Formation of Fe(III)-minerals by Fe(II)-oxidizing photoautotrophic bacteria

It has been suggested that Fe(II)-oxidizing photoautotrophic bacteria may have catalyzed the precipitation of an ancient class of sedimentary deposits known as Banded Iron Formations. In order to evaluate this claim, it is necessary to define and understand this process at a molecular level so that putative Fe-isotope “biosignatures” in ancient rocks can be interpreted. In this report, we characterize the substrates and products of photoautotrophic Fe(II)-oxidation by three phylogenetically distinct Fe(II)-oxidizing bacteria. In every case, dissolved Fe(II) is used as the substrate for oxidation, and there is no evidence for active dissolution of poorly soluble Fe(II)-minerals by biogenic organic ligands. Poorly crystalline Fe(III) (hydr)oxide mineral phases are initially precipitated, and as they age, rapidly convert to the crystalline minerals goethite and lepidocrocite. Although the precipitates appear to associate with the cell wall, they do not cover it entirely, and precipitate-free cells represent a significant portion of the population in aged cultures. Citrate is occasionally detected at nanomolar concentrations in all culture fluids, whereas an unknown organic molecule is always present in two out of the three bacterial cultures. Whether these molecules are released by the cell to bind Fe(III) and prevent the cell from encrustation by Fe(III) (hydr)oxides is uncertain, but seems unlikely if we assume Fe(II)-oxidation occurs at the cell surface. In light of the energetic requirement the cell would face to produce ligands for this purpose, and given the local acidity metabolically generated in the microenvironment surrounding Fe(II)-oxidizing cells, our results suggest that Fe(III) is released in a dissolved form as an inorganic aqueous complex and/or as a colloidal aggregate prior to mineral precipitation. The implication of these results for the interpretation of Fe-isotope fractionation measured for this class of bacteria (Croal et al., 2004) is that equilibrium processes involving free biological ligands do not account for the observed fractionation.