Heme Proteins

The iron in heme proteins is present within a porphyrin, which consists of four pyrrole rings linked by methene bridges; in most hemoproteins the porphyrin is protoporphyrin IX, with four methyl, two vinyl, and two propionyl substituents as shown in Figure 36-1. Heme Fe is always coordinated to the four nitrogen atoms of the tetrapyrrole structure; the fifth, and in some cases the sixth, coordination position is occupied by amino acid residues of the protein. Heme itself is Fe²⁺–protoporphyrin IX, whereas the corresponding Fe³⁺–protoporphyrin IX complex is defined as hemin. In some heme proteins, such as cytochrome c, the porphyrin is covalently linked to amino acid residues of the protein.

Heme proteins are involved in a great number of functions (see Box 36-1). The heme proteins include the oxygen carriers hemoglobin and myoglobin and many heme-containing enzymes such as cytochrome oxidase, peroxidases, catalases, and cytochrome P450s, which are involved in the activation of O₂ or in the metabolism of peroxides (ROOHs). The iron atoms of these heme proteins possess an unoccupied sixth coordination site. This sixth site serves as the binding site for O₂, other inorganic ligands (e.g., nitric oxide [NO]), or organic ligands (e.g., through carbon–iron bonds in cytochrome P450s) (Crichton, 2009). Examples of such heme-containing enzymes are listed in Box 36-1 and include monooxygenases (hydroxylases), dioxygenases, and peroxidases. In contrast, in most of the electron transporting cytochromes, the sixth coordinate position of the Fe²⁺ of the heme is occupied by an amino acid residue of the protein (often a histidine or methionine residue). In contrast to cytochrome a₃ and cytochrome P450, which both bind oxygen and transfer electrons, most cytochromes only transfer electrons, with the iron in the heme alternating between the Fe²⁺ and Fe³⁺ states.

Iron-Sulfur Cluster Proteins

The most common types of iron-sulfur centers are 2Fe-2S and 4Fe-4S clusters, which consist of equal numbers of iron and sulfide ions; the iron atoms of each cluster are also coordinated to the sulfur atoms of four cysteine residues in the protein, as shown in Figure 36-2. When both the sulfide atoms and the cysteine residues are considered, one can see that each Fe atom is tetrahedrally coordinated by four sulfur atoms. Most Fe-S cluster proteins are involved in electron transfer reactions. These include proteins of the mitochondrial electron transport chain as well as various mini–electron transport systems, such as the adrenodoxin (i.e., ferredoxin 1) component of mitochondrial cytochrome P450 systems. However, a number of Fe-S proteins have functions other than electron transfer, including catalysis and acting as biological sensors. The mitochondrial enzyme aconitase contains a 4Fe-4S center that acts as a Lewis acid in the dehydration reaction in which citrate is converted to isocitrate. As discussed later in this chapter, there is also a cytosolic aconitase, which is known as iron regulatory protein 1 (IrRP1) that acts as a biosensor for iron levels within the cell.

Mononuclear Nonheme Iron Proteins

A number of enzymes contain a single Fe atom, which is coordinately bound to imidazole (histidine) or carboxylate (glutamate and aspartate) ligands on the protein; these enzymes are involved in reactions that use O₂ as substrate. The aromatic amino acid hydroxylases, which require tetrahydrobiopterin as a cofactor, constitute the first subgroup. A second subgroup requires a keto acid substrate (usually α-ketoglutarate) and a reducing agent (ascorbate). In these reactions, one of the oxygen atoms from O₂ is transferred to the substrate, which is hydroxylated, and the other oxygen atom is transferred to the keto acid cosubstrate, which undergoes oxidative decarboxylation. (See Chapter 27 for examples and an explanation of the role of ascorbic acid in these reactions.) Finally there are some other mononuclear iron-containing enzymes, which add both oxygen atoms from O₂ to a single substrate; these dioxygenases include the 15-lipoxygenase involved in eicosanoid synthesis and cysteine dioxygenase involved in cysteine catabolism and taurine synthesis (see Chapters 14 and 18).

