Cross-linking mass spectrometry (XL-MS) is a powerful tool for probing protein structures and protein-protein interactions. While chemical cross-linkers target specific residues with defined chemistry, photo-cross-linkers offer superior reactivity but have been hampered by incomplete mechanistic understanding and lack of robust analytical framework. Here, we demonstrate that diazirine-based photo-cross-links are inherently MS-cleavable, generating composite backbone and side-chain fragments, which have nevertheless confounded spectral interpretation. Yet by leveraging the side-chain fragmentation fingerprints (sFFP), we develop a machine learning model and subsequently, a rule-based filtering algorithm. When integrated with existing search platforms, our workflow significantly improves ion coverage and reduces false discovery rate for site identification. We further develop a homo-bifunctional diazirine cross-linker, allowing for cross-linking on-demand. This reagent captures transient tetrameric assemblies of human HSP90β and reveals structural transitions in association equilibrium under heat stress, details otherwise inaccessible with chemical cross-linking. Together, this work establishes a transformative framework in XL-MS, combining the temporal resolution of photo-activation with analytical confidence for residue-level structural insights.
Jiang et al. (Mon,) studied this question.