Iron metabolism

A Knowledge of Iron Metabolism is essential for a complete understanding of microbial growth and survival. In microbes, iron is necessary for processes such as electron transport, nitrogen fixation, removal of toxic forms of oxygen, synthesis of DNA precursors, tRNA modifications, and syntheses of certain amino acids and tricarboxylic acid cycle intermediates; some bacteria can even oxidize iron to obtain energy. Iron is such a valuable and versatile nutrilite that, of all the organisms on earth, only certain lactic acid bacteria can manage without it. 

The utility of iron stems primarily from the fact that its redox couple (FeII/FeIII) can have a range of potentials of _300 to _700mV, depending on the nature of the ligands and the environment surrounding the coordinated iron ions. A major consideration in iron metabolism is that, although iron is abundant on the surface of the earth, it is relatively unavailable. At neutral pH and in an oxidizing environment, which includes most common microbial habitats, iron exists in the _3 valence state and, as such, is extremely insoluble. For a microbe inhabiting an animal host, iron availability is restricted by the presence of iron storage and transport proteins, such as ferritin, lactoferrin, and transferrin. Thus, although relatively low concentrations of iron (5_M) are generally sufficient for maximum growth yields, bacteria frequently find themselves in iron-deficient environments and must devote a significant amount of their resources to obtaining this metal. Also important is the fact that it is possible for bacteria to suffer from iron overload and iron assimilation must, therefore, be precisely controlled. This article covers prokaryotic iron metabolism only. Descriptions of fungal iron assimilation are included in the reviews by Guerinot (1994) and Leong and Winkelmann (1998). I. IRON UPTAKE Bacteria can obtain iron from a variety of sources but, regardless of its origin, iron must be transported through the several microbial surface layers to reach the cytoplasm. For gram-negative bacteria, these layers minimally include an outer membrane, a monolayer of peptidoglycan, and an innermost cytoplasmic membrane. The peptidoglycan cell wall is located in the periplasm, the space between the outer and inner membranes. 

Gram-positive cells, in contrast, may contain only a thick, highly cross-linked peptidoglycan cell wall external to the cytoplasmic membrane. Agreat variety of iron transport systems can be distinguished on the basis of the iron source and the form in which iron is mobilized, but, in general, they follow a pattern (Fig. 55.1). Thus, for iron complexed to a carrier, first, passage through the outer membrane requires an outer membrane receptor protein whose synthesis is iron-regulated. A receptor protein has specificity for a given iron–carrier complex, binds that complex only, and is sometimes synthesized in abundance only when that particular iron complex is available. In the two best-studied cases, receptor proteins have been shown to be gated pores. Second, to permit entry of complexed or free iron into the periplasm, cytoplasmic membrane proteins TonB, ExbB, and ExbD are required. These proteins function as a group to utilize the electrochemical potential of the cytoplasmic membrane to open the gated receptor pores, permitting iron and iron-chelates to pass into the periplasm. Generally, each bacterium has just one set of TonB/Exb proteins, capable of interacting with multiple receptors, but Vibrio cholerae and Pseudomonas aeruginosa have two distinct TonB systems. TonB is anchored in the cytoplasmic membrane but spans the periplasmic space, so as to be able to physically contact all receptors that require it. ExbB has three transmembrane domains and is present primarily in the cytoplasm, while ExbD has only one transmembrane segment and extends into the periplasm. Determination of the details by which TonB and its accessory proteins act is an area of active research. A straightforward mechanism would have ExbB sense the proton motive force and, with ExbD, use it to induce a conformational change in TonB, that, in turn, could bring about, by physical contact, allosteric changes in outer membrane receptors. 

