Hereditary Hemochromatosis and Transferrin Receptor 2

Abstract

Background

Multicellular organisms regulate the uptake of calories, trace elements, and other nutrients by complex feedback mechanisms. In the case of iron, the body senses internal iron stores, iron requirements for hematopoiesis, and inflammatory status, and regulates iron uptake by modulating the uptake of dietary iron from the intestine. Both the liver and the intestine participate in the coordination of iron uptake and distribution in the body. The liver senses inflammatory signals and iron status of the organism and secretes a peptide hormone, hepcidin. Under high iron or inflammatory conditions hepcidin levels increase. Hepcidin binds to the iron transport protein, ferroportin (FPN), promoting FPN internalization and degradation. Decreased FPN levels reduce iron efflux out of intestinal epithelial cells and macrophages into the circulation. Derangements in iron metabolism result in either the abnormal accumulation of iron in the body, or in anemias. The identification of the mutations that cause the iron overload disease, hereditary hemochromatosis (HH), or iron-refractory iron-deficiencey anemia has revealed many of the proteins used to regulate iron uptake.

Scope of the review

In this review we discuss recent data concerning the regulation of iron homeostasis in the body by the liver and how transferrin receptor 2 (TfR2) affects this process.

Major conclusions

TfR2 plays a key role in regulating iron homeostasis in the body.

General significance

The regulation of iron homeostasis is important. One third of the people in the world are anemic. HH is the most common inherited disease in people of Northern European origin and can lead to severe health complications if left untreated.

1. Iron homeostasis in the body

Iron is an essential nutrient required for a variety of biochemical processes such as respiration, metabolism, and DNA synthesis. Cells and organisms possess carefully regulated but poorly understood mechanisms for iron absorption and metabolism. Iron homeostasis in the body appears to be regulated primarily at the level of iron uptake through the intestine. Intestinal iron absorption is controlled by at least two kinds of regulators, those sensing iron stores in the body and those sensing the disparity between erythropoiesis and iron supply [1–6]. When the body is iron replete and under steady-state erythropoiesis, the intake of iron through the intestine is equal to the loss of iron through the urine, bleeding, and desquamation of intestinal tissue.

1.a. Hepcidin, a key regulator of iron homeostasis

Hepcidin, a peptide synthesized by the liver and encoded by the gene, HAMP, plays a major role in iron homeostasis [7–11]. The mature form of the peptide is 20–25 amino acids with four disulfide bonds. The finding that Hamp knockout mice suffer from massive iron overload initially raised the possibility that hepcidin might be an iron stores regulator involved in communication of liver iron status to the intestine [9]. In patients with hepatic adenomas expressing high levels of hepcidin mRNA, the anemia was resolved upon the removal of the tumor or liver transplantation [12]. Similarly, mice that are engineered to overproduce hepcidin or injected with hepcidin are severely anemic [13–15]. These results indicate the importance of hepcidin in the regulation of iron uptake by the intestine.

The mechanism by which hepcidin regulates iron levels in the body is through its binding to ferroportin (FPN), the only known transporter for the efflux of iron out of cells and ultimately into the blood. The binding of hepcidin to FPN results in the internalization and degradation of both FPN and hepcidin [16]. Thus iron export from the intestine is decreased when the liver is stimulated to synthesize hepcidin.

1.b. Regulation of hepcidin mRNA

The discovery that hepcidin is a key regulator of iron homeostasis led to an intensive investigation of the control of hepcidin synthesis. Most of the studies concentrated on the induction or the suppression of hepcidin mRNA because mRNA levels are easy to measure and until recently, no antibodies or assays were sensitive enough to measure endogenous hepcidin in serum. Studies on the transcription of hepcidin show that it is regulated by both inflammation and iron and involves at least two or more interrelated signaling pathways (Figure 1).

Figure 1. Model for the stimulation of hepcidin expression in hepatocytes by inflammation and by iron levels in the body.

