Ins and outs: recent advancements in membrane protein-mediated prokaryotic ferrous iron transport

Abstract

Iron is an essential nutrient for virtually every living organism, especially pathogenic prokaryotes. Despite its importance, however, both the acquisition and the export of this element require dedicated pathways that are oxidation state dependent. Due to its solubility and kinetic lability, reduced ferrous iron (Fe²⁺) is useful to bacteria for import, chaperoning, and efflux. Once imported, ferrous iron may be loaded into apo and nascent enzymes and even sequestered into storage proteins under certain conditions. However, excess labile ferrous iron can impart toxicity as it may spuriously catalyze Fenton chemistry, thereby generating reactive oxygen species and leading to cellular damage. In response, it is becoming increasingly evident that bacteria have evolved Fe²⁺ efflux pumps to deal with conditions of ferrous iron excess and to prevent intracellular oxidative stress. In this work, we highlight recent structural and mechanistic advancements in our understanding of prokaryotic ferrous iron import and export systems, with a focus on the connection of these essential transport systems to pathogenesis. Given the connection of these pathways to the virulence of many increasingly antibiotic-resistant bacterial strains, a greater understanding of the mechanistic details of ferrous iron cycling in pathogens could illuminate new pathways for future therapeutic developments.

Graphical Abstract

Introduction

Iron is a versatile and necessary micronutrient for nearly all organisms, as it is essential for biological processes such as nitrogen fixation, respiration, oxygen transport, and even DNA biosynthesis.¹¹ Notwithstanding its importance, this essential element is difficult to obtain from today’s oxygen-rich environment, and this conundrum is especially tricky for bacteria that are unable to control their surroundings or their elemental feedstocks. For pathogenic bacteria within a host, populations of both ferric (Fe³⁺) and ferrous (Fe²⁺) iron may exist under different physiological conditions and in different forms (labile, inert, chelated, loosely-bound, etc.). In oxygen-replete (oxic) environments, the formation of iron hydroxides/oxides that are notoriously insoluble occurs spontaneously and represents a thermodynamic sink. To access soluble Fe³⁺, bacteria have adapted in order to scavenge iron from their surroundings using low molecular weight siderophore-based approaches for downstream biological processes.^{1, 12} Infectious bacteria may also release proteins that will bind to either extracellular free ferric iron ions/molecules, or will utilize transferrin/lactoferrin-binding proteins that will sequester ferric iron from their respective proteins.^{13, 14} Additionally, pathogenic bacteria may scavenge iron protoporphyrin IX (heme b) from erythrocytes and muscle tissue, if available.^{11, 15, 16} Both siderophore- and heme-based iron acquisition strategies are actively studied by multiple research groups, have been nicely reviewed (see e.g., Miethke and Marahiel; Contreras et al.),^17–21 and have been the target of recent exploits to attenuate bacterial virulence.^{22, 23} However, another form of iron cycling exists that is less well studied but equally as important.

Bacteria may also experience oxygen-depleted (anoxic) and/or acidic conditions that support reducing environments within the host in which Fe²⁺ is the dominant iron oxidation state.^24–26 Examples of these locations include acidic portions of the gut,²⁷ such as the stomach, reducing environments such as the mammalian duodenum,²⁸ and anoxic locations such as the biofilms of the sub-gingival tissue or the lungs of late-stage cystic fibrosis (CF) patients.^{29, 30} In these locations, ferrous iron is often prevalent, and the utilization of ferrous iron acquisition strategies for virulence becomes paramount.^{14, 24, 25} Even though ferrous iron transport is critical under these conditions, the mechanisms by which bacteria obtain ferrous iron are not nearly as well elucidated, necessitating a closing of this knowledge gap.

Regardless of the precise acquisition pathway, bacteria use several common approaches to control iron homeostasis: (1) iron import systems that are tuned to acquire iron from the environment; (2) iron-responsive gene regulation; (3) internal storage systems for use when iron is scarce; (4) redox stress resistance systems; and (5) iron efflux systems (Fig. 1).¹ This tight level of control is required, as excess and misregulated levels of iron can result in Fenton-like chemistry.^{31, 32} Specifically, cytoplasmic, labile iron may serve as a catalyst to reduce H₂O₂, producing reactive oxygen species (ROS) such as hydroxyl radicals that can damage proteins, membranes, and DNA.^{31, 33} As homeostasis wanes, further destruction of [Fe-S] clusters and/or their release into the cytosol can upend electron transport and even trigger bacterial ferroptosis.^33–37 These results highlight the importance of iron homeostasis for the survival of unicellular pathogens.

Figure 1.

Cartoon schematic of strategies employed by bacteria to maintain iron homeostasis.¹ (1) Import of Fe²⁺ increases labile intracellular Fe²⁺ pools. Intracellular labile Fe²⁺ can be (2) sensed by Fe-responsive gene regulatory systems including the ferric uptake regulator (FUR; PDB ID 4RB2; purple spheres represent bound Mn²⁺),⁶ the fumarate and nitrate reductase regulator (FNR), the peroxide operon regulator (PerR), and the iron response regulator (Irr) for repression or derepression of import and export machinery; (3) stored in ferritins (PDB ID 1EUM; gray spheres represent Fe-binding sites),¹⁰ bacterioferritins, and DNA-binding proteins from starved cells (Dps) for use when the cells face iron replete conditions; and/or (4) used in conjunction with flavins to resist redox stress introduced from reaction with ROS. Finally, (5) iron intoxication or iron excess may be relieved by efflux via Fe²⁺ export systems.

If famine becomes feast, bacteria may also use several approaches to compensate for high intracellular iron levels including the attenuation of iron import, the storage of excess iron, the alteration of global metal transport, and even the efflux of iron.^{32, 38} Many of these strategies have been extensively reviewed (see e.g., Andrews et al. and Cornelis et al.),^{1, 12} with the exception of ferrous iron exporters. The concept of ferrous iron “intoxication” is beginning to gain traction in the field, and ferrous iron efflux pumps are generally accepted to be active transporters that are utilized by bacteria to export Fe²⁺ in order to alleviate excess labile intracellular iron and to lower oxidative stress.³² Although ferrous iron exporters are thought to be found across all kingdoms of life, including most pathogenic bacteria,⁹ many ferrous iron efflux systems are only beginning to be understood, and their contribution to virulence remains unclear.

This perspective highlights recent advancements in our biochemical, mechanistic, and structural characterizations of ferrous iron import and efflux. Much of our understanding derives from in vitro recombinant work complemented by studies at the organismal level. Unfortunately, as the majority of these transport systems are membrane proteins and/or membrane protein complexes, structural advancements in this field have been slow but significant (Table 1).

