Files
Abstract
Mass spectrometry (MS) is the most widely utilized analytical tool for the large-scale study of an organisms proteome. This has created lofty expectations for the field of proteomics; however the study of biological systems continues to be daunting task due to the extreme complexity and wide dynamic range of protein expression. Multidimensional separation techniques have been incorporated into MS-based proteomic workflows to overcome this challenge. Unfortunately, improvements in separating power come at the expense of MS analysis time, thus implementation of highly efficient separation strategies becomes necessary to achieve high throughput. In this work, we describe the development of high efficiency reversed-phase liquid chromatography (LC) separation methods. The use of superficially porous column packing materials permitted fast LC separations, and optimization of data-dependent acquisition (DDA) parameters allowed for the collection of high quality MS/MS spectra when experiment time was reduced. The use of formic acid and ammonium formate (FA/AF) as a mobile phase modifier was found to be compatible with electrospray ionization, and provided a significant improvement in peptide separations over formic acid alone. This combination of high efficiency LC separations and optimized DDA parameters lead to a significant reduction in experiment time and substantial increases in proteome coverage. The efficiency of 1D gel electrophoresis and LC (GeLC) separations were evaluated to determine how to maximize protein identifications in a fixed instrument time format. This work demonstrates that the number of gel slices collected in GeLC analysis has very little impact on protein identifications. The most significant factor is GeLC protein identification efficiency is the percentage of instrument time dedicated to LC gradient elution conditions. These newly developed, high efficiency GeLC separation methods were applied to the proteomic analysis of the canine prostate gland. Canines and humans are the only two large mammals that spontaneously generate prostate cancer, thereby suggesting a potential predictive model of the human disease. Our work identified several proteins with association to prostate cancer development and progression, suggesting the canine could be a relevant predictive model of androgen-insensitive, highly aggressive human prostate cancer.