Abstract/Summary

Iron is essential for the growth of nearly all bacteria. However, it is a dangerous metal because it has the ability to catalyse the generation of reactive oxygen species (ROS) through the Fenton chemistry. The oxidation status of iron in the environment is largely determined by pH and oxygen levels, with the poorly soluble ferric (Fe3+) form persisting with high pH and high O2, and the more soluble ferrous form (Fe2+) favoured by low pH or low O2. EfeUOB is a bacterial ferrous-iron transport system, induced by acidic pH aerobically, that exhibits maximal activity under low pH, low-iron conditions. The FeoABC system is also a ferrous-iron transporter, and is widespread in bacteria; it functions during low oxygen conditions, and unlike EfeUOB is not considered to be active aerobically. This study explores the differences in the activities of the two ferrous-iron transporters (Feo and Efe) of E. coli in their responses to environmental factors, in particular O2 and hydrogen peroxide. E. coli mutants devoid of iron-transport systems were employed along with low-copy number plasmids, carrying either efeUOB or feoABC under control of their natural promoters or inducible surrogate promoters. Results showed that the Feo system exhibits weak activity aerobically, and relatively-strong activity aerobically with reductant; in contrast, Efe shows weak anaerobic and strong aerobic activity. These effects were also apparent when the genes were under control of inducible Para or Prha promoters. The difference in O2 dependence of Feo and Efe appeared to be caused by their distinct responses to H2O2, with Feo being activated aerobically by exogenous catalase, and Efe being activated by provision of H2O2. 55Fe uptake experiments supported the growth phenotype data, showing that Efe-mediated iron uptake requires H2O2 whereas Feo-mediated iron uptake requires the absence of H2O2. The results thus indicate that Feo-dependent iron uptake is sensitive to H2O2, rather than O2 as reported previously. It is speculated that the H2O2-sensitivity of Feo results from peroxide-mediated oxidation of the three highly-conserved Cys residues in the permease (or ‘Gate’) domain of FeoB (the predominant Feo component). Indeed, alteration of these residues by site-directed mutagenesis (see below) led to complete loss of activity. Also, a convincing three-dimension model of the FeoB permease domain (based on a Gate-motif-containing nucleoside transporter) indicates that these three Cys residues are closely positioned and line the predicted uptake channel, and are thus well located to act as Fe2+ ligands. The H2O2 dependence of Efe correlates well with the reported peroxidase activity of EfeB (a periplasmic DyP-peroxidase), and the suggested ferrous-oxidation mechanism of EfeU-like ferric-permease iron-uptake systems (as established for Ftr1p/Fet3p of yeast). H2O2 assays clearly showed that the Efe system mediates consumption of H2O2 in a strain lacking catalases and alkylhydroperoxidase, further supporting the suggestion that H2O2 is required by Efe to oxidise iron prior to EfeU uptake. Thus, the combined presence of EfeUOB and FeoABC allows ferrous-iron uptake under distinct conditions relating to peroxide abundance; this would thus explain the need for the combined presence of these two alternate ferrous-iron uptake systems in E. coli strains, and many other bacteria. The requirement for FeoA (small, highly conserved, cytoplasmic SH3 protein) and FeoC (small, cytoplasmic winged helix-turn-helix [4Fe-4S] protein, with no apparent role in transcription control, particularly found to be associated with γ-Proteobacteria Feo systems) along with FeoB (cytosolic G-protein domain coupled to a permease domain), was also explored. FeoA was found to be essential, along with FeoB. Although, FeoC was not essential, it did consistently enhance activity of FeoAB under aerobic conditions, particularly with reductant. FLAG-tagged versions of Feo components were generated to enable estimation of their abundance in response to O2 regime; the addition of the tags did not, generally, affect Feo activity but did enable detection by Western blotting. The results showed that FeoB levels are far higher anaerobically than aerobically – this effect was largely independent of FeoC and the FtsH protease, which contrasts with previous reports in Salmonella. . Nine highly-conserved FeoB residues within the permease domain were altered by SDM. The FeoB-C403S, -C432S, -C677S and -E582Q variants failed to exhibit Feo-enhanced growth under iron restriction indicating an essential role for these residues. These four residues were found to be well positioned to act as Fe2+ ligands in a FeoB model structure. In contrast, the FeoB-E488Q mutant retained good Feo activity, although this was slightly reduced with FeoC aerobically, and slightly raised aerobically. This residue is predicted to be located in a cytosolic loop and thus may interact with the N-terminal G-protein domain. The FeoB-C772S/H773G and -C763S/C764S (residues in the C-terminal cytoplasmic subdomain of FeoB) variants showed greater Feo activity in the presence of FeoC, suggesting that this subdomain interferes with FeoC activation of Feo activity under aerobic conditions and that FeoC may interact with the C-terminal subdomain of FeoB. Further work is required to discover how FeoC activates Feo activity aerobically, how FeoB levels are reduced aerobically in an apparent posttranslational fashion, whether the conserved FeoB Cys residues are indeed subject to oxidation and what factors influence FeoA activation of FeoB activity. The FeoB protein was overexpressed using a baculovirus system, purified to homogeneity in a monodisperse form and subject to crystallisation