Single and multiple phenotype QTL analyses of downy mildew resistance in interspecific grapevines

• Published in: Theoretical and applied genetics Vol. 131; no. 5; pp. 1133 - 1143
• Main Authors: Divilov, Konstantin, Barba, Paola, Cadle-Davidson, Lance, Reisch, Bruce I.
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01.05.2018; Springer; Springer Nature B.V
• Subjects: Agriculture; Bayes Theorem; Bayesian analysis; Biochemistry; Biomedical and Life Sciences; Biotechnology; Breeding; Disease resistance; Disease resistance (Plants); Disease Resistance - genetics; Diseases and pests; Downy mildew; Genetic aspects; Genotype & phenotype; Grapes; Health aspects; Hypersensitive response; Inoculation; Interspecific; Life Sciences; Linear Models; Mathematical models; Mildew; NBS-LRR gene; Original; Original Article; Peronospora; Phenotype; Phenotypes; Plant Biochemistry; Plant Breeding/Biotechnology; Plant Diseases - genetics; Plant Diseases - microbiology; Plant Genetics and Genomics; Plant Leaves - anatomy & histology; Prevention; Quantitative Trait Loci; Ribonucleic acid; RNA; Sporulation; Trichomes - growth & development; Vitis - genetics; Vitis - microbiology
• ISSN: 0040-5752; 1432-2242
• DOI: 10.1007/s00122-018-3065-y

Key message Downy mildew resistance across days post-inoculation, experiments, and years in two interspecific grapevine F 1 families was investigated using linear mixed models and Bayesian networks, and five new QTL were identified. Breeding grapevines for downy mildew disease resistance has traditionally relied on qualitative gene resistance, which can be overcome by pathogen evolution. Analyzing two interspecific F 1 families, both having ancestry derived from Vitis vinifera and wild North American Vitis species, across 2 years and multiple experiments, we found multiple loci associated with downy mildew sporulation and hypersensitive response in both families using a single phenotype model. The loci explained between 7 and 17% of the variance for either phenotype, suggesting a complex genetic architecture for these traits in the two families studied. For two loci, we used RNA-Seq to detect differentially transcribed genes and found that the candidate genes at these loci were likely not NBS-LRR genes. Additionally, using a multiple phenotype Bayesian network analysis, we found effects between the leaf trichome density, hypersensitive response, and sporulation phenotypes. Moderate–high heritabilities were found for all three phenotypes, suggesting that selection for downy mildew resistance is an achievable goal by breeding for either physical- or non-physical-based resistance mechanisms, with the combination of the two possibly providing durable resistance.