Comparison of Non-labeling Quantitative Proteomics Techniques Using Ion Trap and QTOF Mass Spectrometers

摘要

Introduction: Relative quantitation of proteins within complex samples is an increasingly important aspect of proteomics. Two common non-labeling workflows are often cited: ion trap or QTOF mass spectrometry combined with ion intensity or spectral counting versus comparison of peptide peak height intensities. The purpose of this study was to compare these two commonly utilized quantitative methodologies using a moderately complex mixture of known peptides with spike-ins. Importantly, the comparison was made using a QTOF or ion trap with identical front end components and methods. A commercially available, standard 48-protein mixture was trypsinized and C18 cleaned and then several proteins spiked in at two- and four-fold molar concentrations. Samples were analyzed using identical HPLC parameters, including a chip cube source on either an Ion Trap (n(velence)3) or QTOF mass spectrometer (n(velence)5). The exact same chip column was used on both instruments, rendering the front end portion of the analyses identical. Ion trap data were quantitated using total peptide precursor ion intensity or peak height intensity in a database searching program. Peptides were analyzed in MS mode on the QTOF and differentially abundant mass features were determined using Genespring MS. Targeted MSMS analyses of differentially regulated features was used to assign protein identifications using the QTOF. Results: In a relatively rapid (40 minute) single MS/MS run on either instrument, 40 of the 48 proteins are identified with high confidence (high peptide scores and at least 2 peptides per protein) and 4 proteins by a single, high-quality peptide. In a comparison of non-spiked versus 4x-spiked all 5 spike proteins, but no non-spike proteins, were significantly different (P < 0.05). Comparisons between other samples (0 vs 2x and 2x vs 4x) showed that the significance of the differences observed appears to be correlated with the number of peptides detected. In general, there appeared to be a correlation between a high number of peptides detected and low standard deviations for ion trap data. Two-fold differences were detected in 4 out of 5 proteins in the ion trap. In order to determine the limitations of quantitating high abundant proteins, we spiked serotransferrin at 1.5x concentration, which was also detected by the ion trap. The results underscore the importance of running technical replicates in either system.