Genomic prediction contributing to a promising global strategy to turbocharge gene banks.

The 7.4 million plant accessions in gene banks are largely underutilized due to various resource constraints, but current genomic and analytic technologies are enabling us to mine this natural heritage. Here we report a proof-of-concept study to integrate genomic prediction into a broad germplasm evaluation process. First, a set of 962 biomass sorghum accessions were chosen as a reference set by germplasm curators. With high throughput genotyping-by-sequencing (GBS), we genetically characterized this reference set with 340,496 single nucleotide polymorphisms (SNPs). A set of 299 accessions was selected as the training set to represent the overall diversity of the reference set, and we phenotypically characterized the training set for biomass yield and other related traits. Cross-validation with multiple analytical methods using the data of this training set indicated high prediction accuracy for biomass yield. Empirical experiments with a 200-accession validation set chosen from the reference set confirmed high prediction accuracy. The potential to apply the prediction model to broader genetic contexts was also examined with an independent population. Detailed analyses on prediction reliability provided new insights into strategy optimization. The success of this project illustrates that a global, cost-effective strategy may be designed to assess the vast amount of valuable germplasm archived in 1,750 gene banks.

Multi-Phase US Spread and Habitat Switching of a Post-Columbian Invasive, Sorghum halepense.

Johnsongrass (Sorghum halepense) is a striking example of a post-Columbian founder event. This natural experiment within ecological time-scales provides a unique opportunity for understanding patterns of continent-wide genetic diversity following range expansion. Microsatellite markers were used for population genetic analyses including leaf-optimized Neighbor-Joining tree, pairwise FST, mismatch analysis, principle coordinate analysis, Tajima's D, Fu's F and Bayesian clusterings of population structure. Evidence indicates two geographically distant introductions of divergent genotypes, which spread across much of the US in <200 years. Based on geophylogeny, gene flow patterns can be inferred to have involved five phases. Centers of genetic diversity have shifted from two introduction sites separated by ~2000 miles toward the middle of the range, consistent with admixture between genotypes from the respective introductions. Genotyping provides evidence for a 'habitat switch' from agricultural to non-agricultural systems and may contribute to both Johnsongrass ubiquity and aggressiveness. Despite lower and more structured diversity at the invasion front, Johnsongrass continues to advance northward into cooler and drier habitats. Association genetic approaches may permit identification of alleles contributing to the habitat switch or other traits important to weed/invasive management and/or crop improvement.

Oil, fatty acid, flavonoid, and resveratrol content variability and FAD2A functional SNP genotypes in the U.S. peanut mini-core collection.

Peanut seeds contain high amounts of oil and protein as well as some useful bioactive phytochemicals which can contribute to human health. The U.S. peanut mini-core collection is an important genetic resource for improving seed quality and developing new cultivars. Variability of seed chemical composition within the mini-core was evaluated from freshly harvested seeds for two years. Oil, fatty acid composition, and flavonoid/resveratrol content were quantified by NMR, GC, and HPLC, respectively. Significant variability was detected in seed chemical composition among accessions and botanical varieties. Accessions were further genotyped with a functional SNP marker from the FAD2A gene using real-time PCR and classified into three genotypes with significantly different O/L ratios: wild type (G/G with a low O/L ratio <1.7), heterozygote (G/A with O/L ratio >1.4 but <1.7), and mutant (A/A with a high O/L ratio >1.7). The results from real-time PCR genotyping and GC fatty acid analysis were consistent. Accessions with high amounts of oil, quercetin, high seed weight, and O/L ratio were identified. The results from this study may be useful not only for peanut breeders, food processors, and product consumers to select suitable accessions or cultivars but also for curators to potentially expand the mini-core collection.

Assessment of oil content and fatty acid composition variability in two economically important Hibiscus species.

