Lipopolysaccharide-Defective Mutants of the Wilt Pathogen Pseudomonas solanacearum

Abstract

Lipopolysaccharide (LPS)-defective mutants of Pseudomonas solanacearum were used to test the hypothesis that differences in LPS structure are associated with the ability or inability of different strains to induce a hypersensitive response (HR) in tobacco. To obtain these mutants, LPS-specific bacteriophage of P. solanacearum were isolated and used to select phage-resistant mutants of the virulent, non-HR-inducing strain K60. The LPS of 24 of these mutants was purified and compared with that of K60 and its HR-inducing variant, B1. Upon sodium dodecyl sulfate-polyacrylamide gel electrophoresis, LPS from K60 and other smooth strains separated into many evenly spaced bands that migrated slowly, whereas LPS from B1 and most phage-resistant strains separated into one to three bands that migrated rapidly. Carbohydrate analysis showed that the LPS of the phage-resistant strains lacked O-antigen sugars (rhamnose, xylose, and N-acetylglucosamine) and could be grouped into (i) those that had all core sugars (rhamnose, glucose, heptose, and 2-keto-3-deoxyoctonate), (ii) those that had no core rhamnose, and (iii) those that lacked all core sugars except for 2-keto-3-deoxyoctonate. The LPS composition of 10 of the rough, phage-resistant mutants was similar to that of the HR-inducing strain, B1, yet none of them induced the HR. Only 2 of 13 mutant strains tested caused wilting of tobacco, and these had rough LPS but produced large amounts of extracellular polysaccharide, unlike most LPS-defective mutants. The evidence did not support the hypothesis that the initial interaction between rough LPS and tobacco cell walls is the determining factor in HR initiation.