A EL-MALEH¹, C KOEHLER², J HUANG², AM MYERS² and E HACK¹

¹Department of Biological and Nutritional Sciences, University of Newcastle, Newcastle upon Tyne NE1 7RU, UK; ²Department of Biochemistry, Iowa State University, Ames, IA 50011, USA

Background and objectives
Maize with Texas male-sterile cytoplasm (cms-T) is highly susceptible to the southern corn leaf blight disease caused by race T of Cochliobolus heterostrophus (anamorph Bipolaris maydis). This pathogen produces a host-specific toxin, BmT toxin, which binds to the T-URF13 (URF13) protein of cms-T maize mitochondria, permeabilizing the inner mitochondrial membrane. The T-URF13 protein is the product of a gene unique to cms-T maize mitochondria, T-urf13. Expression of the T-urf13 gene in Escherichia coli, yeast, insect cells and tobacco can confer sensitivity to BmT toxin or to methomyl, an insecticide that mimics the effects of BmT toxin [1, 2]. Our aim was to find out whether tobacco genetically engineered to contain T-URF13 in its mitochondria would be a host for C. heterostrophus race T, which would imply that toxin sensitivity determines disease susceptibility even in a plant that is very different from the natural host.

Results and conclusions
The sequences coding for the mitochondrial targeting peptide of Neurospora crassa ATP synthase subunit 9 and T-urf13 were fused, as in experiments with yeast [3]. Expression was controlled by a double CaMV 35S promoter and the AMV coat-protein RNA leader. The chimaeric gene was introduced into tobacco using Agrobacterium tumefaciens. Primary transgenic tobacco lines were used as female parents in outcrosses, whose progeny were used for these experiments.

For pathogenicity tests, a small amount of C. heterostrophus race T mycelium was placed on leaf surfaces of cultured tobacco shoots. On some transgenic shoots, a halo appeared around the inoculation point and slowly enlarged. Eventually, fungal mycelium grew from the lower surface of the leaf and secondary infections occurred on the stems. On non-transgenic shoots, no host-pathogen interaction was evident. Susceptible shoots were cloned for further investigations. The presence in them of the T-urf13 gene was confirmed by PCR analysis.

When the fungus was allowed to grow on medium supporting cultured shoots, susceptible shoots were systemically bleached, indicating systemic spread of toxin, while control shoots were unaffected. To test the effect of toxin directly, shoot culture protoplasts were treated with partially purified BmT toxin and then with the fluorescent dye DiOC6, which accumulates only in coupled mitochondria. Toxin concentrations that eliminated mitochondrial fluorescence of protoplasts from susceptible shoots had no effect on control protoplasts.

Shoots from a susceptible line were rooted in soil and inoculated with race T culture filtrate. The treated areas became bleached after about 16 h and necrotic at 24 h. There was no evidence of systemic spread of the toxin. This localized response to toxin suggests that tobacco plants can resist its effects. There was no visible effect of treating control plants with the culture filtrate, or toxin-sensitive plants with culture filtrate of a toxin-non-producer isolate. Attempts to infect potted plants with the fungus were unsuccessful.

Our results show that expression in tobacco of a gene coding for the maize mitochondrial T-urf13 gene fused to a mitochondrial targeting peptide can confer sensitivity to BmT toxin and, moreover, a degree of susceptibility to the fungus C. heterostrophus race T, which is normally a pathogen of maize. Susceptibility was only observed, however, in shoot cultures. This could be because these cultures have a generally low ability to resist pathogens, or because they are more sensitive to the toxin than plants transferred to soil.

References
1. Siedow JN, Rhoads DM, Ward GC, Levings CS III, 1995. Biochimica et Biophysica Acta 1271, 235-240.
2. Chaumont F, Bernier B, Buxant R et al., 1995. Proceedings of the National Academy of Sciences, USA 92, 1167-1171.
3. Huang J, Lee S-H, Lin C et al. 1990. EMBO Journal 9, 339-347.