Relationship between transmission ratio distortion and genetic divergence in intraspecific rice crosses

APOK3, a pollen killer antidote in Arabidopsis thaliana

The principles of heredity state that the two alleles carried by a heterozygote are equally transmitted to the progeny. However, genomic regions that escape this rule have been reported in many organisms. It is notably the case of genetic loci referred to as gamete killers, where one allele enhances its transmission by causing the death of the gametes that do not carry it. Gamete killers are of great interest, particularly to understand mechanisms of evolution and speciation. Although being common in plants, only a few, all in rice, have so far been deciphered to the causal genes. Here, we studied a pollen killer found in hybrids between two accessions of Arabidopsis thaliana. Exploring natural variation, we observed this pollen killer in many crosses within the species. Genetic analyses revealed that three genetically linked elements are necessary for pollen killer activity. Using mutants, we showed that this pollen killer works according to a poison-antidote model, where the poison kills pollen grains not producing the antidote. We identified the gene encoding the antidote, a chimeric protein addressed to mitochondria. De novo genomic sequencing in twelve natural variants with different behaviors regarding the pollen killer revealed a hyper variable locus, with important structural variations particularly in killer genotypes, where the antidote gene recently underwent duplications. Our results strongly suggest that the gene has newly evolved within A. thaliana. Finally, we identified in the protein sequence polymorphisms related to its antidote activity.

Artificial Selection in Domestication and Breeding Prevents Speciation in Rice

Speciation has long been regarded as an irreversible process once the reproductive barriers had been established. However, unlike in natural populations, artificial selection might either accelerate or prevent speciation processes in domesticated species. Asian cultivated rice is a target crop for both domestication and artificial breeding; it contains two subspecies of indica and japonica, which usually produce sterile inter-subspecific hybrids due to reproductive barriers. In this study, we constructed the evolutionary trajectory of a reproductive isolation system S5, which regulates fertility in indica-japonica hybrids via three adjacent genes, based on the data of 606 accessions including two cultivated and 11 wild rice species. Although hybrid sterility haplotypes at S5 lead to establishment of a killer-protector reproductive barrier, origin of wide-compatibility haplotypes by complex hybridization and recombination provides an opposing force to reproductive isolation and thus prevents speciation during domestication. Analysis in a diallel set of 209 crosses involving 21 parents showed that the wide-compatibility genotypes largely rescued fertility of indica-japonica hybrids, indicating that the wide-compatibility gene would enable gene flow to maintain species coherence. This counteracting system indicates that combined effects of natural evolution and artificial selection may result in reversible processes of speciation in rice, which may also have implications for genetic improvement of rice.

Transmission ratio distortion is frequent in Arabidopsis thaliana controlled crosses

The equal probability of transmission of alleles from either parent during sexual reproduction is a central tenet of genetics and evolutionary biology. Yet, there are many cases where this rule is violated. The preferential transmission of alleles or genotypes is termed transmission ratio distortion (TRD). Examples of TRD have been identified in many species, implying that they are universal, but the resolution of species-wide studies of TRD are limited. We have performed a species-wide screen for TRD in over 500 segregating F2 populations of Arabidopsis thaliana using pooled reduced-representation genome sequencing. TRD was evident in up to a quarter of surveyed populations. Most populations exhibited distortion at only one genomic region, with some regions being repeatedly affected in multiple populations. Our results begin to elucidate the species-level architecture of biased transmission of genetic material in A. thaliana, and serve as a springboard for future studies into the biological basis of TRD in this species.

Genetic Properties of a Nested Association Mapping Population Constructed With Semi-Winter and Spring Oilseed Rapes

Nested association mapping (NAM) populations have been widely applied to dissect the genetic basis of complex quantitative traits in a variety of crops. In this study, we developed a Brassica napus NAM (BN-NAM) population consisting of 15 recombination inbred line (RIL) families with 2,425 immortal genotypes. Fifteen high-density genetic linkage maps were constructed by genotyping by sequencing (GBS) based on all RIL families, with further integration into a joint linkage map (JLM) having 30,209 unique markers in common with multiple linkage maps. Furthermore, an ultra-density whole-genome variation map was constructed by projecting 4,444,309 high-quality variants onto the JLM. The NAM population captured a total of 88,542 recombination events (REs). The uneven distribution of recombination rate along chromosomes is positively correlated with the densities of genes and markers, but negatively correlated with the density of transposable elements and linkage disequilibrium (LD). Analyses of population structure and principal components revealed that the BN-NAM population could be divided into three groups with weak stratification. The LD decay distance across genome varied between 170 and 2,400 Kb, with LD decay more rapid in the A than in the C sub-genome. The pericentromeric regions contained large LD blocks, especially in the C sub-genome. This NAM population provides a valuable resource for dissecting the genetic basis of important traits in rapeseed, especially in semi-winter oilseed rape.