Elevation of interleukin 6 (IL6) during inflammation stimulates hepcidin transcription through binding to the IL6-receptor (IL6R)-gp130 complex, which stimulates signaling through the phosphorylation of JAK leading to the phosphorylation of STAT3. The phosphorylated STAT3 dimer binds to two identified BMP responsive elements in the promoter region of hepcidin stimulating transcription. Elevation of iron levels in the body leads to the upregulation of BMP6 and Tf-saturation. BMP6 signals through binding to hemojuvelin (HJV), BMP receptors and neogenin (Neo) and signals through the SMAD4/SMAD1/5/8 pathway to stimulate hepcidin expression. Two BMP responsive elements in the promoter region of hepcidin have been identified. Less is known about the Tf-TfR2-HFE signaling pathway (see text). Mutations in HFE result in a blunted pSMAD response. The p38 MAPK-ERK1/2 signaling pathway is proposed to mediate this response.

Inflammatory processes increase interleukin 6 (IL6) and interleukin 1 (IL1), which activate HAMP expression through a JAK (just another kinase)-STAT3 (signal transducer and activator of transcription)-mediated pathway [17–21]. The promoter region of hepcidin contains a STAT3 binding site [21]. Thus, stimulation of hepcidin expression through the inflammatory pathway, which results in decreased iron absorption and sequestration of iron in macrophages, provides an explanation of the anemia of chronic disease [12, 22].

Iron sensing by the liver is mediated through at least the bone morphogenetic protein (BMP) signaling pathway involving BMPs, hemojuvelin (a co-receptor), matriptase-2 (a serine protease), BMP receptors, and the SMADs. BMP2, BMP4, and BMP6 are predominantly synthesized in the liver by endothelial and stellate cells, and to a lesser extent by hepatocytes [23, 24]. These BMPs bind to hemojuvelin and a subset of BMP type 1 receptors (ALK-2 and ALK-3) and BMP type 2 receptor (ActRIIA) [25]. Lack of functional hemojuvelin results in low levels of hepcidin mRNA, and severe iron-overload in patients and mice [26, 27]. Hemojuvelin, in turn, can be degraded by matriptase-2 cleavage [28]. Mutations in matriptase-2 fail to turn off hepcidin signaling, resulting in iron-refractory iron-deficiency anemia (IRIDA) [29–31]. Matriptase-2 mutations in a subset of patients with IRIDA have been identified [32–37]. Activation of the BMP receptors results in the phosphorylation of SMAD 1/5/8, which in conjunction with SMAD4 traffic to nucleus where they bind to BMP regulatory elements to activate hepcidin transcription [38–40]. In addition, neogenin binds to hemojuvelin and activates SMAD signaling, although this observation is controversial [25, 41–44]. Mice fed high iron diets show increased SMAD1/5/8 phosphorylation [45, 46] implying that this pathway is sensitive to iron. Of the BMPs synthesized by the liver, BMP6 mRNA increases with iron loading, providing an iron-mediated regulation of hepcidin expression [47].

High iron diets also increase the iron-loading of Tf (Tf-saturation). Since lack of functional transferrin receptor 2 (TfR2) or the HH protein, HFE, also results in lower hepcidin mRNA levels and iron overload, these molecules also are involved in iron sensing. Signaling TfR2 is proposed to be through the activation of p38 MAPK/ERK (mitogen-activated protein kinase/extracellular-signal regulated kinase) [48, 49], but this is controversial. A study using an in vivo model of iron-overload failed to detect activation of the p38 MAPK/ERK pathway [50].

An intact SMAD signaling pathway appears to be necessary for Tf, STAT3 and p38 MAPK/ERK signaling [40]. Recent evidence indicates that an elevation in BMP6 is the result of chronic but not acute iron-loading and an elevation in Tf-saturation results from acute iron-loading [50, 51]. Increases in either BMP6 or Tf-saturation increase hepcidin mRNA. The inflammatory and p38 MAPK/ERK pathways are dependent on an intact SMAD signaling pathway. In the absence of SMAD4, high iron levels in the liver fail to induce hepcidin mRNA but do increase BMP6 levels [47]. Upregulation of hepcidin by inflammation requires intact BMP-regulatory elements in the hepcidin promoter [52, 53]. How the inflammatory, SMAD and p38 MAPK/ERK pathways are integrated remains to be resolved.

