60.1.1

The Essentiality of Iron and Overview of Intestinal Iron Homeostasis

Iron is abundant in the Earth’s soils and as such, it is ubiquitously present in life forms on this planet. Most species, including humans, require iron for numerous biological functions, yet iron in excess is toxic; as such, iron levels must be precisely managed. Iron is necessary for several critical physiological functions, including oxygen transport, regulation of cell growth and differentiation, energy production, and control of gene expression. Mammals have developed sophisticated regulatory mechanisms that govern iron absorption, storage, and recycling. Iron associates with many proteins and it mediates important functions, including, electron transfer reactions. Some iron-binding proteins have no known enzymatic function, but many others are enzymes, which contain iron as iron-sulfur clusters, heme moieties, or in other molecular configurations. Iron-binding proteins lacking enzymatic activity include those that transport oxygen (hemoglobin and myoglobin). Some iron-sulfur cluster-containing enzymes are involved in energy (i.e., ATP) production via their role in single-electron transfer reactions (e.g., in the cytochromes of the electron transport chain). Heme-containing proteins also participate in a variety of electron transfer reactions in association with various cofactors (e.g., cytochrome P450 complexes). Other proteins transiently bind iron and facilitate its movement across the plasma and intracellular membranes (e.g., divalent metal-ion transporter 1 [DMT1], ferroportin 1 [FPN1). The nonredundant roles of iron in these and other proteins exemplify the importance of iron in mammalian physiology.

Overall body iron homeostasis is controlled by the peptide hormone hepcidin, which functions to decrease serum iron levels by inhibiting the absorption of dietary iron, and iron release from stores in reticuloendothelial (RE) macrophages of the spleen, liver and bone marrow, and hepatocytes. Hepcidin is induced by high body iron content, infection, and inflammation. Conversely, during iron deficiency and hypoxia and when erythropoietic demand increases, hepcidin production is very low, and additional regulatory mechanisms are invoked to increase intestinal iron absorption. Normally, absorption of dietary iron reflects body iron requirements; however, homeostatic dysregulation of iron absorption occurs in numerous disease states, with pathological consequences. Humans have not evolved an efficient iron excretory system, so the overall control of body iron content is ultimately determined by assimilation of dietary iron by the intestine. Absorption of dietary iron typically matches nonspecific losses, which occur via menstrual blood flows, desquamation of skin cells and exfoliation of GI epithelial cells, and in bile and urine. A normal human will absorb 25–50 g of dietary iron over their lifetime. In the adult males, this represents ~ 1 mg of iron/day, while adult females absorb 1.5–2 mg/day to compensate for iron losses associated with menstruation, pregnancy, and lactation. Furthermore, although some species can excrete greater quantities of iron than humans (e.g., rodents), the same basic regulatory schemes that have been described in man appear to operate in other mammals as well.

Physiological studies of iron absorption have been carried out for many decades, but it is only within the last 20 years that the application of molecular and genetic techniques has led to an understanding of basic mechanistic details regarding the transfer of iron across the intestinal epithelium. Anatomically, iron is absorbed predominantly by intestinal epithelial cells (IECs) in the duodenum and proximal jejunum. Absorption is restricted to the differentiated enterocytes on the upper half of the villus. Most dietary iron exists in the Fe ^{3 +} , or ferric form, which is highly insoluble. Ferric iron must thus be reduced to the more soluble ferrous form (Fe ^{2 +} ) before it can be efficiently absorbed. One possible brush-border membrane (BBM) ferrireductase is duodenal cytochrome b (DCYTB), although other candidate proteins have also been identified. Moreover, dietary (e.g., ascorbic acid) and endogenous (e.g., gastric acid) factors also promote reduction of ferric iron. Once reduced, ferrous iron is transported across the BBM via DMT1. The fate of iron within the enterocyte is determined by body iron status. If iron stores are replete, most absorbed iron will be stored within ferritin and later lost when enterocytes are exfoliated from the villus tip. However, if body iron stores are low, iron will be transported out of the cell across the basolateral membrane (BLM) via the iron exporter FPN1. The export (or transfer) process requires that iron be oxidized for binding to transferrin (TF) in the interstitial fluid and plasma, and this likely occurs via the iron oxidase (i.e., ferroxidase [FOX]) hephaestin (HEPH), which is expressed on the basolateral surface of enterocytes. Iron export from duodenal enterocytes is, in part, regulated by the liver-derived, peptide hormone hepcidin, which binds to FPN1 causing its internalization and degradation. Hepcidin expression in hepatocytes is regulated by physiological signals that are transduced to the HAMP gene (encoding hepcidin) by the hemochromatosis (HFE) protein, transferrin receptor 2 (TFR2) and hemojuvelin (HJV), among others. Mutations in the genes encoding these proteins lead to impaired hepcidin production, resulting in inappropriately increased intestinal iron absorption and consequent systemic iron loading, eventually resulting in tissue and organ damage and associated pathological phenotypes. Each of these steps in iron absorption and regulatory mechanisms will be considered in greater detail in subsequent sections.

Iron homeostasis in undifferentiated enterocytes of the intestinal crypts is quite different from what occurs in their mature counterparts. These immature epithelial cells are not specialized for vectorial iron (or nutrient) transport and act rather more like a generic cell. They are rapidly growing and dividing, so they have a high demand for iron, which they absorb as TF-bound iron from the serosal side, unlike mature (i.e., fully differentiated) enterocytes which get much of their iron from dietary components on the luminal side. Both transferrin receptor 1 (TFR1)-dependent and TFR1-independent diferric TF uptake have been demonstrated in cryptal enterocytes. As the cells differentiate and migrate upward along the villus, they take on an absorptive phenotype and subsequently lose the capacity to absorb iron from the circulation.

