downloads.hindawi.comdownloads.hindawi.com/journals/specialissues/852308.pdf · 2019-08-07 ·...

Crosstalk between Erythropoiesis and Iron MetabolismGuest Editors: Stefano Rivella, Elizabeta Nemeth, and Jeffery Lynn Miller

Advances in Hematology

Crosstalk between Erythropoiesis andIron Metabolism


Crosstalk between Erythropoiesis andIron Metabolism

Guest Editors: Stefano Rivella, Elizabeta Nemeth,and Jeffery Lynn Miller

Copyright © 2010 Hindawi Publishing Corporation. All rights reserved.

This is a special issue published in volume 2010 of “Advances in Hematology.” All articles are open access articles distributed under theCreative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided theoriginal work is properly cited.


Editorial Board

Camille N. Abboud, USARafat Abonour, USAMaher Albitar, USAKamran Alimoghaddam, IranAyad Al-Katib, USAGeorge F. Atweh, USAMichelle Baccarani, ItalyPeter Bader, GermanyMaria R. Baer, USASamir K. Ballas, USAMeral Beksac, TurkeyYazid Belkacemi, FranceLeif Bergsagel, USAIvan Bertoncello, AustraliaHenny H. Billett, USANeil Blumberg, USABrian Bolwell, USAKevin D. Bunting, USAFederico Caligaris-Cappio, ItalySuparno Chakrabarti, IndiaNelson J. Chao, USAFrancesco Dazzi, UKConnie J. Eaves, CanadaStefan Faderl, USAKenneth A. Foon, USAFrancine Foss, USAGo”sta Gahrton, SwedenVarsha Gandhi, USAArnold Ganser, GermanyGunther A. Gastl, AustriaAlan M. Gewirtz, USANicola Gokbuget, GermanyJohn M. Goldman, UKElvira Grandone, ItalyCharles S. Greenberg, USAPeter Greenberg, USAThomas G. Gross, USA

Michael L. Grossbard, USAZafer Gulbas, TurkeyThomas M. Habermann, USARobert G. Hawley, USADonna E. Hogge, CanadaDebra A. Hoppensteadt, USAPeter A. Jacobs, South AfricaSundar Jagannath, USAYuzuru Kanakura, JapanStefan Karlsson, SwedenSimon Karpatkin, USAThomas Kickler, USAMartin Klabusay, Czech RepublicVladimir Koza, Czech RepublicKarl-Anton Kreuzer, GermanyShaji Kumar, USAAbdullah Kutlar, USABoris Labar, CroatiaMyriam Labopin, FranceWing H. Leung, USAMark R. Litzow, USASagar Lonial, USAHeinz Ludwig, AustriaBashir A. Lwaleed, UKRita Maccario, ItalyEstella M. Matutes, UKOwen McCarty, USAJohn Meletis, GreeceRuben A. Mesa, USAEmili Montserrat, SpainJan S. Moreb, USASuneel D. Mundle, USAAndreas Neubauer, GermanyKenneth Nilsson, SwedenLuigi Daniele Notarangelo, USANiels Odum, DenmarkJohannes Oldenburg, Germany

Angela Panoskaltsis-Mortari, USAHelen A. Papadaki, GreeceLouis M. Pelus, USAChristian Peschel, GermanyPeter J. Quesenberry, USAMargaret V. Ragni, USAPranela Rameshwar, USAJohn Rasko, AustraliaPaolo Rebulla, ItalyLawrence Rice, USAJohn Roback, USAAldo M. Roccaro, USAJorge Enrique Romaguera, USAFrits R. Rosendaal, The NetherlandsJacob M. Rowe, IsraelGiuseppe G. Saglio, ItalyFelipe Samaniego, USAJesus Fernando San Miguel, SpainDavid C. Seldin, USAOrhan Sezer, GermanyJohn D. Shaughnessy, USAShimon Slavin, IsraelEdward F. Srour, USALuen Bik To, AustraliaKensei Tobinai, JapanKunihiro Tsukasaki, JapanJoseph M. Tuscano, USABenjamin Van Camp, BelgiumDavid Varon, IsraelDavid H. Vesole, USAGeorgia Vogelsang, USARichard Wells, CanadaJan Westin, SwedenJane N. Winter, USAEmanuele Zucca, Switzerland

Contents

Crosstalk between Erythropoiesis and Iron Metabolism, Stefano Rivella, Elizabeta Nemeth,and Jeffery Lynn MillerVolume 2010, Article ID 317095, 2 pages

Crosstalk between Iron Metabolism and Erythropoiesis, Huihui Li and Yelena Z. GinzburgVolume 2010, Article ID 605435, 12 pages

Targeting the Hepcidin-Ferroportin Axis in the Diagnosis and Treatment of Anemias, Elizabeta NemethVolume 2010, Article ID 750643, 9 pages

Ferroportin and Erythroid Cells: An Update, Luciano Cianetti, Marco Gabbianelli, and Nadia Maria SposiVolume 2010, Article ID 404173, 12 pages

Iron Loading and Overloading due to Ineffective Erythropoiesis, Toshihiko Tanno and Jeffery L. MillerVolume 2010, Article ID 358283, 8 pages

β-Thalassemia: HiJAKing Ineffective Erythropoiesis and Iron Overload, Luca Melchiori, Sara Gardenghi,and Stefano RivellaVolume 2010, Article ID 938640, 7 pages

Iron Chelation Therapy in Myelodysplastic Syndromes, Emanuela Messa, Daniela Cilloni,and Giuseppe SaglioVolume 2010, Article ID 756289, 8 pages

Red Blood Cell Transfusion Independence Following the Initiation of Iron Chelation Therapy inMyelodysplastic Syndrome, Maha A. Badawi, Linda M. Vickars, Jocelyn M. Chase, and Heather A. LeitchVolume 2010, Article ID 164045, 5 pages

Iron Overload in Sickle Cell Disease, Radha Raghupathy, Deepa Manwani, and Jane A. LittleVolume 2010, Article ID 272940, 9 pages

Diamond Blackfan Anemia at the Crossroad between Ribosome Biogenesis and Heme Metabolism,Deborah Chiabrando and Emanuela TolosanoVolume 2010, Article ID 790632, 8 pages

Unexplained Aspects of Anemia of Inflammation, Elizabeth A. Price and Stanley L. SchrierVolume 2010, Article ID 508739, 5 pages

Hindawi Publishing CorporationAdvances in HematologyVolume 2010, Article ID 317095, 2 pagesdoi:10.1155/2010/317095

Editorial

Crosstalk between Erythropoiesis and Iron Metabolism

Stefano Rivella,1 Elizabeta Nemeth,2 and Jeffery Lynn Miller3

1 Division of Hematology-Oncology, Department of Pediatrics, Weill Cornell medical College, New York, NY 10021, USA2 David Geffen School of Medicine at UCLA, CHS 37-131, 10833 Le Conte Avenue, Los Angeles, CA 90095-1690, USA3 Molecular Medicine Branch, NIDDK, NIH, Bethesda, MD 20892, USA

Correspondence should be addressed to Stefano Rivella, [email protected]

Received 14 July 2010; Accepted 14 July 2010

Copyright © 2010 Stefano Rivella et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Oxygen delivery is essential to sustain all vital functions.Vertebrate organisms have, therefore, evolved sophisticatedand tightly regulated mechanisms to ensure continuous redcell synthesis. Chronic or acute blood losses trigger multiplefeedback mechanisms aimed at increasing erythropoiesis andminimizing the damage associated with profound anemiaand hypoxia.

Iron is an element of paramount importance for ery-thropoiesis, being an essential part of the hemoglobinmetalloprotein that ensures oxygen absorption, transport,and delivery. Therefore, it is not surprising that ironabsorption and transport are closely linked to the demandsof erythropoiesis. However, the mechanisms underlying thisinterconnected relationship have only recently started beingunderstood.

The focus of this special issue is on iron metabolism,erythropoiesis, and their close association, with a specialemphasis on proteins regulating iron metabolism and howerythropoiesis affects their expression. The mechanistic viewwill be integrated with the clinical observations in diseasesuch as myelodysplastic syndromes, sickle cell anemia,Diamond-Blackfan anemia, and anemia of inflammation.And finally, in light of these new discoveries, new potentialtherapies for well-known disorders such as beta-thalassemiawill be discussed.

In the last decade, new discoveries have fueled the field ofiron metabolism. The papers by H. Li and Y. Z. Ginzburg[1],E. Nemeth [2], L. Cianetti et al. [3], and T. Tanno and J. L.Miller [4] provide a comprehensive and updated overview ofthe proteins involved in regulating dietary iron absorption,plasma iron concentrations, tissue iron distribution, anderythropoiesis. The review by H. Li and Y. Z. Ginzburg[1] introduces the mechanisms of crosstalk between iron

and erythropoiesis and their function in different diseases.E. Nemeth [2] describes the role of the peptide hormonehepcidin in iron physiology and in many iron-related dis-orders and highlights the potential of hepcidin agonists andantagonist as future therapeutic tools for iron disorders. Thepaper by L. Cianetti et al. [3] describes the hepcidin receptorand iron transporter ferroportin and its role during normaland pathological erythroid differentiation. The paper by T.Tanno and J. L. Miller [4] describes the relationship betweeniron regulation and ineffective erythropoiesis, focusing onpotential mechanisms for pathological iron overloading.

Improved knowledge on proteins that interconnect ironmetabolism and erythropoiesis is opening new prospectiveto develop new therapies, such as in beta-thalassemia.The review by L. Melchiori et al. [5] provides importantevidence that use of Jak2 inhibitors or hepcidin agonists mayameliorate the abnormal erythropoiesis and iron overload inβ-thalassemia.

The interdependence of erythropoiesis and iron reg-ulation extends beyond normal metabolism to includeseveral pathological conditions. Diseases that are primarilyassociated with erythroid defects often lead to iron overload,maldistribution, or deficiency. The reviews by E. Messa etal. [6] and by M. A. Badawi et al. [7] summarize thehistorical progress and recent clinical developments usingiron chelation in patients with myelodysplastic syndromes.Similarly, the paper by R. Raghupathy et al. [8] provides anoverview of transfusional iron overload patients with sicklecell disease, highlighting how to prevent iron overload in thisdisorder.

Diamond-Blackfan anemia is a rare congenital pure red-cell aplasia that presents during infancy. D. Chiabrandoand E. Tolosano’s [9] review focused upon heme and

2 Advances in Hematology

ribosomal biogenesis in this disease and why deficiency of theheme exporter Feline Leukemia Virus subgroup C Receptor(FLVCR1) may cause a related phenotype.

Finally, the review by E. A. Price and S. L. Schrier [10]provides a comprehensive and updated overview of anemiaof inflammation, focusing on many of key clinical issues thatremain to be clarified, such as understanding of mechanismsof anemia of inflammation in “noninflammatory” diseases,optimal methods for distinguishing anemia of inflammationfrom iron deficiency anemia, and understanding the contrib-utory role of various pathologic mechanisms in individualhuman diseases that lead to anemia of inflammation.

We hope that this special issue will stimulate curiosity,interest, and new research directions in exploring the fieldof iron and erythropoiesis, with the ultimate goal to developnew drugs and improve the management of disorders associ-ated with altered iron metabolism, abnormal erythropoiesis,and anemia of inflammation.

Stefano RivellaElizabeta Nemeth

Jeffery Lynn Miller

References

[1] H. Li and Y. Z. Ginzburg, “Crosstalk between iron metabolismand erythropoiesis,” Advances in Hematology, vol. 2010, ArticleID 605435, 12 pages, 2010.

[2] E. Nemeth, “Targeting the hepcidin-ferroportin axis in thediagnosis and treatment of anemias,” Advances in Hematology,vol. 2010, Article ID 750643, 9 pages, 2010.

[3] L. Cianetti, M. Gabbianelli, and N. M. Sposi, “Ferroportin anderythroid cells: an update,” Advances in Hematology, vol. 2010,Article ID 404173, 12 pages, 2010.

[4] T. Tanno and J. L. Miller, “Iron loading and overloading due toineffective erythropoiesis,” Advances in Hematology, vol. 2010,Article ID 358283, 8 pages, 2010.

[5] L. Melchiori, et al., “β-thalassemia: hiJAKing ineffectiveerythropoiesis and iron overload,” Advances in Hematology,vol. 2010, Article ID 938640, 7 pages, 2010.

[6] E. Messa, D. Cilloni, and G. Saglio, “Iron chelation therapyin myelodysplastic syndromes,” Advances in Hematology, vol.2010, Article ID 756289, 8 pages, 2010.

[7] M. A. Badawi, L. M. Vickars, J. M. Chase, and H. A.Leitch, “Red blood cell transfusion independence followingthe initiation of iron chelation therapy in myelodysplasticsyndrome,” Advances in Hematology, vol. 2010, Article ID164045, 5 pages, 2010.

[8] R. Raghupathy, D. Manwani, and J. A. Little, “Iron overload insickle cell disease,” Advances in Hematology, Article ID 272940,9 pages, 2010.

[9] D. Chiabrando and E. Tolosano, “Diamond blackfan anemiaat the crossroad between ribosome biogenesis and hememetabolism,” Advances in Hematology, vol. 2010, Article ID790632, 8 pages, 2010.

[10] E. A. Price and S. L. Schrier, “Unexplained aspects of anemiaof inflammation,” Advances in Hematology, vol. 2010, ArticleID 508739, 5 pages, 2010.


