Research Article: Heme Oxygenase-1: A Critical Link between Iron Metabolism, Erythropoiesis, and Development

Date Published: November 20, 2011

Publisher: Hindawi Publishing Corporation

Author(s): Stuart T. Fraser, Robyn G. Midwinter, Birgit S. Berger, Roland Stocker.


The first mature cells to arise in the developing mammalian embryo belong to the erythroid lineage. This highlights the immediacy of the need for red blood cells during embryogenesis and for survival. Linked with this pressure is the necessity of the embryo to obtain and transport iron, synthesize hemoglobin, and then dispose of the potentially toxic heme via the stress-induced protein heme oxygenase-1 (HO-1, encoded by Hmox1 in the mouse). Null mutation of Hmox1 results in significant embryonic mortality as well as anemia and defective iron recycling. Here, we discuss the interrelated nature of this critical enzyme with iron trafficking, erythroid cell function, and embryonic survival.

Partial Text

Adult humans require 4-5 mg of iron per day [1]. The predominant use of iron is in its reduced ferrous state (Fe2+) in heme complexed with two α-globin and two β-globin chains to form hemoglobin. Reduced iron in heme is uniquely amenable for the transport of oxygen and carbon monoxide, while hemoglobin can also bind allosteric ligands such as carbon dioxide and nitric oxide. When red blood cells are lysed, iron in the released hemoglobin can become oxidized to its ferric (Fe3+) form, and heme no longer binds tightly to hemoglobin and may be released [2, 3]. The hydrophobicity of the resulting free heme allows it to cross cell membranes leading to oxidative stress within cells [4, 5]. If left uncontrolled, heme and its iron can contribute to the cellular labile iron pool and act as a pro-oxidant by catalytically amplifying the production of oxidants inside the cell via Fenton chemistry whereby Fe2+ is oxidized by hydrogen peroxide to Fe3+, a hydroxyl radical, and hydroxyl anion. Amplification of oxidant production by redox-active iron can directly lead to lipid, protein, and DNA damage and ultimately cell death. However, free iron can be rapidly neutralized by a number of metabolic pathways induced by the iron itself. These include the induction of iron efflux systems or upregulation of ferritin, a cytosolic protein that sequesters iron in a redox-inactive form, thereby limiting its pro-oxidant capacity. Free heme has also been shown to have proinflammatory properties. With the vast amount of heme used in erythroid cells, mechanisms must be put into place to allow the body to cope with free iron and heme released during red blood cell clearance in the spleen.

Heme oxygenases are the initial and rate-limiting enzymes in the breakdown of heme (iron protoporphyrin IX) that itself plays an essential role in the transport of oxygen and mitochondrial electron transport as a cofactor of hemoglobin, myoglobin, and cytochromes. Degradation of heme generates carbon monoxide, iron, and biliverdin, the latter of which is subsequently converted to bilirubin by biliverdin reductase (Figure 1). Heme oxygenases bind heme in a 1 : 1 molar complex where turnover of each heme molecule requires three molecules of O2 and seven electrons derived from NADPH supplied by cytochrome P450 reductase. Heme oxygenases are unusual in their ability to use heme both as a substrate and prosthetic group for its own degradation. However, heme oxygenases are not heme proteins in the classic sense of the cytochromes or peroxidases [6]. The removal of toxic heme is not the only function of heme oxygenases, as the products of their enzymatic activity have now being recognized to play significant roles in vascular biology, iron recycling, and cellular protection against oxidative stress and diseases associated with such stress [7, 8].

The erythroid or red blood cell lineage is the main consumer of iron and heme in the human body. This lineage is also the first cell type to mature during mammalian embryogenesis [15]. In mice, at least five distinct forms of erythropoiesis, or red blood cell production, have been identified (Figure 2). The erythrocytes produced during these distinct waves may vary according to cell size and globin gene expression profile, but all serve to transport oxygen and metabolic waste products via hemoglobin. Systems regulating both heme production and catabolism are, therefore, required from early embryonic development and throughout adulthood to prevent the release of free heme and subsequent oxidative stress.

Macrophages play crucial roles in erythropoiesis from the cradle of the bone marrow to the grave of the spleen. In erythropoietic tissues, such as the fetal liver and bone marrow, erythroblasts are found in intimate proximity to supportive macrophages in multicellular structures termed erythroblastic islands (EBIs) [21]. These structures are thought to play crucial roles in regulating erythroid cell production with a central macrophage often interacting with numerous erythroblasts [22]. Central EBI macrophages have been postulated to transfer iron to the developing erythroblasts to complete hemoglobin synthesis. This idea is strengthened by the initial description of EBI macrophages as containing iron detected by Perls’ staining [21]. However, developing erythroblasts also express high levels of the transferrin receptor, CD71, demonstrating the vast appetite for iron that these cells have [23, 24].

Heme oxygenase-1-deficient mice have been generated by targeted null mutation of the Hmox1 gene. The original authors, other groups, and our own group (unpublished data) have noticed abnormally low number of Hmox1−/− mice at birth, strongly suggesting that the Hmox1 deficiency is lethal in a significant number of embryos [29, 30]. Approximately 20% of the expected number of Hmox1−/− pups are born live [29]. The infrequent Hmox1−/− animals that survive until adulthood are smaller in size compared with their littermate wild-type mice and exhibit a range of pathological conditions including splenic, hepatic, and renal fibrosis as well as increased expression of indicators of stress and inflammation [31]. Accumulation of iron is observed in tissues throughout the body of Hmox1−/− mice, strongly suggesting that these animals suffer from defects in the trafficking of iron from peripheral tissues back to sites of erythropoiesis [32].