1.c. Erythropoietic regulators of iron uptake

The regulators for communicating the erythropoietic state of the individual are only beginning to be understood. Early physiological studies demonstrated that a soluble factor(s) in the blood is involved. Initially, iron-loaded Tf, ferritin, serum TfR1 generated from the proteolytic cleavage of the full-length transmembrane TfR1 were proposed as candidate factors [54–59]. Tf, ferritin, and TfR1 are found in serum and fluctuate with iron status of the individual. The amount of serum ferritin increases in iron-overloaded individuals. Although under most situations the concentration of Tf remains constant, Tf-saturation increases with iron overload. In addition to liver biopsy for iron-staining, ferritin levels and Tf-saturation are used clinically to evaluate iron stores in the body [60]. There are two notable exceptions to the correlation of these proteins with iron stores within an organism. Mice lacking or having very low levels of Tf suffer from iron overload [61], implying a possible role of Tf in the sensing of iron stores. This finding fits with the hypothesis that intestinal iron absorption is regulated according to Tf-saturation levels. In the absence of Tf, dietary iron is transported into the blood without regulation resulting in the severe iron-overload seen in the hypotransferrinemic mouse. The second exception is hyperferritinemic individuals. A mutation in the stem-loop structure of L-ferritin results in unregulated ferritin synthesis, leading to high serum ferritin levels and cataracts [62, 63]. Importantly, these people do not suffer from iron overload [63]). Thus, serum ferritin levels are not likely to be a key part of the sensing mechanism for iron absorption. Serum TfR1 fluctuates with erythropoietic activity and iron status of the organism [54]. Approximately 80% of serum TfR1 is generated from the maturation of erythroid cells [64]. One argument against a role for serum TfR1 as an erythroid regulatory factor is that it is generated after cells no longer need iron for hemoglobin biosynthesis.

Anemia also increases iron uptake by the intestine. Factors in the plasma of blood beside hepcidin are proposed to either act on the liver to suppress hepcidin transcription, thus allowing for greater iron uptake by the intestine by FPN or on the intestine to upregulate FPN. Several candidate molecules are proposed to regulate iron homeostasis during increased erythropoiesis. Thalassemias, which are characterized by defects in globin synthesis are used as a model to identify erythroid factors involved in the regulation of iron uptake. In this disorder, hepcidin levels are low [65, 66], which could account for the iron overload in thalassemic patients. Tanno and colleagues identified two genes with elevated mRNAs in murine models of thalassemias [67]. They include GDF15 (growth differentiation factor-15) and TWSG1 (twisted in gastrulation 1). GDF15 fulfills the criteria of a soluble factor, which regulates iron homeostasis. It is a member of the TNF-β family of proteins and in high concentrations can lower hepcidin mRNA. Further studies indicate that although GDF15 is elevated in a number of dyserythropoietic diseases, it is not elevated in normal blood donors who become anemic through the loss of blood [68]. TWSG1, also a member of the TGF-β1 family of cytokines, is capable of inhibiting hepcidin transcription and is elevated in thalassemic mice, murine hepatocytes and human cell lines [69]. The role of TWSG1 in normal iron homeostasis remains to be clarified.

1.d. Other roles of the liver in iron homeostasis

The liver plays a major role in iron homeostasis in the body in addition to secreting hepcidin. The Kupffer cells of the liver take up senescent red blood cells and hemoglobin through the hemoglobin-haptoglobin receptor (CD163), salvage the iron released from hemoglobin and secrete the iron as Fe²⁺ via FPN. Ceruloplasmin (Cp), a copper containing ferrioxidase, facilitates the efflux of iron from cells as well as the loading of iron into Tf [70–73]. Hepatocytes take up Tf through TfR1 and most likely through TfR2 [74]. They also take up other forms of non-Tf-bound iron, including heme via the heme-hemopexin receptor [75], and are capable of storing large quantities of iron in ferritin and hemosiderin, a breakdown product of ferritin. In addition to taking up and storing iron, hepatocytes synthesize Tf and Cp. Thus the liver and, in particular, the hepatocyte is thought to sense and reflect the bodily iron stores [76].

