Automated extraction of backbone deuteration levels from amide hydrogen/deuterium mass spectrometry experiments to understand EX1 protein unfolding and determine protein binding interfaces

H/2H exchange coupled with mass spectrometry is a powerful, fast tool to study protein-protein interactions and dynamic motions of proteins. Unfortunately, natural isotopic abundances create difficulty in analyzing H/2H exchange, and side chain deuteration (when present) further adds a complicating factor that also dilutes the observable results. Both isotopic profiles and side chain exchange can be modeled by sums of binomial equations with high correlation to measured results, which creates the ability to overcome the difficulty of analysis by mathematical extraction. Back exchange during the sample preparation stage also minimizes real differences between experimental states, but it too can be explicitly modeled and corrected for.; A Fourier Transform deconvolution method is used to extract natural isotopic abundances and it reveals the backbone amide exchange levels. This data is then processed to determine the different levels of backbone exchange by three available methods; centroids, populations and peak widths. Centroids measure the average number of deuteriums incorporated into a sample; populations measure the specific fraction of the total with a particular number of exchanged deuterons; and peak widths measure the linear interpolated distance between the first and last peak in a mass spectrometry profile that meets a certain count requirement. Of these methods, only the first was widely available before the deconvolution method was implemented. This mathematical calculation was implemented in a small computer program and configured to handle mass spectrometry data in a high-throughput manner, and it is available to the public under the name DEX. These improvements in mathematical analysis of H/2H exchange data gave better understanding and improved detection of EX1 unfolding of local regions of protein domains where the unfolding is rapid and incomplete, making identification of it previously difficult. The software package and analysis tools also led to a more rapid analysis of the PKA RIIbeta-C interface for various domains and cofactors. This led to a greater understanding of the different ways the regulatory isoforms of the catalytic subunit interact and regulate phosphorylation.