Dinuclear Nonheme Iron Proteins

A family of proteins containing carboxylate-bridged dinuclear iron sites is involved in functions as diverse as the oxidation of Fe²⁺ to Fe³⁺ (e.g., the H-chain of ferritin), the catalysis of hydroxylation, epoxidation, and desaturation reactions; and the conversion of ribonucleotides into deoxyribonucleotides (e.g., ribonucleotide reductase) (Nordlund and Eklund, 1995; Sazinsky and Lippard, 2006). The common link in all of these dinuclear iron proteins is that they all contain a four helical bundle protein fold, which surrounds a (μ-carboxylato) diiron core with the two iron atoms separated by 0.4 nm or less, and that these iron atoms react with dioxygen (O₂). An example of this class of proteins is ribonucleotide reductase; ribonucleotide reductase catalyzes the reduction of the four common ribonucleotides to their corresponding deoxyribonucleotides, which is essential for deoxyribonucleic acid (DNA) synthesis. The structure of the diiron binding site is represented in Figure 36-3 (Nordlund and Eklund, 1995): each Fe²⁺ ion is octahedrally coordinated by four carboxylate or imidazole groups of the protein. The reaction of the apoenzyme with O₂ results in formation of an Fe³⁺–O–Fe³⁺ bridge and generation of a tyrosyl free radical in the protein, which plays an essential role in catalysis.

The major plasma protein involved in the transport of iron is transferrin (Crichton, 2009; Sargent et al., 2005; Wessling-Resnick, 2000). Transferrin, a glycoprotein synthesized principally in the liver, is composed of a single polypeptide chain of about 680 amino acids with two iron-binding sites. The closely related lactoferrin is found in human milk as well as in secretions such as tears and saliva, and is secreted by neutrophils; its principal function is as an iron scavenger preventing the proliferation of invading microorganisms. The transferrins have a characteristic bilobal structure, as shown for the diferric form of human lactoferrin in Figure 36-4, A. The similar amino- and carboxy-terminal lobes of transferrin are organized into two distinct domains, and each lobe contains one iron-binding site. The hexadentate iron binding site (Figure 36-4, C), consists of four protein ligands (an aspartate, a histidine, and two tyrosine residues) plus two oxygen atoms from the synergistically bound carbonate anion (CO₃^2–). Together they form a nearly ideal metal coordination sphere, with tight Fe³⁺ binding (K_d = 10^–19 to 10^–20 M^–1). Neither Fe³⁺ nor CO₃^2– is bound significantly in the absence of the other. The multicopper oxidase ceruloplasmin and its enterocytic homolog hephaestin oxidize Fe²⁺ to Fe³⁺ before iron is incorporated into apotransferrin (the iron-free form of transferrin). Iron release from transferrin is described later.

Binding of iron by transferrin results in striking conformational changes in each of the two lobes (Figure 36-4, B). In the absence of bound Fe³⁺, the apotransferrin molecule is flexible, with the two domains of each lobe free to swing apart and adopt an open conformation (Figure 36-4, B, left). In the presence of Fe³⁺ and CO₃^2– they take up a pincerlike closed position around the iron atom, bringing all six ligands involved in iron binding together (Baker et al., 2003).

In healthy individuals, transferrin is present in the plasma at a concentration of 25 to 50 μM but is only approximately 30% saturated with iron. The distribution of the plasma transferrin forms pool is 27% diferric, 34% monoferric (23% with iron bound to the N-lobe and 11% with iron bound to the C-lobe), and 39% apotransferrin. Thus plasma has significant excess iron-binding capacity. Transferrin is distributed throughout most of the extracellular fluids of the body.

