Metabolomic studies for the interaction Glycine max-Fusarium tucumaniae

2. SCANDIANI MM; LUQUE A; O´DONNELL K; AOKI T; SPAMPINATO C; CERVIGNI GDL

Durban

Congreso; World Soybean Research Conference IX; 2013

Sudden-death syndrome (SDS) of soybean can be caused in Argentina by 4 differentFusarium species: F. brasiliense, F. crassistipitatum, F. tucumaniae and F. virguliforme.Fusarium tucumaniae and F. virguliforme are the primary etiological agents of soybean SDSin Argentina and United States, respectively. The resistance mechanism to these pathogens insoybean is complex. Metabolomic technology was explored to phenotype resistance to F.tucumaniae. Two soybean cultivars (CV) with contrasting levels of resistance to SDS wereinoculated with an isolate of F. tucumaniae grown on sorghum grain, using the layer method.Uninoculated controls were included for both genotypes. Plants were grown for 7, 10, 14 and25 days. Four independent biological replicates were harvested at indicated times. Foliardisease, root rot incidence and severity, shoot height, shoot and root weights were rated ateach time. Areas under disease severity progress curves were also calculated 25 days afterinoculation (DAI). All data were subjected to statistical analysis. Means were compared byleast significant differences (p<0.05). The resistant cultivar showed lower foliar disease androot rot incidence and severity, higher plant height, and root and shoot weights, than thesusceptible cultivar. Uninoculated controls remained healthy. Experiments were extended toroot metabolite profiling by gas chromatography mass spectrometry (GC-MS). Metabolitelevels were normalized to the ribitol internal standard. Compounds were putatively identifiedby comparison of their retention index and mass spectrum with those present in thecommercial mass spectra library NIST. Data from two biological replicates from tolerant andsusceptible CVs obtained at 7, 10, 14 and 25 DAI were evaluated by principal components(PC) and principal coordinates analysis (ACoP). Analyses were either performed forindividuals or pools. Results obtained allow us to identify and monitor the relative levels of30 metabolites, including amino acids, organic acids, soluble sugars, secondary metabolites,inorganic and nitrogen compounds. These metabolites were more abundant at 7 and 10 DAI.The first two PCs (PC1 and PC2) explained 76.97 %, 84.88% and 79.11% of the variance ofthe susceptible, tolerant and pooled CV profiles, respectively. As these values are lower thanexpected (> 90%), the first four PCs had to be considered. In the susceptible CV, inorganicphosphate and sucrose showed the highest weight in PC1 and PC2, respectively. However, inthe tolerant CV, inositol and sucrose were the most important variables in PC1 and malonicacid, citric acid, inorganic phosphate and myoinositol heavily contributed to PC2. When thesedata were pooled and analyzed, inorganic phosphate and sucrose were associated with PC2but no variables could be identified in PC1. In accordance with PC analysis, the first twoACoP factors (F1 and F2) explained the 63.90%, 67% and 43.34% of the variance of thesusceptible, tolerant and pooled CV profiles, respectively. Four DAI metabolic profiles couldbe clearly separated in the tolerant CV, but 7 and 10 DAI remained close together in thesusceptible CV and in the pool analyses.