TonB-dependent receptors have a heptapeptide (TonB box) near their amino termini that is thought to serve as a recognition sequence (Table 55.1). Third, transport across the cytoplasmic membrane employs members of the ABC supertransporter family. In this case, transport components consist of a peripheral cytoplasmic membrane ATPase, present in two copies and possessing a defining ATP binding site motif, and two hydrophobic cytoplasmic membrane proteins. These latter proteins can be similar but distinct polypeptides, one large fusion protein, or a homodimer. In addition, iron transporters have a component exterior to the cytoplasmic membrane. For gram-negative cells, the component is a periplasmic binding protein, which recognizes the iron substrate and presents it to the cytoplasmic membrane complex. In gram-positive cells, the periplasmic protein is, in some cases, replaced by a lipoprotein tethered in the cytoplasmic membrane but protruding into the peptidoglycan region. In summary, entry of iron into the cytoplasm of gramnegative bacteria requires an outer membrane receptor protein, a TonB system, and an ABC transporter. Passage through the outer and cytoplasmic membranes depends on the proton motive force and ATP, respectively. Outer membrane receptor proteins bind just one specific iron complex, TonB systems have broad specificity, and ABC transporters often recognize several iron-complexes, provided these are structurally related. A. Siderophore-mediated systems A major mechanism by which bacteria obtain iron is through the production and secretion of small (600–1000 Da), iron-chelating molecules termed siderophores. Many different siderophores have been isolated and characterized but they can generally be classified as being either of the catecholate or hydroxamate type (Fig. 55.2). 

Siderophores are synthesized in iron-deficient environments; after release from cells, siderophores bind iron and the siderophore– iron complex is subsequently internalized using the general type of transport system that has been described. Siderophore biosynthesis is an interesting and growing area of research. Siderophores, generally, are built up of amino acids and hydroxy acids and, although they contain amide bonds, their synthesis does not involve ribosomes. Instead, a thiotemplate process with strong similarities to that employed for synthesis of certain peptide antibiotics is used. Nonribosomal peptide synthetases link activated precursors to the enzyme-bound cofactor 4_-phosphopantetheine (P-pant). The thioesterified precursors are then covalently linked by amide bonds in the sequence determined by their position on the synthetase. The synthetase enzymes can be huge; in some cases, molecular weights of greater than 350,000 have been reported. Also, studies on siderophore biosynthesis resulted in the discovery of a new class of enzymes, the phosphopanthetheinyl transferases, which donate the P-pant to the peptide synthetases. This essential posttranslational modification of synthetases permits them to bind activated precursors and then join the precursors together to yield the final product. Other molecules, which do not meet the rigorous definition of siderophores, can also provide iron to cells by means of specific outer membrane receptor proteins acting in concert with binding proteindependent ABC transporters. Citrate, when present in environmental concentrations of greater than 0.1mM, is one such molecule, as are dihydroxybenzoic acid, a precursor of the common siderophore enterobactin, and dihydroxybenzoylserine, a breakdown product of enterobactin (Fig. 55.2). Release of iron from siderophores is not well understood. Ferrisiderophores enter the cytoplasm and free siderophores, or modified versions, are released back into the medium. Because siderophores (i) bind FeII less avidly than FeIII and (ii) the cytoplasm is a reducing environment, enzymatic reduction of iron is thought to be the release mechanism. In vitro experiments have demonstrated the presence of a variety of ferrisiderophore reductase activities that employ reduced flavins to convert iron to the ferrous state. These enzymes have broad specificity and their abundance is not affected by iron availability. 

The Fes protein of E. coli (ferrienterobactin esterase) is the one case where a specific iron-regulated protein is necessary for release of siderophore-bound iron. This enzyme hydrolyzes Ent but this esterase activity may be secondary to its primary reductase role. How newly synthesized siderophores are released is similarly unclear for most microbes. A Mycobacterium smegmatis mutant has been isolated that appears to be defective in siderophore secretion: the defective protein has no homologs in gene banks. For E. coli, the entS (ybdA) gene product encodes a protein with strong homology to proton motive force-dependent membrane efflux pumps. Secretion of the siderophore enterobactin but not its breakdown products is reduced in cells bearing entS mutations. The commonly held idea that siderophore secretion is facilitated by having the siderophore biosynthetic enzymes form a complex that is loosely associated with cytoplasmic membrane remains unproven. In fact, several recent studies argue that the enzymes do not form a complex in vivo. B. Uptake of ferrous and ferric ions Ferrous iron transport is useful to microbes capable of inhabiting oxygen-restricted environments, such as swamps, intestines, and marshes, or acidic locales, where reduced iron is stable and soluble. An E. coli ferrous iron transport system ( feo) has been identified; two proteins, FeoA and FeoB, participate in FeII uptake. FeoB is a large cytoplasmic membrane protein with a nucleotide binding site, suggesting that ATP hydrolysis is the energy source for transport. The FeoA protein is small (less than 10,000 Da) and of unknown function. Salmonella typhimurium, a facultative anaerobe like E. coli, has feo genes and Methanococcus jannischii has a homolog of feoB. 

