From reversible to irreversible fragments vs caspase-6

The caspase family of cysteine proteases is critical for initiating and executing programmed cell death, and some members also play a role in inflammation. Humans have a dozen caspases, all of which cleave substrate peptides immediately after an aspartic acid residue. Developing chemical probes for specific family members has been challenging for several reasons, not least because of similar active sites. In a newly published open-access J. Am. Chem. Soc. paper, Michelle Arkin, Adam Renslo, and colleagues at UCSF have done just that. (Michelle spoke about this work at the CHI DDC meeting last month.)

 

The researchers were particularly interested in caspase-6, which has been implicated in neuroinflammation. They recognized that, in addition to the active site cysteine, caspase-6 has another cysteine not found in other caspases. C264 sits at the far end of the active site, on the smaller of the two separate protein subunits that come together at the active site. They used Tethering (previously described here) to screen a library of 1500 disulfide-containing fragments. Since C264 is on the small subunit and the active site cysteine is on the large subunit, it was easy to determine specificity using mass spectrometry.

 

There were quite a few hits against the small subunit of caspase-6, including compound 1. The researchers replaced the disulfide moiety with an irreversible vinylsulfonamide warhead and made a few other modifications to the core to obtain compound 2a. This showed sub-micromolar inhibition of the enzyme after just 15 minutes, and crystallography revealed that it bound covalently to C264, as expected. The crystal structure also suggested where to grow the molecule, leading to compound 3a, which was a low nM inhibitor at 15 minutes. The researchers determined the k_inact and K_i values, which are important parameters for irreversible inhibitors. The molecule was also quite selective, showing little to no inhibition of nine other human caspases at 10 µM in biochemical assays.

 

Compound 3a effectively blocked caspase-6 mediated cleavage of the substrate lamin A in a cellular assay. To assess cellular specificity, the researchers made an alkyne-containing version of compound 3a which could be used in pull-down experiments using click chemistry. This revealed a couple dozen protein targets, of which only eight could be competed by the parent compound 3a. Interestingly, all except caspase-6 were membrane-bound proteins, and none were proteases.

 

This is a nice, concise paper reminiscent of the early work that ultimately led to sotorasib. A key liability of the series is the vinylsulfonamide warhead, which is probably too intrinsically reactive for in vivo studies. The researchers did try replacing this with the clinically validated acrylamide warhead, but unfortunately the resulting molecule was not active. Nonetheless, compound 3a is worth considering as a chemical probe, and the researchers note that further optimization efforts are ongoing. I look forward to seeing the evolution of these molecules.