Table 1.

The structures of known and proposed prokaryotic Fe²⁺ importers, exporters, and their associated fragmentary domains.

Protein	Organism	Function	PDB ID
MntH	Deinococcus radiodurans	Import	5KTEa, 6D9Wa, 6C3Ib, 6BU5c, 6D91d
EfeO	Yersinia pestis	Import	5TW9
EfeB	Escherichia coli	Import	2Y4Df, 3O72g
NFeoBe	Escherichia coli	Import	3HYRf, 5FH9f, 3I8Sf, 3I92h
NFeoB	Gallionella capsiferriformans	Import	3W5If, 3W5Ji
NFeoB	Klebsiella pneumoniae	Import	2WIAf, 2WIBi, 2WICj
NFeoB	Legionella pneumophila	Import	3IBY
NFeoB	Methanocaldococcus jannaschii	Import	2WJHf, 2WJGi, 2WJIj,2WJJk
NFeoB	Pyrococcus furiosus	Import	3K53
NFeoB	Streptococcus thermophilus	Import	3B1Zf, 3B1Yi, 3B1Wi, 3LX8i, 3B1Xj, 3SS8l, 7BWVl, 7BVUl, 3LX5m, 3B1Vm, 3TAHn
NFeoB	Thermatoga maritima	Import	3A1Vf, 3A1Si, 3A1Ti, 3A1Uj, 3A1Wo
FieF	Escherichia coli	Export	3H90p, 2QFIp
FieF	Shewanella oneidensis	Export	3J1Za, 5VRFp, 7KZXq, 7KZZr
MamBCTDs	Desulfamplus magnetovallimortis	Export	6QFJ
MamMCTD	Desulfamplus magnetovallimortis	Export	6QEK
MamMCTD	Magnetospirillum gryphiswaldense	Export	3W5Xf, 3W5Yf, 6GP6t, 6GMTu, 6GMVv

^aInward-facing, open, apo;

^bInward-facing, occluded, apo;

^cOutward-facing, open, Mn²⁺-bound;

^dOutward-facing, open, apo;

^eNFeoB is the soluble G-protein-like domain at the N-terminus of FeoB;

^fApo form;

^gHeme-bound form;

^hGCP-bound form;

ⁱGDP-bound form;

^jGMP-PNP-bound form;

^kGDP-bound selenomethionine form;

^lGDP:AlF₄⁻- and K⁺-bound form;

^mmantGMP-PNP-bound form;

ⁿmantGDP-bound form;

^oΔGDI form;

^pZn²⁺-bound form;

^qApo form complexed with fragment antigen-binding (Fab) protein;

^rZn²⁺-bound form complexed with Fab;

^sCTD represents the C-terminal, cytosolic domain of Mam proteins;

^tCu²⁺-bound form;

^uCd²⁺-bound form;

^vNi²⁺-bound form

However, with the major leap forward in the accessibility of alternative strategies for, in particular, the structural determination of particulate proteins (e.g., cryo-EM), we anticipate advancements in this field to accelerate at a rapid pace to complement what is known at the in vivo and in vitro levels. We believe a greater understanding at the atomic level of the underpinnings that control the cycling of ferrous iron within pathogenic bacteria could afford the targeting of these systems for the attenuation of virulence, similar to strategies that have been developed for ferric siderophore-based and heme-based iron acquisition systems.

Ferrous Iron Importers

The first step in the bacterial ferrous iron cycle is import across a lipid bilayer. The relevant reported ferrous iron transporters that have been described include the MntH, ZupT, EfeUOB, IroT, YfeABCDE, FutABC, and Feo systems (Fig. 2). Despite the ostensibly large number of ferrous iron importers that have been characterized, few appear to be dedicated to the uptake of only ferrous iron, and even fewer (chiefly Feo) are widespread across the prokaryotic domain. In this section, we summarize and synopsize what is currently known about bacterial ferrous iron import.

Figure 2.

Cartoon overview of the major bacterial ferrous iron importers. From left to right are the YfeABCDE (orange), EfeUOB (brown), MntH (blue), FeoABC (purple), ZupT (green), IroT (red), and FutABC (yellow) systems. Ferrous iron ions are represented by gray spheres, manganese is represented by a purple sphere, zinc is represented by a green sphere, and a proton is represented by a blue sphere. In several cases (e.g. MntH, ZupT), Fe²⁺ is not the primary metal substrate but rather an auxiliary substrate.

Promiscuous iron importers

The H⁺-dependent manganese transport system (MntH) membrane protein is encoded by the gene mntH and may transport Fe²⁺ in addition to its chief function as a Mn²⁺ importer. Despite its name, MntH has ≈ 36% sequence identity to the eukaryotic natural resistance-associated macrophage protein 1 (NRAMP1),³⁹ a transmembrane divalent metal, proton-dependent co-transporter with broad metal specificity, suggesting that other divalent metals such as Fe²⁺ may be MntH substrates. The Deinococcus radiodurans R1 MntH structure has been determined (PDB ID 6C3I) and the overall architecture is composed of 10 α-helices that resemble a pore (Fig. 3).⁴⁰ The Mn²⁺-bound MntH structure reveals that Mn²⁺ is coordinated by the backbone carbonyl of Ala53 and the side chains of Asp56, Asn59, and Met230 as well as two water molecules.⁴⁰ In addition to its canonical substrate (Mn²⁺), studies utilizing inducible Escherichia coli mntH have observed the intracellular uptake of radioactive Fe²⁺ in a proton-dependent manner, establishing that the MntH transmembrane protein is also able to transport Fe²⁺.³⁹ MntH-mediated metal transport activity is higher in acidic conditions, and protons accumulate intracellularly, indicating that dedicated symport is operative.⁴¹ However, despite the ability of MntH to transport Fe²⁺, experiments done with radioactive Mn²⁺ demonstrated a preference for Mn²⁺ as a substrate over Fe²⁺. In fact, MntH is capable of binding and transporting a wide swath of divalent metals in a proton-dependent manner with the following affinities: Mn²⁺ > Cd²⁺ > Co²⁺ > Fe²⁺ > Zn²⁺ > Ni²⁺ > Cu²⁺. Such a broad specificity demonstrates the promiscuous nature of MntH. In fact, given MntH’s sequence identity and similar broad specificity to that of the eukaryotic NRAMP family of proteins, it has been suggested that MntH may represent a NRAMP bacterial ancestor, leading many to refer to MntH as an NRAMP-like transporter.^{39, 42}

Figure 3.

