Natural products (NPs) are a major source of bioactive molecules for drug discovery, yet their development and translation are often limited by inefficient and ambiguous target identification. Although mass spectrometry-based proteomics has advanced rapidly, upstream sample preparation remains a critical bottleneck for high-throughput target deconvolution. Here, we report μPAS (micro proteomics automation system), an automated and miniaturized proteomic sample preparation platform that integrates protein reduction, alkylation, digestion, and TMTpro labeling into a single streamlined workflow. By achieving a 3- to 7-fold reduction in digestion and labeling volumes, μPAS improves throughput and cost efficiency, reducing TMT reagent consumption by 2-7.5-fold while maintaining high digestion efficiency (>90% within 4 h) and TMTpro labeling efficiency (>96%). The platform demonstrates consistent intra- and inter-batch reproducibility, with Pearson correlation coefficients exceeding 0.96. Using three model compounds, μPAS was benchmarked against three complementary target identification strategies, enabling automated target discovery. Application of μPAS to a 96-sample workflow enabled systematic target deconvolution for 18 NPs lacking well-defined targets. Key candidate targets, including HIF1AN, FECH, and TXNRD1, were further validated using Western blot-based thermal shift assays, confirming target engagement. Collectively, these results establish μPAS as a robust and scalable platform for high-throughput NP target discovery, facilitating mechanistic elucidation of NP bioactivity.
Wu et al. (Mon,) studied this question.