Published in Crop Sci. 44:1917-1919 (2004).
© 2004 Crop Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
SYMPOSIUM ON GENOMICS AND PLANT BREEDING: THE EXPERIENCE OF THE INITIATIVE FOR FUTURE AGRICULTURAL AND FOOD SYSTEMS
SYMPOSIUM SUMMARY
 |
INTRODUCTION
|
|---|
"To reach an objective, use all the information you have" (Duvick). (Undated attributions are personal communications during the symposium discussion reported on the preceding pages).
Genomics emerged recently as a term to describe investigations of the whole genome using biotechnologies. Given that substantial resources have been devoted to plant genomics and molecular genetics, it is timely to discuss how they can assist plant breeding. This symposium summary serves to emphasize and document ideas, suggestions, and recommendations made by panelists and audience participants.
The panelists were scientists who had received competitive funding for crop genomics research and had articulated a vision and a role for genomics in plant breeding from their particular perspective. Some panelists and audience participants emphasized how genomics could be applied directly to crop improvement, while others emphasized its role in understanding the fundamental biological questions of adaptation and response to selection, which one day may make breeding more efficient. Panelists generally expected new genomic applications will become available as the price of technology continues to drop and as a greater understanding of the plant genome leads to new insights into its manipulation, though views on the usefulness of genomics for crop improvement varied from enthusiasm to skepticism. Moreover, "integration of genomics and plant breeding may become increasingly important if transgenics become unavailable because of lack of public acceptance" (Knapp). Three areas were viewed as most important for application of genomics and molecular genetics: molecular-assisted breeding; gene and genome sequencing and gene networks; and use of genetic diversity.
|
Marker-Assisted Breeding
|
|---|
The use of molecular markers for marker-assisted selection (MAS) received early attention by the plant breeding community and, consequently, has been the approach most used. Validation data are still being obtained and optimal strategies to capitalize on the use of MAS are still being formulated (Cooper et al., this symposium, 1907–1913). Even so, MAS is increasingly efficient, with a steady evolution in the types of markers used. A new generation of molecular markers based on single nucleotide polymorphisms (SNPs) should permit relatively low cost, high-throughput analyses of entire breeding populations (Dubcovsky, this symposium, 1895–1898).
Successful use of MAS requires markers linked to traits of interest. Associations between markers and simply inherited traits with a strong impact on the phenotype are the easiest associations to make (the "low-hanging fruit"—Walsh), but these are precisely the type of trait with which breeders have always been the most successful. Consequently, use of MAS for simply inherited traits can be justified only when it replaces more expensive or tedious assays, or results in increased precision in the identification of desired genotypes (Cooper et al., this symposium, 1907–1913). Two examples of increased precision included the manipulation of tight linkages within Capsicum (Jahn) and wheat (Dubcovsky). In the future, markers linked to simply inherited traits of interest will need to be "resolved to the level of candidate genes" (Cooper et al., this symposium, 1907–1913), making the process more efficient.
Cautious optimism was voiced about MAS of complex traits. Although molecular markers have been successfully associated with quantitative trait loci (QTLs), these associations have had very limited usefulness in plant breeding programs. Complex traits are the most difficult to handle during a breeding program, but are responsible for most breeding progress in critical traits such as yield, yield stability, and adaptation (Nelson et al., this symposium, 1901–1904; Goodman, this symposium, 1913–1914). "We [plant breeders] get paid for the phenotype, yet the individual contribution of yield components to the overall phenotype has not been adequately modeled in a genomics context" (Cooper). Long-term efforts are required to investigate the nature of complex traits at the molecular level, before selection of complex traits by molecular markers can be fully realized. "Molecular genetics will not add much to routine breeding practices until this (the inability of molecular techniques to manipulate complex traits) is overcome" (Goodman, this symposium, 1913–1914).
|
Gene and Genome Sequencing and Gene Networks
|
|---|
Most panelists agreed that more genome sequencing of crop species is essential for improved and continued application of plant genomics to plant breeding. In fact, the sequence of crop genomes may have greater eco nomic impact for developing countries than the efforts of the human genome project (Martienssen, this symposium, 1898–1899). Genome sequence for various crops would improve the quality of molecular markers used for MAS by targeting the gene of interest, rather than a nearby sequence. This limitation of linkage is being resolved as sequence data are becoming available, making it possible to use the gene itself as its own marker (Dubcovsky, this symposium, 1895–1898; Nelson et al., this symposium, 1901–1904; Martienssen, this symposium, 1898–1899). These markers are "based directly on ... variation at the gene responsible ... [for] the trait. Examples of perfect markers (in wheat) include genes for gluten strength, genes for starch properties, genes for hardness, vernalization genes, and the Lr genes for leaf rust resistance" (Dubcovsky, this symposium, 1895–1898).
Polymerase chain reaction (PCR)-based markers are desirable as they can be automated, but every type of PCR-based molecular marker requires up-front DNA sequence information. In fact, SNPs require sequence knowledge of multiple alleles, i.e., sequencing genes a single time is not enough. Continuing efforts to sequence expressed genes will provide data for SNP markers for individual alleles, making MAS more efficient (Dubcovsky). In addition, "other genomic research will provide information regarding nucleotide sequences which may prove more valuable than many of the transcribed (i.e., expressed) genes. Approaches for cloning gene-rich regions of the genome will provide sequence information of for both genes and their controlling regions" (Paterson et al., this symposium, 1900–1901).
