Plant Pathology Department, University of Florida, Gainesville 32611, USA;
1Present address: Division of Plant Pathology, Indian Agricultural Research Institute, Pusa, New Delhi 10012, India

Background and objectives
Some strains of Xanthomonas campestris pv. malvacearum (Xcm) in West Africa are virulent on all commercial cotton lines in the USA, including the entire set of host differential lines used to distinguish 19 races of the cotton blight pathogen. There is some evidence to indicate that some combinations of cotton R genes provide low levels of resistance to the African Xcm strains. Use of a combination of R genes in the same plant, also called 'pyramiding', has been effective against North American strains of Xcm. It may be desirable to combine weak or defeated R genes in cotton, particularly those R genes with measurable residual resistance, with strong R genes that are not yet defeated, such as B12 [1]. The problem is how to assay and select recombinant plants carrying both the strong and the weak R genes.

The molecular basis of race specificity of Xcm has been shown to be due to gene-for-genes (plural) recognition by cotton R genes of several avirulence/pathogenicity (avr/pth) genes [2]. Agrobacterium tumefaciens was recently shown to be useful in transient expression assays of a variety of avirulence genes from both Xanthomonas and Pseudomonas. In the process of using A. tumefaciens to transiently express different Xcm avr/pth genes at very high levels in cotton plant cells, we were surprised to detect apparent residual resistance activity of some cotton R genes against an avr/pth gene cloned from an African Xcm strain that had 'defeated' the R genes. The residual resistance appeared to be gene-for-gene specific. Transient expression assays using individual avr/pth genes should allow detection of weak individual R genes in complex genetic backgrounds and facilitate pyramiding of weak or defeated R genes.

Materials and methods
An avr/pth gene, pthN, was previously cloned from African strain XcmN and was found to contribute to the pathogenicity of XcmN on cotton [3]. However, pthN was not properly assayed for avirulence on S-295 because there are no known strains of Xcm virulent on 8-295. Normally, assays for avirulence are performed by placing the genes to be tested in a virulent strain background and then inoculating test plants. Avirulence genes cause the normally virulent strain to elicit a hypersensitive response (HR). In the present study we used A. tumefaciens AG L-1 as the background test strain. pthN was cloned onto a T-DNA transient expression vector with a strong promoter and mated into AG L-1. AG L-1 by itself gives a very faint chlorotic phenotype several days after inoculation into cotton leaves. Cultures of AGL-1 carrying pthN were grown in a medium suitable for vir gene induction, resuspended to a final OD600 of 2.4, and then used for inoculation of cotton leaves. The plants were scored for presence or absence of the HR 2 days after inoculation into cotton leaves.

Results and conclusions
AG L-1 with pthN driven by a strong promoter caused an HR on 8-295 and 101-102B. However, it did not cause an HR on AcB2 or Acala 44. The results indicate weak recognition of pthN from Race 19 African strain XcmN by cotton R genes in 101-102B, defeated by XcmN. The A. tumefaciens system provided a more sensitive assay for weak R genes or those with residual effects than could be obtained using natural pathogenic strains and races.

1. Wallace TP, EI-Zik KM, 1989. Crop Science, 29, 1114-1119.
2. DeFeyter R, Yang Y, Gabriel DW, 1993. Molecular Plant-Microbe Interactions 6, 225-237.
3. Chakrabarty PK, Duan YP, Gabriel DW, 1997. Phytopathology 8, 1160-1167.