 At this point, you too might be feeling really excited and figuring we should just throw all the old-style drug design overboard and focus on new advanced design. There are lots of companies doing that, but there are some caveats. This is an example of such a modern design. This is actually two proteins. This particular part is a CD28 receptor on our T-cells in the human body. This particular drug was developed by a company called Teginero. It had a development name called TGN1412, very common. They're not marketing it yet, so it doesn't need a fancy name. The idea was to create a drug that would kickstart the immune system to help fight tumor cells, for instance. If I could target certain receptors on the T-cells, maybe I can get those T-cells to work more. Very smart idea. What they did is that they took this antibody and literally replaced residues here to get this to match perfectly with the CD28 receptor. If this works, we should then be able to have the entire immune system kickstart it if we just adapt the dose correctly and get at least a five or tenfold more killer effect than the T-cells. The T-cells will then hopefully primarily kill the tumors. Really smart idea. When they developed this drug, they took it through the normal preclinical test, animal tests, and everything, and then they were ready for phase one trials. In phase one trials, you usually have students or other volunteers come in. They get 50 pounds, and you normally start with a very low dose, say a factor 100 or even 500 in this case, lower than the dose you expect will have an effect. So that if you see any side conditions, we will start to see those before we crank up the dose. But in this case, the company Paracel did this test in March 2006, I think it was. When they started to inject the first dose, what happened after just a few minutes is that the test subjects pretty much fell over. They had very severe reactions in their immune system. Their own T-cells went berserk and started killing all the leukocytes, and they had multiple organ failures. Their hands swelled up. Their heads swelled up to almost twice the size for a few of them. For out of a miracle, none of them died, but several of them lost limbs, such as their fingers or so. Obviously, the phase one studies was canceled immediately. I think the entire project was canceled. As far as I know, the generic went bankrupt. But what happened? Well, it's not entirely easy to perform a postmortem on the product, because these things were not shared entirely publicly. But having looked a little bit at that, this might have been related to this design. Because remember what I told you. They designed a project, a protein here, to perfectly fit the human CD28 receptor. But then all the actual tests here, they were performed in mice. And the mouse CD28 receptor only has something like 60% residue identities of the human one. So that's not going to be a key that fits the lot perfectly. In fact, you probably have to use quite a lot of force to activate the mouse immune system. And by force, I mean a relatively high dose. And now we adjust the dose until we get the expected response from the mouse immune system. And then we believe that that is a safe dose, but that is an excessively high dose because we didn't have a good fit. Now we reduce that by a factor of 500, and we think it's a super safe dose and inject that in humans. But what has rather happened now that in humans, every single receptor is going to fit perfectly. We will activate this on all cylinders to 100% and the immune system will go crazy. So instead of a factor of 502 low dose, we probably rather had a factor of 10,002 high dose. And that's likely what led to the leukocytes dying and everything. That we can't continue such a project. But that particular feature might very well be possible to rescue in a future drug. The important take home message here is not what went wrong, but that as we're doing new types of drug design, we're going to start hitting new types of problems, errors, challenges, side effects that we've never seen before. And this was, of course, one of them. As always, the danger is not the side effects we are aware of. But for each such effect like this one, there are probably in the other 100 new effects that we're still not aware of. So beware of this. If you start working on protein design, there will be completely new pitfalls that I have never thought of and neither you likely.