Structure of Deinococcus radiodurans MntH bound to Mn²⁺ (PDB ID 6BU5). The MntH structure is composed of 10 α-helices that adopt a pore-like structure. The metal binding site (right) reveals that MntH coordinates Mn²⁺ via Asp56, Asp59, Met230, and the backbone carbonyl of Ala53. Two additional molecules of water coordinate Mn²⁺ but are not displayed for clarity. ‘N’ and ‘C’ represent the amino and carboxy termini, respectively.

MntH has been shown to be important for some pathogenic strains, although a consensus remains opaque. For example, Makui et al. found that a frameshift deletion of the mntH gene in E. coli did not affect cell growth in either minimal or nutrient-rich media.³⁹ This finding led to the conclusion that E. coli MntH is not required for cell growth, or virulence, meaning that E. coli relies heavily on metal transporters other than MntH in order to meet its needs for certain divalent metals.³⁹ Similarly, in vivo studies of Yersinia pestis-infected mice found that a ΔmntH strain had no significant loss of virulence in a bubonic mouse model, and only a double knock out of yfe and mntH led to a 133-fold loss in virulence; however, no virulence attention was not observed in the pneumonic plague model in mice.⁴³ Virulence testing of Mycobacterium tuberculosis MntH using bone marrow derived macrophages demonstrated no significant loss of virulence.^{43, 44} Another study done on Brucella abortus, the leading cause of undulant fever in humans, showed that a deletion of the mntH gene led to attenuated growth in both rich and minimal media but was later rescued by introducing high concentrations of Mn²⁺.⁴³ Given the established role of Mn²⁺ in protecting bacteria from oxidative stress,³⁸ these findings may link MntH to a role in oxidative stress resistance. Consistent with this notion, deletion of the mntH gene within Yersinia pseudotuberculosis, and Salmonella typhimurium led to higher sensitivity to H₂O₂ under manganese-limited conditions, signifying a MntH dependence of oxidative stress resistance in bacteria.^{24, 41, 45} While MntH is the main Mn²⁺ transporter in B. abortus and potentially other species of bacteria,⁴⁶ it is unclear to what extent auxiliary Mn²⁺ transporters, and even a secondary function in Fe²⁺ transport, contribute to virulence, thus warranting further characterization.

The zinc uptake transporter ZupT is a part of the ZRT-, IRT-like protein (ZIP) family of proteins involved chiefly in Zn²⁺ transport, but these proteins may also transport Fe²⁺ in addition to other divalent metal ions.^{47, 48} The crystal structure of Bordetella bronchiseptica ZIP with bound Zn²⁺ has been determined and reveals a membrane-imbedded 8 α-helical pore with two Zn²⁺ ions forming an intriguing binuclear metal center (PDB ID 5TSA).⁴⁹ While there is currently no structure of ZupT, it is speculated based on sequence conservation to that of B. bronchiseptica that ZupT may be composed of a similar architecture. In vivo radioactive Zn²⁺ transport studies have demonstrated definitively that ZupT mediates Zn²⁺ uptake in E. coli;⁴⁸ however, whether or not ZupT has a binuclear metal center like other members of the ZIP family has yet to be determined. Like MntH, ZupT also exhibits substrate promiscuity, having been shown to transport Fe²⁺, Co²⁺, and Mn²⁺, but ZupT appears to retain its highest affinity for Zn²⁺.^{47, 48} While the mechanism of how ZupT transports divalent metals is still not clearly understood, it is suggested to function as a permease, similar to other ZIP family proteins. In addition, ZupT maintains a constitutive and low expression level even in the presence of the aforementioned divalent metals.⁴⁷ In terms of its connection to virulence, deletion of the zupT gene alone does not disable infection in pathogens.^{50, 51} For example, in vivo studies of Salmonella enterica demonstrate that virulence is decreased only when both zupT and an additional zinc ABC transporter gene, znuABC (zinc uptake ABC) are deleted, and not zupT alone. These observations suggest that ZupT is not the primary Zn²⁺ transporter in bacteria, and thus is not strictly required for colonization.⁵⁰ Additionally, phenotypic changes do not appear to be related to Fe²⁺ homeostasis, suggesting that ZupT only plays a minimal role at best in maintaining intracellular stores of ferrous iron.⁴⁷

The Yersinia ferrous iron transport (Yfe) ABCDE and the Salmonella iron transport (Sit) ABCD systems have homology and are involved in the import of Mn²⁺ and Fe²⁺.^{52, 53} The Yfe system is an inner membrane ATP-binding cassette (ABC) transporter that utilizes ATP for the active transport of both Mn²⁺ and Fe²⁺. Studies have shown that Yfe can transport both metals but has a higher affinity for Mn²⁺, suggesting that Fe²⁺ transport may be an auxiliary function in S. enterica.⁵⁴ YfeA is a periplasmic protein predicted to bind both Mn²⁺ and Fe²⁺, while YfeC and -D are inner membrane permeases and YfeB is an ATPase.⁵² In addition to its typical ABC transporter architecture that consists of a four protein complex including two cytosolic nucleotide-binding domains and two transmembrane α-helical domains,⁵⁵ the Yfe system bears an additional gene, yfeE, which is not part of the same operon as yfeABCD but encodes for a separate inner membrane protein YfeE that has an unknown function.^{24, 52, 56} Knockout studies report that a yfeE deletion in Yersinia pestis did not result in decreased cell growth, suggesting that YfeE may only be accessory to the Yfe system and not essential for homeostasis.⁵² Interestingly, in vivo knockout studies reveal that a deletion of the yfeABCD operon resulted in only a modest decrease of Y. pestis cell growth in iron-deficient conditions, indicating that Yfe does not act as a primary Fe²⁺ transporter.⁵² In contrast, metal transport studies conducted of the pathogenic E. coli SitABCD system demonstrated a higher preference for Fe²⁺ than Mn²⁺, and Fe²⁺ was transported with higher affinity than Fe³⁺.^{54, 57} Taken together, these findings highlight the need for additional studies into the specific determinants of substrate recognition within these transporters.