Disturbances in intestinal iron absorption can have significant clinical consequences. In humans, the most common pathological condition that perturbs iron absorption is hereditary hemochromatosis (HH). HH is a group of genetic diseases in which hepcidin production is impaired leading to dysregulated (and increased) intestinal iron absorption and iron release from stores, eventually resulting in tissue iron overload and associated damage. The most common form of HH results from mutations in the gene encoding HFE, but several less common forms have also been described. Pathological conditions associated with reduced iron absorption, though less common, are still important from a clinical perspective. Impaired iron absorption frequently occurs in patients with malabsorptive disorders, such as celic disease or inflammatory bowel disease (IBD), in individuals who have undergone gastric bypass surgery for morbid obesity, in those who take proton-pump inhibitors for chronic gastric reflux, and in older adults with achlorhydria. Iron-refractory anemias, resulting from impairments in iron absorption, have also been described, but to date, only one of these has been attributed to an intestinal iron transport molecule. These clinical disorders will be considered in greater detail below.

60.1.2

Chapter Overview

This chapter will describe the molecular details of intestinal iron absorption, a process not fully understood, but which is being actively investigated by a large number of groups around the world. The brief overview of the importance of iron as a nutrient and of the basics of the intestinal transport process presented above will be expanded upon in subsequent sections. Dietary and endogenous factors that affect the efficiency of intestinal iron absorption will be considered, along with physical and morphological attributes of the GI system that also regulate this process. The steps involved in transferring iron across the BBM, through the cytoplasm and out of the enterocyte for loading onto transferrin will be delineated in detail. How this complex process is altered during development, both pre- and postnatally, will be outlined. Moreover, systemic and local control of intestinal iron absorption will be considered, as well as how this process is either primarily or secondarily perturbed in various disease states.

60.3.1

Other Influences on Intestinal Iron Absorption

60.4.1

General Characteristics of Iron Absorption

Iron absorption occurs through the differentiated epithelial cells (enterocytes) of the mid and upper villus, and predominantly in the proximal part of the small intestine. The proteins involved in this process are listed and briefly described in Table 60.1 . The immature enterocytes of the intestinal crypts handle iron quite differently from their mature counterparts (as noted above). They are not absorptive in nature and take up a large amount of TF-bound iron from the blood, unlike mature enterocytes. As the cells mature and differentiate, they acquire their absorptive characteristics but lose the capacity to take up iron from the circulation. Iron metabolism in crypt cells will not be considered further here.

Iron movement across the enterocyte is usually considered in terms of three phases or steps ( Fig. 60.1 ): (1) iron uptake—the movement of iron from the intestinal lumen across the BBM and into the enterocyte; (2) an intracellular phase where iron is stored or utilized within the enterocyte or directed to the BLM for subsequent export; and (3) the export of iron across the BLM and into the interstitial fluids which bathe the serosal side of enterocytes, the so-called transfer step. The small intestine is able to utilize different forms of iron and it is the BBM proteins that must deal with this variety. The utilization of inorganic, or nonheme, iron (which is likely bound to low molecular weight organic acids) is particularly important and has been most extensively studied, but iron can also be delivered to enterocytes as heme, ferritin, or lactoferrin (as discussed later). Irrespective of the dietary source, it is likely that much of the iron taken up into enterocytes enters a common cytosolic “labile” iron pool and is subsequently exported via a common FPN1-mediated pathway.

Diagram of overall body iron homeostasis. Absorption of dietary by the duodenum and proximal jejunum is depicted as well as iron delivery to tissues and cells of the body. Small amounts of total dietary iron are absorbed daily and passive iron losses more or less match these, although no regulated excretory processes exist. A larger amount of iron is recycled internally each day, deriving from the breakdown of senescent red blood cells by macrophages of the reticuloendothelial system. Most recycled iron is destined for the bone marrow for the production of hemoglobin in new erythrocytes. The liver produces the iron regulatory peptide hormone hepcidin in response to increased body iron stores, infection, and inflammation. Hepcidin decreases serum iron levels by inhibiting iron export from intestinal epithelial cells, RE macrophages, and hepatocytes.

Iron absorption is a rapid process. Following the administration of a radioactive dose of iron into the lumen of the proximal small intestine, radioactivity usually appears in the circulation within 15 s, and within minutes, 60%–80% of the total amount ultimately absorbed is transferred into the body. This initial phase of rapid iron uptake is followed by a slower rate of transfer that continues for 12–48 h. It is likely that this iron was initially retained within enterocytes in ferritin. Some iron stored in ferritin, however, never makes it into the circulation since it is lost when mucosal cells are exfoliated at the end of their lifespans. During iron deficiency, not only is the total amount of iron absorbed greater but also the fraction retained within the enterocyte is much smaller.

The quantity of iron absorbed by the intestine is related to the luminal iron concentration in a biphasic manner. At low iron concentrations, there is a direct relationship between the iron concentration and absorption, but with increasing amounts of luminal iron, proportionally smaller amounts of the nutrient are absorbed, that is, the process is saturable. This in turn reflects specific iron uptake processes. However, iron absorption never fully saturates, and with very high doses of iron, a linear relationship between the luminal iron concentration and the amount absorbed is again observed. These findings suggest that iron absorption is a saturable, carrier-mediated process at physiological intraluminal iron concentrations, but that a high-capacity, nonspecific process (possibly paracellular flux) is operating at higher concentrations. This same phenomenon has been described for the absorption of other essential minerals, notably calcium. Very large doses of oral iron can overcome feedback mechanisms, which normally limit iron absorption, likely reflecting the nonspecific component. The saturable process represents what we now recognize as the normal physiological DMT1/FPN1-mediated iron absorption pathway that will be described in detail below.

