Population Genomic Analyses Based on 1 Million SNPs in Commercial Egg Layers

Identifying signatures of selection can provide valuable insight about the genes or genomic regions that are or have been under selective pressure, which can lead to a better understanding of genotype-phenotype relationships. A common strategy for selection signature detection is to compare samples from several populations and search for genomic regions with outstanding genetic differentiation. Wright''s fixation index, F_ST, is a useful index for evaluation of genetic differentiation between populations. The aim of this study was to detect selective signatures between different chicken groups based on SNP-wise F_ST calculation. A total of 96 individuals of three commercial layer breeds and 14 non-commercial fancy breeds were genotyped with three different 600K SNP-chips. After filtering a total of 1 million SNPs were available for F_ST calculation. Averages of F_ST values were calculated for overlapping windows. Comparisons of these were then conducted between commercial egg layers and non-commercial fancy breeds, as well as between white egg layers and brown egg layers. Comparing non-commercial and commercial breeds resulted in the detection of 630 selective signatures, while 656 selective signatures were detected in the comparison between the commercial egg-layer breeds. Annotation of selection signature regions revealed various genes corresponding to productions traits, for which layer breeds were selected. Among them were NCOA1, SREBF2 and RALGAPA1 associated with reproductive traits, broodiness and egg production. Furthermore, several of the detected genes were associated with growth and carcass traits, including POMC, PRKAB2, SPP1, IGF2, CAPN1, TGFb2 and IGFBP2. Our approach demonstrates that including different populations with a specific breeding history can provide a unique opportunity for a better understanding of farm animal selection.     

Genome-wide association study reveals the genetic determinism of growth traits in a Gushi-Anka F2 chicken population

Chicken growth traits are economically important, but the relevant genetic mechanisms have not yet been elucidated. Herein, we performed a genome-wide association study to identify the variants associated with growth traits. In total, 860 chickens from a Gushi-Anka F₂ resource population were phenotyped for 68 growth and carcass traits, and 768 samples were genotyped based on the genotyping-by-sequencing (GBS) method. Finally, 734 chickens and 321,314 SNPs remained after quality control and removal of the sex chromosomes, and these data were used to carry out a GWAS analysis. A total of 470 significant single-nucleotide polymorphisms (SNPs) for 43 of the 68 traits were detected and mapped on chromosomes (Chr) 1–6, -9, -10, -16, -18, -23, and -27. Of these, the significant SNPs in Chr1, -4, and -27 were found to be associated with more than 10 traits. Multiple traits shared significant SNPs, indicating that the same mutation in the region might have a large effect on multiple growth or carcass traits. Haplotype analysis revealed that SNPs within the candidate region of Chr1 presented a mosaic pattern. The significant SNPs and pathway enrichment analysis revealed that the MLNR, MED4, CAB39L, LDB2, and IGF2BP1 genes could be putative candidate genes for growth and carcass traits. The findings of this study improve our understanding of the genetic mechanisms regulating chicken growth and carcass traits and provide a theoretical basis for chicken breeding programs.Subject terms: Genome-wide association studies, Genetic linkage study, Development, DNA sequencing, Animal breeding     

Genome scan linkage analysis identifies quantitative trait loci affecting serum clinical–chemical traits in Korean native chicken

Alterations in robustness- and health-related traits lead to physiological changes, such as changes in the serum clinical chemical parameters in individuals. Therefore, clinical–chemical traits can be used as biomarkers to examine the health status of chickens. The aim of the present study was to detect the quantitative trait loci (QTLs) influencing eight clinical–chemical traits (glucose, total protein, creatinine, high-density lipoprotein cholesterol, total cholesterol, glutamic oxaloacetic transaminase, glutamic pyruvic transaminase, and α-amylase) in an F₁ nuclear families comprising 83 F₀ founders and 585 F₁ progeny of Korean native chickens. Genotypic data on 135 DNA markers representing 26 autosomes have been generated for this resource pedigree. The total length of the map was 2729.4 cM. We used a multipoint variance component linkage approach to identify QTLs for the traits. A significant QTL affecting serum α-amylase levels was identified on chicken chromosome (GGA) 7 [logarithm of odds (LOD) = 3.02, P value = 1.92 × 10^?4]. Additionally, we detected several suggestive linkage signals for the levels of total cholesterol, glutamic oxaloacetic transaminase, glutamic pyruvic transaminase, and creatinine on GGA 4, 12, 13, and 15. In this study, serum α-amylase levels related significant QTL was mapped on GGA7 and cholesterol, glutamic oxaloacetic transaminase, glutamic pyruvic transaminase, and creatinine traits related suggestive QTLs were detected on GGA4, 12, 13 and 15, respectively. Further verification and fine mapping of these identified QTLs can provide valuable information for understanding the variations of clinical chemical trait in chickens.     

