We choose to clone the rscSTU genes

of SBW25 for complementation experiments because SBW25 genome is sequenced (in contrast to the hrcU gene of MFN1032) and the rscRST genes present more than 90% of identity with the hrcRST genes of MFN1032. The phenotypes of the resulting strain, MFN1030-pBBR-rscSTU, are summarised in Table 2 (Results are means of at least three independent experiments). D. discoideum growth inhibition and cHA were restored in MFN1030-pBBR-rscSTU, with levels similar to those characteristic of wild type MFN1032. Macrophages lysis was partly restored in MFN1030-pBBR-rscSTU with a level corresponding to half of that of the wild type. Introduction of parental plasmid pBBR1MCS-5 in MFN1030 (MFN1030-pBBR1MC-5 strain) did not modify MFN1030 phenotypes. Table 2 Phenotypes of MFN1032, MFN1030, MFN1030-pBBR- rsc STU and MFN1030-pBBR1MCS-5 BIBF-1120 Phenotypes GSK2245840 clinical trial Strains MFN1032 Rabusertib concentration MFN1030 MFN1030-pBBR-rscSTU MFN1030-pBBR1MCS-5 Cell-associated hemolytic activity (% cHA at 28°C) 69 ± 10 9 ± 7 69 ± 3 12 ± 4 D. discoideum growth inhibition (%) 100 11 ± 3 100 9 ± 2 Macrophages lysis (% LDH release) 40 ± 3 0 24 ± 2 0 Discussion cHA seems dependent on strain origin and not only on T3SS basal part homology All clinical P. fluorescens strains had cHA while environmental strains of

Pseudomonas did not. Nevertheless, hrpU-like operons of SBW25, MF37 (environmental strains) and MFN1032 are highly homologous (more than 90% identity for the HrcR protein) [15]. This was confirmed by complementation of MFN1030 by the SBW25 genes. Even if hrpU-like operon genes are essential to the cHA of MFN1032, as demonstrated by MFN1030 mutant and complementation results, other factors that depend on the origin of the strain, like the T3SS upper part components or the T3SS effectors, are necessary for red blood cell lysis. In C7R12 and SBW25 the functionality or mechanism of T3SS are not fully understood. On the contrary, P. syringae DC3000 has a

functional T3SS with HrpZ as a translocation protein. In our conditions, T3SS of this phytopathogen was not able to induce cHA. This result C225 confirms the inability of HrpZ to cause RBC lysis as described by Lee [31]. Moreover, none of the clinical strains induced HR on tobacco leaves, while C7R12 did. This suggests that the hrpU-like operons have a function in the hemolytic P. fluorescens clinical strains different from that in the biocontrol and phytopathogenic strains, which are able to induce T3SS mediated HR. These findings are in concordance with those of Mavrodi et al. who demonstrated the presence of stable divergent lineages of T3SS in Pseudomonas fluorescens strains [23]. P. fluorescens clinical strains inhibit D. discoideum growth D. discoideum growth inhibition is not a common feature in this species and was rarely found in P. fluorescens environmental strains, even if our panel is too low to be representative. The majority of environmental P.