Embryonic survival appears to be linked to HO-1 gene dosage. The mother of the Japanese HO-1-deficient patient was found to be heterozygous for Hmox1 and had experienced several fetal deaths [33]. Supporting this, polymorphisms in the human HMOX1 gene have been associated with an increased risk for idiopathic recurrent miscarriage [36]. HO-1 is expressed widely throughout the placenta including the syncytiotrophoblasts and endothelial cells [37]. Carbon monoxide produced locally by heme catabolism by HO-1 is thought to contribute to the regulation of vascular tone within the placenta. HO-1 expression is also thought to stimulate trophoblast invasion into the placenta [38]. To investigate possible placentation defects, Zhao and colleagues examined Hmox1+/− embryos and found that their placentae were reduced in weight [30]. This was attributed to apoptosis in the spongiotrophoblast layer. HO-1 expression during ontogeny has been examined in the rat, where the enzyme is expressed in the syncytiotrophoblasts of the placenta as well as the endodermal layer of the yolk sac [39]. Expression was also detected in immature macrophages in the yolk sac and later in macrophages in the fetal liver [39]. Similarly, during the first 48 hours of development of the zebra fish embryo, Hmox1 mRNA was detected in yolk syncytial layer, blood cells, as well as the lens [40]. Hmox1 deficiency may, therefore, affect development of the mouse yolk sac through changes in the endoderm, as well as resulting in macrophage defects. Collectively, these data implicate roles for HO-1 in placenta formation, development of the yolk sac, and function of macrophages in the embryo itself. Interference with any of these developmental steps could lead to embryonic lethality.

The spleen encounters the greatest amount of heme, arising from the clearance of aged or damaged erythrocytes by red pulp macrophages. Spleens from Hmox1−/− mice are enlarged and show decreased amounts of free iron, as assessed by Perls Prussian blue staining [32]. Macrophages isolated from such spleens are reduced in number and show intoxication from heme. These macrophages are able to engulf and destroy aged or damaged red blood cells. However, they fail to detoxify the heme released from the breakdown of the engulfed erythrocytes due to the lack of Hmox1 [32]. This is thought to lead to macrophage death and splenomegaly. Spleens from Hmox1−/− mice are also highly fibrotic, again a defect most likely associated with the lack of macrophages. This results in profound changes in iron recycling. In the wild-type setting, macrophages in the spleen and liver are the main iron trafficking cells. In the absence of HO-1, proteins that serve to sequester free hemoglobin and heme (i.e., haptoglobin and hemopexin, resp.) are elevated in the liver and kidney [32]. In the case of haptoglobin, this may be a consequence of the drastic decrease in the expression of CD163, the scavenger receptor that normally removes haptoglobin-hemoglobin complexes from the circulation [32]. Instead, hepatocytes and the proximal tubular epithelial cells of the kidney become the predominant cell types regulating the reutilization of iron derived from erythrocytes in HO-1−/− mice, likely relying on the enzymatic activity of HO-2 [32]. This suggests that compensatory pathways are enhanced in the absence of the main system to detoxify the body of heme.

One of the most intriguing questions in this field is why most Hmox1−/− embryos die in utero while a small fraction are born live and can survive until late adulthood? Knockout mice lacking genes essential for erythropoiesis are often 100% embryonic lethal (e.g., all Eklf-deficient embryos die in utero [43]) or survive until adulthood with milder defects. A phenotype resulting in most embryos dying in utero with a small percentage surviving until late adulthood is, therefore, highly unusual. The fact that a small number of Hmox1−/− mice and HO-1-deficient individuals are born live suggests that mechanisms, such as genetic or environmental modifiers, can rescue these defects sufficiently to allow survival until birth. Is the lethality seen in HO-1 knockout embryo due to death of macrophages intoxicated by heme they are unable to catabolize following engulfment of embryonic erythroid cells? Or is HO-1 playing a more important role in epithelial cells such as the syncytiotrophoblasts or visceral endoderm of the yolk sac? Alternatively, environmental factors derived from the mother may somehow rescue this typically lethal phenotype. Resolving these modifiers will shed light onto the function of HO-1. A clear reason for the survival of a fraction of the HO-1−/− mice to birth and even one year of age, while others die in utero, is still not readily apparent.

During evolution, the utility of both iron and heme has been harnessed to drive basic biochemical activities within our cells. Erythroid cells, in particular, have used these compounds extremely effectively to transport gases throughout our bodies. However, this is a double-edged sword as the very activities that make iron and heme useful are the same that make them potentially toxic to our cells. HO-1 is essential for ameliorating its toxicity. This is highlighted by the embryonic lethality of HO-1 deficiency. The activity of this enzyme is, therefore, a critical tipping point between health and anemia or embryonic death. Increasing our knowledge of the roles of HO-1 in embryonic development may provide a better understanding of how the enzyme provides protection against oxidative stress-associated diseases.




0 0 vote
Article Rating
Notify of
Inline Feedbacks
View all comments