Asymmetries,heterosis, and phenotypic profiles of red junglefowl,White Plymouth Rocks,and F<Subscript>1</Subscript> and F<Subscript>2</Subscript> reciprocal crosses

During the domestication of farm animals, humans have manipulated genetic variation for growth and reproduction through artificial selection. Here, data are presented for growth, reproductive, and behavior traits for the red junglefowl, a line of White Plymouth Rock chickens, and their F₁ and F₂ reciprocal crosses. Intra- and intergenerational comparisons for growth related traits reflected considerable additive genetic variation. In contrast, those traits associated with reproduction exhibited heterosis. The role of sexual selection was seen in the evolution of prominent secondary sexual ornaments that lend to female choice and male-male competition. The large differences between parental lines in fearfulness to humans were only mitigated slightly in the intercross generations. Whereas, overall F₁ generation heterosis was not transferred to the F₂, there was developmental stability in the F₂, as measured by relative asymmetry of bilateral traits. Through multigenerational analyses between the red junglefowl and the domestic White Plymouth Rocks, we observed plasticity and considerable residual genetic variation. These factors likely facilitated the adaptability of the chicken to a broad range of husbandry practices throughout the world.     

Cloning of TPO gene and associations of polymorphisms with chicken growth and carcass traits

Thyroid peroxidase (TPO), which located on the apical membrane surface of thyrocytes, is the key enzyme involved in thyroid hormone synthesis, mainly catalyses the iodination of tyrosine residues and the coupling of iodotyrosines on thyroglobulin to form thyroxine and triiodothyronine. The objectives of this study were to identify genetic polymorphisms of the chicken TPO gene and to analyze potential association between single nucleotide polymorphisms (SNPs) and growth and carcass traits in chicken. Partial sequences of TPO gene were cloned firstly. The nucleotide sequence was found to have 72 % identity with that of humans. The chicken TPO amino acid sequence was 71 %. Through polymerase chain reaction-restriction fragment length polymorphism and DNA sequencing methods, three novel mutations of the chicken TPO gene were detected in the F₂ resource population from Gushi chickens and Anka broilers. The association analysis indicated that all of the three SNPs showed association with chicken growth at different periods. The g.29996C>T polymorphisms was significantly associated with body weight, breast bone length, pectoral angle at 12 weeks, claw weight and leg muscle weight (P < 0.05). In addition, individuals with the TT genotype had higher value for almost all the traits than CC and CT genotype. Meanwhile for CLW, the additive effects were significant (P < 0.05). Hence, we suggest that genotype TT can be regarded as a potential molecular marker for later growth and carcass traits in chicken.     

Genome-wide association study on chicken carcass traits using sequence data imputed from SNP array