2. Hereditary hemochromatosis

Lack of proper communication regarding the amount of iron required by the body and intestinal iron uptake results in derangements in iron metabolism. In particular, hereditary hemochromatosis (HH) is a disease of iron overload leading to iron accumulation in specific organs including the liver, heart, pancreas, and pituitary. Early studies demonstrated that intestinal iron absorption by HH patients is abnormally elevated [77]. Excess iron in affected tissues catalyzes oxidative damage, resulting in cirrhosis, hepatoma, cardiomyopathy, diabetes, hypogonadotropic hypogonadism, and arthritis [60].

The most prevalent form of HH results from a point mutations in HFE. HH causing mutations in HFE result in autosomal recessive form of HH type 1. The HFE (C282Y) is the most common inherited disease in Caucasians (reviewed in [78]). The carrier frequency is about 1 in 9 in people of Northern European origin in the U.S., and the disease affects approximately 1 in 400–10,000 individuals. In a North American screen of ~10,000 participants the prevalence of HFE(C282Y) homozygotes in non-Hispanic whites was 0.44% with other ethnic groups lower [79]. While carrying this mutation may be an advantage with an iron-poor diet, it presents a distinct disadvantage for an iron-sufficient diet. The frequency and importance of HH has only recently been appreciated because it was difficult to diagnose prior to PCR technology and identification of the point mutation in HFE responsible for the most common form of HH.

Other forms of HH arise from mutations in hepcidin, hemojuvelin, TfR2, and FPN [9, 80–83]. Mutations in hepcidin and hemojuvelin result in juvenile hemochromatosis, the most severe form of HH (HH type 2A and 2B respectively). The pathological symptoms of iron-overload are apparent as early as the second and third decade of life. In contrast mutations in HFE (HH type 1) results in a milder form of iron overload usually apparent in the fifth to sixth decade of life. Mutations in TfR2 (HH type 3) are rare and result in mild to intermediate iron-overload. HH types 1–3 result in primary iron loading in the hepatocytes, implying a key role for this cell type in iron homeostasis. Mutations in FPN (HH type 4A & 4B) lead to autosomal dominant diseases with different phenotypes depending on the particular mutation (reviewed in [84, 85]. Mutations in FPN that fail to fold properly or transport iron are dominant-negative mutations. The HH type 4A mutations result in low Tf-saturation and iron accumulation in liver macrophages rather than hepatocytes. The flatiron mouse mutation, arising from an ethylnitrosurea-induced mutagenesis screen, is a mouse model of HH type 4A [86]. Dominant positive mutations in FPN, include mutations that can still transport iron but fail to bind hepcidin or that are not downregulated in response to binding hepcidin (HH type 4B). Patients with HH type 4B have high Tf-saturation and high levels of iron in hepatocytes similar to HH types 1–3.

3. Identification of transferrin receptor 2

In most tissues, TfR1 is responsible for the majority of cellular iron uptake through interaction with iron-bound Tf [87]. This homodimeric membrane receptor binds two molecules of iron-loaded Tf with high affinity and is internalized into acidified endosomes. The combination of TfR1 and low pH facilitates the release of iron from Tf [88, 89]. The iron is then transported across the vesicle membrane for utilization within the cell and/or storage. The TfR1-Tf complex recycles back to the cell surface where apo-Tf is released at the higher pH in extracellular space of tissues. Although TfR1-mediated endocytosis is the major pathway for cellular iron uptake, cells can obtain iron through TfR1 independent pathways using diferric Tf or inorganic iron (reviewed by Aisen and colleagues [87]).