Review Article

Crosstalk between Iron Metabolism and Erythropoiesis

Huihui Li and Yelena Z. Ginzburg

Lindsley F. Kimball Research Institute, New York Blood Center, 310 East 67th Street, 1–38, NY 10065, USA

Correspondence should be addressed to Yelena Z. Ginzburg, [email protected]

Received 21 January 2010; Accepted 25 March 2010

Academic Editor: Stefano Rivella

Copyright © 2010 H. Li and Y. Z. Ginzburg. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Iron metabolism and erythropoiesis are inextricably linked. The majority of iron extracted from circulation daily is used forhemoglobin synthesis. In the last 15 years, major advances have been made in understanding the pathways regulating ironmetabolism. Hepcidin is a key regulator of iron absorption and recycling and is itself regulated by erythropoiesis. Whileseveral viable candidates have been proposed, elucidating the “erythroid regulator” of hepcidin continues to generate significantexperimental activity in the field. Although the mechanism responsible for sensing iron demand for erythropoiesis is stillincompletely understood, evaluating diseases in which disordered erythropoiesis and/or iron metabolism are showcased hasresulted in a more robust appreciation of potential candidates coordinated erythroid iron demand with regulators of iron supply.We present data drawn from four different conditions—iron deficiency, congenital hypotransferrinemia, beta-thalassemia, andhereditary hemochromatosis—both in human and non-human models of disease, together suggesting that erythroid iron demandexerts a stronger influence on circulating iron supply than systemic iron stores. Greater understanding of the interplay betweenthe key factors involved in the regulation of iron metabolism and erythropoiesis will help develop more effective therapies fordisorders of iron overload, iron deficiency, and hemoglobin synthesis.

1. Introduction

Iron is an essential element for almost all living organisms,from mammals and lower vertebrates down to unicellu-lar organisms. It forms the core of molecules such ashemoglobin and myoglobin and is necessary for cytochromeproduction. Two to three million red blood cells (RBCs)are produced every second and require 30–40 mg of irondelivered to the erythron to make 30 pg of hemoglobin percell, a total of 6 g of hemoglobin daily. The pool of ironbound to transferrin [Tf-Fe(III)] is 10 times smaller thanthe daily iron requirements, requiring rapid turn aroundto ensure sufficient delivery of iron. Daily iron required forerythropoiesis is predominantly derived from recycling ofheme iron by macrophages erythrophagocytosing senescentRBCs. Because the majority of iron in most organisms canbe found in the hemoglobin compartment, erythropoiesisdominates iron metabolism and the two are inextricablyintertwined. This dynamic process of iron trafficking forerythropoiesis requires significant crosstalk to prevent irondeficiency or iron overload and protect the organism from

developing anemia as well as from the potential toxicity ofexcess iron. For instance, iron absorption increases, oftendramatically, when erythropoiesis occurs at a higher thannormal rate to accommodate the higher iron demand.Chronic blood loss, for example, results in both a maximalstimulation of iron absorption and less iron per cell leadingto smaller RBCs with less hemoglobin before progressingto decreased RBC counts and anemia. Conversely, diseasestates of excess iron are often associated with expanded RBCsize and higher cellular hemoglobin concentrations as a wayof sequestering iron into a nontoxic compartment. Lastly,diseases in which anemia and excess iron coexist exhibitcomplicated regulation schema that are still incompletelyunderstood.

Iron directed to the erythroid compartment is restrictedto transferrin-bound iron (Tf-Fe(III)) and its ability to bindtransferrin receptor (TfR1) is a well worked out paradigm.Transferrin is the second most abundant serum protein, afteralbumin, and takes up iron from duodenal enterocytes whenit is absorbed and from macrophages when iron is recycledfrom senescent RBCs. Iron absorption and recycling are


regulated by hepcidin, a peptide hormone thought to be themain regulator of iron flows in the body. Hepcidin exerts itsfunction by binding to the only known iron export protein,ferroportin (FPN-1), found on hepatocytes, macrophages,duodenal enterocytes, and placental cells, all involved in ironmetabolism. Hepcidin binds FPN-1, causes its internaliza-tion and degradation, and results in cessation of iron releasefrom cells [1]. Regulation of hepcidin has been extensivelystudied in recent years. It is known that hepcidin is regulatedby iron, hypoxia, inflammation, and erythropoiesis. Whileevidence exists for an “erythroid regulator” of hepcidin, themechanisms by which this is accomplished are still underinvestigation.

Understanding the crosstalk between iron regulationand erythroid proliferation and maturation is a hotbed ofresearch activity by multiple groups from many differentangles. Much can be learned from various disease statesand experimental models of disease. This review attemptsto catalog some of these for the purpose of elucidating thecurrent state of knowledge on this subject.

2. Iron Metabolism, Iron Deficiency,and Anemia

Iron deficiency anemia (IDA) is one of the most com-mon diseases worldwide and is typically associated with amicrocytosis and hypochromasia. The presence of small (lowMCV) and pale (low MCH) RBCs is typically indicativeof low cellular hemoglobin and results from defects inheme or globin synthesis. While defects in globin synthesisare typically caused by genetic defects (e.g., α- and β-thalassemia), heme synthesis defects most commonly resultfrom iron deficiency. In the pediatric population, this latermanifestation results from nutritional iron deficiency, whilein adults, the cause is more likely blood loss. As a betterunderstanding of erythropoiesis evolved in the middle ofthe last century, the ease of diagnosing and treating IDAresulted in a great deal of comfort on the part of clinicians.However, there are several clinical scenarios that showcasethe complex relationship between iron supply, iron demand,and erythropoiesis.

The amount of iron delivered to each erythroid precursordepends on the amount of monoferric and diferric trans-ferrin found in circulation as well as the density of TfR1on the cell surface. Typically, each erythroid precursor hasover a million TfR1s on its membrane because of its largeiron requirement—for hemoglobin synthesis—relative to allother cells in the body [2]. In IDA, TfR1 membrane densityincreases further [3, 4] and drives up the concentration ofsoluble TfR1. Soluble TfR1 is normally found in the serum asa truncated version of membrane-bound TfR1 in quantitiesproportional to the amount found on the cell surface [5, 6].Increased density of membrane-bound TfR1 enables the cellto increase avidity for iron during iron deficiency. IncreasedTfR1 and soluble TfR1 levels can also be found during stresserythropoiesis when the normally ample amount of systemiciron may require supplementation to enable a higher rate ofheme synthesis.

The mechanism by which cells alter their TfR1 expressioninvolves iron regulatory proteins (IRPs) which have a highaffinity for iron response elements (IREs) present in themRNA of target genes involved in iron homeostatis. If anIRE is on the 5′ untranslated region, mRNA is more likelyto be degraded whereas if it is on the 3′ untranslatedregion, it is more likely to be stabilized as a consequenceof IRP binding. Thus, in an iron depleted state, whenIRPs are able to bind mRNA, ferritin mRNA with a 5′

untranslated region IRE is more likely to be degraded whileTfR1 mRNA with a 3′ untranslated region IRE proceedsto translation more readily. IRP2 knockout mice developmicrocytic hypochromic anemia probably as a result of thereduced TfR1 expression in erythroid precursors [7, 8].As a consequence of IRE/IRP regulation, iron deficiencyis associated with high TfR1 expression and low ferritinconcentrations [9].

Iron uptake starts when Tf-Fe(III) binds TfR1 (Figure 1).Under normal circumstances, the affinity of TfR1 for diferrictransferrin is greater than for monoferric transferrin. How-ever, this greater affinity wanes as the iron supply is dimin-ished [10, 11]. Monoferric transferrin is the predominantform of transferrin in circulation when transferrin saturationis lowered [12]. The loss of preference for diferric transferrinmay itself be a result of a proportionally higher concentrationof monoferric transferrin, and relatively fewer diferric trans-ferrin molecules, in circulation. Each molecule of monoferrictransferrin delivers less iron to erythroid precursors thandiferric transferring [10]. This enables a greater numberof erythroid precursors to receive a smaller portion of theiron pool to offset the potential of developing anemia. Thisfinding is consistent with the fact that MCV drops beforehemoglobin decreases during the progressive worsening ofiron deficiency. What controls this apportioning of iron toerythroid precursors is not completely understood.

Iron gains entry into erythroid precursors when Tf-Fe(III) complexes with TfR1 and is internalized into aclathrin-coated pit which matures into an endosome with alowered pH, facilitating the release of iron from transferrinand iron transport across the membrane by DMT1, a proton-coupled divalent metal transporter [13]. Iron-depleted trans-ferrin (apotransferrin) is then recycled back to the cellsurface where the restored physiologic pH 7.4 results in thedissociation of apotransferrin from TfR1. The entire cycle iscompleted within minutes and recurs 100–200 times in theduration of a single transferrin molecules life cycle in thebody [14].

Iron in erythroid precursors is then transported to themitochondria for heme synthesis. Heme itself functions as atranscriptional regulator. It can induce heme oxygenase 1, amolecule which reciprocally induces heme degradation [15].Heme not participating in hemoglobin synthesis results ina downregulation of IRP2 which reduces TfR1 expressionon the cell surface and thus the amount of iron enteringcells [16]. Both of these functions prevent excess heme fromaccumulating in erythroid precursors. In order to preventexcess globin synthesis, heme deficiency represses globinsynthesis by activating the heme regulated inhibitor whichphosphorylates eIF2a and prevents the conversion of GTP

Advances in Hematology 3

Ferritin

Heme Globin

Hemoglobin

Heme

Hepcidin(normal)Normal

+

(a)

Ferritin

Heme Globin

HemoglobinHeme

Hepcidin(low)Iron deficiency anemia

? +

(b)

Ferritin

Heme Globin

HemoglobinHeme

Hepcidin(low)Congenital hypotransferrinemia

+?

(c)

Ferritin

HemeGlobin

HemoglobinHeme

Hepcidin(low)β-thalassemia

? +

(d)

Ferritin

Heme Globin

Hemoglobin

Heme

Hepcidin(low)Hereditary hemochromatosis

Transferrin receptor 1

Transferrin

Iron

Ferroportin-1B

FLVCR

? +

(e)

Figure 1: Model relationship between iron delivery, relative abilities to synthesize heme and globin, and heme and iron export in variousdiseases associated with concurrent iron and erythroid pathology.

from GDP, shutting down mRNA translation at the β-globinlocus control region [17–19]. ALAS2, the first enzyme in theheme synthesis pathway, also has a 5′ untranslated regionIRE, resulting in the reduction in heme synthesis in an irondeficient state. Heme itself can also repress ALAS synthesis inthe liver [20]. Interestingly, during terminal differentiation oferythroid precursors, the regulation of ferritin and ALAS2 isdisparate, with significant impairment of ferritin expressionand undeterred ALAS2 translation [21]. The mechanismregulating this effect appears to preferentially channel ironfor heme synthesis in erythroid precursors. These negativefeedback mechanisms and reciprocal regulation of hemeand iron transport during erythroid precursor regulation isessential to maintain all components in proper balance.

Only a single molecule, FPN-1, has been identified tofunction as an iron exporter. FPN-1 is regulated by hepcidin.Additionally, FPN-1 mRNA has IREs on the 5′ untranslatedregion [22–24]. In iron deficiency, these two regulatorymechanisms predict opposite results; low hepcidin drivingFPN-1 up but IRP binding to the IRE on the 5′ untranslatedregion would be expected to drive FPN-1 expression down.

Hepcidin levels are low in iron deficiency while FPN-1on the basolateral membranes of duodenal enterocytesis increased to absorb and transfer iron to circulatingtransferrin [25–27]. Because increased FPN-1 is observed iniron deficient mice, a non-IRE containing transcript of FPN-1 was hypothesized to enable iron absorption to proceedin the iron deficient state. Recently, an additional FPN-1(FPN-1B) was discovered which lacks an IRE and is notsubject to repression by IRP in iron deficiency [28]. FPN-1B is generated from an alternative upstream promoter andaccounted for the increased expression of FPN-1 in theduodena of iron deficient mice. Furthermore, FPN-1B wasalso identified on erythroid precursors [28]. Why these cellsneed to export iron is not completely understood but it mayenable erythroid precursors to sense systemic iron status andallow these precursors to respond to hepcidin levels. FPN-1Bexpression is diminished during the later stages of erythroiddifferentiation, that is once the erythroid precursor beginsto produce hemoglobin, and possibly results in lower MCVand MCH during iron deficiency when hepcidin levels arelow.


Heme export has also recently been demonstrated inerythroid precursors. Based on data in cats, the functionof feline leukemia virus, subgroup C receptor (FLVCR)was reported to be a heme exporter [29]. In infected cats,FLVC envelope protein induced blockade of FLVCR, leadingto pure red cell aplasia due to a block of differentiationat the CFU-E or proerythroblast stage and resulting inreticulocytopenia and anemia [30, 31]. Studies show thatall bone marrow cells are infected but the infection onlyresults in an erythroid lineage phenotype, implying thatFLVCR is uniquely important in erythroid development.FLVCR overexpression in mice results in a mild microcytichypochromic anemia, suggesting that, since hypochromasiaand microcytosis only result from heme or hemoglobindeficiency, FLVCR is needed to maintain heme and globinbalance and avoid accumulation of free heme or excess globinin the cytoplasm [32]. The absence of FLVCR results inproerythroblast differentiation arrest and apoptosis, likelydue to heme toxicity, and points to the significance ofexport of excess iron and iron-containing compounds. Whathappens to FLVCR during iron deficiency is yet to bedetermined but hepcidin expression in Flvcr−/− mice isincreased with associated decrease in FPN-1 and a higherconcentration of iron in the cell.

Although hepcidin expression is expected to be low iniron deficiency, some clinical situations are more complex.Most recently, studies involving both human subjects (withiron refractory, iron deficiency anemia, or IRIDA) as well asmice (mask phenotype) present a situation in which highlevels of hepcidin expression and iron deficiency coexist.These situations enabled scientists to uncover the role ofmembrane-bound serine protease type 6 (TMPRSS6) in theregulation of hepcidin during iron deficiency. TMPRSS6 isexpressed in the liver. Mask mice have microcytic anemiadue to iron deficiency caused by decreased iron absorptionfrom high hepcidin levels. Positional cloning experimentsuncovered the splicing error in Tmprss6 [33]. Similarly,patients with IRIDA have an autosomal recessive mutationin the gene encoding TMPRSS6 resulting in hypochromicmicrocytic anemia, low serum transferrin saturation, andinappropriately elevated hepcidin concentrations [34–37].Recent studies suggest that TMPRSS6 normally acts to down-regulate hepcidin expression by cleaving membrane-boundhemojuvelin (HJV) [37]. (See page 8 for a more extensivediscussion of the regulation of hepcidin expression.)

