Today high Fe(II) environments are relegated to oxic–anoxic habitats with opposing gradients of O2 and Fe(II); however, during the late Archaean and early Proterozoic eons, atmospheric O2 concentrations were much lower and aqueous Fe(II) concentrations were significantly higher. In current Fe(II)-rich environments, such as hydrothermal vents, mudflats, freshwater wetlands or the rhizosphere, rusty mat-like deposits are common. The presence of abundant biogenic microtubular or filamentous iron oxyhydroxides readily reveals the role of FeOB (iron-oxidizing bacteria) in iron mat formation. Cultivation and cultivation-independent techniques, confirm that FeOB are abundant in these mats. Despite remarkable similarities in morphological characteristics between marine and freshwater FeOB communities, the resident populations of FeOB are phylogenetically distinct, with marine populations related to the class Zetaproteobacteria, whereas freshwater populations are dominated by members of the Gallionallaceae, a family within the Betaproteobacteria. Little is known about the mechanism of how FeOB acquire electrons from Fe(II), although it is assumed that it involves electron transfer from the site of iron oxidation at the cell surface to the cytoplasmic membrane. Comparative genomics between freshwater and marine strains reveals few shared genes, except for a suite of genes that include a class of molybdopterin oxidoreductase that could be involved in iron oxidation via extracellular electron transport. Other genes are implicated as well, and the overall genomic analysis reveals a group of organisms exquisitely adapted for growth on iron.
Iron is the fourth most abundant element in the Earth's crust, constituting approximately 4.3% of the Earth's continental crust. It is potentially one of the most abundant energy sources on Earth through its role as an electron donor for chemolithoautotrophic growth coupled to Fe(II) oxidation. It is estimated that 3×1011 mol of Fe(II) is released each year through hydrothermal venting in the Earth's oceans . Additional inputs come from low-temperature leaching of Fe(II) from fresh oceanic basalts at crustal spreading centres and active seamounts , and from leaching of Fe(II) from terrestrial soils and rocks in groundwater, freshwater springs, wetlands and riverine inputs . From a microbial perspective, this is a potentially vast food source, yet it is as ephemeral as it is abundant. This is because chemical oxidation of Fe(II) by O2 is rapid at circumneutral pH; at pH 7, the t1/2 of Fe in fully aerated freshwater is <15 min . This is in accordance with the rate equation for iron oxidation:
where k is a rate constant [(8.0±2.5)×1013 min−1·atm−1·mol−2·l−2; 1 atm=101.325 kPa] at 20°C, and PO2 is the partial pressure of oxygen . Thus under neutrophilic conditions (pH 6–7), only microbes that are well adapted to growing at low oxygen concentrations are able to take advantage of Fe(II) as an energy source. A group of microbes has long been associated with high-iron environments, such as iron-containing springs, and suspected of being able to utilize Fe(II) as an energy source. However, until relatively recently, these ‘iron bacteria’ have proven refractory to laboratory culture, thus a good understanding of their metabolism, and general physiology, has remained elusive. For this reason, our knowledge of oxygen-dependent iron oxidation at neutral pH has lagged behind our understanding of other important forms of microbial lithotrophy such as sulfur oxidation, ammonia oxidation or hydrogen oxidation.
Growth on iron is thought to be a very ancient metabolism on Earth . The ability of microbes to grow on iron, under either acidic or circumneutral conditions is spread throughout different phylogenetic groupings of Bacteria and Archaea . Global-scale iron-deposition events, e.g. banded iron formations, were common between 2 and 3 billion years ago, and are conclusive evidence for O2 concentrations in the ocean and atmosphere being far below present levels . The so-called ‘great oxidation event’ (GOE), approximately 2.4 billion years ago, led to atmospheric oxygen levels that are a few per cent of modern levels, but as much as three orders of magnitude higher than they had been for much of the Archaean eon. There is ongoing debate  about fluctuations in oxygen levels both before and after the GOE, but suffice it to say there was a significant period in Earth's history from the mid-late Archaean to the early-mid Proterozoic when O2 concentrations were low (<50 μM or much lower), and Fe(II) concentrations were high (tens to hundreds or even thousands of μM). These periods of Earth's history would have been ideal for the growth of the microaerobic FeOB (iron-oxidizing bacteria) that still exist today.