Linkage Map of Lissotriton Newts Provides Insight into the Genetic Basis of Reproductive Isolation

Linkage maps are widely used to investigate structure, function, and evolution of genomes. In speciation research, maps facilitate the study of the genetic architecture of reproductive isolation by allowing identification of genomic regions underlying reduced fitness of hybrids. Here we present a linkage map for European newts of the Lissotriton vulgaris species complex, constructed using two families of F2 L. montandoni × L. vulgaris hybrids. The map consists of 1146 protein-coding genes on 12 linkage groups, equal to the haploid chromosome number, with a total length of 1484 cM (1.29 cM per marker). It is notably shorter than two other maps available for salamanders, but the differences in map length are consistent with cytogenetic estimates of the number of chiasmata per chromosomal arm. Thus, large salamander genomes do not necessarily translate into long linkage maps, as previously suggested. Consequently, salamanders are an excellent model to study evolutionary consequences of recombination rate variation in taxa with large genomes and a similar number of chromosomes. A complex pattern of transmission ratio distortion (TRD) was detected: TRD occurred mostly in one family, in one breeding season, and was clustered in two genomic segments. This is consistent with environment-dependent mortality of individuals carrying L. montandoni alleles in these two segments and suggests a role of TRD blocks in reproductive isolation. The reported linkage map will empower studies on the genomic architecture of divergence and interactions between the genomes of hybridizing newts.

Spodoptera frugiperda (Lepidoptera: Noctuidae) host-plant variants: two host strains or two distinct species?

The moth Spodoptera frugiperda is a well-known pest of crops throughout the Americas, which consists of two strains adapted to different host-plants: the first feeds preferentially on corn, cotton and sorghum whereas the second is more associated with rice and several pasture grasses. Though morphologically indistinguishable, they exhibit differences in their mating behavior, pheromone compositions, and show development variability according to the host-plant. Though the latter suggest that both strains are different species, this issue is still highly controversial because hybrids naturally occur in the wild, not to mention the discrepancies among published results concerning mating success between the two strains. In order to clarify the status of the two host-plant strains of S. frugiperda, we analyze features that possibly reflect the level of post-zygotic isolation: (1) first generation (F1) hybrid lethality and sterility; (2) patterns of meiotic segregation of hybrids in reciprocal second generation (F2), as compared to the meiosis of the two parental strains. We found a significant reduction of mating success in F1 in one direction of the cross and a high level of microsatellite markers showing transmission ratio distortion in the F2 progeny. Our results support the existence of post-zygotic reproductive isolation between the two laboratory strains and are in accordance with the marked level of genetic differentiation that was recovered between individuals of the two strains collected from the field. Altogether these results provide additional evidence in favor of a sibling species status for the two strains.

Punctuated distribution of recombination hotspots and demarcation of pericentromeric regions in Phaseolus vulgaris L

High density genetic maps are a reliable tool for genetic dissection of complex plant traits. Mapping resolution is often hampered by the variable crossover and non-crossover events occurring across the genome, with pericentromeric regions (pCENR) showing highly suppressed recombination rates. The efficiency of linkage mapping can further be improved by characterizing and understanding the distribution of recombinational activity along individual chromosomes. In order to evaluate the genome wide recombination rate in common beans (Phaseolus vulgaris L.) we developed a SNP-based linkage map using the genotype-by-sequencing approach with a 188 recombinant inbred line family generated from an inter gene pool cross (Andean x Mesoamerican). We identified 1,112 SNPs that were subsequently used to construct a robust linkage map with 11 groups, comprising 513 recombinationally unique marker loci spanning 943 cM (LOD 3.0). Comparative analysis showed that the linkage map spanned >95% of the physical map, indicating that the map is almost saturated. Evaluation of genome-wide recombination rate indicated that at least 45% of the genome is highly recombinationally suppressed, and allowed us to estimate locations of pCENRs. We observed an average recombination rate of 0.25 cM/Mb in pCENRs as compared to the rest of genome that showed 3.72 cM/Mb. However, several hot spots of recombination were also detected with recombination rates reaching as high as 34 cM/Mb. Hotspots were mostly found towards the end of chromosomes, which also happened to be gene-rich regions. Analyzing relationships between linkage and physical map indicated a punctuated distribution of recombinational hot spots across the genome.

