15NH/D-SOLEXSY Experiment for Accurate Measurement of Amide Solvent Exchange Rates: Application to Denatured Drkn SH3
Veniamin Chevelkov,Yi Xue,D. Krishna Rao,Julie D. Forman-Kay,Nikolai R. Skrynnikov
DOI: https://doi.org/10.1007/s10858-010-9398-8
2010-01-01
Journal of Biomolecular NMR
Abstract:Amide solvent exchange rates are regarded as a valuable source of information on structure/dynamics of unfolded (disordered) proteins. Proton-based saturation transfer experiments, normally used to measure solvent exchange, are known to meet some serious difficulties. The problems mainly arise from the need to (1) manipulate water magnetization and (2) discriminate between multiple magnetization transfer pathways that occur within the proton pool. Some of these issues are specific to unfolded proteins. For example, the compensation scheme used to cancel the Overhauser effect in the popular CLEANEX experiment is not designed for use with unfolded proteins. In this report we describe an alternative experimental strategy, where amide 15N is used as a probe of solvent exchange. The experiment is performed in 50% H2O–50% D2O solvent and is based on the (HACACO)NH pulse sequence. The resulting spectral map is fully equivalent to the conventional HSQC. To fulfill its purpose, the experiment monitors the conversion of deuterated species, 15ND, into protonated species, 15NH, as effected by the solvent exchange. Conceptually, this experiment is similar to EXSY which prompted the name of 15NH/D-SOLEXSY (SOLvent EXchange SpectroscopY). Of note, our experimental scheme, which relies on nitrogen rather than proton to monitor solvent exchange, is free of the complications described above. The developed pulse sequence was used to measure solvent exchange rates in the chemically denatured state of the drkN SH3 domain. The results were found to correlate well with the CLEANEX-PM data, r = 0.97, thus providing a measure of validation for both techniques. When the experimentally measured exchange rates are converted into protection factors, most of the values fall in the range 0.5–2, consistent with random-coil behavior. However, elevated values, ca. 5, are obtained for residues R38 and A39, as well as the side-chain indole of W36. This is surprising, given that high protection factors imply hydrogen bonding or hydrophobic burial not expected to occur in a chemically denatured state of a protein. We, therefore, hypothesized that elevated protection factors are an artefact arising from the calculation of the reference (random-coil) exchange rates. To confirm this hypothesis, we prepared samples of several short peptides derived from the sequence of the drkN SH3 domain; these samples were used to directly measure the reference exchange rates. The revised protection factors obtained in this manner proved to be close to 1.0. These results also have implications for the more compact unfolded state of drkN SH3, which appears to be fully permeable to water as well, with no manifestations of hydrophobic burial.