3. Human Congenital Hypotransferrinemia

Transferrin is the main serum iron transporter in allvertebrates; it takes up iron from duodenal enterocyteswhere iron is absorbed and from macrophages when ironis recycled from senescent RBCs and delivers it to cellsby binding TfR1. Congenital hypotransferrinemia is a rarehereditary disease characterized by a severe deficiency ofserum transferrin. The defect results in iron deficient ery-thropoiesis and hypochromic anemia as well as iron overloadin nonhematopoietic tissues (Figure 1). The excess iron isunavailable for erythropoiesis in this disease, suggesting thatTf-Fe(III) uptake by TfR1 is the only known means of iron

delivery for erythropoiesis [38]. Fewer than a dozen caseshave been described in the literature [39–44]. By analyzingcase reports of this disease, we have learned that transferrinproduction occurs in excess of that required for normalerythropoiesis for which a minimum of only 10–20 mg/dL oftransferrin is required [41], 20-fold less than levels typicallyfound in circulation in normal individuals. The rarity ofthis disease in addition to the excess transferrin producedin normal individuals underscore the essential nature of thiscompound for normal physiologic function.

Treatment for individuals diagnosed with congenitalhypotransferrinemia involves the use of recurrent plasmainfusions, plasma-derived human transferrin, or recombi-nant human transferrin. A case study reporting (1-2 g)transferrin infusions every 3-4 months resulted in patientimprovement [39, 41]. This case demonstrates that effec-tiveness of a single dose of transferrin is between 4 and9 months [41] despite a half life of only 7.6 days [14]. Inanother case, monthly plasma infusions provided sufficienttransferrin to maintain a hemoglobin of 12 g/dL and enabledphlebotomy as simultaneous therapy for iron overload [42].In this case, reticulocytosis was observed within 10–14 daysof infusion followed by a rise in hemoglobin. Anothercase study reported the treatment of a boy with congenitalhypotransferrinemia using monthly plasma infusion whichresulted in a normal hemoglobin concentration [43, 44].This last case provided amounts of plasma and patienthematocrit and enabled us to calculate some specific values.

For example, if 300 mL of plasma was transfused, andplasma contains an average transferrin concentration of200 mg/dL = 2 mg/mL, then 300 mL × 2 mg/mL = 600 mgtransferrin per 300 mL plasma infusion. If total bloodvolume is approximately 5 L and hematocrit is roughly 30%(in an anemic patient), then plasma volume = 5000 mL ×0.7 = 3500 mL. 600 mg transferrin transfused into a volumeof 3500 mL = 600 mg/35 dL = 17 mg/dL. With a t1/2of 7.6 days, the end of 1 week results in <10 mg/dL oftransferrin, the minimum concentration required to sustainerythropoiesis.

Because the circulating levels after one week are belowminimum transferrin requirements for erythropoiesis, theefficacy of monthly plasma infusions implies that the amountof transferrin required for steady-state erythropoiesis is wellbelow levels needed during stress erythropoiesis. In fact, theonset of a growth spurt at age 10 was associated with a dropin hemoglobin and an increase in reticulocytosis. The patientwas treated with an increased rate of plasma infusions—weekly doses for two months—which lead to a reboundof hemoglobin and increased serum transferrin concentra-tion approximately 3- to 5-fold from previously. Hepcidinexcretion was negligible and increased over the course oftwo months of accelerated plasma infusions, suggesting thaterythropoietic drive, as evidenced by worsening anemia andreticulocytosis, resulted in hepcidin suppression which wasrelieved by increasing iron supply to meet the demand ofaccelerated erythropoiesis [45].

A hypotransferrinemic (hpx/hpx) mouse model isdescribed and characterized in the literature [46, 47]resulting from a splicing defect in the mouse transferrin


gene on chromosome 9 [48]. Approximately 1% of normaltransferrin protein is found in circulation in hpx/hpx mice[47]. These mice exhibit hypochromic microcytic anemia,low transferrin levels, severe growth retardation and a robustresponse to mouse plasma or purified transferrin injections,a pattern of characteristics reminiscent of human congenitalhypotransferrinemia. Little iron is found in the bone marrowand spleen but massive iron overload develops in non-hematopoietic tissues such as the liver, heart, endocrineorgans, and kidney [46, 47].

Although Tf-Fe(III) uptake occurs via binding to TfR1 inerythroid precursors (as well as other cell types), nontrans-ferrin bound iron (NTBI), as in many diseases of iron over-load, predominantly results in parenchymal iron depositionin non-hematopoietic cells. When transferrin levels are low,as in hpx/hpx mice, complete transferrin saturation occursquickly. Once transferrin saturation approaches 100% andits iron binding capacity is exceeded, labile plasma iron (LPI)can be found in circulation [49]. LPI is a redox active formof NTBI which is taken up by cells in a dysregulated manner,can cause free radical damage resulting in the morbidity andmortality of iron overload diseases, and is unavailable forerythropoiesis. LPI is the presumed cause of iron overloadin non-hematopoietic tissues in hpx/hpx mice.

Transferrin replacement in hpx/hpx mice relieves anemia,decreases parenchymal iron deposition in the liver, andreduces the high rate of iron absorption in the duodenum[50]. When transferrin-treated hpx/hpx mice are analyzedafter transferrin concentration returns to undetectable lev-els, iron absorption increases despite persistent normalhemoglobin levels and decreased liver nonheme iron stores.This finding implies that transferrin concentration (or per-haps by extension transferrin saturation) itself has an effecton iron absorption, independent of the effect of hemoglobinconcentration [50]. In fact, hepcidin expression, now knownto control iron absorption, is under the regulation of Tf-Fe(III) [51] and hepcidin expression is low in hpx/hpx mice[52], resulting in increased iron absorption. In hpx/hpxmice, as in β-thalassemic mice, hepcidin levels do not reflectthe degree of systemic iron overload, suggesting that acompeting signal is counter-regulating hepcidin expressionand provides further evidence for the existence of an“erythroid regulator” of hepcidin.

The authors suggest that one week after transferrininjections, when transferrin levels have returned to baselinein hpx/hpx mice, “in spite of normalized hemoglobinlevels. . .low transferrin levels lead to ineffective (iron-deficient) erythropoiesis.” [50] In fact, iron deficiency itself isassociated with an increased rate of erythropoiesis and a pre-ponderance of erythroid precursors that become quiescentwithout completing the maturation cycle [53]. In addition,hpx/hpx mice exhibit extramedulary erythropoiesis notablyin the liver [47] and splenomegaly [46], findings reminis-cent of β-thalassemia, a disease characterized by anemia,iron overload, and ineffective erythropoiesis. Hpx/hpx miceabsorb somewhat more iron relative to β-thalassemic mice,which have a comparable anemia and reticulocytosis [54],suggesting that not the hemoglobin levels but the amountof transferrin itself, relative to the degree of erythropoietic

demand, may influence hepcidin expression. In supportof this, when hpx/hpx mice are transfused to suppressendogenous erythropoesis, the degree of iron absorptionapproached normal [55] possibly by reducing the pressureon Tf-Fe(III) delivery for erythropoiesis [50]. While furtherexperimentation is necessary, these findings suggest thatiron deficiency results in ineffective erythropoiesis as aconsequence of relatively insufficient circulating transferrinto accommodate the degree of erythropoiesis [56].

4. β-thalassemia

β-thalassemias are caused by mutations in the β-globingene resulting in reduced or absent β-chain synthesis. Arelative excess of α-globin chain synthesis leads to increasederythroid precursor apoptosis, causing ineffective erythro-poiesis which in turn results in extramedullary expansionand splenomegaly. Together with shortened RBC survival,these abnormalities result in anemia. The clinical phenotypeis heterogeneous due to genotypically different mutations,combination inheritance with hemoglobinopathies, andadditional modifying factors. Patients with β-thalassemiamajor, the most severe form of β-thalassemia, exhibit verylimited synthesis of β-globin and require life-long RBC trans-fusions to ameliorate anemia and suppress extramedullaryerythropoiesis. Without transfusions, expanded erythro-poiesis results in hepatosplenomegaly and bone deformitiesdue to expansion of the intraosseous marrow compartment.Patients with β-thalassemia intermedia show a milder clinicalpicture with more β-globin chains synthesized and requireonly intermittent transfusions. Although chronic transfusiontherapy has been standard practice for the last half-century,advances in understanding the mechanisms of this diseasehave lead to some interesting findings and promise tocontinue evolving therapy in β-thalassemia.

Patients with β-thalassemia have increased intestinal ironabsorption which, in addition to transfusion dependence,contributes to iron overload. If left untreated, iron overloadresults in progressive iron deposition, leading to multipleorgan dysfunction and accounts for the majority of deathsin this disease. LPI has been proposed as the cause of thismorbidity. As mentioned previously, because no physiologicmeans of iron excretion exist, increased iron absorptionand RBC turn over from transfusion lead to saturation ofplasma transferrin. When transferrin iron binding capacity isexceeded and NTBI is detectable in circulation, iron traffick-ing is dysregulated and results in the clinical manifestationsof iron overload. Despite iron overload in patients with β-thalassemia, hepcidin levels are not increased.

Insufficient hepcidin expression, relative to the degree ofiron overload, is implicated as the cause of iron overloadobserved in β-thalassemia. We and others have demon-strated that hepcidin expression is relatively low for thedegree of liver iron stores in untreated β-thalassemic micerelative to controls [57, 58]. Because hepcidin functions bybinding FPN-1 and preventing iron release from duodenalenterocytes and macrophages, increased hepcidin levels areexpected in diseases of iron overload to prevent continuediron absorption and release from stored recycled iron. In


order to analyze the variability in hepcidin levels in β-thalassemic patients, serum samples were collected frompatients and demonstrated that hepcidin levels increasein those with higher hemoglobin and decrease in thosewith high serum TfR1 levels, increased erythropoietin,and detectable NTBI [59]. These findings are consistentwith other reports that correlate increased erythropoieticactivity with hepcidin suppression as well as the response toincreased hepcidin levels following transfusion.

Several researchers have assessed the erythropoieticregulation of hepcidin from a clinical and basic scienceperspective. For example, transfusions in β-thalassemiamajor patients, used to suppress endogenous erythropoiesis,resulted in an increase in urinary hepcidin concentrations[60]. Furthermore, the exposure of HepG2 cells to sera frompatients with β-thalassemia major after transfusion resultedin higher hepcidin levels relative to the cells exposed to sera ofthe same patients prior to the next transfusion [61, 62]. Thesefindings suggest that hepcidin suppression in β-thalassemicpatients results from the secretion of a soluble factor, theconcentration of which is proportional to the degree oferythroid activity. This “erythroid regulator” of hepcidin hasnot yet been elucidated but is of great importance in diseasesin which anemia and iron overload coexist. This signal isstronger than the regulation of hepcidin by iron and leadsto the exacerbation of iron overload, the very complicationassociated with clinical deterioration and mortality in β-thalassemia.

Hepcidin suppression in diseases of iron overload withineffective erythropoiesis exacerbates the degree of ironoverload by increasing iron absorption. This excess irondeposits in the parenchyma of non-hematopoietic tissue.Because further iron absorption ultimately exceeds trans-ferrin iron-carrying capacity, suppressed hepcidin results inthe formation of NTBI which is not available for erythro-poiesis. Prior experiments demonstrate that phlebotomy,erythropoietin administration, and hemolysis resulted indecreased hepcidin expression [26, 63, 64]. To examine themechanism of hepcidin suppression in these conditions,separating the effect of erythropoiesis from that of anemiaand iron stores is important. Ablation of erythropoiesisaccomplished by chemotherapeutic agents, radiation, anderythropoietin-blocking antibodies prevents hepcidin sup-pression in response to hemolysis, bleeding, or epogeninjection [64, 65]. Although ablation of erythropoiesis resultsin increased hepcidin expression, compensated hemolysis(without ablation of erythropoiesis) does not affect hepcidinexpression despite an equivalent degree of anemia andhepatic iron deposition [64]. These studies further demon-strate that iron requirements for erythropoiesis influencehepcidin expression to a greater degree than anemia or non-hematopoietic iron stores in the body.

New evidence regarding the influence of erythropoiesison hepcidin levels involves growth differentiation factor15 (GDF15) and twisted gastrulation 1 (TWSG1) in β-thalassemic patients and mice, respectively [66, 67]. Thesefactors have recently been identified as possible solubleerythroid factors that regulate hepcidin [67]. GDF15 andTWSG1 levels are significantly elevated in β-thalassemic

patients and mice, respectively. Sera from β-thalassemicpatients with upregulated GDF15 suppress hepcidin mRNAexpression in primary human hepatocytes, and depletion ofGDF15 reverses hepcidin suppression. These findings suggestthat increased serum GDF15 may be the factor throughwhich ineffective erythropoiesis and/or increased erythroidprecursor apoptosis influence hepcidin expression and ironhomeostasis in β-thalassemic patients [66]. Alternatively,TWSG1, a protein synthesized at early stage of erythro-poiesis, shows an indirect effect on hepcidin suppression[67]. It is possible that ineffective erythropoiesis in β-thalassemia modifies erythroid parameters such as GDF15and TWSG1 which, together or through several differentsignaling pathways, induce inappropriate hepcidin inhibitionand maldistribution of iron.

Our laboratory hypothesized that this insufficient hep-cidin secretion and maldistribution of iron in β-thalassemiamay result from inadequate circulating transferrin to deliveriron for erythropoiesis. This hypothesis is informed bypreliminary experiments in Hbbth1/th1 mice, a model of β-thalassemia intermedia, that demonstrates low non-hemeiron in the bone marrow relative to control mice [57].High-dose iron dextran in these mice results in a dose-dependent increase in extramedullary erythropoiesis. Thelack of medullary erythroid response to iron suggests thatincreased transferrin concentration may be necessary toaccommodate the degree of erythoid expansion observed inβ-thalassemia. Chronic treatment with transferrin injectionsin Hbbth1/th1 mice results in increased hemoglobin pro-duction, decreased reticulocytosis and erythropoietin levels,reverses splenomegaly, and elevates hepcidin expression[56]. Transferrin injections also change the proportion oferythroid precursors to more mature relative to immatureprecursors, lower the rates of apoptosis in mature erythroidprecursors, and reduce the amount of extramedullary ery-thropoiesis in the liver and spleen in Hbbth1/th1 mice. Thesefindings imply that exogenous transferrin results in moreiron delivery for erythropoiesis and restores more normal(less ineffective) erythropoiesis as a result.

