COLD SPRING HARBOR, NY (GenomeWeb) – Researchers have begun to tease out variants that influence human longevity and viability.
In a proof-of-concept study presented at the Biology of Genomes meeting here, Columbia University's Hakhamanesh Mostafavi and his colleagues searched for and uncovered gene variants, such as one within APOE, that become less common with age in large cohorts, suggesting that these variants affect survival. They also linked having a higher polygenic risk for conditions like high cholesterol to lower survival.
Such variants, he added, could provide insight into the evolutionary fitness of humans. "We know very little about variants that underlie adaptation in humans," Mostafavi said.
As he noted in his talk, Mostafavi and his colleagues reasoned that if a variant has no effect on longevity, it should be present at the same frequency across a cohort of various ages. But if it does adversely influence longevity, it should be less common among older members of the cohort.
With this in mind, they developed a method to test whether the frequency of an allele varied across ages in large cohorts, after controlling for differences in ancestry.
In particular, they applied this method to a cohort of nearly 58,000 people of European ancestry from the Resource for Genetic Epidemiology Research on Aging (GERA). In this cohort, the strongest signal emanated from the APOE gene at the ε4 allele, which is known to increase the risk of late-onset Alzheimer's disease.
The frequency of the known deleterious variant dropped in the cohort around the age of 70, which Mostafavi noted correlated with the age of onset of Alzheimer's disease. This indicated that this variant affects longevity after the age of 70, Mostafavi said.
They also applied their approach to a cohort of nearly 96,000 individuals of European ancestry from the UK Biobank who had provided death information about their parents. Using their children as a proxy, Mostafavi and his colleagues uncovered a signal near CHRNA3 that was associated with paternal age at death.
Variants near CHRNA3, he noted, have been associated with smoking quantity. He added that in the 1950s, some 80 percent of men in Britain smoked, but that those with this allele smoked more.
But as many conditions are affected by variation in multiple genes, Mostafavi and his colleagues also created polygenic risk scores for individuals in their cohorts based on genome-wide association study results for about 40 genes. They then examined whether these risk scores were associated with survival.
Within their UK Biobank cohort, they found that variants linked to increased total cholesterol, LDL cholesterol, body-mass index, and coronary artery disease risk were all negatively associated with paternal survival.
Variants associated with later puberty timing, meanwhile, were linked to a higher chance of survival among fathers. This is consistent with findings from epidemiological studies, Mostafavi said.
He and his colleagues further replicated these findings in a subset of the UK Biobank that included individuals who were not of British ancestry.
Similarly, maternal survival was also influenced by puberty timing as well as the age at which she had her first child. Both delayed menarche and later age at first child were linked to greater longevity.
As having a first child at a later age was also associated with having a lower number of children, Mostafavi said this suggested that there is a trade-off between fertility and longevity. This further indicated to him that translating survival to fitness isn't straightforward.
Still, Mostafavi said that "we have a much more comprehensive picture of viability in humans."
An earlier version of this work is available as a preprint at BioRxiv, though Mostafavi noted that they intend to update that soon.