mass spectrometry

0:03 This first unit is really just an exemplary example, a tiny uh puzzle almost taken out of context um for showing you what the module will be like. And uh we'll uh take a step back next week and do some of the the background in a bit slower pace. Uh but for today, I want to go uh straight into this. Um, the puzzle we'll look at is essentially sequencing proteins where here this means reading what the protein is made of. And I'll tell you a little bit of what proteins are to start with. Uh, and then we'll look at the the abstract problem that's behind this. That's the turnpike problem.

0:44 Okay. So, what is protein sequencing? Well, for that we should uh know what proteins are, I guess. Um proteins is a whole class of biological molecules and uh in a way they're among the most interesting ones that we have. Uh you may know protein as uh a nutrient. So meat is protein and many other uh uh things we eat consist of protein and our muscles are made of protein and much of our cells are made of protein. Uh but they can actually do a lot more. Um, first of all, what they are are long chains of little beads, and the beads are amino acids. There's more chemistry behind how the chain is formed, but you could really think of it as kind of a a chain of pearls, which are connected with a very strong piece of string. So, they're really hard to break in in that uh chain uh connection. They're very stable. Um, but also these little uh circles that are drawn here, they're not all the same. These amino acids uh come in very many different colors or flavors if you want. Uh at least 20 are the standard standard kind of variations. Um and they have different properties. For example, some of them are uh electrically charged or magnetic and so they uh attract each other which means uh in um in water as these mostly exist in our body they fold into three-dimensional lumps or knots or shapes.

2:24 And here you see this is uh the red and blue part. These are two different proteins but they together uh form this almost knotted structure and that's hemoglobin um which somehow uh manages to to uh capture oxygen. I don't know if the green things that's not oxygen that's way too big. Okay, don't don't ask me about the chemical details of how hemoglobin does this treat of capturing oxygen and then uh moving it via your blood cells, red blood cells through your body and then releasing it at the right time. Um I'm a bit out of my depth here as well. The point is that proteins are essentially a linear shape. And if you know these uh the colors of these beads, if you know the sequence of amino acids, then many of the other uh functions that the that the protein has can be worked out from there or at least you can try to make these proteins synthetically if you know the sequence of amino acids.

3:27 Um the structure is usually uh that the function is determined by this three-dimensional structure which is yet another question how to get there. Um there's a few things listed of what proteins do in our bodies. Uh is a wide range. Um our hair are proteins. So they can be fairly stiff and uh uh really stable structures but they're also working as enzymes. the the chemicals that break down food even in your uh saliva and uh there's hemoglobin as I mentioned some hormones like insulin are um are proteins uh the sequencing of insulin was one of uh the Nobel prizes won for proteins one of the few people that be uh were twice awarded the Nobel Prize and the first he got I think this was okay I'm not I'm not going to say wrong things. Um there was a Nobel Prize for sequencing the amino acid sequence of insulin uh with obvious applications for all the diabetes patients.

4:35 So proteins do a lot of things and that means in bioinformatics uh there's many uh many things we want to know about them. We want to know the amino acid sequence. There's a long puzzle that's kind of solved of how to get from the sequence to the three-dimensional structure. where um uh alpha fold seems to work well enough that people are using this to at least guess what the structure will be and what the function will be. Uh they interact as you can see here. So there's there's lots to study about them too much um for us. We'll just focus on this uh on this problem of determining these amino acids in the sequence. To make this a bit more precise, there's al these uh unpronouncable names. These are the standard 20 amino acids that you usually find. Um they have a name. They have a a short code. For us, they're just part of an alphabet. So, a protein, its sequence is a string of of these letters. So, every letter is is one amino acid. And because there's too many, usually you need um a slightly longer name than a single letter in in our alphabet.

5:54 Uh and they have different chemical properties. There's the sum formula of what they consist of, but of course the structure is uh also important. And they have different molecular masses that we'll come back to soon. Okay. So much for proteins.

6:14 The key question is um for for this for this unit if you're given such a molecule you know somehow it exists in some cell uh how do you get at the the sequence of amino acids and now let's completely forget the problem of how you isolate this specific protein that you want to look at out of this messy uh environment with all the million other substances. uh suppose you already have a way of isolating that um how would you get to the sequence? The in practice most successful option is actually a a completely crazy indirect one um which comes from from the fact that most proteins are made by your cells according to a blueprint which is written in your DNA. And so if you read the DNA, it's actually easier to work out simulating that process what the protein will look like than actually measuring the protein itself. This is it sounds nuts to me. And the reason this works is just that our technology for doing the letter is so much better. Uh inherently it seems much more complicated.

7:28 But there's some uh proteins where this also doesn't work. If you know, if you know learn anything about biology, then it is that every rule is broken in some example. But whatever weird thing you can come up with, you'll find some some example of some weird creature that evolved to do with something different. And for proteins that also exists, um there's some that are not made from genes. uh and maybe at least for these or also because you want to do it more directly. There's also lab techniques for sequencing proteins if you just have the final product.