Chicken carcass traits are economically important for the chicken industry. Detecting which genes affect chicken carcass traits is of great benefit to the genetic improvement of this important agricultural species. To investigate the genetic mechanism of carcass traits in chickens, we carried out a genome-wide association study (GWAS). A total of 435 Chinese indigenous chickens were phenotyped for carcass weight (CW), eviscerated weight with giblets (EWG), and eviscerated weight (EW) after slaughter at 91 days and were genotyped using a 600-K single nucleotide polymorphism (SNP) genotyping array. Twenty-four birds were selected for sequencing, and the 600 K SNP panel data were imputed to sequence data with the 24 birds as the reference. Univariate GWASs were performed with GEMMA software using the whole genome sequence data imputed from SNP chip data. Finally, 3, 25, and 63 suggestively significant SNPs were identified to be associated with carcass weight (CW), eviscerated weight with giblets (EWG), and eviscerated weight (EW), respectively. Six candidate genes, RNF219, SCEL, MYCBP2, ETS1, APLP2, and PRDM10 were detected. SCEL and MYCBP2 were potentially associated with these three traits, RNF219 and APLP2 were potentially associated with EWG and EW, and ETS1 and PRDM10 were only potentially associated with EWG and EW, respectively. Compared with forefathers’ research, 10 reported QTLs associated with CW were located within a 5-Mb distance near the SNPs with P value lower than 1×10^?5. This study enriched the knowledge of the genetic mechanisms of chicken carcass traits.     

A genome-wide scan for quantitative trait loci affecting limb bone lengths and areal bone mineral density of the distal femur in a White Duroc × Erhualian F<Subscript>2</Subscript> population

Background

Limb bone lengths and bone mineral density (BMD) have been used to assess the bone growth and the risk of bone fractures in pigs, respectively. It has been suggested that limb bone lengths and BMD are under genetic control. However, the knowledge about the genetic basis of the limb bone lengths and mineralisatinon is limited in pigs. The aim of this study was to identify quantitative trait loci (QTL) affecting limb bone lengths and BMD of the distal femur in a White Duroc × Erhualian resource population.

Results

Limb bone lengths and femoral bone mineral density (fBMD) were measured in a total of 1021 and 116 F₂ animals, respectively. There were strong positive correlations among the lengths of limb bones and medium positive correlations between the lengths of limb bones and fBMD. A whole-genome scan involving 183 microsatellite markers across the pig genome revealed 35 QTL for the limb bone lengths and 2 for femoral BMD. The most significant QTL for the lengths of five limb bones were mapped on two chromosomes affecting all 5 limb bones traits. One was detected around 57 cM on pig chromosome (SSC) 7 with the largest F-value of more than 26 and 95% confidence intervals of less than 5 cM, providing a crucial start point to identify the causal genes for these traits. The Erhualian alleles were associated with longer limb bones. The other was located on SSCX with a peak at 50–53 cM, whereas alleles from the White Duroc breed increased the bone length. Many QTL identified are homologous to the human genomic regions containing QTL for bone-related traits and a list of interesting candidate genes.

Conclusion

This study detected the QTL for the lengths of scapula, ulna, humerus and tibia and fBMD in the pig for the first time. Moreover, several new QTL for the pig femoral length were found. As correlated traits, QTL for the lengths of five limb bones were mainly located in the same genomic regions. The most promising QTL for the lengths of five limb bones on SSC7 merits further investigation.

     

Genome-wide analyses reveal population structure and identify candidate genes associated with tail fatness in local sheep from a semi-arid area

Under a climate change perspective, the genetic make-up of local livestock breeds showing adaptive traits should be explored and preserved as a priority. We used genotype data from the ovine 50 k Illumina BeadChip for assessing breed autozygosity based on runs of homozygosity (ROH) and fine-scale genetic structure and for detecting genomic regions under selection in 63 Tunisian sheep samples. The average genomic inbreeding coefficients based on ROH were estimated at 0.017, 0.021, and 0.024 for Barbarine (BAR, n = 26), Noire de Thibar (NDT, n = 23), and Queue fine de l'Ouest (QFO, n = 14) breeds, respectively. The genomic relationships among individuals based on identity by state (IBS) distance matrix highlighted a recent introgression of QFO into the BAR and a genetic differentiation of NDT samples, possibly explained by past introgression of European gene pools. Genome-wide scan for ROH across breeds and within the BAR sample set identified an outstanding signal on chromosome 13 (46.58–49.61 Mbp). These results were confirmed using F_ST index, differentiating fat vs. thin-tailed individuals. Candidate genes under selection pressure (CDS2, PROKR1, and BMP2) were associated to lipid storage and probably preferentially selected in fat-tailed BAR animals. Our findings suggest paying more attention to preserve the genetic integrity and adaptive alleles of local sheep breeds.     