In the past, there was considerable debate over the issue of whether brush-border uptake or basolateral efflux was the rate-limiting step in iron absorption (summarized in Ref. ). Teleologically, it makes sense that the basolateral transfer step is rate limiting as the amount of iron crossing the BLM cannot exceed that crossing the BBM. The intestinal epithelium essentially acts as a buffer to control dietary iron assimilation. Iron can move across the BBM and into the enterocyte, but whether it proceeds any further depends on the export capacity of the BLM. Iron not absorbed immediately is stored within ferritin, but this iron can ultimately be utilized later if needed to meet metabolic demands. Although early kinetic and physiological investigations provided results that were somewhat conflicting, the consensus from these studies was that basolateral transfer was indeed rate limiting. In recent years, with improved understanding of the molecular basis of iron absorption and its regulation, strong support for the regulatory role of the basolateral transit of iron has been provided. This will be considered in greater detail below.

60.4.2

BBM Uptake of Iron

60.4.2.3

Heme Absorption

Heme, principally derived from myoglobin and to a lesser extent hemoglobin, is considered to be an important source of iron, particularly in diets that contain red meat. Heme is liberated from the porphoryin ring by proteases in the intestinal lumen and it is this released heme that is presented to the BBM. Heme iron is considered to be more bioavailable than nonheme iron, and because the iron is effectively bound within the porphyrin ring, heme iron absorption is largely unaffected by substances that inhibit the absorption of nonheme iron. In fact, it has been suggested that although only 10% of the iron in a typical Western diet may be in the form of heme, the heme pool contributes up to 50% of the total iron absorbed.

Despite the important contribution of heme to overall dietary iron intake, the mechanism of heme absorption remains very poorly understood. Early studies provided evidence for the presence of a specific heme-binding protein on the BBM in several species, although its identity remains unknown. Associated morphological studies suggested that after binding to the BBM, heme was internalized in endocytic vesicles, which ultimately fused with lysosomes where the heme was degraded. There is also evidence that heme might be transported directly across the BBM. Recently, a low-affinity heme transporter designated heme-carrier protein 1 (HCP1) was identified ( Fig. 60.2 ). HCP1 has some characteristics consistent with a role in heme transport in the gut, including strong expression on the BBM and increased expression in response to hypoxia, a known stimulus for iron absorption. However, subsequent studies have shown that HCP1 is also a high-affinity folate transporter (termed the proton-coupled folate transporter; PCFT). Patients with mutations in HCP1/PCFT are phenotypically folate deficient rather than iron deficient, so whether this protein contributes to heme absorption in the physiological setting is unclear. The mechanism by which dietary heme is absorbed across the BBM has thus not been clarified.

Iron homeostasis in duodenal enterocytes. Two adjacent, fully differentiated enterocytes are depicted showing the major proteins responsible for intestinal iron transport. The pathway for inorganic iron is shown on the left , beginning with iron reduction by a reductase, possibly DCYTB, followed by transport into the cell by DMT1 and culminating in iron release from the basolateral surface by FPN1 and oxidation by HEPH (and/or CP) for binding to transferrin in the interstitial fluid. A sodium‑hydrogen exchanger (NHE3) on the BBM maintains an electrochemical H ⁺ gradient, which is utilized by DMT1 to drive iron into cells (the sodium gradient is maintained by the basolateral Na ⁺ /K ⁺ ATPase—not shown). Alternative sources of dietary of iron are shown above the enterocyte to the right , including heme, lactoferrin and ferritin. The uptake pathway for heme is still not understood in detail, but HCP1 or another unidentified receptor may be involved. Heme can be broken down in endosomes by heme oxygenase (HO) and iron released into the cytoplasm via DMT1, or heme may be exported intact via FLVCR and bound to hemopexin for distribution to body tissues and cells. The entry and processing pathways for ferritin and lactoferrin are not understood in detail, but in both cases, iron appears to be able to enter cells associated with these molecules. Iron absorbed by all pathways and released into the cytosolic labile iron pool inside enterocytes is pumped across the BLM by FPN1. Also depicted are two mechanisms that regulate the expression of genes related to intestinal iron homeostasis. The first, shown in the cell on the left , becomes active during iron deficiency with concurrent hypoxia, where HIF2α is stabilized and it, along with its binding partner ARNT, enter the nucleus and induce expression of DCYTB, DMT1, and FPN1. The second regulatory mechanism is shown in the cell on the right and is commonly referred to as the IRE/IRP system. This system involves two iron-sensing cytosolic proteins (the iron-regulatory proteins), which can bind to stem loop structures (iron-response elements) in transcripts encoding iron transport-related proteins and increase (by stabilizing the transcript at the 3′ end and enhancing translation) or decrease (by blocking translation at the 5′ end) protein expression.

After crossing the apical membrane, heme has two potential fates: being catabolized within the enterocyte or trafficked intact across the BLM and into the circulation. The latter does not seem to be a major route, as studies which followed the fate of heme with a radioactive iron center showed that the vast majority of the heme-derived iron appears in the circulation bound to transferrin, that is, it appears to exit the cell via the same pathway as nonheme iron. Despite this, two heme export proteins have been identified that could function in the enterocyte: FLVCR (feline leukemia virus, subgroup C, receptor) and ABCG2 (also referred to as the breast cancer resistance protein). FLVCR plays a role in reducing excess heme levels in erythroid cell precursors and in heme release from macrophages. ABCG2 appears to facilitate heme efflux from immature erythroid cells. Both proteins are expressed in enterocytes and intestinal cell lines (with ABCG2 being localized to the BBM), but whether these proteins play a role in heme iron absorption or in protecting IECs against heme-related toxicity remains to be determined.

