Genome-Wide Association and Interaction Study on Quantitative Traits of CSF Phosphorylated Tau in ADNI cohort

Abstract

Alzheimer's disease (AD) is a most common and particularly complex neurodegenerative disease commonly characterized by a progressive decline in cognition and memory impairments. CSF Aß42 and tau levels have emerged as useful biomarkers for disease and Quantitative Traits (QTs) for genetic studies of AD. In this study, CSF Phosphorylated Tau (P-tau) were used both in genome-wide association studies (GWAS) and genome-wide interaction study (GWIS) as QTs. The GWAS results shows that genes identified by Total Tau (T-tau), also have been replicated on QTs of P-tau, and a new GWAS loci of rs157582 have been identified by GWAS of CSF P-tau. The three genes of APOE, APOC1, and TOMM40 identified by p-Tau in this GWAS are previously confirmed AD susceptibility genes, while our GWAS have replicated them. Therefore, we may argue that QTs of P-tau is more sensitive and more significant than tau. This also indirectly proves the view of P-tau may be a more specific marker than T-tau for AD. Furthermore, the results demonstrate that the data, phenotypes, quality control methods and GWAS method we used to be rationality and validity, which making the subsequent GWIS based on this data more meaningful. In GWIS, 8 pairs of SNPs showed significant interaction effects on the CSF P-tau (corrected p-value <0.05). These 8 pairs of SNPs revealed by GWIS codes from 8 novel genes. Among them, NAT1, ST18, CTNNA2 and IMMP2L have been reported as AD risk genes. NCKAP5, LHFPL3, PRLHR are related to schizophrenia (SCZ). And the gene of AOAH go unreported be related to AD or SCZ. All the work illustrates the plausibility of T-tau as QTs on our dataset, while we also found more significant SNPs by GWIS, which can be used as an effective supplement when AD risk loci cannot be identified by GWAS. The 8 pairs of SNP interactions findings warrant replication and can be used for the analysis of other complex diseases.