Inference of domestication history and differentiation between early‐ and late‐flowering varieties in pearl millet

Pearl millet (Pennisetum glaucum) is a staple crop in Sahelian Africa. Farmers usually grow varieties with different cycle lengths and complementary functions in Sahelian agrosystems. Both the level of genetic differentiation of these varieties and the domestication history of pearl millet have been poorly studied. We investigated the neutral genetic diversity and population genetic structure of early‐ and late‐flowering domesticated and wild pearl millet populations using 18 microsatellite loci and 8 nucleotide sequences. Strikingly, early‐ and late‐flowering domesticated varieties were not differentiated over their whole distribution area, despite a clear difference in their isolation‐by‐distance pattern. Conversely, our data brought evidence for two well‐differentiated genetic pools in wild pearl millet, allowing us to test scenarios with different numbers and origins of domestication using approximate Bayesian computation (ABC). The ABC analysis showed the likely existence of asymmetric migration between wild and domesticated populations. The model choice procedure indicated that a single domestication from the eastern wild populations was the more likely scenario to explain the polymorphism patterns observed in cultivated pearl millet.     

Phylogeny and origin of pearl millet (Pennisetum glaucum [L.] R. Br) as revealed by microsatellite loci

During the last 12,000 years, different cultures around the world have domesticated cereal crops. Several studies investigated the evolutionary history and domestication of cereals such as wheat in the Middle East, rice in Asia or maize in America. The domestication process in Africa has led to the emergence of important cereal crops like pearl millet in Sahelian Africa. In this study, we used 27 microsatellite loci to analyze 84 wild accessions and 355 cultivated accessions originating from the whole pearl millet distribution area in Africa and Asia. We found significantly higher diversity in the wild pearl millet group. The cultivated pearl millet sample possessed 81% of the alleles and 83% of the genetic diversity of the wild pearl millet sample. Using Bayesian approaches, we identified intermediate genotypes between the cultivated and wild groups. We then analyzed the phylogenetic relationship among accessions not showing introgression and found that a monophyletic origin of cultivated pearl millet in West Africa is the most likely scenario supported by our data set.     

Levels and patterns of nucleotide variation in domestication QTL regions on rice chromosome 3 suggest lineage-specific selection

Oryza sativa or Asian cultivated rice is one of the major cereal grass species domesticated for human food use during the Neolithic. Domestication of this species from the wild grass Oryza rufipogon was accompanied by changes in several traits, including seed shattering, percent seed set, tillering, grain weight, and flowering time. Quantitative trait locus (QTL) mapping has identified three genomic regions in chromosome 3 that appear to be associated with these traits. We would like to study whether these regions show signatures of selection and whether the same genetic basis underlies the domestication of different rice varieties. Fragments of 88 genes spanning these three genomic regions were sequenced from multiple accessions of two major varietal groups in O. sativa--indica and tropical japonica--as well as the ancestral wild rice species O. rufipogon. In tropical japonica, the levels of nucleotide variation in these three QTL regions are significantly lower compared to genome-wide levels, and coalescent simulations based on a complex demographic model of rice domestication indicate that these patterns are consistent with selection. In contrast, there is no significant reduction in nucleotide diversity in the homologous regions in indica rice. These results suggest that there are differences in the genetic and selective basis for domestication between these two Asian rice varietal groups.     

Association mapping, patterns of linkage disequilibrium and selection in the vicinity of the PHYTOCHROME C gene in pearl millet

Key message

Linkage analysis confirmed the association in the region of PHYC in pearl millet. The comparison of genes found in this region suggests that PHYC is the best candidate.

Abstract

Major efforts are currently underway to dissect the phenotype–genotype relationship in plants and animals using existing populations. This method exploits historical recombinations accumulated in these populations. However, linkage disequilibrium sometimes extends over a relatively long distance, particularly in genomic regions containing polymorphisms that have been targets for selection. In this case, many genes in the region could be statistically associated with the trait shaped by the selected polymorphism. Statistical analyses could help in identifying the best candidate genes into such a region where an association is found. In a previous study, we proposed that a fragment of the PHYTOCHROME C gene (PHYC) is associated with flowering time and morphological variations in pearl millet. In the present study, we first performed linkage analyses using three pearl millet F₂ families to confirm the presence of a QTL in the vicinity of PHYC. We then analyzed a wider genomic region of ~100 kb around PHYC to pinpoint the gene that best explains the association with the trait in this region. A panel of 90 pearl millet inbred lines was used to assess the association. We used a Markov chain Monte Carlo approach to compare 75 markers distributed along this 100-kb region. We found the best candidate markers on the PHYC gene. Signatures of selection in this region were assessed in an independent data set and pointed to the same gene. These results foster confidence in the likely role of PHYC in phenotypic variation and encourage the development of functional studies.     

