Background

Site-directed ligand discovery was first reported in 2000 and utilizes surface-exposed cysteine residues to covalently trap disulfide-linked fragments that bind at an adjacent pocket (Erlanson et al., 2000). Since then, irreversible cysteine-targeting inhibitors have grown immensely in popularity and now represent front-line therapies in multiple oncology indications, with notable examples targeting BTK and EGFR as well as previously undruggable targets such as KRAS(G12C). Following these developments, covalent fragment-based ligand discovery has emerged as an efficient route towards the design of target-specific covalent inhibitors (Resnick et al., 2019). Libraries of cysteine-reactive fragments are now commercially available from a range of compound vendors and utilize a variety of reaction types for covalent bond formation such as conjugate addition and nucleophilic substitution. Protein mass spectrometry was adopted early on as a useful assay for conducting covalent fragment screening, providing a direct readout on the extent and the site of fragment binding. However, even the earliest reports of covalent fragment-based ligand discovery emphasized the challenge that fragment-dependent electrophilicity presents (Nonoo et al., 2012): the intrinsic reactivity of covalent fragments can vary by several orders of magnitude within a library. Fragments with high intrinsic reactivity will covalently bind to proteins regardless of their structure and this reactivity-driven binding does not constitute a useful starting point for fragment optimization. In pioneering work by Statsyuk, it was demonstrated that, through careful design of the reactive warhead architecture, it is possible to limit the variations in intrinsic reactivity, resulting in more robust screening outcomes (Kathman et al., 2014). However, in many cases it is desirable to screen a wide variety of warhead structures, maximizing the exploration of chemical space and enlarging the pool of accessible targets (Keserű et al., 2016).

Quantitative irreversible tethering (qIT) is a plaform for screening cysteine-reactive fragments that has been designed to overcome the intrinsic reactivity problem (Craven et al., 2018). The fluoregenic probe CPM (7-diethylamino-3-(4'-Maleimidylphenyl)-4-Methylcoumarin) is used to quantify the rate of reaction between covalent fragments and cysteine residues. Covalent fragments are simultaneously screened against glutathione (GSH) as well as the target protein and the rates of reactions (k_obs) determined in high-throughput. GSH functions as a control cysteine containing biomolecule that lacks significant secondary structure. The rate of reaction between a fragment and GSH represents a measure of the covalent fragment’s intrinsic reactivity. Hit fragments are those that react significantly faster with the target protein than with GSH, as quantified by the rate enhancement factor (REF = k_obs(target)/k_obs(GSH). Screening by REF analysis ensures that qIT can be carried out in high-throughput and does not suffer from high false-positive/negative hit rates because it accounts for intrinsic fragment reactivity. The key advantage of screening by quantitative irreversible tethering is that, in accounting for intrinsic fragment reactivity, true hits can be readily distinguished from non-selective reactive molecules which otherwise lead to high levels of false positives. The platform is compatible with a broad range of cysteine-reactive fragments and is applicable to most cysteine-containing soluble protein domains. Additionally, the high-throughput nature and low cost of using a plate-based fluorescence readout means that quantitative irreversible tethering projects can be performed rapidly and with high efficiency.

Since the original publication of qIT (Craven et al., 2018), we have updated the screening protocol to make it more applicable to a wider variety of proteins by changing the buffer system, more high-throughput by reformatting from 96-well to 384-well format and easier to implement by changing the environment of operation from 4 °C to room temperature. In our lab we have successfully applied this updated protocol to a range of protein classes including kinases, GTPases, oxidoreductase and proteases.