Direct evidence of acid-driven protein desolvation

Water and its ability to modulate the protonation states of biomolecules govern the physical chemistry of life, dictating their metabolic functions. However, how amino acid protonation alters protein hydration and solubility is an open question since Kuntz and Kauzmann proposed pH-driven protein desolvation in 1974. Here, in a series of high-resolution cryoelectron microscopy structures of a protein complex at different pH values (from pH 9.0 to 3.5), we examined thousands of observable hydration sites. Cryoelectron microscopy data, in agreement with constant-pH molecular dynamics simulations, show that nearly half of protein-bound waters exchanged with the bulk solvent upon acidification, with ~100 waters lost per pH unit per molecule. The loss of waters was most significant around the side chains of glutamate and aspartate residues while specific polar residues, mostly asparagine, anchored persistent waters. A positionally conserved hydration layer was observed across all pH conditions, accounting for 40% of resolved waters. Those waters displayed denser packing than less persistent waters, forming a pH-independent solvation shell. Acid-induced water exchange also displaced bound iron, providing a mechanistic link between solvation and metal release. Our findings demonstrate the core principles of acid-driven protein desolvation, resolving a 50-y-old biochemical hypothesis.