The general goal of plant breeding is the improvement of plants for human benefit, fulfilling the needs of both producers and consumers. However, the exponential growth of human population, the effects of the global warming on the crop environments and the need to stop the negative impact of agriculture on the ecosystems, are demanding new and urgent specific goals (of particular social value) to maintain the quality of life on earth. In short, a more sustainable and environmental friendly agriculture that increases crop yield taking into account the tolerance to biotic and abiotic stresses and the efficiency of water and nutrient uptake. This will allow colonizing marginal areas for agriculture and decreasing its dependence on pesticides, herbicides, fertilizers and water. The great scientific advances of twenty-first century should help to pace these new challenges.
M.J. Asins (*), G.P. Bernet, I. Villalta, and E.A. Carbonell IVIA, Apdo. Oficial, 46113, Moneada, Valencia, Spain e-mail: [email protected]
S.M. Jain and D.S. Brar (eds.), Molecular Techniques in Crop Improvement, DOI 10.1007/978-90-481-2967-6_1, © Springer Science+Business Media B.V. 2010
Plant breeding has a long history of integrating the latest innovations in biology and genetics to enhance crop improvement (Moose and Mumm 2008). The large advance in science has stimulated shifts in funding at public institutions to enhance intellectual capacity and infrastructure for molecular genetics and genomics, often, ironically, at the expense of conventional plant breeding (Knight 2003; Brummer 2004). Likely linked to this, plant breeding has often moved from public institutions to commercial companies. There is a need of research linking molecular methods with breeding objectives to fully realize the potential of recent advances in biotechnology and genomics (Guimaraes and Kueneman 2006; National Research Council 2008)
Nowadays, there are two different strategies for molecular improvement depending on the origin of genetic variability: hybridization of genotypes (the classical one), and genetic transformation and TILLING (Targeting Induced Local Lesions in Genomes, Slade et al. 2005) of an elite genotype. The former uses natural genetic variability, and allows the obtaining of multiple combinations integrated by multiple variants of (usually) small effect each. The latter obtains new forms yielding few variants that contain, in the case of transformation, one or few copies of a functionally important gene or DNA fragment (supposedly of large effect on the phenotype) and in the case of TILLING, a mutated allele of the gene of interest. Both alternatives manage very different amounts of genetic variability (large and small, respectively) and provide complementary information (from forward and reverse genetics, respectively). Although the latter, if successful, might seem more attractive due to the rapid genetic gain in the trait of interest, breeding programs must cope with multiple agronomic traits (yield, quality, timing, resistance ...), most of them controlled by multiple genes (some of them linked or with pleiotropic effects) and interactions (with other genes and the environment). Therefore, at least the knowledge of the position of the genes controlling these traits is relevant.
Only by the joint analysis of segregation of marker genotypes and of phenotypic values of individuals or lines, it is possible to detect and locate loci affecting quantitative traits ("quantitative trait loci" or "QTLs"). QTLs are, a priori, difficult to identify due to the lack of discrete phenotypic segregation and because the pheno-typic effect of each gene associated with a complex trait is relatively small. QTL analysis in segregating populations, involves selecting and hybridizing parental lines that differ in one or more quantitative traits and analyzing the progeny in order to link the QTL to known DNA markers. A breeder can use this knowledge to advantage, for instance by using indirect selection. When selection is (partly) based on genetic information retrieved through the application of molecular markers this is called marker-assisted selection (MAS). It can be employed to enhance plant breeding efforts and to speed up the creation of cultivars. Also, it unveils masked, interesting wild alleles and makes for an easier introduction of genetic material from related and unrelated wild species, without the drawbacks ("linkage drag") that are associated with the introduction of "wild genes" through conventional methods, facilitating germplasm enhancement and pre-breeding.
QTL analysis is usually seen as the methodology of choice to study quantitative traits, which show continuous variation and are theoretically controlled by many genes. But, a priori, any trait (not only the quantitative traits) can be controlled by more than one gene what makes QTL analysis an universal tool to obtain information on the number and position of genes controlling it. We may detect just one QTL in a given segregating population but by using this genomic scanning tool, other hypothesis (more than one locus) can be tested, too.
QTL analysis has clear limitations that have been reviewed by several authors (Doerge and Rebai 1996; Beavis 1994; Kearsey and Farquhar 1998; Doerge 2002; Asins 2002; Melchinger et al. 2004; Holland 2007). A major problem, pointed by Goring et al. (2001), come from the fact that the chromosomal position and geno-type-phenotype relationship of a locus cannot both be reliably estimated by the use of a single data set of currently realistic size, at least for loci of small effect size. An approach put forward by Sen and Churchill (2001) breaks the QTL problem into two distinct parts: the relationship between the QTL and the quantitative trait, and the location of the QTL. So the initial focus is placed on estimation of the unknown QTL genotypes and then on allowing the search for different models and their comparisons with the information gained from completing the QTL genotype information (Doerge 2002). Other improvements concerning the experimental design (segregating populations from multiple parents and environments) and the development of new statistical methodologies for locating multiple QTLs (reviewed by Zou and Zeng 2008) will enhance our capability of detecting more QTLs per trait, including those of small effect. Besides, recent advances in statistical methods to control the QTL false-discovery rate should provide a better balance between declaring too many false-positive QTL and sacrificing power to detect those that have smaller effects (Benjamini and Yekutieli 2005).
Advances in the application of QTL analysis to plants are also becoming evident.
The most obvious are marker-assisted selection (MAS) in breeding and pre-breeding and QTL cloning. Other areas where QTL analysis is contributing decisively are: (1) understanding of complex but agronomically important traits such as plant-pathogen interaction and adaptability to marginal areas (genotype-environment interaction); (2) plant genomics, providing information on candidate polymorphisms by connecting the QTLs that control phenotypes, metabolites, proteins, genes and regulatory elements; and (3) germplasm enhancement allowing its efficient utilization in pre-breeding through genotyping for candidate polymorphisms. All of them, and particularly MAS, have been a major impetus to quantitative genetics research and breeding.
The present chapter focuses on recent advances of three applications of QTL analysis in plants: (1), the genetic integration of agronomical, physiological and gene expression traits (the scientific value of QTL analysis) (2) MAS in breeding programs and (3) utilization of wild germplasm to improve quantitative traits, with breeding tomato for salt tolerance as an example.
Was this article helpful?