New types of covalent fragments

As covalent drugs become more accepted, covalent fragment libraries are becoming more popular: we’ve previously written about both reversible and irreversible fragments. One potential limitation is the number of different types of covalent modifiers, or warheads. The program DOCKovalent, for example, only considers ten classes. György Keserű and collaborators at the Hungarian Academy of Sciences and the University of Ljubljana go some way towards expanding this list in a recent paper in Arch Pharm Chem Life Sci.

The researchers were interested in heterocyclic electrophiles. For heterocycles, they considered pyridines, pyrimidines, pyrazines, imidazoles, pyrazoles, oxazoles, isoxazoles, and thiazoles. For electrophiles, they considered chloride, bromide, iodide, nitrile, vinyl, and ethynyl groups, often at different positions around a given heterocycle. So for example, they chose 2, 3, and 4-chloropyridine. Not every electrophile was available or easily synthesized for every heterocycle, so in total they assembled a library of 84 different fragments.

The library was tested for aqueous stability, and all but six fragments had half-lives of at least 24 hours at pH 7.4. Next, the researchers examined the intrinsic reactivities of their molecules by reacting them with glutathione, a physiologically relevant thiol. As might be expected, the different molecules showed a wide range of different reactivities, all of which are reported in the paper. This is a useful list: ultimately one wants a warhead with low or modest reactivity for better selectivity.

Next, the researchers tested their fragments against MurA from Staphylococcus aureus and Escherichia coli; this enzyme is important for bacterial cell wall biosynthesis, and contains an active-site cysteine that has previously been shown to be sensitive to electrophiles. The reactivity patterns were similar between the two enzymes, but they did differ somewhat from glutathione reactivity, suggesting the possibility of molecular recognition. Dose response assays were performed on the most potent molecules, most of which had IC₅₀ values in the mid to high micromolar range. These results expand on research we highlighted six years ago showing that halopyridines could covalently modify a cysteine-dependent enzyme.

The researchers also examined the mechanism of their fragments by doing time-dependence and dilution experiments. Some of the results are quite unexpected, suggesting that, for example, 4-iodopyridine is a reversible modifier, which is hard to understand mechanistically. Perhaps, like the “universal fragment” 4-bromopyrazole, the molecule does not act covalently, though the time dependence observed suggests otherwise.

This is a nice example of how to create and assess a fragment library with a particular mechanism in mind, reminiscent of the metal-chelating fragments described by Seth Cohen and colleagues. Finally, I’d like to note that the first author, Aaron Keeley, is part of the FragNet training program. He and his fellow trainees will soon be looking for the next phase in their careers, so if you’re hiring keep them in mind!