The latter finding is of interest in that M. jannischii is classified in the domain Archaea, not in the domain Bacteria. Lastly, certain aerotolerant bacteria, such as the gram-positive organism Streptococcus mutans, obtain iron by using a reductase exposed on the cell exterior to convert surface- bound FeIII to FeII. A ferrous ion transporter then delivers the iron to the cytoplasm. A ferric iron acquisition system of the ABC type is present in a number of gram-negative genera including Serratia, where it was discovered and termed Sfu type transport, Haemophilus, Yersinia, Actinobacillus, and Neisseria. A periplasmic binding protein accepts FeIII and presents it to the cytoplasmic components of the transporter, which internalize the iron. Ferric iron transporters can also function in concert with uptake systems that have outer membrane components. In these cases, they become the terminal portion of the assimilation path. Thus, iron is removed from transferrin and lactoferrin at the outer membrane (see following), a process which requires a receptor and active TonB system, and then translocated into the cytoplasm by the ferric transporter. Also, iron can be transferred from citrate to a periplasmic binding protein for entry into the cytoplasm.  C. Uptake of iron from heme The vast majority of iron in animals is present intracellularly as heme (Hm). 

Heme, in turn, serves as a prosthetic group of proteins, primarily hemoglobin (Hb), but also myoglobin and Hm-containing proteins such as cytochromes. Cell lysis is necessary for release of these proteins. Freed Hb, such as that present following hemolysis of erythrocytes, is bound by the serum glycoprotein haptoglobin (Hp). Hb not bound by Hp can oxidize, in the process releasing Hm from globin. This extracellular Hm is bound by the plasma protein hemopexin and, less specifically, by serum albumin. Therefore, extracellular Hm, available as a possible source of iron (and Hm), occurs infrequently by itself as it is usually bound to Hb, Hb-Hp complexes, and hemopexin. In no case has a siderophore been able to scavenge iron from Hm. However, as will be described, each Hm source is capable of being utilized by one or another group of bacteria. These uptake systems vary greatly. An additional complication is that there is such variability among strains of a given species that it is often not possible to list the systems present in a given species. For instance, E. coli lab strains cannot use either Hm or Hb as iron sources but pathogenic E. coli strains can often use both. Iron assimilation pathways that recognize free Hm resemble those for iron-siderophore complexes; they require (i) a single, ligand-specific outer membrane receptor protein that is TonB-dependent and (ii) for cytoplasmic membrane passage, an ABC transporter. For several of these systems, including those of Yersinia enterocolitica, Vibrio cholerae, and Shigella dysenteriae, it has been deduced that the entire heme molecule enters the cytoplasm. (In these cases, heme supported porphyrin growth requirements, as well as providing iron.) Little is known regarding the intracellular release of iron from Hm. HemS of Y. enterocolitica may provide this function, and HmnO of Corynebacterium diphtheriae, one of only a few gram-positive species able to assimilate Hm, may be a Hm oxygenase activity. Several genera can remove Hm from Hb. Neisseria and Haemophilus spp. have TonB-dependent Hbbinding proteins in their outer membranes. 