How additional transferrin is able to improve erythro-poiesis is incompletely understood. One hypothesis mayinvolve the observation that although the hemoglobin con-centration and the number of RBCs increase after transferrininjections, RBCs are smaller and contain less hemoglobin.Similar phenotype of decreased MCV and MCH with normalhemoglobin values due to increased number of RBCs isobserved in T f R+/− mice [68]. This lower MCV, both inour transferrin-treated Hbbth1/th1 mice and T f R+/− mice,is reminiscent of iron deficient erythropoiesis in normalsubjects treated with recombinant erythropoietin [69]. This“functional iron deficiency,” a term used to describe statesof iron restricted erythropoiesis induced by exogenouserythropoietin, applies also to states of ineffective erythro-poiesis with elevated endogenous erythropoietin [70]. Theunderlying concept, that the rate of iron supply is insufficientto meet the demands of erythropoiesis, applies to othercircumstances such as expanded erythropoiesis observed inβ-thalassemia. The finding of protoporphyrin in RBCs ofpatients with β-thalassemia, in whom transferrin saturation


Table 1: Characteristics of Hereditary Hemochromatosis.

Gene Mutation Inheritance Hepcidin levels Pathophysiology

Type I HFE C282Y (6p21), H63D AR Low Increased iron absorption

Type II –Juvenilehemochromatosis

HJV; hepcidin 1q21; 19q13 AR Low Increased iron absorption

Type III TFR2 7q22 AR Low Increased iron absorption

Type IV∗ FPN1 2q32 AD High Increased iron absorption∗Mutation results in 2 similar forms of disease: either a hemochromatosis-like illness with increased iron in hepatocytes due to hepcidin resistance (gain-of-function mutation) or reduced macrophage iron export with normal transferrin saturation called “classic ferroportin disease” (loss-of-function mutation).

is an ample 40%, provides evidence that this is the case[71]. In fact, old literature estimated that the daily ironrequirement in β-thalassemia may be as high as 150 mg,values which could make iron demand by the expandederythron greater than its available supply [72] and possiblytrigger hepcidin suppression in order to stimulate an increasein iron absorption.

5. Hereditary Hemochromatosis

Hereditary hemochromatosis (HH) is a genetically inheriteddisorder of iron metabolism. Although gene frequency isas high as 5–7%, low penetrance results in only 1:300 to1:400 affected individuals. Four types of disorders exist,all of which result in increased intestinal iron absorp-tion as a consequence of inadequate hepcidin, or hep-cidin insensitivity, relative to the degree of systemic iron(Table 1).

The most common type of HH, Type I HH, is char-acterized by a mutation in the HFE gene. HomozygousC282Y mutation, which accounts for more than 80% HHpatients [73], disturbs formation of a disulfide bond in the α3domain of HFE, prevents its binging to β2-microglobulin—a protein involved in the regulation of iron absorption, andmarkedly reduces the appearance of HFE on the cell surface.Disease is occasionally present as a compound heterozygotemutation with a second synergistic H63D mutation whichalso leads to the reduction of cell surface HFE [74]. Althoughthe molecular function of HFE remains uncertain, recentevidence points to its role in hepcidin regulation [75].Because hepcidin is believed to be the central regulatorof iron flows in the body, the involvement of HFE inhepcidin regulation is consistent with its presumed functionin iron absorption. In the recent past, major advances havebeen made in understanding the molecular mechanism ofhepcidin regulation. The bone morphogenic protein (BMP)pathway—involved in cell proliferation, differentiation, andapoptosis—has been identified as a critical regulator ofhepcidin expression [76, 77]. BMP receptor activation resultsin Smad phosphorylation which translocates to the nucleusand activates transcription of target genes [78] (Figure 2).Hemojuvelin (HJV) was identified as an iron-specific BMPcoreceptor and stimulant of the BMP pathway in ironoverload states [79]. Another form of HJV—soluble HJV(sHJV)—acts as a BMP antagonist [80] and leads to reducedhepcidin expression [77] as a means of negative feedback(Figure 2).

Expression of hepcidin is inappropriately low in patientswith Type I HH and H f e−/− mice [81–83]. HFE is one ofseveral membrane proteins involved in communicating sys-temic iron status to the hepatocyte and affects HJV/BMP sig-naling pathway to positively influence hepcidin production(Figure 2). Suppressed HFE levels in HH prevent appropriatesensing and result in a dampened hepcidin response to ironload, consequent increased iron absorption and increasedtransferrin saturation.H f e(−/− ) mice confirm the associ-ation of HFE gene mutation, iron overload, and increasederythropoiesis. Compared to wild type, H f e−/− mice exhibitmassive iron deposition and elevated transferrin saturation[84]. The role of HFE in influencing iron absorption is notcompletely understood. HFE associates with TfR1 under lowiron conditions [85] and is displaced when TfR1 binds Tf-Fe(III) [86–88] (Figure 2). Unlike TfR1, TfR2 lacks an IREand is thus not regulated by levels of plasma iron. As serumiron concentration increases, TfR2 expression exceeds that ofTfR1 and Tf-Fe(III) binds both TfR1 and TfR2, increasingTfR2 stability [89, 90] on the membrane and induces HFEbinding to TfR2. This HFE/TfR2 complex interacts withHJV, the iron-specific BMP co-receptor, and potentiates theBMP signaling pathway and hepcidin transcription [91](Figure 2).

Thus, both TfR2 and HFE/TfR1 complex function as themain Tf-Fe(III) sensors [90, 92] and communicate systemiciron status to the hepatocyte resulting in altered hepcidinsecretion. As expected, iron overload diseases are alsoassociated with mutation in genes coding these intermediaryproteins. Type II HH results from HJV or hepcidin muta-tions. As those with HFE mutations, patients with mutatedHJV/HAMP also exhibit enhanced iron absorption and rapidiron accumulation at a young age [93]. HJV mutations, likemutations in hepcidin itself, result in nearly absent hepcidinexpression and clinically result in the most severe formof HH termed Juvenile Hemochromatosis. H jv−/− micedevelop a substantial decrease in hepcidin expression withconcurrently increased FPN-1 expression and early onsetiron deposition [94]. Type III HH results from a mutation inthe gene encoding TfR2. TfR2 depletion leads to hepatic irondeposition and decreased hepcidin expression in zebrafishembryos [95]; mice models exhibit elevated ferritin andtransferrin saturation in addition to these features [96].

Lastly, type IV or FPN-1 mutation is an autosomaldominant HH. Because mutations in the gene encodingFPN-1 may affect its presentation on the cell membrane aswell as its ability to bind hepcidin, there are two clinical


TfR

1

HFE T

fR2

HJV

BM

PR

BMP

Tf-Fe (III)

apoTf

sHJV

R-smadR-smad-PSmad4

Ferritin

Hepcidin

Ferritin

Duodenalenterocyte

Fe (III)

FPN-1

Tf-Fe(III)

Tf-Fe (III)

Normal Fe state

(a)

TfR

1

HFE

TfR

2 HJV

BM

PR

BMP

Tf-Fe (III)

apoTf

sHJV

R-smadR-smad-PSmad4

Ferritin

Hepcidin

Ferritin

Fe (III)Fe (III)

FPN-1

Tf-Fe (III) Tf-Fe (III)

Tf-Fe (III)

Hepatocyte

Low Fe stateDuodenalenterocyte

(b)

TfR

1

HFE T

fR2

HJV

BM

PR

BMP

Tf-Fe (III)

apoTf

R-smadR-smad-PSmad4

Ferritin

Hepcidin

Ferritin

Fe (III)Fe (III)

FPN-1

Tf-Fe (III) Tf-Fe (III)

Tf-Fe (III)

Hereditaryhemochromatosis Duodenal

enterocyte

(c)

Figure 2: Hepcidin regulation through the BMP pathway, demonstrating effects of a functional response to systemic iron status and thedysregulation of proteins resulting in iron overload.

features of this genetic disorder [97]. When FPN-1 mutationleads to reduced FPN-1 surface expression (loss-of-functionmutation), patients develop low transferrin saturation andKupffer cell iron loading due to limited iron export functionand become anemic when treated with phlebotomy. Thesepatients are unable to mobilize their iron stores due tomutant FPN-1. The flatiron mouse, with a heterozygousmissense mutation in FPN-1, demonstrates iron loadingof Kupffer cells, high serum ferritin, and low transferrinsaturation, similar to patients with classic “ferroportindisease” described above [98]. Although FPN-1 on duodenalenterocytes is likely also affected, the export of 2–4 mg ofiron during iron absorption may be easier to accomplishdespite the mutation than the 20 mg of iron recycled dailyby macrophages [99].

The other type of FPN-1 mutation results in normalFPN-1 cell surface expression with functional iron exportand is characterized by insensitivity to hepcidin (gain-of-function mutation). This mutation is associated with hightransferrin saturation and hepatocyte iron loading [100].As proof of principle, De Domenico developed an FPN-1mutated cell line that shows a normal iron efflux activity butdoes not respond to increased hepcidin [99]. This lack ofresponse of mutated FPN-1 to hepcidin is associated withincreased duodenal iron absorption, increased transferrinsaturation, and increased iron deposition in patient hepato-cytes. The phenotype of “hepcidin-resistant” HH is similarto the hepcidin deficient phenotype in other types of HH.

Overall, HH mutations are associated with hepcidin sup-pression or hepcidin insensitivity. Hepcidin injections inhibit


the increased iron absorption in the duodena of H f e−/−

mice [101] and forced expression of hepcidin corrects thehemochromatosis phenotype [83]. Unlike in β-thalassemia,sera from patients with HH do not result in hepcidinsuppression in HepG2 cells [102] and induce an increasedhepcidin gene expression in normal hepatocytes [61]. Thesefindings demonstrate mutant HFE on hepatocytes results ininsufficient hepcidin stimulation in HH patients leading toincreased iron absorption and could be the major reason foriron overload in this disease.

Increased hemoglobin, MCV, and MCH have beendemonstrated in patients with HH [103] and H f e−/− mice[84], suggesting that increased iron absorption influenceserythropoiesis. A systematic analysis of erythroid param-eters in HH has not been reported. In one study, 94patients homozygous for C282Y were analyzed and foundto have a 7–9% increase in hemoglobin, MCV, and MCHcompared to normal controls; no increase in RBC countwas observed [103]. This represents approximately 5 g ofadditional hemoglobin and enabled these individuals tosequester 170 mg of iron. This net increase in the hemeand hemoglobin synthesis is not completely understoodsince normal erythroid precursors control heme synthesisby exerting a negative feedback mechanism on ALAS byheme itself [20]. However, ALAS2, with a 5′ untranslatedregion IRE, may result in continued heme synthesis duringiron overload. This compensation to safely sequester iron byincreasing hemoglobin, MCV, and MCH above normal levelsrequires additional experimentation to more completelyunderstand the mechanisms of physiologic set points in ironutilization for hemoglobin synthesis.

6. Summary

Erythropoiesis and iron metabolism must be closely coordi-nated to ensure adequate supply of iron for erythropoiesis.Conversely, the inability to prevent excess production of allhemoglobin intermediates including iron, heme, and globinis hazardous and results in its own pathologic consequences.Hepcidin is a key regulator of iron absorption and recy-cling and is itself under the regulation by erythropoiesisas evidenced by its suppression in diseases of ineffectiveerythropoiesis despite systemic iron overload. Transferrinand TfR1/2 are intimately involved in the regulation ofhepcidin; the dysregulation of these components results inthe inappropriate trafficking of iron in the organism. Howsensing of the body’s iron status occurs is still incompletelyunderstood although it is becoming clear that the totalsystemic iron is not as important as iron availability forerythropoiesis. A tremendous degree of reserve is builtinto the system to accommodate for minor excesses anddeficiencies and the physiology of the “normal state” isuncovered only when the pathology exceeds the body’s abilityto compensate by using these reserves. The distribution ofiron between cells is dependent on the amount of total avail-able iron but also relies on the ability of the body to deliverthat iron for erythropoiesis. Greater understanding of theinterplay between the key factors involved in the regulationof iron metabolism and erythropoiesis will help develop

more effective therapies for disorders of iron overload, irondeficiency, and hemoglobin synthesis.

References

[1] E. Nemeth, M. S. Tuttle, J. Powelson, et al., “Hepcidinregulates cellular iron efflux by binding to ferroportin andinducing its internalization,” Science, vol. 306, no. 5704, pp.2090–2093, 2004.

[2] P. Ponka, C. Beaumont, and D. R. Richardson, “Functionand regulation of transferrin and ferritin,” Seminars inHematology, vol. 35, no. 1, pp. 35–54, 1998.

[3] J. H. Ward, J. P. Kushner, F. A. Ray, and J. Kaplan, “Transferrinreceptor function in hereditary hemochromatosis,” TheJournal of Laboratory and Clinical Medicine, vol. 103, no. 2,pp. 246–254, 1984.

[4] Y. Kohgo, Y. Niitsu, T. Nishisato, et al., “Externalization oftransferrin receptor in established human cell lines,” CellBiology International Reports, vol. 11, no. 12, pp. 871–879,1987.

[5] Y. Kohgo, Y. Niitsu, H. Kondo, et al., “Serum transferrinreceptor as a new index of erythropoiesis,” Blood, vol. 70, no.6, pp. 1955–1958, 1987.

[6] H. A. Huebers, Y. Beguin, P. Pootrakul, D. Einspahr, and C.A. Finch, “Intact transferrin receptors in human plasma andtheir relation to erythropoiesis,” Blood, vol. 75, no. 1, pp. 102–107, 1990.

[7] B. Galy, D. Ferring, B. Minana, et al., “Altered body irondistribution and microcytosis in mice deficient in ironregulatory protein 2 (IRP2),” Blood, vol. 106, no. 7, pp. 2580–2589, 2005.

