B-249 Assessing Targeted Mass Spectrometry Assay Performance for Putative Biomarker Proteins on a New Hybrid Nominal Mass Instrument

D L Plubell,W Phipps,A Emmanuel,G Merrihew,L R Heil,P M Remes,C C Jacob,C P Martins,S M Peterman,A N Hoofnagle,M J MacCoss
DOI: https://doi.org/10.1093/clinchem/hvae106.606
IF: 12.114
2024-10-27
Clinical Chemistry
Abstract:BackgroundThe development of targeted mass spectrometry-based proteomics assays that monitor biomedically relevant proteins is an important step in bridging discovery experiments to highly multi-analyte clinical studies. Targeted assays are currently unable to scale to thousands of peptide targets from hundreds of proteins, requiring extensive refinement down assays for the minimum useful analytes. Large parallel reaction monitoring (PRM) assays are now possible with a new hybrid nominal-mass instrument. The scale of assay is achievable with this instrument by using real-time alignment, while being sensitive and fast enough to handle many concurrent targets. To assess the performance of this new instrument we constructed a large scale PRM assay for proposed neurodegenerative disease marker proteins, and an assay for proteins associated with amyloid tissue typing.MethodsTo evaluate the performance of this instrument on proteins of interest in cerebrospinal fluid, we performed an analysis on a pool from individuals diagnosed as Alzheimer's, Parkinson's, or cognitively normal. A calibration curve of cerebrospinal fluid diluted into chicken serum at 14 dilutions and individual human cerebrospinal fluid were digested with a protein aggregation and capture based digestion protocol. CSF peptides were separated by 60 minute gradient on C18 IonOpticks column, and amyloid tissue peptides separated by 30 minute gradient on C18 PepMap column with a Vanquish Neo UHPLC and analyzed by data-independent acquisition (DIA) and PRM with a new hybrid nominal mass instrument designed for targeted tandem mass spectrometry (tMS2). Scheduled methods were generated using a Skyline external tool, transitions were refined to minimize interference and maximize the LLOQ. DIA data was searched, and data extracted and analyzed in Skyline. PRM data was extracted and quantified with Skyline; CSF calibration curve LLOQ is calculated by bilinear fit.ResultsFor the large scale cerebrospinal fluid (CSF) assay the selection of protein targets was based on literature from neurodegenerative disease cohort studies. We target 1065 peptides from 104 proteins, including proteins currently measured by ELISA in clinical research, e.g.: APP, TAU, and APOE. Sensitivity and reproducibility of the assay is assessed using figures of merit from replicate matched-matrix calibration curves. Over 80% of proteins have at least one peptide with a LLOQ below 30% dilution in the matched matrix, and of all targeted peptides over 55% had a LLOQ below 30% dilution. To demonstrate the application of the real-time alignment on the new instrument we constructed an assay for samples where not every analyte is shared. An assay for amyloidosis associated proteins was constructed for a total of 180 peptides mapping to 52 proteins. This includes 11 proteins most commonly associated with amyloid deposition; APCS, APOA4, APOE, CLU, LECT2, IGKC, IGLC2/4/7, SAA1, TTR, and VTN. From the preliminary set of 12 previously typed samples we observe peptide abundances consistent with the assigned type.ConclusionsThis novel instrument provides a new approach for building large targeted mass spectrometry assays. We demonstrate the ability of the real-time alignment to adjust scheduled retention time windows, ensuring that chromatographic shifts do not lead to a loss of peptide detection and quantification.
medical laboratory technology
What problem does this paper attempt to address?