Identifying Protein Quantitative Trait Loci in the Cerebrospinal Fluid Proteome

Abstract

AbstractBackgroundGenome‐wide association studies (GWAS) on dementia have successfully identified variants conferring risk to disease. Still, how such variants impact exact biological mechanisms underlying disease remains largely unclear. Protein quantitative trait loci (pQTLs) measured in cerebrospinal fluid (CSF) may provide insight into these neurobiological mechanisms, since CSF protein levels can reflect ongoing processes in the brain and have been related to genetic variants. We used a CSF pQTL approach in individuals along the Alzheimer’s disease (AD) spectrum to reveal intermediate molecular pathways by which genetic variance influences neurobiological processes. Improving the functional interpretation of genetic variants in a matrix relevant to neurological disease helps deciphering biological mechanisms underlying neurological traits.MethodWe selected 243 subjects (106 controls/67 mild cognitive impairment/70 AD‐dementia, age 66.8±8.1 years, 53.9%female) from the EMIF‐AD MBD study, who had genetic data and CSF data (n = 1,351 proteins, tandem mass tag (TMT) mass spectrometry) available. Association signals between genetic variants and CSF proteins were tested using linear regression, adjusted for principal components (PC1‐3), age and sex using the genome‐wide significance threshold (P<5.0e−8). CSF pQTLs were identified as novel if the association has not been previously reported in Yang et al., (2021), Sasayama et al., (2017) or Kauwe et al., (2014).ResultsTwenty‐four out of 1,351 CSF proteins (1.78%) showed genome‐wide significant associations with in total 147 SNPs (Table 1; Figure 1). We identified 25 CSF pQTLs, representing three cis‐acting (IGH, LTF and ADAMTS8) and 22 trans‐acting pQTLs. Results included mostly novel CSF pQTL associations, exception for the cis association of LTF with LTF protein levels. Two novel cis‐acting pQTLs were previously reported in a study on human lymphocytes (IGH on IGHV2‐70D) (Theusch et al., 2020) and prefrontal cortex brain data (ADAMTS8 on ADAMTS8) (Ng et al., 2017).ConclusionsThese results contribute to the functional interpretation of genetic variance on neurobiological processes, by pinpointing inflammation and tissue remodeling as potential contributing mechanisms in AD.