The more recently identified TfR2 is a second, distinct TfR and could be responsible for the non-TfR1 mediated uptake of Tf into the liver as reported earlier. TfR2 clearly plays a critical role in iron homeostasis, because mutations in TfR2 result in a rare form of HH [90]. TfR2 can support growth in transfected Chinese hamster ovary (CHO) cells lacking endogenous Tf receptors, given Tf as an iron source [74]. The role of TfR2 in Tf-mediated iron-uptake by the liver has been called into doubt recently [91]. These studies only found a minimal contribution of TfR2-mediated iron uptake in a hepatoma cell line, HUH7 cells. Both people and mice with mutations in TfR2 or lacking TfR2 suffer from iron-overload in the liver [92, 93]. These findings indicate that the uptake of iron into the liver by TfR2 is not the primary function for this receptor and that TfR2 may have additional functions. The expression of both hepcidin and TfR2 in hepatocytes [92, 94, 95] and the observation that TfR2 is regulated by iron-loaded Tf, led to the hypothesis that TfR2 might sense the levels of Tf-saturation in the blood [96]. TfR2 would then activate hepcidin transcription, in addition to its ability to take up Tf (Figure 1).

4. Distinguishing features of transferrin receptor 1 and 2

TfR2, like TfR1, is a type II membrane glycoprotein with a large C-terminal ectodomain and small N-terminal cytoplasmic domain [97, 98]. TfR2 shares 45% amino acid sequence identity with TfR1 in the extracellular region, has two cysteines in the ectodomain proximal to the transmembrane domain that form intersubunit disulfide bonds like TfR1, and has an internalization sequence in its cytoplasmic domain [97, 98].

Clear differences exist in the regulation of the two TfRs. In humans and mice, TfR2 is expressed predominantly in liver, erythroid precursors [99, 100], and also expressed in erythroid cell lines [100], while TfR1 is expressed ubiquitously [93, 98]. Intracellular iron controls TfR1 mRNA stability. In contrast changes in intracellular iron levels do not influence the level of TfR2 mRNA [96–98, 101]. TfR2 expression is controlled at the transcriptional level by the erythroid transcription factor GATA-1 [100]. The transcriptional regulation of TfR2 in the liver has not been characterized extensively. Tfr2 mRNA levels increase dramatically from embryonic day 13 to postnatal day 4 in the developing mouse liver [102]. The transcription factor C/EBP-α is abundant in the liver and induces transcription in Tfr2 promoter reporter assays [102].

The two receptors are functionally different. Although TfR2 can mediate cellular iron-uptake in transfected cells, TfR2 is not sufficient to compensate for the function of TfR1 because mice in which Tfr1 is deleted die as embryos [103]. The lack of ability for TfR2 to compensate for TfR1 could be due to its restricted expression pattern. The affinity of TfR2 for iron-loaded Tf is 27 nM, 27 fold lower than TfR1 [98, 104, 105]. A lower affinity of Tf could make TfR2 more sensitive to Tf-saturations in the blood, although the binding constants measured in vitro are still too high to be sensitive to Tf-saturation. While both receptors have internalization motifs, there are no sequence similarities in their respective cytoplasmic domains. Evidence suggests that the two receptors can internalize by independent mechanisms [106]. Both TfR1 and TfR2 bind to HFE and Tf, but the interacting domains of HFE with TfR2 are different from those with TfR1. The crystal structure of the ectodomains of TfR1 and HFE indicates that TfR1 interacts with α1 and α2 domains of HFE [107]. Extensive site-directed mutagenesis studies indicate that the binding site of HFE on TfR1 substantially overlaps with that of Tf-binding site [108, 109]. In contrast, co-precipitation studies using cell lines transfected with TfR2 and HFE chimeras indicates that TfR2 interacts with α3 domain of HFE and the binding of HFE to TfR2 does not interfere with Tf binding [110]. HFE interacts with the extracellular domain of TfR2 proximal to the transmembrane domain (TfR2 residues 104–250) [110, 111]. The TfR2-HFE complex in the liver regulates hepcidin expression [112, 113].

5. Mutations in transferrin receptor 2 result in hereditary hemochromatosis

The role of TfR2 in a rare form of HH was initially surprising because northern analysis of TfR2 mRNA tissue distribution indicated that TfR2 is predominantly expressed in the liver [97, 98] and early erythroid precursors [99, 100]. One report using antibodies to TfR2 indicates that TfR2 is located extensively throughout the intestine [114] and another report shows TfR2 in the crypt cells of the intestine [115]. Both used immunohistological staining. These reports are inconsistent with the lack of TfR2 mRNA in the intestine [97, 98].

