Mitochondrial iron trafficking and the integration of iron metabolism between the mitochondrion and cytosol

Abstract

The mitochondrion is well known for its key role in energy transduction. However, it is less well appreciated that it is also a focal point of iron metabolism. Iron is needed not only for heme and iron sulfur cluster (ISC)-containing proteins involved in electron transport and oxidative phosphorylation, but also for a wide variety of cytoplasmic and nuclear functions, including DNA synthesis. The mitochondrial pathways involved in the generation of both heme and ISCs have been characterized to some extent. However, little is known concerning the regulation of iron uptake by the mitochondrion and how this is coordinated with iron metabolism in the cytosol and other organelles (e.g., lysosomes). In this article, we discuss the burgeoning field of mitochondrial iron metabolism and trafficking that has recently been stimulated by the discovery of proteins involved in mitochondrial iron storage (mitochondrial ferritin) and transport (mitoferrin-1 and -2). In addition, recent work examining mitochondrial diseases (e.g., Friedreich's ataxia) has established that communication exists between iron metabolism in the mitochondrion and the cytosol. This finding has revealed the ability of the mitochondrion to modulate whole-cell iron-processing to satisfy its own requirements for the crucial processes of heme and ISC synthesis. Knowledge of mitochondrial iron-processing pathways and the interaction between organelles and the cytosol could revolutionize the investigation of iron metabolism.

The mitochondrion is mostly appreciated for its role in energy transduction. However, it is less well known that this organelle can be considered a focal point when it comes to the metabolism of the most common transition metal in cells, namely iron (1). It is the reversible oxidation states of iron that enable the mitochondrion to catalyze electron transport via heme- and iron sulfur cluster (ISC)-containing proteins and use this process in energy trans-duction. Considering this alone, it is no wonder that the mitochondrion plays such a critical role in iron metabolism. In fact, the mitochondrion is the sole site of heme synthesis and a major generator of ISCs, both of which are present in mitochondria and cytosol (2). Although the molecular pathways involved in the generation of heme and ISCs are well known, only more recently have some of the molecular players responsible for mitochondrial iron transport been identified. Clearly, these molecular circuits are vital for the supply of the metal ion that is needed for generating the final biologically important end-products, namely heme and ISCs. In contrast to the mitochondrion, at the whole-cell level the molecular pathways and regulation of iron uptake and storage have been well characterized. Hence, these will be first briefly described to provide basic background on the field of iron metabolism (3, 4) before examining what is known regarding the mitochondrion.

Cellular Iron Metabolism and Transport

Iron is an essential metal for the organism because of its unparalleled versatility as a biological catalyst. Consequently, iron is a crucial element required for growth. However, the very chemical properties of iron that allow this versatility also create a paradoxical situation, making acquisition by the organism very difficult. Indeed, at pH 7.4 and physiological oxygen tension, the relatively soluble iron(II) is readily oxidized to iron(III), which upon hydrolysis forms insoluble ferric hydroxides. As a result of this virtual insolubility and potential toxicity because of redox activity, iron must be constantly chaperoned. In fact, specialized molecules for the acquisition, transport, and storage of iron in a soluble, nontoxic form have evolved to meet the organism's iron requirements. Over the last 15 years there has been a wide variety of unique molecules discovered that play a role in iron metabolism, and the most relevant of these to this review are listed in Table S1.

Because of its redox properties, iron can catalyze the production of reactive oxygen species (ROS) that can be highly toxic (3). Therefore, under normal physiological conditions, iron is specifically transported in the blood by diferric transferrin (Tf) (4, 5). All tissues acquire iron by the binding of Tf to the transferrin receptor 1 (TfR1), with this complex then being internalized by receptor-mediated endocytosis (4, 5) (Fig. 1A). Recent studies have demonstrated that the internalization of the Tf-containing endosome via the cytoskeleton is under control of intracellular iron levels (6, 7). In part, this uptake mechanism is mediated by myotonic dystrophy kinase-related Cdc42-binding kinase alpha (MRCKα) (6). This molecule plays a role in organizing the actin cytoskeleton and is up-regulated by iron-depletion through the iron regulatory protein (IRP)–iron-responsive element (IRE) interaction (see below) (6). MRCKα colocalizes with Tf-TfR1 complexes following their internalization and it has been shown that attenuation of MRCKα expression causes a significant decrease in Tf-mediated iron uptake (7). Additionally, it is known that Sec15l1, which is involved in the mammalian exocyst complex (8), plays a role in iron uptake from Tf via its role in exocytosis (9, 10). In fact, Sec15l1 is linked to the Tf cycle through its interaction with Rab11 (a GTPase involved in vesicular trafficking) and a mutation in Sec15l1 leads to the anemia found in hemoglobin-deficit mice (9, 10).

Fig. 1.

Schematics of cellular iron uptake. (A) The process or iron uptake and utilization. (B) The “kiss and run” hypothesis (see text).

