FX125L is a first-in-class anti-inflammatory small drug molecule that has prospects for the treatment of asthma, COPD and related conditions. Along with back-ups and analogues, it was discovered by medicinal chemists in the Fox group and biologists in the Grainger group in Cambridge. In 2007 FX125L was taken into clinical trials by Funxional Therapeutics Ltd (FXT) and in 2012 was bought by large pharmaceutical company Boehringer Ingleheim (BI) for an undisclosed sum.
The success of the project has been based on not only good fortune but also using the right approach to take advantage of our good fortune.
Simple and scalable synthesis. The synthesis of drug molecules is not a single problem. When compounds are first designed and synthesised they are often made using a route designed to be flexible and allow many variants to be made. High through-put synthesis allows large collections of related compounds to be made using near identical reaction conditions. If you find the ideal "hit" molecule you then have to make more of that one compound. The route you use to make a large amount of one compound is often different from that one that you would use to make a small amount of many compounds. So you change the route to make it cheaper, safer, more efficient, more reproducable, and more environmentally firendly. If you end up with a successful best selling drug molecule it makes sense to try to devise a third even better and cheaper synthesis. All of this takes a lot of investment in time and money. Anything that can be done to use a more scalable first "discovery" synthesis would be a great advantage when selling on the project. It must be more difficult to sell a clinical trials candidate if non-one knows how to make enough of it. If you have a scalable synthesis to begin with it all seems more attractive. "But.....!" I hear you read. Yes, I know, the requirement for in-built scalablity will reduce the number and speed of synthesis of compounds that can be tested. Don't worry, there is a solution......
Complicated and less-scalable biology(!) This doesn't sound good, does it? It's not what it sounds like and it has massive benefits. Step back a minute from the complexities of molecular/chemical biology and consider what we are trying to do. We are trying to make molecules that have an effect on a biological system. So what should we use as a measure of success when rating molecules? How about using that effect itself. Not a simplified assay that we assume will have the result we want but the effect itself. The assumption that "binding here" or "inhibiting there" is the key to curing a disease is a big step. While the journey down the reductionist pathway of a biological drug mechanism from a phenotype to an interaction with a single enzyme or receptor is no doubt challenging, interesting and a method of gathering more data about your molecules, making that journey "in reverse" can lead you up blind alleys. Being able to turn a steering-wheel doesn't make you able to drive a Formula 1 car - it's much more complicated. If you want to find the best F-1 driver you had better test them in an F-1 car. Understanding how they do it is a separate issue. It's the same with molecules. Make your testing regime as similar to the outcome you actually want as possible. It cuts out false positives by integrating and effectively balancing all of the inter-related biological effects of the test molecule on the system. It helps cut out molecules earlier which, while showing activity against the drug's molecular target (enzyme or receptor), would fail in the clinic because of unwanted activities and interactions. This type of assay is, at a scientific level, more complicated and, because it can be more exacting to perform, less scalable. However the information that you gather is much more valuable because it is a better indicator of the effect that you really want. It may seem slower but you win in the end. Remember, you haven't had to spend years finding and validating the molecular target, and then cloning the protein and designing the assay - you can, for instance, use self-replicating cells instead. The fact that the biology may be a bit slower helps the chemists as well - they can spend more time designing better routes to better molecules rather than just making as many easy-to-synthesise molecules as they can, as fast as they can, in order to feed a hungry high-throughput screening assay.
So it comes down to this: if you have to design and make a molecule yourself, do something simple. If however you can use a pre-existing system (e.g. a cell) to test your molecule, go for something that, although seemingly complicated, will tell you far more information for little extra effort.
And if that didn't convince you....... the FX125L story has a big "bonus extra feature" at the end. Using the sensible synthesis and functional assay combination described above we found what we believe could be a near-ideal anti-inflammatory molecule. This is however not the only thing we found. We found a new mechanism. Like most medicinal programs this project started with an idea to inhibit a certain interaction (in this case the binding of the chemokine CCL2 to its receptor) but we never succeeded with this. Of course to begin with we didn't know this, we just knew that the compounds we had made inhibit cell migration. While this could have reasonably been explained by chemokine receptor antagonism it was discovered that FX125L (or its analogues) do not bind to chemokine receptors. They still worked, but not how it had been planned. So this is the point - the discovery of this class of small molecule anti-inflammatory drug led to the discovery of a new biological mechanism: FX125L is an SSTR2 (somatostatin receptor) agonist. Not a classical one, but a ligand all the same, and one with unpredicted "somatotaxin" activity . The functional assay, far from being a "non-target-oriented" assay is, in fact, a method of discovering the new biological intervention points desperately needed to jump-start a sluggish pharmaceutical industry.
Special thanks to: There are many people who have helped on this voyage of discovery but I would like to mention a couple in particular: firstly Dr David Grainger who initiated and led the project. David's name is now all over the web because of FXT and his combination of science and business ideas. Perhaps less well known but equally important has been Dr Jill Reckless. Jill has been my "mirror" in biology and has organised and run most of the assays that have shown FX125L to be such a great molecule. It has been a great sharing this medicinal chemistry project with two such great scientists.
David Grainger's (DrugBaron blog) tells the FX125L story here
David Cavalla's (Numedicus blog) view on BI sale and functional assays here