"No gene acts alone" (Walsh), yet interactions between genes—so called "gene-networks"—are little-understood at this time. Accurate gene-to-phenotype models will depend on better understanding of these intergenic interactions. Global gene expression studies will be very valuable to help address problems intractable until now. This will require, in addition, a greater understanding of whole plant physiology and yield, and the functional interactions and properties of genes (Cooper et al., this symposium, 1907–1913). Until better gene prediction models are in place, breeding programs will not be able to rely exclusively on MAS, but must supplement genomics-based efforts with analyses of phenotypic measurements from replicated field trials.
|
Use of Genetic Diversity
|
|---|
Many wild plant species are related to modern crops and contain useful traits which are not found in adapted varieties, including additional disease resistances, stress tolerance, and even genes for increased production (Brummer, this symposium, 1904–1907; Paterson et al., this symposium, 1900–1901). Use of broad-based genetic diversity in breeding will benefit from MAS to follow specific genes from unadapted relatives and/or different plant species as they are bred into elite varieties (Paterson et al., this symposium, 1900–1901; Dubcovsky, this symposium, 1895–1898; Brummer, this symposium, 1904– 1907). It will be important to assay as much diversity as possible, which in turn requires that the germplasm collections be well maintained and curated (Jahn, Goodman). "Many of the answers to our questions about crop productivity, and more generally plant biology, lie in [these] germplasm resources" (Jahn).
|
Can Genomics be Useful to Plant Breeding?
|
|---|
The most frequent answer was a cautious "yes" particularly when MAS was considered. To what extent, has genomics been useful to plant breeding programs to date? The best example may be the wheat MAS project, which not only developed molecular markers but also facilitated their transfer to breeders who use them: "MAS programs are good examples of implementation projects that have the potential to facilitate the transfer of valuable genes identified in basic research programs into public varieties" (Dubcovsky, this symposium, 1895–1898).
Nearly all panelists agreed that the largest potential benefits of plant genomics are still years away. For example, "It is going to take as long to breed an engineered gene into a successful cultivar as it takes for a natural gene. If better crop performance such as yield is the ultimate goal, in forty years from now, traditional plant breeding methods would have been the best investment for today's dollar" (Goodman; Goodman, this symposium, 1913–1914). Although not all panelists predicted a forty-year time lag, all perceived that use of today's genomics information in plant breeding is limited.
What are the factors that limit current genomics applications in plant breeding? Some insights from the symposium included the following.
It will take time to develop a way to use genomics that is a practical improvement over very successful current methods. (Cooper)
Few researchers or students are conversant at an adequate professional level in both plant genomics and plant breeding. (Brummer, Duvick)
Plant genomic research has centered on model species; many critical crop traits are simply not represented in these models (Paterson; Havey, this symposium, 1893–1895; Nelson et al., this symposium, 1901–1904). "Direct analysis of crop genomes will continue to be essential to elucidate the unique features that make them important as crops (such as cotton fibers and underground peanut pegs, that is, traits not found in model species)" (Paterson).
Plant breeding infrastructure and human resources have been lost, limiting capacity to take advantage of genomics (Duvick, Goodman, Pratt).
Plant breeding capacity to use genomics is particularly lacking for "orphan" crops—that is, crops grown on small acreages or grown primarily in developing countries (Havey, this symposium, 1893–1895; Nelson et al., this symposium, 1901–1904). To the extent that the "model crops" model is valid, there is hope that genomics tools developed for any given species can be directly applied to related orphan crops.
While there were diverse opinions among the participants on whether and how to use genomics and molecular genetic approaches in plant breeding, there was a strong consensus that the plant breeding infrastructure and resources to permit genomics research to be applied to crop improvement are inadequate and declining. "The paradox of the genomics age is that funding for plant breeding programs is decreasing at the same time that the potential of genomics is being realized ... with the result that many technological advances in genomics may not be applied to cultivar development at all" (Brummer, this symposium, 1904–1907). Currently, "one costly outcome of reductions to (plant breeding) programs at public universities is the loss of an educational environment which provides hands-on breeding experience" (Jahn). Without skilled plant breeding programs constantly at work—using genomics and "all the information we have" (Duvick)—to meet ever-changing local and global challenges, plant genomics research will miss its pay-day.
|
ACKNOWLEDGMENTS
|
|---|
The editors thank the panelists and audience members for their participation; and Thomas Stalker, Crop Science Dep., North Carolina State Univ., and Edward Kaleikau, CSREES, USDA, for indispensable assistance in planning and preparations. In addition to panelists, audience participants in the discussion of the symposium topic were D. Duvick, Iowa State Univ.; S. Kaeppler, Univ. of Wisconsin; R. Pratt, The Ohio State Univ.; D. Stuthman, Univ. of Minnesota; B. Walsh, Univ. of Arizona; and two anonymous participants. The editors are responsible for any misrepresentations in the discussion report. The symposium was supported by a CSREES, USDA Innovation Grant.
TOP
INTRODUCTION
Marker-Assisted Breeding
