Reversible covalent inhibitors: the best of both worlds?

What’s not to love about covalent inhibitors? Well, unfortunately, quite a bit when you start thinking about it. Reversible inhibitors offer an extra layer of subtlety, they are tuneable, time-dependent and can modulate individual functions of proteins or biological systems by using several inhibitors in combination. That is as long as you can make them potent and selective enough; covalent inhibitors tend to solve the first problem in spades, once you make your covalent bond it’s staying there usually knocking out the enzyme target. Yet the greatest strength is the greatest weakness, unfortunately by improving the potency of your compounds the problems with selectivity are multiplied. If your covalent inhibitor goes to the wrong protein first well tough; this great if you’re a beta-lactam antibiotic trying to kill a bacterium, but killing things isn't what tool compounds are for. 

But what if you could make a reversible covalent inhibitor? Potentially you could gain the potency and, if well designed, the selectivity as well. Maybe this would still be difficult to incorporate into a drug, but as a basis for a sophisticated tool compound it could be extremely useful. Taunten et.al. have developed such a system that targets the amino acid cysteine. 

After several test experiments they found cyanoacrylamides to be the most reliable substrates, for there reactivity and reversibility, against mercaptoethanol and other simple thiols. They then developed several potential inhibitors against RSK2 (ribosomal kinase 2) based around a known kinase inhibitor. The aim being that one of the cyanoacrylamide inhibitors would bond covalently to nearby Cys-436. After profiling many different kinases (442 of them) only 6 showed greater than 70% inhibition. Of those 6 the top inhibitor targeted RSK2 with at least an 80-fold higher affinity and usually greater than 400; that’s pretty impressive selectivity for a kinase inhibitor and they were able to crystalise the molecule within the protein to gain further insights into its binding mode. Several control experiments were also performed to ensure that it was the presence of a cysteine that was conferring the potency and selectivity. Interestingly they noted that bonding of their inhibitor caused the protein to denature and this seemed to encourage inhibitor release however there were problems with permanent covalent bonds forming to highly reactive cysteines on other proteins.

I really like this idea, it explores the possibilities of a tool compound offering the potential to create selective and highly potent compounds with the benefits of the traditional small-molecule inhibitor, while at the same time remaining sensibly drug-like. I think covalent inhibitors are an under explored area and that a reversible covalent inhibitor gets rid of lots of the objections to pure covalent inhibitors (even though are some  clinical candidates and others have been phenomenally successful). But part of me also thinks this is yet another concept for targeting protein kinases. I hope the authors manage to get it to work for other protein targets as well especially ones that are currently under explored. Indeed the authors do hint that they want to do this in the discussion, I for one hope they are successful.