Step by step: towards a better understanding of the genetic architecture of Alzheimer's disease

Abstract

Alzheimer's disease (AD) is considered to have a large genetic component. Our knowledge of this component has progressed over the last 10 years, thanks notably to the advent of genome-wide association studies and the establishment of large consortia that make it possible to analyze hundreds of thousands of cases and controls. The characterization of dozens of chromosomal regions associated with the risk of developing AD and (in some loci) the causal genes responsible for the observed disease signal has confirmed the involvement of major pathophysiological pathways (such as amyloid precursor protein metabolism) and opened up new perspectives (such as the central role of microglia and inflammation). Furthermore, large-scale sequencing projects are starting to reveal the major impact of rare variants - even in genes like APOE - on the AD risk. This increasingly comprehensive knowledge is now being disseminated through translational research; in particular, the development of genetic risk/polygenic risk scores is helping to identify the subpopulations more at risk or less at risk of developing AD. Although it is difficult to assess the efforts still needed to comprehensively characterize the genetic component of AD, several lines of research can be improved or initiated. Ultimately, genetics (in combination with other biomarkers) might help to redefine the boundaries and relationships between various neurodegenerative diseases.

Fig. 1. Ideogram of the 90 loci with genome-wide significance extracted from the GWAS catalog and classified as tier 1, tier 2 or tier 3.

For each locus, the figure shows the P-value categories for the association with AD in the six main GWASs published since 2019 (IGAP2, PGC1, IGAP2 + UKB, GR@ACE, PGC2, and EADB, [–23, 26, 27]): P ≤ 5 × 10⁻⁸, P ≤ 1 × 10⁻⁴, P > 1 × 10⁻⁴, or NA (not available). See the additional note for details of the methods.

Fig. 2. Number of loci identified in the different GWASs.

Number of loci identified as a function of the sample size (left) or the effective sample size (right) (cases and controls) in the main GWASs published since 2009 (EADI, GERAD, CHARGE, GERAD+, ADGC, IGAP1, IGAP1 + UKB (2017), IGAP1 + UKB (2018), PGC1, IGAP2, IGAP2 + UKB, GR@ACE, PGC2, and EADB [–27]. The colors indicate the presence or absence of proxy cases in the GWAS. The effective sample size was computed per study included in the meta-analyses, and then summed across studies [124]. The effective sample size for the proxy UKB study was computed by dividing the raw number of proxy cases and proxy controls by four [24, 117].