Interspecific tests of allelism reveal the evolutionary timing and pattern of accumulation of reproductive isolation mutations

Despite extensive theory, little is known about the empirical accumulation and evolutionary timing of mutations that contribute to speciation. Here we combined QTL (Quantitative Trait Loci) analyses of reproductive isolation, with information on species evolutionary relationships, to reconstruct the order and timing of mutations contributing to reproductive isolation between three plant (Solanum) species. To evaluate whether reproductive isolation QTL that appear to coincide in more than one species pair are homologous, we used cross-specific tests of allelism and found evidence for both homologous and lineage-specific (non-homologous) alleles at these co-localized loci. These data, along with isolation QTL unique to single species pairs, indicate that >85% of isolation-causing mutations arose later in the history of divergence between species. Phylogenetically explicit analyses of these data support non-linear models of accumulation of hybrid incompatibility, although the specific best-fit model differs between seed (pairwise interactions) and pollen (multi-locus interactions) sterility traits. Our findings corroborate theory that predicts an acceleration ('snowballing') in the accumulation of isolation loci as lineages progressively diverge, and suggest different underlying genetic bases for pollen versus seed sterility. Pollen sterility in particular appears to be due to complex genetic interactions, and we show this is consistent with a snowball model where later arising mutations are more likely to be involved in pairwise or multi-locus interactions that specifically involve ancestral alleles, compared to earlier arising mutations.

Sequencing, assembling, and correcting draft genomes using recombinant populations

Current de novo whole-genome sequencing approaches often are inadequate for organisms lacking substantial preexisting genetic data. Problems with these methods are manifest as: large numbers of scaffolds that are not ordered within chromosomes or assigned to individual chromosomes, misassembly of allelic sequences as separate loci when the individual(s) being sequenced are heterozygous, and the collapse of recently duplicated sequences into a single locus, regardless of levels of heterozygosity. Here we propose a new approach for producing de novo whole-genome sequences—which we call recombinant population genome construction—that solves many of the problems encountered in standard genome assembly and that can be applied in model and nonmodel organisms. Our approach takes advantage of next-generation sequencing technologies to simultaneously barcode and sequence a large number of individuals from a recombinant population. The sequences of all recombinants can be combined to create an initial de novo assembly, followed by the use of individual recombinant genotypes to correct assembly splitting/collapsing and to order and orient scaffolds within linkage groups. Recombinant population genome construction can rapidly accelerate the transformation of nonmodel species into genome-enabled systems by simultaneously producing a high-quality genome assembly and providing genomic tools (e.g., high-confidence single-nucleotide polymorphisms) for immediate applications. In populations segregating for important functional traits, this approach also enables simultaneous mapping of quantitative trait loci. We demonstrate our method using simulated Illumina data from a recombinant population of Caenorhabditis elegans and show that the method can produce a high-fidelity, high-quality genome assembly for both parents of the cross.

Bridging the genotyping gap: using genotyping by sequencing (GBS) to add high-density SNP markers and new value to traditional bi-parental mapping and breeding populations

Genotyping by sequencing (GBS) is the latest application of next-generation sequencing protocols for the purposes of discovering and genotyping SNPs in a variety of crop species and populations. Unlike other high-density genotyping technologies which have mainly been applied to general interest "reference" genomes, the low cost of GBS makes it an attractive means of saturating mapping and breeding populations with a high density of SNP markers. One barrier to the widespread use of GBS has been the difficulty of the bioinformatics analysis as the approach is accompanied by a high number of erroneous SNP calls which are not easily diagnosed or corrected. In this study, we use a 384-plex GBS protocol to add 30,984 markers to an indica (IR64) × japonica (Azucena) mapping population consisting of 176 recombinant inbred lines of rice (Oryza sativa) and we release our imputation and error correction pipeline to address initial GBS data sparsity and error, and streamline the process of adding SNPs to RIL populations. Using the final imputed and corrected dataset of 30,984 markers, we were able to map recombination hot and cold spots and regions of segregation distortion across the genome with a high degree of accuracy, thus identifying regions of the genome containing putative sterility loci. We mapped QTL for leaf width and aluminum tolerance, and were able to identify additional QTL for both phenotypes when using the full set of 30,984 SNPs that were not identified using a subset of only 1,464 SNPs, including a previously unreported QTL for aluminum tolerance located directly within a recombination hotspot on chromosome 1. These results suggest that adding a high density of SNP markers to a mapping or breeding population through GBS has a great value for numerous applications in rice breeding and genetics research.

