Potential applications of Agrobacterium virulence gene promoters in plant-protecting microbial inoculants

Abstract

The concept behind this project was to continue the development of strains of Agrobacterium tumefaciens that were capable of producing pesticidal proteins in response to plant wounding, thereby killing the invading organism. To this end, vir induction was studied in A. tumefaciens and a protocol to elicit the maximum response was developed. In order for this concept to work, it was necessary to determine whether vir induction was occurring at plant wound sites and a method for showing this was developed, the results suggesting that indeed vir induction did occur. The stability of two types of plasmid was also analysed in this bacterium to ascertain how stable the proposed 'microbial inoculant' would be in the field. The results suggested that IncW plasmids should not be used in the final product. The activities of two chitinases from Serratia marcescens were analysed and it was found that both chitinases were effective in controlling some types of fungus. In addition it was found that the expression of chiB in Escherichia coli led to the appearance of a filamentous phenotype at intermediate temperatures. A construct was made that linked the virE promoter to the chiB gene. This plasmid was introduced into A. tumefaciens but did not function as expected. However, other constructs were demonstrated to be inducible although they were only partially successful in controlling the fungi that the strains were assayed against. The Bacillus thuringiensis 5-endotoxin gene, crylA(c), was cloned and various constructs were made to examine the effects of various regions of the native promoter. One constract was made that linked the virB promoter to crylA(c) and this was introduced into A. tumefaciens. The resulting strain was capable of inducible δ-endotoxin expression and was also capable of controlling the larvae of the tobacco homworm, Manduca sexta.