Evolutionary Adaptation in the Human Lineage

Positive natural selection, or the tendency of beneficial traits to increase in prevalence (frequency) in a population, is the driving force behind adaptive evolution. For a trait to undergo positive selection, it must have two characteristics. First, the trait must be beneficial; in other words, it must increase the organism's probability of surviving and reproducing. Second, the trait must be heritable so that it can be passed to an organism's offspring. Beneficial traits are extremely varied and may include anything from protective coloration, to the ability to utilize a new food source, to a change in size or shape that might be useful in a particular environment. If a trait results in more offspring who share the trait, then that trait is more likely to become common in the population than a trait that arises randomly. At the molecular level, selection occurs when a particular DNA variant becomes more common because of its effect on the organisms that carry it.

Charles Darwin and Alfred Wallace (1858) famously proposed that positive selection could explain the many marvelous adaptations that suit organisms to their environments and lifestyles, and this simple process remains the central explanation for all evolutionary adaptation yet today. Positive selection is by no means the only component of evolution, however. In humans, at least, the great majority of mutations are thought to be selectively neutral, conferring neither benefit nor cost on their bearers (Hellmann et al., 2003). The frequency of some of these neutral genetic variants (alleles) increases simply by chance, and the resulting "genetic drift" is thought to be the most common process in human evolution (Kimura, 1968). Moreover, when selection does occur, it is most often in the form of negative, or purifying, selection, which removes new deleterious mutations as they arise, rather than promoting the spread of new traits (Kreitman, 2000).

As advantageous alleles that are under positive selection increase in prevalence, these alleles leave distinctive signatures, or patterns of genetic variation, in the DNA sequence. Consider a population of individuals for which, before selection, there are hundreds of thousands of varied chromosomes in the population, all with different combinations of genetic variants. Now, say that an advantageous allele arises as a mutation on one copy of a chromosome. Through succeeding generations, the descendants of this copy, including the selected allele and nearby "hitchhiking" alleles, become more and more common through a process called a "selective sweep" (Figure 1). Note that the entire chromosome is not passed down as a unit, however; rather, because of recombination, segments of the chromosome are inherited. Thus, while the selected allele and hitchhiking alleles increase in prevalence in a selective sweep, at the same time, the segment that includes the selected allele is slowly reduced in size by recombination. Investigators are interested in the types of signals that can be detected in a selective sweep, as well as their properties and technical challenges (Nielsen, 2005; Sabeti et al., 2006).

Within the last decade, our ability to probe our own species for evidence of selection has increased dramatically due to the flood of genetic data that have been generated. Starting with the complete sequence of the human genome (Lander et al., 2001), which provides a framework and standard reference for all human genetics, key data sets include the completed or near-completed genomes of several related species (e.g., chimpanzee, macaque, gorilla, and orangutan), a public database of known genetic variants in humans, and surveys of genetic variation in hundreds of individuals in multiple populations (Chimpanzee Sequencing and Analysis Consortium, 2005; Gibbs et al., 2007; Sherry et al., 2001; International HapMap Consortium, 2007). With these new data, it is now possible to scan the entire human genome in search of signals of natural selection.

Although the study of natural selection in humans is still in an early stage, the new data, building on decades of earlier work, are beginning to reveal some of the landscape of selection in our species. In fact, researchers have identified many genetic loci at which selection has likely occurred, and some of the selective pressures involved have been elucidated. Three significant forces that have been identified thus far include changes in diet, changes in climate, and infectious disease.

While these instances of selection illustrate the power this line of research has to answer important biological and historical questions, in most cases, little or nothing of the underlying story is understood. For the great majority of selective sweeps, the pressure that drove selection, the trait selected for, and even the specific gene involved are unknown. Understanding these will require case-by-case study, identifying the possible causal mutations within each region based on strength of signal and function (e.g., mutations that alter amino acids or gene regulatory regions), and then finding the biological effects of each.

Such detailed investigations are underway, and they are intriguing. For example, a strong signal of selection in Asia localizes to amino acid substitution in the gene EDAR Sabeti et al., 2007). Mutations in EDAR cause defects in the development of hair, teeth, and exocrine glands in both mice and humans. Meanwhile, there is also evidence for selection at other genes in the same pathway in humans, as well as in stickleback fish (Colosimo et al., 2005), where the pathway regulates scale development. The phenotypic variation for this mutation is only just being elucidated, but it has already been linked to thicker head hair in Asia and has been shown to affect gene activity in the molecular pathway (Bryk et al., 2008; Fujimoto et al., 2008), although what trait was actually under selection is not yet clear. In another case, Scott Williamson and his colleagues found the strongest signal of selection in Europe and Asia at the gene DTNA, a component of the dystrophin complex (Williamson et al., 2007). While the target polymorphism and genetic variation have yet to be elucidated, the dystrophin complex is known to be important in the architecture of muscle tissue, as well as in the pathogenesis of many infectious agents, including arenaviruses and mycobacterium leprae. Another candidate gene for selection, LARGE, is also important for dystrophin function, and it has been shown to be critical for entry of various arenaviruses, including Lassa virus (Sabeti et al., 2007).

Understanding the biology behind these cases, and the many others like them, will not be easy, and it will require contributions from diverse fields, including genetics, molecular biology, developmental biology, and the study of model organisms (Figure 2). Nevertheless, the potential rewards are high. Through the study of natural selection in humans, researchers hope to learn more about how our species has changed over time, about the challenges the species has faced and how it has overcome them, and about past and present causes of disease.