Tuning Kinase Drug Residence Times for Reversible Covalent Inhibitors

Drug residence time, the length of time that an inhibitor remains bound to its target, is increasingly being recognized as a key parameter impacting pharmacodynamic activity. Longer residence times translate into higher efficacy and fewer side effects because a lower circulating dose of drug is required to maintain an effective level of target engagement. One approach being pursued for achieving longer residence times is development of inhibitors that bind their targets covalently. However, irreversible covalent inhibitors can increase the risk of off-target effects and of immune responses to permanent covalent adducts. Covalent inhibitors can bind reversibly as well, which reduces some of this risk, and researchers at Principia Biopharma and University of California, San Francisco (UCSF) recently collaborated on a powerful approach for tuning the residence time of reversible covalent kinase inhibitors (Nat Chem Biol. 2015 Jul;11(7):525-31).

A team led by Michael Bradshaw at Principia built upon an approach for targeting a non-catalytic active site cysteine residue with a cyano-acrylamide based electrophile (CAE) that the lead investigator from UCSF, Jack Taunton, developed for the ribosomal protein S6 kinase RSK2. The team focused initially on Cys 481 in Bruton’s tyrosine kinase (BTK), a highly targeted kinase for B-cell malignancies and autoimmune disorders. The first BTK inhibitor drug, ibrutinib, was approved last year for chronic lymphocytic leukemia and mantle cell lymphoma.

They used a structure guided approach, starting with scaffolds for which BTK co-crystal structures were available, to design a vector for the CAE. To provide good access to Cys 481, they had to invert the orientation of the CAE relative to the previous RSK2 inhibitor. Based on the hypothesis that kinetically trapping covalent complex would increase residence time, they tested different  amine linkers between the scaffold and the CAE and varied the branched chain alkyl groups used to cap the electrophilic carbon. They assessed residence time of inhibitors in vitro and in cells by measuring the durability of their binding following addition of an excess of a fluorescent active site BTK probe.

Analysis of a series of 21 compounds in which the CAE was attached to a pyrazolopyrimidine scaffold with different linkers and alkyl capping groups revealed a graded continuum of BTK residence times ranging from a few hours to seven days. For comparison, the non-covalent inhibitor dasatinib inhibits BTK with an IC₅₀ of 0.5 nM, and its residence time is only 30 minutes. They also measured inhibitor potencies using enzymatic assays, and found little correlation between residence times and potencies, which were all in the sub-nanomolar to low nanomolar range.

Importantly, the investigators showed that the covalent inhibitors dissociated rapidly and intact from denatured BTK, indicating that interactions with the native kinase were required to trap the inhibitor in the active site and that the binding was truly reversible. They went on to demonstrate that more durable binding in vitro translates into longer residence time in cells. And, most impressively, they showed that the inhibitor with the week-long residence time remained bound to more than 40% of the BTK in rat peripheral blood mononuclear cells 24 hours after dosing the animal, when the serum level had dropped to less than 5% of the IC₅₀ value. This demonstrates the therapeutic value of long residence time: achieving effective target engagement without the need to sustain a high circulating level of drug.

The team then went on to apply a similar approach to another kinase, targeting Cys486 of the fibroblast growth receptor (FGFR1), which has previously been targeted by an irreversible covalent inhibitor. They produced a series of 11 compounds that potently inhibited FGFR (IC₅₀ < 10 nM) and showed a wide range of residence times ranging again from a few hours to almost a week. Cys486 is located in a different region of the FGFR active site than the Cys residues targeted in BTK and RSK2, indicating that varying the CAE linker and capping group may be a general approach for tuning kinase inhibitor residence time.

The results of this study are certainly impressive in terms of the week-long inhibitor residence times achieved, which makes them the most durable reversible inhibitors known. But the real impact is in the ability to develop a series inhibitors based on the same scaffold with a continuum of residence times, which is very likely to be applicable to other kinases and perhaps to different enzyme targets as well. (About 40% of kinases have what is considered to be an ‘approachable’ cysteine in the ATP binding pocket.) This would make it possible to analyze how residence time affects drug efficacy and side effects in different disease models, and optimize the residence time for specific diseases or even patient groups.

Learn more about residence time in our Application Note:
Determination of Drug-Kinase Residence Time with the Transcreener ADP² FP Assay