Genetic mapping and prediction for novel lesion mimic in maize demonstrates quantitative effects from genetic background, environment and epistasis

Alper Adak, Seth C. Murray, Claudia Irene Calderón, Valentina Infante, Jennifer Wilker, José I. Varela, Nithya Subramanian, Thomas Isakeit, Jean-Michel Ané, Jason Wallace, Natalia de Leon, Matthew A. Stull, Marcel Brun, Joshua Hill & Charles D. Johnson

Theoretical and Applied Genetics July 2023; vol. 136, Article number: 155

Figure: Disease Lesion Mimics Mutants of Maize

(Gurmukh S. Johal 2007)

Key message

A novel locus was discovered on chromosome 7 associated with a lesion mimic in maize; this lesion mimic had a quantitative and heritable phenotype and was predicted better via subset genomic markers than whole genome markers across diverse environments.

Abstract

Lesion mimics are a phenotype of leaf micro-spotting in maize (Zea mays L.), which can be early signs of biotic or abiotic stresses. Dissecting its inheritance is helpful to understand how these loci behave across different genetic backgrounds. Here, 538 maize recombinant inbred lines (RILs) segregating for a novel lesion mimic were quantitatively phenotyped in Georgia, Texas, and Wisconsin. These RILs were derived from three bi-parental crosses using a tropical pollinator (Tx773) as the common parent crossed with three inbreds (LH195, LH82, and PB80). While this lesion mimic was heritable across three environments based on phenotypic (Hpp = 0.68) and genomic (Hgg = 0.91) data, transgressive segregation was observed. A genome-wide association study identified a single novel locus on chromosome 7 (at 70.6 Mb) also covered by a quantitative trait locus interval (69.3–71.0 Mb), explaining 11–15% of the variation, depending on the environment. One candidate gene identified in this region, Zm00001eb308070, is related to the abscisic acid pathway involving in cell death. Genomic predictions were applied to genome-wide markers (39,611 markers) contrasted with a marker subset (51 markers). Population structure explained more variation than environment in genomic prediction, but other substantial genetic background effects were additionally detected. Subset markers explained substantially less genetic variation (24.9%) for the lesion mimic than whole genome markers (55.4%) in the model, yet predicted the lesion mimic better (0.56–0.66 vs. 0.26–0.29). These results indicate this lesion mimic phenotype was less affected by environment than by epistasis and genetic background effects, which explain its transgressive segregation.

See https://link.springer.com/article/10.1007/s00122-023-04394-y