A Gene Family Derived from Transposable Elements during Early Angiosperm Evolution Has Reproductive Fitness Benefits in Arabidopsis thaliana

The benefits of ever-growing numbers of sequenced eukaryotic genomes will not be fully realized until we learn to decipher vast stretches of noncoding DNA, largely composed of transposable elements. Transposable elements persist through self-replication, but some genes once encoded by transposable elements have, through a process called molecular domestication, evolved new functions that increase fitness. Although they have conferred numerous adaptations, the number of such domesticated transposable element genes remains unknown, so their evolutionary and functional impact cannot be fully assessed. Systematic searches that exploit genomic signatures of natural selection have been employed to identify potential domesticated genes, but their predictions have yet to be experimentally verified. To this end, we investigated a family of domesticated genes called MUSTANG (MUG), identified in a previous bioinformatic search of plant genomes. We show that MUG genes are functional. Mutants of Arabidopsis thaliana MUG genes yield phenotypes with severely reduced plant fitness through decreased plant size, delayed flowering, abnormal development of floral organs, and markedly reduced fertility. MUG genes are present in all flowering plants, but not in any non-flowering plant lineages, such as gymnosperms, suggesting that the molecular domestication of MUG may have been an integral part of early angiosperm evolution. This study shows that systematic searches can be successful at identifying functional genetic elements in noncoding regions and demonstrates how to combine systematic searches with reverse genetics in a fruitful way to decipher eukaryotic genomes.     

Classic Selective Sweeps Revealed by Massive Sequencing in Cattle

Human driven selection during domestication and subsequent breed formation has likely left detectable signatures within the genome of modern cattle. The elucidation of these signatures of selection is of interest from the perspective of evolutionary biology, and for identifying domestication-related genes that ultimately may help to further genetically improve this economically important animal. To this end, we employed a panel of more than 15 million autosomal SNPs identified from re-sequencing of 43 Fleckvieh animals. We mainly applied two somewhat complementary statistics, the integrated Haplotype Homozygosity Score (iHS) reflecting primarily ongoing selection, and the Composite of Likelihood Ratio (CLR) having the most power to detect completed selection after fixation of the advantageous allele. We find 106 candidate selection regions, many of which are harboring genes related to phenotypes relevant in domestication, such as coat coloring pattern, neurobehavioral functioning and sensory perception including KIT, MITF, MC1R, NRG4, Erbb4, TMEM132D and TAS2R16, among others. To further investigate the relationship between genes with signatures of selection and genes identified in QTL mapping studies, we use a sample of 3062 animals to perform four genome-wide association analyses using appearance traits, body size and somatic cell count. We show that regions associated with coat coloring significantly (P<0.0001) overlap with the candidate selection regions, suggesting that the selection signals we identify are associated with traits known to be affected by selection during domestication. Results also provide further evidence regarding the complexity of the genetics underlying coat coloring in cattle. This study illustrates the potential of population genetic approaches for identifying genomic regions affecting domestication-related phenotypes and further helps to identify specific regions targeted by selection during speciation, domestication and breed formation of cattle. We also show that Linkage Disequilibrium (LD) decays in cattle at a much faster rate than previously thought.     