[8] S. S. Cooperman, E. G. Meyron-Holtz, H. Olivierre-Wilson,M. C. Ghosh, J. P. McConnell, and T. A. Rouault, “Microcyticanemia, erythropoietic protoporphyria, and neurodegenera-tion in mice with targeted deletion of iron-regulatory protein2,” Blood, vol. 106, no. 3, pp. 1084–1091, 2005.

[9] N. M. Sposi, L. Cianetti, E. Tritarelli, et al., “Mechanismsof differential transferrin receptor expression in normalhematopoiesis,” European Journal of Biochemistry, vol. 267,no. 23, pp. 6762–6774, 2000.

[10] H. A. Huebers, E. Csiba, E. Huebers, and C. A. Finch,“Competitive advantage of diferric transferrin in deliveringiron to reticulocytes,” Proceedings of the National Academy ofSciences of the United States of America, vol. 80, no. 1, pp. 300–304, 1983.

[11] M. Cazzola, P. Pootrakul, G. Bergamaschi, H. A. Huebers,M. Eng, and C. A. Finch, “Adequacy of iron supply forerythropoiesis: in vivo observations in humans,” The Journalof Laboratory and Clinical Medicine, vol. 110, no. 6, pp. 734–739, 1987.

[12] H. A. Huebers, B. Josephson, E. Huebers, E. Csiba, and C.A. Finch, “Occupancy of the iron binding sites of humantransferrin,” Proceedings of the National Academy of Sciencesof the United States of America, vol. 81, no. 14, pp. 4326–4330,1984.

[13] F. Canonne-Hergaux, A.-S. Zhang, P. Ponka, and P.Gros, “Characterization of the iron transporter DMT1(NRAMP2/DCT1) in red blood cells of normal and anemicmk/mk mice,” Blood, vol. 98, no. 13, pp. 3823–3830, 2001.

[14] J. H. Katz, “Iron and protein kinetics studied by means ofdoubly labeled human crystalline transferrin,” The Journal ofClinical Investigation, vol. 40, pp. 2143–2152, 1961.


[15] J. F. Reichard, G. T. Motz, and A. Puga, “Heme oxygenase-1 induction by NRF2 requires inactivation of the transcrip-tional repressor BACH1,” Nucleic Acids Research, vol. 35, no.21, pp. 7074–7086, 2007.

[16] H. Ishikawa, M. Kato, H. Hori, et al., “Involvement ofheme regulatory motif in heme-mediated ubiquitination anddegradation of IRP2,” Molecular Cell, vol. 19, no. 2, pp. 171–181, 2005.

[17] T. Tahara, J. Sun, K. Nakanishi, et al., “Heme positivelyregulates the expression of β-globin at the locus controlregion Via the transcriptional factor Bach1 in erythroid cells,”The Journal of Biological Chemistry, vol. 279, no. 7, pp. 5480–5487, 2004.

[18] G. Kramer, J. M. Cimadevilla, and B. Hardesty, “Specificityof the protein kinase activity associated with the hemincontrolled repressor of rabbit reticulocyte,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 73, no. 9, pp. 3078–3082, 1976.

[19] E. McEwen, N. Kedersha, B. Song, et al., “Heme-regulatedinhibitor kinase-mediated phosphorylation of eukaryotictranslation initiation factor 2 inhibits translation, inducesstress granule formation, and mediates survival upon arsen-ite exposure,” The Journal of Biological Chemistry, vol. 280,no. 17, pp. 16925–16933, 2005.

[20] S. Sassa and T. Nagai, “The role of heme in gene expression,”International Journal of Hematology, vol. 63, no. 3, pp. 167–178, 1996.

[21] M. Schranzhofer, M. Schifrer, J. A. Cabrera, et al., “Remod-eling the regulation of iron metabolism during erythroiddifferentiation to ensure efficient heme biosynthesis,” Blood,vol. 107, no. 10, pp. 4159–4167, 2006.

[22] S. Abboud and D. J. Haile, “A novel mammalian iron-regulated protein involved in intracellular iron metabolism,”The Journal of Biological Chemistry, vol. 275, no. 26, pp.19906–19912, 2000.

[23] A. T. McKie, P. Marciani, A. Rolfs, et al., “A novel duodenaliron-regulated transporter, IREG1, implicated in the basolat-eral transfer of iron to the circulation,” Molecular Cell, vol. 5,no. 2, pp. 299–309, 2000.

[24] X. B. Liu, P. Hill, and D. J. Haile, “Role of the ferroportiniron-responsive element in iron and nitric oxide dependentgene regulation,” Blood Cells, Molecules & Diseases, vol. 29,no. 3, pp. 315–326, 2002.

[25] T. Ganz, G. Olbina, D. Girelli, E. Nemeth, and M. Westerman,“Immunoassay for human serum hepcidin,” Blood, vol. 112,no. 10, pp. 4292–4297, 2008.

[26] G. Nicolas, C. Chauvet, L. Viatte, et al., “The gene encodingthe iron regulatory peptide hepcidin is regulated by anemia,hypoxia, and inflammation,” The Journal of Clinical Investi-gation, vol. 110, no. 7, pp. 1037–1044, 2002.

[27] A. Bondi, P. Valentino, F. Daraio, et al., “Hepatic expressionof hemochromatosis genes in two mouse strains afterphlebotomy and iron overload,” Haematologica, vol. 90, no.9, pp. 1161–1167, 2005.

[28] D.-L. Zhang, R. M. Hughes, H. Ollivierre-Wilson, M. C.Ghosh, and T. A. Rouault, “A ferroportin transcript thatlacks an iron-responsive element enables duodenal anderythroid precursor cells to evade translational repression,”Cell Metabolism, vol. 9, no. 5, pp. 461–473, 2009.

[29] J. G. Quigley, Z. Yang, M. T. Worthington, et al., “Iden-tification of a human heme exporter that is essential forerythropoiesis,” Cell, vol. 118, no. 6, pp. 757–766, 2004.

[30] N. G. Testa, D. Onions, O. Jarrett, F. Frassoni, and J. F.Eliason, “Haemopoietic colony formation (BFU-E, GM-CFC) during the development of pure red cell hypoplasiainduced in the cat by feline leukaemia virus,” LeukemiaResearch, vol. 7, no. 2, pp. 103–116, 1983.

[31] J. L. Abkowitz, K. M. Sabo, B. Nakamoto, et al., “Diamond-Blackfan anemia: in vitro response of erythroid progenitorsto the ligand for c-kit,” Blood, vol. 78, no. 9, pp. 2198–2202,1991.

[32] S. B. Keel, R. T. Doty, Z. Yang, et al., “A heme exportprotein is required for red blood cell differentiation and ironhomeostasis,” Science, vol. 319, no. 5864, pp. 825–828, 2008.

[33] X. Du, E. She, T. Gelbart, et al., “The serine proteaseTMPRSS6 is required to sense iron deficiency,” Science, vol.320, no. 5879, pp. 1088–1092, 2008.

[34] K. E. Finberg, M. M. Heeney, D. R. Campagna, et al.,“Mutations in TMPRSS6 cause iron-refractory iron defi-ciency anemia (IRIDA),” Nature Genetics, vol. 40, no. 5, pp.569–571, 2008.

[35] M. A. Melis, M. Cau, R. Congiu, et al., “A mutationin the TMPRSS6 gene, encoding a transmembrane serineprotease that suppresses hepcidin production, in familial irondeficiency anemia refractory to oral iron,” Haematologica,vol. 93, no. 10, pp. 1473–1479, 2008.

[36] F. Guillem, S. Lawson, C. Kannengiesser, M. Westerman, C.Beaumont, and B. Grandchamp, “Two nonsense mutationsin the TMPRSS6 gene in a patient with microcytic anemiaand iron deficiency,” Blood, vol. 112, no. 5, pp. 2089–2091,2008.

[37] L. Silvestri, A. Pagani, A. Nai, I. De Domenico, J. Kaplan,and C. Camaschella, “The serine protease matriptase-2(TMPRSS6) inhibits hepcidin activation by cleaving mem-brane hemojuvelin,” Cell Metabolism, vol. 8, no. 6, pp. 502–511, 2008.

[38] D. R. Richardson and P. Ponka, “The molecular mechanismsof the metabolism and transport of iron in normal andneoplastic cells,” Biochimica et Biophysica Acta, vol. 1331, no.1, pp. 1–40, 1997.

[39] N. Goya, S. Miyazaki, S. Kodate, and B. Ushio, “A family ofcongenital atransferrinemia,” Blood, vol. 40, no. 2, pp. 239–245, 1972.

[40] R. L. Hamill, J. C. Woods, and B. A. Cook, “Congenitalatransferrinemia: a case report and review of the literature,”American Journal of Clinical Pathology, vol. 96, no. 2, pp. 215–218, 1991.

[41] A. Hayashi, Y. Wada, T. Suzuki, and A. Shimizu, “Studieson familial hypotransferrinemia: unique clinical course andmolecular pathology,” American Journal of Human Genetics,vol. 53, no. 1, pp. 201–213, 1993.

[42] E. Beutler, T. Gelbart, P. Lee, R. Trevino, M. A. Fernandez,and V. F. Fairbanks, “Molecular characterization of a caseof atransferrinemia,” Blood, vol. 96, no. 13, pp. 4071–4074,2000.

[43] S. Goldwurm, C. Casati, N. Venturi, et al., “Biochemical andgenetic defects underlying human congenital hypotransfer-rinemia,” The Hematology Journal, vol. 1, no. 6, pp. 390–398,2000.

[44] P. Trombini, T. Coliva, E. Nemeth, et al., “Effects of plasmatransfusion on hepcidin production in human congenitalhypotransferrinemia,” Haematologica, vol. 92, no. 10, pp.1407–1410, 2007.


[45] C. Finch, “Regulators of iron balance in humans,” Blood, vol.84, no. 6, pp. 1697–1702, 1994.

[46] S. E. Bernstein, “Hereditary hypotranferrinemia withhemosiderosis, a murine disorder resembling human atrans-ferrinemia,” The Journal of Laboratory and Clinical Medicine,vol. 110, no. 6, pp. 690–705, 1987.

[47] C. C. Trenor III, D. R. Campagna, V. M. Sellers, N. C.Andrews, and M. D. Fleming, “The molecular defect inhypotransferrinemic mice,” Blood, vol. 96, no. 3, pp. 1113–1118, 2000.

[48] J. I. Huggenvik, C. M. Craven, R. L. Idzerda, S. Bernstein, J.Kaplan, and G. S. McKnight, “A splicing defect in the mousetransferrin gene leads to congenital atransferrinemia,” Blood,vol. 74, no. 1, pp. 482–486, 1989.

[49] B. P. Esposito, W. Breuer, P. Sirankapracha, P. Pootrakul,C. Hershko, and Z. I. Cabantchik, “Labile plasma iron iniron overload: redox activity and susceptibility to chelation,”Blood, vol. 102, no. 7, pp. 2670–2677, 2003.

[50] K. B. Raja, D. J. Pountney, R. J. Simpson, and T. J. Peters,“Importance of anemia and transferrin levels in the regu-lation of intestinal iron absorption in hypotransferrinemicmice,” Blood, vol. 94, no. 9, pp. 3185–3192, 1999.

[51] L. Lin, E. V. Valore, E. Nemeth, J. B. Goodnough, V. Gabayan,and T. Ganz, “Iron transferrin regulates hepcidin synthesisin primary hepatocyte culture through hemojuvelin andBMP2/4,” Blood, vol. 110, no. 6, pp. 2182–2189, 2007.

[52] C. N. Roy, D. A. Weinstein, and N. C. Andrews, “2002E. Mead Johnson Award for research in pediatrics lecture:the molecular biology of the anemia of chronic disease: ahypothesis,” Pediatric Research, vol. 53, no. 3, pp. 507–512,2003.

[53] I. Cavill, “Erythropoiesis and iron,” Best Practice & Research:Clinical Haematology, vol. 15, no. 2, pp. 399–409, 2002.

[54] K. B. Raja, R. J. Simpson, and T. J. Peters, “Intestinaliron absorption studies in mouse models of iron-overload,”British Journal of Haematology, vol. 86, no. 1, pp. 156–162,1994.

[55] S. S. Buys, C. B. Martin, M. Eldridge, J. P. Kushner, andJ. Kaplan, “Iron absorption in hypotransferrinemic mice,”Blood, vol. 78, no. 12, pp. 3288–3290, 1991.

[56] H. Li, A. C. Rybicki, S. M. Suzuka, et al., “Transferrin therapyameliorates disease in β-thalassemic mice,” Nature Medicine,vol. 16, no. 2, pp. 177–182, 2010.

[57] Y. Z. Ginzburg, A. C. Rybicki, S. M. Suzuka, et al., “Exoge-nous iron increases hemoglobin in β-thalassemic mice,”Experimental Hematology, vol. 37, no. 2, pp. 172–183, 2009.

[58] S. Gardenghi, M. F. Marongiu, P. Ramos, et al., “Ineffectiveerythropoiesis in β-thalassemia is characterized by increasediron absorption mediated by down-regulation of hepcidinand up-regulation of ferroportin,” Blood, vol. 109, no. 11, pp.5027–5035, 2007.

[59] A. Kattamis, I. Papassotiriou, D. Palaiologou, et al., “Theeffects of erythropoetic activity and iron burden on hepcidinexpression in patients with thalassemia major,” Haematolog-ica, vol. 91, no. 6, pp. 809–812, 2006.

[60] S. L. Kearney, E. Nemeth, E. J. Neufeld, et al., “Urinaryhepcidin in congenital chronic anemias,” Pediatric Blood andCancer, vol. 48, no. 1, pp. 57–63, 2007.

[61] O. Weizer-Stern, K. Adamsky, N. Amariglio, et al., “Down-regulation of hepcidin and haemojuvelin expression in thehepatocyte cell-line HepG2 induced by thalassaemic sera,”British Journal of Haematology, vol. 135, no. 1, pp. 129–138,2006.

[62] E. H. J. M. Kemna, A. E. R. Kartikasari, L. J. H. van Tits,P. Pickkers, H. Tjalsma, and D. W. Swinkels, “Regulationof hepcidin: insights from biochemical analyses on humanserum samples,” Blood Cells, Molecules, and Diseases, vol. 40,no. 3, pp. 339–346, 2008.

