Achieving depth, quantitative accuracy and throughput are key targets of method development, with significant challenges in plasma proteomics. Here, suppressive effects of plasma’s dynamic range have driven development towards improved detection of low-abundance proteins, including new sample workflows, many which are biofluid- and species-agnostic. Alongside this, increasing attention has been paid towards adapting LC-MS methods from complex samples (i.e. lysate) to plasma, where 75% of protein mass is estimated as 2000 and 7 proteins, respectively.
Herein, we explored eight sample methods for plasma spanning neat (SP3, Strap), depletion (Perchloric Acid (PerCA)) and enrichment strategies (MagNet HILIC/SAX, Enrich, Nanomics), and compared their performance in human plasma and sera, alongside rat plasma. We analysed these on an Orbitrap Astral (Thermo) using two plasma-optimised DIA methods, a discovery-maximised MS-method (65 minute, 25cm C18 ‘Aurora’ series, Ionopticks), and a throughput-maximised MS-method (30 minutes, 15cm C18 ‘Aurora’ series, Ionopticks).
We identified (Spectronaut, DirectDIA (v19)) a non-redundant 2,801 human proteins (40,433 precursors) and 3,663 rat proteins (56,911 precursors) using our discovery-maximised method, with 1,834 human proteins (31,382 precursors) and 2,938 rat proteins (42,438 precursors) with our throughput-maximised method. For human plasma, all neat methods identified >1100 reproducible proteins in the discovery-maximised, and >900 in the throughput-maximised (Strap). In all biofluids and methods MagNet SAX and HILIC beads, separately or combined, saw 40-50% increase in precursors compared to neat preparations. Furthermore, species-specific differences in dynamic range moderated improvements in some methods, whilst performance differences between plasma or serum resulted in >20% differences in unique protein identifications. Furthermore, using the DIA-analyst suite, > 85% of proteins were significant in either species across pairwise comparisons between methods.
Our study highlights sample and instrument workflows dramatically influence the resultant proteome within species, between species, and between biofluids, whereby no ‘Cinderella’ situation exists for a perfect fit in plasma proteomics.