New insights into the regulation of iron homeostasis


  • Department of Medicine III, Medical University of Vienna, Vienna, Austria (R. Deicher, W. H. Hörl).

Walter H. Hörl, MD, PhD, FRCP, Division of Nephrology and Dialysis, Department of Medicine III, Währinger Gürtel 18-20, A-1090 Vienna, Austria. Tel.: +43–1−404004390; fax: +43–1−404004392; e-mail:


Hepcidin evolves as a potent hepatocyte-derived regulator of the body's iron distribution piloting the flow of iron via, and directly binding, to the cellular iron exporter ferroportin. The hepcidin-ferroportin axis dominates the iron egress from all cellular compartments that are critical to iron homeostasis, namely placental syncytiotrophoblasts, duodenal enterocytes, hepatocytes and macrophages of the reticuloendothelial system. The gene that encodes hepcidin expression (HAMP) is subject to regulation by proinflammatory cytokines, such as IL-6 and IL-1; excessive hepcidin production explains the relative deficiency of iron during inflammatory states, eventually resulting in the anaemia of inflammation. The haemochromatosis genes HFE (the human leukocyte antigen-related gene), TfR2 (the transferrin receptor-2 gene) and HJV (the haemojuvelin gene) potentially facilitate the transcription of HAMP. Disruption of each of the four genes leads to a diminished hepatic release of hepcidin consistent with both a dominant role of hepcidin in hereditary haemochromatosis and an upstream regulatory role of HFE, TfR2 and HJV on HAMP expression. The engineered generation of hepcidin agonists, mimetics or antagonists could largely broaden current therapeutic strategies to redirect the flow of iron.


For decades, the story of iron has been a tale of absorption, storage and excretion. Until recently, it appeared appropriate to approach the body's iron balance through measurements of plasma ferritin. A concentration of 1 µg L−1 was believed to equal an amount of 120 µg of storage iron per kg of body weight [1]. However, plasma ferritin, virtually free of iron, is secreted by reticuloendothelial cells in response to the highly variable intracellular iron content and the plasma concentration decreases disproportionately to erythropoietic stimuli [2]. Three different regulators, namely dietary, stores and the erythropoietic regulator were hypothesized some 50 years ago to govern intestinal iron absorption [3] but, to date, a more complex regulatory network which governs iron traffic rather than deposition has emerged. Each day approximately 20 mg of iron transits to the erythron via transferrin, macrophages recycle approximately the same amount of iron from senescent erythrocytes and 1–2 mg of iron traverse the enterocytes by way of regulated absorption. These figures highlight the key players of iron trafficking, namely duodenal enterocytes, reticuloendothelial macrophages, hepatocytes, which store the bulk of the body's iron load, and maturing erythroid progenitors. Together, they collaboratively pilot the flow of iron [4–6].

A number of recent reports point to hepcidin as a major, evolutionary conserved denominator of iron distribution (Fig. 1) [7], although the defensive-like structure of this peptide rather suggests a primary function as a mediator of innate immunity. Hepcidin is primarily expressed by the liver in response to acute-phase reactions, any further expression depends on the degree of hepatic iron storage, and hypoxia and/or anaemia strongly down-regulate hepatic hepcidin release. Each of these three pathways is subjected to a hitherto only partially understood control. Furthermore, extrahepatic hepcidin gene expression has been documented [8,9] and it is conceivable that knowledge of the regulatory network that guides the flow of iron will quickly expand.

Figure 1.

Feedback between hepatocytes and enterocytes in the regulation of dietary iron absorption. Modified according to Fleming RE, Curr Opin Gastroenterol 2005; 21: 201–206. Dcytb, enterocyte ferrireductase; DMT1, divalent metal transporter-1; Fe, iron; HAMP, the hepcidin gene; HFE, the HFE protein; HJV, haemojuvelin; TfR1 and 2, transferrin receptor-1 and -2.

