Date Published: March 25, 2018
Publisher: John Wiley and Sons Inc.
Author(s): Serena Dato, Mette Soerensen, Francesco De Rango, Giuseppina Rose, Kaare Christensen, Lene Christiansen, Giuseppe Passarino.
In human longevity studies, single nucleotide polymorphism (SNP) analysis identified a large number of genetic variants with small effects, yet not easily replicable in different populations. New insights may come from the combined analysis of different SNPs, especially when grouped by metabolic pathway. We applied this approach to study the joint effect on longevity of SNPs belonging to three candidate pathways, the insulin/insulin‐like growth factor signalling (IIS), DNA repair and pro/antioxidant. We analysed data from 1,058 tagging SNPs in 140 genes, collected in 1825 subjects (1,089 unrelated nonagenarians from the Danish 1905 Birth Cohort Study and 736 Danish controls aged 46–55 years) for evaluating synergic interactions by SNPsyn. Synergies were further tested by the multidimensional reduction (MDR) approach, both intra‐ and interpathways. The best combinations (FDR<0.0001) resulted those encompassing IGF1R‐rs12437963 and PTPN1‐rs6067484, TP53‐rs2078486 and ERCC2‐rs50871, TXNRD1‐rs17202060 and TP53‐rs2078486, the latter two supporting a central role of TP53 in mediating the concerted activation of the DNA repair and pro‐antioxidant pathways in human longevity. Results were consistently replicated with both approaches, as well as a significant effect on longevity was found for the GHSR gene, which also interacts with partners belonging to both IIS and DNA repair pathways (PAPPA,PTPN1,PARK7, MRE11A). The combination GHSR‐MREA11, positively associated with longevity by MDR, was further found influencing longitudinal survival in nonagenarian females (p = .026). Results here presented highlight the validity of SNP‐SNP interactions analyses for investigating the genetics of human longevity, confirming previously identified markers but also pointing to novel genes as central nodes of additional networks involved in human longevity.
Association studies can identify the association of individual gene variants to a given phenotype. Nevertheless, such analysis is unable to explain the biological complexity of several diseases and complex phenotypes such as human aging and longevity.
In this work, we used a combination of different approaches for understanding the relationships between SNP variants in the predisposition to become long‐lived. First, we computed case–control based SNP‐SNP interaction without pathway constraints using the whole data set, and we plotted interaction networks from significant SNP combinations. Then, we performed pathway‐based case–control analysis, looking for epistatic interactions inside original assigned pathways and among different pathways by applying a MDR approach. Finally, we analysed those pairs of SNPs significantly enriched in cases respect to controls for their influence on survival in the oldest old.
Many reasons for the missing replicability in genetic association studies of survival to ages ≥90 years have been suggested, as it was recently reviewed by Sebastiani and coworkers, who indicated demographical reasons, that is the definition of the age windows for survival for birth cohort, as the main bias source when comparing different association studies on human longevity (Sebastiani, Nussbaum, Andersen, Black & Perls, 2016). Another possible explanation of the missing replicability and missing heritability may be related to the complexity of longevity (Dato et al., 2017), which harbours many heterogeneity sources, including the effect of rare variants, not captured by standard genomewide genotyping, and interactions between different loci, an often‐cited reason for the lack of success in genetic studies of complex diseases (Moore, 2003). With the aim of exploring this poorly investigated genetic effect, we re‐analysed a genetic data set previously described and used for single‐SNP and gene set analyses (Debrabant et al., 2014; Soerensen et al., 2012), for SNP‐SNP interactions. The findings obtained indeed indicate that an epistatic analysis approach is very much applicable for candidate gene/pathway data and hence might contribute to the knowledge concerning the genetics of human longevity.
Serena Dato drafted the manuscript and generated the conception of the study, and participated in the data acquisition and interpretation of data. Mette Soerensen generated the study design and participated in the data acquisition, data management and interpretation of data. Francesco De Rango carried out the statistical analyses and participated in the interpretation of data. Giuseppina Rose participated in the interpretation of data and revision of the manuscript. Lene Christiansen participated in generating the study design, acquisition of data and interpretation of data. Giuseppe Passarino participated in the interpretation of data and revision of the manuscript. Kaare Christensen participated in the acquisition of data and interpretation of data. All authors have revised the manuscript and given their final approval.