1University of Oklahoma, Botany and Microbiology, Norman, Oklahoma 73019, USA; 2ProTech, Inc., 4408 Brookfield Drive, Norman, Oklahoma 73072, USA

Background and objectives
Pathogenic fungi are major problems in agriculture. Chemical control using existing chemistries may not be feasible in the future because of economic and environmental issues. Genetic engineering provides an opportunity to protect plants from fungal diseases and to reduce use of synthetic fungicides. The focus of our research is to identify, purify and characterize antifungal activities, especially antifungal proteins and peptides, and to isolate genes encoding these proteins/peptides. These genes can be engineered into plants to increase the resistance of crop plants to fungal attack, decreasing the use of environmentally unfriendly pesticides. The major factor limiting the application of this technology is the identification and isolation of useful genes/gene products. Plants themselves are a potential source of new antifungal activities. For example, two potent antifungal proteins have been isolated from radish [1]. The genes for these proteins were cloned and constitutively expressed in transgenic tobacco plants. The transgenic tobacco showed enhanced resistance to pathogenic fungi. Plants from the tropical rainforests of Central and South America are a source of novel pesticidal proteins. These plants are under constant selective pressure to develop defence-related proteins because of the diversity of pathogenic fungi and intensity of fungal attack under the moist tropical conditions present in the rainforest. We have examined a number of these species from different families for antifungal activity.

Materials and methods
Seeds, leaves and roots of rainforest species were collected and frozen on dry ice. Frozen plant tissue was homogenized in buffer, filtered and centrifuged to give a crude extract. Part of the crude extract was dialysed extensively to remove low molecular-weight compounds, leaving a protein-containing fraction. Both crude extracts and dialysed fractions were tested for biological activity using the liquid antifungal assay developed by Duvick et al. [2]. Selected extracts which showed activity in the dialysed fractions were fractionated further to isolate and characterize the antifungal activity.

Results and conclusions
Crude and dialysed extracts of over 100 tropical species have been tested for antifungal activity. Crude extracts of 13 plants showed complete to very strong inhibition against Fusarium chlamydosporum. Several of these extracts retained their activity after extensive dialysis, suggesting that the protein-containing fraction from these extracts accounted for the activity. We have chosen to purify the antifungal proteins from several of the plants exhibiting the best activity. Extracts of these plants strongly inhibited the growth of a number of pathogenic fungi including Aspergillus flavus, Fusarium chlamydosporum and Fusarium moniliforme. Several antifungal activities have been partially purified and characterized. The potential applications of these activities, and the genes encoding these proteins to increase resistance to fungal pathogens through a transgenic approach, are being examined.

1. Terras FRG, Eggermont K, Kovaleva V et al., 1995. Plant Cell. 7, 573-588.
2. Duvick JP, Rood T, Rao AG, Marshak DR, 1992. Journal of Biological Chemistry 267, 18814-18820.