Investigating incipient speciation in Arabidopsis lyrata from patterns of transmission ratio distortion

Our understanding of the development of intrinsic reproductive isolation is still largely based on theoretical models and thorough empirical studies on a small number of species. Theory suggests that reproductive isolation develops through accumulation of epistatic genic incompatibilities, also known as Bateson-Dobzhansky-Muller (BDM) incompatibilities. We can detect these from marker transmission ratio distortion (TRD) in hybrid progenies of crosses between species or populations, where TRD is expected to result from selection against heterospecific allele combinations in hybrids. TRD may also manifest itself because of intragenomic conflicts or competition between gametes or zygotes. We studied early stage speciation in Arabidopsis lyrata by investigating patterns of TRD across the genome in F2 progenies of three reciprocal crosses between four natural populations. We found that the degree of TRD increases with genetic distance between crossed populations, but also that reciprocal progenies may differ substantially in their degree of TRD. Chromosomes AL6 and especially AL1 appear to be involved in many single- and two-locus distortions, but the location and source of TRD vary between crosses and between reciprocal progenies. We also found that the majority of single- and two-locus TRD appears to have a gametic, as opposed to zygotic, origin. Thus, while theory on BDM incompatibilities is typically illustrated with derived nuclear alleles proving incompatible in hybrid zygotes, our results suggest a prominent role for distortions emerging before zygote formation.

Conflictual speciation: species formation via genomic conflict

A remarkable suite of forms of genomic conflict has recently been implicated in speciation. We propose that these diverse roles of genomic conflict in speciation processes can be unified using the concept of 'conflictual speciation'. Conflictual speciation centers on the evolution of reproductive isolation as a byproduct of antagonistic selection among genomic elements with divergent fitness interests. Intragenomic conflicts are expected to readily generate Dobzhansky-Muller incompatibilities, due to population-specific interactions between opposing elements, and thus they could be especially important in speciation. Moreover, selection from genomic conflicts should be relatively unrelenting across ecological and evolutionary time scales. We explain how intragenomic conflicts can promote, or sometimes constrain, speciation, and describe evidence relating conflicts to the evolution of reproductive isolation.

OsGA20ox1, a candidate gene for a major QTL controlling seedling vigor in rice

Seedling vigor is among the major determinants of stable stand establishment in direct-seeded rice (Oryza sativa L.) in temperate regions. Quantitative trait loci (QTL) for seedling vigor were identified using 250 recombinant inbred lines (RILs) derived from a cross between two japonica rice cultivars Kakehashi and Dunghan Shali. Seedling heights measured at 14 days after sowing were 20.3 and 29.4 cm for Kakehashi and Dunghan Shali, respectively. For the RILs, the height ranged from 14.1 to 31.7 cm. Four putative QTLs associated with seedling height were detected. qPHS3-2, the major QTL that was located on the long arm of chromosome 3, accounted for 26.2 % of the phenotypic variance. Using progeny of the near isogenic lines (NILs) produced by the backcross introduction of a chromosome segment carrying this major QTL into an elite cultivar Iwatekko, we fine-mapped qPHS3-2 to a 81-kb interval between two markers, ID_CAPS_01 and RM16227. Within this mapped region, we identified the gene OsGA20ox1, which is related to gibberellin (GA) biosynthesis. The relative expression levels of GA20ox1 in seedlings of Dunghan Shali and NILs were higher than that of Iwatekko. Concomitantly, the amount of endogenous active GA was higher in Dunghan Shali and the NILs compared to the level detected in Iwatekko. These results indicate that OsGA20ox1 is a strong candidate gene for major QTL controlling seedling vigor in rice.