 Small drugs are amazing in a ton of ways, but they're also relatively limited. You only have a few hydrophobic, large aromatic rings or so, and then we have to build diversity with that. Compare that to proteins. Remember the diversity and the stability, the details we had in protein structure. It's completely amazing in comparison. So, couldn't we try to design protein drugs? That would likely give us much more freedom, and if we have more freedom, I might be able to design something that is significantly more specific. Well, in a few cases we almost already did that, right? Remember those glycophorein helices in a membrane? When I told you that normally you had a glycine and then three other amino acids in the glycine, and if we carefully put the right polar site chains on those two helices, we could get them to stick together. Basically what I am doing here is that I'm designing a helix to stick together with something else that just happens to be a copy of the same helix, but it's the same principle. It's not a coincidence that I use that as an example, because that particular example is actually very important. Not the glycophorein system, but another system that you've heard of. Remember the tyrosine kinase receptors? Not the receptor on the inside, not the tyrosine kinase on the inside now, but that transmembrane helix. And do you remember that I told you that under some conditions they have to dimerize for the effect to happen, and either they dimerize too much or too little if there are mutants in these helices, and that can lead to severe disease, in particular tumor growth. Wouldn't it be amazing if I could somehow add another helix and start to influence this process? That's an old idea, and many researchers have tried it, most of them have failed historically. So the idea is that if our healthy helices here would be green, the transmembrane part. Normally you would have some sort of ligand appearing extracellularly that will drive the two extracellular domain here to dimerize. That causes the transmembrane region to form pairs, dimerize, and that leads to a signal on the inside. Then what happens is that the extracellular part here should let go or at least dissociate, leave each other again, and then the green helices should leave each other. But for some reason if there is now a mutant in the blue helices here, the blue helices will keep sticking together, or maybe not stick together at all from the start, and that is going to lead to errors on the signaling on the inside. The idea here is that if we could define a red intercepting peptide, that red helix could have higher affinity to the blue helix than other blue helices. And in that case I could stop that bad connection between two blue helices happening and stop the signaling and thereby stop the tumor growth. This is not science fiction. Bill de Grado actually did this some 15 years ago together with Joanna Slaski. And their idea was surprisingly simple. When it comes to helices, you've seen this. There are pretty much only two ways helices cross each other, right? We know that from the Protein Data Bank. We also have the GX3G motif from the like foreign helices I just showed you. So what they did is that they collected statistics from the PDB and grouped all pairs of helices in clusters based on how they pack against each other. Then they took those as templates, and for each of these templates like we have here, one of the two sequences will have to be my blue one. So I start by filling in the blue one. The other helix is free. Well, it's not completely free. I have to have a hydrophobic sequence so that it will insert. But then I try to optimize a hydrophobic sequence to get that to bind to the blue sequence. And then they kept iterating this a bunch of times and basically let the computer do this for them. The end result here is that they actually did find a bunch of helices that had dimerization effect on the blue ones. And what that means is that at least in theory this worked in their experiments and cell lines. You now have a helix where we can start to specifically interfere with the binding process and either prevent things from sticking together in the membrane or possibly even aid them to stick together more if we have to. This is a bit of science fiction for now, but it's not more science fiction that people are able to do this at least on a research basis. It's probably going to take a while before this is regularly used in industry. This discovery, if I recall correctly, was even sold off to a company, but I haven't seen them use it since. So who knows, it might have failed later on in clinical tests or so. But the concept of tailor-making designed proteins is something that's only starting to happen. I bet it's going to be super hot in the future.