 Imagine that we were to take those four helical bundles and carefully pick something on the inside. The easiest thing way to do that I think is to show this with those helical wheels I showed you. So here we've taken amino acids along the helix and you're carefully making every third and fourth rest you roughly hydrophobic. We're putting them on the inside or the inside according to our model and then you just make the rest of them hydrophilic. The idea is that those should go on the outside and if I now synthesize these helices have a small four helical bundle like that express them as a gene. If I'm really lucky and when we throw them in water they would end up in a form roughly like this and they would spontaneously have a hydrophobic effect that would drive those four hydrophobic surfaces to face each other. That's not a guarantee right? Remember that's our model that's how we hope the protein will behave when we kind of instruct it to behave that way but you would certainly need to check that with the structure. But if we do that maybe you could somehow artificially create the binding site for something particular. And in this case what you have here on the right is a protein where researchers I think it was Pittsburgh a few years ago tried to no U-PEN sorry my bad Philadelphia U-PEN a few years ago tried to create a synthetic blood protein. This is an extremely important area for a number of reasons that blood today is a fresh product you can't really freeze it for a long time or anything and that means that we're constantly in need of blood donors for all the different blood groups and if there is ever an emergency they might urgently need to go out to radio and call for more blood donors. There are also cases where there might be religious reasons why people don't want to receive blood and perhaps the most important one it's simply complicated to store all this blood have a procedure for getting blood donors to come in and having staff to take care of them and draw blood from them. So the idea is that maybe if you could synthetically make a protein that has the same capability roughly as hemoglobin that it would bind oxygen in the lungs it would release oxygen in the muscle tissue and then be compatible with your blood groups and avoid setting up your immune system. A protein is in theory a great way to do that so we just need to create a protein that would bind say a heme group or something with the iron and then we're all set. Historically what people have tried to do is that they've tried to engineer hemoglobin and myoglobin to do that unfortunately it this far it hasn't really worked well it has occasionally worked in the lab but in practice either it has been binding oxygen too weakly and then it's not going to be efficient or it's been binding oxygen too hard and then it doesn't release the oxygen in the muscle tissue and unfortunately in many cases they appear to bind nitrogen too and when they're binding nitrogen that frequently leads to tissue damage in the vessels and everything. So there have been a whole string of these companies that have set out a new product they've had a patent they've gone to clinical trials and then a year or two later the clinical trials have been a disaster and they filed for bankruptcy. So the last decade or so there's rather been renewed interest in trying to engineer proteins that so instead of copying myoglobin or hemoglobin the idea is that we could try to make a protein that's much smaller maybe just a four helix bundle and get this one to bind a molecule that would bind the iron that would bind the oxygen. Unfortunately I'm not sure how well this study went either I haven't seen any follow-ups since the Nature paper roughly a decade ago but I posted that paper for you to give you an idea of the early stages of protein design that we're doing today. Does that mean that protein design doesn't work? No it just means that it's hard. There are several other cases where protein design has been remarkably efficient in particular this when trying to match hydrophobic and hydrophilic residues. Have you ever used any low-fat spreads or for instance trying to use creme fraiche that you're boiling? The reason you can do that is that we're using emulsifiers the same type of emulsifiers that you have in milk that means that in milk you have a few percent of milk fat although it's a water solution. What the dairy industry has done very systematically is using in particular milk emulsifiers and try to use them to dissolve larger amounts of fat for instance if you have creme fraiche or something that you would like to fry in. Now ideally we would like less than 40% fat but it's also the fat that makes it possible to fry in various products such as low-fat spreads. Historically that hasn't worked if you ever when I was a student and we were poor eventually you would try to fry in low-fat spread and if you've ever done that that's a great exercise. What would happen when you heat it something just turns into pieces in the frying pan and you eventually end up destroying the pan if you try to fry it. What's happening there is that these emulsifiers are proteins and when you're heating a protein eventually the protein will de-ratuate. So what you're doing is that you're destroying the protein that is the emulsifier that dissolves the fat in the water and when that has happened you're going to have separate phase separation. You will have one phase of fat and one phase of water and it's all that water that makes it look like some sort of strange solution with pieces in it or so. What the dairy industry has then tried to do is try to derive more and more stable emulsifiers ideally proteins then that would be stable at 100 degrees centigrade so that you could boil or fry in these low-fat products and they've actually been remarkably successful at this partly by just finding better emulsifiers and partly by systematically changing amino acids and making sure that you have a protein that is A still an emulsifier but B significantly more stable than the traditional ones. So usually the best recipe for successful we're designing protein is to stick as close as possible to nature do some minor adjustments but if you can try to avoid starting from scratch and designing the protein entirely yourself. We'll get back to that towards the end of the class when I talk about protein design.