Ferroportin: the iron exporter

In the duodenum and proximal jejunum, the nonheme dietary Fe3+ is reduced to Fe2+ by the cytochrome b-like ferrireductase Dcytb. The Fe2+ is gathered from the lumen of the intestine and crosses the apical enterocyte brush border membrane through the divalent metal transporter-1 (DMT1) [10,11] (Fig. 1). The expression of both Dcytb and DMT1 is strongly affected by the iron concentration within the enterocyte [12,13]. Mutations in the DMT1 gene result in inhibition of intestinal iron absorption and causes severe iron deficiency and microcytic anaemia [10,14]. The absorbed iron may be incorporated into intracellular ferritin and is lost when the enterocyte is ultimately sloughed at the villous tip [4]. The subsequent movement of iron from the enterocyte into the bloodstream is mediated by the iron exporter ferroportin (iron-regulated transporter-1) [15–18] (Fig. 1).

Ferroportin is located along the entire basolateral membrane of enterocytes (Fig. 1[16,17,19]). The ferroxidase hephaestin, a ceruloplasmin-like protein, promotes the conversion of Fe2+ to Fe3+ and facilitates the release of iron from enterocytes [20]. Ferroportin is also located in high levels in tissue macrophages in the liver (Kupffer cells), spleen and bone marrow, predominantly in the intracellular vesicular compartment [19]. This protein also serves as an iron exporter in circulating phagocytic cells that recycle iron from senescent erythrocytes [21,22]. Greater than 60% of total iron is present in erythrocytes. Thus, efficient heme iron recycling is critical in iron homeostasis [23]. Following erythrophagocytosis of senescent red blood cells, haemoglobin is degraded and iron is released from heme through the action of heme oxygenase [24,25]. The rapid increase in HO-1, the inducible form of heme oxygenase, as well as a rapid increase of ferroportin mRNA and protein indicate that iron is rapidly extracted from heme and exported by ferroportin after erythrophagocytosis [26]. Iron homeostasis is maintained by a co-ordinate expression of ferroportin in both intestinal and phagocytic cells [19]. Ferroportin is required for normal iron export both from enterocytes and from macrophages [27]. It is the sole iron exporter identified in mammals [15–17]. Inactivation of the ferroportin gene SLC40A1, under experimental conditions, results in iron accumulation in enterocytes, Kupffer cells and splenic macrophages [18]. Ferroportin may also play a role in iron detoxification in airway epithelial cells in the lung. However, it localizes subcellularly to the apical membrane of these cells. Iron uptake and release occur across the apical but not basolateral membrane of the airway epithelial cells [28].

Iron egress from the enterocyte at the basolateral surface requires oxidation of ferrous iron to the ferric state by hephaestin [20]. Hephaestin localizes predominantly to apical intracellular membranes, as well as the basolateral membrane of villous enterocytes, suggesting either an intracellular ferroxidase activity of hephaestin or some other function of the molecule to mobilize Fe3+ to serum transferrin [29]. The exact mechanism of how iron reaches transferrin is not fully understood. It is conceivable that hepaestin oxidizes iron in the near vicinity to apotransferrin and the latter may enter cytosolic compartments to acquire iron [30]. Apotransferrin, mono- and diferric transferrin all circulate in human serum; the distribution of iron between the three isoforms depends upon the transferrin saturation with an increasing amount of monoferric transferrin at lower degrees of transferrin saturation. The ratio of mono- to diferric transferrin has been proposed to somehow alter the possible impact of hepatic surface HFE protein on hepatocyte hepcidin release (see below, [31]).

Hepcidin: the iron allocator

Hepcidin controls the whole body iron content. Circulating levels of hepcidin negatively regulate intestinal iron absorption and release of iron from enterocytes and macrophages [32,33]. Hepcidin is a small 25-amino-acid peptide with antimicrobial properties. It is highly inducible by bacterial challenge, structurally resembles the defensin family [34], and is mainly synthesized by the liver, secreted into the bloodstream and excreted through the kidneys. Hepcidin can be isolated from human urine [35] and plasma ultrafiltrate [36]. The hepcidin gene encodes an 84-amino-acid pre-propeptide which is translated in the liver, but is also detectable in human serum [37]. The mature bio-active 25-amino-acid form is found in plasma and urine [35,36]. Hepcidin acts as the principal iron-regulatory hormone to maintain iron homeostasis [38,39]. Hepatic hepcidin expression is decreased in iron deficiency [38,40,41] and during stimulated erythropoiesis [40], allowing intestinal iron absorption and the release of iron from macrophages under these conditions. In contrast, hepatic hepcidin expression is increased in iron overload [38,42,43] and during inflammation [40,44]. Under these conditions, hepcidin serves to inhibit the absorption of iron from the bowel and the release of iron from macrophages. The increase in hepcidin is associated with hypoferremia [40,41,44,45]. Hepcidin may play an important role in defence against infection by depriving micro-organisms of a ready source of iron [46].

