MSACL 2023 Abstract

4D Proteomic Profiling of 200 Individuals Plasma with DIA-MS and Internal Standards for a Health Surveillance Panel-A High-Throughput Workflow Qin Fu1, Robin Park2, Ali Haghani1, Tomasz Siwarga2, Michael Krawitzky4, Jonathan Krieger3, Tharan Srikumar3, Blandine Chazarin1, Esthelle Hoedt1, Sandy Joung1, Susan Cheng1; Christopher Adams4 and Jennifer E. Van Eyk1 1Cedars-Sinai Smidt Heart Institute, Cedars Sinai Medical Center, Los Angeles, CA; 2Bruker, San Diego, CA; 3Bruker Ltd., Milton, ON; 4Bruker, Inc., San Jose, CA Qin Fu, PhD (Presenter) Cedars Sinai Medical Center

Presenter Bio: Qin Fu earned her PhD degree in Genetics from the University of Minnesota and received post-doctoral training in Immunology at the University of California, San Francisco. She has research experience in human genetics, autoimmunity, drug resistance, cardiovascular diseases, and biomarker discovery and validation. Her current research interests include automated MS sample preparation, multiplexed targeted MS assays, and biomarker discovery. Dr. Fu is the Director of the High-Throughput Laboratory in the Advanced Clinical Biosystems Research Institute at Cedars Sinai Medical Center.

Introduction
There is a profound need in establishing a defined plasma proteomic baseline of a healthy population and establish normal reference ranges for quantifiable peptides. Human plasma contains thousands of proteins and makes contact with virtually all cells in the human body. Therefore, the plasma proteome is reflective of an individual’s health status. It is feasible to develop an analytic method for a Health Surveillance Panel (HSP) comprised of selected and relevant plasma proteins. We have selected plasma proteins with important biological functions and involvement in diverse pathways with 50% being FDA or LDT tests. We investigated 4D proteomic profiles of 200 healthy individuals by employing the following: 1) automated sample preparation; 2) 83 spiked in stable isotopic labeled heavy peptides (representing HSP proteins) as internal standards; 3) timsTOF Pro dia-PASEF (Parallel Accumulation Serial Fragmentation); 4) dia-PASEF MS acquisition and implementation of automated LC-MS/MS quality controls; 5) novel library building algorithm to calibrate and normalize CCS (collisional cross-sectional) values; 6) CCS-enabled DIA-NN for neural networks search algorithm (Tims DIANN) for analyzing data.

Objectives
This is the first in-depth study of 4D plasma profiling in a healthy population. Development of a robust, precise, quantitative and automated 4D proteomic profiling workflow to profile healthy individuals with dia-PASEF MS methodology, spiked internal standards and to establish the proteomics healthy population baseline (n=200).

Methods
Development of the workflow, reproducibility and linearity of the tryptic peptides were carried out using a gender-pooled plasma from 100 females and 100 males. Protein denaturation, reduction, alkylation, digestion and desalting were performed on a Beckman i7 automated workstation. Digested individual plasma samples (800 ng) with 83 SIL heavy peptides (57 HSP proteins) were injected onto an PepSep column attached to an EVOSEP One coupled to a Bruker timsTOF Pro mass spectrometer. Plasma samples were injected in triplicate. The workflow was validated on plasma samples from 200 healthy individuals enrolled in Coronavirus Risk Associations and Longitudinal Evaluation (CORALE) Study and run robustness was evaluated based on quality control runs of predigested plasma (n=5) as an end-to-end workflow control (system suitability).

Results
Healthy individuals’ plasma (n=200) were baseline of the CORALE Study and who self-reported as being healthy and were not COVID-19 positive comprising 45% female, age ranged 21 to 69 years and composed of four ethnic groups: 7% African American or black, 22% White Hispanic, 37% Asian and 34% Caucasian. For data analysis using our newly developed DIA data analysis suite that builds spectral libraries from either real-time data or existing peptide libraries from different sources (e.g. multiple in-house DDA libraries, and libraries in the public domain). DIA data was analyzed using our improved CCS-enabled version of DIA-NN that calculates ion mobility-based scores for both target and decoy precursors and build features for deep learning. The spectral library tool calibrates CCS values and re-normalizes fragment ion intensities, which is used in the CCS-enabled DIA-NN for neural networks. We applied our complete workflow to plasma from 200 individuals. The pooled QC plasma (system suitability) was used for MS and full workflow performance. The reproducibility, CV, LLOQ and linearity were established for 83 SIL peptides in the plasma matrix. The 4-D proteomic baseline profiled 200 baseline healthy individuals using timsTOF dia-PASEF MS methodology, novel data analysis and bioinformatic package including library building/calibrating and an improved CCS-enabled DIA-NN search algorithm.

Conclusion
Development of a robust, precise, quantitative, and automated proteomic profiling workflow with HSP SIL peptides as internal standards using a dia-PASEF -MS methodology, novel library building and CCS-enabled DIA-NN algorithm to profile 200 individuals 4D proteomic plasma baseline.