Within the endosome, the affinity of Tf for iron is decreased by the low pH generated through the activity of a proton pump (11, 12). Importantly, the TfR1 facilitates liberation of iron from Tf in the pH range attained by the endosome (pH 5–5.5) (13). In vitro, iron release from Tf requires a “trap,” such as pyrophosphate (13), but a physiological chelator serving this role has not been identified. In erythroid cells, once iron(III) is released from Tf in the endosome, it is thought to be reduced to iron(II) by a ferrireductase in the endosomal membrane known as the six-transmembrane epithelial antigen of the prostate 3 (14, 15) (Fig. 1A). After this step, iron(II) is then transported across the endosomal membrane by the divalent metal transporter-1 (DMT1) (16) and forms, as generally believed, the cytosolic low M_r labile or chelatable iron pool (17). This pool of iron is thought to supply the metal for storage in the cytosolic protein ferritin and for metabolic needs, including iron uptake by the mitochondrion for heme and ISC synthesis.

Iron can also be released from the cell by the transporter, ferroportin1 (18) (Fig. 1A). Ferroportin1 expression can be regulated by the hormone of iron metabolism, hepcidin. Hepcidin is a key regulator of systemic iron metabolism (18) and is transported in the blood bound to α2-macroglobulin (18). Hepcidin secretion by the liver is stimulated by high iron levels and also inflammatory cytokines, such as interleukin-6 (19).

Regulation of Cellular Iron Homeostasis

The chelatable iron pool is thought to control the activity of IRPs-1 and -2 (Fig. 1A). The IRPs are RNA-binding proteins that bind to IREs in the 3′- and 5′-untranslated regions in mRNAs of molecules playing crucial roles in the uptake, utilization, export, and storage of iron (e.g., TfR1, ferritin, etc.) (20). IRP-1 performs two functions: (i) regulating iron homeostasis via binding IREs and (ii) having cytosolic aconitase activity when containing an [4Fe-4S] cluster (21). IRP-2 is thought to be the principal RNA-binding protein in vivo and is regulated by iron-dependent proteasomal degradation (22–24). Although the IRP-IRE mechanism plays crucial roles in the regulation of iron metabolism, complete understanding of the homeostatic mechanisms involved in iron metabolism in relation to communication between the cytosol and mitochondrion have yet to be deciphered, and are considered in the following sections.

Intracellular Iron Transport and Communication with the Mitochondrion.

Once iron is transported out of the endosome via DMT1, it enters the chelatable or labile iron pool (Fig. 1A) that traditionally was thought to consist of low M_r complexes (e.g., iron-citrate) (17, 25, 26). The only strong evidence that such a pool exists comes from studies with chelators that mobilize iron from cells (27, 28). However, it is just as likely that these compounds remove iron from organelles and proteins as it is that they chelate iron from genuine cytosolic low M_r complexes (29).

Studies using reticulocytes, which are highly active in terms of iron uptake, demonstrate that these cells contain very little iron as low M_r complexes (30). In fact, the only low M_r iron present had kinetics of iron uptake consistent with an end-product rather than an intermediate (30). Furthermore, the inhibitor of heme synthesis, succinylacetone, led to a reduction in this low M_r iron, suggesting it was heme or a heme-containing molecule (30). These studies, coupled with the findings in previous investigations by others, led to a hypothesis that iron is always transported, at least in erythroid cells, bound within hydrophobic pockets of proteins that act as intermediates (or chaperones) and prevent cytotoxic redox chemistry (30).

As yet, such iron chaperone molecules remain elusive, although Vyoral et al. (31) identified a high M_r intermediate that appeared to donate its iron to ferritin after incubation of K562 cells with Tf. More recently, a protein known as poly (rC)-binding protein 1 has been identified that donates iron to ferritin and may play an important role in this process (32). Although work identifying chaperones that transport iron remain preliminary, it is notable that for copper, which is the second most-abundant transition metal ion in mammalian tissues, much evidence of chaperone-mediated transport has been described (33, 34). Like iron, copper is also cytotoxic because of its redox activity, and is never found free at significant concentrations, but is always bound to transporters, chaperone molecules, and target proteins (33–36). In fact, copper uptake and efflux involves chaperones as well as organelle interactions (33–36).

Considering the arguments above, it has been shown that highly efficient iron targeting to the mitochondrion is evident in erythroid cells where ferrochelatase inserts iron(II) into protoporphyrin IX [PPIX (11)]. Because Tf-bound iron is efficiently used for heme synthesis (11, 30) and no low M_r cytoplasmic iron-transport intermediate has been found in reticulocytes, an intimate direct transfer of iron from Tf to the mitochondrion was proposed to occur (37, 38). This idea has developed in more recent years and has led to the “kiss and run” hypothesis (11) (Fig. 1B). This model suggests that a direct transfer of iron from the Tf-containing endosome to the mitochondrion occurs, by-passing the cytosol (11, 28, 30, 39). The precise molecular details involved in a possible contact between the endosome and mitochondrion remain unknown. However, molecular motors, docking complexes, myosin Vb (40), and molecules involved in regulating the cytoskeleton, namely MRCKα (5), and also vesicular docking [e.g., Sec15l1 (9, 10)], point to mechanisms that may facilitate kiss and run (29).

