Submitted to: Euphytica
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 9/20/2016
Publication Date: N/A
Citation: N/A

Interpretive Summary: Sugarcane orange rust, caused by the fungus Puccinia kuehnii, is a serious disease of sugarcane, especially in Florida. The most effective way to combat orange rust is to breed varieties that are genetically resistant to the disease. The breeding process would be faster if DNA markers were available to quickly identify plants that are resistant, without having to infect them with orange rust, and wait for visual symptoms to appear. The objective of this research was to identify DNA markers for sugarcane orange rust resistance, using high-throughput next-generation sequencing methods. Over 49,000 DNA markers from 71 sugarcane varieties that were either highly resistant or highly susceptible to orange rust were examined. Forty-eight markers that looked promising were tested again on the original 71 varieties, and 495 additional varieties from the germplasm collection at Canal Point, Florida. Thirteen of these markers were found to be associated with orange rust resistance/susceptibility. This research has provided DNA markers which can be immediately used to select superior parents for orange rust resistance, and has also given evidence that resistance to orange rust may be controlled by a number of genes with small to medium effects.

Technical Abstract: Sugarcane orange rust, caused by the fungus Puccinia kuehnii, is a serious disease of sugarcane. The most effective strategy to combat the disease is to develop resistant sugarcane cultivars. Phenotypic screening for resistance to orange rust is a laborious and time-consuming process, and breeders would greatly benefit from molecular markers linked to resistance. The objective of this research was to identify, via association mapping, markers linked to resistance to orange rust. From the germplasm available in the breeding nursery at Canal Point, Florida, 726 genotypes were screened for orange rust resistance via artificial inoculation of field-grown plants. A frequency distribution was generated from the disease reaction scores, and 39 susceptible and 37 resistant genotypes, comprising the upper and lower 5 percent of scores, respectively, were chosen for marker analysis. These individuals were genotyped via capture sequencing using 32,555 sugarcane probes anchored to the sorghum genome. Principal components analysis of the genomic kinship matrix suggested some degree of population structure in the dataset. Several methods were used to account for population structure and/or kinship as part of the association analysis. The most appropriate methods (determined by the genomic inflation factor) were mixed model analysis, structured association analysis, and inclusion of principal components. Amongst the 75 most significant markers identified by the three methods (the top 25 from each method), 27 were identified by at least two methods, resulting in 48 unique SNPs. These were validated on the surviving 566 members of the population, using high resolution melting of small PCR amplicons. Thirteen of these markers were statistically significant for the quantitative measure of orange rust severity, and ten markers were significant for the qualitative measure of ‘status’ (resistant/susceptible). Combined, the markers explained 42.1% of the variation for orange rust severity, and 36.2% of the variation for orange rust status. The maximum amount of variation explained by any single marker was approximately 11%. This research has provided molecular markers which can be immediately deployed to select superior parents for orange rust, and has also provided valuable insight for future research towards genetic control of orange rust in sugarcane.