Links between heme biosynthesis and intestinal iron absorption may also include 5-aminolaevulinic acid (ALA). A low-iron diet causes low levels of hepatic nonheme iron, low hepcidin mRNA, reduced urinary ALA excretion and enhanced intestinal iron absorption in mice. Iron-loaded animals have shown markedly increased liver nonheme iron, increased hepcidin mRNA, increased urinary ALA excretion and decreased intestinal iron absorption. Hypoxia caused a decrease of hepatic hepcidin mRNA, reduced urinary ALA excretion and increased intestinal iron absorption. Additionally, ALA may act as a modulator in controlling intestinal iron absorption, as injection of ALA to iron-deficient animals or hypoxic mice reduced their enhanced intestinal iron absorption to normal levels [47].

Hepcidin targets ferroportin

Hepcidin inhibits cellular iron export through binding directly to the iron exporter ferroportin (Fig. 1) and inducing its internalization and degradation in HEK-293 cells [48]. Microdomains positive for ferroportin were detected at the plasma membrane of macrophages [16]. However, a recent study by Delaby et al. shows that ferroportin is expressed in vesicular compartments that can reach the plasma membrane of macrophages [49]. Furthermore, when ferroportin expression was up-regulated, the protein expression was strongly enhanced at the plasma membrane of macrophages, and hepcidin dramatically reduced macrophage ferroportin protein levels. At present it is unclear whether hepcidin exerts any regulatory role on the subcellular distribution of ferroportin within macrophages [49]. Serial deletion of the N-terminal amino acids of intact hepcidin causes a progressive loss of bioactivity of the peptide with almost complete loss when all five N-residues are deleted [50]. Human urine contains two predominant hepcidin forms, comprised of 20 and 25 amino acids each, which differ only by N-terminal truncation [51]. The N-terminal peptides alone do not internalize ferroportin, nor modified hepcidin molecules when the C-terminus is deleted or the disulphide pattern is altered by replacing pairs of cysteines with alanines [50]. These data indicate that the N-terminus within the intact molecules is responsible for the hepcidin activity of the peptide with respect to ferroportin internalization and degradation.

Frazer et al.[52] observed a lag period of approximately 4 days after an erythropoietic stimulus before intestinal iron absorption increased. Administration of an intraperitoneal dose of phenylhydrazine in rats caused rapid haemolysis with a maximal decline in haemoglobin and haematocrit after 3 days and recovery thereafter. Hepatic hepcidin mRNA expression decreased dramatically within a few days, reaching almost undetectable levels by the 4th and 5th days. The reduction in hepcidin expression was accompanied by a transient increase in the expression of enterocyte DMT1, Dcytb and ferroportin, and also by prominent hepatic iron deposits, primarily in Kupffer cells, from the 4th day [52]. Once the increased need for iron has been recognized, there is a drop in hepatic hepcidin expression, followed rapidly by an increase in intestinal iron absorption. Similar data have been reported after phenylhydrazine-induced haemolysis in mice [53]. Up-regulation of duodenal DMT1, Dcytb, ferroportin mRNA and protein also occurs in dietary iron-deficiency anaemia [54].