Tuareg, a novel miniature-inverted repeat family of pearl millet (Pennisetum glaucum) related to the PIF superfamily of maize

Miniature-inverted repeat transposable elements (MITEs) are abundantly repeated in plant genomes and are especially found in genic regions where they could contribute regulatory elements for gene expression. We describe with molecular and cytological tools the first MITE family reported in pearl millet: Tuareg. It was initially detected in the pearl millet ortholog of Teosinte-branched1, an important developmental gene involved in the domestication of maize. The Tuareg family was amplified recently in the pearl millet genome and elements were found more abundant in wild than in domesticated plants. We found that they shared similarity in their terminal repeats with the previously described mPIF MITEs and that they are also present in other Pennisetum species, in maize and more distantly related grasses. The Tuareg family may be part of MITEs activated by PIF-like transposases and it could have been mobile since pearl millet domestication. Electronic supplementary material Electronic supplementary material is available for this article at and accessible for authorised users. O. Robin contributed the FISH and fiber-FISH hybridizations.     

Comparative analysis of QTLs affecting domestication traits between two domesticated x wild pearl millet ( Pennisetum glaucum L., Poaceae) crosses

Comparative mapping of Quantitative trait loci (QTLs) involved in domestication of adaptative syndrome traits of pearl millet was realized at the intra-specific level using two F(2) populations derived from domesticated ( Pennisetum glaucum ssp. glaucum) x wild ( Pennisetum glaucum ssp. monodii) crosses. The two domesticated parents analyzed differ in their geographical origins, agronomic characteristics and life cycles. In both populations, two regions of the genome were identified on linkage groups 6 and 7, that controlled most of the key morphological differences. The importance of these two linkage groups reveals their central role both in the developmental control of spikelet structure and in the domestication process of this crop. In contrast, QTLs involved in traits that are components of yield and measure differences in resource allocation (such as the shape of the spike, the number of spikes per plant and plant height) show a low level of correspondence among our two crosses. The results of the comparative mapping between cereals, although preliminary, reveal that genes involved in seed-shattering could correspond in maize, rice and sorghum. The evolutionary significance of our results, and especially the relationships between genome organization and cereal domestication, are discussed. The potential use of these results in pearl millet genetic-resources enhancement are presented.     

Contributions of flowering time genes to sunflower domestication and improvement

Determining the identity and distribution of molecular changes leading to the evolution of modern crop species provides major insights into the timing and nature of historical forces involved in rapid phenotypic evolution. In this study, we employed an integrated candidate gene strategy to identify loci involved in the evolution of flowering time during early domestication and modern improvement of the sunflower (Helianthus annuus). Sunflower homologs of many genes with known functions in flowering time were isolated and cataloged. Then, colocalization with previously mapped quantitative trait loci (QTLs), expression, or protein sequence differences between wild and domesticated sunflower, and molecular evolutionary signatures of selective sweeps were applied as step-wise criteria for narrowing down an original pool of 30 candidates. This process led to the discovery that five paralogs in the flowering locus T/terminal flower 1 gene family experienced selective sweeps during the evolution of cultivated sunflower and may be the causal loci underlying flowering time QTLs. Our findings suggest that gene duplication fosters evolutionary innovation and that natural variation in both coding and regulatory sequences of these paralogs responded to a complex history of artificial selection on flowering time during the evolution of cultivated sunflower.     

A draft genome of sweet cherry (Prunus avium L.) reveals genome‐wide and local effects of domestication

Sweet cherry (Prunus avium L.) trees are both economically important fruit crops but also important components of natural forest ecosystems in Europe, Asia and Africa. Wild and domesticated trees currently coexist in the same geographic areas with important questions arising on their historical relationships. Little is known about the effects of the domestication process on the evolution of the sweet cherry genome. We assembled and annotated the genome of the cultivated variety “Big Star*” and assessed the genetic diversity among 97 sweet cherry accessions representing three different stages in the domestication and breeding process (wild trees, landraces and modern varieties). The genetic diversity analysis revealed significant genome‐wide losses of variation among the three stages and supports a clear distinction between wild and domesticated trees, with only limited gene flow being detected between wild trees and domesticated landraces. We identified 11 domestication sweeps and five breeding sweeps covering, respectively, 11.0 and 2.4 Mb of the P. avium genome. A considerable fraction of the domestication sweeps overlaps with those detected in the related species, Prunus persica (peach), indicating that artificial selection during domestication may have acted independently on the same regions and genes in the two species. We detected 104 candidate genes in sweep regions involved in different processes, such as the determination of fruit texture, the regulation of flowering and fruit ripening and the resistance to pathogens. The signatures of selection identified will enable future evolutionary studies and provide a valuable resource for genetic improvement and conservation programs in sweet cherry.     