The dominant fate of heme within the enterocyte is degradation and this is mediated by heme oxygenases (HOs). The majority of studies have concluded that HO-1 is the main enzyme involved in processing newly absorbed heme, but this has not been proven definitively. A recent study showed that lack of intestinal HO-1 did not impair heme absorption in mice, although the authors suggested that mice are poor utilizers of dietary heme. Recent localization data have also implicated HO-2 in the process and have suggested that this is the more likely of the two enzymes involved in releasing iron from intracellular heme. It is possible that both enzymes contribute to heme degradation in enterocytes. Whichever the case though, after iron is released from heme, it is predicted to enter the same transit pool as nonheme iron and thus be transported across the BLM via FPN1.

Like nonheme iron, heme iron transport in the intestine is regulated by iron demand, being enhanced when iron requirements increase. However, heme iron absorption is not regulated over as wide a range as nonheme iron absorption. This seems somewhat surprising since basolateral transfer is considered rate limiting for absorption and both heme and nonheme iron appear to utilize the same efflux pathway. This apparent discrepancy can be reconciled by the finding that it is the degradation of heme that is likely rate limiting for absorption. While HO-2 shows limited regulation, HO-1 can be induced by iron deficiency, providing a mechanism for the (more limited) regulation of heme iron absorption.

60.4.3

The Intracellular Phase of Iron Absorption

The fate of iron after it crosses the BBM is poorly understood. Within the enterocyte, as with all cells, iron is considered to enter an exchangeable “pool,” the nature of which is not well defined. It seems likely that iron is present in the cytoplasm as chelates with small organic acids such as citrate or amino acids, or it could be bound with low affinity to intracellular proteins. Membrane permeant iron chelators such as desferrioxamine inhibit iron absorption, presumably by chelating iron in this intracellular pool. For another redox-active and potentially toxic transition metal, copper, several specific and high-affinity intracellular- binding proteins, or chaperones, have been identified ; however, with one exception (see below), intracellular iron chaperones have not been identified to date. Irrespective of what iron is bound to in the intracellular pool, it is clear that this iron can have one of three fates: (1) if body iron requirements are high, then the metal can move quickly across the BLM and into the bloodstream; (2) if body iron demand is low, the metal is stored within ferritin in the enterocyte; or (3) enterocytes may also utilize newly absorbed iron for metabolic purposes, but it is likely that most of the iron required in enterocytes is obtained when the cells are produced from stem cells in the intestinal crypts. Furthermore, the intracellular iron pool can influence posttranscriptional regulation of genes encoding proteins involved in cellular iron metabolism, as discussed below.

The best characterized pool of iron in the enterocyte is that within the iron-storage protein ferritin. The 24 subunits of apo -ferritin form a hollow sphere which can accept up to 4500 atoms of iron. Iron not required for basolateral transport and in excess of cellular metabolic needs is incorporated into ferritin, but this iron can subsequently be accessed if metabolic demand increases, at least in the short term. If the iron is not reutilized, the ferritin is lost from the body in the feces after the enterocytes are exfoliated into the intestinal lumen. When very large amounts of iron are incorporated into ferritin, ferritin complexes form and ultimately fuse with lysosomes, leading to the degradation of the protein shell; the resulting amorphous mixture of peptides and iron oxide is known as hemosiderin. The translation of ferritin mRNA is regulated by iron such that increased cellular iron levels lead to the synthesis of ferritin, which sequesters the iron and prevents it from accumulating in the cytoplasm to toxic levels. In this sense, ferritin would be expected to play a passive role in iron absorption. However, it has been shown that overexpression of ferritin in cultured cells depletes cytosolic iron, and there is evidence that a similar effect can occur in enterocytes in vivo under certain pathological circumstances. Thus, the possibility that ferritin can actively modulate the passage of iron across the enterocyte cannot be excluded. A recent study, in fact, suggests that ferritin functions in parallel with hepcidin to prevent excessive iron absorption.

Although there is limited evidence to indicate that protein chaperones might play a role in the intracellular trafficking of iron, it was recently demonstrated that poly (rC)-binding protein 1 (PCBP1) was able to facilitate iron loading onto ferritin. More recent work demonstrated that PCBP2, a PCBP1 homolog, interacts with DMT1 and FPN1, thus possibly functioning in a similar manner to copper chaperones, but whether this occurs in enterocytes has not yet been investigated.

60.4.4

BLM Transfer of Iron

The movement of iron across the BLM of the enterocyte (the transfer phase) plays a critical role in iron absorption. It is the BLM that is the main interface with the body and as such, it receives information about body iron requirements. This in turn makes it the primary site for the regulation of dietary iron intake. The membrane passage of iron is mediated by the iron export protein ferroportin 1 (FPN1), but an iron oxidase is also required for efficient transport. The iron export‑iron oxidation functional couple is considered in detail below.

