Exploration of the Use of the “Bialaphos Genes” for Improving Bioherbicide Efficacy

Abstract

We are studying the possibility of altering the virulence and host range of a phytopathogen by transferring and expressing certain genes from the soil-dwelling saprophyte,Streptomyces hygroscopicus, in a plant pathogen model,Xanthomonas campestrispv.campestris(XCC). The genes, referred to herein as the “bialaphos genes,” encode the production of bialaphos, a potent glutamine-synthetase-inhibiting herbicide. This cluster of genes was originally isolated from several biosynthetically blocked mutants ofS. hygroscopicusand constructed into a plasmid vector, pBG9. We have transferred a fragment of the gene cluster into pLAFR3, a plasmid that functions in bothEscherichia coliand XCC and contains a tetracycline resistance marker. The resulting plasmid, named pIL-1, was used to transformE. coliand was incorporated into XCC by conjugation. The transfer of the fragment was confirmed by Southern analysis. The genes were maintained in XCC for about 47 generations in the absence of selection for tetracycline, and no changes in cultural phenotypes were seen in the transformed XCC (XCC/pIL-1). The XCC/pIL-1 cells were pathogenic to their natural hosts cabbage and broccoli, but induced an altered hypersensitive response in the nonhosts bean, pepper, sunflower, and tobacco. The pathogenic host-reaction, induced by the parent XCC, XCC/pLAFR3, and XCC/pIL-1, was a typical black rot disease in inoculated leaves of the two hosts. The nonhost reaction on the nonhost leaves was necrotic hypersensitivity, induced by XCC and XCC/pLAFR3, or the inhibition of hypersensitivity accompanied by only chlorosis at sites inoculated with XCC/pIL-1. We hypothesize that the altered hypersensitivity phenotype may be due to the transformed XCC becoming more compatible with the nonhosts, a step toward acquiring nonhost-virulence, or due to disruption of the normal expression of the hypersensitivity and pathogenicity genes in the transformed XCC. More work is needed to confirm that the introduced genes are being expressed in XCC. With further understanding, this approach may provide a useful model to study host range, virulence, and strain improvement of plant pathogens for biological control of weeds.