A genome-wide screening and SNPs-to-genes approach to identify novel genetic risk factors associated with frontotemporal dementia

Abstract

Frontotemporal dementia (FTD) is the second most prevalent form of early onset dementia after Alzheimer's disease (AD). We performed a case-control association study in an Italian FTD cohort (n = 530) followed by the novel single nucleotide polymorphisms (SNPs)-to-genes approach and functional annotation analysis. We identified 2 novel potential loci for FTD. Suggestive SNPs reached p-values ∼10(-7) and odds ratio > 2.5 (2p16.3) and 1.5 (17q25.3). Suggestive alleles at 17q25.3 identified a disease-associated haplotype causing decreased expression of -cis genes such as RFNG and AATK involved in neuronal genesis and differentiation and axon outgrowth, respectively. We replicated this locus through the SNPs-to-genes approach. Our functional annotation analysis indicated significant enrichment for functions of the brain (neuronal genesis, differentiation, and maturation), the synapse (neurotransmission and synapse plasticity), and elements of the immune system, the latter supporting our recent international FTD-genome-wide association study. This is the largest genome-wide study in Italian FTD to date. Although our results are not conclusive, we set the basis for future replication studies and identification of susceptible molecular mechanisms involved in FTD pathogenesis.

Keywords: Association study; Case-control; Frontotemporal dementia; Functional annotation; Genetic risk factors; Population.

Fig. 1

Manhattan plot of the association analysis. The genome-wide significance level threshold was: p-value = 2.18 × 10⁻⁸. Single nucleotide polymorphisms on chromosomes 2 and 17 reached strongly suggestive p values. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Suggestive loci and regional plot analysis. (A) Highlight of the 2p16.3 locus holding 7 suggestive single nucleotide polymorphisms (SNPs) within and downstream of LOC730100. (B) Regional plot for the 2p16.3 locus. The 7 suggestive SNPs are in high linkage disequilibrium (LD) with each other but no other SNPs in the surrounding region. (C) Highlight of the 17q25.3 locus holding 7 suggestive SNPs across CEP131, ENTHD, and C17orf89. (D) Regional plot for the 17q25.3 locus. The 7 suggestive SNPs are in high LD with 5 further SNPs at this locus identifying a 45.2 Kb long haplotype block. In regional plot, each circle represents a SNP, y-axis is the −log10 association p value for frontotemporal dementia association and x-axis represents the physical position on the chromosome (build 36, hg18). The circles are filled with colors according to the linkage disequilibrium (LD; r2) between the given SNP and the lead SNP (purple square). (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

SNPs-to-genes analysis. The Venn diagram shows the number of genes that were significant after FDR correction (p-value < 0.05) and their level of overlap across the 3 methods (GATES, sPCA, and SKAT). Abbreviations: FDR, false discovery rate; GATES, extended Simes' procedure; SKAT, sequential kernel machine association test; sPCA, supervised principal components analysis.