[63] G. Nicolas, L. Viatte, M. Bennoun, C. Beaumont, A. Kahn,and S. Vaulont, “Hepcidin, a new iron regulatory peptide,”Blood cells, Molecules & Diseases, vol. 29, no. 3, pp. 327–335,2002.

[64] M. Vokurka, J. Krijt, K. Sulc, and E. NeCas, “Hepcidin mRNAlevels in mouse liver respond to inhibition of erythropoiesis,”Physiological Research, vol. 55, no. 6, pp. 667–674, 2006.

[65] M. Pak, M. A. Lopez, V. Gabayan, T. Ganz, and S. Rivera,“Suppression of hepcidin during anemia requires erythro-poietic activity,” Blood, vol. 108, no. 12, pp. 3730–3735, 2006.

[66] T. Tanno, N. V. Bhanu, P. A. Oneal, et al., “High levelsof GDF15 in thalassemia suppress expression of the ironregulatory protein hepcidin,” Nature Medicine, vol. 13, no. 9,pp. 1096–1101, 2007.

[67] T. Tanno, P. Porayette, O. Sripichai, et al., “Identification ofTWSG1 as a second novel erythroid regulator of hepcidinexpression in murine and human cells,” Blood, vol. 114, no.1, pp. 181–186, 2009.

[68] J. E. Levy, O. Jin, Y. Fujiwara, F. Kuo, and N. C. Andrews,“Transferrin receptor is necessary for development of ery-throcytes and the nervous system,” Nature Genetics, vol. 21,no. 4, pp. 396–399, 1999.

[69] C. Brugnara, G. M. Colella, J. Cremins, et al., “Effects ofsubcutaneous recombinant human erythropoietin in normalsubjects: development of decreased reticulocyte hemoglobincontent and iron-deficient erythropoiesis,” The Journal ofLaboratory and Clinical Medicine, vol. 123, no. 5, pp. 660–667, 1994.

[70] I. Cavill and I. C. Macdougall, “Functional iron deficiency,”Blood, vol. 82, no. 4, p. 1377, 1993.

[71] P. Pootrakul, J. Wattanasaree, M. Anuwatanakulchai, andP. Wasi, “Increased red blood cell protoporphyrin in tha-lassemia: a result of relative iron deficiency,” AmericanJournal of Clinical Pathology, vol. 82, no. 3, pp. 289–293, 1984.

[72] Y. Beguin, S. M. Stray, M. Cazzola, H. A. Huebers, and C.A. Finch, “Ferrokinetic measurement of erythropoiesis,” ActaHaematologica, vol. 79, no. 3, pp. 121–126, 1988.

[73] J. N. Feder, A. Gnirke, W. Thomas, et al., “A novel MHCclass I-like gene is mutated in patients with hereditaryhaemochromatosis,” Nature Genetics, vol. 13, no. 4, pp. 399–408, 1996.

[74] J. N. Feder, Z. Tsuchihashi, A. Irrinki, et al., “The hemochro-matosis founder mutation in HLA-H disrupts β2- microglob-ulin interaction and cell surface expression,” The Journal ofBiological Chemistry, vol. 272, no. 22, pp. 14025–14028, 1997.

[75] A. Piperno, D. Girelli, E. Nemeth, et al., “Blunted hepcidinresponse to oral iron challenge in HFE-related hemochro-matosis,” Blood, vol. 110, no. 12, pp. 4096–4100, 2007.

[76] J. Truksa, H. Peng, P. Lee, and E. Beutler, “Bone mor-phogenetic proteins 2, 4, and 9 stimulate murine hepcidin1 expression independently of Hfe, transferrin receptor 2(Tfr2), and IL-6,” Proceedings of the National Academy ofSciences of the United States of America, vol. 103, no. 27, pp.10289–10293, 2006.

[77] J. L. Babitt, F. W. Huang, Y. Xia, Y. Sidis, N. C. Andrews,and H. Y. Lin, “Modulation of bone morphogenetic proteinsignaling in vivo regulates systemic iron balance,” Journal ofClinical Investigation, vol. 117, no. 7, pp. 1933–1939, 2007.


[78] R.-H. Wang, C. Li, X. Xu, et al., “A role of SMAD4 iniron metabolism through the positive regulation of hepcidinexpression,” Cell Metabolism, vol. 2, no. 6, pp. 399–409, 2005.

[79] J. L. Babitt, F. W. Huang, D. M. Wrighting, et al., “Bonemorphogenetic protein signaling by hemojuvelin regulateshepcidin expression,” Nature Genetics, vol. 38, no. 5, pp. 531–539, 2006.

[80] L. Lin, Y. P. Goldberg, and T. Ganz, “Competitive regulationof hepcidin mRNA by soluble and cell-associated hemoju-velin,” Blood, vol. 106, no. 8, pp. 2884–2889, 2005.

[81] K. R. Bridle, D. M. Frazer, S. J. Wilkins, et al., “Disruptedhepcidin regulation in HFE-associated haemochromatosisand the liver as a regulator of body iron homoeostasis,” TheLancet, vol. 361, no. 9358, pp. 669–673, 2003.

[82] M. Muckenthaler, C. N. Roy, A. O. Custodio, et al., “Regula-tory defects in liver and intestine implicate abnormal hep-cidin and Cybrd1 expression in mouse hemochromatosis,”Nature Genetics, vol. 34, no. 1, pp. 102–107, 2003.

[83] G. Nicolas, L. Viatte, D.-Q. Lou, et al., “Constitutive hepcidinexpression prevents iron overload in a mouse model ofhemochromatosis,” Nature Genetics, vol. 34, no. 1, pp. 97–101, 2003.

[84] X. Y. Zhou, S. Tomatsu, R. E. Fleming, et al., “HFE geneknockout produces mouse model of hereditary hemochro-matosis,” Proceedings of the National Academy of Sciences ofthe United States of America, vol. 95, no. 5, pp. 2492–2497,1998.

[85] J. N. Feder, D. M. Penny, A. Irrinki, et al., “The hemochro-matosis gene product complexes with the transferrin receptorand lowers its affinity for ligand binding,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 95, no. 4, pp. 1472–1477, 1998.

[86] M. J. Bennett, J. A. Lebron, and P. J. Bjorkman, “Crystalstructure of the hereditary haemochromatosis protein HFEcomplexed with transferrin receptor,” Nature, vol. 403, no.6765, pp. 46–53, 2000.

[87] A. M. Giannetti and P. J. Bjorkman, “HFE and transferrindirectly compete for transferrin receptor in solution and atthe cell surface,” The Journal of Biological Chemistry, vol. 279,no. 24, pp. 25866–25875, 2004.

[88] J. A. Lebron, M. J. Bennett, D. E. Vaughn, et al., “Crystalstructure of the hemochromatosis protein HFE and charac-terization of its interaction with transferrin receptor,” Cell,vol. 93, no. 1, pp. 111–123, 1998.

[89] M. B. Johnson and C. A. Enns, “Diferric transferrin regulatestransferrin receptor 2 protein stability,” Blood, vol. 104, no.13, pp. 4287–4293, 2004.

[90] A. Robb and M. Wessling-Resnick, “Regulation of transferrinreceptor 2 protein levels by transferrin,” Blood, vol. 104, no.13, pp. 4294–4299, 2004.

[91] P. J. Schmidt, P. T. Toran, A. M. Giannetti, P. J. Bjorkman,and N. C. Andrews, “The transferrin receptor modulatesHfe-dependent regulation of hepcidin expression,” CellMetabolism, vol. 7, no. 3, pp. 205–214, 2008.

[92] T. Goswami and N. C. Andrews, “Hereditary hemochro-matosis protein, HFE, interaction with transferrin receptor2 suggests a molecular mechanism for mammalian ironsensing,” The Journal of Biological Chemistry, vol. 281, no. 39,pp. 28494–28498, 2006.

[93] C. Camaschella and E. Poggiali, “Rare types of genetichemochromatosis,” Acta Haematologica, vol. 122, no. 2-3, pp.140–145, 2009.

[94] V. Niederkofler, R. Salie, and S. Arber, “Hemojuvelin isessential for dietary iron sensing, and its mutation leads tosevere iron overload,” The Journal of Clinical Investigation,vol. 115, no. 8, pp. 2180–2186, 2005.

[95] P. G. Fraenkel, Y. Gibert, J. L. Holzheimer, et al., “Transferrin-a modulates hepcidin expression in zebrafish embryos,”Blood, vol. 113, no. 12, pp. 2843–2850, 2009.

[96] D. F. Wallace, L. Summerville, E. M. Crampton, D. M. Frazer,G. J. Anderson, and V. N. Subramaniam, “Combined deletionof Hfe and transferrin receptor 2 in mice leads to markeddysregulation of hepcidin and iron overload,” Hepatology,vol. 50, no. 6, pp. 1992–2000, 2009.

[97] A. Fernandes, G. C. Preza, Y. Phung, et al., “The molecularbasis of hepcidin-resistant hereditary hemochromatosis,”Blood, vol. 114, no. 2, pp. 437–443, 2009.

[98] I. E. Zohn, I. De Domenico, A. Pollock, et al., “The flatironmutation in mouse ferroportin acts as a dominant negative tocause ferroportin disease,” Blood, vol. 109, no. 10, pp. 4174–4180, 2007.

[99] I. De Domenico, D. M. Ward, E. Nemeth, et al., “Themolecular basis of ferroportin-linked hemochromatosis,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 102, no. 25, pp. 8955–8960, 2005.

[100] I. De Domenico, D. M. Ward, G. Musci, and J. Kaplan, “Ironoverload due to mutations in ferroportin,” Haematologica,vol. 91, no. 1, pp. 92–95, 2006.

[101] A. H. Laftah, B. Ramesh, R. J. Simpson, et al., “Effect ofhepcidin on intestinal iron absorption in mice,” Blood, vol.103, no. 10, pp. 3940–3944, 2004.

[102] B. A. C. van Dijk, C. M. M. Laarakkers, S. M. Klaver, etal., “Serum hepcidin levels are innately low in HFE-relatedhaemochromatosis but differ between C282Y-homozygoteswith elevated and normal ferritin levels,” British Journal ofHaematology, vol. 142, no. 6, pp. 979–985, 2008.

[103] J. C. Barton, L. F. Bertoli, and B. E. Rothenberg, “Peripheralblood erythrocyte parameters in hemochromatosis: evidencefor increased erythrocyte hemoglobin content,” The Journalof Laboratory and Clinical Medicine, vol. 135, no. 1, pp. 96–104, 2000.


Review Article

Targeting the Hepcidin-Ferroportin Axis in the Diagnosis andTreatment of Anemias

Elizabeta Nemeth

David Geffen School of Medicine at UCLA, CHS 37-131, 10833 LeConte Avenue, Los Angeles, CA 90095-1690, USA

Correspondence should be addressed to Elizabeta Nemeth, [email protected]

Received 2 November 2009; Accepted 23 November 2009


Copyright © 2010 Elizabeta Nemeth. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

The hepatic peptide hormone hepcidin regulates dietary iron absorption, plasma iron concentrations, and tissue iron distribution.Hepcidin acts by causing the degradation of its receptor, the cellular iron exporter ferroportin. The loss of ferroportin decreasesiron flow into plasma from absorptive enterocytes, from macrophages that recycle the iron of senescent erythrocytes, and fromhepatocytes that store iron, thereby lowering plasma iron concentrations. Malfunctions of the hepcidin-ferroportin axis contributeto the pathogenesis of different anemias. Deficient production of hepcidin causes systemic iron overload in iron-loading anemiassuch as beta-thalassemia; whereas hepcidin excess contributes to the development of anemia in inflammatory disorders and chronickidney disease, and may cause erythropoietin resistance. The diagnosis of different forms of anemia will be facilitated by improvedhepcidin assays, and the treatment will be enhanced by the development of hepcidin agonists and antagonists.

1. Hepcidin-Ferroportin InteractionRegulates Iron Homeostasis

Hepcidin is a small peptide hormone secreted by hepato-cytes, circulating in blood plasma and excreted in urine[1]. Like other peptide hormones, hepcidin is synthesizedinitially as a larger 84-amino acid preprohepcidin thenprocessed in hepatocytes by the signal peptidase to 60-aminoacid prohepcidin that lacks iron-regulatory activity [2]. Priorto secretion, prohormone convertases cleave prohepcidin ata polybasic motif to generate the mature bioactive 25-aminoacid hepcidin [3]. Other cell types including macrophagesand adipocytes also contain hepcidin mRNA but their localand systemic contribution to the production of bioactivehepcidin has not been established with certainty. Hepcidinplays an essential role in maintaining iron homeostasis, andthe dysregulation of its production underlies many iron dis-orders. Chronic excess of hepcidin causes iron-restricted ane-mia [4], whereas hepcidin deficiency results in iron overloadwith iron deposition in the liver and other parenchyma [5].

Hepcidin acts by regulating the cellular concentration ofits receptor, ferroportin. Ferroportin is the sole known cel-lular iron exporter and is essential for iron homeostasis [6].

This multispanning membrane protein is expressed in tissueswhich transport large amounts of iron (Figure 1): duodenalenterocytes which absorb dietary iron, macrophages of thespleen and liver which recycle iron from old erythrocytes,hepatocytes which store and release iron according to bodyneeds, and placental trophoblast which transports iron frommaternal to fetal circulation [7–9].

When ferroportin is located in the cell membrane itallows efflux of iron from the cells into plasma. Hepcidinbinding to the extracellular face of ferroportin triggers inter-nalization and degradation of the ligand-receptor complex[10]. Removal of ferroportin from the membrane stopscellular iron export leading to decreased supply of iron intoplasma (Figure 1). Without the constant iron influx, theplasma iron pool is rapidly depleted by the iron-consumingcells, most prominently erythroid precursors. In mice, asingle injection of synthetic hepcidin caused a rapid dropin serum iron [11], and this lasted for 2 days, presumablyuntil sufficient amount of ferroportin was resynthesized.Decreased ferroportin concentration in cell membranes, asseen during chronic overproduction of hepcidin, can leadto iron-restricted erythropoiesis. Interestingly, ferroportinis also expressed in erythroid precursor cells [12], but its


physiological role or the effect of hepcidin on developingerythrocytes remains to be determined.