Communication Between the Cytosol and Mitochondrion: Regulation of Iron Uptake.

If a direct system of iron transfer exists between the endosome and mitochondrion, regulation of iron metabolism by the cell must be coordinated with iron uptake and transport to the mitochondrion. Such homeostatic mechanisms are not fully understood, but may work in concert with the IRP-IRE mechanism to regulate iron homeostasis. Evidence that regulatory processes exist that couple cytosolic and mitochondrial iron metabolism can be deduced from studies in vitro and in vivo. For example, when heme synthesis is inhibited in the mitochondrion using succinylacetone, iron continues to enter the organelle (30, 41, 42). This finding could be interpreted in two ways. First, it may indicate little communication and coupling between the cytosolic and mitochondrial iron-metabolism machineries, as iron continues to be transported into the organelle irrespective of the lack of heme synthesis. Second, it could suggest that iron continues to enter the mitochondrion because of a failure to generate heme, which leads to a signal-to-import iron in an effort to rescue heme synthesis.

The latter concept suggests a coupling between the cytosolic- and mitochondrial iron-processing pathways and is supported by several lines of evidence. For example, Tf and Tf-dependent iron uptake are increased when heme synthesis is inhibited using reticulocytes in vitro (41, 42), and this is accomplished by an increase in TfR1 cycling rate (42). This response occurs despite on-going mitochondrial iron-loading because of the inhibition of heme synthesis. Hence, the lack of heme generation appears to result in a compensatory increase in Tf-bound iron uptake. Other evidence comes from studies in vivo using the muscle creatine kinase conditional frataxin knockout mouse (43), where conditional frataxin-deletion in cardiomyocytes leads to depressed ISC and heme synthesis, mitochondrial iron-loading, TfR1 up-regulation, and thus increased iron uptake from Tf (44, 45) (Fig. 2).

Fig. 2.

Model of alteration in iron uptake in Friedreich's ataxia. Frataxin-deficiency results in increased mitochondrial-targeted iron uptake and cytosolic iron-deficiency (see text).

Under these latter conditions, in the absence of frataxin, the defect in ISC and heme synthesis appears to lead to a “rescue response,” where iron uptake is increased and targeted away from the cytoplasm and ferritin toward the mitochondrion. This response leads to a relative cytosolic iron-deficiency and mito-chondrial iron-loading (44, 45). In the latter case, it was hypothesized that a signal (potentially caused by decreased ISC or heme levels) was sent from the mitochondrion to the cytoplasm to up-regulate iron uptake to compensate for the depressed ISC and heme synthesis resulting from frataxin-deficiency (45). Although increased directional targeting of iron to the mitochondrion occurred, leading to iron-loading in this organelle, it does not lead to an effective reversal of the mitochondrial defect (44, 45), because frataxin, which plays a crucial role in ISC and heme synthesis, is deleted in cardiomyocytes (43, 46–48). Similarly, Li et al., using cultured cells in vitro, have also demonstrated that there is a cytosolic iron-deficit in fibroblasts and lymphoblasts from Friedreich's ataxia (FA) patients, as shown by increased IRP1/2-RNA-binding activity (49). Finally, overexpression of mitochondrial ferritin (Ftmt) in the mitochon-drion leads to mitochondrial iron-loading and cytosolic iron-deprivation (50). Hence, there is evidence that alterations in mitochondrial iron homeostasis lead to changes in cellular iron metabolism, suggesting communication between compartments and a modulating influence of the mitochondrion.

The discussion above demonstrates that cytosolic iron metabolism is regulated, at least in part, by the mitochondrial demand for iron that is critical for heme and ISC synthesis. This finding appears logical, because the mitochondrion is a focal point of iron processing. Such compensatory alterations in iron trafficking, demonstrating communication between the cytosol and mitochondrion, are also found in other conditions where mitochondrial iron-processing is disturbed (51). For example, similar mitochondrial iron-loading occurs in patients with a splice mutation in the ISC scaffold protein, Iscu (52), or with a deficiency in glutaredoxin-5 that is involved in ISC/heme synthesis (53, 54). Moreover, in a patient with a glutaredoxin-5 deficiency, this problem led to decreased ferritin and increased TfR1 expression (54), which was a similar metabolic phenotype to that found in frataxin knockout mice (44, 45). Further studies are necessary to define at the molecular level how this communication occurs between the cytosol and the mitochondrion, and may be facilitated by a better understanding of the mechanisms of mitochondrial iron import and release.

