Integrating evolutionary and functional approaches to infer adaptation at specific loci

Abstract

Inferences about adaptation at specific loci are often exclusively based on the static analysis of DNA sequence variation. Ideally,population-genetic evidence for positive selection serves as a stepping-off point for experimental studies to elucidate the functional significance of the putatively adaptive variation. We argue that inferences about adaptation at specific loci are best achieved by integrating the indirect, retrospective insights provided by population-genetic analyses with the more direct, mechanistic insights provided by functional experiments. Integrative studies of adaptive genetic variation may sometimes be motivated by experimental insights into molecular function, which then provide the impetus to perform population genetic tests to evaluate whether the functional variation is of adaptive significance. In other cases, studies may be initiated by genome scans of DNA variation to identify candidate loci for recent adaptation. Results of such analyses can then motivate experimental efforts to test whether the identified candidate loci do in fact contribute to functional variation in some fitness-related phenotype. Functional studies can provide corroborative evidence for positive selection at particular loci, and can potentially reveal specific molecular mechanisms of adaptation.

Figure 1

A hypothetical example that illustrates the challenge of documenting mechanistic connections between allelic variation in gene function and fitness. Panel A shows the simplest case, where alternative alleles at an enzyme-encoding gene A have different catalytic efficiencies (= biochemical phenotypes) that translate into differences in physiological performance (= physiological phenotype), which in turn translate into fitness differences. Functional variation in the enzyme encoded by gene A is unaffected by genetic background; there is no epistasis with other loci such as gene B. Consequently, there is a simple mapping of variation at gene A to biochemical phenotype, physiological performance, and ultimately, fitness. In the example shown in Panel B, the molecular basis of fitness variation is slightly more complex as the biochemical phenotype is determined by the epistatic interaction between the enzymes encoded by genes A and B. Nonetheless, each two-locus genotype is associated with a distinct phenotype that ultimately determines its fitness ranking. In the example shown in Panel C, the performance phenotype is continuously distributed due to contributions of many genes of individually small effect. In such cases, it will be far more difficult to relate variation at specific genes to specific fitness rankings. Modified from Koehn et al. (1983).

Figure 2

Polymorphic amino acid residues in the PGI enzyme of Colias eurytheme, and variation in site-frequency spectra across the coding region of the underlying gene. (A) Homology-based model of a single PGI monomer (green) showing segregating amino acid sites (369 & 375) located in an exterior inter-helical loop that is positioned at the intersubuint interface of the dimeric enzyme. This region is part of a peptide chain (yellow) that connects active site residues Glu 361 and His 392. (B) A sliding window analysis of Tajima’s D based on synonymous site polymorphism across exons of the Pgi gene in C. eurytheme. Codons 369 and 375 are located in exon 9 (step size = 25bp and window length = 70bp, which is half the average exon length). Modified from Wheat et al. (2006).

Figure 3

Homology-based structural model of deer mouse hemoglobin (Hb) showing the location of 12 polymorphic amino acid sites that exhibit allele frequency differences between high and low altitude populations. Mutations in the α- and β-chain subunits are shown in panels A and B, respectively. These represent candidate sites for the adaptive fine-tuning of Hb-O₂ affinity between highland and lowland populations. Based on data reported in Storz et al. (2009 in Storz et al. (2010).

Figure 4

O₂-equilibrium curves of deer mouse hemoglobins showing allelic differences in Hb-O₂ affinity. (A) Curves for high-altitude mice that express the βI Hb isoform (product of the d¹ β-globin allele) in the presence and absence of allosteric cofactors (2,3-DPG and Cl⁻ ions); (B) Curves for low-altitude mice that express the βI Hb isoform (product of the d⁰ β-globin allele) under the same experimental treatments; (C) Summary of allelic differences in O₂ affinity and cofactor sensitivity between the β-chain hemoglobin isoforms of high- and low-altitude mice. Based on data reported in Storz et al. (2009).

