Duplications of Human Longevity-Associated Genes Across Placental Mammals

Abstract

Natural selection has shaped a wide range of lifespans across mammals, with a few long-lived species showing negligible signs of ageing. Approaches used to elucidate the genetic mechanisms underlying mammalian longevity usually involve phylogenetic selection tests on candidate genes, detections of convergent amino acid changes in long-lived lineages, analyses of differential gene expression between age cohorts or species, and measurements of age-related epigenetic changes. However, the link between gene duplication and evolution of mammalian longevity has not been widely investigated. Here, we explored the association between gene duplication and mammalian lifespan by analyzing 287 human longevity-associated genes across 37 placental mammals. We estimated that the expansion rate of these genes is eight times higher than their contraction rate across these 37 species. Using phylogenetic approaches, we identified 43 genes whose duplication levels are significantly correlated with longevity quotients (False Discovery Rate (FDR) < 0.05). In particular, the strong correlation observed for four genes (CREBBP, PIK3R1, HELLS, FOXM1) appears to be driven mainly by their high duplication levels in two ageing extremists, the naked mole rat (Heterocephalus glaber) and the greater mouse-eared bat (Myotis myotis). Further sequence and expression analyses suggest that the gene PIK3R1 may have undergone a convergent duplication event, whereby the similar region of its coding sequence was independently duplicated multiple times in both of these long-lived species. Collectively, this study identified several candidate genes whose duplications may underlie the extreme longevity in mammals, and highlighted the potential role of gene duplication in the evolution of mammalian long lifespans.

Keywords: Heterocephalus glaber; Myotis myotis; gene duplication; mammalian longevity; truncated pseudogenes.

Fig. 1.

Placental mammals and human longevity-associated genes investigated in this study. (A) The distribution of LQs across 37 placental mammals. The species are categorized into four groups: 0 < LQ ≤ 1, 1 < LQ ≤ 2, 2 < LQ ≤ 4, and LQ > 4. (B) GO enrichment analysis of 293 human longevity-associated genes. The circles (GO terms) with the same color indicate that they belong to the same parental term, and the connections between circles indicate that the GO terms share common genes. The size of a circle indicates the number of genes in that GO term.

Fig. 2.

Expansions and contractions of 287 human longevity-associated genes across 37 placental mammals. The time-calibrated phylogenetic tree indicating the relationship amongst 37 mammals was obtained from TimeTree (v5). The values on the phylogenetic tree represent the number of genes expanded and contracted on each node. The bars on the tips represent the number of genes expanded and contracted in each species, respectively. “Sig Exp” and “Sig Con” indicate the number of genes significantly expanded and contracted in each species (P < 0.05), respectively.

Fig. 3.

Phylogenetic correlation between duplication levels of human longevity-associated genes and LQ. (A) Correlation coefficients of 287 genes after the phylogeny correction. The color code indicates the significance level (FDR). Forty-three genes shown on the plot exhibit significant correlation between their duplication levels and mammalian LQs. (B) Boxplot showing the duplication levels of 43 significantly correlated genes across 37 species. For each gene, the species with high duplication levels (>20) were not included in the plot.

Fig. 4.

Scatterplots showing the association between gene duplication level and LQ for the genes RGN and HESX1. High duplication levels of RGN and HESX1 are observed in the genomes of a few relatively short-lived species.

Fig. 5.

Scatterplots showing the association between gene duplication level and LQ for the genes CREBBP, PIK3R1, HELLS, and FOXM1. Their strong positive correlation is mainly driven by their high duplication levels in two extremely long-lived species, H. glaber and M. myotis.

Fig. 6.

Sequence and expression analyses of four gene candidates in H. glaber and M. myotis. (A) Distribution of duplicated copies of four gene candidates (CREBBP, PIK3R1, HELLS, FOXM1) in the H. glaber and M. myotis genomes. The genome scaffolds shorter than 5 Mb were not included. (B) Alignments of PIK3R1 pseudogenes from H. glaber (5) and M. myotis (13). Only the ∼600 bp common regions are shown. (C) Expression of all the pseudogenes of these four candidates in H. glaber and M. myotis. The values on the bar plot indicate the number of expressed and nonexpressed pseudogenes of each gene in the brain, liver, or kidney samples of H. glaber and M. myotis.