Considering the regulation of whole-cell iron metabolism via molecular defects in the mitochondrion described above, it is intriguing to consider the possibility that similar mechanisms of communication exist for other organelles that play key roles in iron metabolism. For example, the lysosome is involved in ferritin degradation, recycling of stored iron, and autophagy of other organelles, including the iron-rich mitochondrion (55). The rupture of a small number of lysosomes is an early upstream event in many cases of apoptosis, particularly oxidative stress-induced apoptosis; necrosis results from a major lysosomal rupture. Consequently, it has been suggested that regulation of the lysosomal content of redox-active iron appears essential for cell survival (56).

Mitochondrial Iron Import

Although DMT1 is responsible for the export of iron(II) from the endosome, the itinerary of the metal from the cytosol to the inner mitochondrial membrane is not well understood. However, it is known that the inner membrane of the mitochondrion contains proteins capable of transporting iron into the mitochondrial matrix. Foury and Roganti suggested a role for the eukaryotic, mitochondrial solute carriers, Mrs3 and Mrs4, in mediating mitochondrial iron metabolism in yeast (57). A further study demonstrated that these proteins are essential for yeast growth under iron-limiting conditions, suggesting that, at least in this organism, an additional iron transport mechanism is present (58). A recent investigation of the function of Mrs3 and -4 in small mitochondrial particles showed these proteins transport iron(II) along a concentration gradient (59).

Studies with the zebrafish mutant with profound anemia, frascati, led to the discovery of the Mrs3 and -4 homolog, termed mitoferrin-1 (SLC25A37) and mitoferrin-2 (SLC25A28) (60). These homologous proteins are important for mitochondrial iron uptake in erythroid and nonerythroid cells, respectively (60, 61) (Fig. 1A). Mitoferrin-1 is localized on the inner mitochondrial membrane and functions as an essential importer of iron for mitochondrial heme and ISC in erythroblasts and is necessary for erythropoiesis (60). Mitoferrin-1 is highly expressed in erythroid cells and in low levels in other tissues, whereas mitoferrin-2 is ubiquitously expressed (61). The half-life of mitoferrin-1 (but not mitoferrin-2) increases in developing erythroid cells and this may be part of a regulatory mechanism aiding mitochondrial iron uptake (61). Recently, Abcb10, a mitochondrial inner-membrane ATP-binding cassette transporter, was found to physically interact with mouse Mfrn1, and thereby enhance the stability of the protein and increase mitochondrial iron-import (62). Interestingly, Abcb10 has been suggested to play some role in heme synthesis and can be rapidly induced by the transcription factor, GATA-1, which plays a role in erythroid differentiation (63).

It is still unknown how iron is transported across the outer mitochondrial membrane and clearly other transporters may yet be identified. Considering this, a large-scale computational screen identified three potential transporters that may be involved in mitochondrial iron metabolism, namely SLC25A39, SLC22A4, and TMEM14C (64). In fact, targeted knockdown of these genes in zebrafish resulted in profound anemia. Furthermore, silencing of Slc25a39 in murine erythroleukemia cells impaired iron incorporation into PPIX to form heme (64).

It is also notable that a mutation in the transmembrane mitochondrial protein, sideroflexin1, is responsible for the skeletal abnormalities and hematological phenotype in the flexed-tail mouse model, namely hypochromic, microcytic anemia, and siderotic granules in erythrocytes (65). Because of its predicted five-transmembrane domains, this molecule has been suggested to be a transporter essential for mitochondrial iron metabolism. Indeed, it has been speculated to transport molecules into or out of the mitochondrion for iron utilization or heme synthesis (65). Sideroflexin may not necessarily transport iron and could mediate the uptake of other metabolites essential for heme synthesis (e.g., pyridoxine) that is necessary for the formation of pyridoxal phosphate, a coenzyme required for the first step in heme production (65).

Mitochondrial Iron Metabolism

Three Major Pathways: Heme Synthesis, ISC Synthesis, and Storage.

Once iron is transported into the mitochondrion it can then be used for heme synthesis, ISC synthesis, or stored in Ftmt. It is essential that mitochondrial iron is maintained in a safe form to prevent oxidative damage, as mitochondria are a major source of cytotoxic ROS (3). Hence, it is likely that as in the cytosol, iron is carefully transported within the mitochondrion in a form distal to the aqueous environment, deep in the hydrophobic pockets of communicating proteins that form iron transport pathways. Although the molecular nature of these circuits remains unclear, the pathways that use iron are well known and are discussed below.

Mitochondrial iron storage: mitochondrial ferritin.

