Protein complex prediction using Rosetta, AlphaFold, and mass spectrometry covalent labeling

Abstract

Structural mass spectrometry offers several techniques for the characterization of protein structures. Covalent labeling (CL) in combination with mass spectrometry can be used as an analytical tool to study and determine structural properties of protein-protein complexes. Degrees of modification obtained from CL experiments for specific labeled residues can be compared between the unbound and bound states of complexes. This analysis can yield insights into structural features of these protein assemblies, specifically the proximity of specific residues to the protein-protein interface. However, this data is sparse and does not unambiguously elucidate protein structure. Thus, computational algorithms are needed to deduce structure from the CL data. In this work we present a novel hybrid method that combines models of protein complex subunits generated with AlphaFold with differential CL data via a CL-guided protein-protein docking in Rosetta. In a benchmark set, the RMSD (root-mean-square deviation) of the best-scoring models was below 3.6 Å for 5/5 complexes with inclusion of CL data, whereas the same quality was only achieved for 1/5 complexes without CL data. The average improvement in RMSD observed upon inclusion of CL data was 5.2 Å. This study suggests that our integrated approach can successfully use data obtained from CL experiments to distinguish between nativelike and non-nativelike models. Significance Statement: Structural mass spectrometry can be a powerful and versatile approach to characterize the structure of protein complexes. Data obtained from covalent labeling mass spectrometry can provide insights into higher order protein structure (particularly with respect to residue interactions and solvent accessibility) but needs to be supplemented by computational techniques to elucidate accurate, atomic-detail structural information. Here, we present a method to combine bioanalytical data obtained from covalent labeling with models generated using AlphaFold to accurately predict protein-protein complexes in Rosetta. Differential covalent labeling data can be used to determine the proximity of residues to the binding interface of complexes which we utilized to analyze computational models and improve structure prediction algorithms.