60.5.1

Stimuli That Alter Iron Absorption

Changes in the two largest pools of iron within the body—storage iron and iron in the erythroid mass—exert the greatest influence on iron absorption. Absorption is increased when body iron stores are low or when the erythropoietic rate is increased, and vice versa. Absorption is also increased during chronic hypoxia, pregnancy and the suckling period (as discussed in detail below). The discussion here will focus largely on nonheme iron absorption. The absorption of heme iron is also affected by variations in iron status and erythropoiesis, but the magnitude of change is smaller. Moreover, after heme has entered the enterocyte and iron has been released from the protoporphyrin ring, the liberated iron is predicted to follow the same pathway into the body as recently absorbed nonheme iron, and thus, this component of intestinal iron transit should be regulated similarly for both dietary sources of iron.

The level of body iron stores is a major factor regulating nonheme iron absorption in a normal individual. In both humans and in animals, an inverse relationship exists between iron stores and absorption, with depleted stores leading to increased absorption and elevated stores reducing it. Finch coined the term the “store regulator” to describe the regulation of iron absorption by body iron levels. The body appears to have an optimum level of storage iron and in a normal adult male human this is approximately 1 g. If stores drop below this level, absorption is increased until they are replenished, whereas if stores are elevated, absorption will decline until equilibrium is regained. Indeed, under normal conditions, once iron stores have reached the optimum level, it is difficult to induce the body to accumulate more even if the iron content of the diet is increased significantly.

The other major physiological process influencing iron absorption is erythropoiesis. The largest pool of iron in the body is in the hemoglobin of circulating erythrocytes, and consequently the erythroid marrow is the single most significant sink for iron. The rate of erythropoiesis has been shown to have a direct relationship with iron absorption in animal studies. In humans, iron absorption is also enhanced when erythropoiesis is stimulated by bleeding or acute hemolysis or following androgen or erythropoietin treatment, and decreased following hypertransfusion or radiation treatment. The capacity of the body to increase iron absorption in response to an increased erythroid iron demand is very high, and in IDA, for example, individuals may absorb over 20 mg/day providing dietary iron supply is sufficient. The magnitude of this response is far greater than that typically seen with the store regulator, and thus the term “erythroid regulator” has been proposed. As will become apparent below, experimental advances have blurred the distinction between the store and erythroid regulators. The erythropoietic effect on absorption is very strong and under many circumstances it can override the effect of changes in body iron stores. In β-thalassemia for example, high iron absorption continues in the setting of elevated body iron stores due to the high marrow demand for iron, and in the iron-loading disorder hereditary hemochromatosis (HH), venesection therapy will stimulate absorption even though iron stores may be greatly elevated. Nevertheless, these regulatory factors each exert their effects and, for example, the high drive to increase iron absorption in β-thalassemia can be blunted when body iron levels become very high.

The adequacy of iron supply to the marrow and the rate of effective erythropoiesis are important determinants of iron absorption. However, an increased rate of erythropoiesis per se need not stimulate iron absorption. For example, in chronic hemolytic states, the absorption rate is usually normal despite increased erythroid activity. In this situation, as long as the return of iron to the plasma from the destruction of senescent red cells or from storage is sufficient to provide for the increased erythroid activity, absorption will not be increased. Another example is found in a murine model of mild β-thalassemia where erythroid iron demand is increased, but iron absorption is not. However, once the ability to meet the marrow’s requirements by catabolism of red cells or mobilization of storage iron is exceeded, an increase in iron absorption will result. When erythroid demand is very high and erythropoiesis is markedly ineffective (for example, in iron loading anemias such as β-thalassemia and sideroblastic anemia), internal iron traffic cannot meet the marrow demand and iron absorption is elevated.

Iron absorption is also known to be increased by a number of other stimuli that are worth mentioning briefly. Absorption is increased under hypoxic conditions and returns to normal after the hypoxic stimulus is removed. Part of the effect of low oxygen tensions can be ascribed to alterations in the rate of erythropoiesis, but a component of the response is independent of these effects. For example, early studies showed that an effect of hypoxia on iron absorption can be observed before a response from the erythroid marrow, suggesting that hypoxia can exert direct effects on the intestinal enterocyte. Studies in recent years have provided a mechanism for this by describing effects of hypoxia on both the body iron homeostatic machinery in the liver and on the expression of genes encoding iron transport proteins in the intestine. Absorption can also be increased in the perinatal period. Moreover, during pregnancy, there is a high demand for iron to provide for an expansion of the maternal erythroid mass and the iron requirements of the developing fetus. Iron absorption is increased several fold to meet these requirements, but it seems likely that the drive for this increase is provided by reduced maternal iron stores and hypoxia rather than by a pregnancy-specific factor. Finally, during the suckling period, mammals absorb iron from breast milk extremely efficiently. Iron demands are certainly very high at this time due to rapid growth and no doubt provide a strong driving force for absorption, but the immature intestine has some quite different properties from the adult intestine so the passage of iron across the intestinal epithelium during suckling may use pathways that are not available in the adult. Each of these physiologic situations that increase iron absorption will be considered in more detail below.

60.5.2

Hepcidin and the Regulation of Body Iron Homeostasis

The liver-derived, peptide hormone hepcidin is now considered the main systemic regulator of body iron homeostasis, including intestinal iron absorption, particularly in the setting of increased body iron stores and during infection and inflammation. Hepcidin was originally identified in plasma and urine as an antimicrobial peptide and was subsequently linked to iron homeostasis by two groups. In one of these investigations, inactivation of the Usf2 gene in mice inadvertently disrupted expression of the adjacent Hamp1 (hepcidin) gene and the mice subsequently developed significant iron overload. In the other study, the Hamp1 gene was identified as being strongly upregulated in the liver of mice that had been loaded with iron. These studies showed that the expression of hepcidin was inversely related to body iron intake and implied that hepcidin was a repressor of iron absorption. This was confirmed when it was demonstrated that mice engineered to overexpress hepcidin developed a severe IDA. Other studies also showed that there was a close inverse relationship between the expression of hepcidin by the liver and the expression of iron transporters in the duodenum. The importance of hepcidin in maintaining body iron balance is well exemplified by patients with mutations in the HAMP gene. Such patients develop an early onset, severe iron-loading disorder (juvenile hemochromatosis) as a result of an uncontrolled high level of intestinal iron absorption. Iron absorption is increased when hepcidin levels are reduced and disturbances in hepcidin production underly many inherited iron-loading diseases, as will be discussed below.