Dedicated iron importers

The ferric uptake (Fut) ABC system is primarily found in cyanobacteria and is an inner membrane ABC transporter system that appears to be a dedicated iron transporter;^{56, 58, 59} however, there is no consensus as to iron oxidation state specificity. The Fut system contains two separate periplasmic proteins named FutA1 and FutA2, with FutB being an inner membrane pore and FutC an ATPase.⁵⁹ Evidence suggests each part of the ABC transporter may have differing iron specificities, with FutB preferentially binding Fe³⁺, while the periplasmic proteins FutA1 and FutA2 preferentially bind Fe²⁺.^58–60 However, the crystal structure of a Trichodesmium erythraeum FutA1 homolog reveals that FutA1 binds Fe³⁺ coordinated by a strongly Lewis acidic, octahedral ligand sphere (3 Tyr hydroxyls/hydroxides and 3 oxygens from water molecules), which was confirmed by electron paramagnetic resonance spectroscopy.⁶¹ It is important to note that these experiments were performed in oxic conditions; thus, it is possible that FutA1 is capable of binding Fe²⁺ via a different ligand set under anoxic conditions, but these experiments remain unrealized.⁶¹

While there is still no consensus on the Fut mechanism, Kranzler et al. have suggested that the Fut system works in conjunction with the alternate respiratory terminal oxidase (ARTO) membrane protein in a reductive iron uptake pathway, similar to a ferric reductase approach (vide infra).⁵⁸ It is hypothesized that FutA1 and/or FutA2 bind Fe³⁺ and either deliver ferric iron to FutB for transport, or the Fut system interacts with ARTO and reduces Fe³⁺ to Fe²⁺, allowing the ferrous iron to be transported in another manner. While a consensus on the substrate oxidation state remains unresolved, knockout of any components of the Fut system led to modest decreases of iron accumulation within Synechocystis 6803.⁵⁸ While cyanobacteria are generally not regarded as pathogenic, this information could be useful and extrapolated to infectious homologs.

A more common pathogenic iron transport system is the elemental ferrous iron uptake (EfeUOB) system. This system resembles the Fe²⁺ transporter Ftr1P found in yeast, with EfeU being an inner transmembrane pore, and EfeO and -B both being periplasmic. EfeO has a C-terminal peptidase domain (PDB ID 5TW9) and a predicted N-terminal cupredoxin-like domain, while EfeB is a homodimer in which each protomer is composed of twelve α-helices and eight β-strands (PDB ID 2Y4D). Crystal structures of E. coli EfeB reveal the protein to be complexed with a b-type heme, showing that it is part of the heme peroxidase superfamily (PDB ID 3O72).⁶² Because of the high sequence homology to Ftr1P, it has been suggested that EfeU acts as a permease to transport both Fe²⁺ and Fe³⁺, with a higher affinity for Fe²⁺.^{63, 64} Interestingly, the efe operon is induced at low pH conditions, possibly due to the fact that these acidic environments have higher concentrations of Fe²⁺ compared to Fe³⁺. Despite its connection to Fe²⁺ import, deletions to the efe operon do not seem to restrict cell growth (at least in E. coli), indicating that other systems may be the primary drivers of Fe²⁺ uptake, or that these systems are capable of compensating in lieu of the Efe system.⁶²

A more specialized set of dedicated Fe²⁺ transporters are the iron transporter (IroT) membrane proteins, which are homologs of the previously defined “more regions allowing vacuolar colocalization N” (MavN) proteins. IroT is found chiefly in the Legionella genus of bacteria, including L. pneumophila, the human pathogen that causes Legionnaires’ disease.⁶⁵ This bacterium infects macrophage cells and relies on the formation of an intracellular compartment named the Legionella containing vacuole (LCV). In order to meet the iron requirements of the L. pneumophila, IroT is recruited to the LCV membrane.^{66, 67} IroT is predicted to contain eight transmembrane α-helices and has also been shown to transport a wide array of divalent metals in addition to Fe²⁺, including Mn²⁺, Co²⁺, and Zn²⁺, with equivalent proficiency at least in vitro; however, IroT is suggested to function only for Fe²⁺ transport in vivo.⁶⁸ While the exact mechanism of IroT-mediated metal transport has not been elucidated, several mutagenesis experiments display the importance of key amino acids for metal translocation. For example, an ExxE motif that is conserved amongst IroT proteins has been found to affect iron transport using an E439A variant.⁶⁸ However, this motif is notably found within the lumen side of the LCV, indicating that it may play a role in either initial iron binding or mediating iron translocation through the pore.⁶⁷ An additional variant of the predicted transmembrane His (H412A) showed a considerable decrease in growth within the LCV, implying that divalent metals may be coordinated by His412 during metal transport.⁶⁸ Interestingly, in vivo knockout studies of LCV-independent L. pneumophila cells show impaired growth, suggesting that IroT may also be incorporated to the bacterial membrane for iron uptake.⁶⁵ However, knockout of the LCV IroT affected growth more drastically, indicating a heavier reliance on this strategy in the vacuole.⁶⁸ Regardless, these results strongly suggest that targeting this Fe²⁺ transporter could be a viable strategy in the fight to curb Legionnaires’ disease.

The ferrous iron transport (Feo) system is the most widely distributed and dedicated Fe²⁺ uptake system across the prokaryotic domain. Unlike many of the aforementioned iron transporters, the Feo system has been shown to have the highest affinity for Fe²⁺,^{25, 56, 69–71} and Feo is connected most widely to the virulence of pathogenic prokaryotes.^{56, 65, 71, 72} Interestingly, studies of FeoB2, a second protein annotated as FeoB that is found in mostly the Porphyromonas class of bacteria, indicate that this protein is able to transport Mn²⁺ and not Fe²⁺, whereas FeoB1 in P. gingivalis is strictly an Fe²⁺ transporter.⁷³ These results suggest FeoB2 may behave similar to MntH; however, structural and metal specificity studies of FeoB2 have yet to be carried out in vitro in great detail. The Feo system, initially identified in E. coli, is made up of three separate proteins: FeoA, FeoB, and FeoC. FeoA and FeoC are both small (ca. 8 kDa each), cytosolic proteins, while FeoB is a large (ca. 80 kDa) transmembrane, nucleotide-hydrolyzing protein (most commonly GTP) with eukaryotic G-protein orthologs.²⁵

The functions of both FeoA and FeoC remain enigmatic. FeoA has been structurally characterized and bears a conserved SH3-like domain, which is characterized by a small β-barrel.⁷⁴ The conservation of this protein-protein interaction fold has led to the hypothesis that FeoA acts as either a GTPase activating protein (GAP; increases the rate of GTP hydrolysis), or a guanine exchange factor (GEF; helps to exchange hydrolyzed GDP for GTP). However, in order for this function to be operative, direct protein-protein interactions must be made, and no such interactions have been characterized at the atomic level, although they have been hypothesized and inferred.⁷⁴ FeoC has also been structurally characterized and contains a conserved helix-turn-helix (HTH) motif,⁷⁵ suggesting that FeoC may function as a transcriptional regulator. Adding to its complexity, FeoCs from both E. coli and K. pneumoniae have been shown to bind a redox-active [4Fe-4S] cluster under strict anoxic conditions.⁷⁶ This cluster is O₂ sensitive and may be used as an oxygen sensor to regulate ferrous iron transport, somewhat similarly to the fumarate and nitrate reduction regulatory protein (FNR).⁷⁶ However, there are FeoC proteins that do not have the ability to bind [Fe-S] clusters, and there are no reports indicating that FeoC binds to DNA; therefore, the function of FeoC remains to be elucidated.²⁵