Figure 5

Temporal changes in allele and genotype frequencies at the Eda gene in four replicate freshwater populations of threespined stickleback. (A) Changes in frequency of the ‘low plated’ Eda allele in four replicate ponds (different colored lines). All samples are from the first (F₁) cohort of offspring, except the June and July 2007 samples, which are from the second (F₂) pond generation. (B). Approximate life history stages through the course of the experiment. Fish were stained with Alizarin red to highlight bone. (C). Genotype frequencies averaged across all four ponds. Purple, homozygous complete genotype (CC); orange, heterozygote genotype (CL); green, homozygote low genotype (LL). Vertical bars show standard errors on the basis of n = 4 ponds. From Barrett et al. (2008).

Figure 6

Graphical representation of haplotype homozygosity in the region surrounding independently derived cis-regulatory mutations that are associated with adult persistence of lactase activity in different human populations. (A) Tracts of homozygous SNP genotypes flanking the causative SNP (G/C-14010) are shown for a sample of human subjects from Kenya and Tanzania. Homozygosity tracts in red are associated with the derived SNP allele (C-14010) that contributes to the lactase persistence phenotype, and homozygosity tracts in blue are associated with the wildtype SNP allele (G-14010). (B) Tracts of homozygous SNP genotypes flanking the causative SNP (C/T-13910) are shown for a sample of human subjects from Europe and Asia. Homozygosity tracts in green are associated with the derived SNP allele (T-13910) that contributes to the lactase persistence phenotype, and homozygosity tracts in blue are associated with the wildtype SNP allele (C-13910). In both the African and Eurasian populations, the extended lengths of homozygosity tracts associated with the derived SNP alleles appear to reflect the fact that these variants were recently driven up to high frequency by positive directional selection, so recombination has not yet whittled down the size of the ancestral haplotypes in which the adaptive mutations originated. Nucleotide positions along the x-axis are relative to start codon of the lactase gene. From Tishkoff et al. (2007).

Figure 7

An adaptive Agouti allele that contributes to cryptic coloration of deer mice from the Nebraska Sand Hills appears to have arisen after the formation of the dune field habitat (a minimum of 8,000 yrs ago). (A) The Agouti allele class (the ‘wideband’ haplotype) that harbors the causative mutation(s) harbors far less polymorphism than the wildtype allele class. Rows are observed haplotypes for exon 2 and flanking regions; columns represent polymorphic nucleotide positions (black = ancestral, white = derived). Arrows indicate the position of a derived amino acid deletion that is associated with the wideband phenotpe. Numbers of segregating sites and nucleotide diversities (S and π, respectively) are given for both allele classes. (B) Posterior probability distribution for the estimated age of the adaptive wideband allele in Sand Hills deer mice (assuming N_e = 10,000 and 2 generations per year). From Linnen et al. (2009).

Figure 8

Global allele frequencies and haplotype patterns at the SLC24A5 gene in humans. (A) Pie charts of allele frequencies at the rs1426654 SNP, where frequencies of the derived and ancestral alleles are shown in red and blue, respectively. (B) A representation of haplotype patterns for 500 kb around the SLC24A5 gene, centered on rs1426654. Each box represents a single population sample, and observed haplotypes are plotted as thin horizontal lines using the same haplotype color-code for each sample. The derived allele of the the rs1426654 SNP is mainly found on the red haplotype. From Coop et al. (2009).

Figure 9

Effect of SLC24A5 genotype on skin pigmentation in two admixed human populations, African-Americans and African-Caribbeans. (A) Variation in skin pigmentation index with estimated African ancestry and SLC24A5 genotype. Each point represents a single individual. Lines show regressions constrained to have equal slopes, for each of the three SLC24A5 genotypes (GG, AG, and AA). (B) Histograms showing the distribution of pigmentation scores associated with each genotype after adjusting for the estimated percentage of African ancestry. Values shown are the differences between the measured melaning index and the calculated regression line for the GG genotype. From Lamason et al. (2005).