 So already in the first part of the 20th century we knew that there was protein in biological membranes in addition to lipids. The only problem is that it was very hard to get it out. The first protein structures, do you remember when we determined those? In the 1960s, hemoglobin and myoglobin. The problem is that we can't really determine membrane protein structures that way because the membrane protein, if we draw that red, the membrane protein doesn't exist in solvent, but it's going to exist in this lipid bilayer environment, right? And now I feel a little bit bad because I'm drawing these ideal parallel lipid tails, but it would take too long to draw them properly. I can't crystallize that protein because this protein has to exist in an infinite bilayer. And second, these are lipids. That's basically oil. Have you ever seen oil crystals? Technically, some lipid membranes will actually crystallize, but it's exceptionally hard and was impossible for anybody to crystallize membrane proteins for a long time. In fact, it was not until 1980 that Hartmut Michel published an idea that he took a normal membrane protein, one called bacterial rhodopsin, but then he took this, he used some chemical reactions to first get this out of the bilayer, salivating it in oil, essentially, but then he was adding something else. Instead of those normal lipids, he added detergents, simple small molecules like the one you might have in washing powder or so. And they were also, they were anionic. What then happens is that these, if you put many of these in solvent, they're going to form a so-called micelle. So they will put their heads outwards and then the tails inwards. Sorry about that. It's quite disordered in here, but the key difference here, do you see it's not a bilayer, it's a small spherical blob or something, roughly spherical. What then happens to the entire protein is that that protein that we previously had in a bilayer, we will now have that in a blob of detergent, not all the way around it, roughly like that. The upper and lower part here is fine because there we expect to have water, right? Here I would have water and here I will have water. So this detergent creates a way to solubilize the membrane protein, technically it's still hydrophobic on the surface. If we're lucky, we haven't destroyed the structure too much. What Michel, what Hartmut Michel then did is that he could get these to crystallize. And what basically happens then is that you get a large periodic copy of the protein, a crystal. And whenever needed between these, you would have the micelles, the detergents. You probably get the idea, right? And in particular the water soluble parts here, they're again, they're water soluble, so they're going to be quite happy to pack against each other. With a crystal like that, he was able to determine, to crystallize a membrane protein and then they determined the first structure of a protein called bacterial rhodopsin. It's a very special protein.