Exploring the precursor ion exclusion feature of liquid chromatography-electrospray ionization quadrupole time-of-flight mass spectrometry for improving protein identification in shotgun proteome analysis

Abstract

In shotgun proteome analysis by liquid chromatography-tandem mass spectrometry (LC-MS/MS), not all coeluting peptides at a given retention time are subjected to MS/MS due to the limitation of spectral acquisition speed of a mass spectrometer. In this work, precursor ion exclusion (PIE) in an electrospray ionization (ESI) quadrupole time-of-flight (QTOF) mass spectrometer was explored as a means of mitigating the undersampling problem. This strategy is based on running replicates of the sample where the precursor ions detected in the initial run(s) are excluded for MS/MS in the subsequent run. Four PIE methods as well as running replicates without PIE were investigated and compared for their effectiveness in identifying peptides and proteins. In the analysis of an MCF-7 breast cancer cell lysate digest by three replicate 2 h gradient LC-ESI runs, the first PIE method used a list of precursor ions detected in the initial run(s) for exclusion and identified a total of 572 proteins from the three runs combined with an average of 3.59 peptides matched to a protein. The second PIE method involved in the generation of a list of m/ z values of precursor ions along with their retention time information from the initial run(s), followed by entering these ions with retention times into the ion exclusion program of the QTOF control software for exclusion at a predefined retention time window (i.e., +/-150 s). In comparison to the first PIE method, this method reduced the possibility of excluding different peptide ions of the same m/ z (within a mass tolerance window) eluted at different retention windows. A total of 657 proteins were identified with an average of 3.75 peptides matched to a protein. The third PIE method studied relied on the exclusion of the precursor ions of peptides identified through database search of the MS/MS spectra generated in the initial run(s). This selective PIE method identified a total of 681 proteins with an average of 3.68 peptides matched to a protein. The final PIE method investigated involves the expansion of the selective PIE list by including nonidentifiable peptide ions found in the database search. This complete PIE method identified a total of 726 proteins with an average of 3.66 peptides per protein. In the case of three replicate runs without PIE, a total of 460 proteins were identified with an average of 3.51 peptides matched to a protein. Thus, the use of an optimal PIE strategy significantly increased the number of proteins identified from replicate runs (i.e., 726 vs 460 or a 58% increase). It is further demonstrated that this PIE strategy also improves protein identification efficiency in the analysis of a yeast whole cell lysate digesta less complex proteome digest. A total of 533 proteins identified from five replicate runs with complete PIE, compared to 353 proteins identified from the five replicate runs without PIE, representing a 51% increase in the number of proteins identified.