In an ideal world, where we can rationally design our drugs to target only the one receptor or enzyme we aim for and there are no side-effects or adverse reactions, we could find the ideal compound for every application and then dial up the prescription depending on the patient – the ultimate personalized medicine.
Meanwhile back in the real world, the messy, non-selective and unexpected actions of drugs reign. And even where you see good selectivity by a drug, in many cases, especially in the cancer field, the therapy is found to be ineffective, because the disease – the cancer, the virus, whatever – has found a way to get along without that component that was previously thought indispensible.
In short, it has become at least useful to see exquisite selectivity as a missed opportunity to get two birds with one stone. This is particularly true in the kinase field, where many hopes for treating cancer have been vested. A recent article by Richard Morphy in the Journal of Medicinal Chemistry talks about this. A good number of drugs were in any case non-selective or at least with limited selectivity, but it turned out that as long as you got the right off-target kinases as well as your one, there were not any ill-effects. In fact, Morphy notes there has been an increase in programs specifically looking for multiple targets from a single drug.
This is an immensely difficult thing to achieve, as each change to a molecule will result in a cascade of changes in the inhibitory profile, which is overlooking the changes that will be seen in physiochemical and, as a result, pharmacokinetic properties. So actually tuning the activity to exactly what you want is nigh on impossible and in most cases, you have to look at what you found and take your best shot at which one is “the best”, as it is not a trivial matter to determine how much each activity against a particular kinase contributes to the overall efficacy of the drug. It is quite common, for example, to find you have binding with a particular kinase but when you run a functional assay, you find that the compound has little or no activity. Adding in the complexities of the human body in all its pharmacological glory makes determining what you just did when you gave someone that drug tricky.
But when we make a kinase inhibitor, we don’t only have to worry about other kinases that our super-drug might target. Oh no, that would be too easy. It might also inhibit – or even stimulate – activity elsewhere in the complex human machinery. Drug discovery programs often look at cytochrome P450 inhibition, as that is a huge red flag for a drug due to the potential for drug-drug interactions to cause problems (see St John’s wort for an example). But it is much more wide spread than that, according to research by Brian Roth and Brian Shoichet and reported in a recent C&E News, research that may go some way to explain some of the side effects drugs have.
So we see that attempting to tease out amazingly selectivity over one kinase from another is a thankless task, given how many other things it may end up interacting with. All we can really do is try to avoid interactions with particularly important targets and then hope for success in the toxicology screens. That sounds a bit cavalier, but in reality, the clinical effects of these compounds will be muted by other factors such as protein binding, clearance and the like. And at the end of the day, what other choice do you have?