Transferrin Receptors

Transferrin receptors are involved in the cellular uptake of iron from the circulation, and they have highest affinity for diferric transferrin. Transferrin receptor 1, is a transmembrane glycoprotein composed of two identical 95-kDa monomers linked by a pair of disulfide bridges. Each monomer consists of 760 amino acid residues organized into an amino-terminal cytoplasmic domain, required for intracellular trafficking of the transferrin–transferrin receptor complex, a short membrane-spanning segment, and a large extracellular domain. The three-dimensional structure of the human transferrin receptor 1-binding extracellular domain has been determined (Lawrence et al., 1999) as has that of the human diferric transferrin–transferrin receptor 1 complex (Cheng et al., 2004), shown in Figure 36-5. Each receptor dimer can bind two transferrin molecules, one to each subunit. The receptor is fundamental to the cellular binding and uptake of iron-bearing transferrin. A second transferrin receptor, transferrin receptor 2, has been identified, which binds transferrin with lower affinity than transferrin receptor 1. It appears to play an important role in the regulation of systemic iron homeostasis.

To prevent the unwanted effects of iron-catalyzed free radical generation or iron oxidation and its subsequent precipitation, it is imperative that the cell stores excess iron safely. The iron-storage proteins ferritin and hemosiderin fulfill this cellular “housekeeping” function in all cells; in addition they serve as an iron reservoir in specialized cells (macrophages of the reticuloendothelial system and parenchymal cells of the liver) from which body iron pools can be replenished as iron is used or lost (Liu and Theil, 2005; Crichton, 2009; Arosio and Levi, 2010). The major storage sites of ferritin in normal subjects are liver, spleen, and skeletal muscle. Hemosiderin, which is thought to be the lysosomal degradation product of ferritin, represents a very small fraction of normal body iron stores, and is present mostly in macrophages. It can increase dramatically in iron overload.

The iron-free apoferritin molecule is composed of 24 subunits with molecular weights around 20 kDa each and has a total molecular mass of approximately 500 kDa. The apoferritin subunits form a hollow protein shell with an external diameter of 12 to 13 nm, delimiting a cavity of diameter 7 to 8 nm, within which iron can be stored in a nontoxic, water-soluble, yet bioavailable form. In mammalian ferritins, the iron core resembles the mineral known as ferrihydrite. The ferritin subunits (Figure 36-6, A) are roughly cylindrical, 5 nm long and 2.5 nm wide, with approximately 80% of the amino acid sequence present in five α-helices. Ferritin with its 24 subunits has a high degree of symmetry, and Figure 36-6, B and C, show a ferritin molecule viewed down the fourfold and threefold axes of symmetry, respectively (Crichton and Declercq, 2010).

Human ferritins are made up of two subunits of distinct amino acid sequence, known as H and L. The H subunits are characterized by a (μ-carboxylato) diiron core with ferroxidase activity, oxidizing Fe²⁺ to Fe³⁺, whereas the L subunits, which lack this center, are thought to be involved in the nucleation of the iron core of ferritin. With two types of subunits, 25 “isoferritins” can be built. Variations in the proportions of the two subunits in ferritins of different tissues exist and reflect the tissue-specific functions of ferritin. For example, in liver and spleen, which play a major role in iron storage, ferritins have a much higher proportion of the L subunit, whereas in human heart and brain, where iron detoxification is more important, heteropolymers rich in H subunits are predominant.

The effective storage of iron in human ferritins requires contributions from both subunit types, which play complementary roles in storing, detoxifying, and maintaining Fe³⁺ in a soluble form. In the first step of iron incorporation, Fe²⁺ penetrates the apoferritin shell through channels at the surface of the protein and is oxidized at dinuclear iron centers (known as ferroxidase centers) localized in the H subunits. Fe³⁺ then migrates to nucleation centers within L subunits, in the interior of the protein shell. Once a nucleus of iron core has formed at the nucleation sites, deposition of iron on this inorganic core accomplishes the further incorporation of iron. The maximum storage capacity of each ferritin molecule is approximately 4,500 iron atoms. The mechanism of iron release from ferritin is not at all well understood, although its mobilization appears to require degradation of the ferritin molecule, within both the proteasome and the lysosome (De Domenico et al., 2006).