Association of IGF1 and KDM5A polymorphisms with performance, fatness and carcass traits in chickens

Two functional and positional candidate genes were selected in a region of chicken chromosome 1 (GGA1), based on their biological roles, and also where several quantitative trait loci (QTL) have been mapped and associated with performance, fatness and carcass traits in chickens. The insulin-like growth factor 1 (IGF1) gene has been associated with several physiological functions related to growth. The lysine (K)-specific demethylase 5A (KDM5A) gene participates in the epigenetic regulation of genes involved with the cell cycle. Our objective was to find associations of selected single-nucleotide polymorphisms (SNPs) in these genes with performance, fatness and carcass traits in 165 F₂ chickens from a resource population. In the IGF1 gene, 17 SNPs were detected, and in the KDM5A gene, nine SNPs were detected. IGF1 SNP c.47673G?>?A was associated with body weight and haematocrit percentage, and also with feed intake and percentages of abdominal fat and gizzard genotype × sex interactions. KDM5A SNP c.34208C?>?T genotype × sex interaction affected body weight, feed intake, percentages of abdominal fat (p?=?0.0001), carcass, gizzard and haematocrit. A strong association of the diplotype × sex interaction (p?<?0.0001) with abdominal fat was observed, and also associations with body weight, feed intake, percentages of carcass, drums and thighs, gizzard and haematocrit. Our findings suggest that the KDM5A gene might play an important role in the abdominal fat deposition in chickens. The IGF1 and KDM5A genes are strong candidates to explain the QTL mapped in this region of GGA1.     

Novel SNPs in the <Emphasis Type="Italic">PRDM16</Emphasis> gene and their associations with performance traits in chickens

The PR domain containing 16 (PRDM16) is a member of the Prdm family, and is known to regulate cell differentiation. In the present study, DNA pool sequencing methods were employed to screen genetic variations in the chicken PRDM16 gene. The results revealed four novel single nucleotide polymorphisms (SNPs): NC_006108.2: g.92188G>A, XM_417551: c.1161C>T (Ala/Ala, 387aa), c.1233C>T (Ser/Ser, 411aa) and c.1433G>A (Ser/Asn, 478aa). The BglI polymerase chain reaction–restriction fragment length polymorphism (PCR-RFLP) was used to detect c.1161C>T, while HhaI Forced PCR-RFLP methods were used to detect 1233C>T and c.1433G>A in 964 chickens. The chickens comprised 38 grandparents, 66 F₁ parents and 860 F₂ birds derived from an F₂ resource population of Gushi chickens crossed with Anka broilers. The associations of the polymorphisms in the chicken PRDM16 gene with performance traits were analyzed in the 860 F₂ chickens. The results indicated that the three SNPs were significantly associated with growth, fatness and meat quality traits in the chickens. In particular, the polymorphisms of the missense SNP (c.1433G>A) had positive effects on chicken body weight and body size at different stages. It affected also fatness traits significantly. Comparison of the different genotypes of c.1433G>A showed that the GG genotype favored chicken growth and fatness traits.     

Gene interactions and genetics of blast resistance and yield attributes in rice (Oryza sativa L.)

Blast disease caused by the pathogen Pyricularia oryzae is a serious threat to rice production. Six generations viz., P₁, P₂, F₁, F₂, B₁ and B₂ of a cross between blast susceptible high-yielding rice cultivar ADT 43 and resistant near isogenic line (NIL) CT13432-3R, carrying four blast resistance genes Pi1, Pi2, Pi33 and Pi54 in combination were used to study the nature and magnitude of gene action for disease resistance and yield attributes. The epistatic interaction model was found adequate to explain the gene action in most of the traits. The interaction was complementary for number of productive tillers, economic yield, lesion number, infected leaf area and potential disease incidence but duplicate epistasis was observed for the remaining traits. Among the genotypes tested under epiphytotic conditions, gene pyramided lines were highly resistant to blast compared to individuals with single genes indicating that the nonallelic genes have a complementary effect when present together. The information on genetics of various contributing traits of resistance will further aid plant breeders in choosing appropriate breeding strategy for blast resistance and yield enhancement in rice.     