FeoB is the most important component of the Feo system, and this large, polytopic transmembrane protein is hypothesized to bind and to transport Fe²⁺ in a GTP-dependent manner. The recently released AlphaFold database proposes that FeoB (Uniprot ID: P33650; AlphaFold ID: AF-P33650-F1) is composed of three domains: a soluble N-terminal G-protein-like domain (termed NFeoB) (6 β-strands that form a central β-sheet surrounded by 5–6 α-helices and a GDI domain featuring 5 α-helices), a transmembrane domain made up of 10–11 α-helices, and a carboxylate-rich periplasmic loop (Fig. 4).⁵ However, there is no full-length structure of FeoB; instead, most research has historically focused on soluble NFeoB, which has been structurally characterized in the apo and guanosine-bound forms.^{69, 77, 78} Further, GTPase activity analyses have been done on both NFeoB and full-length FeoB proteins, with the full-length protein displaying higher activity (in some cases), implicating the existence of important contacts between the soluble domain and the transmembrane region.^{70, 79, 80} Even so, full-length FeoB still exhibits slow GTP hydrolysis rates (maximal reported k_cat^GTP ca. 0.1 s⁻¹), calling into question whether Fe²⁺ is translocated via an active or passive process.^{70, 80, 81} While the precise mechanism of Fe²⁺ import remains to be elucidated, Seyedmohammad et al. proposed that FeoB features a Cys-lined, GTP-gated channel as opposed to a transporter;⁸¹ alternatively, our lab has postulated that a Met-rich pathway may be involved in metal translocation,²⁵ which is a common theme amongst many Fe²⁺ transporters. These conflicting proposals underscore the need for additional studies to define the mechanism of Fe²⁺ transport by FeoB.

Figure 4.

Predicted AlphaFold model of Escherichia coli FeoB. FeoB (Uniprot ID P33650; AlphaFold ID AF-P33650-F1) is predicted to contain three distinct regions: a soluble N-terminal G-protein-like domain (NFeoB; teal), a transmembrane region made up of 10–11 α-helices, and a soluble periplasmic loop (both in purple).⁵ ‘N’ and ‘C’ represent the amino and carboxy termini, respectively.

Adding to FeoB’s complexity, recent discoveries have indicated that despite structural similarity between FeoB and eukaryotic G-proteins, some FeoBs may actually be NTPases.^{82, 83} Recent results from Shin et al. have demonstrated that a subset of NFeoB proteins have the ability to hydrolyze ATP in addition to GTP.^{82, 83} It has been proposed that the adenosine/guanosine discriminatory factor is a single amino acid difference within the FeoB G5-motif of the G-domain.⁸² This result is both intriguing and surprising, as NFeoB does not have structural similarity to known ATPases. Whether or not this motif is a defining factor of nucleotide distinction remains to be seen, as it is not yet known how NFeoB binds adenosine nucleotides.

While a consensus on mechanism has yet to be reached, a number of in vivo studies have demonstrated the connection between the presence of Feo and pathogenesis. Particularly emphasizing its role as a major virulence factor, the infectivity of problematic pathogens such as E. coli O78:H10, Helicobacter pylori, Y. pestis, Campylobacter jejuni, and even Porphyromonas gingivalis (to name a few) all have shown significant decreases upon the deletion (either in part or in whole) of the Feo system. While outside of the scope of this perspective, a more comprehensive report on Feo virulence has been laid out previously by Lau et al.⁵⁶ Thus, as the most widely distributed and dedicated Fe²⁺ uptake system across the prokaryotic domain, a greater understanding of the function of Feo at the atomic level would make major impacts in targeting this essential system in an effort to fight bacterial pathogenesis.

Ferric reductases and ferrous iron transport

Intriguingly, the widely-distributed Feo system is expressed when bacteria live in both anoxic and oxygen-replete environments, despite the reactivity of Fe²⁺ in the presence of oxygen.⁸⁴ While living under oxic conditions, many bacteria notably employ Fe³⁺ transporters to uptake ferric siderophores complexes. While these chelators bind and solubilize Fe³⁺ effectively, ferric siderophore complexes are notoriously stable and recalcitrant to iron release. To circumvent this issue, many bacteria use soluble (flavin-based) and/or membrane-bound (heme-based) ferric iron reductases, which reduce Fe³⁺ to Fe²⁺, creating a more labile feedstock into ferrous iron uptake pathways. Given its presence in most bacteria, it has thus been hypothesized that ferric iron reductases and the Feo system may work in conjunction with one another to fulfill bacterial iron needs. Indeed, a study has implicated the involvement of membrane-bound ferric reductases in Feo-mediated uptake pathways.⁸⁵ Given that membrane-bound ferric reductases are typically localized to the cytoplasmic membrane, it is tempting to speculate that these small, α-helical hemoproteins could be co-localized with FeoB to orchestrate ferrous iron import under oxic conditions; however, more work is needed to probe this hypothesis.⁸⁶ While ferrous iron uptake is principal to the survival of bacteria, a careful control of the intracellular iron levels must be achieved. In order to prevent the risk of intracellular oxidative damage, ferrous iron efflux systems are expressed and utilized when intracellular labile iron levels exceed a certain threshold.

Ferrous Iron Efflux Systems

Although iron is a necessary nutrient for nearly all bacteria, excess iron can be both toxic and dangerous, especially O₂-reactive ferrous iron. In oxygen-rich environments, this toxicity is commonly attributed to oxidative damage resulting from Fenton-like chemistry that creates reactive radicals from Fe²⁺ redox chemistry;³¹ however, the precise cause of damage under anoxic conditions has yet to be determined. To prevent toxicity, iron is typically sequestered by proteins, including enzymes, and chaperones, and excess iron can trigger the expression of iron storage vessels such as heme-containing bacterioferritins, traditional ferritins, and even the DNA-binding proteins from starved cells (Dps)-family of mini-ferritins.⁹ However, should intracellular iron exceed this buffering capacity, its egress is warranted, and versatile ferrous iron efflux systems have been recently identified in bacteria and are used to attenuate deleterious iron concentrations. These exporters are divided into four main families: (1) cation diffusion facilitator (CDF) proteins, (2) major facilitator superfamily (MFS) proteins, (3) membrane-bound ferritin-like proteins, and (4) P-type ATPases. In this section, we summarize and synopsize what is currently known about bacterial ferrous iron export.