Like a typical ferritin, Ftmt shows ferroxidase activity and binds iron (66). In contrast to cytoplasmic ferritin, Ftmt is encoded by an atypical intronless gene (66). However, its role is unclear, particularly considering its tissue distribution pattern. Indeed, Ftmt was found at the highest levels in the testes and the erythroblasts of sideroblastic anemia patients (67, 68). Mitochondrial ferritin has also been detected in the heart, brain, spinal cord, kidney, and pancreas (67). Unlike cytosolic ferritin, Ftmt is not highly expressed in the liver and spleen, suggesting that it plays a distinct role. These findings led to the hypothesis that Ftmt plays a role in protection from iron-dependent oxidative damage (67).

The role of Ftmt in iron metabolism was examined by employing a stably transfected cell line that hyper-expresses Ftmt (50). These studies showed that overexpression of Ftmt resulted in increased IRP-1/2-RNA-binding activity, decreased cytosolic and mitochondrial aconitase activity (suggesting decreased ISC synthesis), decreased cytoplasmic ferritin, increased TfR1 expression, decreased heme synthesis, decreased frataxin expression, and increased iron-loading of Ftmt (50). Hence, Ftmt overexpression leads to partitioning of iron away from heme and ISC synthesis in the mitochondrion. This effect not only alters mitochondrial iron metabolism, but also whole-cell iron metabolism (50), leading to a cytosolic iron-deficiency that reduces the proliferation of neoplastic cells hyper-expressing Ftmt in vivo (69). As already discussed, very similar alterations of gene expression also occur after ablation of frataxin expression (44, 45) and indicate communication between the cytosol and mitochondrion.

Iron-sulfur cluster biosynthesis.

Being a major site of ISC assembly, the mitochondrion plays a pivotal role in the biosynthesis of ISC proteins (1, 2). ISCs consist of iron and sulfide anions (S^2-), which assemble to form [2Fe-2S] or [4Fe-4S] clusters (2). These clusters form cofactors in proteins that perform vital functions, such as electron transport, redox reactions, metabolic catalysis, and other such functions (70). In eukaryotic cells, more than 20 molecules facilitate maturation of ISC proteins in the mitochondria, cytosol, and nucleus (2). Functional defects in ISC proteins or components involved in their biogenesis lead to human diseases (71, 72). Biosynthesis of ISCs and their insertion into apo-proteins requires both the mitochondrial and cytosolic machinery. The first molecule identified in the ISC machinery was the enzyme NifS from Azobacter vinlandii, which is a cysteine desulfurase that participates in ISC assembly as a sulfur donor (73). This enzyme is highly conserved and in humans is known as Nfs1 (74).

ISCs are assembled on scaffold proteins and then transferred to apo-proteins. In Escherichia coli, this scaffold protein is known as ISC assembly protein U (IscU) and leads to sequential assembly leading to [2Fe-2S] units that then form a [4Fe-4S] cluster (75). In yeast cells, these reactions are accomplished by Isu1 and -2 (76), but in humans, the function of IscU is performed via the splicing of IscU mRNA to lead to two transcripts which generate a cytosolic (Iscu1) or mitochondrial (Iscu2) isoform. The maturation of extramitochondrial ISC proteins requires the mitochondrial ISC assembly system (77). The mitochondrion contributes a yet-to-be discovered compound necessary for the biogenesis of ISCs outside of this compartment (i.e., in the cytosol or other organelles) (70).

The mechanism involved in iron delivery to Iscu1 is not clear, but it has been suggested to involve frataxin as an iron donor (2). Moreover, frataxin has been shown to play a critical role in the early stages of ISC genesis (78), and this is discussed in detail below. In yeast, Isu1 only provides a cluster for de novo cluster production from which an HSP70-type chaperone system transfers the new clusters to apo-proteins (2). Yet another molecule, ABCB7, appears to mediate cytosolic ISC biogenesis (79). The maturation of cytosolic ISCs is inhibited by mutations in ABCB7 and this causes the disease, X-linked sideroblastic anemia with cerebellar ataxia (XLSA/A). This condition is characterized by loss of cytosolic ISC proteins, defects in heme metabolism, and increased mitochondrial iron levels (79).

Heme biosynthesis.

The third major mitochondrial metabolic pathway that utilizes iron is that of heme synthesis that is exclusive to this organelle (1, 11). Heme is synthesized by a pathway composed of eight sequential reactions in the mitochondrion and cytoplasm (1, 11). The first and last three steps in the heme biosynthesis pathway take place in the mitochondrion. The first enzyme in the path-way, 5-aminoleuvulinate synthase (ALAS), catalyzes the condensation of glycine and succinyl CoA to form 5-aminolevulinate (1, 11). There are two genes for ALAS, the ubiquitously expressed ALAS1 and the erythroid-specific ALAS2, which is under the regulation of iron via the IRP system (1). In nonerythroid cells, the rate of heme synthesis is dependent on the rate of 5-aminolevulinate synthesis via ALAS1 (11). In contrast, in erythroid cells, the rate-limiting step may be the delivery of Tf-iron (11).

