Native MS: turning up the voltage

At the FBDD meeting in San Diego last month there was some discussion of the connection between protein-ligand binding in the gas phase (ie, in native mass-spectrometry experiments) vs in solution; the discussion continued in the comment section here. Valerie Vivat, head of the mass spectrometry, molecular & cell biology department at NovAliX, has now provided a thorough response.

When analyzing non-covalent complexes by mass spectrometry under non-denaturing conditions, one must keep in mind indeed how interactions stabilizing the complex in solution will be affected by the ion transfer in the gas phase. Basically, interaction based on hydrophobic effect is lost while the electrostatic-based interactions (Van der Waals, H-bonds, ionic interaction) are preserved (and are even strengthened due to the absence of solvent shielding). Regarding water molecules now, during the ESI process, biomolecules are transferred in the gas phase as partially hydrated ions and complete ion desolvation is subsequently achieved through low energy collisions with the residual gas molecules in the mass spectrometer interface. This complete desolvation is required for accurate mass measurements. Consequently, in most cases, molecular mass measurement of the complexes is accurate enough to rule out the possibility that water molecules remain after ions are transferred from the atmospheric pressure to the deep vacuum of the TOF analyzer, even in the case of crystallographic water molecules.

Regarding gas phase stability : Reports in the literature mention that collision-induced dissociation experiments and the determination of Vc50s can be used to monitor gain or loss in polar interaction. Indeed, we have now accumulated a variety of in house examples showing that increase in the gas phase stability of a complex is correlated with a gain in polar interactions as shown by X-ray or ITC. Monitoring the complex gas phase stability is thus an attractive feature to quickly evaluate, for example, gain or loss in polar interaction of analog series targeting the same protein binding site.

Importantly, the gas phase stability is not an indication of the complex binding affinity. Evaluation of complex affinity (after complexes are formed in solution) is done under instrumental conditions showing no dissociation of the non-covalent complexes. For complexes stabilized by electrostatic-based interactions (and thus rather stable in the gas phase), it is indeed much easier to find optimal instrument settings compared to complexes stabilized by hydrophobic effect. However, this does not preclude complexes mainly stabilized in solution by hydrophobic effect to be detected at all. For example, nuclear hormone receptors in complex with fatty acids or phospholipids are readily detected with appropriate settings. In this case, the polar head of the ligands is likely to act as a stable anchor in the gas phase, resulting in non-covalent complexes which remain intact during the ca. 1 millisecond flight of the ion from the ionization source to the detector.

To sum up, native MS can be run under two different modes: 

1.     Low energy conditions where no complex dissociation occurs are used to assess the binding affinity of protein / ligand complexes,

2.     High energy conditions inducing complex dissociation provide insight into the extent of polar interaction involved between the protein and the ligand. This type of experiment makes sense to compare a series of molecules binding to the same protein binding site.

Valerie makes some interesting points here, in particular the statement that "the gas phase stability is not an indication of the complex binding affinity." Still, I do like the fact that native MS could be used as a quick way to sort enthalpic from entropic binders. What do you think?