The major site of hepcidin synthesis is hepatocytes in the liver. Most circulating hepcidin is likely derived from these cells, but some studies have shown that small amounts of hepcidin can be produced locally in other tissues, including, but not limited to, macrophages and adipocytes. Of particular interest in the context of iron absorption is the demonstration that hepcidin is expressed in gastric parietal cells, where it may play a role in acid secretion, which promotes iron absorption in the duodenum. Hepcidin is produced as a preprotein of 85 amino acids, and it is subsequently processed by furin-mediated cleavage and secreted from the cell. In the plasma, hepcidin peptides of 20, 22, and 25 amino acids have been detected, but it is the 25 amino acid form that is responsible for the bulk of its biological activity.

Hepcidin exerts its profound effects on body iron metabolism by acting on FPN1 to reduce iron export from cells. While this activity has only been shown directly for a few cell types, it is likely that hepcidin acts on most body cells in this way. Duodenal enterocytes are one of the major targets of hepcidin and this explains its effects on iron absorption. Other important targets include RE macrophages and hepatocytes. Hepcidin interacts with FPN1 on the plasma membrane and facilitates the internalization and degradation of FPN1, thus reducing iron efflux. Although full details have not been elucidated, the ubiquitination of multiple cytoplasmic lysines is required for the internalization of FPN1 following hepcidin binding. Early reports suggesting that hepcidin binding leads to phosphorylation of tyrosine residues on FPN1 by Jak2 kinase have not been independently substantiated. After internalization, the protein is targeted for lysosomal degradation. Since FPN1 is expressed on the enterocyte BLM, the action of hepcidin provides a molecular explanation for why the basolateral transfer step is rate limiting for iron absorption and why hepcidin levels are inversely related to iron absorption.

60.5.3

The Regulation of Hepcidin

Consistent with its role as the central regulator of body iron traffic, hepcidin synthesis is decreased when iron demands are high, for example, during iron deficiency, when erythropoiesis is stimulated and in pregnancy, leading to increased iron absorption and release of storage iron. Conversely, hepcidin production is increased when the body is iron replete, resulting in reduced intestinal absorption and the storage of excess iron in RE macrophages and hepatocytes. Hepcidin expression is also stimulated by infection and inflammation, resulting in the hypoferremia associated with the anemia of chronic inflammation.

Changes in the level of circulating hepcidin likely explain variations in iron absorption, but an important question relates to how hepcidin production is coordinated with body iron status. Important clues to how this occurs have come from the analysis of inherited iron-loading disorders, collectively known as hereditary hemochromatosis (HH). In a normal individual with an increased body iron load, iron absorption is reduced to prevent excessive tissue iron accumulation; however, in hemochromatosis patients, this does not occur and iron absorption is inappropriately increased in relation to metabolic needs. Mutations in the HAMP gene lead to an early-onset, severe form of iron loading known as juvenile hemochromatosis. Juvenile hemochromatosis can also be caused by mutations in the hemojuvelin ( HJV ) gene. However, the most common form of inherited iron loading has a far milder phenotypic presentation and an adult-onset, which results from mutations in the HFE gene. On average, approximately 1 in 200 individuals in populations of northern European origin are homozygous for the major HFE mutation (C282Y) that increases the risk for developing pathologic iron overload. A rarer form of adult-onset hemochromatosis results from mutations in the transferrin receptor 2 ( TFR2 ) gene, but the presentation of these patients, both clinically and histologically, is very similar to patients with HFE-associated hemochromatosis. Despite variations in the severity of these four forms of HH, they share several key features: (1) iron absorption is increased in each case; (2) iron is deposited in the liver in a periportal distribution and predominantly in hepatocytes; (3) RE macrophages are relatively iron deficient (particularly in the early stages of the diseases); and (4) the mutated gene in each case is strongly expressed in hepatocytes. These commonalities suggested that these genes encode proteins that function in the same regulatory pathway ( Fig. 60.3 ). Consistent with this interpretation, subsequent studies showed that all of these diseases share the critical feature of inappropriately low or absent hepcidin expression in the setting of increased body iron levels.

Pathways for the regulation of hepcidin in hepatocytes. A variety of systemic stimuli that reflect body iron requirements act on hepatocytes to alter transcription of the HAMP gene. The BMP/SMAD signaling pathway appears to play a central role in HAMP regulation. Proteins, such as HJV and TMPRSS6, that are defective in human disorders of iron homeostasis, act through this pathway to increase or decrease hepcidin expression, respectively. HFE and TFR2 are also mutated in human iron loading disorders, but precisely how they alter HAMP expression is unclear. The effects of hypoxia and inflammatory cytokines are better defined. Hepcidin secreted by hepatocytes into the circulation travels to enterocytes, RE macrophages and other cell types to regulate how much iron they release into the plasma via interaction with FPN1 (which essentially serve as the hepcidin receptor).