The Hibiscus genus encompasses more than 300 species, but kenaf (Hibiscus cannabinus L.) and roselle (Hibiscus sabdariffa L.) are the two most economically important species within the genus. Seeds from these two Hibiscus species contain a relatively high amount of oil with two unusual fatty acids: dihydrosterculic and vernolic acids. The fatty acid composition in the oil can directly affect oil quality and its utilization. However, the variability in oil content and fatty acid composition for these two species is unclear. For these two species, 329 available accessions were acquired from the USDA germplasm collection. Their oil content and fatty acid composition were determined by nuclear magnetic resonance (NMR) and gas chromatography (GC), respectively. Using NMR and GC analyses, we found that Hibiscus seeds on average contained 18% oil and seed oil was composed of six major fatty acids (each >1%) and seven minor fatty acids (each <1%). Hibiscus cannabinus seeds contained significantly higher amounts of oil (18.14%), palmitic (20.75%), oleic (28.91%), vernolic acids (VA, 4.16%), and significantly lower amounts of stearic (3.96%), linoleic (39.49%), and dihydrosterculic acids (DHSA, 1.08%) than H. sabdariffa seeds (17.35%, 18.52%, 25.16%, 3.52%, 4.31%, 44.72%, and 1.57%, respectively). For edible oils, a higher oleic/linoleic (O/L) ratio and lower level of DHSA are preferred, and for industrial oils a high level of VA is preferred. Our results indicate that seeds from H. cannabinus may be of higher quality than H. sabdariffa seeds for these reasons. Significant variability in oil content and major fatty acids was also detected within both species. The variability in oil content and fatty acid composition revealed from this study will be useful for exploring seed utilization and developing new cultivars in these Hibiscus species.

Screening of the entire USDA castor germplasm collection for oil content and fatty acid composition for optimum biodiesel production.

Castor has tremendous potential as a feedstock for biodiesel production. The oil content and fatty acid composition in castor seed are important factors determining the price for production and affecting the key fuel properties of biodiesel. There are 1033 available castor accessions collected or donated from 48 countries worldwide in the USDA germplasm collection. The entire castor collection was screened for oil content and fatty acid composition by nuclear magnetic resonance (NMR) and gas chromatography (GC), respectively. Castor seeds on the average contain 48.2% oil with significant variability ranging from 37.2 to 60.6%. Methyl esters were prepared from castor seed by alkaline transmethylation. GC analysis of methyl esters confirmed that castor oil was composed primarily of eight fatty acids: 1.48% palmitic (C16:0), 1.58% stearic (C18:0), 4.41% oleic (C18:1), 6.42% linoleic (C18:2), 0.68% linolenic (C18:3), 0.45% gadoleic (C20:1), 84.51% ricinoleic (C18:1-1OH), and 0.47% dihydroxystearic (C18:0-2OH) acids. Significant variability in fatty acid composition was detected among castor accessions. Ricinoleic acid (RA) was positively correlated with dihydroxystearic acid (DHSA) but highly negatively correlated with the five other fatty acids except linolenic acid. The results for oil content and fatty acid composition obtained from this study will be useful for end-users to explore castor germplasm for biodiesel production.

Population structure and marker-trait association analysis of the US peanut (Arachis hypogaea L.) mini-core collection.

Peanut (Arachis hypogaea L.) is one of the most important oilseed and nutritional crops in the world. To efficiently utilize the germplasm collection, a peanut mini-core containing 112 accessions was established in the United States. To determine the population structure and its impact on marker-trait association, this mini-core collection was assessed by genotyping 94 accessions with 81 SSR markers and two functional SNP markers from fatty acid desaturase 2 (FAD2). Seed quality traits (including oil content, fatty acid composition, flavonoids, and resveratrol) were obtained through nuclear magnetic resonance (NMR), gas chromatography (GC), and high-performance liquid chromatography (HPLC) analysis. Genetic diversity and population structure analysis identified four major subpopulations that are related to four botanical varieties. Model comparison with different levels of population structure and kinship control was conducted for each trait and association analyses with the selected models verified that the functional SNP from the FAD2A gene is significantly associated with oleic acid (C18:1), linoleic acid (C18:2), and oleic-to-linoleic (O/L) ratio across this diverse collection. Even though the allele distribution of FAD2A was structured among the four subpopulations, the effect of FAD2A gene remained significant after controlling population structure and had a likelihood-ratio-based R ( 2 ) (R ( LR ) ( 2 ) ) value of 0.05 (oleic acid), 0.09 (linoleic acid), and 0.07 (O/L ratio) because the FAD2A alleles were not completely fixed within subpopulations. Our genetic analysis demonstrated that this peanut mini-core panel is suitable for association mapping. Phenotypic characterization for seed quality traits and association testing of the functional SNP from FAD2A gene provided information for further breeding and genetic research.

Genetic diversity and population structure analysis of accessions in the US historic sweet sorghum collection.

Sweet sorghum has the potential to become a versatile feedstock for large-scale bioenergy production given its sugar from stem juice, cellulose/hemicellulose from stalks, and starch from grain. However, for researchers to maximize its feedstock potential a first step includes additional evaluations of the 2,180 accessions with varied origins in the US historic sweet sorghum collection. To assess genetic diversity of this collection for bioenergy breeding and population structure for association mapping, we selected 96 accessions and genotyped them with 95 simple sequence repeat markers. Subsequent genetic diversity and population structure analysis methods identified four subpopulations in this panel, which correlated well with the geographic locations where these accessions originated or were collected. Model comparisons for three quantitative traits revealed different levels of population structure effects on flowering time, plant height, and brix. Our results suggest that diverse germplasm accessions curated from different geographical regions should be considered for plant breeding programs to develop sweet sorghum cultivars or hybrids, and that this sweet sorghum panel can be further explored for association mapping.