The mutations in TfR2 associated with HH are all recessive. They fall into three categories: amino acid deletions, single amino acid substitutions, and nonsense mutations producing a truncated protein. The most severe mutation results in a transcript encoding the first 63 residues of TfR2. An interesting disease-causing mutation in TfR2 is a single aminoacid substitution of a valine to isoleucine in the cytoplasmic domain (V22I) [116]. Most mutations of the cytoplasmic domain of proteins disrupt signaling or trafficking rather than ligand-binding. The recessive nature of the disease combined with extreme truncation of the resultant protein product is consistent with a loss of function rather than a gain of function. The generation of the Tfr2 knockout mouse model confirms this prediction [92]. Mice with a complete or liver-specific knockout of Tfr2 have no detectable Tfr2 protein in their livers and develop significant liver iron-overload and elevated Tf-saturation [117, 118]. The iron overload in the hepatocyte-specific knockout mouse and the Tfr2 knockout mice are similar to the phenotype of the most common mutation (245X) in TfR2-associated HH demonstrating that mutations in TfR2 causing HH are due to lack of functional protein [92, 117]. All result in lower hepcidin expression compared to wild-type mice with equivalent liver-iron loading. Conversely, the hepatocyte-specific expression of Tfr2 in TfR2-deficient mice increases hepcidin expression and decreases hepatic iron levels and Tf-saturation [112]. Collectively, these results demonstrate that lack of expression of TfR2 in hepatocytes accounts for the increased iron absorption and iron accumulation in the liver seen in this form of hemochromatosis.

6. Regulation of transferrin receptor 2 in the liver

Several studies shed some light on the regulation of TfR2 in the liver [96, 101, 104]. Two studies demonstrate stabilization of TfR2 by iron-loaded Tf [96, 101]. Western blot analysis shows that TfR2 increases in a time- and dose- dependent manner after addition of iron-loaded Tf to the culture medium [96]. In cells exposed to iron-loaded Tf, the amount of TfR2 returns to control levels within 8 hours after removal of iron-loaded Tf from the medium indicating a higher turnover rate than TfR1 [96]. However, TfR2 does not increase when non-Tf bound iron (FeNTA) or apo-Tf is added to the medium [96], which suggests that the Tf binding stabilizes TfR2 [96]. The response to iron-loaded Tf appears to be hepatocyte specific. Non-hepatic cell lines which either endogenously express TfR2 such as K562 cells or are transfected with a plasmid coding for TfR2 do not show a similar response [96, 101]. Quantitative real-time PCR analysis shows that TfR2 mRNA levels do not change in cells exposed to iron-loaded Tf [96, 101]. Rather, the increase in TfR2 is due to an increase in the half-life at the protein level in cells exposed to iron-loaded Tf [96]. The binding of Tf to TfR2 and the cytoplasmic domain of TfR2 appear to be largely responsible for the stabilization of TfR2 by Tf [119, 120]. TfR2 levels increased in response to Tf while TfR1 levels were downregulated over 10 fold [96]. These results support a role for TfR2 in monitoring iron levels by sensing changes in the concentration of iron-loaded Tf.

Animal studies are consistent with the stabilization of TfR2 by iron-loaded Tf rather than iron loading of hepatocytes. The Hfe-deficient and hypotransferrinemic mice both have iron-overloaded livers [101]. Rats fed an iron-deficient diet have lower TfR2 levels than rats fed a high iron diet [101]. Tfr2 is higher in Hfe knockout mice compared to wild-type littermates [101]. Conversely, hypotransferrinemic mice have extremely low levels of Tf and reduced Tfr2 supporting the role of iron-loaded Tf in the regulation of Tfr2 [101]. The regulation of Tfr2 in these mice is consistent with tissue culture studies showing that intracellular iron-levels do not correlate with TfR2 levels. Rather, regulation of TfR2 does correlate with the levels of Tf-saturation in the serum or cell culture medium. Because both the tissue culture and the animal studies show that TfR2 levels are sensitive to Tf over a physiological range of Tf-saturation in the blood, TfR2 is proposed to be part of the iron-sensing machinery in the liver for bodily iron levels as reflected by Tf-saturation. Consistent with this hypothesis, mice with a disease-causing mutation in Tfr2 have decreased hepcidin mRNA levels [104], and hepcidin mRNA in Tfr2 mutant mice is decreased relative to their liver iron levels.

