Alzheimer's Disease

Genomics and Beyond

Fuhai Song*,†; Guangchun Han*; Zhouxian Bai*,†; Xing Peng*,†; Jiajia Wang*,†; Hongxing Lei*,‡,1    * CAS Key Laboratory of Genome Sciences and Information, Beijing Institute of Genomics, Chinese Academy of Sciences, Beijing, PR China
† University of Chinese Academy of Sciences, Beijing, PR China
‡ Center of Alzheimer's Disease, Beijing Institute for Brain Disorders, Beijing, PR China
1 Corresponding author: email address: leihx@big.ac.cn

1 Introduction

Alzheimer's disease (AD) affects a large population in the senior community, likely 10 million in China alone (Han et al., 2014; Lei, 2010). The pathological hallmarks of AD include extracellular deposit of Aβ amyloid plaques derived from APP and intraneuronal neurofibrillary tangles (NFTs) from hyperphosphorylation of tau. Much of the efforts in AD research have been devoted to molecular pathways centered at Aβ or tau. The vast majority of novel treatment strategies are also targeting either Aβ or tau. Nevertheless, promising results from animal models have not translated well in human clinical trials (Callaway, 2012). Thus, revolutionary ideas outside of the hallmarks are desperately needed.

Technology developments in genomics have provided a variety of tools to investigate AD at the whole system level (Fig. 1). Earlier genetic linkage or association studies in both early-onset AD (EOAD) and late-onset AD (LOAD) have gradually evolved into genome-wide association studies (GWASs) and more recently developed into whole-exome sequencing (WES) and whole-genome sequencing (WGS; Bettens, Sleegers, & Van Broeckhoven, 2013; Guerreiro, Bras, & Hardy, 2013). Transcriptome studies are still in the process of transiting from microarray to RNA-seq (Sekar et al., 2015). Several types of epigenetic and epigenomic studies have been conducted on AD, including DNA methylation, histone modification, and microRNA (Lunnon & Mill, 2013). As a valuable supplement to the traditional studies on human subjects and animal models, induced pluripotency stem cell (iPSC) technology has been increasingly used to bridge the gap in translational medicine (Israel et al., 2012). In this review, we will survey representative investigations applying genomics technologies to the study of AD and relevant processes such as aging and neurological disorders.

2 GWASs on the Primary Phenotype of AD

The quality of GWASs depends heavily on the sample size. In the most recent large-scale GWAS of AD (Lambert et al., 2013), over 74K samples were included in the meta-analysis (all of European ancestry), including ~ 54K samples in stage 1 (previously published datasets) and ~ 20K samples in stage 2 (new datasets). Over 7M imputed single-nucleotide polymorphisms (SNPs) were used in the meta-analysis of stage 1 datasets. In stage 2, only SNPs showing moderate significance in stage 1 (P < 1 × 10− 3) were genotyped (~ 11K SNPs). In the meta-analysis of stage 1 datasets, the significance of APOE and nine other previously identified genes was confirmed (Table 1), including CR1, BIN1, CLU, PICALM, ABCA7, CD2AP, EPHA1, MS4A4E, and CD33. Additionally, five new loci reached genome-wide significance in the discovery datasets, namely, SORL1, PTK2B, DSG2, HLA-DRB5, and SLC24A4. When combining the discovery and replication datasets, seven new loci reached genome-wide significance, including INPP5D, MEF2C, NME8, FERMT2, ZCWPW1, CELF1, and CASS4. Among these 22 loci, CD33 and DSG2 did not reach genome-wide significance in the combined datasets. These genes are involved in immune response, lipid metabolism, synaptic function, and Aβ/tau-related pathways.

Besides testing the significance at the SNP level, gene-level test may reveal new genes affecting AD susceptibility. In a follow-up work of the above-described study (Escott-Price et al., 2014), gene-level analysis identified two genes with no previous reports, TP53INP1 and IGHV1-67. The three most significant SNPs in TP53INP1 represented at least two partially independent signals, while the two most significant SNPs in IGHV1-67 represented the same signal. As for the gene function, TP53INP1 is involved in apoptosis and IGHV1-67 is involved in immune response.

We have adopted a different strategy in our reanalysis of public GWAS datasets. We...