Cracking the code of cellular protein-protein interactions: Alphafold and whole-cell crosslinking to the rescue

Abstract

Integration of experimental and computational methods is crucial to better understanding protein-protein interactions (PPIs), ideally in their cellular context. In their recent work, Rappsilber and colleagues (O'Reilly et al, 2023) identified bacterial PPIs using an array of approaches. They combined whole-cell crosslinking, co-fractionation mass spectrometry, and open-source data mining with artificial intelligence (AI)-based structure prediction of PPIs in the well-studied organism Bacillus subtilis. This innovative approach reveals architectural knowledge for in-cell PPIs that are often lost upon cell lysis, making it applicable to genetically intractable organisms such as pathogenic bacteria.

Figure 1. The Integrative workflow to interrogate cellular protein–protein interactions used by O'Reilly et al (2023)

(A) Protein complexes were crosslinked and evidence was provided that in‐cell interactions are retained and can be annotated using co‐fractionation mass spectrometry, network biology, and crosslinking mapping. (B) These interactions were modeled and predicted using crosslink‐scored Alphafold models, providing functional hypotheses for the direct interaction of unknown proteins in Bacillus subtilis. (C) Biochemical validation of novel complexes involved measurements of analytical ultracentrifugation, growth strain kinetics, and bacterial 2‐hybrid assays.