Towards developing a metabolic-marker based predictive model for Phytophthora nicotianae tolerance in citrus rootstocks

Root rot of citrus trees caused by Phytophthora nicotianae is responsible for severe economic losses in citriculture. Use of resistant rootstocks is an effective method of managing this problem, however, breeding and selection of new citrus rootstocks is a time-consuming undertaking. The objective was to develop a method for the rapid assessment of rootstocks for P. nicotianae tolerance, using a metabolomics approach to identify metabolic markers for the phenotypic trait of tolerance. Sixteen citrus rootstocks were inoculated with P. nicotianae in the greenhouse for determination of relative tolerance/susceptibility. Healthy citrus roots from four tolerant and four susceptible rootstocks were used for metabolite analysis with the objective of identifying potential biomarkers. Organic solvent extractions of the roots were prepared and analysed by mass-spectrometry based liquid chromatography, which produced 367 ion features (retention time and m/z). Orthogonal partial least squares discriminant analysis of peak abundance using MarkerLynx software allowed for the identification of ion features that differentiate tolerant and susceptible rootstocks. Using descriptive and inferential statistics based on the ion features of uninoculated tolerant vs. susceptible rootstocks, applying logistic regression, 14 top markers were identified and two of them (22.03_259.0975 and 22.21_313.1445: retention time (rt) and mass to charge ratio (m/z) were accepted as potential metabolic markers. A model that can potentially predict tolerance in citrus rootstocks with >98% accuracy is presented.The authors acknowledge financial support to conduct this research from Citrus Research International, South Africa. Supplementary funding was provided by the University of Pretoria (South Africa). We also thank Dr. Wilhelm Botha for the morphological and molecular characterisation of the pathogen (Pathogen Accession Number: Ph 453 PPRI 23883 Agricultural Research Council, Plant Protection Research Institute, Pretoria, South Africa).https://link.springer.com/journal/421612019-07-01am2019BiochemistryPlant ScienceStatistic