Remarkably, both of these genera have additional TonB-dependent receptors that function in iron acquisition from Hb- Hp complexes. Each of these Hb-Hp receptors may consist of two different proteins. A different mechanism for the initial steps in obtaining iron from Hm or Hb is found in Serratia marcescens. This organism uses an ABC transporter to secrete a small protein (HasA) that functions as a hemophore (Hbp). That is, HasA is an extracellular Hm binding protein that is necessary for uptake of Hm, either free or bound to Hb. It shuttles its bound Hm to an outer membrane receptor (HasR). E. coli strains harboring the virulence plasmid pColV-K30 also secrete a Hbp. Unlike HasA, Hbp is autotransported out of the cell and is bifunctional. It has protease activity, degrading Hb as well as binding and transporting the released Hm. A chromosomally encoded outer membrane receptor protein (ChuA), required for the well-known E. coli pathogen O157:H7 to use Hb or Hm, may be the Hbp-Hm receptor. Only Haemophilus strains are known to utilize Hm associated with hemopexin. The system is not well characterized but three genes are required and one appears to encode a large secreted Hbp (HxuA). HxuA binds Hm-hemopexin, removes the Hm, and carries Hm to an outer membrane receptor. The other two genes encode proteins concerned with the secretion of HxuA. Like all Haemophilus Hm transport systems (free Hm, Hb:Hp, Hm:albumin), the Hm-hemopexin system requires a functional TonB protein. D. Acquisition of iron from transferrin and lactoferrin Transferrin (TF) and lactoferrin (LF) are extracellular iron transport molecules present in the fluids of many vertebrates; each of these related glycoproteins can bind two ferric ions. Many siderophores are capable of removing iron from these glycoproteins. Neisseria and Haemophilus produce no siderophores, however, and, instead, they have specific TonB-dependent outer membrane receptors that bind these transport proteins. An ABC transporter necessary for all nonheme iron uptake pathways (TF, LF, or iron chelates) is present in Neisseria; iron from these sources passes through the outer membrane and is bound by periplasmic protein FbpA, prior to entry into the cytoplasm. Outer membrane TF and LF receptors are unusual in that they appear to be bipartite; one protein has the characteristics of a typical TonB-dependent receptor, while the second is a surface-exposed lipoprotein. These receptors in Neisseria and Haemophilus show high specificity for glycoproteins of their normal hosts, such as humans. 

E. Low affinity iron transport The high affinity iron transport systems described immediately above are generally not expressed in environments containing more than 5–10 _M iron. Remarkably, the means by which bacteria assimilate iron in such iron-replete environments, oxic or anoxic, is not known. This so-called low affinity iron uptake may first require that iron be in the FeII form. Ascorbic acid, which reduces FeIII, stimulates low affinity iron uptake. FeII could be assimilated either by (i) the repressed levels of the feo system proteins or (ii) a cytoplasmic membrane transporter with broad specificity for divalent cations, such as the CorA protein of E. coli and Salmonella typhimurium. Also, several types of small molecules (monocatecholates, _-keto acids, and _-hydroxy acids) can, in certain genera, function as siderophores; they could mobilize extracellular ferric iron and, using a variety of transporters, provide iron. 

II. IRON-DEPENDENT REGULATION Genes encoding proteins necessary for high affinity iron uptake are regulated by iron availability. The key regulatory protein in most bacteria, gram-positive and gram-negative, is Fur. Fur, a small, histidine-rich polypeptide, is an aporepressor; in the presence of its corepressor FeII, it binds DNA. The holorepressor binds operators termed iron boxes, which are AT rich 19 bp sequences with dyad symmetry (Table 55.1). Iron boxes are located in promoters of iron-regulated genes, such that steric hindrance prevents the repressor and RNA polymerase from binding simultaneously. Adequate intracellular iron supplies thus prevent transcription of genes for transport proteins and siderophore biosynthetic enzymes. In E. coli, there are approximately 50 such Fur-regulated genes. Negative control of genes by Fur does not completely explain the regulatory effects of iron. Although most iron-regulated genes are repressed by Fur in iron-replete conditions, some are positively controlled by Fur and some are induced by iron only in the absence of Fur. A subset of the E. coli proteins positively regulated by Fur was recently demonstrated to be controlled indirectly by a small RNA (RyhB) that could function as an antisense RNA. RyhB synthesis is negatively regulated by Fur; in the absence of holorepressor, RyhB is made and blocks synthesis of relevant proteins at a post-transcriptional initiation step. This complexity also arises because Fur can influence the expression of other regulatory systems and because some iron-regulated genes are also subject to other global control molecules. This latter observation becomes understandable when the key role of iron in metabolism is considered. For instance, iron requirements should depend in part on whether the fueling reactions being utilized require an electron transport system (respiration), with its requisite Hm and iron–sulfur proteins and propensity to generate toxic oxygen species, or not (fermentation). Gram-positive organisms whose DNA has a high GC content, like Corynebacterium and Streptomyces, have no Fur but, instead, accomplish iron-regulated transcriptional control with DtxR proteins. 

