We are thrilled to share our latest publication in Nature Communications, “A Machine-Learning-Based Chemoproteomic Approach to Identify Drug Targets and Binding Sites in Complex Proteomes”. The publication demonstrates the utility of our Limited Proteolysis (LiP) technology and workflow for drug target identification. Our next-generation proteomics approach combines LiP with machine-learning-based data analysis to enable the identification of small molecule drug targets in complex proteomes. The work was performed in collaboration with ETH Zurich, Bayer, and BASF and its findings are summarized below:
Understanding a Compound’s Mechanism of Action
Target identification is a critical step in the development and optimization of drug candidates. While in phenotypic drug discovery (PDD) target identification provides essential information for the elucidation of the mechanism of action (MoA) of the bioactive compound, the profiling of potential (off)targets in a target-based drug discovery setting may help to predict potential adverse effects.
Overcoming Limitations of Traditional Approaches
Current target identification approaches such as affinity-based chemoproteomics or thermal proteome profiling (TPP) have shown their utilities but have clear caveats concerning compound modification and binding site identification respectively.
To address these shortcomings, we developed a novel workflow based on Limited Proteolysis (LiP). A major advantage of LiP is its unique focus on the detection of signature peptides that discern ligand binding. Combining drug dose-titration, LiP, quantitative DIA-MS (Data Independent Acquisition – Mass Spectrometry), and a machine-learning-based data analysis framework, we devise an integrated pipeline (LiP-MS) which enables the identification of drug targets including the prediction of binding sites and affinity.
Innovation through Collaboration
We demonstrated the ability of LiP-MS for unbiased drug target identification across several compound classes in different biological matrices. First, we showcased that this approach can be applied to a specific kinase inhibitor (selumetinib) as well as an unspecific kinase inhibitor (staurosporine). Subsequently, we characterized the specificity of LiP-MS by robustly identifying the target proteins of two natural product-derived phosphatase inhibitors (calyculin A and fostriecin). Last but not least, in collaboration with Bayer, we identified the target(s) of a novel fungicide candidate (BAYE-004) and predicted the putative MoA by pinpointing the compound binding site.
LiP-MS – The Solution for Drug Target Identification
In conclusion, LiP-MS deploys orthogonal biophysical principles in comparison to existing methods (namely protein structural alterations and steric hindrance) for the identification of drug-protein interactions with peptide-level resolution. These capabilities make LiP-MS a powerful identification strategy and valuable addition to the target identification toolbox.
If you would like to view the original publication or find out more about the application of our LiP technology to our service offerings, please click on the icons below: