Small Molecule Screening Strategies from Lead Identification to Validation

The genomics revolution has resulted in a vast amount of knowledge about the association between genetic perturbation and different disease phenotypes¹. This has led to the annotation of numerous proteins as potential drug targets and spurred interest in the discovery and development of compounds that inhibit these targets². In this context, different screening strategies have emerged as important approaches for hit compound discovery, target validation, measuring binding affinities, selectivity, toxicity, and in-cell target engagement, as well as compound reactivity, stability, and pharmacokinetic properties. This collection captures some of the methodological advances in this area with the goal of showcasing a multitude of strategies that can be used to screen and evaluate the biochemical and/or biological activity of small molecules.

Protein kinases represent a large enzyme family with over 500 members. These enzymes are key regulators of all cellular processes, and their malfunction has been linked to numerous diseases³. Therefore, they have been the focus of intense interest in drug discovery and, by extension, methodology development. Thus, four articles in this collection report on advances in this space. Festa et al. describe a screening platform for kinase inhibitor discovery that uses self-assembled nucleic acid programmable protein arrays (NAPPA)⁴. These arrays display full-length human kinases and allow for multiplexed screening of compounds across a large kinase collection, thus allowing not only for identifying hits but for profiling selectivity as well. Tieman et al. describe a phosphoantibody microarray-based method developed to screen for cell-wide changes in phosphorylation status as a readout of kinase inhibitor activity⁵. This can lead to insights into the nature of the pathways that have become resistant to inhibition, which is an important parameter relevant for understanding drug response and efficacy. Chen et al. describe a flow cytometry assay for screening kinase inhibitor collections to identify compounds that affect T-cell receptor (TCR) signaling⁶. This strategy led to the rapid identification of kinases involved in TCR signaling, indicating their potential as targets for immunotherapeutic drug development. Lastly, the activity of many kinases is regulated through protein-protein interactions (PPIs), suggesting that compounds that target these PPIs could be of interest. Nandha Premnath et al. describe a method called REPLACE (replacement with partial ligand alternatives using computational enrichment) that can be used to develop non-ATP competitive PPIs of kinases by converting features of peptide binders into small molecules that engage the same binding surface⁷.

In addition to kinases, many other enzymes play roles in disease and have been pursued in drug discovery. Therefore, methods optimized for enzymes other than kinases are also of interest. For example, Stockman et al. describe nuclear magnetic resonance (NMR) spectroscopy-based activity assays to discover and evaluate inhibitors of nucleoside ribohydrolases from the parasite Trichomonas vaginalis⁸. These assays can also be used for the determination of IC₅₀ values. To screen for ATPase inhibitors, Radnai et al. developed a semi-high throughput screening method that uses a coupled enzyme assay format involving two enzymes (pyruvate kinase and lactate dehydrogenase) that ultimately translates ATPase activity into intrinsic fluorescence of nicotinamide adenine dinucleotide (NADH) as a readout⁹. Unlike all the enzymes described thus far that are processive and catalyze multiple steps, proprotein convertase subtilisin/kexin type 9 (PCSK9) is a single-turnover protease that is autoinhibited by its substrate. This makes screening for inhibitors of PCSK9 challenging, and Chorba et al. describe a high-throughput cell-based method that links PCSK9 cleavage activity to the secreted luciferase activity as a readout, thus enabling screening for inhibitors¹⁰.

In addition to finding compounds that inhibit the activity of various targets, assays that confirm that the compound engages its intended target in cells are also of high interest. Axelsson et al. describe a strategy for validating target engagement in adherent cells based on combining high-content imaging by immunofluorescent antibodies and a cellular thermal shift assay (CETSA)¹¹. This assay uses an isothermal heat challenge to cells treated with a large compound collection to identify stabilizing compounds. A method that can assess target engagement and activity of degrader molecules, developed by Riching et al., uses HiBiT/LgBiT tagging technology to produce a luminescent protein and monitor degradation parameters via a decrease in luminescence¹². Lastly, Takakusagi et al. describe a method for validating protein-drug interactions that is based on using phage-based peptide displays to screen for binding between drug-binding peptide motifs and small molecules¹³. The peptide display screen is combined with bioinformatic analysis to validate target binding.

Another layer of small molecule validation is the characterization of their effects in animal models. This collection includes two such methods, one that uses zebrafish (Danio rerio) models of human influenza A virus infection to evaluate the efficacy of compounds with antiviral activity, as well as the host immune response¹⁴, and another that employs Caenorhabditis elegans as a model for profiling compound toxicity¹⁵. Both of these studies highlight the need for optimized methods that can evaluate efficacy and toxicity in vivo.

In conclusion, drug discovery and development are formidable processes with numerous bottlenecks, challenges, and a high probability of failure. Additionally, continued disease burden, be it in the area of infectious diseases where the number of drug-resistant pathogens has been growing at an alarming rate or in the areas of increasing prevalence of chronic diseases and diseases associated with the aging population, demands our focused attention and action. This collection highlights several methods that aim to accelerate drug discovery, validation, and characterization. Going forward, we expect to see further innovation in this space, especially those aimed at improving the throughput of screening, increasing the diversity of chemical libraries, and developing in vivo models compatible with high-throughput screening (HTS). Furthermore, workflows that provide a holistic view of potency, selectivity, specificity, efficacy, and toxicity on a faster and smaller scale could be especially enabling.

M.K. is a paid consultant for Life Science Editors.

M.K. is partly supported by the NCI T32 CA236754.