Native mass spectrometry of membrane protein lipid and ligand complexes using de novo designed WRAP domains

Membrane proteins engage in dynamic transient interactions with surrounding lipids along with their substrates. Capturing these concomitant contacts often remains a challenging problem. Here, we demonstrate that generative deep learning-designed WRAP domains can preserve weak interactions of membrane protein complexes in native mass spectrometry. Using WRAP-fused GlpG, AqpM, and OmpA as model systems, we show that lipid interactions can be retained and characterized without detergent micelles. We use the approach to show that WRAP-OmpA selectively binds phosphatidylethanolamine lipids within cavities formed at the protein–WRAP interface, while simultaneously accommodating weak chitobiose ligand binding. These findings establish WRAPs as versatile vehicles to probe ligand and lipid interactions of membrane proteins.