Multi-tissue high-throughput proteomics profiling for neurodegenerative disease Yun Ju Sung1,2, Chengran Yang1, Herve Rhinn3, Joanne Norton1, Fengxian Wang1, Joseph Bradley1, Fabiana Farias1, Bruno A. Benitez1, Oscar Harari1, Carlos Cruchaga1,4 1:Department of Psychiatry, Washington University School of Medicine, St Louis, MO, USA; 2:Division of Biostatistics, Washington University School of Medicine, St. Louis, MO, USA; 3:Department of Bioinformatics. Alector, Inc. 151 Oyster Point Blvd. #300 South San Francisco CA, USA; 4:The Charles F. and Joanne Knight Alzheimer Disease Research Center, Washington University School of Medicine, St. Louis, MO, USA Background: Alzheimer disease (AD) is the most common cause of dementia characterized by the hallmark pathologies: amyloid-beta (Aβ) and tau. Autosomal-dominant AD (ADAD) individuals carry pathogenic mutations in the APP, PSEN1, or PSEN2 genes, dominantly inheriting AD. Our group and others recently identified several rare variants of triggering receptor expressed on myeloid cells 2 (TREM2) that strongly increases the risk of developing AD. Methods: Deep proteomics profiling was obtained (SOMAscan; 1,305 proteins) from brain (n=450), cerebrospinal fluid (CSF, n=1,301), and plasma (n=648) tissue. These neurologically relevant tissues were obtained from well-characterized Knight-ADRC and DIAN participants with comprehensive clinical information about AD pathology and cognition. Results: After stringent QC and data cleaning, we performed analysis for 1079 proteins in the brain (n=370), 713 proteins in CSF (n=699), and 931 proteins in plasma (n=486). We identified 12, 117, and 26 AD-specific proteins (with Bonferroni-corrected statistical significance) in the brain, CSF, and plasma, respectively. Twenty-seven of the proteins identified in CSF showed differential levels in brain and plasma (P < 0.05) and higher prediction accuracy than the well-accepted pTau/Aβ42 ratio (AUC=0.87 vs. 0.81, P=4.1×10-4) regardless APOE status. We also identified 27, 38, and 69 TREM2-specific proteins in the brain, CSF, and plasma. Twenty-three of the proteins identified in plasma showed differential levels in CSF and brain. A prediction model of these 23 proteins was able to discriminate TREM2 carriers from controls (AUC=0.94) and other AD cases (AUC=0.91). Finally, we identified 371 ADAD-specific proteins in the brain, among which 74 proteins showed differential levels in CSF and plasma. Pathway analysis of AD-specific proteins are involved in angiogenesis (FDR P=9.2×10-8), hemostasis (FDR P=1.6×10-12) and growth factor signaling (FDR P=9.5×10-10). TREM2-specific proteins are associated with growth factors including VEGF (FDR P=1.9×10-9), PDGF (FDR P=2.9×10-6), EGF (FDR P=3.6×10-6) and immunological response. ADAD-specific proteins converge in immunological response pathways (FDR P=1.1×10-40) including cytokine-mediated signaling (FDR P 6.2×10-37) and DAP12-mediated pathway (FDR P 6.2×10-21). Conclusions: This is the first high-throughput proteomics study for genetically defined AD cases and TREM2 risk variant carriers. Multiple proteins significantly associated with status for each AD group were found in CSF, brain, and plasma. These findings not only help create novel prediction models but also point to specific pathways implicated in AD, elucidating the functional mechanisms underlying the genetic architecture of this neurodegenerative disease.
Washington University School of Medicine in St Louis
You must be logged in and own this session in order to