ABSTRACT
Erwinia herbicola pv. gypsophilae induces gall formation in gypsophila that is dependent on the existence of a pathogenicity plasmid (pPATHEhg). We previously demonstrated the presence of several hrp genes on this plasmid. By employing transposon mutagenesis and sequencing, a functional hrp gene cluster on the pPATHEhg has now been characterized completely. The hrp genes of E. herbicola pv. gypsophilae are remarkably similar to and colinear with those of Erwinia amylovora and Pantoea stewartii and generally showed 60 to 90% nucleotide or deduced amino acid identity. E. herbicola pv. gypsophilae, however, lacks hrpW, which is present in E. amylovora. Additionally, E. herbicola pv. gypsophilae mutants deficient in harpin production retained pathogenicity and were slightly reduced in their ability to elicit a hypersensitive response (HR) in tobacco. The "disease specific" region, dspA/EB/F, exhibited 60 to 74% identity with the dspA/EB/F loci of E. amylovora and P. stewartii, respectively. Mutations in dspA/E abolished pathogenicity of E. herbicola pv. gypsophilae but not HR elicitation on tobacco. Inactivation of HrpL reduced plant-induced transcription of dspA/E by three orders, indicating Hrp-dependent regulation.
Subject(s)
Bacterial Proteins/genetics , Erwinia/genetics , Genes, Bacterial , Plant Tumors/microbiology , Blotting, Western , DNA Transposable Elements , DNA, Bacterial/analysis , Erwinia/pathogenicity , Molecular Sequence Data , Mutagenesis, Insertional , Nucleic Acid Conformation , Operon , Physical Chromosome Mapping , Sequence Analysis, DNA , Sequence Homology, Nucleic AcidABSTRACT
The host range of the gall-forming bacterium Erwinia herbicola pv. gypsophilae (Ehg) is restricted to the gypsophila plant whereas E. herbicola pv. betae (Ehb) incites galls on beet as well as gypsophila. The pathogenicity of Ehg and Ehb was previously shown to be dependent on a plasmid (pPATH). Transposon mutagenesis was used to generate mutants on the cosmid pLA150 of the pPATH from Ehg824-1. A cluster of nonpathogenic mutations flanked by two IS1327 elements was identified on a 3.2-kb NdeI DNA fragment. All mutants were restored to pathogenicity by complementation in trans with the wild-type Ehg DNA. DNA sequence analysis of the 3.2-kb NdeI fragment revealed a single open reading frame (ORF) of 2 kb as well as a potential ribosome binding site and a putative hrp box upstream to the ORF. The ORF had no significant homology to known genes. Southern analysis also revealed the presence of DNA sequences that hybridized to the ORF in the beet pathovar Ehb4188. This gene was isolated and sequenced. Marker exchange mutants generated in the ORF of Ehb eliminated the pathogenicity of Ehb on gypsophila but fully retained its pathogenicity on beet. Since the putative gene appeared to encode a host-specific virulence factor for gypsophila it was designated as hsvG.
Subject(s)
Bacterial Proteins/genetics , Erwinia/genetics , Erwinia/pathogenicity , Genes, Bacterial , Plasmids/genetics , Amino Acid Sequence , Base Sequence , Cloning, Molecular , Gram-Negative Bacteria/genetics , Gram-Negative Bacteria/pathogenicity , Magnoliopsida/microbiology , Molecular Sequence Data , Plant Diseases/genetics , Sequence Analysis, DNA , Sequence Homology, Amino Acid , Species Specificity , Virulence/geneticsABSTRACT
The pathogenicity-associated plasmid (pPATH) of Erwinia herbicola pv. gypsophilae (Ehg), which is present only in pathogenic strains, contains a gene cluster encoding indole-3-acetic acid and cytokinin biosynthesis. The transposon-reporter Tn3-Spice was used to generate nonpathogenic mutants on two overlapping cosmids, pLA150 and pLA352, of the pPATH. A cluster of such mutations, which spanned 16 kb, mapped approximately 15 kb from the gene cluster involved in phytohormone biosynthesis. Non-pathogenic mutants also failed to elicit the hypersensitive reaction (HR) on tobacco. Pathogenicity and HR were restored concomitantly to these mutants by in trans complementation with wild-type Ehg DNA. A 3.8-kb HindIII DNA fragment that complemented the hrp mutants was sequenced and six complete and two partial open reading frames (ORFs) were identified. Comparison of the deduced amino acid sequences of the eight ORFs showed striking homology and co-linearity with hrp genes of E. amylovora as well as with other plant and mammalian pathogenic bacterial genes encoding proteins of the type III secretion system. Limited DNA sequencing at various sites on the remaining 11-kb region of pLA352 also showed high identity to Hrp proteins of E. amylovora, E. stewartii, and Pseudomonas syringae. These results suggest that hrp genes are mandatory for gall formation by E. herbicola pv. gypsophilae.
