Assessment of reproducibility in depletion and enrichment workflows for plasma proteomics using label-free quantitative data-independent LC-MS.

Quantitation in plasma-based proteomics necessitates the reproducible removal of highly abundant proteins to enable the less abundant proteins to be visible to the mass spectrometer. We have evaluated immunodepletion (proteoprep20) and enrichment (Bio-Rad beads), as the current predominant approaches. Label-free analysis offers an opportunity to estimate the effectiveness of this approach without incorporating chemical labels. Human plasma samples were used to quantitatively assess the reproducibility of these two methods using nano-LC-data-independent acquisition MS. We have selected 18 candidate proteins and a comparison of both methodologies showed that both of the methods were reproducible and fell below 20% residual SD. With the same candidate proteins, individual inter-day variability for the samples was also processed, allowing us to monitor instrument reproducibility. Overall, a total of 131 proteins were identified by both methods with 272 proteins identified by enrichment and 200 identified by immunodepletion. Reproducibility of measurements of the amount of protein in the processed sample for individual proteins is within analytically acceptable standards for both methodologies. This enables both methods to be used for biomarker studies. However, when sample is limited, enrichment is not suitable as larger volumes (>1.0 mL) are required. In experiments where sample is not limited then a greater number of proteins can be reliably identified using enrichment.