Laftah et al. [55] injected synthetic hepcidin into normal, iron-deficient, and iron-overloaded HFE knockout mice. Hepcidin inhibited duodenal iron absorption in all three groups of animals but did not affect the proportion of iron transferred to the circulation. Rivera et al.[56] demonstrated that a single intraperitoneal injection of hepcidin caused a rapid fall of serum iron within 1 h in a dose-dependent manner. Serum iron remained suppressed for more than 48 h after the hepcidin injection. The slow recovery of hypoferremia to normal iron levels, 96 h after the synthetic hepcidin injection, suggests slow resynthesis of membrane ferroportin after internalization and degradation by hepcidin binding. After the injection, radio-labelled hepcidin accumulated in the ferroportin-rich organs, liver, spleen and duodenum, which is consistent with a blockade of an iron export from tissue stores and macrophages by hepcidin. The lower accumulation of radio-labelled hepcidin in the liver, as compared with duodenal enterocytes and macrophages, supports the idea of lower ferroportin expression and/or activity in hepatocytes and may explain preferential iron accumulation in liver cells in iron-storage disease [56]. In a tetracycline-regulated transgenic mouse model, short-term and long-term expression of hepcidin in the liver leads to hypoferremia and iron-limited erythropoiesis [57]. In haemochromatotic mice, induction of chronic hepcidin expression decreases the expression of macrophage ferroportin but not of hepatic ferroportin, resulting in a reduced iron content of hepatocytes and increased iron load of tissue macrophages and duodenal cells [57].

Mice lacking hepcidin develop severe iron overload [58], whereas mice with a particularily high hepcidin expression develop severe iron-deficiency anaemia [59,60]. Mice with hepcidin-producing tumours developed more severe anaemia and iron deficiency compared with mice with control tumours, indicating that hepcidin excess exacerbates tumour-associated anaemia [61]. In hepcidin-deficient mice, ferroportin is strongly up-regulated in both enterocytes and macrophages [62–64]. Patients with hepatic adenomas and autonomous production of hepcidin may develop hypoferremia and severe iron-refractory anaemia. The resection of the adenomas reverses these haematological abnormalities [60].

Iron load, inflammation and anaemia affect the hepcidin-ferroportin axis

Excessive hepcidin production occurs in patients with inflammatory and infectious disorders, resulting in anaemia of inflammation [44,60]. Increased hepcidin expression during inflammation and infection explains sequestration of iron in the macrophages and inhibition of intestinal iron absorption, the two hallmarks of the anaemia of inflammation. Ferroportin expression in reticuloendothelial system (RES) cells of the spleen, liver and bone marrow is down-regulated by inflammation [60]. Thus, iron sequestration in the RES that accompanies inflammation is not only caused by ferroportin internalization and degradation [49] but also by down-regulation of ferroportin [60]. Anaemia of chronic disease is characterized by changes in iron homeostasis, with increased uptake and retention of iron within RES cells resulting in normocytic or microcytic iron-refractory anaemia. In patients with anaemia of chronic disease, the proliferation and differentiation of erythroid precursors are impaired and are linked to the inhibitory effects of proinflammatory cytokines [5]. Hepcidin contributes to anaemia of chronic disease through the effects on iron metabolism and inhibition of erythroid progenitor proliferation and survival [65]. Hepcidin production is stimulated by inflammation caused by the injection of lipopolysaccharide (LPS), Freund adjuvant and/or turpentine [38,40,45,66,67]. Both iron overload and LPS injection stimulate transcription of the gene that encodes hepcidin (HAMP).

However, the mechanisms inducing transcription of HAMP are different between inflammation and iron overload. Reduced transcription of HAMP and an inappropriate low production of hepcidin may explain the iron overload observed in patients with primary iron-overload syndromes, i.e. hereditary haemochromatosis. The various heritable forms of the iron-storage disease are owing, at least in part, to the inability of iron to stimulate the hepatic production of hepcidin appropriately [6,68,69]. In contrast, up-regulation of HAMP during inflammation occurs even in HFE-knockout mice and in animals with mutation in TfR2– the gene of transferrin receptor-2. Disruptions of both genes define different variants of hereditary haemochromatosis. These data indicate that hepcidin production may occur during inflammation in the absence of HFE and/or TfR2 activation. Nemeth et al.[44,45] suggested that the only cytokine that stimulates hepatic hepcidin production is interleukin-6 (IL-6). Infusion of IL-6 caused increased urinary hepcidin excretion and hypoferremia in human volunteers. Up-regulation of HAMP is impaired in IL-6-knockout mice after LPS stimulation [70], but not abolished, suggesting that other proinflammatory cytokines are also able to induce hepcidin production. Recent data from Lee et al. showed that hepcidin transcription was stimulated not only by IL-6 but also by IL-1α and IL-β in hepatocytes from IL-6 and HFE knockout mice, and mice with a hypomorphic transferrin receptor 2 mutation [71]. The maximal response of hepcidin expression occurs 6 h after the LPS injection [72]. Thus, hepcidin transcription by proinflammatory cytokines does not require HFE and/or TfR2 stimulation, suggesting direct up-regulation of HAMP by IL-6, IL-1α and IL-1β[69,71]. IL-1 induces hypoferremia [73] and up-regulation of ferritin [74] and it may play a primary role in the anaemia of chronic inflammation [75], probably by its stimulation of hepcidin [71].