7. The role of transferrin receptor 2 in iron homeostasis

The lack of functional TfR2 or HFE results in HH. TfR2 and HFE interact [110, 111]. Further studies demonstrate that TfR2 forms a complex with iron-loaded Tf and HFE to regulate hepcidin expression in hepatoma cell lines stably expressing HFE, and in primary hepatocytes [113]. In mice, both Tfr2 and Hfe are required for the proper regulation of hepcidin expression in vivo. Hfe but not Tfr2 is limiting in the formation of the Hfe/Tfr2 complex to regulate hepcidin expression [112]. In addition, one recent study shows that TfR2 is critical for iron-sensing and hepcidin regulation in human primary hepatocytes [121], and another study using patients with HFE and TfR2 hemochromatosis indicates that TfR2 plays a prominent and HFE plays a contributory role in the regulation of hepcidin in response to a dose of oral iron [122]. These observations suggest that HFE and TfR2 are part of an upstream pathway that regulates hepcidin expression, and supports a model proposed by Schmidt et al. [123] in which TfR1 serves to sequester HFE from interaction with TfR2, thereby reducing signaling for hepcidin expression. When the Tf-saturation increases, iron-loaded Tf competes with HFE to bind to TfR1, which releases HFE from binding to TfR1 and allows HFE to bind to TfR2, thereby Tf-TfR2-HFE complex forms and initiates regulation of hepcidin expression (Figure 2).

Figure 2. Model for the function of HFE and TfR2 in hepcidin signaling.

Upon Tf binding to transferrin receptor 1 (TfR1), HFE is released and binds to transferrin receptor 2 (TfR2). The HFE-TfR2 complex stimulates SMAD signaling either directly by binding to the BMPR receptor complex or indirectly, through stimulating a MAPK-ERK complex that stimulates SMAD phosphorylation.

The mechanism by which the TfR2/HFE complex regulates hepcidin expression remains to be determined. Studies in humans with mutations in HFE show that the regulation of hepcidin is blunted. Hepcidin is still modulated by iron stores both in patients and in Hfe knockout mice [124, 125]. Mice lacking functional Hfe with overexpression of Tfr2 still have lower hepcidin levels indicating that elevated TfR2 levels cannot compensate for lack of HFE [112]. Tfr2 knockout mice have a more severe iron overload phenotype than Hfe knockout mice [126] indicating that TfR2 may have additional functions in the regulation of iron homeostasis. Mice with combined deletions of both Tfr2 and Hfe display a more severe iron overload phenotype compared to Hfe knockout mice or to Tfr2 knockout mice [126]. These results are consistent with a more severe form of iron overload in a report of a patient with mutations in both TFR2 and HFE [127].

The BMP/SMAD signaling pathway plays an important role in the regulation of hepcidin expression [45, 128]. BMP6 is an endogenous regulator of hepcidin expression [129, 130]. HFE is not necessary for regulation of BMP6 in response to iron, but HFE-deficiency triggers iron overload by decreasing hepatic BMP/SMAD signaling [131], and impairing downstream signals of BMP6 triggered by iron [132]. Both Tfr2-deficient and Hfe-deficient mice have similar pSMAD levels as the wild type mice, but Tfr2 and Hfe double knockout mice have lower phospho-SMAD levels, and all of them have lower phospho-ERK1/2 levels [126], indicating that the p38MAPK/ERK1/2 and SMAD signaling pathways could be involved in the regulation of hepcidin by TfR2 and HFE. Further in vitro studies support this hypothesis. Ramey and colleagues showed that iron-loaded Tf, in cooperation with circulating serum factors, stimulates hepcidin expression through both the p38MAPK/ERK1/2 and BMP/hemojuvelin (HJV) pathways [133]. Poli and colleagues show that iron-loaded Tf in a complex with TfR2 and HFE induce furin expression and the p38MAPK/ERK signaling pathway. Furin participates maturation of hepcidin and thereby regulates hepcidin expression [49]. Taken together, these studies indicate that TfR2 may regulate hepcidin expression through the p38MAPK/ERK and BMP/SMAD signaling pathways. No p38MAPK/ERK activation was detected in mouse models of acute and chronic iron-loading leaving the role of p38MAPK/ERK signaling controversial [50].