2. Regulation of Hepcidin

Hepcidin is homeostatically regulated by iron and erythro-poietic activity. Increased plasma and stored iron stimulatehepcidin production, which in turn blocks dietary ironabsorption and further iron loading (Figure 1). Hepcidinis suppressed in iron deficiency [13], allowing increasedabsorption of dietary iron and replenishment of iron stores.The feedback loop between iron and hepcidin ensures thestability of plasma iron concentrations.

As would be expected for the iron-regulatory hormone,hepcidin production is also regulated by the process whichconsumes most iron, erythropoiesis [14]. Increased erythro-poietic activity suppresses hepcidin production which allowsthe release of stored iron from macrophages and hepatocytes,and increased iron absorption, all resulting in greater supplyof iron for hemoglobin synthesis.

Hepcidin production is also pathologically increased ininflammation and infection [15]. Resultant hypoferremiamay represent a host defense strategy to limit iron availabilityto microorganisms, but can also lead to iron dysregulationand iron-restricted anemia in inflammatory diseases.

3. Molecular Mechanisms ofHepcidin Regulation

3.1. Iron. Hepcidin is likely regulated by both circulatingiron-transferrin and intracellular iron stores. Although therespective mechanisms of sensing extracellular and intra-cellular iron are not well understood, they both appear toutilize the bone morphogenetic protein (BMP) pathway toalter hepcidin expression. Several BMPs have been shownto increase hepcidin production in vitro and in vivo [16],but BMP6 has recently emerged as the principal endogenousBMP regulating hepcidin. BMP6 knockout mice developsevere iron overload but no other significant abnormalities[17, 18].

In other biological settings, BMP signaling is known tobe modulated by coreceptors and antagonists. Hemojuvelin(HJV), a GPI-linked membrane protein, appears to be theco-receptor specialized for iron regulation [19]. The solubleform of hemojuvelin acts as an antagonist, probably bybinding BMPs, but the biological role of this interaction hasnot been documented yet [20]. HJV mutations in humans ormice result in severe iron overload similar to that caused byablation of hepcidin, without any other apparent problems[21]. Additional molecules, including a protease TMPRSS6[22] and a large multifunctional transmembrane proteinneogenin [23], were shown to interact with HJV and modifyits cell-surface expression. Whether these molecules or theirinteraction with HJV is modulated by iron or other signalsremains to be determined.

The mechanism by which intracellular iron regulateshepcidin expression is still unclear. However, expression ofBMP6 mRNA was recently shown to increase with iron

loading in mice, raising the possibility that BMP6 is asignal reflecting iron stores [24]. Hepcidin regulation byextracellular iron is better understood, and a tentative modelis emerging but needs further experimental support. Sensingof iron-transferrin concentrations apparently depends ontransferrin receptor 2 (TfR2) and HFE, two molecules whichare mutated in the adult form of hereditary hemochromato-sis. TfR2 is a homolog of TfR1, but is primarily expressedin the liver, the site of hepcidin expression. Binding ofiron-transferrin to TfR2 stabilizes the protein [25], whichresults in elevated ERK1/2 and Smad1/5/8 signaling [26].HFE, a protein similar to MHC I type molecules, appearsto function as a shuttle between TfR1 and TfR2 dependingon iron-transferrin concentrations [27]. Because the bindingsites of HFE and iron-transferrin on TfR1 overlap, higheriron-transferrin concentrations displace HFE from TfR1allowing HFE to associate with TfR2, which presumablypotentiates signaling pathways downstream of TfR2. HFEand TfR2, however, do not appear to be required for hepcidinregulation by iron stores, as mice and humans with HFE andTfR2 mutations are still capable of decreasing hepcidin levelsafter iron depletion [28].

3.2. Erythropoiesis. Increased erythropoietic activity is apotent suppressor of hepcidin production. A single injectionof erythropoietin in humans caused a dramatic decrease inserum hepcidin within 24 hours [29], and a mouse modelshowed a dose-dependent decrease in hepcidin mRNA aftererythropoietin administration [30]. Epo by itself does notappear to be a direct regulator of hepcidin expression becausepretreatment of mice with carboplatin, a cytotoxic inhibitorof erythropoiesis, completely abrogated the effect of Epo onhepcidin [14]. Similarly, mouse models of anemias causedby bleeding or hemolysis showed that hepcidin suppressiondepended on intact erythropoietic activity [14, 31].

How erythropoiesis affects hepcidin production is notclear, but the mediators could include the production ofsoluble factors by the erythroid precursors in the bonemarrow, decreased circulating or stored iron, and hypoxia.Two proteins produced by erythroid precursors, growthdifferentiation factor 15 (GDF15) and twisted gastrulationprotein (TWSG1), have been proposed to mediate hepcidinsuppression in anemias with ineffective erythropoiesis [32,33]. GDF15, a member of the TGF-β superfamily, andTWSG1, a BMP-binding protein, are both produced bydeveloping erythroblasts. The two proteins were shown tosuppress hepcidin mRNA in vitro [32, 33]. Very high levelsof GDF15 were detected in β-thalassemia and congenitaldyserythropoietic anemia type I, and elevated Twsg1 expres-sion was found in a mouse model of thalassemia. However,physiologic hepcidin suppression in response to bleeding orto anemias with effective erythropoiesis is likely mediated byother mechanisms. Whatever the signal, erythroid activityappears to suppress hepcidin, at least in part, by modulatingthe BMP pathway. Erythropoietin administration in mice,which caused the expected decrease in hepcidin levels, wasalso found to reduce Smad signaling [30].

The physiological relevance of hepcidin regulation byhypoxia is still uncertain. Alterations of the HIF pathway in


Erythropoietic signal

Spleen

Bonemarrow

RBC

Liver

Duodenum

Hepcidin

Hepcidin

Iron signal

Hepcidin

Fpn

Fpn

Fpn

Inflammation

PlasmaFe-Tf

Figure 1: Hepcidin-ferroportin interaction determines the flow of iron into plasma. Hepcidin concentration is in turn regulated by iron,erythropoietic activity, and inflammation.

vivo can affect hepcidin expression [34] but whether HIFregulates hepcidin transcription directly or mostly indirectlyis still unresolved. It is possible that the main effect ofhypoxia on iron homeostasis is to increase erythropoietinproduction in the kidney, which would lead to proliferationof erythroblasts and suppression of hepcidin by putativeerythroid factors.

3.3. Inflammation. Hepcidin synthesis is rapidly increasedby infection and inflammation, causing retention of ironin macrophages and decreased iron absorption [15]. Theresulting hypoferremia is presumably a component of innateimmune responses that deprive invading microbes of ironand other essential nutrients. Serum hepcidin was foundto be greatly increased in patients with inflammationdefined as CRP > 10 mg/dL, patients with sepsis, burns,inflammatory bowel disease, and multiple myeloma [13, 35–37]. Of inflammatory mediators regulating hepcidin, IL-6was shown to be a prominent inducer in vitro and in vivo,and it stimulates hepcidin transcription through a STAT-3dependent mechanism [38, 39]. Other cytokines such as IL-1 may also directly regulate hepcidin production, at leastin mice, as IL-6 knockout mice with chronic inflammationincreased hepcidin mRNA similarly to wild-type mice [40].Direct regulation of hepcidin synthesis in myeloid cells bymicrobial molecules acting through toll-like receptors hasalso been proposed [41] but it is not clear how much thismechanism contributes to hepcidin production locally orsystemically, and whether it also applies to hepatocytes.

4. The Role of Hepcidin in Anemias

Hepcidin expression is the result of the interplay betweenmultiple stimuli, including iron, inflammation, and erythro-poiesis. Accordingly, hepcidin concentrations in differentforms of anemia vary widely, and may have diagnosticpotential in differentiating between the various types ofanemia. Furthermore, in cases where hepcidin is a causativefactor in anemia, hepcidin-targeted therapies may improvetreatment options for the patients.

4.1. Iron Deficiency Anemia (IDA). In pure iron deficiencyanemia, serum and urinary hepcidin concentrations aregreatly decreased and are frequently undetectable by cur-rently available assays [13]. The low expression is presumablydue to the lack of transcriptional stimulation by iron as wellas active suppression by erythroid factors. Hepcidin appearsto be a sensitive indicator of iron deficiency even in theabsence of anemia. Decreased hepcidin, together with lowtransferrin saturation and serum ferritin, is observed priorto a detectable decrease in Hb or Hct ([13], E. Nemethunpublished). Hepcidin measurements could improve thescreening of blood donors, in whom deferral is currentlybased on low Hct or Hb levels, the relatively late sequelaeof iron deficiency. Identifying donors who are already iron-deficient prior to the blood donation should reduce the fre-quency of frank iron deficiency and anemia in blood donors.

Hepcidin is readily detectable in urine [13] and thus maybe an excellent candidate for the development of a simple


field test for iron deficiency. Apart from the possible usefor easy screening of infants and small children, a field-friendly test would be particularly useful in areas whichlack access to clinical laboratories. Accurate detection ofiron deficiency has become an important issue in endemicmalarial regions. “The Pemba trial,” a randomized controlledtrial in Zanzibar [42] which evaluated the impact of iron,folic acid, and zinc supplementation on morbidity andmortality in young children, found an increased risk of severeillness and death in children supplemented with iron/folicacid. Furthermore, the risk appeared to be restricted to iron-replete children. Thus, a simple, quick hepcidin test wouldhelp distinguish between iron-deficient/anemic children whocan benefit from iron supplementation, and iron-replete oralready infected children in whom iron supplementationmight be harmful.

4.2. Iron-Refractory Iron Deficiency Anemia (IRIDA). Iron-refractory iron deficiency anemia is a hereditaryhypochromic, microcytic anemia, refractory to treatmentwith oral iron, and only partially responsive to parenteraliron. The molecular basis of the disease was only recentlydescribed. IRIDA is caused by increased hepcidin productiondue to mutations in the hepcidin suppressor, TMPRSS6also called matriptase-2 [43, 44]. The disease is thus aphenotypic opposite of juvenile hemochromatosis causedby disruptive hepcidin mutations. TMPRSS6 is a membraneprotease which was reported to act by cleaving membrane-associated HJV [22], which presumably results in decreasedBMP pathway activity. When TMPRSS6 is mutated,hepcidin expression is increased, chronically inhibitingiron absorption and resulting in the development of irondeficiency anemia. How TMPRSS6 expression and activityare regulated in relation to iron homeostasis still remains tobe determined.

Interestingly, TMPRSS6 locus is proving to be highlypolymorphic. If some common polymorphisms conferchanges in TMPRSS6 expression or activity, this couldin turn affect hepcidin expression and the control ofiron metabolism. Recently, genome-wide association stud-ies identified association between certain TMPRSS6 vari-ants and serum iron, transferrin saturation, MCV andhemoglobin levels [45, 46]. These findings suggest that evensubtle variants in hepcidin regulators have an influenceon iron status, erythropoiesis, and associated disorders inthe general population. Improved definition of genetic riskfactors for iron deficiency or iron overload could help avoidunnecessary burdens associated with chronic treatment.

4.3. Iron-Refractory Anemia Associated with Hepcidin-producing Tumors. Another rare form of iron-refractoryiron deficiency anemia is associated with hepatic adenomas.Originally reported in the setting of type 1a glycogen storagedisease, patients had large hepatic adenomas associatedwith moderate to severe unremitting microcytic anemiaand hypoferremia, unresponsive to oral iron and onlypartially responsive to parenteral iron [47]. The anemiaand hypoferremia rapidly resolved after tumor resection.A sample of tumor tissue overexpressed hepcidin mRNA

which was suppressed in the surrounding liver, suggestingautonomous production of hepcidin by the tumor. A patientwith a similar presentation in the setting of a large hepaticadenoma but no history of glycogen disorder was recentlyreported [48].

4.4. Iron-Loading Anemias. In iron-loading anemias, suchas β-thalassemia and congenital dyserythropoietic anemias,urinary and serum hepcidin are severely decreased in theabsence of transfusions [49–51]. The low hepcidin in turnallows excessive iron absorption and development of sys-temic iron overload, similar to hereditary hemochromatosis.The signal causing hepcidin suppression in iron-loadinganemias appears to be generated by high erythropoieticactivity and outweighs the effects of the resulting ironoverload on hepcidin regulation. As mentioned earlier,GDF15 and TWSG1 are two erythroid factors that maycontribute to hepcidin suppression in syndromes withineffective erythropoiesis [32, 33].

Hepcidin diagnostics may be useful for iron-loadinganemias to identify the patients at higher risk of iron toxicitydue to severely decreased hepcidin levels. Moreover, futurehepcidin agonists may be sufficient to prevent the life-threatening iron overload in these patients. The first promis-ing evidence of the beneficial effect of hepcidin comes fromthe th3/+ mouse model of β-thalassemia in which moderatetransgenic expression of hepcidin resulted in lower spleenand liver iron content, decreased inefficient erythropoiesis inthe spleen, lower spleen weight, and even improvement ofhematological parameters (Gardenghi et al., submitted).

In chronically transfused patients, hepcidin concentra-tions are much higher than in nontransfused patients, pre-sumably due to both increased iron load and the alleviationof ineffective erythropoiesis [49, 51]. Interestingly, Origaet al. [51] showed that nontransfused patients have liveriron concentrations similar to those of regularly transfusedthalassemia major patients. However, because of the differenthepcidin levels, the cellular distribution of iron in the liverdiffered in these two groups. In nontransfused thalassemia,iron was deposited in hepatocytes, whereas higher hepcidinlevels in transfused patients resulted in macrophage ironloading. As a consequence of this difference in cellulariron distribution, serum ferritin levels were much lower innontransfused patients, and did not adequately reflect thepatients’ liver iron load. Considering that high hepcidinshifts iron distribution to macrophages and decreases intesti-nal iron absorption, it is possible that hepcidin agonists couldbe useful even in transfused thalassemia patients. Trappingiron in macrophages where it is less toxic may postpone irondeposition and consequent damage in the parenchyma, butthis remains to be investigated.