Largely unlinked gene sets targeted by selection for domestication syndrome phenotypes in maize and sorghum

The domestication of diverse grain crops from wild grasses was a result of artificial selection for a suite of overlapping traits producing changes referred to in aggregate as ‘domestication syndrome’. Parallel phenotypic change can be accomplished by either selection on orthologous genes or selection on non‐orthologous genes with parallel phenotypic effects. To determine how often artificial selection for domestication traits in the grasses targeted orthologous genes, we employed resequencing data from wild and domesticated accessions of Zea (maize) and Sorghum (sorghum). Many ‘classic’ domestication genes identified through quantitative trait locus mapping in populations resulting from wild/domesticated crosses indeed show signatures of parallel selection in both maize and sorghum. However, the overall number of genes showing signatures of parallel selection in both species is not significantly different from that expected by chance. This suggests that while a small number of genes will extremely large phenotypic effects have been targeted repeatedly by artificial selection during domestication, the optimization part of domestication targeted small and largely non‐overlapping subsets of all possible genes which could produce equivalent phenotypic alterations.     

Quantitative trait loci associated with traits determining grain and stover yield in pearl millet under terminal drought-stress conditions

Drought stress during the reproductive stage is one of the most important environmental factors reducing the grain yield and yield stability of pearl millet. A QTL mapping approach has been used in this study to understand the genetic and physiological basis of drought tolerance in pearl millet and to provide a more-targeted approach to improving the drought tolerance and yield of this crop in water-limited environments. The aim was to identify specific genomic regions associated with the enhanced tolerance of pearl millet to drought stress during the flowering and grain-filling stages. Testcrosses of a set of mapping-population progenies, derived from a cross of two inbred pollinators that differed in their response to drought, were evaluated in a range of managed terminal drought-stress environments. A number of genomic regions were associated with drought tolerance in terms of both grain yield and its components. For example, a QTL associated with grain yield per se and for the drought tolerance of grain yield mapped on linkage group 2 and explained up to 23% of the phenotypic variation. Some of these QTLs were common across stress environments whereas others were specific to only a particular stress environment. All the QTLs that contributed to increased drought tolerance did so either through better than average maintenance (compared to non-stress environments) of harvest index, or harvest index and biomass productivity. It is concluded that there is considerable potential for marker-assisted backcross transfer of selected QTLs to the elite parent of the mapping population and for their general use in the improvement of pearl millet productivity in water-limited environments. Received: 15 November 2000 / Accepted: 12 April 2001     

Methylation analysis revealed salicylic acid affects pearl millet defense through external cytosine DNA demethylation

The cytosine DNA methylation and demethylation have a role in regulating plant responses to the environment by affecting the promoter regions of most plant defense-related genes through the CpG islands or the CCGG motifs. Salicylic acid, a defense and signaling plant hormone, is seen playing crucial role in the variation of the methylome. In this study, the effects of salicylic acid and feeding of the millet headminer (Heliocheilus albipunctella de Joannis) on pearl millet DNA methylome changes were evaluated through MSAP epigenotyping during panicle development. The results showed that millet headminer feeding increased the level of genomic methylation while application of salicylic acid caused DNA demethylation occurring mostly at external cytosine and accompanied by a decrease of the number of larvae per panicle. This suggests that hemimethylation (external cytosine methylation) has key role in regulating defense responses and conferring tolerance to pearl millet through salicylic acid application.     

MADS-box out of the black box

The compelling elegance of using genome‐wide scans to detect the signature of selection is difficult to resist, but is countered by the low demonstrated efficacy of pinpointing the actual genes and traits that are the targets of selection in nonmodel species. While the difficulty of going from a suggestive signature to a functional nucleotide polymorphism should not prevent researchers from using genome scans, it does lessen their long‐term utility within and across study systems. In a new study published in this issue of Molecular Ecology ( Mariac et al. 2011 ), researchers have gone a long way towards increasing the relevance of genome‐wide scans for selection via two approaches: (i) they tailored the markers used in the scan to target a family of developmental genes that were good candidates for controlling a trait of interest and (ii) they used an independent mapping population to confirm the association of the gene with polymorphism in the trait of interest. All of this was completed in the nonmodel system of pearl millet (Pennisetum glaucum) and may provide a road map for other researchers hoping to pin down solid candidate genes for selected traits in natural or cultivated systems. Outside of these broad methodological innovations, the paper specifically focuses on a trait (flowering time) that varies across an environmental gradient (rainfall). This environmental gradient potentially serves as a model for environmental change over time, and allele frequencies at the gene can therefore be used to track how populations of pearl millet will adapt to future climate shifts at the genetic level.     

