Using Functional Genomics to Identify Molecular Markers for Fire Blight Resistance (Erwinia Amylovora) in Apple (Malus)

Fire blight, caused by Erwinia amylovora (Ea), is a destructive disease of Malus (apple), Pyrus (pear) and some woody ornamentals in the rose family (Rosaceae). The goal of this project is to use a functional genomics approach to develop tools to breed fire blight resistant apples. Suppression subtractive hybridization (SSH) and cDNA-AFLP analysis were used to identify 650 expressed sequence tags (ESTs) associated with fire blight from Ea-challenged apple leaf tissue. ESTs were ranked for their potential impact on resistance based on bioinformatics and inferences drawn from model systems. Simple sequence repeat (SSR) and single nucleotide polymorphism (SNP) markers derived from highly ranked fire blight-associated ESTs were mapped in a ‘M.9’ × ‘Robusta 5’ population in which a major QTL for fire blight resistance has been located on Linkage Group 03. Highly ranked fire blight-associated ESTs were mapped to this QTL, as well as to the positions corresponding to the location of at least two QTLs reported in other populations. A secretory class III peroxidase mapped to the LG03 fire blight resistance QTL and a serine/threonine protein kinase mapped at a 4 cM distance. To date, 6 out of 28 candidate fire blight resistance gene markers that have been mapped, have co-located to or near known fire blight resistance QTLs. This research will facilitate new methods of marker-assisted selection to efficiently breed superior apple cultivars with fire blight resistance. INTRODUCTION Fire blight, caused by the bacterial pathogen Erwinia amylovora (Ea), is a destructive disease of apple and pear with infections occurring on blossoms, fruits, vegetative shoots, woody tissues, and rootstock crowns. The disease can kill young trees outright or result in permanent structural damage to mature trees. Epidemics of fire blight develop explosively, and current control measures are not adequate in the face of diseasea present address: Istituto Agrario San Michele all'Adige, E. Mach Foundation, Via E. Mach 1, 38010 San Michele all’Adige (TN), Italy 415 Proc. 1 st IS on Biotechnol. of Fruit Species Eds.: M.-V. Hanke et al. Acta Hort. 839, ISHS 2009 favoring weather conditions. The most effective control, the antibiotic streptomycin, is now ineffective in many areas because of development of streptomycin-resistant strains of Ea (Loper et al., 1991). In addition, there are significant environmental and public health concerns regarding the use of antibiotics in agriculture (McManus et al., 2002). As alternative controls including biologicals and resistance inducers provide inadequate partial control, the need for fire blight-resistant cultivars is pressing. Because of the long-term, perennial nature of apple orchards (ca. 20 years), the deployment of durable disease resistance is critical in apple. Gene pyramiding is a commonly used approach to enhance the durability of resistances and marker-assisted breeding is an efficient method to pyramid resistance genes. Major gene resistance to fire blight has not been identified in apple, unlike resistance against other major diseases such as apple scab. To date, two promising QTLs for fire blight resistance have been identified from M. × robusta 5 (Peil et al., 2007, 2008) and the commercial cultivar ‘Fiesta’ (Calenge et al., 2005; Khan et al., 2006). The goal of this project was to use functional genomics analysis to identify candidate fire blight resistance genes and then to map these genes to determine their association with quantitative trait loci (QTLs) for fire blight resistance (Fig. 1). The transcript profiling methods of suppression subtractive hybridization (SSH) and cDNAAFLP analysis were used to identify over 650 expressed sequence tags (ESTs) associated with fire blight resistance and susceptibility (Malnoy et al., 2007; Norelli et al., in press). To narrow this large number of genes to a smaller number of useful markers for cultivar improvement, a candidate gene approach was used to first identify “candidate” fire blight resistance genes and then determine their co-localization with known fire blight resistance QTLs (Peace and Norelli, in press). ESTs with a high likelihood of playing a role in resistance were identified by bioinformatics and inferences drawn from the scientific literature. Once candidates had been identified, DNA markers, such as simple sequence repeat (SSR) and single nucleotide polymorphism (SNP), were developed for the genes, mapped, and their locations compared to the two known QTLs for fire blight resistance. In future work, markers closely flanking the QTLs will be developed into “breeder friendly” markers for marker assisted breeding of fire blight resistant apple rootstock and scion cultivars. Additionally, mapping results will be used to select candidate resistance genes for further functional studies by transgenic analysis (Fig. 1). MATERIALS AND METHODS Bioinformatics From 5,395 apple ESTs identified from Ea-challenged leaves of ‘Red Delicious’ and ‘G.41’ apple rootstock, we identified 883 unique contigs (581 clusters and 302 singletons) or “fire blight unigenes” (Baldo et al., 2007). An additional 200,115 ESTs from presumably non-infected apple tissues were downloaded from the GenBank database and clustered into 34,982 unique contigs (23,870 clusters and 11,112 singletons) or “non-fire blight unigenes”. To identify ESTs unique to Ea-challenged tissue, we selected SSH and cDNA-AFLP ESTs with the worst BLAST match (highest E values) to non-fire blight unigenes. To select ESTs similar to genes associated with bacterial resistance in model systems, we selected fire blight associated ESTs with the best BLAST match (lowest E values) to Arabidopsis ESTs associated with resistance to Pseudomonas syringae pv. tomato or with systemic acquired resistance. Mapping of Candidate Genes The ‘M.9’ × ‘Robusta 5’ map described in Celton et al. (2008) was used to accurately identify the location of a QTL for fire blight resistance on Linkage Group 3 (LG03) and for all candidate gene mapping. The most highly ranked candidate resistance genes were screened by bin mapping to identify candidates located near QTLs for fire blight resistance, either from ‘Robusta 5’ directly, or ‘Fiesta’ (Calenge et al., 2005; Khan et al., 2006) by map alignment. SSR motifs in candidate genes were detected using the