Analysis of the role played by HJV in hepcidin regulation has proved particularly rewarding. HJV is one of a family of proteins known as repulsive guidance molecules (RGMs). Other members of this family, unlike HJV, are involved in development of the central nervous system, and like HJV, have been shown to act as cofactors for the signaling of the bone morphogenetic proteins (BMPs). HJV is a BMP coreceptor and signals through this pathway to increase hepcidin expression. The BMPs are members of the TGF-β superfamily of ligands and act through cell-surface BMP receptors to activate the multifunctional SMAD signaling pathway. Mice in which the key signaling component SMAD4 is deleted develop severe iron overload, indicating the importance of this pathway in iron homeostasis. A series of subsequent studies has shown that BMP/SMAD signaling is critical for HAMP expression and that many of the factors that influence hepcidin levels do so through their effects on this pathway. A range of BMPs have been shown to stimulate hepcidin gene transcription, but it appears that BMP6 is likely the most important isoform in vivo . BMP6 is a strong stimulator of hepcidin expression and Bmp6-deficient mice have reduced hepcidin expression and display body iron loading. Furthermore, BMP6 expression positively correlates with body iron levels, providing evidence that BMP6 could be an important link between body iron requirements and hepcidin expression. Interestingly, BMP6 appears to be expressed most strongly in nonparenchymal cells of the liver, implying that it might exert its effects on hepcidin expression in a paracrine manner.

HJV exists in both a membrane-bound and a soluble form (sHJV). While the membrane-bound form acts as a BMP coreceptor to stimulate hepcidin expression, sHJV has been shown to antagonize signaling via the BMP-SMAD pathway, presumably by competitively inhibiting the binding of membrane-bound HJV to BMPs. The release of sHJV requires the protease furin and furin levels are increased by iron depletion and hypoxia, providing another potential link between body iron status and hepcidin expression. Two other proteins have been shown to play important roles in modulating the signaling of HJV through the BMP/SMAD pathway. The first is a membrane protein called neogenin that appears to stabilize membrane-bound HJV and inhibit sHJV secretion. The demonstration that neogenin null mice have low hepcidin levels and develop iron overload is consistent with this. The other protein is a membrane-bound serine protease known as matriptase-2 (or TMPRSS6). The physiological substrate of matriptase-2 appears to be membrane-associated HJV and neogenin may be required for HJV cleavage by TMPRSS6. After HJV is cleaved, it can no longer act as a BMP coreceptor and hepcidin levels are reduced. This implies that the normal function of TMPRSS6 is to repress hepcidin production via the BMP-SMAD pathway.

It is well established that patients with HFE-related hemochromatosis absorb too much dietary iron relative to body iron load, and more recent studies have shown that hepcidin levels are low in such patients. However, how HFE acts to modulate hepcidin expression remains poorly understood. HFE may exert its effects on BMP/SMAD signaling, either directly or indirectly through the MAPK/ERK pathway. More recently, HFE has been shown to bind to the BMP receptor complex. Patients with HFE mutations are not as severely affected as those with HJV or HAMP mutations, indicating that HFE is an important, but not essential, part of the hepcidin regulatory pathway. This could have been predicted from earlier physiological data that showed that iron absorption was effectively regulated in hemochromatotic patients (likely carrying HFE mutations), but that at any given body iron load, iron absorption was higher in iron-loading patients than in normal controls. HFE thus appears to affect the sensitivity of the hepcidin-mediated absorption response to changes in body iron levels. A similar role can be ascribed to TFR2. Patients with mutations in TFR2 also have a relatively mild form of hemochromatosis, with a phenotype similar to subjects with HFE mutations, and reduced, but not absent, hepcidin expression. Again, how TFR2 influences hepcidin expression is not known, but MAPK/ERK signaling has been implicated. Interestingly, mice lacking both HFE and TFR2 have lower hepcidin levels and a more severe iron-loading phenotype that in animals with either deletion alone, suggesting that HFE and TFR2 exert independent effects on hepcidin expression.

Although BMP6 and furin are regulated by body iron levels and provide potential mechanisms by which body iron requirements might be relayed to the HAMP gene, it has also been postulated that HFE and TFR2 are involved in this iron-sensing process. The basis of this prediction is that both proteins have strong links to circulating diferric transferrin, which delivers iron to most cells. TFR2 binds to and is stabilized by circulating diferric transferrin, and HFE has been shown to interact with TFR1, which also binds diferric transferrin. The level of circulating diferric transferrin (or relative transferrin saturation) is thus thought to be an important signal for modulating hepcidin expression. Diferric transferrin levels increase with iron loading and decrease with iron deficiency, and a positive correlation between the concentration of diferric TF and hepcidin expression has been demonstrated. A working model has been proposed in hepatocytes whereby diferric TF displaces HFE from the HFE/TFR1 complex, leaving HFE free to interact with TFR2 and initiate signaling through the BMP/SMAD pathway or perhaps via some other mechanism. This is an attractive hypothesis, but it does not fit well with a number of in vivo studies. Despite considerable work in this area, much of it using proteins overexpressed in cell lines, we still do not have a clear picture of how the iron-sensing mechanism functions, particularly in vivo, nor indeed whether or not diferric TF provides the key regulatory signal.

