RNAseq Analysis of Brain Aging in Wild Specimens of Short-Lived Turquoise Killifish: Commonalities and Differences With Aging Under Laboratory Conditions

Abstract

A vast body of studies is available that describes age-dependent gene expression in relationship to aging in a number of different model species. These data were obtained from animals kept in conditions with reduced environmental challenges, abundant food and deprivation of natural sensory stimulation. Here we compared wild- and captive-aging in the short-lived turquoise killifish (Nothobranchius furzeri). These fish inhabit temporary ponds in the African savannah. When the ponds are flooded, eggs hatch synchronously, enabling a precise timing of their individual and population age. We collected brains of wild fish of different ages and quantified the global age-dependent regulation of transcripts using RNA-seq. A major difference between captive and wild population is that wild population had unlimited access to food, and hence grew to larger sizes and reached asymptotic size more rapidly, enabling the analysis of age-dependent gene expression without the confounding effect of adult brain growth. We found that the majority of differentially-expressed genes show the same direction of regulation in wild and captive population. However, a number of genes were regulated in opposite direction. Genes down-regulated in the wild and up-regulated in captivity were enriched for terms related to neuronal communication. Genes up-regulated in the wild and down-regulated in captive conditions were enriched in terms related to DNA replication. Finally, the rate of age-dependent gene regulation was higher in the wild animals suggesting a phenomenon of accelerated aging.

A deep neural network provides an ultraprecise multi-tissue transcriptomic clock for the short-lived fishNothobranchius furzeriand identifies predicitive genes translatable to human aging

“…In this context, it is worth stressing that in the application of the N-TICK to the wild brain samples, we obtained overestimation of the predicted age but also a slope that, with respect to the true one, is inferior to 1. It should be noted that wild fish reach asymptotic size much earlier than captive ones [56] and in the age range explored, the former ones do not show growth [55]. Although three points in lifespan are not sufficient to reliably interpret this phenomenon, we can speculate that the N-TICK predictions correctly capture the heterochronic and non-linear nature of growth and aging in the wild vs. captivity.…”

“…The spread of the errors on the wild samples is larger than that of the captive ones. This reduction in precision is expected since wild animals show basal differential expression of almost one third of expressed genes as compared to captive-raised fish and some genes are regulated in opposite directions by age in the two conditions [55]. Furthermore, it should be considered that a constant source of error in the age of the wild fish arises from the uncertainty in their age estimation that is obtained from daily otolith increments [56].…”