Premium
Functional interpretation of genetic risk loci for dementia using a protein quantitative trait loci (pQTLs) approach in cerebrospinal fluid
Author(s) -
Reus Lianne M.,
Jansen Iris E.,
Tijms Betty M.,
Visser Pieter Jelle,
Lemstra Afina W.,
Pijnenburg Yolande A.L.,
van Der Flier Wiesje,
Teunissen Charlotte E.,
Campo Marta Del
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.040774
Subject(s) - dementia , genome wide association study , frontotemporal dementia , genetic association , cohort , dementia with lewy bodies , bonferroni correction , quantitative trait locus , single nucleotide polymorphism , genetics , biology , medicine , oncology , disease , bioinformatics , genotype , gene , statistics , mathematics
Background Genome‐wide association studies have successfully identified genetic risk loci for dementia,but through which biological mechanisms these factors contribute to dementia remains elusive. Cerebrospinal fluid (CSF) protein levels reflect ongoing biological processes that take place in the brain. Studying which CSF proteins associate with dementia risk loci could reveal intermediate molecular pathways connecting genetic variance to the development of disease.We aimed to identify protein quantitative trait loci (pQTLs) in ante‐mortem CSF in different types of dementia, including Alzheimer’s disease (AD), dementia with Lewy bodies (DLB) and frontotemporal dementia (FTD). Method We included 513 subjects (144 non‐demented controls, 211 AD, 56 DLB, 102 FTD, age 64 ± 9 years, 36%female) from the Amsterdam Dementia Cohort and European Medical Information Framework for AD (EMIF‐AD) PreclinAD cohort who had genetic and CSF proteomic data available. Genetic data was generated using the Illumina Genome Screening Array(GSA). CSF proteins (n = 665) were measured using an antibody‐based proteomic array (Olink). We included 43 genetic risk loci for AD, 7 for DLB and 9 for FTD. Association signals between SNPs and CSF protein levels were determined using linear regression, adjusted for principal components, age and sex, stratified for diagnosis (i.e., non‐demented controls and diagnostic group combined). Results were corrected for multiple comparisons with the Bonferroni method. Result Variants in three AD ( CR1 , ZCWPW1 , HESX1 ) and one DLB ( GBA ) risk loci were associated with levels of seven CSF proteins (Figure 1). The CR1 risk locus (rs4844610) was associated with higher CR2 CSF levels ( P = 5.47 × 10 −7 ), ZCWPW1 (rs2906657) with lower PILRB ( P = 2.28 × 10 −32 ) and HESX1 (rs186108507) with higher RETN ( P = 6.00 × 10 −8 ). Within DLB, GBA (rs12742181) was associated with higher CSF protein levels of ANG, CD798, CXCL13 and TNFRS13B ( P = 7.79 × 10 −7 , P = 8.83 × 10 −8 , P = 4.87 × 10 −6 , P = 4.47 × 10 −8 respectively). No significant associations were identified between FTD risk loci and CSF protein levels. Conclusion We identified four pQTLs in ante‐mortem CSF, providing a neurobiological link by which genetic risk loci potentially contribute to the development of AD and DLB. Dissecting the contribution of risk loci to neurobiological processes aids in understanding the complex disease mechanisms underlying different types of dementia. Our efforts are currently being extended to functional validation and genetic replication in independent cohorts.