Cation diffusion facilitator (CDF) proteins

Cation diffusion facilitator (CDF) proteins export a range of divalent metal ions including Zn²⁺, Cd²⁺, Co²⁺, Ni²⁺, Fe²⁺, and Mn²⁺.^87–89 Bacterial CDF proteins have been structurally characterized and are generally homodimers that feature a transmembrane domain composed of six α-helices connected to a C-terminal cytoplasmic domain (Figs. 5 and 6).^{2, 87, 90} To date, only the structures of metal-bound (Zn²⁺), full-length CDF proteins from E. coli and S. oneidensis are known (Table 1).^{2, 91} Phylogenetic and functional studies have identified factors for substrate selectivity, leading to a division of CDF proteins into three groups: those that transport only Zn²⁺, those that transport Fe²⁺ and Zn²⁺, or those that transport only Mn²⁺.⁹² However, nearly fifteen years have elapsed since the most recent effort to characterize metal selectivity based on phylogenetics,⁹² and an updated analysis is warranted. In general, transport by CDF proteins is stimulated by pH changes and chemiosmosis,⁹³ and Fe²⁺ translocating CDF proteins use the proton motive force (PMF) for efflux.^{89, 94}

Figure 5.

Cartoon overview of the major ferrous iron efflux systems in bacteria. Cation diffusion facilitator proteins (CDF; green) generally form a homodimer featuring an N-terminal transmembrane domain (TM) followed by a soluble C-terminal domain (C) and use the proton motive force (PMF) to transport divalent metal ions. Major facilitator superfamily (MFS; red) proteins are implicated in the transport of ferric/ferrous iron citrate and citrate alone (teal).³ MFS proteins often have distinct N- and C-terminal domains. Iron-transporting MFS proteins are homologous to proton antiporters, and it is proposed that an exchange with the iron citrate and protons occur in this system.⁸ Membrane-bound ferritin-like proteins (yellow) feature a soluble, dimeric, N-terminal ferritin-like Er domain (Er) and a C-terminal vacuolar iron transporter domain (VIT1). P_1B-ATPase proteins (blue) are composed of a transmembrane domain (TM), an ATP-binding domain (ATP-BD), and an actuator domain (AD). Ferrous iron ions are represented by gray spheres, protons are represented by blue spheres, and phosphate is represented by a red sphere. Figure adapted from Pi and Helmann.⁹

Figure 6.

The X-ray crystal structure of the iron transporting E. coli FieF (YiiP) protein in the presence of Zn²⁺ (PDB ID 2QFI). FieF is often considered a model for other CDF proteins.² The homodimer adopts a “Y”-like shape, where each protomer (colored green and teal separately) includes a transmembrane (TM) domain and a soluble, C-terminal domain (C). The TM domain features a six-helical bundle, and the C-terminal domain features two α-helices and a three-stranded β-sheet. The structure of FieF in the presence of Zn²⁺ reveals multiple metal binding sites: one in the TM domain (coordinated by Asp45, Asp49, His153, and Asp157), one in the linker between the TM and C domains (coordinated by Asp68 and His75), and an interesting di-zinc site in the C-terminal domain (coordinated by His232, His261, His283, and Asp285). It is possible that Fe²⁺ may bind in similar locations. ‘N’ and ‘C’ represent the amino and carboxy termini, respectively.

The field is conflicted regarding metal binding and selectivity.⁹ This conflict has likely stemmed from and even influenced metal binding characterizations during binding studies. For example, E. coli ferrous iron efflux protein (FieF, previously reported as YiiP) was first identified as a Zn²⁺ transporter,^{89, 94} as FieF expression was induced by the presence of Zn²⁺ in a concentration-dependent manner. However, this behavior did not explain the inherent contradiction that ΔfieF strains did not exhibit a decrease in Zn²⁺ tolerance, suggesting that FieF may serve to transport a different cation.⁹⁵ Later, FieF was found to transport iron based on studies in which iron accumulation was decreased upon in trans expression of FieF in ΔfieF E. coli,⁹⁰ and biophysical studies demonstrated that E. coli FieF selectively bound both Zn²⁺ and Cd²⁺.⁹⁶ This discord in metal binding specificity of FieF highlights the possibility that FieF recognizes and/or transports a diverse set of metals. Despite the challenges in determining its function, E. coli FieF is still considered a model of both CDF protein structure and function, having been studied structurally by X-ray crystallography (Fig. 6) and even recently by cryo-EM to lend insight into the mechanism of metal efflux. Specifically, cryo-EM studies have shown the rotating and “rocking” motions of the transmembrane helices that are necessary for Zn²⁺ transport,^{91, 97, 98} which may be involved in Fe²⁺ efflux by extrapolation.

Other iron transporting CDF proteins have been identified in both pathogenic and non-pathogenic bacteria but remain poorly understood. For example, Shewanella oneidensis FeoE, a homolog of E. coli FieF, transports Fe²⁺ to maintain cell vitality during iron respiration,⁹⁹ but biochemical characterization is lacking. In Magnetospirillum gryphiswaldense, MamB and MamM comprise a heterodimeric CDF protein complex that is relevant for transporting Fe²⁺ into magnetosome membrane vesicles, which are thought to drive biomineralization of magnetite crystals.^100–103 The opportunistic pathogen P. aeruginosa encodes AitP, a CDF that directly participates in the homeostasis of both Fe²⁺ and Co²⁺,¹⁰⁴ but little else is known beyond this extent. In fact, despite the identification of CDF proteins across several pathogenic bacterial strains, and their likely contribution to virulence, a consensus model for metal specificity of CDF proteins has yet to be reached.⁹ Thus, further investigation is warranted into the CDF metal-transport mechanism.