The subsequent four steps of heme biosynthesis take place in the cytosol, following which coproporphyrinogen III is decarboxylated and then oxidized in the mitochondrial intermembrane space to produce PPIX (1, 11). Coproporphyrinogen may be transported into mitochondria by either peripheral-type benzodiazepine receptors (80) or potentially ABCB6 (81). The seventh step, which is catalyzed by protoporphyrinogen IX oxidase generates PPIX. In the final reaction of the pathway, iron(II) is inserted into PPIX by the ISC protein, ferrochelatase, in the mitochondrial matrix (11).

Mitochondrial Iron Export.

Apart from importing iron, the mitochondrion synthesizes heme and ISCs that subsequently are transported out to the cytosol. These export processes are important to understand, as decreased release of iron as heme or ISCs—or their precursors—can contribute to mitochondrial iron-loading, as found in the frataxin knockout mouse (45). A candidate iron exporter has been identified, namely mammalian mitochondrial ABC protein 3 (MTABC3; also known as ABCB6) (82). This protein can rescue mitochondrial iron-loading, respiratory dysfunction, and mitochondrial DNA damage in atm1-deficient yeast cells (82). Of relevance, the human ortholog of atm1 is ABCB7, which can complement atm1 deficiencies in yeast cells, enabling the maturation of ISC-containing proteins in the cytosol (79). It has been suggested that ABCB7 transports ISCs to the cytoplasm (79), although this has not been directly shown.

It is unknown how heme is exported from the mitochondrion. However, its low solubility and highly toxic nature suggest an efficient heme-carrier must be involved. Considering this, heme-binding protein 1 has been identified as a candidate for this carrier (83). The expression of this molecule is ubiquitous, but it is also increased during erythroid differentiation and high levels are found in the liver (83). Heme-binding protein 1 binds one heme per mole of protein (83) and, although it could be a candidate for heme transport, direct evidence for this is lacking.

Diseases of Mitochondrial Iron Metabolism

Mechanisms of iron transport across mitochondrial membranes have evolved to supply the necessary iron to mitochondria and also maintain the balance of cytosolic iron (1). Mitochondrial iron levels must be tightly controlled as iron induces ROS, which can damage the organelle (84). The importance of these tight controls is highlighted by the fact that alterations in mitochondrial iron homeostasis have pathological consequences (1, 70). In recent years, cellular and animal models of mitochondrial iron dysfunction have provided valuable information in identifying new proteins to elucidate the pathways involved in mitochondrial iron homeostasis. Interesting examples of mitochondrial diseases that have provided important insight into the processes of mitochondrial iron metabolism are Friedreich's ataxia and sideroblastic anemia. These conditions are discussed in detail in the following sections.

Friedreich's Ataxia and the Metabolic Role of Frataxin.

FA is a rare disease leading to severe neuro- and cardio-degeneration and is caused by a deficiency of frataxin (85). The decrease in frataxin expression is caused by an intronic GAA-repeat expansion in intron-1 of FRDA. Frataxin is a vital protein that is highly expressed in tissues rich in mitochondria (e.g., heart and neurons) (86), with deletion leading to lethality (87).

The suggested functions for frataxin all relate to iron metabolism (Fig. 3) and include ISC and heme biogenesis, as well as iron storage. Frataxin was linked to ISC proteins by the observation that there was a deficiency in ISC cluster proteins in knock-out mice, FA patients, and yeast (43, 46–48). Frataxin has also been implicated in iron scavenging (88), regulating oxidative stress (89), and as an iron chaperone (90). To date, the cumulative evidence suggests frataxin is involved in the maintenance of iron homeostasis (44, 45). Frataxin is an inner mitochondrial membrane and mitochondrial-matrix protein (85). However, as there does not appear to be any structural feature that would anchor frataxin to the mitochondrial membrane, it is possible that membrane-associated frataxin is bound to other proteins in a macromolecular complex containing ferrochelatase (91).

Fig. 3.

Schematic of the possible functions of frataxin. (A) Frataxin may perform a mitochondrial iron-storage function similar to ferritin. (B) Frataxin may function as an iron chaperone by binding iron and then delivering it. (C) The frataxin ortholog, CyaY may function as an “iron-sensing”-negative regulator that inhibits the rate of ISC assembly under conditions of high mitochondrial iron and low ISC apo-acceptor availabilities. (D) Frataxin may also function as a “metabolic switch” that allows the mitochondrion to favor heme or ISC/heme syntheses, depending on frataxin and protoporphyrin IX levels (91, 99) (see text).

Recent insight into frataxin function has come from studies of frataxin orthologs in yeast (i.e., Yfh1) and bacteria (i.e., CyaY) that share a high degree of sequence identity to the human protein (92). A consensus is that the functions of frataxin share a requirement for iron-binding (92–94). Although there are several proposals, the exact function of frataxin remains controversial and is discussed in the following sections.

Frataxin and iron storage.