Hemosiderin is a lysosomal storage form of iron, which is found particularly in conditions of iron overload, such as hereditary hemochromatosis and transfusion-dependent hemoglobinopathies such as the thalassemias. The hemosiderins are thought to be the product of lysosomal degradation of ferritin (Crichton, 2009) The iron cores in hemosiderins isolated from iron-loaded animals and humans are ferrihydrite-like, similar to those in ferritin, but are less available to chelation than are those in ferritins (Ward et al., 2000).

Use of Iron for Erythropoiesis

Erythroblasts (nucleated, developing red blood cells) are generated by the bone marrow from stem cells. An erythroblast eventually loses its nucleus to become a reticulocyte and ultimately a mature red blood cell called an erythrocyte, by which stage hemoglobin synthesis has ceased. Transferrin-bound iron is taken up by erythroblasts and transported into the mitochondria, where it is incorporated into protoporphyrin IX by the mitochondrial enzyme ferrochelatase to form heme. Each day in the normal adult approximately 2 × 10¹¹ red blood cells are produced, and this requires more than 2 × 10²⁰ atoms, or almost 24 mg, of iron. The majority of this (~17 mg) is incorporated into hemoglobin.

Recycling of Iron from Senescent or Damaged Red Blood Cells

Once released from the bone marrow, mature erythrocytes (red blood cells) circulate for approximately 120 days, and thereafter are recycled by cells of the mononuclear phagocytic system. As the erythrocytes circulate in the blood, a number of physicochemical changes occur. After senescent erythrocytes are taken up and lysed within the lysosomes of mononuclear phagocytic cells (macrophages), the heme is broken down by heme oxygenase and the iron released from heme breakdown is exported from the lysosomes into the cytosol via NRAMP1 (natural resistance-associated macrophages protein 1, a divalent metal transporter homologous to DMT1). Export of iron from the macrophages occurs via ferroportin in conjunction with iron reduction by plasma ceruloplasmin and its loading onto plasma transferrin.

As iron is used for erythropoiesis and recovered from senescent erythrocytes by the mononuclear phagocytic system, a small amount of iron also moves in and out of iron storage forms, ferritin and hemosiderin, in these cells. In addition, a fraction of the newly formed erythrocytes are destroyed within the bone marrow itself (ineffective erythropoiesis) and their iron is released for reuse.

Central Role of Transferrin in Iron Transport and Cellular Uptake

As mentioned earlier, transferrin has two iron-binding sites and transports iron between the erythroid marrow and the mononuclear phagocytic cells, as well as other tissues. The overall saturation of plasma transferrin with iron is normally about one third; so apotransferrin, the two monoferric transferrins, and diferric transferrin are all normally found in plasma.

Transferrin-dependent Cellular Iron Uptake

Transferrin-bound iron uptake by cells is mediated by expression of transferrin receptors on their plasma membranes. At physiological pH, the affinity of transferrin for its receptor correlates directly with the degree of iron occupancy. Apotransferrin has negligible affinity for the transferrin receptor at pH 7.4, whereas diferric transferrin has the highest affinity. Iron is delivered to cells via the transferrin-mediated cell cycle, as summarized in the upper left-hand corner of Figure 36-8. The diferric transferrin binds to its receptor and the complex is invaginated into clathrin-coated pits to form vesicles that fuse with endosomes. This delivers the vesicle contents into the interior of the endosome, where the pH is reduced to around 5 to 6 by the action of an ATP-dependent proton pump. Iron release from diferric transferrin is facilitated at acidic pH by protonation of the bound carbonate, thereby weakening the coordination of the iron, and by triggering conformational changes in both transferrin and the transferrin receptor, which stabilize the transferrin molecule in its apo-transferrin conformation. The Fe³⁺ is then reduced to Fe²⁺ by members of the six-transmembrane epithelial antigen of the prostate (STEAP) family of metalloreductases (Ohgami et al., 2005, 2006) for transport of Fe²⁺ out of the late endosome/lysosome into the cytosol via the divalent cation transporter, DMT1.

