Multi-Platform Proteomic Analysis of Alzheimer’s Disease Cerebrospinal Fluid and Plasma Reveals Network Biomarkers Associated with Proteostasis and the Matrisome

Abstract

Robust and accessible biomarkers that can capture the heterogeneity of Alzheimer’s disease and its diverse pathological processes are urgently needed. Here, we undertook an investigation of Alzheimer’s disease cerebrospinal fluid (CSF) and plasma from the same subjects using three different proteomic platforms—SomaLogic SomaScan, Olink proximity extension assay, and tandem mass tag-based mass spectrometry—to assess which protein markers in these two biofluids may serve as reliable biomarkers of AD pathophysiology observed from unbiased brain proteomics studies. Median correlation of overlapping protein measurements across platforms in CSF (r∼0.7) and plasma (r∼0.6) was good, with more variability in plasma. The SomaScan technology provided the most measurements in plasma. Surprisingly, many proteins altered in AD CSF were found to be altered in the opposite direction in plasma, including important members of AD brain co-expression modules. An exception was SMOC1, a key member of the brain matrisome module associated with amyloid-β deposition in AD, which was found to be elevated in both CSF and plasma. Protein co-expression analysis on greater than 7000 protein measurements in CSF and 9500 protein measurements in plasma across all proteomic platforms revealed strong changes in modules related to autophagy, ubiquitination, and sugar metabolism in CSF, and endocytosis and the matrisome in plasma. Cross-platform and cross-biofluid proteomics represents a promising approach for AD biomarker development.

Novel brain-penetrant inhibitor of G9a methylase blocks Alzheimer’s disease proteopathology for precision medication

“…Recently, our mechanistic study of inflammation control revealed that constitutively active G9a is the upstream regulator of METTL3/m 6 A-mediated translation in chronically inflamed macrophages that mimic AD-related neuroinflammation 18 . In agreement with findings that translational regulation was broadly perturbed in various neurodegenerative diseases, 19,20,21 and loss of proteostasis is an AD hallmark, 22,23 our in vivo ChaC-MS identification of G9a interactions with translation regulators such as METTL3 revealed a new, noncanonical function of G9a in the translational or posttranslational regulation of AD pathogenesis. Because translational mechanisms ultimately determine function-related protein abundance or modification states 24 , we developed a new mechanism-based therapeutic for AD, i.e., MS1262, a brain-penetrant inhibitor of G9a, to block G9a-mediated translation of AD-related proteins.…”

“…6B). For example, the level of SPARC-related modular calcium-binding protein 1 (SMOC1), an Aβ plaque-associated synaptic protein 23,107 was found elevated in AD CSF nearly 30 years before the onset of symptoms 28 . Other CSF markers 28 showing MS1262 affected/reversed expression included the following (Fig.…”

Quantitative Mass Spectrometry Analysis of Cerebrospinal Fluid Protein Biomarkers in Alzheimer’s Disease

“…Significant correlations of these peptides to the established biomarker and cognitive measures indicate the potential of these measurements to classify or stage disease progression. The targeted SRM measurement correlations largely agree with those observed from unbiased discovery proteomics 43 and parallel reaction monitoring 23 experiments.…”

“…The comparison of existing biomarkers to the SRM peptide measurements can be accomplished by correlation, where the degree of correlation indicates how similar a peptide measurement is to the established immunoassay-measured biomarkers of Aβ(1-42), total Tau, and pTau as well as cognition (MoCA score). In Figure 6A, we demonstrate that 57 of the 58 The targeted SRM measurement correlations largely agree with those observed from unbiased discovery proteomics 43 and parallel reaction monitoring 23 experiments.…”

“…In Figure 6A, we demonstrate that 57 of the 58 biomarker peptides have significant correlation to at least one of the above biomarkers, or the ratio of total Tau/Aβ. Individual correlation scatterplots and linear fit lines for three of the peptides (SMOC1: AQALEQAK, YWHAZ: VVSSIEQK, and VGF: EPVAGDAVPGPK) are provided in The targeted SRM measurement correlations largely agree with those observed from unbiased discovery proteomics 43 and parallel reaction monitoring 23 experiments.…”