Next-generation transcriptome sequencing, SNP discovery and validation in four market classes of peanut, Arachis hypogaea L.

Single-nucleotide polymorphisms, which can be identified in the thousands or millions from comparisons of transcriptome or genome sequences, are ideally suited for making high-resolution genetic maps, investigating population evolutionary history, and discovering marker-trait linkages. Despite significant results from their use in human genetics, progress in identification and use in plants, and particularly polyploid plants, has lagged. As part of a long-term project to identify and use SNPs suitable for these purposes in cultivated peanut, which is tetraploid, we generated transcriptome sequences of four peanut cultivars, namely OLin, New Mexico Valencia C, Tamrun OL07 and Jupiter, which represent the four major market classes of peanut grown in the world, and which are important economically to the US southwest peanut growing region. CopyDNA libraries of each genotype were used to generate 2 × 54 paired-end reads using an Illumina GAIIx sequencer. Raw reads were mapped to a custom reference consisting of Tifrunner 454 sequences plus peanut ESTs in GenBank, compromising 43,108 contigs; 263,840 SNP and indel variants were identified among four genotypes compared to the reference. A subset of 6 variants was assayed across 24 genotypes representing four market types using KASP chemistry to assess the criteria for SNP selection. Results demonstrated that transcriptome sequencing can identify SNPs usable as selectable DNA-based markers in complex polyploid species such as peanut. Criteria for effective use of SNPs as markers are discussed in this context.

Use of Targeted Amplicon Sequencing in Peanut to Generate Allele Information on Allotetraploid Sub-Genomes.

The use of molecular markers in plant breeding has become a routine practice, but the cost per accession can be a hindrance to the routine use of Quantitative Trait Loci (QTL) identification in breeding programs. In this study, we demonstrate the use of targeted re-sequencing as a proof of concept of a cost-effective approach to retrieve highly informative allele information, as well as develop a bioinformatics strategy to capture the genome-specific information of a polyploid species. SNPs were identified from alignment of raw transcriptome reads (2 × 50 bp) to a synthetic tetraploid genome using BWA followed by a GATK pipeline. Regions containing high polymorphic SNPs in both A genome and B genomes were selected as targets for the resequencing study. Targets were amplified using multiplex PCR followed by sequencing on an Illumina HiSeq. Eighty-one percent of the SNP calls in diploids and 68% of the SNP calls in tetraploids were confirmed. These results were also confirmed by KASP validation. Based on this study, we find that targeted resequencing technologies have potential for obtaining maximum allele information in allopolyploids at reduced cost.

Effect of the high-oleic trait on roasted peanut flavor in backcross-derived breeding lines.

The high-oleic trait of peanut (Arachis hypogaea L.) has been suggested to have a positive impact on the roasted peanut sensory attribute. A series of lines derived by backcrossing the high-oleic trait into several existing cultivars were compared with their parent cultivars at locations in Florida, Georgia, North Carolina, and Texas. Breeders grew their high-oleic lines and parents in three-replicate tests at one or two locations. The Florida high-oleic line F435-2-3-B-2-1-b4-B-B-3-b3-b3-1-B was grown at each location. The test included normal- and high-oleic variants of F435, GK 7, NC 7, NC 9, Sunrunner, Tamrun 96, and Tamspan 90. Sound-mature kernel samples were roasted, ground into paste, and evaluated by a sensory panel using a 14-point flavor intensity unit (fiu) scale. Background genotype had an effect (P < 0.01) on the heritable sensory attributes roasted peanut, sweet, and bitter. Oleate level had a positive effect on roasted peanut intensity, increasing it by 0.3 fiu averaged across all seven background genotypes. However, the magnitude of improvement varied across background genotypes. The high-oleic trait had no effect or increased the intensity of the roasted peanut attribute in each background genotype. The increase was greatest in Tamrun 96 (+0.6 fiu, P < 0.05) and Spanish genotypes Tamspan 90 (+0.4 fiu, P < 0.05) and F435 (+0.4 fiu, P < 0.10). A change of 0.5 fiu or more should be perceptible to consumers. Interaction between oleate level and background genotype was detected for sweet (P < 0.10) and bitter (P < 0.01) attributes. The trait had an increasing effect on the bitter attribute only in the background genotype of Tamspan 90 (+0.7 fiu, P < 0.01). There was a nonsignificant increase in bitterness in the other Spanish background genotype, F435. Changes in bitterness in runner- and Virginia-type backgrounds were close to zero. Incorporation of the high-oleic trait into peanut cultivars is likely to improve the intensity of the roasted peanut attribute, but it may also increase the bitter attribute in Spanish genotypes.

Identification of QTLs for pod and kernel traits in cultivated peanut by bulked segregant analysis

Bulked segregant analysis was used to identify simple sequence repeat (SSR) markers associated with pod and kernel traits in cultivated peanut, to permit rapid selection of superior quality genotypes in the breeding program. SSR markers linked to pod and kernel traits were identified in two DNA pools (high and low), which were established using selected F2:6 recombinant individuals resulting from a cultivated cross between a runner (Tamrun OL01) and a Spanish (BSS 56) peanut. To identify quantitative trait loci (QTLs) for pod and kernel-related traits, parents were screened initially with 112 SSR primer pairs. The survey revealed 8.9 percent polymorphism between parents. Of ten SSR primer pairs distinguishing the parents, five (PM375, PM36, PM45, pPGPseq8D9, and Ah-041) were associated with differences between bulks for seed length, pod length, number of pods per plant, 100-seed weight, maturity, or oil content. Association was confirmed by analysis of segregation among 88 F2:6 individuals in the RIL population. Phenotypic means associated with markers for three traits differed by more than 40 percent, indicating the presence of QTLs with major effects for number of pods per plant, plant weight, and pod maturity. The SSR markers can be used for marker assisted selection for quality and yield improvement in peanut. To the best of our knowledge, this is the first report on the identification of SSR markers linked to pod - and kernel- related traits in cultivated peanut.