Distinct genomic signatures of adaptation in pre- and postnatal environments during human evolution

Abstract

The human genome evolution project seeks to reveal the genetic underpinnings of key phenotypic features that are distinctive of humans, such as a greatly enlarged cerebral cortex, slow development, and long life spans. This project has focused predominantly on genotypic changes during the 6-million-year descent from the last common ancestor (LCA) of humans and chimpanzees. Here, we argue that adaptive genotypic changes during earlier periods of evolutionary history also helped shape the distinctive human phenotype. Using comparative genome sequence data from 10 vertebrate species, we find a signature of human ancestry-specific adaptive evolution in 1,240 genes during their descent from the LCA with rodents. We also find that the signature of adaptive evolution is significantly different for highly expressed genes in human fetal and adult-stage tissues. Functional annotation clustering shows that on the ape stem lineage, an especially evident adaptively evolved biological pathway contains genes that function in mitochondria, are crucially involved in aerobic energy production, and are highly expressed in two energy-demanding tissues, heart and brain. Also, on this ape stem lineage, there was adaptive evolution among genes associated with human autoimmune and aging-related diseases. During more recent human descent, the adaptively evolving, highly expressed genes in fetal brain are involved in mediating neuronal connectivity. Comparing adaptively evolving genes from pre-and postnatal-stage tissues suggests that different selective pressures act on the development vs. the maintenance of the human phenotype.fetal ͉ human disease ͉ mitochondria ͉ placenta ͉ thyroid

Genomic signatures of diet-related shifts during human origins

“…One test to see whether these diet-related signals are biologically meaningful is to perform branch-specific enrichments, where only regions under positive selection in each species are used in the enrichments. These analyses show that the metabolic categories for each species can be quite different, and that even for the same category the specific genes showing evidence of selection are usually distinct subsets on different species' branches [61,62,71]. A weaker signal for glucose metabolism-related categories is seen on the chimpanzee branch, but the specific genes involved differ on the human and chimpanzee branches.…”

“…(b) Evidence for changes in energy transport In brain tissues, there is a consistent pattern of changes in expression of genes critical to aerobic energy metabolism [67,69,71,91]. This includes categories such as oxidative phosphorylation, electron transport and other nuclearencoded genes that function in the mitochondria.…”

A red queen model of personality

“…For example, the protocadherin gene family provides a molecular code which enables highly detailed neuronal wiring to be encoded genetically (Wu & Maniatis, 1999). Such a capacity for fine-scaled neural circuitry to be specified by the genome means it is likely that genes of this type have been favored by natural selection in humans, given that variations across other taxonomic groups are also thought to underpin species-specific adaptations (Noonan et al, 2004;Uddin et al, 2007;Wu, 2005). In line with Red Queen predictions, just as regions of the human genome that are mutation "hotspots" (where most mutations occur) are disproportionately genes involved in immune function, the protocadherin gene families in particular have especially high mutation rates (Chuang & Li, 2004).…”

Statistical methods for detecting natural selection from genomic data

“…In addition, statistical methods have been used even when they appeared to be inapplicable, e.g., when d S and d N were considered to be saturated (Studer et al, 2008). Biological significance of the results was speculated even when they were obtained without conducting statistical tests (Uddin et al, 2008;Goodman et al, 2009;Goodman and Sterner, 2010) or the proportion of positive results was smaller than the significance level adopted. Positively selected genes identified were sometimes randomly associated with functions and gene duplications (Studer et al, 2008), and little overlap was observed among positively selected regions obtained from the analyses of similar data using different methods.…”