Major facilitator superfamily (MFS)

The major facilitator superfamily (MFS) of proteins is the largest group of secondary active transport proteins that are linked to pathogenesis and involved in translocation (both import and efflux) of nutrients, small molecules, and metal ions (Fig. 5).^{3, 105} Subdivided into three categories based on the direction and type of energy sources used to support substrate transport,³ MFS-mediated transport may be carried out by a facilitator (uniporter) that supports diffusion across the membrane based on a concentration gradient, a symporter (cotransporter in the same direction), or an antiporter (exchanger in opposite directions).³ Generally, MFS proteins are composed of 12 transmembrane α-helices arranged in distinct N- and C-terminal domains with two-fold pseudosymmetry that are connected by an unstructured linker and cycle through distinct conformations to complete transport.³ Structural analysis of an iron transporting MFS protein remains elusive, but other MFS transporters have been well-studied by crystallography.^{3, 106–109} These studies have demonstrated that helices 1, 4, 7, and 10 are oriented toward the center of the two domains to form a path for transport, while the interface between the N- and C-terminal domains is formed by α-helices 2, 5, 8, and 11; α-helices 3, 6, 9, and 12 compose the outer surface of the protein to provide structural support.³ Substrates of MFS transporters are a diverse but essential set of molecular building blocks, and evidence has linked MFS transport to the uptake of amino acids, ions, lipids, nucleosides, peptides, and even iron complexes.³

The iron-citrate efflux transporter (IceT; formerly MdtD) is a member of the Salmonella typhimurium MFS family that facilitates the efflux of citrate or iron-citrate and has sequence homology to proteins belonging to the drug:proton antiporter-2.⁸ S. typhimurium causes foodborne illness, as the pathogen requires intestinal inflammation in the host for survival.¹¹⁰ IceT is encoded by mdtD, part of the mdtABCD-baeSR gene cluster that also expresses the multidrug transporter (Mdt) domains A, B, and C as well as BaeSR, a two-component system that influences antibiotic tolerance and export in S. tyhpimurium.^{9, 111} Both of these operons comprising this gene cluster are promoted by the increase of intracellular ROS.⁸ In addition, IceT expression has been linked to tolerance to streptonigrin, an iron-sensitive antibiotic.^{8, 112} Expression of IceT has also been shown to reduce cellular iron levels, which has been proposed to be the result of obstruction of Feo import in addition to export of iron citrate.⁸ While this mode of export is intriguing, cytoplasmic interactions between iron and citrate have not been previously reported, it is unclear whether and to what extent IceT and Feo may interact or have cross-talk, and the mechanism by which Fe²⁺ is selected over Fe³⁺ for chelation via citrate has yet to be established.³² Considering that IceT is not homologous to previously characterized MFS transporters or other metal-citrate transporters, a better structural and mechanistic understanding of IceT would shed light onto this unique subset of MFS transporters present in problematic pathogens.³²

Membrane bound ferritin-like proteins

Membrane bound ferritin A (MbfA) belongs to the erythrin-vacuolar iron transport (Er-VIT1) family and is involved in ferrous iron efflux.^113–115 A member of the ferritin-like superfamily, the Er-VIT1-like proteins generally consist of two major domains: a soluble N-terminal ferritin-like domain (also called Er) that houses a diiron binding site, and a membrane-imbedded VIT1 domain that is proposed to function in iron transport to vacuoles based on sequence homology to Arabadopsis VIT1.^{113, 115} In bacteria that utilize MbfA, expression is regulated by the transcriptional iron response regulator (Irr) protein.^{114, 116} In addition, biochemical studies of Agrobacterium tumefaciens suggest that MbfA expression is increased in response to low pH (5.5) and H₂O₂ as well as iron.^{113, 115} Similarly, Bradyrhizobium japonicum MbfA expression is promoted by high iron conditions and has been shown to remediate iron excess, as cells lacking mbfA had elevated iron levels.¹¹⁷ Furthermore, MbfA appears to be relevant in some organisms for general intracellular iron homeostasis, rather than simply correcting for major iron imbalances.¹¹⁷ Taken together, these findings indicate a role for MbfA in iron export to reduce potential cellular damage.^{9, 117}

P_1B-type ATPases

P_1B-type ATPases are a subdivision of the P-type ATPase family of membrane proteins that are found across all kingdoms, including pathogenic prokaryotes, and are utilized to transport a number of metal substrates including Fe²⁺.^{4, 118, 119} The architecture of a P_1B-type ATPase is generally conserved and features three domains: (1) several α-helices (typically ≥ 6) that comprise a transmembrane region for metal transport, (2) a soluble ATP-binding domain (ATP-BD) that includes nucleotide binding (N-) and phosphorylation (P-) subdomains, and (3) a soluble actuator domain (AD) that is hypothesized to communicate changes in the ATP-BD, aid in dephosphorylation, and alter protein dynamics (Figs. 5 and 7).⁴ Two P_1B-type ATPases have been structurally characterized using X-ray crystallography: the Cu⁺-transporting Legionella pneumophilia CopA, and the Zn²⁺-transporting Shigella sonnei ZntA.^{7, 120} Additionally, many (but not all) P_1B-type ATPases have a variable number of soluble metal binding domains (MBDs) that can be found at the N- and/or C-termini, may function as allosteric sites, and may even participate in metal hand-off to the transmembrane domain (Fig. 7).^{4, 121, 122} Structural comparisons to other P-type ATPases as well as bioinformatics-based studies of hundreds of sequences demonstrated conservation of the structure and multiple residues that contact ATP/ADP or facilitate interdomain interactions across the P_1B-type subfamily, indicating a common and general mechanism despite the wide distribution of substrate diversity.^{4, 123, 124}

Figure 7.

Structural elements of P_1B-ATPases. A. Cartoon topology of a P_1B-type ATPase. Soluble N- and/or C-terminal metal-binding extensions (MBDs; gray) are proposed to regulate P_1B-ATPase activity. The transmembrane region may feature N-terminal accessory helices (cylinders labeled A and B; salmon), but six helices are generally conserved across P_1B-ATPases (cylinders 1–6; red). The soluble actuator domain (AD; green) generally lies between helices two and three, and the ATP-binding domain (ATP-BD; blue) is generally located between helices four and five. Figure adapted from Smith et al.⁴ B. X-ray crystal structure of the archetypal metal-transporting P_1B-type ATPase CopA (PDB ID 3RFU) highlighting the fold of the 6 α-helices comprising the transmembrane region (TM; red), the ATP-BD (blue), and the AD (green).⁷ C and D. P_1B-ATPases can have one or several metal binding domains (MBDs) at either terminus. C. The CopA C-terminal MBD (PDB ID 3FRY) features a ferredoxin-like domain-swapped dimer colored with α-helices in purple and β-strands in pink. D. The CzcP N-terminal MBD (PDB ID 4U9R) adopts a duplication of the ferredoxin-like fold, and a second metal binding site located at the domain-domain interface. ‘N’ and ‘C’ represent the amino and carboxy termini, respectively.