The protein sequence of DtxR is unlike that of Fur but DtxR, nonetheless, functions in a Fur-like manner. FeII-DtxR binding sites are 19 bp palindromes (Table 55.1) and are positioned in the promoters of genes they regulate. Some iron transport systems are produced only in iron-deficient environments that contain their cognate ferrisiderophores. That is, the systems are synthesized only if the intracellular iron concentration, as sensed by Fur, is low and if their specific ferrisiderophore is bound to the cell. As established in the ferridicitrate system of E. coli, the specific outer membrane receptor must be present, as must a functional TonB system. Binding of ferrisiderophore to the receptor transmits a signal to a cytoplasmic membrane protein that has both periplasmic and cytoplasmic domains. The signal is then passed to a cytoplasmic sigma-factorlike protein that, upon activation, associates with RNA polymerase and stimulates transcription of the necessary transport genes. Unique regulatory elements for these systems include an outer membrane receptor capable of interacting with the cytoplasmic membrane protein, the cytoplasmic membrane protein, and its partner sigmalike factor. Among Pseudomonads, this dual regulation of specific iron transport systems by Fur and surface binding of the relevant ferrisiderophore is common. Regulation of iron assimilation not only conserves bacterial carbon and energy resources but also prevents overaccumulation of iron. Iron overload is a legitimate concern in bacteria as excess free iron can promote the formation of the toxic superoxide (O_ 2) and hydroxyl (OH ) radicals. The significance of DNA damage brought about by iron-generated hydroxyl radicals is demonstrated by the finding that E. coli Fur_ mutants, which assimilate too much iron, that are also defective in DNA repair, cannot survive in oxic environments. The intracellular oxidation state of iron is also used for regulation. Several proteins that control genes whose functions are related to oxygen levels contain iron–sulfur clusters; the iron present in these clusters is used as a sensing device. The E. coli Fnr protein, a transcriptional activator needed for anaerobic growth, is active only when its Fe–S center is reduced. In contrast, the sensor protein for superoxide stress, SoxR, is a positive regulator that is active with an oxidized Fe–S cluster. 

III. INTRACELLULAR “FREE” AND STORED IRON Two types of iron storage proteins, ferritin (FTN) and bacterioferritin (BFR), are found in bacteria. BFRs are a subfamily of FTNs, distinguished by the fact that they contain heme units. Like eukaryotic ferritins, the bacterial proteins are large (ca. 500,000 Da) and composed of 24 subunits. The actual role of these proteins is uncertain. Unlike eukaryotic FTNs, which can accommodate over 4000 iron atoms, maximum iron contents for FTN and BFR are much less. BFR is synthesized primarily during periods of slow or no growth and may serve as an iron donor upon resumption of growth. FTN, on the other hand, is present at a constant low level throughout the growth cycle and, for at least Escherichia coli and Campylobacter jejuni, has been shown to protect against iron-catalyzed oxidative damage. Amajor uncertainty regarding bacterial iron metabolism is the distribution and form of much of the intracellular iron. Even the quantity of iron in a bacterium is unclear; for the well-studied organism E. coli, estimates of the number of iron atoms/cell range from 100,000 to 750,000. In a variety of bacteria, Hm-containing proteins account for less than 10% of the total iron, in contrast to the situation in animals. Iron-sulfur proteins also contain 10% or less of the bacterial iron, as do FTN and BFR when cells grown under iron-deficient conditions are examined. The remainder of the bacterial iron, the majority, exists in a mobile pool about which little is known. It is this mobile iron that presumably is responsible for iron regulation and for the toxic effects of iron overload. Whether or not iron passes through this pool before being placed by ferrochelatase into protoporphyrin IX to form Hm, before incorporation into Fe–S clusters of proteins by unknown means, or before being converted to FeIII by the ferroxidase activity of BFR and FTN and deposited in the core of these molecules is unknown. IV. IRON IN PRIMARY FUELING REACTIONS Some bacteria can use the oxidation of iron compounds as their primary energy source. Bacteria capable of using inorganic, rather than organic, molecules for their fueling reactions are termed chemolithotrophs and iron-oxidizing bacteria are a major group in this nutritional category. Iron-oxidizing bacteria typically live in acidic, aerobic environments rich in both reduced iron and sulfur compounds; they grow poorly at pH values greater than 4. Among other things, low pH is critical in keeping FeII from being spontaneously oxidized to FeIII. 

