Effect of artificial ROS manipulation on M. oryzae , C. miyabeanus and R. solani infection. For continuous generation of H2O2 in situ, detached leaves were infiltrated with mixtures of glucose oxidase (GO; 100 units ml-1) plus glucose (G; 2 mM), or xanthine oxidase (XO; 0.1 units ml-1) plus xanthine (X; 1 mM). Control plants were treated with buffer solution only (50 mM phosphate, pH = 6.5). Alternatively, plants were infiltrated with 3-aminotriazole (3-AT; 10 mM) or catalase (CAT; 1100 units ml-1) with MES buffer-treated plants as corresponding controls. Two hours later, 10 μl droplets of conidial suspension of M. oryzae or C. miyabeanus were carefully applied to the center of the infiltrated area. For infection with R. solani, 8-mm mycelium-overgrown agar plugs were used. After 4 days of incubation under laboratory conditions, M. oryzae and C. miyabeanus symptom development was assessed using digital image analysis for quantification of necrotic leaf areas. The intensity of the R. solani symptoms was evaluated 60 h post-inoculation and graded into five categories based on the leaf area affected as described in the Methods section. In all graphs, bars represent the mean and SD of twenty-four leaf segments. Different letters indicate statistically significant differences between treatments (M. oryzae and C. miyabeanus, Fisher's LSD test, α = 0.05; R. solani, Mann-Whitney, α = 0.05). Photographs depicting representative symptoms were taken 96 hpi in case of M. oryzae and C. miyabeanus challenge, and 60 hpi in case of challenge with R. solani.