Frataxin has been proposed to function as a mitochondrial iron storage protein (95) (Fig. 3A). The frataxin ortholog, Yfh1, was reported to be capable of self-assembling into oligomers and then into higher-order multimers in the presence of increasing iron(II) (96). The iron-dependent oligomerization of Yfh1 is associated with the development of ferroxidase activity that appears to exist only in the oligomeric/multimeric structure (92). This finding was proposed to allow Yfh1 oligomers/multimers to store iron in a mineralized and redox-inactive iron(III) state (92, 97). However, because the iron-dependent oligomerization of Yfh1 only occurs in the absence of Ca²⁺ and Mg²⁺, which are typically abundant in the mitochondrion (92, 93, 96), this diminishes support for the iron-storage hypothesis.

Spontaneously oligomerized recombinant human frataxin appears capable of binding and storing iron (97, 98). However, recombinant human frataxin monomers, unlike those of Yfh1, cannot be induced to oligomerize in vitro by iron (98). Moreover, the spontaneous oligomerization of human frataxin appears to be dependent on heterologous overexpression (98). Such considerations argue against an in vivo iron-storage function for human frataxin.

The observation that the manipulation of mitochondrial iron levels does not affect frataxin expression levels is not consistent with a significant role in mitochondrial iron storage (99). Theoretical support for the generality of the “iron-storage” hypothesis was also lessened upon the discovery of Ftmt in higher organisms (66, 100). The existence of Ftmt appears to make redundant any significant iron-storage role for human frataxin. However, yeast cells do not express ferritins (100), and it is possible that Yfh1 may possess an additional iron-storage capacity not shared by frataxin orthologs in higher organisms.

Frataxin as an iron chaperone.

Perhaps the most promising emerging role for frataxin is as an iron chaperone (Fig. 3B) for ISC and heme biosyntheses. Frataxin has been observed to interact with, and presumably donate iron to, iron-dependent proteins involved in ISC and heme biosyntheses (90, 91). For example, Yfh1 appears capable of interacting with the central ISC assembly complex comprising the scaffold protein, Isu, and the cysteine desulfurase, Nfs1, in a manner enhanced by iron(II) (90, 101).

Recently, it has been suggested that frataxin interacts with Isu1 via a highly conserved tryptophan residue (W131a) in its conserved β-sheet region (102). Analogously, human frataxin demonstrates iron-enhanced interactions with Isu and ferrochelatase (91, 103). Importantly, the interaction of frataxin with either Isu or ferrochelatase appears to increase the rate of ISC synthesis (90) or the ferrochelatase-catalyzed insertion of iron(II) into PPIX (91), respectively. Such observations suggest frataxin may function as an iron-donor to Isu and ferrochelatase. Emerging data indicate that the nature of frataxin's facilitatory interactions with the ISC and heme synthesis machinery may be more complex than just simple iron-donation. As discussed in the next section, a possible primary function of frataxin's interactions may be to additionally negatively regulate the respective biosynthetic interactions.

Frataxin as an iron-sensing negative-regulator.

A recent study with CyaY suggests that the protein may act as an “iron-sensor” (Fig. 3C) that negatively regulates the rate of ISC biosynthesis under conditions of high iron and low ISC apo-acceptor availabilities (89). If we extend this model to eukaryotic systems, a deficiency in frataxin expression is presumably deleterious because the rate of ISC biosynthesis may exceed the availability of ISC apo-acceptors, resulting in the overproduction of ISCs that are unstable in an unbound form (89). Essentially, this model suggests that over and above functioning as an iron donor in ISC biosynthesis, frataxin may exert “kinetic control” over the rate of ISC biosynthesis, depending on the relative availabilities of iron and ISC apo-acceptors. This possible iron-sensing role is consistent with the relatively low (i.e., micromolar) affinities of iron-binding by bacterial, yeast and human frataxin orthologs (92). The applicability of this model remains to be examined in mammals. It also needs to be assessed whether any such kinetic control is elicited by frataxin's interaction with ferrochelatase during heme biosynthesis.

Frataxin as an expression-regulated “metabolic switch.”

An extension of the notion of frataxin as a negative regulator to frataxin's functional role in heme biosynthesis may help to explain the decline in frataxin levels during erythroid differentiation (99) (Fig. 3D). That is, because frataxin expression is markedly decreased during Friend cell hemoglobinization (99), it is possible that frataxin may be down-regulated during erythroid differentiation to allow higher rates of heme synthesis, potentially at the expense of decreased levels of ISC synthesis (99). This hypothesis is supported by the observation that the immediate precursor for heme synthesis, PPIX, down-regulates frataxin expression (99). Hence, increased PPIX levels, which indicate a requirement for heme synthesis, lead to decreased frataxin expression and a diversion of iron from other mitochondrial pathways (i.e., ISC synthesis or iron storage) to heme biogenesis (99).