Genome-wide association studies for mottled eggs in chickens using a high-density single-nucleotide polymorphism array

Mottled eggs in layer chickens are gaining increasing attention because of the economic impact on the egg industry caused by the reduced sale value of commodity eggs. However, the genetic architecture underlying mottled eggs is not well understood. The genetic architecture underlying the mottled egg trait was investigated using genome-wide association studies (GWAS) by high-density arrays, using a total of 407 pink eggs and 799 blue eggs from an F₂ resource population generated by crossing Dongxiang Blue-shelled and White Leghorn chickens. The mottled egg score in blue eggs was found to be higher than that in pink eggs. The single-nucleotide polymorphism heritability of mottled egg at laying day and storage for 7 days was 0.18 and 0.20, respectively. Bivariate GWAS provided 29 significant loci, mainly located on GGA2, GGA3, GGA8, GGA10, GGA15, GGA17, and GGA23, affecting mottled egg on laying day. Candidate genes RIMS2, SLC25A32, RIMBP2, VPS13B, and RGS3 were obtained for mottled eggshell by bivariate GWAS and gene annotation. Our findings provide new insights into the genetic architecture of mottled egg in hens, and demonstrate that a genomic selection method would be profitable for breeding out the mottled egg trait.     

Dopamine Receptor Genes and Evolutionary Differentiation in the Domestication of Fighting Cocks and Long-Crowing Chickens

The chicken domestication process represents a typical model of artificial selection, and gives significant insight into the general understanding of the influence of artificial selection on recognizable phenotypes. Two Japanese domesticated chicken varieties, the fighting cock (Shamo) and the long-crowing chicken (Naganakidori), have been selectively bred for dramatically different phenotypes. The former has been selected exclusively for aggressiveness and the latter for long crowing with an obedient sitting posture. To understand the particular mechanism behind these genetic changes during domestication, we investigated the degree of genetic differentiation in the aforementioned chickens, focusing on dopamine receptor D2, D3, and D4 genes. We studied other ornamental chickens such as Chabo chickens as a reference for comparison. When genetic differentiation was measured by an index of nucleotide differentiation (N _ST) newly devised in this study, we found that the N_ST value of DRD4 for Shamo (0.072) was distinctively larger than those of the other genes among the three populations, suggesting that aggressiveness has been selected for in Shamo by collecting a variety of single nucleotide polymorphisms. In addition, we found that in DRD4 in Naganakidori, there is a deletion variant of one proline at the 24^th residue in the repeat of nine prolines of exon 1. We thus conclude that artificial selection has operated on these different kinds of genetic variation in the DRD4 genes of Shamo and Naganakidori so strongly that the two domesticated varieties have differentiated to obtain their present opposite features in a relatively short period of time.     

A ∼ 4.1 kb deletion in IRX1 gene upstream is completely associated with rumplessness in Piao chicken

Piao chicken, a Chinese indigenous rumpless chicken breed, lacks pygostyle, caudal vertebra, uropygial gland and tail feathers. The rumplessness in Piao chicken presents an autosomal dominant inheritance pattern. However, the molecular genetic mechanisms underlying the rumplessness in Piao chicken remains unclear. In this study, whole-genome resequencing was performed for 146 individuals from 10 chicken breeds, including 9 tailed chicken breeds and Piao rumpless breed. Tailbone CT scan for Piao chickens and WL chickens, revealed that some Piao chicken tails were normal in number, and for a few Piao chickens tail length and tail bone numbers were between the rumpless and the normal tailed chickens. The results showed that the rumpless phenotype has not been completely fixed in Piao chicken breed. Using selection signature analysis and structural variation detection, we found a 4174 bp deletion located in the upstream region of IRX1 gene on chromosome 2 related to rumpless phenotype. Structural variation genotyping showed that the deletion was present in all 32 rumpless Piao chickens (del/del, wild/del) and absent from all 112 tailed chickens included in the dataset for the other 9 breeds and 2 tailed Piao chickens (wild/wild). In summary, all rumpless Piao chickens tested here carry this deletion mutation, to show a complete linkage association with rumplessness trait. We suggested that the 4174 bp deletion could be causative for rumpless phenotype in Piao chicken since this is the only mutation to show the complete linkage disequilibrium with rumplessness on whole genome level across all of 146 chickens from the 10 breeds. This study could facilitate a better understanding of the genetic characteristics of Piao chicken.     