The apo-transferrin has an extremely high affinity for the transferrin receptor in the more acidic environment of the endosome. By remaining bound to the receptor, apo-transferrin largely escapes degradation by the lysosomal enzymes, and the majority of the apo-transferrin–transferrin receptor complexes are returned to the cell surface after iron is released. At the cell surface, the slightly alkaline pH allows the apo-transferrin to dissociate, leaving the transferrin receptor attached to the plasma membrane. Then the apo-transferrin is returned to the circulation where it is once again able to bind iron and participate in further cycles of iron delivery via the transferrin cycle, which ensures iron uptake by most cells.

Transferrin and Ferroportin-Dependent Iron Export

Export of ferrous iron from cells occurs via ferroportin, which works together with ferroxidases for iron loading onto plasma transferrin. Ferroportin levels are regulated by hepcidin, which is a central regulatory molecule of systemic iron homeostasis.

Ferroportin, the Cellular Iron Exporter

One of the most significant discoveries of the last decade has undoubtedly been the recognition of the key role of ferroportin, the only known cellular iron exporter, in systemic iron homeostasis. Ferroportin, the product of the SLC40A1 gene, was independently identified by three laboratories in 2000 and shown to play a key role in the export of iron from a number of different cell types (McKie et al., 2000; Donovan et al., 2000; Abboud and Haile, 2000). Ferroportin is a transmembrane protein with 10 potential membrane-spanning regions. In polarized epithelial cells, it localizes to the basolateral membrane. It is expressed in many cells that play a critical role in mammalian iron metabolism, including placental syncytiotrophoblasts, duodenal enterocytes, hepatocytes, and reticuloendothelial macrophages. Targeted disruption of the mouse SLC40A1 gene clearly established the unique and nonredundant functions of ferroportin in iron release from cells (Donovan et al., 2005). As illustrated in Figure 36-8 (lower center), ferroportin transports Fe²⁺ out of the cell, in concert with the ferroxidases (hephaestin in enterocytes and ceruloplasmin in other cell types). The ferroxidases facilitate the extraction of iron from the ferroportin channel and the subsequent loading of Fe³⁺ onto apotransferrin (De Domenico et al., 2007). Ferroportin expression is regulated by hepcidin, which triggers its degradation. Ferroportin mRNA translation is regulated by the IrRE/IrRP system in macrophages but not in erythroblasts.

The presence of excess free heme in cells is highly toxic, and there is some evidence that heme efflux pathways exist, notably involving the feline leukemia virus, subgroup C receptor.

Ferritin Transport

Although the physiological significance of ferritin transport is unknown, some proteins have been suggested to serve a role in ferritin uptake. Ferritin can enter cells via the Scara5 (scavenger receptor class A, member 5) and TIM-2 (T-cell immunoglobulin and mucin domain containing 2) receptors (see Figure 36-8, lower left). Mechanisms of ferritin efflux are unidentified.

Intracellular Iron Trafficking, Utilization, and Storage

Iron uptake systems feed into the poorly characterized intracellular transit pool, usually referred to as the labile iron pool (see Figure 36-8), which has been described as “somewhat like the Loch Ness monster—only to disappear from view before its presence, or indeed its nature, can be confirmed” (Crichton, 1984). The labile iron pool represents iron in transit between storage, transport, functional, or recycled forms. Although the exact chemical nature of labile iron remains uncertain, it is thought to represent approximately micromolar concentrations of iron (Breuer et al., 2008). This pool reflects the readily available iron within the cell and is believed to be the key regulator of internal and external iron exchange.

The labile iron pool can be utilized (see Figure 36-8) for direct incorporation into iron proteins within the cytosol, or iron can be transported into the mitochondria by the inner mitochondrial membrane protein mitoferrin (see Figure 36-8, right of center). The majority of iron for cellular utilization is channeled through mitochondria-localized pathways that are involved in the insertion of iron into heme and Fe-S cluster prosthetic groups. Mitochondria represent the major subcellular site of iron utilization, so it is not surprising that they play a central role in the control of cellular iron metabolism (Sheftel and Lill, 2009). However, iron management within mitochondria remains poorly understood.