This latter hypothesis is supported by the observation that an increase in frataxin levels relative to ferrochelatase (i.e., above a molar ratio of 1:1 frataxin:ferrochelatase dimer) results in decreased rates of heme synthesis in vitro (91). It has also been observed that iron-bound human frataxin has a higher putative binding affinity for ferrochelatase (17 nM) than Isu (480 nM) (91). These observations provide a basic mechanism that supports the hypothesis that frataxin may allow metabolic switching between ISC and heme synthesis pathways depending on expression levels relative to those of Isu and ferrochelatase (91, 99).

A frataxin metabolon?

Frataxin's putative ability to modulate iron-dependent biochemical reactions through protein-protein interactions is suggestive of the possibility that it may form one or more metabolons, or protein complexes, with proteins involved in ISC and heme biosyntheses (94). The conserved tryptophan residue-131 in frataxin, which is responsible for its interaction with Isu1, suggests the association is crucial for its function, which is underlined by the fact that mutating this residue results in a loss of mitochondrial aconitase activity (102). This hypothesis suggests that the loss of the interaction with Isu1 results in an impairment of ISC synthesis and supports the notion of a functional protein complex involving frataxin. In general, metabolons are dynamic protein complexes that greatly enhance the efficiency of metabolic reactions through processes, such as substrate channeling. On the basis of the possibility that frataxin may act as an iron-donor, it might be expected that frataxin could be part of a mitochondrial metabolon consisting of ISC assembly components, such as Iscu, and heme biosynthesis enzymes, such as ferrochelatase (Fig. 3).

Sideroblastic Anemias.

The characteristic feature of all sideroblastic anemias is the ring sideroblast. This is a pathological erythroid precursor containing excessive deposits of nonheme iron in mitochondria with peri-nuclear distribution responsible for the ring appearance. With considerable simplification, and from the point of view of pathogenesis, sideroblastic anemias can be divided into those with or without a heme synthesis defect in erythroblasts.

It can be proposed that the combination of several factors plays a role in the pathogenesis of mitochondrial iron accumulation in sideroblastic anemias associated with a heme synthesis defect: (i) iron is specifically targeted toward erythroid mitochondria; (ii) iron cannot be used for heme synthesis because of the lack of PPIX; (iii) there is a lack of heme, the negative regulator of iron uptake from Tf; and (iv) iron can leave erythroid mitochondria only after being inserted into PPIX. A key example of this scenario is X-linked sideroblastic anemia, which is caused by mutations in erythroid-specific ALAS2 (104). As already discussed, a distinct form of X-linked sideroblastic anemia is XLSA/A. This condition is caused by mutations in ABCB7 (79), whose product is thought to transfer an ISC precursor from mitochondria to the cytosol. In XLSA/A, as is the case in ALAS2-associated sideroblastic anemia, decreased levels of heme are likely to contribute to the pathogenesis of ring sideroblast formation. In refractory anemia with ring sideroblasts (RARS), there is no evidence for a defect in PPIX formation in patients’ erythroblasts. In some patients with RARS, acquired mutations in subunits of cytochrome oxidase encoded by mitochondrial DNA have been described (105). It has been proposed that this defect could lead to impaired iron reduction that is needed for heme and ISC synthesis, and without this, mitochondrial iron deposits occur.

It can also be hypothesized that erythroid progenitors of patients with RARS, characterized by genomic instability and premature apoptosis, exhibit anomalous induction of mitochondrial ferritin that would lead to a shift of iron into their mitochondria (50, 69). This would prevent the use of iron for hemoglobin synthesis and cause a ring-sideroblast phenotype. In fact, any metabolic abnormality that markedly affects the synthesis of ISCs or heme is likely to result in mitochondrial iron-loading. Actually, it has recently been shown that mutations in the putative glycine transporter, SLC25A38, lead to a rare form of sideroblastic anemia (106). This iron-loading probably occurs because ALAS catalyzes the reaction of glycine with succinyl CoA to generate 5-aminolevulinate. Without glycine, PPIX synthesis would be inhibited which prevents heme generation. Defects in other metabolic pathways, which affect mitochondrial iron metabolism, can also lead to sideroblastic anemia. An example of this would be patients with mutations in pseudouridine synthase-1, which is involved in the processing of mitochondrial tRNAs (107).

Summary

Although the mitochondrion is a focal point for iron utilization in heme and ISC synthesis, there has been little realization that the mitochondrion can play an important role in orchestrating whole-cell iron metabolism. Indeed, analysis of disease states has enabled understanding of the role of the mitochondrion in regulating cellular iron metabolism. There is strong evidence to suggest that the mitochondrion can modulate the cellular iron uptake machinery to satisfy its demand. Considering this evidence, it is likely that signaling pathways exist that allow communication between the mitochondrion and cytoplasm, enabling the mitochondrial iron processing machinery to be supplied with this metal to allow heme and ISC synthesis.