Effects of protein-ligand interactions on hydrogen/deuterium exchange kinetics: canonical and noncanonical scenarios

Hydrogen/deuterium exchange (HDX) methods are widely used for monitoring protein-ligand interactions. This approach relies on the fact that ligand binding can modulate the extent of protein structural fluctuations that transiently disrupt hydrogen bonds and expose backbone amides to the solvent. It is commonly observed that ligand binding causes a reduction of HDX rates. This reduction can be restricted to elements adjacent to the binding site, but other regions can be affected as well. Qualitatively, ligand-induced HDX protection can be rationalized on the basis of two-state models that equate structural dynamics with global unfolding/refolding. Unfortunately, such models tend to be unrealistic because the dynamics of native proteins are dominated by subglobal transitions and local fluctuations. Ligand binding lowers the ground-state free energy. It is not obvious why this should necessarily be accompanied by a depletion of excited-state occupancies, which would be required for a reduction of HDX rates. Here, we propose a framework that implies that ligand binding can either slow or accelerate amide deuteration throughout the protein. These scenarios are referred to as "type 1" and "type 2", respectively. Evidence for type 1 binding is abundant in the literature, whereas the viability of type 2 interactions is less clear. Using HDX mass spectrometry (MS), we demonstrate that the oxygenation of hemoglobin (Hb) provides a dramatic example of a type 2 scenario. The observed behavior is consistent with cooperative T → R switching, where part of the intrinsic O2 binding energy is reinvested for destabilization of the ground state. This destabilization increases the Boltzmann occupancy of unfolded conformers, thereby enhancing HDX rates. Surprisingly, O2 binding to myoglobin (Mb) also induces elevated HDX rates. These Mb data reveal that type 2 behavior is not limited to cooperative multisubunit systems. Although enhanced protection from deuteration is widely considered to be a hallmark of protein-ligand interactions, this work establishes that an overall deuteration increase also represents a viable outcome. HDX-based ligand screening assays, therefore, have to allow for canonical as well as noncanonical effects.