Department of Chemistry, University of Saskatchewan, Saskatoon, SK S7N SC9, Canada

Background and objectives
Phytoalexins play a significant role in the defence response of plants. These secondary metabolites, which are synthesized de novo in response to diverse forms of stress, including fungal infection, are part of the plant's chemical and biochemical defence mechanisms. However, when pathogenic fungi can effectively disarm the plant by detoxifying phytoalexins, the outcome of the interaction can favour the pathogen and be detrimental to the plant. An enviromnentally advantageous strategy for controlling pathogenic fungi of crucifers could involve the inhibition of the fungal enzymes involved in the detoxification of phytoalexins. Towards this end, we have been investigating the effects of brassica phytoalexins on one of the most significant canola (Brassica napus, B. rapa) pathogens, the blackleg fungus (Leptosphaeria maculans, asexual stage Phoma lingam) [1]. We have demonstrated that P. lingam can metabolize the phytoalexin brassinin by promptly transforming it into harmless metabolites. We have now examined the metabolism of the phytoalexins cyclobrassinin, 1-methoxybrassinin, and spirobrassinin by P. lingam, as well as the antifungal activity of the metabolic products.

Results and conclusions
Cyclobrassinin was incubated with the so-called 'antivirulent' isolate Unity [2], to establish a time-course transformation profile. Analysis of organic extracts of culture samples by HPLC indicated a rapid decrease in the concentration of cyclobrassinin and the concurrent appearance of two additional constituents over a 12-h period. Subsequently, to obtain sufficient quantities of each constituent, larger-scale fungal cultures incubated with cyclobrassinin were extracted, the extract fractionated by chromatography, and each fraction analysed by HPLC. The fractions containing the new constituents were analysed by standard spectroscopic methods for structural elucidation. Based on these results, one of the new constituents was established to be the known phytoalexin brassilexin, and the other new constituent was assigned as a tautomeric mixture. Two days after incubation of the isolate Unity with cyclobrassinin, no brassilexin or other phytoalexins or putative metabolites were detected in any of the cultures, or their extracts. Similar experiments carried out with the 'virulent' isolate BJ-125 incubated with cyclobrassinin for 12 h afforded yet another known phytoalexin, dioxibrassinin. Similar experiments with spirobrassinin indicated that this phytoalexin was not metabolized and had no effect on mycelial growth, whereas methoxybrassinin was completely metabolized to polar metabolites.

These results indicate that spirobrassinin is not metabolized by isolates of P. lingam, whereas cyclobrassinin is detoxified via the phytoalexins brassilexin or dioxibrassinin, depending on the particular fungal 'group'. Our results suggest that two different groups of P. lingam, the so-called virulent and avirulent groups, can metabolize the phytoalexin cyclobrassinin 'mimicking' pathways that may operate in the plant. Considering that fungal pathogens have been co-evolving with plants for multiple generations, the detoxification of phytoalexins by 'mimicry' appears quite plausible. Nonetheless, because in planta only a part of the biosynthetic pathway of cruciferous phytoalexins has been established, such a hypothesis remains to be demonstrated. A clearer picture will eventually unfold upon tracing a complete map of phytoalexin transformation in both cruciferous plants and their pathogenic fungi.

1. Pedras MSC, Khan AQ, Taylor JL, 1997. In Hedin PA et al., eds, Phytochemicals for Pest Control. ACS Symposium Series 658, pp. 155-166.
2. Pedras MSC, Taylor JL, Morales VM, 1995. Phytochemistry 38, 1215-1222.