8:05 And there a very versatile one is is mass spectrometry where essentially you take the molecule, you break it into lot of lots of pieces and you try to measure how much they weigh the different pieces. Um, let me try to play a little bit of this video. Now, you're not going to hear much from it. That's the problem. Uh, because I didn't couldn't get the sound really to work.

8:47 not too bad, but okay. So you can watch the full version at some later point. So that's what what these things look like. Um we don't need the history for this one. But let's look at the explanation of what this machine does because I think it's useful to have a little bit of a at least a sketched idea of what happens to appreciate what we need to do afterwards. >> Move electrons through the sample. So there's a little cathode dra tube. it produces all these electrons. What it's going to do is it's going to pull electrons away from the sample. And then as it does does that, it's going to create a number of positive ions. And so we're going to ionize that sample inside here. Remember, it's still the sample. It's just ionized. It's lost its electrons. And so now we move to the mass analyzer. That's really only going to have two parts in it. It's going to have an electrical field. You can see that's negative because we want to.

9:45 >> So already before that step, what what he's showing down here, um it's a little cut off. Let me move it up a bit. Um what you see down there in in this in this box is already your protein, many copies of it essentially are broken into pieces in all possible ways. you have so many copies and you break it vigorously in into many pieces that it essentially breaks at every possible position. That's going to be at least our abstraction. Uh and then the uh chemical parts that are left over, they'll have protons in the nuclei of the atoms and they'll have electrons and you try to don't ask me how to do that in detail. You somehow get the electrons away with an an electric field. So what's left are positively charged particles because they lost the electrons. And so that you use and what comes what comes now? >> Move the ions into the mass analyzer.

10:51 And now we're going to have a magnet. And what that magnet is going to do is it's going to bend the path of the ions. And so as we're bending the path of the ions, it's just like driving around a corner. If you're really heavy, it's harder for you to make a corner. Let's say if you're in a big semitr. But if you're in a little motorcycle, it's easier to make it. And so this is where we're going to figure out the difference between the mass of those ions. And then finally, we have a detector. That detector is going to be made up of two things. We're going to have an electron multiplier, which is essentially a plate. As electron hits it, it spawns more electrons, which hit the next plate, which spawns more electrons. And so we can really have a small amount of ions or anything hitting that plate. And we're going to get a signal. Now that signal has to be amplified, but eventually we can send that into a computer and we can look at the spectrum coming from those different masses. And so the first thing you have to do is you have to calibrate the machine. What does that mean? You're going to start sending ions through. Okay, so that ion didn't hit the detector. Why is that? It's because the magnet is turned up too high. And so we're going to have to lower the strength of that electromagnet. We run another ion. Okay, now the magnet's not quite uh strong enough. And so now we run another ion.

11:58 Another ion. Another ion. Okay, so we've calibrated it and so it seems like it's working well. Which of these would be heavier? Well, the ones that are heavier are going to be the ones that can't quite make the corner. And so they're going to end up out here and then the lighter ions are going to end up right here. Now, let's actually get to some sampling. And so this is what it's going to look like when we create a spectrum. We're going to have the different weights across the bottom and then we're going to have the intensity. And so wherever the intensity is high, we're going to have peaks. That means we have a lot of ions that are with that specific uh atomic weight. And let's try chlorine. So, we're going to put chlorine through here. Chlorine really only has two stable isotopes. It's going to have uh chlorine 35 and 37. So, he's doing a much simpler example.

12:42 It's a it's a single element, not a whole protein. Uh but it turns out even a single element is not just a single variety. They come in different isotopes. Um and so if you uh if you send this through there'll be different peaks for the different uh different amount of of neutrons in in the core really there. So different weights. Uh maybe that's that's enough of of that. So you uh you get this even for very simple chemical compounds. Um, one thing to keep in mind is what you measure here is not actually the atomic weight of the particles. It's a a ratio of of charge and weight. That's one problem because you can have ions that lose several electrons and they have a larger charge and so they're more influenced by the magnet. That's one of the things we're going to ignore entirely. Um, and amazingly, you can do the same process with much larger molecules. You can still reliably break them into many, many pieces, and you'll get these huge number of peaks for all possible little parts in that molecule.

14:01 That's basically what the lab technique can do for you. Take many copies, break them in all possible ways, and measure how much of these different parts you've got.

14:16 Okay. Um, so so much for that. uh we'll abstract from that a bit. But even at this point, it might sound totally crazy to reconstruct this linear sequence of the molecule from just these peaks of weights.

14:37 And the uh the way to uh to rescue us here is that many uh you will get many peaks also for for segments of the whole protein that are still intact. So you get substrings and you get the sum of the weights in that substring as another peak and I have to somehow work out from the collection of all peaks what the sequence was.