Our understanding of the metal specificity (or even promiscuity in some cases) of P_1B-type ATPases is constantly evolving and has led to the identification of at least seven subclasses (P_1B1-P_1B7) that transport numerous metal substrates,⁴ although Fe²⁺ as a substrate remains controversial. The P_1B1–3 subclasses have been heavily studied historically and are known to be involved most commonly in Cu⁺ and Zn²⁺ homeostasis.⁴ The P_1B4-ATPases exhibit the greatest substrate diversity of all the P_1B-type ATPases.^{7, 119, 121, 125–129} Though originally considered to transport exclusively Co²⁺, increasing evidence has emerged demonstrating that P_1B4-ATPases are also capable of transporting Cd²⁺, Fe²⁺, Ni²⁺, and Zn²⁺.^{119, 130–136} Recent reports have suggested that several of the P_1B4-ATPases may be ferrous iron exporters that function in concert with regulators to mitigate iron intoxication (i.e., excess intracellular iron). Work has shown that the peroxide operon regulator (PerR) and the ferric uptake regulator control transcription of the P_1B4-ATPases B. subtilis peroxide-induced ferrous efflux transporter (PfeT) and the Listeria monocytogenes Fur-regulated virulence determinant FrvA.^{131, 132, 137} PerR appears to regulate export by Streptococcus pyogenes Per-regulated metal transporter A (PmtA).¹³⁵ In addition, expression of several P_1B-type ATPases also appears to be induced by ROS. Specifically, B. subtilis PfeT, M. tuberculosis CtpD, and Group A Streptococcus PmtA expression levels have all been linked to the presence and concentration of H₂O₂.⁹ However, despite this potential connection to ferrous iron export, Michaelis constants (i.e., K_½^Fe) indicate that Fe²⁺ is a weak substrate (ranging from 92 μM to 640 μM) for these transporters compared to other metal ions (K_½^Co ranging from 3.4 to 58 μM), suggesting that Fe²⁺ may only be exported by this subfamily when homeostatic conditions are majorly disrupted.^{131–133, 138} Subclasses P_1B5–7 are less well-characterized but are implicated in the transport of Fe²⁺ and/or Ni²⁺.⁴ Specifically, the P_1B5-ATPase subfamily’s substrate specificity is still unknown, but approximately 25% of P_1B5 proteins have a C-terminal diiron-binding hemerythrin-like domain,⁴ implicating iron as a likely substrate.^{139, 140} Additionally, some members of the P_1B6 subfamily are found downstream of Feo genes, prompting the hypothesis that this subfamily may export excess Fe²⁺ that is imported via FeoB.⁴ Undoubtedly more in vivo and in vitro work are necessary to iron out the metal specificity of these recently identified subfamilies.

Conclusions and Further Perspectives

Currently there are several reported prokaryotic ferrous iron importers, but only three of them (Efe, IroT, and Feo) are specific for iron, and all three remain understudied. These systems, and Feo in particular, have been implicated in bacterial pathogenesis, and further research into these importers could reveal novel approaches and potential therapeutic targets for attenuating virulence. Given its broad distribution, additional studies of Feo in particular may yield the greatest benefit, although advancements on any of these systems would be a boon for the development of therapeutic strategies. Excitingly, novel inhibitors against Staphylococcus aureus FeoB were recently described and may offer insight for future antibiotic development strategies to target this or other Fe²⁺ transporters.^{141, 142} However, like Efe and IroT, there are currently no experimentally-determined structures of the full length FeoB protein, representing a major bottleneck. While some advancement has been made in determining metal-protein interactions for IroT, it is not known how Efe and FeoB bind and transport Fe²⁺. In addition, the roles of auxiliary proteins (i.e., EfeB, EfeO, FeoA, and FeoC) in Fe²⁺ uptake are still largely unknown. Thus, more work including additional structural information is needed to elucidate mechanisms of prokaryotic ferrous iron import, especially to target these uptake systems in pathogens.

The identification of dedicated ferrous iron exporters is a newer concept, but homologs are spread across all kingdoms of life, are necessary for cellular iron homeostasis, and are found in several infectious and problematic prokaryotes. If better characterized, these efflux pumps could present a target for antibiotic development against pathogenic bacteria; however, our knowledge of their structures, mechanisms, and metal selectivities represent major (but not insurmountable) barriers. While bioinformatics analyses have proven especially insightful in expanding our knowledge of P_1B-type ATPases, other classes of Fe²⁺ exporters have not been similarly explored. Additionally, like ferrous iron importers, few experimentally-determined structures are known for the ferrous iron efflux pumps. Critically, the structure of any of these membrane proteins in the presence of Fe²⁺ would greatly expand our knowledge of structure and mechanism and could inform selectivity. Despite these major hurdles, we anticipate that the increased availability and widespread instrumentation of cryo-EM techniques could fill this knowledge gap sooner rather than later.

Abbreviations

ABC

ATP-binding cassette

AD

actuator domain

ARTO

alternate respiratory terminal oxidase

ATP

adenosine 5′-triphosphate

ATP-BD

ATP-binding domain

CDF

cation diffusion facilitator

CF

cystic fibrosis

cryo-EM

cryogenic electron microscopy

DNA

deoxyribonucleic acid

Dps

DNA-binding protein from starved cells

Efe

elemental ferrous iron

Er-VIT1

erythrin-vacuolar iron transport

Fab

fragment antigen-binding

Feo

ferrous iron transport

Fie

ferrous iron efflux

FNR

fumarate and nitrate reduction regulatory protein

Frv

Fur-regulated virulence

Fut

ferric uptake system

GAP

GTPase activating protein

GCP

phosphomethylphosphonic acid guanylate

GDI

guanosine 5′-diphosphate dissociation domain

GDP

guanosine 5′-diphosphate

GEF

guanine exchange factor

GMP-PNP

guanosine 5′-[β,γ-imido]triphosphate

GTP

guanosine 5′-triphosphate

HTH

helix-turn-helix

IceT

iron-citrate efflux transporter

IroT

iron transporter

Irr

iron response regulator

LCV

Legionella containing vacuole

mantGDP

2′/3′-(N-methyl-anthraniloyl)-guanosine-5′-diphosphate

mantGMP-PNP

2′/3′-(N-methyl-anthraniloyl)-guanosine-5′-[(β,γ)-imido]triphosphate

MavN

more regions allowing vacuolar colocalization N

MBD

metal binding domain

MbfA

membrane bound ferritin A

Mdt

multidrug transporter

MFS

membrane facilitator superfamily

MntH

H⁺-dependent manganese transport

NFeoB

N-terminal, G-protein-like domain of the ferrous iron transport protein

NRAMP

natural resistance-associated macrophage protein

PerR

peroxide operon regulator

PfeT

peroxide-induced ferrous efflux transporter

PMF

proton motive force

PmtA

Per-regulated metal transporter A

ROS

reactive oxygen species

Sit

Salmonella iron transport

Yfe

Yersinia ferrous iron transport

Zip

ZRT-, IRT-like protein

Znu

zinc uptake

ZupT

zinc uptake transporter