Improved quantitative plant proteomics via the combination of targeted and untargeted data acquisition

Quantitative proteomics strategies – which are playing important roles in the expanding field of plant molecular systems biology – are traditionally designated as either hypothesis driven or non-hypothesis driven. Many of these strategies aim to select individual peptide ions for tandem mass spectrometry (MS/MS), and to do this mixed hypothesis driven and non-hypothesis driven approaches are theoretically simple to implement. In-depth investigations into the efficacies of such approaches have, however, yet to be described. In this study, using combined samples of unlabeled and metabolically ¹⁵N-labeled Arabidopsis thaliana proteins, we investigate the mixed use of targeted data acquisition (TDA) and data dependent acquisition (DDA) – referred to as TDA/DDA – to facilitate both hypothesis driven and non-hypothesis driven quantitative data collection in individual LC-MS/MS experiments. To investigate TDA/DDA for hypothesis driven data collection, 7 miRNA target proteins of differing size and abundance were targeted using inclusion lists comprised of 1558 m/z values, using 3 different TDA/DDA experimental designs. In samples in which targeted peptide ions were of particularly low abundance (i.e., predominantly only marginally above mass analyser detection limits), TDA/DDA produced statistically significant increases in the number of targeted peptides identified (230 ± 8 versus 80 ± 3 for DDA; p = 1.1 × 10^-3) and quantified (35 ± 3 versus 21 ± 2 for DDA; p = 0.038) per experiment relative to the use of DDA only. These expected improvements in hypothesis driven data collection were observed alongside unexpected improvements in non-hypothesis driven data collection. Untargeted peptide ions with m/z values matching those in inclusion lists were repeatedly identified and quantified across technical replicate TDA/DDA experiments, resulting in significant increases in the percentages of proteins repeatedly quantified in TDA/DDA experiments only relative to DDA experiments only (33.0 ± 2.6% versus 8.0 ± 2.7%, respectively; p = 0.011). These results were observed together with uncompromised broad-scale MS/MS data collection in TDA/DDA experiments relative to DDA experiments. Using our observations we provide guidelines for TDA/DDA method design for quantitative plant proteomics studies, and suggest that TDA/DDA is a broadly underutilized proteomics data acquisition strategy.