Identification of gene pathways implicated in Alzheimer's disease using longitudinal imaging phenotypes with sparse regression

Abstract

We present a new method for the detection of gene pathways associated with a multivariate quantitative trait, and use it to identify causal pathways associated with an imaging endophenotype characteristic of longitudinal structural change in the brains of patients with Alzheimer's disease (AD). Our method, known as pathways sparse reduced-rank regression (PsRRR), uses group lasso penalised regression to jointly model the effects of genome-wide single nucleotide polymorphisms (SNPs), grouped into functional pathways using prior knowledge of gene–gene interactions. Pathways are ranked in order of importance using a resampling strategy that exploits finite sample variability. Our application study uses whole genome scans and MR images from 99 probable AD patients and 164 healthy elderly controls in the Alzheimer's Disease Neuroimaging Initiative (ADNI) database. 66,182 SNPs are mapped to 185 gene pathways from the KEGG pathway database. Voxel-wise imaging signatures characteristic of AD are obtained by analysing 3D patterns of structural change at 6, 12 and 24 months relative to baseline. High-ranking, AD endophenotype-associated pathways in our study include those describing insulin signalling, vascular smooth muscle contraction and focal adhesion. All of these have been previously implicated in AD biology. In a secondary analysis, we investigate SNPs and genes that may be driving pathway selection. High ranking genes include a number previously linked in gene expression studies to β-amyloid plaque formation in the AD brain (PIK3R3, PIK3CG, PRKCA and PRKCB), and to AD related changes in hippocampal gene expression (ADCY2, ACTN1, ACACA, and GNAI1). Other high ranking previously validated AD endophenotype-related genes include CR1, TOMM40 and APOE.

APOE-TOMM40 in the Pharmacogenomics of Dementia

“…The 523 VL poly T resulted in significantly higher expression than the S poly T. These results suggest that the 523 locus may contribute to LOAD susceptibility by modulating the expression of TOMM40 and/or APOE transcription [55]. Recent studies also suggest that the TOMM40 gene rs10524523 ("523") variable length poly T repeat polymorphism is associated to a certain extent with similar AD phenotypes as those reported for APOE, such as brain white matter changes [56,57] or different biomarkers [58][59][60][61]. In addition, the TOMM40 rs2075650 G allele may be a risk factor for the development of depression [62] and sporadic inclusion body myositis [63].…”

The Pathogenic Component of the APOE-TOMM40 Region in Alzheimer’s disease: Its Implications in Metabolomics and Pharmacogenomics

“…The 523 VL poly T resulted in significantly higher expression than the S poly T. These results suggest that the 523 locus may contribute to LOAD susceptibility by modulating the expression of TOMM40 and/or APOE transcription [40]. Recent studies also suggest that the TOMM40 gene rs10524523 ("523") variable length poly T repeat polymorphism is associated to a certain extent with similar AD phenotypes as those reported for APOE, such as brain white matter changes [41,42] or different biomarkers [43][44][45][46]. In addition, the TOMM40 rs2075650 G allele may be a risk factor for the development of depression [47] and sporadic inclusion body myositis [48].…”

APOE and MS4A6A interact with GnRH signaling in Alzheimer's disease: Enrichment of epistatic effects

“…A genomic pathway analysis of an imaging endophenotype for LOAD showed a significant association with vascular smooth muscle contraction [11]. Brain-blood barrier damage can impair amyloid beta clearance from the brain, whereas cerebral blood flow reduction has been associated with the APOE-E4 allele in undemented individuals [18,19].…”

“…For combined uncorrected P values, we found five significant pathways interacting with rs429358 (APOE), one with rs744373 (BIN1), four with rs3851179 (CR1), three with rs3851179 (PICALM), and 11 with rs610932 (MS4A6A). Many of the pathways have been reported with enrichment analysis of individual SNP effects [5,11], enlarging our previous list of enriched pathways for GWAS. We found three Q8 interacting pathways that consistently replicated in all four studies and were significant for the combined corrected P values; most notably, GnRH signaling interacted with both rs429358 and rs610932.…”