Enhanced erythropoiesis stimulates iron absorption and suppresses hepatic hepcidin expression. There are several potential mechanisms for the suppression of HAMP by erythropoiesis, including iron-related downregulation of hepcidin expression in response to heightened iron demand, or tissue hypoxia as a result of reduced levels of circulating hemoglobin. Indeed, the HAMP gene contains two conserved hypoxia-response elements and it has been demonstrated that HIF1α signaling reduces HAMP transcription. While hypoxia-signaling factors may exert some influence on the body’s response (i.e., the hepcidin response) to increased erythropoiesis, there has long been evidence that additional, erythroid-specific, regulatory factors play a more significant role. Some experimental evidence has suggested that growth differentiation factor 15 (GDF15) and twisted gastrulation 1 (TWSG1) could play this role as their expression is increased when erythropoiesis was stimulated and both have been shown to act on hepatic cells to reduce hepcidin expression. Other studies, however, have not found a correlation between GDF15 and hepcidin levels. For example, hepcidin expression was appropriately regulated in response to phlebotomy in Gdf15 -knockout mice, demonstrating that GDF15 is not required for the downregulation of hepcidin expression when erythropoietic demand increases. More recently, another factor, erythroferrone, has emerged as the most likely erythroid regulator and this has now been supported by multiple studies. Precisely how erythroferrone influences hepcidin production is not known, and it remains possible that the hepcidin response to stimulated erythropoiesis involves other, as yet unidentified, regulatory factors.

A final comment should be made on the relationship between inflammation and iron absorption. In response to an inflammatory stimulus, either acute or chronic, there is a reduction in iron absorption and in the donation of iron to the plasma from RE macrophages and other iron storage sites. This is associated with a sequestering of iron within tissues. Inflammation is a strong inducer of hepcidin and the hypoferremia, and ultimately anemia, that accompanies inflammation can be explained by increased plasma hepcidin levels, although the situation is likely to be multifactorial with factors other than hepcidin contributing. Inflammation stimulates the synthesis of a range of proinflammatory cytokines including interleukin-1 (IL-1), IL-6, IL-22, and interferon-α, and it is these signaling molecules that enhance hepcidin expression. IL-6-mediated enhancement of HAMP transcription is mediated by signaling through the JAK-STAT pathway. There is a crosstalk between this pathway and the BMP-SMAD signaling pathway, as a functional SMAD-binding site in the HAMP promoter is necessary for the IL-6 stimulation of hepcidin expression. However, since hepcidin responds strongly to inflammation in Bmp6 -null mice, the necessary SMAD signal must be being supplied by a BMP receptor ligand other than BMP6.

60.5.4

Regulation at the Level of the Mucosa

Iron homeostasis is systemically regulated by the liver-derived, peptide hormone hepcidin, as outlined above, and this hormonal control is important to prevent iron overload, and in the response to infection and inflammation as part of the acute-phase response. There is, however, additional control at the level of the intestinal epithelium, and it is ultimately this epithelium that determines how much iron enters the body. These regulatory mechanisms involve local alterations in the expression of key iron transport molecules, as well as morphological changes in the epithelium, all in response to changes in iron status.

60.6.1

Pregnancy

Iron needs increase during pregnancy and many studies have shown enhanced absorption of dietary iron during gestation. Studies in pregnant humans using either radioactive or stable iron isotopes have shown incremental increases in iron absorption from the first through the third trimesters of gestation. These studies have all suggested that iron absorption increases as the iron demands of the fetus increase. Associated with this increase is a decline in the expression of hepcidin, and this is most likely driven by the reduction in maternal body iron stores. However, not all studies have found correlations between iron and hepcidin levels during gestation, and other factors including steroid hormones associated with pregnancy, could also be involved.

A range of studies on intestinal iron transport during pregnancy have also been performed in laboratory rodents. These studies broadly parallel the investigations in humans and show increases in absorption as pregnancy progresses. In addition to increased absorption, one study in rats documented increased duodenal circumference and villus dimensions during pregnancy. Consistent with an increase in iron absorption, some studies have also documented increased expression of genes encoding iron transport proteins in the rat intestine during pregnancy.

60.6.2

Suckling

Iron homeostasis in neonatal mammals is fundamentally different from iron homeostasis in later life. The capacity of suckling infants to absorb iron is extremely high, as this is a period of very rapid growth with particularly high iron demands. Once mammals wean, however, absorption declines to adult levels. The mechanisms underlying the very high iron absorption during suckling are incompletely understood. At this developmental stage, the small intestine is relatively immature and tight junctions between epithelial cells are not fully developed, raising the possibility that the gut mucosa is “leaky” and that iron may be taken up by nonspecific processes. While this has been proposed as a possible mechanism to explain high neonatal iron absorption, it appears unlikely to be a major contributor as absorption at this stage appears to be specific and most prominent in the proximal small intestine, whereas nonspecific uptake is largely observed in more distal gut segments. More recent studies in mice have revealed that the vast majority of iron absorption during suckling is FPN1 dependent, confirming that it is a specific process. Interestingly, however, during suckling, iron absorption cannot be repressed by an increased level of systemic hepcidin, and this has been proposed to be due to a modification of FPN1, which prevents hepcidin binding. Earlier studies in humans and rodents suggested that there was little homeostatic regulation of iron transport during early infancy, but that regulation began around the time of weaning. Whether the high, unregulated absorption of iron during infancy has evolved to maximize iron intake, or whether it plays some other role, such as removing iron from the intestinal lumen to limit the growth of certain bacteria, remains to be determined. The latter is a possibility as it is generally acknowledged that full-term human infants and rodent pups are born with sufficient iron stores to take them through early postnatal life. One other particularly interesting aspect of iron absorption in early postnatal life is that the role of DMT1 in iron transport from the intestinal lumen into enterocytes appears to be unclear. One study has suggested that DMT1 is not required at this developmental stage in rats, implying that an alternative BBM transporter must be operating. Conversely, a more recent study in mice has presented evidence that DMT1 is involved. Additional studies will thus be required to resolve this apparent discrepancy and further to elucidate the functional role of DMT1 during suckling in humans.