Performing Department
Horticulture
Non Technical Summary
Begonias are popular annual bedding and potted flowering plants. The popularity of begonias is based in two factors: (1) plants display a wide range of ornamental phenotypes, not only for flower color and morphology, but also for leaf color and shape, and (2) they are easy to propagate by leaf or stem cuttings. In the US in 2014, the wholesale prices of begonias (15-state total) were $33.3 million for bedding plants sold in flats, $25.3 million for plants sold in pots, and $10.8 million for plants sold in hanging baskets (total $69.4 million). These numbers rank begonias in the top five of the most popular floricultural crops in all three categories (http://www.usda.gov/nass/PUBS/TODAYRPT/floran15.pdf). To growers, the economic value of begonias therefore is comparable to that of some vegetable crops such as cucumbers ($168 million) and pumpkins ($145 million). Virginia is not one of the top fifteen begonia producers in the US. The begonia production is limitedto several nurseries mainly in the Virginia Beach area, so most of the begonias consumed in Virginia come from other states, such as California, Florida and North Carolina. Nevertheless, the development of new varieties could help to increase the production of this crop in the state of Virginia, helping to the diversification of the economy with stable values such as flower production. Additionally begonias are close related species to the cucurbits family with species economically important such as cucumbers (Virginia is the 10th largest US state producer), gourds, melons, pumpkins, squash, watermelons and zucchinis. Begonia is an excellent out-group to understand some of the traits of these commercially important species such as flower sex ratio and flower architecture. We expect that some of the results of the begonia research will be translated in the improvement of these crops.Begonia has a complex breeding history. For example, the wax begonia, one of the most popular types (Begonia semperflorens-cultorum), comes from the domestication of the wild begonia, Begonia cucullata. This species was hybridized at the end of the 19th century with Begonia schmidtiana and probably other unknown Begonia species in the search for new phenotypesalso being one of the first hybrids produced commercially. Later on, the progeny was backcrossed with these two species creating dozens of lines. Other species such as Begonia roezli (1881) and Begonia gracilis (1894) were used in the first stages of breeding to add the brighter color to the flowers and glossy foliage respectively. Spontaneous mutations also have been described such as B. Vernon (1890) with bronzy-red foliage and scarlet flowers. We propose the genetic study of the domestication of wax begonias as an approach to search for genes involved in specific ornamental traits such as flower and leaf color and flower longevity. Additionally, we will search for genes controlling the flower sex ratio and the inflorescence branching as a way to generate knowledge that could be applied to other crops from the closely related Cucurbitaceae family (e.g. cucumbers). The development of new methods for massively-parallel DNA sequencing, such as the Illumina or PacBio sequencing platforms, make it possible to apply whole genome sequencing (WGS), genotyping-by-sequencing (GBS), and RNA-Seq methods to the genetic study of wax begonias in a cost-effective manner. We will use two different complementary approaches to search candidate genes associated with these traits: (1) Quantitative Trait Loci (QTL) analysis to search for correlations between genomic regions and phenotypes using GBS genetic markers; (2) RNA-Seq analysis to search candidate genes based in differential expression between two or more samples. The result intersection between genome location (QTL analysis) and differentially expressed genes (RNA-Seq analysis) will reduce the list of candidate genes to a manageable size for a posterior functional validation.The results of the proposed study will have a favorable impact in the breeding of this species, accelerating the development of more attractive varieties. Additionally, the use of wax begonias as a model for domestication can help us understand the impact of human manipulation on plant species, and how our search for specific phenotypes (e.g. flower sex ratio and inflorescent branching) can modify the genetic information in these species, by selecting defective alleles for genes (e.g. in the flower development pathway as a way to increase the yield in crops).
Animal Health Component
10%
Research Effort Categories
Basic
90%
Applied
10%
Developmental
(N/A)
Goals / Objectives
ObjectivesThe main objective of this project is to produce a genetic description of the process of domestication and breeding in the species B. semperflorens-cultorum (the wax begonias), highlighting possible alleles and/or chromosome regions that have been selected during this process and that can be involved in specific ornamental traits. Specific objectives for this project are:Objective 1: Development of wax begonia reference genome and other genomic resources (e.g. gene expression atlas) to facilitate the use of approaches such as genomic selection in begonia breeding.Objective 2: Genetic characterization of wild and cultivated accessions to evaluate and describe: (a) the genetic diversity, (b) potential genetic bottlenecks that occurred during the breeding history, and (c) wild species introgressions into old and modern cultivars.Objective 3: Development and Quantitative Trait Loci (QTL) analysis of two mapping populations (B. semperflorens-cultorum x B. cucullata var. cucullata and B. semperflorens-cultorum x B. schmidtiana) to detect marker loci linked to begonia agronomical and ornamental traits.Objective 4: Transcriptomic characterization of B. cucullata var. cucullata, B. schmidtiana, four to six mapping population samples and three B. semperflorens cultivars to search for candidate genes involved in begonia agronomical and ornamental traits.
Project Methods
Plant ResourcesWe will use four different sources to get obtain material:Ornamental Plant Genetic Resource at The Ohio State University (OPGR; http://opgc.osu.edu/) with 150 begonia accessions.Royal Botanical Garden Edinburgh (http://www.rbge.org.uk/science/home) with 200 begonia accessions.American Begonia Society (http://www.begonias.org/)with >100 accessions.Different retail companies suchTaylor Greenhouses (http://www.taylorgreenhouses.com/).We will grow the plants, when needed, at the Virginia Tech greenhouses (http://www.hort.vt.edu/greenhouse/). Seeds germinate in 14 - 21 days, and require between four and six months to produce flowers (one or two generations can be produced per year).For objectives 1 and 2,DNA will be extracted from seedlingsor leaves. For objective 3, we will grow between three and six progenitor plants for the mapping population from cuttings supplied by OPGR. We are planning to obtain F2 seeds in years 1 and 2. We will grow 150 plants per F2 mapping population for the QTL analysis for years 2 and 3. We will reproduce the different lines using vegetative propagation. For objective 4, we will grow 12-15 plants reproduced by cuttings from single plants of each accession during years 1 and 2, except for the bulk segregation analysis, where we will need to have segregating phenotypes for the mapping population, so we will grow the plants during years 3 and 4.DNA will be extracted using the DNeasy Plant Mini Kit® from Qiagen. RNA will be extracted using Qiagen RNeasy Plant Mini Kit®. DNA concentration and integrity will be evaluated with a Qubit® instrument purchased from Thermo Fisher Scientific and a Bioanalyzer 2100 purchased from Agilent, respectively.Objective 1: Development of a wax begonia reference genomeWax begonias were developed through the hybridization of B. cucullata and B. schmidtiana with contributions from B. roezli and B. gracilis 21, although it is probable that some lines have been backcrossed several times with B. cucullata. Because of its hybrid nature, our approach will be focused on one species, B. cucullata (estimated genome size of 293 Mb), and completing the dataset with the resequencing of the three other species (B. schmidtiana, B. roezli, and B. gracilis). For this group, we only have an estimated genome size for B. schmidtiana (372 Mb), but because the average genome size in the genus Begonia is 560 Mb 16, we don't expect to encounter major problems in resequencing the other two species. The design for the whole genome sequencing of these species will be:B. cucullata, hybrid assembly with Illumina and PacBio reads.100X (~30 Gb) of 2x150 Illumina pair ends (insert size 300 bp).30X (~9 Gb) of PacBio reads.B. schmidtiana, B. roezli andB. gracilisIllumina reads.60X (~24 Gb) of 2x150 Illumina pair ends (insert size 300 bp) / each species.Reads will be processed with Fastq-mcf. Illumina reads will be corrected with Musket 29 and PacBio reads with LoRDEC 30. We will use SOAPdenovo2 to assemble the Illumina reads 31, and Sprai and the Celera Assembler 32 to assemble the PacBio reads. Both assemblies will then be evaluated, and the best assembly will be used as the base, using the other read set to re-scaffold