8. The role of transferrin receptor 2 in erythropoiesis

TfR2 is predominantly expressed in hepatocytes, but is also expressed in human erythroid progenitors, erythroblasts [99, 134], UT7 (erythroleukemia) cells [135], and K562 (erythro-myeloblastoid leukemia) cells [96, 98]. The role of TfR2 in immature erythroid cells is largely unexplored because no significant changes in hematopoietic parameters of peripheral blood were originally detected between wild type and Tfr2 deficient mice [93]. A genome-wide association study shows that a TfR2 polymorphism is associated with red blood cell (RBC) hematocrit and mean corpuscular volume [136]. In addition, a recent study demonstrates that TfR2 binds to erythropoietin receptor (EpoR) and is required for efficient erythropoiesis [135]. TfR2 and EpoR are co-expressed during the differentiation of erythroid progenitors. EpoR associates with TfR2 in the endoplasmic reticulum and TfR2 is required for the efficient transport of EpoR to the cell surface [135]. Erythroid progenitors from Tfr2-deficient mice are less sensitive to erythropoietin (Epo), which are presumably compensated by increased circulating Epo levels in their serum. Epo/EpoR is required for survival and proliferation of erythroid progenitors and their terminal differentiation [137–139]. These studies underscore the involvement of TfR2 in erythropoiesis.

Erythropoiesis is the most iron-consuming process in mammals. TfR2 is one of the key regulators of iron homeostasis because it modulates hepcidin expression in the liver. Under physiological conditions, about 25 mg of iron/day (~70% of total iron) are used for hemoglobin biosynthesis. The recycling of hemoglobin iron from senescent erythrocytes by liver and spleen macrophages serves as the major iron source for erythropoiesis. Interestingly, hepcidin production is suppressed by increased erythropoiesis in the bone marrow, which allows more hemoglobin synthesis by increasing iron uptake from the intestine and iron release from the macrophage. The crosstalk between iron homeostasis and erythropoiesis, the role of Tfr2 in erythropoiesis in vivo and the mechanism by which TfR2 regulates erythropoiesis remain to be determined.

9.Future directions

In spite of the rapid progress made towards the understanding of iron homeostasis in the past 10 years key issues remain to be resolved.

1. How does the TfR2-HFE complex stimulate hepcidin transcription? Does the TfR2-HFE complex directly interact with the BMPR signaling complex or indirectly regulate the levels of pSMAD? Stimulation of the p38MAPK/ERK1/2 pathway is the prime candidate for the downstream signaling by TfR2/HFE complex. How is this accomplished?

2. How does BMP6 respond to iron? BMP6 levels positively correlate with iron levels in the liver. How do liver iron levels control BMP6 expression?

3. How does erythropoiesis regulate iron uptake? Is it having direct or indirect effects on the liver suppressing hepcidin levels?

4. What is the role of TfR2 in regulating erythropoiesis?

5. What pathways can be exploited to treat iron overload and iron refractory anemias? Hepcidin, HFE, TfR2, HJV all appear to be specific targets to regulate iron homeostasis. How can these molecules be targeted or regulated to normalize iron levels in disease states?

Highlights.

• Recent data concerning the regulation of iron homeostasis in the body by the liver is discussed.

• Transferrin receptor 2 plays a distinct role in the regulation of hepcidin.

• Transferrin receptor 2 is controlled both transcriptionally and posttranscriptionally at the level of protein stability.

• Transferrin receptor 2 also plays a role in erythropoiesis.

• The regulation of hepcidin transcription is controlled by both iron through Tf saturation and BMP6 and by inflammation.