Integrated analysis of genomics, longitudinal metabolomics, and Alzheimer's risk factors among 1,111 cohort participants

Abstract

Although Alzheimer's disease (AD) is highly heritable, genetic variants are known to be associated with AD only explain a small proportion of its heritability. Genetic factors may only convey disease risk in individuals with certain environmental exposures, suggesting that a multiomics approach could reveal underlying mechanisms contributing to complex traits, such as AD. We developed an integrated network to investigate relationships between metabolomics, genomics, and AD risk factors using Wisconsin Registry for Alzheimer's Prevention participants. Analyses included 1,111 non-Hispanic Caucasian participants with whole blood expression for 11,376 genes (imputed from dense genome-wide genotyping), 1,097 fasting plasma metabolites, and 17 AD risk factors. A subset of 155 individuals also had 364 fastings cerebral spinal fluid (CSF) metabolites. After adjusting each of these 12,854 variables for potential confounders, we developed an undirected graphical network, representing all significant pairwise correlations upon adjusting for multiple testing. There were many instances of genes being indirectly linked to AD risk factors through metabolites, suggesting that genes may influence AD risk through particular metabolites. Follow-up analyses suggested that glycine mediates the relationship between carbamoyl-phosphate synthase 1 and measures of cardiovascular and diabetes risk, including body mass index, waist-hip ratio, inflammation, and insulin resistance. Further, 38 CSF metabolites explained more than 60% of the variance of CSF levels of tau, a detrimental protein that accumulates in the brain of AD patients and is necessary for its diagnosis. These results further our understanding of underlying mechanisms contributing to AD risk while demonstrating the utility of generating and integrating multiple omics data types.

Cerebrospinal fluid metabolomics identifies 19 brain-related phenotype associations

“…The primary data for this study came from two different longitudinal cohort studies of AD with available CSF metabolomics and genotype data: the Wisconsin Alzheimer's Disease Research Center (WADRC) and Wisconsin Registry for Alzheimer's Prevention (WRAP) studies 12,13 . To improve the generalizability of this analysis, only data from cognitively healthy participants were used.…”

“…In the WRAP cohort, DNA was extracted from whole blood samples and genotyped using the Illumina Multi-Ethnic Genotyping Array at the University of Wisconsin Biotechnology Center 13 . Briefly, samples and variants with high missingness (> 5%) and samples with inconsistent genetic and self-reported sex were removed.…”

Metabolic Profiling of Cognitive Aging in Midlife

“…A total of 1,131 metabolites (879 knowns and 252 unknowns) was included in the final analysis in the VETSA. Similar procedures were used to pre-process and QC the metabolomics data in the WRAP (Darst et al, 2019), resulting in a total of 1,097 metabolites in our replication analysis.…”

Current Status of Metabolomic Biomarker Discovery: Impact of Study Design and Demographic Characteristics

“…The metabolome is directly linked to perturbations in cell metabolism caused by environmental stimuli as well as disease, thus metabolomics has been successfully used to measure phenotypic outcomes in vitro and in vivo [15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][40][41][42][43][44][45][46][47][48][49]. Examples include observations of oncometabolites, such as 2-hyroxybutyrate, sarcosine, choline, succinate, lactate, fumarate, and glucose, associated with tumor growth and metastasis that may also function as biomarkers for leukemia, renal carcinoma, breast, brain, or prostate cancer [26][27][28][29][30][31][32].…”