Blog | July 15, 2021

Applying DNA-Encoded Chemistry to the Discovery of Irreversible Covalent Inhibitors

By Anthony D. Keefe, Ph.D. – Senior Vice President, Alliance Management and Innovation

Some of the most commonly used drugs elicit their therapeutic effects by forming covalent bonds with their targets; examples include aspirin and penicillin. Covalent drugs like these offer the advantages of increased  potency and permanent target engagement which in turn results in more sustained effects. Even so, it seemed that until relatively recently there was a palpable fear of actually setting out to discover new covalent drugs. Enthusiasm around this approach seems to now be on the rise again with approval of covalent irreversible oncology drugs targeting BTK and KRAS G12C.

Pharmaceutical companies have spent literally billions building up their compound collections over the last century or so, but the vast majority of these compounds are incapable of covalently engaging their targets.  Accordingly, even if one wanted to set out to directly screen for covalent inhibitors using conventional high-throughput screening approaches it is not immediately clear where one would find the compound collection to do so. Indeed, most recently discovered covalent drugs were derived from non-covalent screening hits that were subsequently appended with electrophiles. 

X-Chem performed a proof-of-concept to see if see if the discovery of covalent inhibitors could be done at scale.  In order to make large numbers of different compounds capable of covalent modes of action we turned to DNA-encoded chemical libraries. Because this technology both synthesizes and screens compounds as mixtures, it is able to more rapidly generate and utilize new collections of compounds when compared to the laborious alternative of generating a collection of compounds by separately synthesizing each and every one. We took an existing X-Chem DNA-encoded chemical library which had twenty-six million different compounds, all of which were terminated in a primary or secondary amine and so were amenable to further functionalization. Compounds with the potential to engage targets covalently need to contain electrophiles.  Installing acrylamide onto this library as a mixture, i.e., in a single tube, was extremely straightforward.  For this project we ultimately installed seven different electrophiles for a total of 182 million different compounds.

The next step was a little more tricky.  Unlike conventional DNA-encoded chemical libraries which may be put through multiple successive rounds of affinity-mediated selection by alternately providing an opportunity to bind to a target, immobilizing it, washing and then liberating the bound members, a covalent bond, once formed, is permanent, which is of course the whole point.  So we had to develop methods that were capable of sufficiently enriching active library members on the basis of a single target-library interaction event.  And we had to deal with the fact that once the desired bond is formed the encoding DNA is now covalently attached to the protein target, complicating the recovery of encoding sequence information.

Our recent paper details how we optimized this process using on-DNA controls including ibrutinib (covalent) and dasatinib (non-covalent) and ultimately arrived at a method utilizing native solution incubation followed by capture on IMAC magnetic beads followed by denaturing washes with guanidine followed by elution with imidazole.

We then applied these methods to the electrophile-displaying libraries we had just made and discovered novel covalent BTK inhibitors as a result. We found distinct active compounds displaying both acrylamide and an epoxide. The epoxides are the first ever inhibitors of BTK that engage the target using this electrophile. The compounds were characterized by a range of techniques at Proteros Biostructures in Munich including reporter displacement, thermal shift, intact protein mass spectrometry and x-ray crystallography.  Potency, covalency and specifity were all confirmed.

If your project calls for a combination of novel equity and a covalent mode-of-action and you are wondering where you can find hundreds of millions of electrophilic compounds for the purposes of screening, then we strongly encourage you to consider DNA-Encoded Chemistry. With this recent experiment, we have demonstrated that this approach can work even for libraries of large numerical size.

Read more about this study in our peer-reviewed publication Novel irreversible covalent BTK inhibitors discovered using DNA-encoded chemistry in Bioorganic Medicinal Chemistry. After these successful proof-of-concept experiments, X-Chem helped its partners by identifying and then licensing multiple irreversible covalent inhibitors for multiple therapeutic targets.

X-Chem’s technology is currently the only demonstrated method available for identifying irreversible target engagers that allows the screening of hundreds of millions of compounds simultaneously. This platform has broad applicability and has proved its value in the discovery of hits and leads with a covalent mode of action. If your comfort with covalent drugs has reached the point where you are ready to act, then X-Chem has the technology, expertise and experience to help you.

Reference

John P. Guilinger, Archna Archna, Martin Augustin Andreas Bergmann, Paolo A. Centrella, Matthew A. Clark, John W. Cuozzo, Maike Däther, Marie-Aude Guié, Sevan Habeshian, Reiner Kiefersauer, Stephan Krapp, Alfred Lammens, Lukas Lercher, Julie Liu, Yanbin Liu, Klaus Maskos, Michael Mrosek, Klaus Pflügler, Markus Siegert, Heather A. Thomson, Xia Tian, Ying Zhang, Debora L. Konz Makino, and Anthony D. Keefe. Novel irreversible covalent BTK inhibitors discovered using DNA-encoded chemistry. Bioorganic & Medicinal Chemistry, 2021.  https://doi.org/10.1016/j.bmc.2021.116223

Back to Blog