1Faculty of Agriculture, Kochi University, Nankoku, Kochi 783, Japan; 2Iwate Biotechnology Research Centre, Kitakami, Iwate 024, Japan

Background and objectives
Bacterial wilt of tomato, Lycopersicon esculentum, caused by Ralstonia solanacearum (Pseudomonas solanacearum), which is one of the most important bacterial plant pathogens, is a devastating disease in the tropics, subtropics and warm regions throughout the world [1]. Control of the disease with chemicals is mostly ineffective or is not adapted to low-income farming systems. Grafting susceptible tomato cultivars on resistant root-stocks is a means of control of R. solanacearum in Japan.

We developed a monitoring system for the behaviour of R. solanacearum and the disease development in an identical host plant using genetically lux-marked R. solanacearum YN5 and a video-intensified microscopy (VIM) camera. We examined relationships between the behaviour of R. solanacearum and the development of bacterial wilt in grafting tomato plants using R. solanacearum YN5 transformed with pNP126 carrying the luxCDABE operon.

Material and methods
R. solanacearum OE1-1 was transformed with pNP126, which carried the luxCDABE operon derived from Vibrio fischeri and a promoter region derived from Burkholderia glumae genomic DNA. Transformants were screened for their bioluminescence, plasmid maintenance and pathogenicity in tomato plants. A bioluminescent transformant, R. solanacearum YN5 was selected and used in the experiments.

Tomato plants used were cultivar Oogata-Fukuju (OF) for susceptibility and cultivar LS-89 (LS) for resistance; the latter was usually used as the rootstock of the susceptible scions in Japan. Two- or three-leaf stage rootstocks and scions were used for grafting. Rootstocks were topped at 10 mm above their cotyledon, and the cuttings were used as scions. For double grafts, rootstocks were topped at 10 mm below their cotyledon and the second grafts were carried out using the middle scions containing the first and second leaves. OF and LS plants were used for scion/rootstock in combinations of LS/OF, OF/LS, OF/OF and LS/LS, and for top scion/ middle scion/rootstock in combinations of LS/OF/LS, OF/LS/OF, LS/LS/LS and OF/OF/OF. The grafted tomato plants were placed in an incubator at 25C and 100% relative humidity for 4 days and then in a shaded room for 3 days. Tomato plants were then inoculated with R. solanacearum YN5 by the root dipping method and were grown in the water culture at 25C in a growth chamber under 20,000 lux for 14 h per day. Bioluminescence derived from R. solanacearum YN5 was observed under a VIM camera equipped with ARGUS 50 (Hamamatsu Photonics, Japan).

Results and conclusions
In nongrafted OF plants and grafted plants in the combinations OF/OF and OF/OF/OF, bioluminescence was first observed in the roots and the collars and then spread into the upper stems with stronger intensity, and the plants wilted heavily. In non-grafted LS plants and grafted plants in the combinations LS/LS and LS/LS/LS, bioluminescence was weakly observed in the roots and the collars, and the plants did not wilt at all. The percentage of wilting tomato plants in combination of OF/LS was 52.2. Bioluminescence in the LS root-stocks of the plants remained weak, but that in the scions of wilting plants became strong. In the combination LS/OF, bioluminescence in the OF rootstocks became strong, and the plants wilted heavily. In the combination OF/LS/OF, bioluminescence, which was first observed strongly in the rootstocks, remained weak in the middle scions and then became strong in the top scions, and the plants wilted heavily. In the combination LS/OF/LS, bioluminescence was first observed weakly in the rootstocks and then became strong in the middle scions; and the percentage of wilting plants was 52.4. The degree of bioluminescence derived from R. solanacearum YN5 correlated with the bacterial population in tomato plants. These results indicate that latently infected R. solanacearum in the resistant root stock was a causal agent of the disease in the susceptible scions and that resistance of LS results from the suppression of bacterial growth and not the absence of bacterial colonization.

1. Hayward AC, 1991. Annual Review of Phytopathology 29, 65-87.