A genome-wide scan for relaxation of constraints in the human lineage affecting specific functional processes
Author(s): Somel, M, Wilson-Sayres, M, Jordan, G, Huerta-Sanchez, E, Fumagalli, M, Ferrer-Admetlla, A, Nielsen, R
Changes in the subsistence mode of a species can lead to adaptive evolution of new functions, while it can also cause relaxed negative selection in previously essential functions. While positive selection in humans has been intensely studied, functional processes subject to relaxed constraints in the human lineage remain largely unknown. Here we present a framework for detecting relaxation of selective constraints that affect a particular functional process specifically in one taxon. Jointly using human and chimpanzee population genomic data with mammalian comparative genomic data, we identify olfactory receptors and proteasome subunits as candidates of relaxed constraints in humans: both gene sets contain high frequency non-synonymous mutations in humans while having conserved amino-acid sequences across other mammals. We further discuss the possible underlying causes of this signal.
Department of Genetics
A quantitative genetics approach to equilibrium assessment and equilibrium fitness estimation; an application to study the polymorphism of the human ACP1 in Europe
Author(s): Álvarez-Castro, JM
Studies of equilibrium and stability under selection are liable to further development. With multiple alleles, a tiny proportion of possible genotype frequencies are possible at polymorphic equilibria. I present a quantitative genetics (QG) method for checking whether a set of genotype frequencies are equilibrium frequencies that also provides estimates of the fitnesses of all genotypes. This method is here applied to a one-locus multiallelic system, but applications to multiple-locus systems with epistasis are also possible. Method Classical QG studies focused on directional selection on the phenotype. Considering instead that the trait is fitness enables tracking genotypic frequencies under other selection regimes. Thus, if the additive variance may be minimized to zero for a set of observed frequencies, then these are putative equilibrium frequencies and the corresponding fitnesses are estimates of the equilibrium fitnesses. Two exclusive properties of NOIA are crucial for performing such analysis: it generalizes previous models and it provides explicit expressions of variance components (Álvarez-Castro and Carlborg 2007, Álvarez-Castro and Yang 2011). The human ACP1 polymorphism in Europe Three alleles of ACP1 (A, B, C) coexist in human European populations and C was claimed to be deleterious using biallelic methods. The multiallelic method proposed here shows instead that the observed genotypic frequencies are consistent with a polymorphic equilibrium with fitnesses in accordance with physiological observations (AA affected by fetal macrosomia). NOIA reveals a fast rise of additive variance for slight fitness increases of AA (not published). This may actually be leading to fixation of A in European populations since fetal macrosomia--with the aid of pregnancy monitoring and reasonably safe caesarean sections--entails nowadays less of a problem than it did some decades ago. ALVAREZ-CASTRO & CARLBORG 2007 Genetics 176 ALVAREZ-CASTRO & YANG 2011 Genetica 139
Evolution of a co-expression network involved in primate brain functions: clues to the evolution of human specific cognitive abilities
Author(s): Nowick, K, Berto, S
Cognitive abilities are among the most striking differences between humans and other primates. As cognition is a complex trait it is beneficial to analyze its evolution on multiple levels. Here we analyzed genome wide expression data of several human, chimpanzee, and rhesus macaque brains to identify molecular changes that may underlie the evolution of human-specific cognitive skills. Because transcription factors (TFs) are main players in regulating gene expression, we focused on TF changes that might drive expression pattern differences between the species. For each TF with a species-specific expression level we identified co-expressed genes, which are potential target genes or interaction partners of the respective TFs. Using this information we derived weighted topological overlap (wTO) networks in which the nodes are the TFs and the weighted links represent the commonality of the TFs’ co-expressed genes. We revealed: 1. The ancestral network, consisting only of links common to all three species, was involved in cognitive functions, such as forebrain development, synaptic plasticity, and learning, as inferred from the Gene Ontology (GO) terms enriched among the genes co-expressed with the most TFs in this network. 2. In the human network, TFs are more interconnected than in the chimpanzee and rhesus macaque network, indicating more complex regulation of genes involved in cognition in the human brain. 3. The TF sub-network constructed only from genes with a human-specific expression change was enriched for GO terms that point to changes in energy metabolism, learning and memory processes in the human brain. Moreover, only in this sub-network, but not in the sub-networks of the other species, we identified an excess of TFs implicated in mental disorders. Taken together, in a network of TFs that is predicted to regulate cognitive functions we identified evolutionary changes that are potentially involved in the evolution of human- specific cognitive abilities.