Chronic inflammation might counteract the effect of HFE mutations and may be of benefit for haemochromatosis patients by HFE-independent up-regulation of HAMP and hepcidin production. However, Beutler et al. [76] found no difference in IL-1 and IL-6 levels between haemochromatosis patients with high iron stores as compared with those patients with low iron stores, indicating that phenotypic differences between such homozygotes are not affected by chronic inflammation. It is noteworthy that in patients with hepatitis C hepcidin mRNA expression did not correlate with markers of hepatic inflammation but with hepatic iron concentration and serum ferritin [77]. The hypoferremic effect of a single dose of hepcidin lasted at least 48 h, suggesting relatively infrequent dosing of hepcidin if considered for treatment of patients with heritable haemochromatosis [56]. Based on the blood levels of hepcidin 1 h after the administration of 50 µg intraperitoneally, Rivera et al.[56] estimated that hepcidin exerts its hypoferremic activity at blood concentrations in the 0·1–1 µM range, which is consistent with the in vitro studies of hepcidin–ferroportin interactions [48].

HFE, TfR2, HJV: upstream regulators of hepcidin?

Hereditary haemochromatosis constitutes a syndrome that is characterized by increased intestinal iron absorption, increased serum iron and deposition of excess iron in parenchymal cells. To date, five different genes appear to be involved ([78], see Table 1), with mutations of the human leukocyte antigen-related gene (HFE) accounting for the majority of cases. Non-HFE variants comprise mutations of the transferrin receptor-2 (TfR2), haemojuvelin (HJV), hepcidin (HAMP) and ferroportin (SLC40A1). Although the precise biological functions of HFE, TfR2 and HJV are currently unknown, some findings suggest a direct regulatory role of these three genes on HAMP expression. Serum levels and hepatic hepcidin expression are inappropriately low in patients with HFE-haemochromatosis [79,80], explaining the persistently increased iron absorption despite the iron overload. Hepatic hepcidin expression is decreased in HFE-knockout mice but over-expression of hepcidin may normalize the hepatic iron deposition [81]. Individuals with TfR2-related haemochromatosis and mice with genetic disruption of TfR2 display inappropriately low levels of hepcidin [67,82]. TfR2 mRNA is highly expressed in the liver, in erythroid precursors as well as in erythroleukaemic cells [83]. Its expression is independent of the cellular iron level but regulated by the cell cycle [84]. TfR2 may mediate the uptake of transferrin-bound iron although the affinity of TfR2 for transferrin is approximately 30-fold less in comparison with TfR1 [85]. It is of note that the hepatic expression of TfR2 persists during iron overload [86], suggesting a modulatory role on hepcidin expression rather than a direct contribution to cellular iron uptake. Finally, subjects with juvenile haemochromatosis, owing to HJV mutations, show inappropriately low urinary hepcidin levels [67,68]. As both HFE- and non-HFE-, namely TfR2- and HJV-, associated forms of haemochromatosis are all consistent with a circulatory iron overload and an absent or largely depressed hepcidin release, it has been hypothesized that hepcidin constitutes the common pathogenic denominator of all forms of hereditary haemochromatosis [87]. Mutations of the HFE, TfR2, HJV and HAMP genes seem to result in a failure to up-regulate hepatic hepcidin expression despite increased body iron. It is intriguing to speculate that HFE, TfR2 and HJV independently, and/or collectively, regulate the expression of hepcidin (Fig. 1) [88]. Loss of function of one of these gene products may be compensated by one of the others; HAMP and HJV appear to be of primary importance owing to the clinically more severe forms of the associated haemochromatosis. Importantly, patients carrying both HFE and TfR2 mutations are phenotypically similar to individuals with solitary HAMP or HJV mutations [89]. This observation supports the notion of a differentiated regulation of hepcidin expression by HFE, TfR2 and HJV.