Association Studies Identify Natural Variation at PHYC Linked to Flowering Time and Morphological Variation in Pearl Millet

The identification of genes selected during and after plant domestication is an important research topic to enhance knowledge on adaptative evolution. Adaptation to different climates was a key factor in the spread of domesticated crops. We conducted a study to identify genes responsible for these adaptations in pearl millet and developed an association framework to identify genetic variations associated with the phenotype in this species. A set of 90 inbred lines genotyped using microsatellite loci and AFLP markers was used. The population structure was assessed using two different Bayesian approaches that allow inbreeding or not. Association studies were performed using a linear mixed model considering both the population structure and familial relationships between inbred lines. We assessed the ability of the method to limit the number of false positive associations on the basis of the two different Bayesian methods, the number of populations considered and different morphological traits while also assessing the power of the methodology to detect given additive effects. Finally, we applied this methodology to a set of eight pearl millet genes homologous to cereal flowering pathway genes. We found significant associations between several polymorphisms of the pearl millet PHYC gene and flowering time, spike length, and stem diameter in the inbred line panel. To validate this association, we performed a second association analysis in a different set of pearl millet individuals from Niger. We confirmed a significant association between genetic variation in this gene and these characters.DOMESTICATION and dispersion of cultivated plants were associated with their adaptation to the agricultural environment. These adaptations led to genetic changes shared by all individuals of a cultivated species (domestication genes) or to variations between varieties within a cultivated species (genes controlling varietal differences). Domestication genes like tb1 (Doebley et al. 1997; Wang et al. 1999) in maize (Zea mays) were selected very early by human populations (Jaenicke-Després et al. 2003). After the first early selection, adaptation of the flowering phenotype to different climatic conditions was certainly a key innovation that enabled colonization of new environments. One of the most well-known examples was the adaptation of maize—a tropical plant—to northern climates. Maize cultivation spread late to northeastern America. By 1000 YBP, only maize was an established staple crop (Fritz 1995). A genetic variant of the Dwarf8 gene led to an earlier flowering phenotype (Thornsberry et al. 2001). This early allele was present at a high frequency in North America and was certainly selected after the domestication of maize under northern climatic conditions (Camus-Kulandaivelu et al. 2006).Pearl millet (Pennisetum glaucum [(L.) R. Br.]), one of the most important West African cereals, was most likely domesticated once in the Sahelian zone of West Africa (Oumar et al. 2008). By 3500 YBP, it was already being cultivated throughout Sahelian and tropical West African countries (D''Andrea et al. 2001; D''Andrea and Casey 2002). The adaptation of pearl millet in West Africa was also associated with an environmental gradient (Haussmann et al. 2006). Pearl millet varieties from tropical coastal West Africa flower very late (up to 160 days from planting to female flowering) as compared to varieties from Sahelian West Africa, which may have a flowering time as short as 45 days (Haussmann et al. 2006). The genetic factors underlying the differences between these varieties are still unknown.Association studies offer new opportunities for assessing the role of a particular gene on a phenotype. Contrary to QTL analysis, association studies have the challenging task of taking an unknown evolutionary history of studied individuals into account. For example, population structure is a common confounding effect in association studies (Pritchard et al. 2000a). Allele frequencies evolve between divergent structured populations via drift, mutation, and selection. Differences in allele frequencies may be correlated with any morphological traits that differentiate two populations. Then a statistical correlation between a gene and a trait is not necessarily associated with a “causative” relationship between the gene and the morphology, which can lead to a high number of false positives. The use of population structure to correct the number of false positives was a significant breakthrough in plant studies (Thornsberry et al. 2001). This approach was recently further refined by also using a matrix of kinship coefficients, which proves efficient when there is a complex structure and familial relationship between individuals (Yu et al. 2006; Kang et al. 2008; Stich et al. 2008). Complex structures and familial relationships are common in inbred cultivated crop material. In the current association study framework (Thornsberry et al. 2001; Yu et al. 2006; Casa et al. 2008; Kang et al. 2008; Stich et al. 2008), population structure was assessed using STRUCTURE software (Pritchard et al. 2000b). This tool is not implemented to deal with selfed inbred materials or inbred species (Pritchard et al. 2000b). Through new methodological developments, population structure analysis can now be performed using Bayesian methods in these particular cases (Gao et al. 2007). The extent to which the power of association studies will differ when dealing with inbred material or selfing species using either Bayesian method has yet to be evaluated.In this study, we developed an association framework for pearl millet to assess the role of flowering pathway genes. We assessed the ability of the method to control the number of false positives, while taking different methodological inferences of population structure that allow inbreeding or not into account. We also assessed the power of the association framework to detect given additive genetic effects. Finally, we applied this method to a set of eight flowering time gene homologs sequenced in pearl millet. We assessed sequence variation in light perception genes (PHYA, PHYB, PHYC, and CRY2) and downstream regulators of flowering (GI, Hd6, Hd1, and FLORICAULA). Variation was detected in the PHYC gene associated with variations in flowering time and morphological traits. This association was noted in two different data sets.     