Discovering and verifying DNA polymorphisms in a mung bean [V. radiata (L.) R. Wilczek] collection by EcoTILLING and sequencing.

BACKGROUND: Vigna radiata, which is classified in the family Fabaceae, is an important economic crop and a dietary staple in many developing countries. The species radiata can be further subdivided into varieties of which the variety sublobata is currently acknowledged as the putative progenitor of radiata. EcoTILLING was employed to identify single nucleotide polymorphisms (SNPs) and small insertions/deletions (INDELS) in a collection of Vigna radiata accessions. FINDINGS: A total of 157 DNA polymorphisms in the collection were produced from ten primer sets when using V. radiata var. sublobata as the reference. The majority of polymorphisms detected were found in putative introns. The banding patterns varied from simple to complex as the number of DNA polymorphisms between two pooled samples increased. Numerous SNPs and INDELS ranging from 4-24 and 1-6, respectively, were detected in all fragments when pooling V. radiata var. sublobata with V. radiata var. radiata. On the other hand, when accessions of V. radiata var. radiata were mixed together and digested with CEL I relatively few SNPs and no INDELS were detected. CONCLUSION: EcoTILLING was utilized to identify polymorphisms in a collection of mung bean, which previously showed limited molecular genetic diversity and limited morphological diversity in the flowers and pod descriptors. Overall, EcoTILLING proved to be a powerful genetic analysis tool providing the rapid identification of naturally occurring variation.

Phylogenetic relationships and genetic diversity of the USDA Vigna germplasm collection revealed by gene-derived markers and sequencing.

Phylogenetic relationships in the USDA Vigna germplasm collection are somewhat unclear and their genetic diversity has not been measured empirically. To reveal interspecific phylogenetic relationships and assess their genetic diversity, 48 accessions representing 12 Vigna species were selected, and 30 gene-derived markers from legumes were employed. Some high-quality amplicons were sequenced. Indels (insertion/deletions) were discovered from the sequence alignments that were specific identifiers for some Vigna species. With regard to revealing polymorphisms, intron-spanning markers were more effective than exon-derived markers. These gene-derived markers were more successful in revealing interspecific polymorphisms than intraspecific polymorphisms at both the DNA fragment and sequence levels. Two different dendrograms were generated from DNA fragment data and sequence data, respectively. The results from these two dendrograms supported each other and showed similar phylogenetic relationships among the Vigna species investigated. The accessions clustered into four main groups and 13 subgroups. Each subgroup represents a subgenus or a species. Phylogenetic analysis revealed that an accession might be misclassified in our collection. The putative misclassified accession was further supported by seed morphology. Limited intraspecific genetic diversity was revealed by these gene-derived markers and/or sequences. The USDA Vigna germplasm collection currently consists of multiple species with many accessions further classified into specific subspecies, but very few subspecies of the total subspecies available exist within the collection. Based on our results, more attention should be paid to the subspecies, wild forms and/or botanical varieties for future curation in order to expand the genetic diversity of Vigna germplasm in the USDA collection.

Genetic diversity of cultivated and wild-type peanuts evaluated with M13-tailed SSR markers and sequencing.

Thirty-one genomic SSR markers with a M13 tail attached were used to assess the genetic diversity of the peanut mini core collection. The M13-tailed method was effective in discriminating almost all the cultivated and wild accessions. A total of 477 alleles were detected with an average of 15.4 alleles per locus. The mean polymorphic information content (PIC) score was 0.687. The cultivated peanut (Arachis hypogaea L.) mini core produced a total of 312 alleles with an average of 10.1 alleles per locus. A neighbour-joining tree was constructed to determine the interspecific and intraspecific relationships in this data set. Almost all the peanut accessions in this data set classified into subspecies and botanical varieties such as subsp. hypogaea var. hypogaea, subsp. fastigiata var. fastigiata, and subsp. fastigiata var. vulgaris clustered with other accessions with the same classification, which lends further support to their current taxonomy. Alleles were sequenced from one of the SSR markers used in this study, which demonstrated that the repeat motif is conserved when transferring the marker across species borders. This study allowed the examination of the diversity and phylogenetic relationships in the peanut mini core which has not been previously reported.