Thiobacillus ferrooxidans, the best studied microbe in this group, oxidizes iron using proteins in the cell envelope. An outer membrane complex oxidizes iron to the ferric form, and a periplasmic protein transfers the electrons to cytochromes in the cytoplasmic membrane; these, in turn, pass the electrons to oxygen, the ultimate electron acceptor, in the cytoplasm. T. ferrooxidans is a major cause of water pollution, specifically, acid mine drainage. Ferrous sulfide is common in many coal and ore sites and its chemical and bacterial oxidation leads to both acidification and addition of dissolved metals to the water. Downstream, as the pH of the mine water becomes less acidic, insoluble ferric precipitates form. In contrast to the relatively few bacteria which can oxidize iron as the initial step in generating energy, in the absence of oxygen, many bacteria, notably Shewanella species and members of the Geobacteriaceae family, can use FeIII as the terminal electron acceptor in fueling reactions. This dissimilatory reduction of iron is a form of anaerobic respiration and is of interest for several reasons. It has the geochemical significance of solubilizing iron and it is likely to represent a very early form of respiration, as all of the last common ancestors of modern organisms can reduce FeIII (using dihydrogen as the electron donor). 

V. IRON AND PATHOGENICITY A major stimulus to current studies on microbial iron metabolism is their medical significance. A brief review of the genera studied in Section I will emphasize this fact. Bacterial pathogens find themselves in an irondeficient environment when they invade a eukaryotic host; conditions facilitating their acquisition of iron would be expected to increase virulence. Several general observations support this idea. Humans with higher than normal iron levels show enhanced susceptibility to bacterial infections. Similarly, at least 20 bacterial pathogens are more virulent when their animal host is injected with iron compounds prior to infection. Because excess iron can have a number of effects, including some that are detrimental to host immune defenses, these data are somewhat ambiguous. They are bolstered, however, by in vitro experiments showing that the antibacterial effects of body fluids can be uniquely reversed by iron supplementation. Bacteria must synthesize an appropriate iron uptake system to overcome the bacteriostatic conditions in their hosts. Pathogens which multiply extracellularly, such as in blood or on mucosal surfaces, and which do not lyse host cells must be able to obtain iron from TF or LF. This can be accomplished by siderophores or by TF- or LF-specific receptor proteins. Septicemic bacteria with defective siderophore systems are, in fact, less virulent. On the other hand, the major source of iron for intracellular pathogens is Hm and siderophore-deficient mutants of these organisms are still virulent. The required iron is taken from Hm-compounds, in this case. For many pathogens, it has been difficult to identify specific iron uptake systems that are essential for their virulence. The multiplicity of means for assimilating iron in every species studied complicates such attempts. However, when all high affinity systems are inactivated by use of mutations in tonB, Salmonella typhimurium, Vibrio cholerae, and Haemophilus influenzae are avirulent. That a specific iron assimilation pathway can be a virulence factor now has been demonstrated for a number of pathogens. The type of transport system critical for virulence varies with the pathogen and can be species specific. Thus, Neisseria meningitidis lacking its outer membrane hemoglobin receptor (HmbR) is attenuated for meningococcal infection in infant rats. Neisseria gonorrhoeae with no functional transferrin receptor is unable to initiate urethritis in humans. High affinity ferrous iron transport (FeoB mediated) is important for Helicobacter pylori colonization of the gastric mucosa of mice and a functional siderophore system is required for Yersinia pestis virulence in mice. Y. pestis strains unable to either synthesize or transport the siderophore yersiniabactin are avirulent. A siderophore system is also a virulence factor in the fish pathogen Vibrio anguillarum. This marine microbe infects salmonid fish; approximately 10 bacteria per fish are sufficient to cause vibriosis, a fatal disease characterized by hemorrhagic septicemia. A plasmid-encoded iron-uptake system that uses the siderophore anguibactin is necessary for virulence. Last, many pathogens utilize the low iron concentrations in the host as an environmental signal to synthesize virulence factors. Fur and DtxR-like proteins control not only iron acquisition systems but synthesis of toxins and hemolysins in these organisms.