Polymorphism pattern at a miniature inverted‐repeat transposable element locus downstream of the domestication gene Teosinte‐branched1 in wild and domesticated pearl millet

Unravelling the mechanisms involved in adaptation to understand plant morphological evolution is a challenging goal. For crop species, identification of molecular causal polymorphisms involved in domestication traits is central to this issue. Pearl millet, a domesticated grass mostly found in semi‐arid areas of Africa and India, is an interesting model to address this topic: the domesticated form shares common derived phenotypes with some other cereals such as a decreased ability to develop basal and axillary branches in comparison with the wild phenotype. Two recent studies have shown that the orthologue of the maize gene Teosinte‐Branched1 in pearl millet (PgTb1) was probably involved in branching evolution during domestication and that a miniature inverted‐repeat transposable element (MITE) of the Tuareg family was inserted in the 3′ untranslated region of PgTb1. For a set of 35 wild and domesticated populations, we compared the polymorphism patterns at this MITE and at microsatellite loci. The Tuareg insertion was nearly absent in the wild populations, whereas a strong longitudinal frequency cline was observed in the domesticated populations. The geographical pattern revealed by neutral microsatellite loci clearly demonstrated that isolation by distance does not account for the existence of this cline. However, comparison of population differentiation at the microsatellite and the MITE loci and analyses of the nucleotide polymorphism pattern in the downstream region of PgTb1 did not show evidence that the cline at the MITE locus has been shaped by selection, suggesting the implication of a neutral process. Alternative hypotheses are discussed.     

Within‐plant variation in reproductive investment: consequences for selection on flowering time

Variation among the leaves, flowers or fruit produced by a plant is often regarded as a nuisance to the experimenter and an impediment to selection. Here, we suggest that within‐plant variation can drive selection on other plant‐level traits. We examine within‐plant variation in floral sex allocation and in fruit set and predict that such variation generates variation in male success among plants, thereby driving selection on flowering time. We tested this prediction in a simulation model estimating selection on flowering time through male fitness when floral sex allocation and/or fruit set vary directionally among flowers on plants. We parameterized the model through a quantitative literature survey of within‐plant change in sex allocation. As predicted, within‐plant variation in floral sex allocation and in fruit set probability can generate selection on flowering time through male fitness. Declining fruit set from first to last flowers on plants, as occurs in many species, selected for early flowering onset through male fitness. This result was robust to self‐incompatibility and to varying returns on male versus female investment. Selection caused by declining fruit set was strong enough to reverse the selection for late flowering that can be caused by intrafloral protandry. Our model provides testable predictions regarding selection on flowering time through male fitness. The model also establishes the intriguing possibility that within‐plant variation may influence selection on other traits, regardless of whether that variation is under selection itself.