Two Pantoea agglomerans type III effectors can transform nonpathogenic and phytopathogenic bacteria into host-specific gall-forming pathogens.

Pantoea agglomerans (Pa), a widespread commensal bacterium, has evolved into a host-specific gall-forming pathogen on gypsophila and beet by acquiring a plasmid harbouring a type III secretion system (T3SS) and effectors (T3Es). Pantoea agglomerans pv. gypsophilae (Pag) elicits galls on gypsophila and a hypersensitive response on beet, whereas P. agglomerans pv. betae (Pab) elicits galls on beet and gypsophila. HsvG and HsvB are two paralogous T3Es present in both pathovars and act as host-specific transcription activators on gypsophila and beet, respectively. PthG and PseB are major T3Es that contribute to gall development of Pag and Pab, respectively. To establish the minimal combinations of T3Es that are sufficient to elicit gall symptoms, strains of the nonpathogenic bacteria Pseudomonas fluorescens 55, Pa 3-1, Pa 98 and Escherichia coli, transformed with pHIR11 harbouring a T3SS, and the phytopathogenic bacteria Erwinia amylovora, Dickeya solani and Xanthomonas campestris pv. campestris were transformed with the T3Es hsvG, hsvB, pthG and pseB, either individually or in pairs, and used to infect gypsophila and beet. Strikingly, all the tested nonpathogenic and phytopathogenic bacterial strains harbouring hsvG and pthG incited galls on gypsophila, whereas strains harbouring hsvB and pseB, with the exception of E. coli, incited galls on beet.

Characterization of nuclear localization signals in the type III effectors HsvG and HsvB of the gall-forming bacterium Pantoea agglomerans.

HsvG and HsvB, two paralogous type III effectors of the gall-forming bacteria Pantoea agglomerans pv. gypsophilae and P. agglomerans pv. betae, determine host specificity on gypsophila and beet, respectively. They were previously shown to be DNA-binding proteins imported into host and non-host nuclei and might act as transcriptional activators. Sequence analysis of these effectors did not detect canonical nuclear localization signals (NLSs), but two basic amino acid clusters designated putative NLS1 and NLS2 were detected in their N-terminal and C-terminal regions, respectively. pNIA assay for nuclear import in yeast and bombardment of melon leaves with each of the NLSs fused to a 2xYFP reporter indicated that putative NLS1 and NLS2 were functional in transport of HsvG into the nucleus. A yeast two-hybrid assay showed that HsvB, HsvG, putative NLS1, putative NLS2, HsvG converted into HsvB, or HsvB converted into HsvG by exchanging the repeat domain, all interacted with AtKAP-α and importin-α3 of Arabidopsis thaliana. Deletion analysis of the NLS domains in HsvG suggested that putative NLS1 or NLS2 were required for pathogenicity on gypsophila cuttings and presumably for import of HsvG into the nucleus. This study demonstrates the presence of two functional NLSs in the type III effectors HsvG and HsvB.

The type III effectors HsvG and HsvB of gall-forming Pantoea agglomerans determine host specificity and function as transcriptional activators.

Pantoea agglomerans pv. gypsophilae (Pag) elicits galls on gypsophila and a hypersensitive response on beet, whereas P. agglomerans pv. betae (Pab) induces galls on both beet and gypsophila. The pathogenicity of both pathovars is dependent on the presence of a plasmid harbouring type III secretion system (TTSS) components and effectors. The HsvG TTSS effectors of Pag (HsvG-Pag) and Pab (HsvG-Pab) determine the host specificity of both pathovars on gypsophila. Here we describe a novel HsvG homologue, HsvB, which determines the host specificity of Pag and Pab on beet. HsvG requires two direct amino acid repeats for pathogenicity on gypsophila, whereas one repeat in HsvB is sufficient for pathogenicity on beet. Exchanging repeats between HsvG-Pag and HsvB-Pab resulted in a switch of host specificities. Transient expression of GFP-HsvG or GFP-HsvB fusions in gypsophila, beet or melon leaves showed that HsvG and HsvB were localized to the nuclei of host and non-host plants. A yeast one-hybrid assay revealed that a single repeat of HsvG or HsvB was sufficient to activate transcription. By employing random binding-site selection and gel-shift assay HsvG was demonstrated to be a double-stranded DNA-binding protein with an ACACC/aAA consensus binding site. These results suggest that HsvG and HsvB are host-specificity determinants and bear the potential to affect the host transcriptional machinery.