School of Biology
Experimental studies of human social learning strategies: exploring sex differences
Author(s): Cross, C, Brown, G, Morgan, T, Laland, K
Objectives Culture is an important driver of recent biological evolution in humans. The mechanisms by which information is transmitted between individuals can be studied at the population level – by cultural evolutionists, and at the individual level – by social psychologists. We combined methods from these two approaches to investigate how sex differences in confidence might lead to sex differences in the use of a copy-when-uncertain social learning strategy. Methods Participants (Study 1: N=97; Study 2: N=89) completed a series of two-alternative forced-choice puzzles and reported their confidence in each answer. They then saw the decisions of some previous participants before being asked again for their answer. Social information use was inferred when participants switched their answer to match that of the majority. We modelled the probability of social information use with participant sex, confidence in initial decision, and accuracy of initial decision as predictors. Results Across both studies, confidence had a large effect on social information use, indicative of a copy-when-uncertain strategy. Accuracy predicted confidence, indicating that this strategy is adaptive. Confidence also differed by sex: women reported lower confidence (independent of any small sex differences in accuracy), which in turn increased their probability of using social information. Conclusions Although both sexes appear to use a ‘copy-when-uncertain’ strategy, women are more likely to feel uncertain. This means that a strategy observed to be used in a population (e.g. copy-when-uncertain) can vary according to individual differences in psychological traits. Further integration of these two levels of explanation is therefore needed.
Institute of Evolutionary Biology and Environmental Studies
Fitness meets fitness: taking an evolutionary approach to physical fitness in humans
Author(s): Postma, E
Evolutionary biologists are usually the first to emphasize that Darwinian fitness has little in common with the meaning that is commonly attached to fitness by the general public, that is physical fitness. However, recent studies in humans and non-human animals are suggestive of an important role for physical fitness in shaping variation in Darwinian fitness via natural and sexual selection, both in the past and in the present. Indeed, it has been argued that it is selection on physical performance that has made us who we are today. To gain a better understanding of variation in (physical) fitness, I use concepts and methods from evolutionary biology and life-history theory and apply these to data on human running performance. Specifically, using a large (longitudinal and cross-sectional) data set for running performance by both men and women over a wide range of distances, I test for sex differences, the effects of ageing and training, and for trade-offs between long- versus short-distance and generalists versus specialists. I will use these findings to infer the selective pressures acting on physical fitness in humans, and to argue that Darwinian and physical fitness may have more in common with each other than is often assumed.
Institute of Integrative Biology & Department of Aquatic Ecology
Genetic constraints underlying human reproductive timing in a pre-modern Swiss village
Author(s): Bürkli, A, Postma, E
The trade-off between reproductive investment in early versus late life is central to life-history theory. Despite abundant empirical evidence in support of different versions of this trade-off, the specific trade-off between age at first reproduction (AFR) and age at last reproduction (ALR) has received little attention, especially in long-lived species with a pronounced reproductive senescence such as humans. Using genealogical data for a 19th-century Swiss village, we (i) quantify natural selection on reproductive timing, (ii) estimate additive genetic (co)variances, and (iii) use these to predict evolutionary responses. Selection gradients were computed using multiple linear regressions, and the additive genetic variance-covariance matrix was estimated using a restricted maximum-likelihood animal model. We found strong selection for both an early AFR and a late ALR, which resulted from selection for an earlier and longer reproductive period (RP, i.e. ALR-AFR). Furthermore, postponing AFR shortened RP in both sexes, but twice as much in women. Finally, AFR and ALR were strongly and positively genetically correlated, which led to a considerable reduction in the predicted responses to selection, or even rendered them maladaptive. These results provide evidence for strong genetic constraints underlying reproductive timing in humans, which may have contributed to the evolution of menopause.
Department of Pathology and Imaging
Genetic genealogy comes of age: advances in the use of deep-rooted pedigrees in human evolutionary research
Author(s): Larmuseau, MHD, Van Geystelen, A, Decorte, R
Research on the recent human evolution will benefit from the implementation of extended genetic genealogical data. The approach to combine deep-rooted pedigrees with genetic information advances the understanding of changes in the human population genetic structure during the last centuries. This recent advance is mainly based on the extensive growth of whole genome sequencing data and available genealogical data of high quality. Moreover, according to the latest genetic genealogical research the historical non-paternity rate in Western Europe is estimated around 1% per generation within the last four centuries, which means that the expected relationship between the legal genealogy and the genetics of DNA donors exists. Therefore, genetic genealogical data will help with three research aims of human evolutionary studies: (I) detecting signals of (past) population stratification and interpreting the population structure in a more objective manner, (II) obtaining the time scale and impact of particular detected gene flow events more accurately and (III) determining temporal genetic differentiation within a population by combining in-depth pedigree data with haploid markers. Each of these research aims will be discussed with examples of the human population in Flanders (Western Europe). At the end, we will discuss the advantages and pitfalls of using genetic genealogy within studies on human evolutionary genomics.
Animal and Plant Sciences
How has the genetic architecture of key life history traits responded to the demographic transition in a human population?
