Figure 1.
Genomic organization of nhaAR region.
(A) Genomic representation of nhaAR and other operons under NhaR regulation in K-12 E.coli strain from http://www.ecocyc.org. All transcription or translation regulators are indicated. (B) GC percent along IAI39 nhaAR region (black line) and mean core genome GC percent (gray line) (C) Organization in various modules of nhaAR region. The modular organization of the region was defined using synteny breaks between 10 pathogenic and commensal E. coli and 2 clades strains from various phylogenetic groups including K-12 and H617 (two group A commensal strains), IAI1 and O26:H11 (two group B1 commensal strains), B367 (a group D commensal strain), 042 (an enteroaggregative group D, E2348/69 (an enteropathogenic group B2 strain), 536 (an extra-intestinal pathogen group B2 strain), O157:H7 Sakaï (an enterohemorrhagic group E strain). The two Escherichia clades (C) used were M863 (CI) and E1118 (CV). Five homologous modules have been defined, nhaAR operon being the third. Dark green: Fragmented or complete hokC toxin-antitoxin system coding genes; turquoise: Fragmented or conserved region associated with putative cerebroside sulfatase coding genes; blue: nhaAR region; pink: Fragmented or conserved region associated with putative adhesin/fimbrial like protein coding genes; yellow: Fragmented or conserved Type 3 secretion system coding genes. IS are also indicated.
Figure 2.
Multiple gains and losses of modules around nhaAR operon.
A parsimonious scenario of gains and losses of the 5 modules defined in figure 1 is presented along the phylogenetic tree of the strains. – indicates losses and + acquisitions (colors of modules as in figure 1).
Figure 3.
MLST and nhaA phylogenetic trees for 121 strains of E. coli.
The trees were reconstructed from (A) multi-locus sequence typing of 8 partial housekeeping genes from the Pasteur scheme [7] representing the species phylogeny and (B) from the nhaA sequences using PHYML [19]. Bootstraps values are indicated. Strains studied and belonging to phylogenetic group B2 (red boxed) are indicated. Branches separating the B2 phylogenetic group strains from the other group strains are indicated in blue.
Figure 4.
Non B2 grew faster than B2 in high pH high osmolarity.
Boxplots of the doubling times (DT) in minutes of 12 B2 and 10 non-B2 representative strains of E. coli in LB, pH 8.5 with 350 mmol/L of sodium. We found a significant difference between B2 strains and non B2 strains using a Welch test (p = 0.001).
Figure 5.
Recombinant nhaA B2 strains have a lower virulence.
Lines represent the mean survival of OF1 mice after subcutaneous injection of 108 cells of the following strains: solid line: K-12 MG1655; dotted line: B2 strains lacking recombination in nhaA (i.e CFT073, 536, S88, RS218) and dashed line: strains showing evidence of recombination in nhaA region (i.e ED1a, E2348/65, SE15, B671, H001, TA103, TA435).
Figure 6.
Impact of nhaAR operon on virulence.
Lines represent the survival of OF1 mice after subcutaneous injection of 108 cells of the following strains. In (A), (B) and (C) black solid lines, K-12 MG1655 and red solid lines, strain 536. In (A) orange, blue and green solid lines, mice injected with mutants 536ΔnhaR, 536ΔnhaA and 536ΔnhaAR, respectively. In B (C) dashed-dotted lines, complemented mutants 536ΔnhaA pGCnhaAR (536ΔnhaAR pGCnhaAR), dashed lines, complemented mutants 536ΔnhaA pGCnhaA (536ΔnhaAR pGCnhaA), solid lines, 536ΔnhaA (536ΔnhaAR) and dotted lines, complemented mutants 536ΔnhaApGC (536ΔnhaAR pGC).