marker-assisted selection in wheat

Chapter 15 – Marker-assisted selection in forestry species 287Over 80 000 ESTs have been sequencedfrom pine (http://pinetree.ccgb.umn.edu/),over 130 000 from poplar (http://poppel.fysbot.umu.se/) and over 100 000 fromspruce (www.arborea.ulaval.ca/en/; www.treenomix.ca/).Comprehensive microarrays (Shena etal., 1996) are now being used in manyof these species, allowing transcriptionprofiling of thousands of genes incontrasting phenotypes in a range of tissuesunder different environmental/stress/developmental regimes. Identification ofcandidate genes from expression profilingis based on the assumption that the genesshowing genotype-specific differences intheir level of expression are causing variationin that trait (Morgante and Salamini, 2003).Microarrays are being used to identify genesthat are regulated differentially in individualswith contrasting wood traits in eucalypts(Moran et al., 2002), symbiosis-regulatedgenes in Eucalyptus globulus-Pisolithustinctorius ectomycorrhiza (Voiblet et al.,2001) and genes involved in embryogenesisin pines (van Zyl et al., 2003). Usingmicroarrays, many genes involved in cellwall biosynthesis have been identifiedin loblolly pine (Whetten et al., 2001;Pavy et al., 2005), eucalypts (Paux et al.,2004) and hybrid aspen (Populus tremulax P. tremuloides) (Hertzberg et al., 2001).A combination of proteomics, whichexamines changes in protein expression indifferent tissue and developmental stages,and microarray technology is also beingused to give a more complete picture ofgene function, for example of droughttolerance in Pinus pinaster (Plomion et al.,2004). This discovery work is uncoveringlarge numbers of candidate genes that areexcellent targets for both QTL mappingand association studies aimed at identifyingmarkers for use in MAS.Family-based genetic linkagemapping and QTL analysisGenetic linkage or recombination mappingrelies on finding sufficient polymorphismusing DNA markers in progeny arraysfrom full-sib pedigrees to identifyassociations between linked loci on a chromosome.Genetic linkage maps have beenconstructed for most of the commerciallyimportant forest tree genera (summarizedin Table 1), and updated information ongenetic linkage maps for forest trees isavailable at http://dendrome.ucdavis.edu/index.php. The number and location ofchromosomal regions affecting a trait(QTL) and the magnitude of their effectcan then be investigated by QTL mapping.QTL are identified by a statistical associationbetween variation in a quantitativetrait and segregation of alleles at a markerlocus in a segregating population (mappingpedigree).Most phenotypic traits of interest for treebreeding are characterized by continuousvariation. Such traits are usually influencedby a number of genes with a small effectinteracting with other genes and the environment.The main traits targeted for QTLmapping are wood properties and traitsrelated to adaptation and growth (reviewedby Sewell and Neale, 2000). These includephysical wood properties that affect thestrength of sawn timber (e.g. wood densityand microfibril angle), and properties thataffect paper pulping, e.g. pulp yield, fibrelength and the relative proportion of cellulose,hemicellulose and lignin, generallymeasured as percentage cellulose. In addition,QTL have been identified for diseaseresistance, growth, flowering, vegetativepropagation, frost tolerance and leaf oilcomposition (Table 2).The detection of QTL requires largesample sizes; the lower the heritability