Table 1.  Some features of hereditary haemochromatosis
GeneSynonymGene productPostulated functionSerum hepcidin
HFEHLA-HHFEHepcidin regulatorLow
TfR2TfR2Iron sensor, hepcidin regulatorLow
HJVHFE2HemojuvelinHepcidin regulatorLow
HAMPLEAP1, HEPCHepcidinIron regulatorLow
SLC40A1FPN1, IREG1FerroportinIron exporter, hepcidin receptorElevated

Mutations of the human ferroportin gene have been reported with substantial clinical differences between the reported families [90–104]. Loss of function mutations appear to result in the accumulation of iron within reticulo-endothelial macrophages, mostly Kupffer cells, and the early predominant deposition of iron within Kupffer cells has been described histologically [90,93,98]. Magnetic-resonance imaging confirms iron accumulation both in the liver and in the spleen [100], which is at variance to the above-mentioned classic forms of haemochromatosis with alterations of the hepatic signal intensity only. Other ferroportin mutants may not respond adequately to hepcidin signalling: a hyporesponsive enterocyte ferroportin may result in the continuous uptake and subsequent outflow of iron into parenchymal storage sites [105,106], despite persistently elevated hepcidin levels [85].

Hepcidin: how to measure?

To date, there are no commercial kits available for the measurement of hepcidin in serum specimen. Instead, serum prohepcidin has been measured by the ELISA technique as a surrogate parameter of hepcidin, but validation of this approach is pending. Hepcidin is translated as an 84-amino-acid prepropeptide, cleaved to the 60-amino-acid peptide prohepcidin and further processed to the 20–25-amino-acid hepcidin, depending on the site of N-terminal truncation. The ELISA measurements using antibodies raised against the N-terminus of prohepcidin revealed that, in addition to hepcidin, the prohormone is released into the circulation [37]. Higher than normal prohepcidin serum levels have been described in uraemia [37] and it is conceivable that serum levels are largely governed by the rate of cleavage or the degree of renal clearance. Prohepcidin serum concentrations may vary irrespectively of the inflammatory state, the individual iron load or the degree of erythropoiesis. It remains unclear whether prohepcidin ELISA assays measure full-length prohepcidin or accumulating cleavage products. Quantification of hepcidin in urine may prove to be a valid approach to evaluate hepatic hepcidin expression. Injection of LPS into healthy human volunteers evoked a rapid increase of plasmas IL-6 and urinary hepcidin excretion within 3–4 h, which was followed by an immediate decrease of serum iron [107]. It is noteworthy that serum prohepcidin levels remained unchanged for more than 20 h after the injection of LPS, although hypoferremia persisted throughout the observation period [107].


The pathophysiology of two disorders of iron homeostasis, namely primary iron overload syndromes and the anaemia of chronic inflammation, has remained enigmatic until the recent discovery of hepcidin. The syndromes of HFE-, TfR2- and HJV-associated haemochromatosis appear to be largely attributable to a circulatory iron overload which results from an absent or largely depressed hepatic hepcidin release. On the other hand, up-regulation and release of hepcidin, as it occurs during acute-phase reactions or chronic inflammatory states, mediates internalization of ferroportin which in turn causes functional or absolute iron deficiency. Hepcidin may thus be regarded as a hepatocyte-derived peptide hormone which evolved to protect the organism from the potential detrimental oxidizing effects of iron overload, to establish the specific needs of rapidly growing tissues like the erythron and to fight the battle against invading iron-consuming pathogens. Notably, stimulation of Toll-like receptors on macrophages and neutrophils led to the expression of hepcidin; tissue staining of hepcidin was demonstrated in colocalization with recruited neutrophils in a murine model of subcutaneous infection with group-A streptococci. Furthermore, bacterial stimulation of macrophages triggered a Toll-like receptor-4-dependent reduction in the iron exporter ferroportin [108]. These observations point to the defensin-like structure of hepcidin and suggest both a phylogenetic conservation of hepcidin as a potent effector molecule of innate immunity and an elaborate gain of function resulting in hepcidin's key role as a central regulator of iron homeostasis.