In modern environments, FeOB are relegated to narrow redox boundary areas where there is a source of anoxic Fe(II)-rich water impinging upon an oxygenated region. Although physically limited in scope, these environments are both common and varied. Examples include freshwater wetlands, slow moving streams, Fe(II)-rich springs, the rhizosphere, saltmarsh and coastal marine sediments and hydrothermal vents . In addition, human-made environments such as water distribution systems, water wells, drainage ditches and mines are commonly inhabited by FeOB, often with detrimental consequences. No matter the specific habitat, oxygen-dependent FeOB face a similar challenge of finding the proper ‘sweet spot’ in opposing gradients of Fe(II) and O2 such that the fluxes of electron donor (Fe2+) and electron acceptor (O2) are optimal for growth. The mechanisms by which they do this are beginning to reveal themselves, but many details are not understood.
Phylogenetic diversity, but morphological unity, among FeOB
A recent review article summarized evidence from a number of phylogenetic surveys of neutrophilic iron mat ecosystems . Together, these studies (references in ) showed that freshwater iron mats contain abundant members of the family Gallionellaceae, which includes the stalk-forming Gallionella ferruginea, as well as non-stalk-forming FeOB such as Sideroxydans spp. A novel class of Proteobacteria, the Zetaproteobacteria, are commonly found at iron mats associated with marine hydrothermal vents; the stalk-forming iron-oxidizer Mariprofundus ferrooxydans is a representative member of this group . However, members of the Gallionellaceae have not been reported from marine environments. Subsequent to the 2010 review , several more reports from freshwater habitats [9–15] and marine environments  have confirmed these observations. Importantly, the realm of Zetaproteobacteria known to be involved in iron oxidation has been expanded into saltmarshes and coastal environments unassociated with hydrothermal venting [17,18].
A unique feature of many FeOB are the distinctive extracellular structures such as biomineralized sheaths and stalks that are produced during growth. Interestingly, these morphotypes appear to occur independently of phylogeny. A nice illustration of this is based on ultrastructural studies of two stalk-forming FeOB: the marine M. ferrooxydans and a freshwater strain related to G. ferruginea . Both organisms produce twisted helical stalks composed primarily of ferrihydrite organized into nanofibrils (Figure 1). Use of high-powered imaging techniques coupled to spectroscopy has revealed the stalks of both Mariprofundus and Gallionella contain an organic matrix, probably with a polysaccharide component, that helps to co-ordinate the precipitation of iron oxides [20–22]. Morphologically, both cells are short curved rods with consistent concave and convex sides (Figure 1). The stalks are excreted, and iron oxidation occurs, within the concavity of the cell. Although the mechanisms of iron oxidation and nanofibril production are not understood, it is clear that the molecular apparatus for iron oxidation and the control of stalk formation, i.e. the biochemical machinery, must be localized within the cell .
Ultrastructural analysis of
M. ferrooxydans PV-1 and a freshwater stalk-forming FeOB
A second striking example of melding of form and function is in sheath formation by two uncultivated iron-oxidizers that live in either marine or freshwater habitats. Leptothrix ochracea has long been recognized as one of the most visually abundant freshwater iron-oxidizers. It forms a tubular sheath composed largely of iron oxides that also has an organic component . Through the use of cultivation-independent molecular techniques, it was shown recently that L. ochracea is phylogenetically related to cultured heterotrophic Leptothrix species which also produce sheaths and can oxidize iron or manganese . Unlike other Leptothrix species, L. ochracea requires high concentrations of Fe(II) for growth, and is not capable of heterotrophic growth. A sheath-forming organism that looks virtually identical with L. ochracea by light microscopy has been observed in iron mats associated with marine hydrothermal vents . Like L. ochracea, it grows as a filament of cells inside the sheath, and the cells produce large quantities of sheath, most of which are empty, apparently as a by-product of iron oxidation. An analysis of these organisms from Loihi Seamount revealed that this group of sheath-formers are not related to L. ochracea (a member of the Betaproteobacteria), but are instead related to M. ferrooxydans and the Zetaproteobacteria (E. Fleming, R. Davis, S. McAllister, C. Chan, C. Moyer, B. Telso and D. Emerson, unpublished work). This example of the remarkable convergence of form and function for genetically distinct FeOB suggests that the physicochemical requirements for growing on Fe(II) can lead to a remarkable consistency in form.