Heritability and identification of QTLs and underlying candidate genes associated with the architecture of the grapevine cluster (Vitis vinifera L.)

Key message

We have identified 19 QTLs for rachis architecture, a key and complex trait for grapevine production. Fifty out of 1,173 genes underlying these QTLs are candidates to be further explored.

Abstract

In the table grape industry, the rachis architecture has economic and management implications. Therefore, understanding the genetics of this trait is key for its breeding. The aim of this work was to identify genetic determinants of traits associated with the cluster architecture. Characterisations of eight traits was performed on a ‘Ruby Seedless’ × ‘Sultanina’ crossing (F₁: n = 137) during three seasons, with and without gibberellic acid (GA₃) applications. The genotypic effects and the genotype × GA₃ interactions were significant for several traits. Rachis length (rl), lateral shoulder length and node number along the central axis were the most prominent traits. On average, the heritability of these traits was ~71 %, with heritability of rl being 76 % as estimated under different seasons. Quantitative trait loci (QTLs) analyses showed that linkage group 5 (LG₅) and LG₁₈ harboured the largest number of QTLs for these traits. According to the variance explained, the main QTL (corresponding to rl) was found on LG₉. These QTLs were supported mainly by a paternal additive effect and revealed possible pleiotropic effects. Based on the grapevine reference genome, we identified 1,173 genes located under these QTL confidence intervals. Fifty of the 891 annotated genes of this list were selected for their further characterisation because of their possible participation in the rachis architecture. In conclusion, the QTLs detected indicate that these traits and their GA₃ responsiveness have a clear genetic basis. Due to the percentage of the total variance explained, they are good candidates to participate in the genetic determination of the cluster architecture.     

Polymorphisms of the interleukin-15 gene and their associations with fatness and muscle fiber traits in chickens

Interleukin-15 (IL-15) is a cytokine that has been proposed to modulate skeletal muscle and adipose tissue mass. In the present study, an F₂ resource population of Gushi chickens crossed with Anka broilers was used to investigate the genetic effects of the chicken IL-15 gene. Two single nucleotide polymorphisms (SNPs) (g.31224G>A and g.31266T>G) were identified in exon 5 of the IL-15 gene by means of polymerase chain reaction?Crestriction fragment length polymorphism (PCR-RFLP) and DNA sequencing. Associations between the two SNPs and chicken fatness and muscle fiber traits were determined using linkage disequilibrium, haplotype construction, and association analysis. Both of the SNPs were associated with abdominal fat weight, leg muscle fiber diameter, and leg muscle fiber density (p?<?0.05). Haplotypes of the two linked SNPs were associated with abdominal fat weight, fat thickness under the skin, and leg muscle fiber diameter (p?<?0.05). The results suggested that the IL-15 gene might be associated with the causative mutation or the quantitative trait locus (QTL) controlling the fatness traits and muscle fiber traits in chickens.     

Genetic control of chicken heterophil function in advanced intercross lines: associations with novel and with known <Emphasis Type="Italic">Salmonella</Emphasis> resistance loci and a likely mechanism for cell death in extracellular trap production

