Chemical proteomics focuses on the drug–target–phenotype relationship for target deconvolution and elucidation of the mechanism of action—key and bottleneck in drug development and repurposing. Majorly due to the limits of using chemically modified ligands in affinity-based methods, new, unbiased, proteome-wide, and MS-based chemical proteomics approaches have been developed to perform drug target deconvolution, using full proteome profiling and no chemical modification of the studied ligand. Of note among them, thermal proteome profiling (TPP) aims to identify the target(s) by measuring the difference in melting temperatures between each identified protein in drug-treated versus vehicle-treated samples, with the thermodynamic interpretation of “protein melting” and curve fitting of all quantified proteins, at all temperatures, in each biological replicate. Including TPP, all the other chemical proteomics approaches often fail to provide target deconvolution with sufficient proteome depth, statistical power, throughput, and sustainability, which could hardly fulfill the final purpose of drug development. The proteome integral solubility alteration (PISA) assay provides no thermodynamic interpretation, but a throughput 10–100-fold compared to the other proteomics methods, high sustainability, much lower time of analysis and sample amount requirements, high confidence in results, maximal proteome coverage (~10,000 protein IDs), and up to five drugs / test molecules in one assay, with at least biological triplicates of each treatment. Each drug-treated or vehicle-treated sample is split into many fractions and exposed to a gradient of heat as solubility perturbing agent before being recomposed into one sample; each soluble fraction is isolated, then deep and quantitative proteomics is applied across all samples. The proteins interacting with the tested molecules (targets and off-targets), the activated mechanistic factors, or proteins modified during the treatment show reproducible changes in their soluble amount compared to vehicle-treated controls. As of today, the maximal multiplexing capability is 18 biological samples per PISA assay, which enables statistical robustness and flexible experimental design accommodation for fuller target deconvolution, including integration of orthogonal chemical proteomics methods in one PISA assay. Living cells for studying target engagement in vivo or, alternatively, protein extracts to identify in vitro ligand-interacting proteins can be studied, and the minimal need in sample amount unlocks target deconvolution using primary cells and their derived cultures.
Graphical abstract: