389

RosBREED: Enabling Marker-Assisted Breeding in Rosaceae A. Iezzoni1, C. Weebadde1, J. Luby2, Chengyan Yue2, 3, E. van de Weg4, G. Fazio5, D. Main6, C.P. Peace6, N.V. Bassil7 and J. McFerson8 1Department of Horticulture, Michigan State University, East Lansing, MI, USA 2Department of Horticultural Science, University of Minnesota, St. Paul, MN, USA 3Department of Applied Economics, University of Minnesota, St. Paul, MN, USA 4Plant Breeding – Wageningen-UR, The Netherlands 5USDA-ARS Cornell University, Geneva, NY, USA 6Department of Horticulture & Landscape Architecture, Washington State University, Pullman, WA, USA

7USDA-ARS National Clonal Germplasm Repository, Corvallis, OR, USA 8Washington Tree Fruit Research Commission, Wenatchee, WA, USA Keywords: MAB pipeline, breeding information management system (BIMS), pedigree-

based analysis (PBA) Abstract

Genomics research has not yet been translated into routine practical application in breeding Rosaceae fruit crops (peach, apple, strawberry, cherry, apricot, pear, raspberry, etc.). Through dedicated efforts of many researchers worldwide, a wealth of genomics resources has accumulated, including EST libraries, genetic and physical maps, QTLs, and whole genome sequences. The potential of genomics approaches to enhance crop improvement, particularly through marker-assisted breeding (MAB), is enormous, but unfulfilled. The U.S. Rosaceae genomics, genetics, and breeding community, with strong international involvement, has united behind the goal of translational genomics and collaborated on the development of large-scale USDA grant proposals. RosBREED, funded for four years from September 2009, incorporates eight teams (Breeding, Socio-Economics, Pedigree-Based Analysis, Breeding Information Management System, Genomics, Genotyping, MAB Pipeline, and Extension) in a transdisciplinary framework that involves significant educational and outreach activities and stakeholder participation. Objectives are to (1) enhance the likelihood of new cultivar adoption, enlarge market potential, and increase consumption of Rosaceae fruits with socio-economics knowledge objectively used in breeding decisions; (2) establish a sustainable technical infrastructure for an efficient MAB Pipeline in Rosaceae; (3) integrate breeding and genomics resources with a standardized breeding information management system incorporating Pedigree-Based Analysis; (4) implement MAB in core RosBREED breeding programs with a common focus on fruit quality traits; and (5) enhance sustainability of cultivar development with MAB technology transfer to current and future U.S. Rosaceae breeders and engagement of key stakeholder groups. Coordinated action is now required to make RosBREED a reality and fulfill the promise of genomics.

INTRODUCTION

Fresh and processed products derived from the Rosaceae plant family (peach, apple, strawberry, cherry, apricot, pear, raspberry, etc.) provide vital contributions to human nutrition, health, and well-being, and collectively their production constitutes the economic backbone of many rural communities in the U.S. Although current domestic production value of these crops is over $7 billion, and global per-capita production and consumption is expanding in both domestic and export markets, the U.S. Rosaceae crop industries face numerous limitations to profitability and sustainability. Overcoming these limitations requires rapid development and deployment of new cultivars with improved characteristics that meet dynamic industry, market, and consumer preferences. The improvement of cultivars by targeted application of genomics research in breeding

Proc. IS on Molecular Markers in Horticulture Eds.: N.V. Bassil and R. Martin Acta Hort. 859, ISHS 2010

 390

programs can exploit the extraordinary diversity of Rosaceae species. Accelerating and increasing efficiency of cultivar development is required, to match consumer demand within constraints of the production and marketing system.

The RosBREED mission and objectives evolved through a process of active engagement among stakeholders, along with development of the U.S. and international Rosaceae research community. Some key outcomes are: creation of the Genome Database for Rosaceae (GDR; http://www.bioinfo.wsu.edu/gdr/); establishment of the National Tree Fruit Technology Roadmap, where genomics, genetics, and breeding was identified as a top priority (http://www.treefruitresearch.com/technology-roadmap/); formation of the U.S. Rosaceae Genomics, Genetics and Breeding Executive Committee (RosEXEC; http://www.bioinfo.wsu.edu/gdr/community/rosexec/index.php); release of the U.S. Rosaceae Genomics, Genetics, and Breeding Initiative White Paper (http://www.bioinfo. wsu.edu/gdr/community/rosexec/documents/RosWP_March_2006.doc); and the Rosaceae Specialty Crops Planning Workshop (http://www.csrees.usda.gov/business/reporting/ stakeholder/pdfs/rosaceae_workshop.pdf) that led to the first community-wide research proposal, RosCAP, submitted in early 2008. While these activities have increased awareness among stakeholder communities of the importance of genomics and genetics to plant breeding, there remains some confusion about genomic approaches and skepticism about the value and return on investment of MAB.

Rosaceae genetics and genomics information, and associated enabling technologies, are developing rapidly, leading to numerous discoveries with potential application (Shulaev et al., 2008; Folta and Gardiner, 2009). In 2009, not only are genetic linkage maps available for all major crop species, but also the apple, peach and strawberry genomes are being sequenced, and hundreds of QTLs and major genes have been identified.

Despite these advances in Rosaceae genetics and genomics, MAB implementation has been minimal in U.S. Rosaceae breeding programs. Most rosaceous crops are clonally propagated for commercial production (often as a scion-rootstock composite plant) and have long generation times before the commercial product (fruit, nut, flower) can be phenotyped. For all of these crops, evaluation of individual plants requires intensive greenhouse and/or extensive field space; thus, phenotyping individual plants is expensive. Application of genetic markers is an obvious opportunity, but has not been implemented due to significant technical, economic, logistical, informational, and sociological barriers. We must overcome these barriers to enable sustainable MAB in rosaceous crop breeding.

The RosBREED project proposal (“Ros” signifying a Rosaceae genomics, genetics, and breeding community initiative, and “BREED” indicating the core focus on breeding programs) was submitted to USDA-ARS CSREES’s Specialty Crops Research Initiative funding opportunity in August 2008, and resubmitted in a revised and expanded form in April 2009. RosBREED is now funded for four years from September 2009 (www.rosbreed.org).

OBJECTIVES OF THE PROJECT

RosBREED’s vision is that integration of modern genomics tools with traditional breeding approaches will transform crop improvement in Rosaceae, significantly improving profitability and sustainability of U.S. rosaceous crop industries and contributing to increased consumption and enjoyment of these fruit, nut and floral products. Our mission is to create a dynamic, sustained program in research, infrastructure establishment, training, and outreach for developing and applying MAB based on improved knowledge of industry value and consumer preferences to accelerate and increase the efficiency of rosaceous cultivar release and successful cultivar adoption.

Objectives 1. Increase the likelihood of new cultivar adoption, enlarge market potential, and increase

consumption of rosaceous fruits by using socio-economics approaches to enhance knowledge of industry values and consumer preferences for fruit traits to inform

  391

breeding. 2. Establish sustainable technical infrastructure for an efficient MAB Pipeline in

Rosaceae, including crop-specific SNP genome scan platforms for breeding-relevant germplasm exploiting the shared ancestry of Rosaceae crops.

3. Integrate breeding and genomics resources by establishing a user-friendly U.S.-wide standardized statistical framework and breeding information management system (BIMS).

4. Implement MAB in core RosBREED programs with a common focus on fruit quality traits.

5. Enhance sustainability of cultivar development by transferring MAB technologies to the public and private community of U.S. Rosaceae breeders through training current and future breeders, while better engaging the production, processing, and marketing sectors, allied scientists, and consumers.

COMPONENTS OF THE PROJECT

The project has five components, each corresponding to a project objective (Fig. 1). Project management is shared among eight teams: Breeding, Socio-Economics, Pedigree-Based Analysis, Breeding Information Management System, Genomics, Genotyping, MAB Pipeline, and Extension, each with a Team Leader, under the leadership of the Project Director (A. Iezzoni).

Trait and Market Segment Breeding Target Establishment (1)

Economic weights of traits will be determined from information on trait values obtained from breeders and key supply chain members. Information will be gained with a mix of experimental research methods, including one-on-one interviews, telephone surveys and experimental auctions. Five crops will be included: apple, peach, strawberry, sweet cherry, and tart cherry. Apple will be used as the model to test auction-based economic approaches and one-on-one interviews involving fruit tasting. Ultimately, a transparent interactive process for determining economic values of production traits and fruit quality traits and setting breeding selection targets will be established. This process will allow Rosaceae breeders to routinely use (and update) economic weighting of production and fruit quality traits to optimize their set of selection targets. Rosaceae producer/processor, marketing group, trade organization, and public awareness of breeding, including efficiencies gained by optimizing selection targets and strategies, is expected to increase as a result of their participation in defining economic weights and through wide dissemination of findings.

Genome Scans and Comparative Genomics (2)

We are developing SNP-based genome-scanning capability for apple, peach, strawberry, and cherry at a resolution of at least one polymorphic SNP marker every 5 cM in any random cross, for use in Pedigree-Based Analysis (Van de Weg et al., 2004; Bink et al., 2007). Genes targeted for SNP development will be those related to fruit quality and shared among species. Specific cases of synteny will be exploited to leverage ancestral relatedness at the genomic level.

SNP markers for genome scans will have two sources: validated SNP markers generated from existing projects and SNPs detected de novo in whole genome sequences using synteny and orthology information. Markers will be genetically and physically mapped prior to their inclusion in the final genome-spanning sets of SNPs for each crop. The SNP markers for apple, peach, strawberry, and sweet and tart cherry will be major sustainable deliverables from this project for the Rosaceae genomics, genetics, and breeding research community; therefore, we will ensure open access to this information in the GDR.

 392

Pedigree-Based Breeding Information Management System (3) A common Breeding Information Management System (BIMS) aligned with the

Pedigree-Based Analysis (PBA) approach will be established for streamlined collection, archiving, analysis, and interpretation of integrated breeding and genomics data. Statistical programs, component databases, and decision-support tools will be incorporated into this broader BIMS framework, which will be hosted on the GDR. The 12 core participating breeding programs of RosBREED will be the alpha-testers for BIMS in 2010-2011, ensuring practicality for the wider U.S. Rosaceae breeding community to which BIMS will be delivered following beta-testing in 2011-2012.

PBA is an existing statistical framework with associated software designed to identify, validate, and use QTL information from pedigree-linked individuals to inform breeding decision-making. U.S. Rosaceae breeders and supporting personnel will be professionally trained in its use and software optimized for the specific needs and opportunities of U.S. Rosaceae breeding programs. Biological and genetic complexities to be considered include polyploidy, inter-locus interactions, and genotype×environment interactions.

One output of this component will be Crop Reference databases of phenotypic, genotypic, and pedigree data – both raw and integrated via PBA – permanently archived on the GDR for the large germplasm sets spanning U.S. breeding programs for apple, peach, strawberry, and cherry. This database will be a powerful and immensely valuable public resource for marker-trait association discovery and validation in Rosaceae, with genome-wide SNP and high-impact trait data collected and analyzed for 480 reference individuals of each crop. Functional genotype-based breeding values for each Crop Reference Set individual will be described. Hundreds of marker-trait associations will be dissected to resolutions <5 cM intervals. In many cases the actual genes likely to be underlying the QTL will be apparent.

Individual breeding programs will establish individual databases for their own pedigree germplasm and traits of interest that build on the crop reference data. The flexible design of PBA and additional power obtained with more data encourages the inclusion of all breeding data for all germplasm – past, present, and future.

Online community-driven websites being developed include: a one-stop warehouse of reported marker-trait associations for Rosaceae breeders; a database of economically weighted fruit quality and other high-impact traits for apple, peach, strawberry, and cherry that incorporates information on heritability, and available genetic variation to provide breeding targets that are objectively defined and quantified; a portfolio of available technologies and service centers for genetic screening (sampling, DNA extraction, genotyping, and data provision) to assist breeding programs in selecting appropriate technologies; and interfaces with genomics data that facilitate development of improved markers.

Decision-support tools will include: a cross planner that allows breeders to design efficient single and multi-generation crosses for combining specific traits and alleles based on expected seedling outcomes under various selection protocols; and a cost-efficiency calculator with inbuilt logistical analysis for identifying optimal marker-assisted seedling selection (MASS) schemes within individual breeding programs.

MAB Pipeline Implementation (4)

An eight-stage MAB Pipeline will be implemented in core breeding programs by mobilizing Pipeline infrastructure established in BIMS. Implementation of all eight stages in breeding programs will fill the gap between available marker-trait associations and routine MAB. Core participating breeding programs will share a common focus of implementing MAB for fruit quality, although the Pipeline approach is generic for any trait.

A stepwise process will introduce breeders to BIMS functions and MAB capability that will ultimately enable programs to efficiently and routinely leverage genomics advances for improved breeding efficiency. In summary, these steps are: 1) use

  393

of BIMS modules that address traditional breeding needs, 2) addition of marker information to guide parent and cross selection, 3) use of marker information for improving seedling selection efficiency, 4) breeder implementation of earlier Pipeline stages based on increasing confidence with genomics interfacing, and 5) routine MAB using the whole Pipeline.

Extension (5)

Meetings and workshops will be conducted to engage, educate, and train Rosaceae breeders and allied scientists, producers/processors, marketing groups, and trade organizations on RosBREED’s socio-economics and genomics research and application. Extension programming will include short courses and hands-on computer training. Graduate students embedded as “breeding trainees” will each receive mentoring from active breeders and other project participants, and will benefit from a unique integrated educational experience. RosBREED personnel will actively participate in the eXtension Plant Breeding and Genomics Community of Practice (http://pbgworks.hort.oregonstate. edu/forum/41), and utilize this resource as a vehicle to create educational content for the larger community and public as well as a communication and document development tool for project participants. Regular external evaluation of extension activities will provide feedback and prompt adjustments in activities.

CONCLUSIONS

The expectations for RosBREED are high. Multi-million dollar funding for this single project creates opportunities unattainable with multiple smaller-scale efforts. Importantly, RosBREED encompasses cross-commodity and trans-disciplinary collaboration, with the shared goal of establishing routine MAB capability for Rosaceae breeding programs. Coordinated action is now required to make RosBREED a reality. Literature Cited Bink, M.C., Boer, M.P., ter Braak, C.J., Jansen, H., Voorrips, R.E. and Van de Weg, W.E.

2007. Bayesian analysis of complex traits in pedigreed plant populations. Euphytica 161:85-96.

Folta, K. and Gardiner S. 2009. Genetics and Genomics of Rosaceae. Springer, New York.

Shulaev, V., Korban, S.S., Sosinski, B., Abbott, A.G., Aldwinckle, H.S., Folta, K.M., Iezzoni, A., Main, D., Arus, P., Dandekar, A.M., Lewers, K., Brown, S.K., Davis, T.M., Gardiner, S.E., Potter, D. and Veilleux, R.E. 2008. Multiple models for Rosaceae genomics. Plant Phys. 147:985-1003.

Van de Weg, W.E., Voorrips, R.E., Finkers, R., Kodde, L.P., Jansen, J. and Bink, M.C.A.M. 2004. Pedigree genotyping: a new pedigree-based approach of QTL identification and allele mining. Acta Hort. 663:45-50.

 394

Figures

MAB Routine

Use

EXTENSION PLATFORM5

Prioritization

PEDIGREE-BASED BREEDING INFORMATION MANAGEMENT SYSTEM3

GENOME SCANS & COMPARATIVE GENOMICS

Available M-T Associations Warehouse

Selection Target

Identifier

Genome Database

Resources

Crop Reference & Breeding Pedigree

Germplasm Databases

Cross Planning

Tool

Technology Portfolio

Seedling Selection Efficiency

Tool

2

MAB PIPELINE IMPLEMENTATION4

TRAIT & MARKET SEGMENT BREEDING TARGET ESTABLISHMENT1

Genetic Screening Efficiency

Improved Markers Validation Utility MAPS

DecisionsMASS Cost Efficiency

& Logistics

MASS Trial Use

Available M-T

associations

Fig. 1. Establishing and implementing the marker-assisted breeding (MAB) Pipeline. Infrastructure that will be established for enabling routine MAB consists of a transparent process of prioritizing breeding selection targets based on values and preferences from industry sectors and consumers to determine economic weights for traits and market segments (1), development of state-of-the-art crop-specific SNP genome scanning capability for breeding-relevant germplasm that exploits genomic connections among Rosaceae crops (2), and creation of a sustainable data management system with statistical, database, and decision support functionalities (3). Using this infrastructure, an eight-stage MAB Pipeline will be implemented that results in routine MAB in Rosaceae breeding programs (4). An Extension platform engages stakeholders – Rosaceae breeders and allied scientists, the next generation of breeders, producers/processors, marketing groups, trade organizations, and consumers (5).