Functionally, the extracellular structures that FeOB produce probably serve at least two purposes. First, they are a mechanism that the cells can use to co-ordinate the rapid precipitation of iron oxyhydroxides owing to the oxidation of Fe(II) to Fe(III); this can prevent the cells from getting entrapped inside an insoluble iron oxide crust. Secondly, being attached to a stalk, or living within a sheath, allows these FeOB to optimally position themselves in dynamic gradients of oxygen and Fe(II). However, not all FeOB produce organized extracellular structures. To date, many of the isolates of freshwater FeOB form only particulate oxides with no distinct morphotype . For strains isolated from the rhizosphere, this may make sense, since the compact soil matrix probably impedes sheath or stalk production. It is likely that these organisms produce an exopolymer that helps to prevent the cells from becoming encrusted in iron oxides. Perhaps these exopolymers have characteristics similar to the ones produced by Gallionella and Mariprofundus, but so far they have not been characterized. At present, we do not know the relative abundances of FeOB that produce extracellular structures compared with those that do not in natural microbial iron mats. It is interesting that, to date, from marine environments, only appendaged FeOB have been observed in pure culture or enrichment culture, leaving open the possibility that marine FeOB have a greater requirement for formation of extracellular structures.
The production of extracellular structures, either stalks or sheaths, must accrue significant benefit to the cell, because production of an exopolymer, and the co-ordination of an extracellular structure, entails a metabolic cost. At present, it is not possible to reliably estimate the metabolic costs of exopolymer production for FeOB; however, compared with the overall metabolism of the cell, it is likely that these are moderate. In theory, the energetic costs of producing polysaccharides are significantly less than that of producing proteins . For example, experiments with Escherichia coli have shown that the ATP cost of polyglucose synthesis is 12.4 mmol of ATP per g of polyglucose compared with 39.1 mmol of ATP per g of protein . We know virtually nothing about specific energetic costs of producing various biomolecules in chemolithotrophs, or what strategies an organism might employ, for example selecting for specific glycosides or glycosidic bonds, that would minimize the energetic costs. Presumably, since FeOB live on such a low-energy substrate, they are adept at minimizing the associated energetic costs with production of extracellular structures.
Potential mechanisms for iron oxidation
The detailed mechanisms of how microaerobic bacteria oxidize iron at circumneutral pH are poorly understood. A general consensus is that iron oxidation requires the transfer of electrons from the site of Fe(II) oxidation outside the cell, or at least from the outer membrane, to the cytoplasmic membrane where the electron-transport chain is located. In acidophilic iron-oxidizing bacteria, there is both genetic and biochemical evidence for this . Similar mechanisms have been proposed for photoferrotrophy  and for nitrate-dependent iron oxidation . The difficulties of working with the microaerobic FeOB have made it challenging to get at these mechanistic questions. Specific challenges include low growth yields coupled with massive production of insoluble iron oxyhydroxides, obligate requirements for Fe(II) for growth and inability to plate the cultures to select for mutants or develop genetic systems.
