Reply to Kitahara and Mulder: An ensemble view of protein stability best explains pressure effects in a T₄ lysozyme cavity mutant

Kitahara and Mulder (1) challenge our interpretation of pressure-perturbation studies of the Leu99Ala (L99A) cavity mutant of the T₄ lysozyme. Kitahara and Mulder suggest we incorrectly discarded a model where the native state (NS) interconverts with a highly structured high-energy state (HES) to explain NMR phenomena observed at elevated pressure. It is first important to be clear as to the meaning of “unfolding.” Nucci et al. (2) note that the responses observed are “…indicative of subglobal and local unfolding events and largely correspond to the cooperative unfolding of secondary structural units and domains…” Also, contrary to what is implied by Kitahara and Mulder, using site-resolved cross-peak intensities (integrals) is a valid means to obtain the apparent free energy difference between the NS and an ensemble of unfolded states. Having said that, the reasons to discount the Kitahara and Mulder two-structure interconversion model in favor of an ensemble of partially unfolded states are many. The pressure sensitivity of the protein is heterogeneous, which argues strongly against the global two-state interpretation. Leu133 is simply provided as representative of many examples of the lack of simple correlation with the HES. Critically, small angle X-ray scattering data by Ando et al. (3) show that the pressure-induced states are more extensive than the uniquely structured and compact NS and HES. Thus, the suppression of the pressure sensitivity by confinement in a reverse micelle and by crystallization is most consistent with our view. Kitahara and Mulder argue that the pressure sensitivity of the native state is relieved by reorganization to the HES where the ring of Phe-114 fills the cavity. In an elegant study, Bouvignies et al. (4) show that binding of benzene is mutually exclusive with the HES. The suppression of unfolding transitions by occupation of the cavity by benzene is therefore consistent with both views. All of this and more was used to support our original interpretation.

In ref. 5, simulations are used to illustrate that some effects in the NMR spectra are consistent with the interconversion of two structures. A multitude of scenarios are equally consistent, so this provides no clarity. They failed to detect “random coil” resonances indicative of high disorder and solvation at high pressure. Although difficult to discern in the supplemental figures to Nucci et al., such resonances are found in our data. This demolishes the two-state interpretation. Their failure to detect the multitude of states may arise from use of a very low volume NMR cell apparently without the benefit of a cryogenic probe. We also show that water does not significantly enter the cavity until pressures greater than 1 kbar are reached. This is directly inconsistent with Kitahara and Mulder, where the HES continues to be populated. Thus, although we agree that interconversion of these states is on a time scale that complicates the NMR spectra, there is overwhelming evidence that the application of pressure to T₄ L99A results in the population of a large number of distinct partially unfolded states and does not simply shift the native state to the HES.