Author(s): Bolund, E, Lummaa, V
A central issue in evolutionary biology concerns the long-term reliability of predictions of evolutionary change. Theory predicts that the genetic architecture of life history traits (summarised by the additive genetic variance-covariance matrix, G) will change over time and thus affect how traits respond to selection. However, studies have found support both for fast changes as well as for a great consistency in G over time. It thus remains unclear how rapidly and in what manner G itself changes in response to changes in selection pressures or environment.
The demographic transition to low mortality and fertility rates in many recent human populations involves a drastic environmental change, but its consequences for the evolutionary potential of traits have rarely been addressed. We use genealogical data from 8 parishes in Finland, from natural high (5-6 offspring) to recent low (< 2 offspring) fertility over 350 years to address this question at the genetic level. We study four key life history traits; age at first and last reproduction, number of offspring and longevity, all of which show significant phenotypic changes during the time period. We use the animal model quantitative genetic approach to study whether and how the genetic architecture underlying these traits has also changed, by comparing the full G matrix in the periods before and after the demographic transition.
First, we establish significant additive genetic variance and heritability for all traits during both time periods. Second, we present the genetic covariances and correlations between all four traits during both periods. Third, we compare the overall G matrix of the two periods to elucidate if and how G has changed during the demographic transition. The results provide a novel insight in how traits can respond to selection in contemporary human populations and whether the potential for such responses might have changed along with the recent demographic and societal changes.
Research Center of Biodiversity and Genetic Resources (CIBIO)
Identifying recent Darwinian selection acting on different human APOL1 variants among diverse African populations in trypanosomiasis endemic areas
Author(s): Ko, W, Rajan, P, Gomez, F, Schelinfeldt, L, Froment, A, Lema, G, Nyambo, TB, Omar, SA, Wambebe, C, Ranciaro, A, Hirbo, JB, Thompson, S, Tishkoff, SA
Disease susceptibility may arise as a consequence of adaptation to infectious disease. Recent findings have suggested that higher rates of chronic kidney disease (CKD) in individuals with recent African ancestry may be attributed to two APOL1 risk alleles (G1 and G2) at the SRA-interacting domain. These two alleles appear to have risen adaptively, possibly due to their protective effects against African trypanosomiasis (African sleeping sickness). In order to explore the distribution of potential functional variation at APOL1, we studied nucleotide variation in 187 individuals across ten geographically and genetically diverse African ethnic groups with differing levels of exposure to two subspecies of Trypanosoma brucei that cause African sleeping sickness. We observed unusually high levels of nonsynonymous polymorphism at the functional domains that are required to lyse parasites. Whereas allele frequencies of G2 are similar across all populations (3-8%), the G1 allele is only common in the Yoruba (39%). We further identified a previously unreported haplotype (termed G3) that contains a nonsynonymous mutation at the membrane-addressing domain of APOL1 and is present in all populations except for the Yoruba. Analyses of long-range patterns of LD indicate evidence for recent selection acting on the G3 haplotype in Fulani (cattle-herding population) from Cameroon that are endemic to Gambian trypanosomiasis, suggesting its possible protective effects against trypanosome infection. Our results indicate that the G1 and G2 variants are geographically restricted and that there are more candidate variants at other functional domains contributing to trypanosomiasis resistance and risk for CKD in African populations.
Investigating the co-evolution between human lactase gene and dairy culture in Europe
Author(s): Papoli Yazdi, H, Gautier, M, Vitalis, R
Genes that have undergone recent positive selection during human evolution can provide new insights into the evolutionary forces that have shaped our species. Human lactase gene LCT which encodes the enzyme responsible for hydrolysis of lactose, the milk sugar is under recent positive selection. The production of lactase is declined after the weaning period; however, in some individuals, lactase expression persists into adulthood. Lactase persistence is found at high frequency in northern and western European populations. Based on Principal Component Analysis (PCA) of genetic diversity of six markers in cattle milk protein genes, Beja-Pereira et al. (2003) proposed that since Neolithic times, there has been co-evolution between LCT in humans and cattle milk protein genes. To test this hypothesis, we used dense SNP datasets for human (441,845 SNPs) and cattle (44,706 SNPs) populations to investigate the signatures of positive selection at LCT and cattle milk protein genes in Europe. To this end, we used a new genome-scan method based on a diffusion approximation for the distribution of allele frequency at migration-selection-drift equilibrium (Vitalis et al, in prep). For LCT, we detected a north-south gradient for the intensity of positive selection. For the cattle data, we could not detect any signature of positive selection in cattle milk protein genes. Furthermore, PCA was performed using the markers surrounding cattle milk protein genes provided the same pattern of geographical variation as with the complete cattle genomic data, showing a north-south gradient of variation across Europe. The results of our study therefore provide no evidence for the co-evolution of lactase and cattle milk protein genes. We argue that the coincidence between the gradient in the selection intensity at LCT in humans and geographic variation in cattle genes simply reflect the migration history of both species during the Neolithic along the Danubian route.