Heterophils, the avian polymorphonuclear leukocyte and the counterpart of mammalian neutrophils, generate the primary innate response to pathogens in chickens. Heterophil performance against pathogens is associated with host disease resistance, and heterophil gene expression and function are under genetic control. To characterize the genomic basis of heterophil function, heterophils from F₁₃ advanced intercross chicken lines (broiler × Leghorn and broiler × Fayoumi) were assayed for phagocytosis and killing of Salmonella enteritidis, oxidative burst, and extracellular trap production. A whole-genome association analysis of single nucleotide polymorphisms at 57,636 loci identified genomic locations controlling these functional phenotypes. Genomic analysis revealed a significant association of extracellular trap production with the SAL1 locus and the SLC11A1 gene, which have both been previously associated with resistance to S. enteritidis. Fine mapping supports SIVA1 as a candidate gene controlling SAL1-mediated resistance and indicates that the proposed cell-death mechanism associated with extracellular trap production, ETosis, likely functions through the CD27/Siva-1-mediated apoptotic pathway. The SLC11A1 gene was also associated with phagocytosis of S. enteritidis, suggesting that the Slc11a1 protein may play an additional role in immune response beyond depleting metal ions to inhibit intracellular bacterial growth. A region of chromosome 6 with no characterized genes was also associated with extracellular trap production. Further characterization of these novel genes in chickens and other species is needed to understand their role in polymorphonuclear leukocyte function and host resistance to disease.     

Asymmetric introgression between Magnolia stellata and M. salicifolia at a site where the two species grow sympatrically

In order to understand the ongoing evolutionary relationships between species, it is important to elucidate patterns of natural hybridization. In the zone where two species are sympatrically distributed, we examined 274 individuals of Magnolia stellata, Magnolia salicifolia, and their putative hybrids by means of 16 nuclear and three chloroplast microsatellite markers. Hybrid classes of individuals were estimated by admixture analyses. Morphological traits were also investigated for 64 of the 274 individuals. Admixture analyses revealed that 66 of the 274 individuals were classified as hybrids, comprising 17 F₁ and 19 F₂ individuals, 27 backcrosses to M. salicifolia, and 3 individuals of unknown origin. Morphological data from the 64 individuals agreed well with their genetic admixture rates. Spatial locations of F₁ and F₂ hybrids at the study site were intermediate between the two purebred species, indicating that the site preferences of hybrids are intermediate. The occurrences of F₂ and backcross hybrids indicate that F₁ hybrids are fertile. The chloroplast DNA haplotypes of all F₁ hybrids corresponded to those detected in M. salicifolia, so that maternal parents of the F₁ hybrids were all M. salicifolia. Furthermore, no hybrid individuals derived from a backcross to M. stellata were detected. These results suggest that the direction of hybridization and the subsequent introgression have been quite asymmetric and that the introgression occurred from M. stellata into M. salicifolia.     

Genetic architecture controlling variation in grain carotenoid composition and concentrations in two maize populations

Key message

Genetic control of maize grain carotenoid profiles is coordinated through several loci distributed throughout three secondary metabolic pathways, most of which exhibit additive, and more importantly, pleiotropic effects.

Abstract

The genetic basis for the variation in maize grain carotenoid concentrations was investigated in two F_2:3 populations, DEexp × CI7 and A619 × SC55, derived from high total carotenoid and high β-carotene inbred lines. A comparison of grain carotenoid concentrations from population DEexp × CI7 grown in different environments revealed significantly higher concentrations and greater trait variation in samples harvested from a subtropical environment relative to those from a temperate environment. Genotype by environment interactions was significant for most carotenoid traits. Using phenotypic data in additive, environment-specific genetic models, quantitative trait loci (QTL) were identified for absolute and derived carotenoid traits in each population, including those specific to the isomerization of β-carotene. A multivariate approach for these correlated traits was taken, using carotenoid trait principal components (PCs) that jointly accounted for 97 % or more of trait variation. Component loadings for carotenoid PCs were interpreted in the context of known substrate-product relationships within the carotenoid pathway. Importantly, QTL for univariate and multivariate traits were found to cluster in close proximity to map locations of loci involved in methyl-erythritol, isoprenoid and carotenoid metabolism. Several of these genes, including lycopene epsilon cyclase, carotenoid cleavage dioxygenase1 and beta-carotene hydroxylase, were mapped in the segregating populations. These loci exhibited pleiotropic effects on α-branch carotenoids, total carotenoid profile and β-branch carotenoids, respectively. Our results confirm that several QTL are involved in the modification of carotenoid profiles, and suggest genetic targets that could be used for the improvement of total carotenoid and β-carotene in future breeding populations.