the assembly with SSPACE 33 and to fill the gaps with GapCloser 31 or PBJelly 34.We will annotate the different genomes using Maker-P 35, and also generate a reference transcriptome as evidence-based support for the annotation. RNA extracted from roots, leaves, stems, male flowers, and female flowers will be sequenced on an Illumina instrument (2x150 bp, 300 bp insert size, 10-20 million reads/sample). Trinity will be used to assemble the reads 36. All the programs will be run on a 64-core server with 256 Gb of RAM and 7 Tb of disk space in the Bombarely laboratory.Objective 2: Genetic characterization of wild and cultivated accessionsWe will use Genotyping-By-Sequencing (GBS) to characterize genetically wild and cultivated begonia accessions. GBS is a relatively cost-effective approach to rapidly generate thousand of polymorphic molecular markers 37. We will use a 5 bp restriction site enzyme (ApeKI), sequencing 0.5-5 million of reads per sample (Illumina; 1x100). Reads will be processed with Fastq-mcf and mapped to the reference genome using Bowtie2 38 or BWA 39 depending of their performance. Sam files will be processed with Samtools 40. DNA sequence polymorphisms will be detected using GATK 41 or FreeBayes 42 with a minimum read depth of 10 and a minimum mapping score of 20. We will estimate the genetic diversity of each sample using a custom Perl script. The population structure will be analyzed using Structure 43 and FineStructure 44. Because several species will be used, we will analyze their phylogenetic relations using SNAPP 45.Objective 3: QTL analysis for begonia agronomical/ornamental traitsThere are several agronomical/ornamental traits that we will analyze in this project and that are related to the domestication and breeding of wax begonias. The wild B. cucullata (B. semperflorens) is a semi-succulent perennial herb with white or pale pink flowers. The wild B. schmidtiana is a compact shrub. It produces many clusters of small white or pinkish flowers 21. There are a number of important traits that have been selected in the last hundred years during the breeding of B. semperflorens-cultorum, such as (1) leaf color (medium, olive or dark green to bronze red or maroon), (2) leaf shape (ovate to broad-ovate), (3) flower color (white, pink, and deep red), (4) flower shape (simple or double), (5) flower longevity (longer-lived), and (6) compact plant size. Additionally we will characterize other traits applicable to other cucurbits such as (7) flower sex ration and (8) inflorescence branching. Our approach to find genomic regions responsible for these traits is to generate two F2 mapping populations derived from a wild species (B. cucullata or B. schmidtiana) crossed with a cultivated accession (B. semperflorens-cultorum with red double flowers and bronze-red leaves). We will analyze between 80 and 120 F2 lines (3 plants/each) for the traits described previously. Phenotypes such as flower and leaf color and morphology will be measured by scanning 5-10 flowers/leaves of each plant. We will also collect other phenotypic data such as number of flowers and leaves, size of the plant and longevity of the flower. One plant per line will be genotyped using GBS. The GBS reads will be processed as was described in Objective 1. QTL analysis will be performed using R/QTL 46.Objective 4: Transcriptomic characterization of begonia accessionsWe will use RNA-seq to study the gene expression in several begonia accessions. We expect that in combination with the QTL results, RNA-seq willelucidate the most relevant genes selected during the begonia breeding. We will use two different sets: (1) two wild and three or four cultivated accessions with different phenotypes with three biological replicates per sample and three plants per biological replicate, and (2) bulks of plants with similar phenotypes from different F2 mapping population lines (two bulks of 5-10 lines for each trait, three traits to analyze: flower color, single/double flower, leaf color). Eight organs/developmental stage will be analyzed: young leaf, mature leaf, male flower bud, female flower bud, male tepals, female tepals, male stamens, and female pistils. Each of the samples will be subjected to Illumina sequencing (1x100 nucleotide reads, 5-10 million reads/sample). The reads will be processed as described previously in the first objective. Reads will be mapped with Tophat2 47 or HISAT 48. The differential expression analysis will be performed using Cufflinks 49 or Stringtie 50. Some of the downstream analyses will involve cluster and Gene Set Enrichment Analysis (GSEA) using several R and Bioconductor software packages such as CummeRBund51, Ballgown52 and TopGo53.