Genomic analysis is one promising method for beginning to understand how FeOB oxidize Fe(II). Recent work by Liu et al.  capitalized on the availability of genomic information from the two freshwater FeOB Gallionella sp. ES-2 and Sideroxydans sp. ES-1, which revealed the presence of homologues of the MtrA, MtrB and CymA genes. These three genes are known to be involved in iron reduction in Shewanella, forming part of a complex that carries out extracellular electron transfer, but, in this case, it is transfer of electrons from the cytoplasmic membrane to the outer membrane where iron reduction occurs . Liu et al.  synthesized the homologous genes from ES-1. The MtrA homologue was cloned into an mtrA-deletion mutant in Shewanella oneidensis MR-1, and was able to partially complement iron reduction in these cells. This same homologue, designated MtoA (to distinguish it from MtrA), was overexpressed and purified from MR-1, and the protein was shown to oxidize Fe(II) in vitro . Although this work does not prove that these genes are expressed and/or act as the primary iron-oxidation pathway in Sideroxydans, this is the first mechanistic evidence for such an activity. This also demonstrates the potential functional overlap between pathways involved in iron reduction with those involved in iron oxidation .
The first analysis of the genome of the marine iron-oxidizer M. ferrooxydans was published recently . The 2.86 Mb genome contains approximately 2920 genes. The genome of M. ferroxydans is consistent with that of a chemolithoautotroph. It contains genes for both Form I and Form II ribulose bisphosphate genes involved in CO2 fixation. It does not contain genes known to be involved in sulfur oxidation, consistent with growth studies showing that it cannot utilize reduced sulfur compounds as growth substrates. The most interesting questions relate to its utilization of Fe(II) as an electron donor. It does not contain homologues of the Mtr genes, discussed above, nor does it have homologues of the genes Rus or CyaA, or those encoding the cytochromes Cyt572 or Cyt579 known to be involved in iron oxidation in acidophilic bacteria [28,34,35]. These latter genes are not found in the genomes of Sideroxydans or Gallionella either (D. Emerson, unpublished work). Interestingly, the M. ferrooxydans genome does contain a cluster of genes that include a molybdopterin oxidoreductase, as well as at least two cytochromes and other genes involved in electron transport (Figure 2). The molybdopterin oxidoreductase family of genes include a number of different reductases and oxidases, including an arsenite oxidase. It has been proposed that this group of genes can form an alternative respiratory complex III, essential for oxidative phosphorylation . Interestingly, this gene cluster shows significant homology with the same cluster of genes in two freshwater iron-oxidizers, Sideroxydans ES-1 and Gallionella ES-2. To date, this is the only instance of a gene cluster encoding redox activity that has consistently high homology between M. ferrooxydans PV-1 and the freshwater FeOB. The PV-1 genome also encodes a cbb3-type oxygen reductase (cytochrome c oxidase), but lacks a caa3-type cytochrome c oxidase. The cbb3 oxygen reductase has a much higher affinity for O2 than the caa3-type enzyme, and is commonly found in microaerophiles . The same pattern is true for both Sideroxydans and Gallionella (D. Emerson, unpublished work).
A model for iron oxidation in
M. ferrooxydans PV-1
Genomic analysis shows that these FeOB are well adapted for growing under microaerobic conditions in Fe(II). It is also becoming apparent that the mechanism for acquiring energy from Fe(II) oxidation almost certainly will utilize some form of extracellular electron transfer. At present, it does not appear that there any universally conserved mechanisms for iron oxidation; instead, there may be consistent modalities that have variable constituents that are tuned to the particular environmental conditions of an organism's habitat . The overall degree to which specific mechanisms are conserved among the neutrophilic oxygen-dependent iron-oxidizers remains to be determined.
Electron Transfer at the Microbe–Mineral Interface: A Biochemical Society Focused Meeting held at University of East Anglia, Norwich, U.K., 2–4 April 2012. Organized and Edited by Jim Fredrickson (Pacific Northwest National Laboratory, U.S.A.), David Richardson (University of East Anglia, U.K.) and John Zachara (Pacific Northwest National Laboratory, U.S.A.).
This work has been funded in part through the National Science Foundation [grant number IOS 0951077], the Office of Naval Research [grant number N00014-08-1-0334] and National Aeronautics and Space Administration (NASA) Experimental Program to Stimulate Competitive Research (EPSCoR) [grant number 10-EPSCoR-0005] administered through the Maine Space Grant Consortium.