4.5. Anemia of Inflammation. Anemia of inflammation (AI)develops in the setting of many infections and inflammatorydisorders, and some malignancies [52]. Elevated hepcidinis believed to be an important mediator of AI but whetherhepcidin is a necessary factor in the AI pathogenesis hasnot yet been established with certainty. IRIDA and mousemodels overexpressing hepcidin demonstrate that elevated


hepcidin is sufficient to cause hypoferremia and anemia [4].Moderate overproduction of hepcidin in transgenic mice orin mice bearing hepcidin-producing tumors caused an iron-restricted anemia [4, 53]. Transgenic mice also had bluntederythropoietic response to EPO, another characteristic of AI.

In humans, elevated hepcidin was observed in a varietyof inflammatory disorders including rheumatologic diseases,inflammatory bowel disease, infections, multiple myeloma,and critical illness [13, 35–37, 54, 55]. The AI phenotypepresumably develops from hepcidin-mediated inhibition ofiron recycling and absorption. Decreased flow of iron intoplasma results in hypoferremia, and because most of theiron in plasma is destined for the bone marrow, lower ironavailability hinders hemoglobin synthesis and erythrocyteproduction.

Detecting elevated hepcidin in the presence of hypo-ferremia and anemia could help distinguish AI from IDA.However, mixed anemia is common, for example, in casesof chronic inflammatory illness with coexisting bleedingor malnutrition. In these conditions, the inflammation-mediated increase in hepcidin would be opposed by theeffects of iron deficiency. Even without blood loss ormalnutrition, when inflammation lasts for years, true irondeficiency may also develop because of the inhibitionof intestinal iron absorption by hepcidin. The ability ofhepcidin measurements to distinguish the contribution ofiron deficiency and inflammation in these conditions willhave to be carefully evaluated. It is possible that even“normal” levels of hepcidin in AI or mixed anemia areinappropriate and perpetuate iron restriction. Obesity, itselfa chronic inflammatory condition, was reported to beassociated with iron deficiency [56], and this can developdespite adequate iron supply and the absence of blood loss.Serum hepcidin concentrations were recently studied inobese premenopausal women [57] and were found to bein the reference range for the healthy iron-replete women.However, when compared to lean volunteers with similariron deficiency, hepcidin levels were several-fold higher inobese women, indicating they were inappropriately highconsidering the obese women’s iron deficiency. Thus, in verychronic mild inflammatory conditions even a mild hepcidinexcess may be sufficient to tip the balance between iron lossand iron uptake and lead to iron deficiency.

Studies of interventions that selectively reduce hepcidinwill clarify how essential the role of hepcidin is in each typeof inflammation-induced anemia. In a mouse model of AIcaused by injections of Brucella abortus, hepcidin antag-onists (neutralizing monoclonal antibody to hepcidin) incombination with erythropoiesis-stimulating agents restorednormal hemoglobin levels [58], even when erythropoiesis-stimulating agents alone were ineffective. Considering thevariety of other possible inflammatory erythroid pathologies,ranging from hemolysis to bone marrow suppression, not allinflammatory conditions may respond equally to antihep-cidin or combination therapies.

4.6. Anemia of Chronic Kidney Disease. Until recently, ane-mia of chronic kidney disease (CKD) was thought to beprimarily due to the deficiency of erythropoietin. Consider-

ing that high amounts of erythropoietin are often needed torestore erythropoiesis, this view has been challenged. This, aswell as the beneficial effect of high doses of intravenous ironin CKD patients has focused attention on the possible roleof hepcidin and iron restriction in the anemia of CKD [59].Renal excretion is a major route of hepcidin clearance. Whenkidney function is normal, urinary hepcidin concentrationscorrelate well with the circulating hepcidin levels, with noapparent regulation of the excretion process. However, lossof kidney function could decrease hepcidin clearance andlead to the accumulation of hepcidin and the developmentof iron-restrictive anemia.

Hepcidin concentrations were indeed reported to beincreased in patients with CKD [54, 60]. Although this couldbe caused by inflammation which frequently accompaniesCKD, even patients without significant inflammation hadelevated hepcidin which progressively increased with theincreasing severity of CKD [60]. Decreased kidney functionis the likely factor contributing to this phenomenon, andsome studies have reported inverse correlation betweenglomerular filtration rate and serum hepcidin [60, 61]. Thus,increased hepcidin levels due to decreased renal clearanceas well as due to inflammation may be a significant factorcontributing to the development of anemia in CKD andshould be considered in the development of new therapiesfor this disease.

4.7. Resistance to Erythropoietin. Hyporesponsiveness totherapeutic erythropoietin has emerged as an important con-sequence of inflammation, especially in chronic kidney dis-eases [62, 63], and may be a consequence of iron restrictionimposed by high hepcidin. As mentioned before, hepcidin-overexpressing transgenic mice had blunted response to Epo[4], and neutralization of hepcidin by monoclonal antibodyin the B. abortus model of AI restored responsiveness toerythropoietin [58]. Hepcidin measurements thus may beuseful for predicting patients’ response to erythropoietin butlarge studies will be necessary to test this concept.

Large pharmacological doses of Epo can sometimesovercome the resistance of AI to erythropoietin [63]. As wasdiscussed before, Epo injections in mice and humans resultedin suppression of hepcidin production [29, 30], and this maybe the mechanism by which high Epo levels overcome ironrestriction. It is therefore conceivable that administrationof anti-hepcidin therapies together with erythropoiesis-stimulating agents may improve patients’ erythropoieticresponse and enable the use of lower erythropoietin dosesto avoid the potential detrimental effects of high Epoconcentrations.

5. The Present and the Future of HepcidinDiagnostics and Therapeutics in Anemias

5.1. Diagnostics. The evaluation of the diagnostic potentialof hepcidin has only recently become possible with thedevelopment of assays for bioactive mature hepcidin inserum and urine. The methodologies include competitiveELISAs using biotinylated or radioiodinated hepcidin as


Table 1: Diagnostic and therapeutic potenital of hepcidin in different forms of anemia.

Condition Expected hepcidin levels Other iron parameters Hepcidin therapy

Iron deficiency anemia Low Low Tsat and ferritin

Iron-refractory iron deficiency anemia High Low Tsat and ferritin Antagonist

Iron-loading anemias Low (unless transfused) High Tsat and ferritin Agonist

Anemia of inflammation High Low Tsat, normal-to-elevated ferritin Antagonist

Mixed anemia (AI/IDA) Normal Low Tsat, low-to-normal ferritin Antagonist

Chronic kidney disease High Variable Antagonist

Erythropoietin resistance High Variable Antagonist

Tsat = transferrin saturation.

tracers [13, 64], and several mass spectrometry-based assays(MALDI-TOF MS, SELDI, and LC–MS/MS) using as internalstandards isotopically labeled hepcidin or truncated hepcidinvariants [55, 65, 66]. Measurements of prohepcidin in serumhave also been reported, but these did not correlate withmature hepcidin concentrations indicating that pro-hepcidinis inadequate as a substitute for measuring biologicallyrelevant hepcidin.

A summary of potential uses of hepcidin assays fordiagnosing different forms of anemia is listed in Table 1.However, the utility of hepcidin for the diagnosis andprognosis of iron disorders is far from understood andneeds to be evaluated in larger clinical studies. Some ofthe problems for interpreting hepcidin levels may includediurnal fluctuations of hepcidin (lower in the morning,higher in the afternoon) [13, 67], and relative sensitivity tothe available iron content of the diet [13].

Hepcidin concentrations may also predict which patientscould benefit from oral iron therapy. Several studies ofhealthy volunteers have shown inverse correlation betweenhepcidin concentrations and radiolabeled iron absorption[68, 69]. If these results are confirmed in appropriatepatient populations, patients with elevated hepcidin levelscould be treated with parenteral iron and thus avoidweeks of potentially burdensome and unsuccessful oral irontherapy.

5.2. Therapeutics. The potential use of hepcidin-targetedtherapeutics in anemias is summarized in Table 1. Althoughno hepcidin therapies are yet available, several candidatesare currently under development. Hepcidin agonists wouldbe useful for preventing or ameliorating iron overload innontransfused β-thalassemias and other iron-loading ane-mias. Small peptides based on hepcidin N-terminal regionhave been shown to act as agonists in mice in vivo (E.Nemeth, unpublished). The small size should allow eventualdevelopment of orally available agonists.

Hepcidin antagonists would be expected to benefitpatients with diseases of hepcidin excess (Table 1), mani-fested as iron-restricted anemia and eventually as systemiciron deficiency. Several approaches have been undertakento develop hepcidin antagonists. Hepcidin-neutralizing anti-body has already been successfully used in vivo in amouse model of AI [58]. Apart from directly interfering

with hepcidin activity, other agents which target pathwaysregulating hepcidin production have also been described.Dorsomorphin, a small-molecule inhibitor of BMP signal-ing, was shown to block hepcidin induction by iron in vivo[70]. Soluble HJV, also acting as an antagonist of BMP sig-naling, decreased hepcidin baseline expression in mice andconcurrently increased liver iron content [71]. Furthermore,some existing therapies may be acting in part by decreasinghepcidin production. Anticytokine therapies such as anti-IL-6 antibody were shown to suppress hepcidin produc-tion and improve anemia [72, 73]. Some erythropoiesis-stimulating agents, such as prolyl hydroxylase inhibitors,could also be effective hepcidin suppressors, not only bystimulating erythropoiesis, but also by interfering with theHIF pathway [74]. Undoubtedly, future studies will assessthe risks and relative benefits of hepcidin-targeted treatmentapproaches.

6. Conclusion

Recent advances in the understanding of the key role ofhepcidin-ferroportin interaction in iron homeostasis and itsdisorders have clarified the pathogenesis of anemias due toiron restriction as well as anemias accompanied by iron-loading. While many mechanistic details of hepcidin andferroportin regulation remain to be worked out, we will soonsee medical applications of these advances. In the comingyears, the role of hepcidin assays in the diagnosis, prognosis,and therapeutic stratification of anemias will be explored.Therapeutic interventions specifically targeting the hepcidin-ferroportin axis for the treatment of anemias are also underdevelopment.

Disclosure

Dr. E. Nemeth is a Cofounder and the Chief Scientific Officerof Intrinsic Life Sciences, LLC, a biotech company developingiron-related diagnostics.

Acknowledgments

This work was supported by the Grant from the NationalInstitute of Health, R01 DK082717.


References

[1] C. H. Park, E. V. Valore, A. J. Waring, and T. Ganz, “Hepcidin,a urinary antimicrobial peptide synthesized in the liver,”Journal of Biological Chemistry, vol. 276, no. 11, pp. 7806–7810, 2001.

[2] B. Gagliardo, N. Kubat, A. Faye, et al., “Pro-hepcidin is unableto degrade the iron exporter ferroportin unless maturated bya furin-dependent process,” Journal of Hepatology, vol. 50, no.2, pp. 394–401, 2009.

[3] E. V. Valore and T. Ganz, “Posttranslational processing of hep-cidin in human hepatocytes is mediated by the prohormoneconvertase furin,” Blood Cells, Molecules, and Diseases, vol. 40,no. 1, pp. 132–138, 2008.

[4] C. N. Roy, H. H. Mak, I. Akpan, G. Losyev, D. Zurakowski, andN. C. Andrews, “Hepcidin antimicrobial peptide transgenicmice exhibit features of the anemia of inflammation,” Blood,vol. 109, no. 9, pp. 4038–4044, 2007.

[5] A. Roettol, G. Papanikolaou, M. Politou, et al., “Mutantantimicrobial peptide hepcidin is associated with severejuvenile hemochromatosis,” Nature Genetics, vol. 33, no. 1, pp.21–22, 2003.

[6] A. Donovan, C. A. Lima, J. L. Pinkus, et al., “The iron exporterferroportin/Slc40a1 is essential for iron homeostasis,” CellMetabolism, vol. 1, no. 3, pp. 191–200, 2005.

[7] S. Abboud and D. J. Haile, “A novel mammalian iron-regulatedprotein involved in intracellular iron metabolism,” Journal ofBiological Chemistry, vol. 275, no. 26, pp. 19906–19912, 2000.

[8] A. Donovan, A. Brownlie, Y. Zhou, et al., “Positional cloningof zebrafish ferroportin1 identifies a conserved vertebrate ironexporter,” Nature, vol. 403, no. 6771, pp. 776–781, 2000.


[10] E. Nemeth, M. S. Tuttle, J. Powelson, et al., “Hepcidin regulatescellular iron efflux by binding to ferroportin and inducing itsinternalization,” Science, vol. 306, no. 5704, pp. 2090–2093,2004.

[11] S. Rivera, E. Nemeth, V. Gabayan, M. A. Lopez, D. Farshidi,and T. Ganz, “Synthetic hepcidin causes rapid dose-dependenthypoferremia and is concentrated in ferroportin-containingorgans,” Blood, vol. 106, no. 6, pp. 2196–2199, 2005.

[12] D.-L. Zhang, R. M. Hughes, H. Ollivierre-Wilson, M. C.Ghosh, and T. A. Rouault, “A ferroportin transcript thatlacks an iron-responsive element enables duodenal and ery-throid precursor cells to evade translational repression,” CellMetabolism, vol. 9, no. 5, pp. 461–473, 2009.

[13] T. Ganz, G. Olbina, D. Girelli, E. Nemeth, and M. Westerman,“Immunoassay for human serum hepcidin,” Blood, vol. 112,no. 10, pp. 4292–4297, 2008.

[14] M. Pak, M. A. Lopez, V. Gabayan, T. Ganz, and S. Rivera, “Sup-pression of hepcidin during anemia requires erythropoieticactivity,” Blood, vol. 108, no. 12, pp. 3730–3735, 2006.

[15] T. Ganz, “Molecular pathogenesis of anemia of chronicdisease,” Pediatric Blood and Cancer, vol. 46, no. 5, pp. 554–557, 2006.

[16] Y. Xia, J. L. Babitt, Y. Sidis, R. T. Chung, and H. Y. Lin, “Hemo-juvelin regulates hepcidin expression via a selective subset ofBMP ligands and receptors independently of neogenin,” Blood,vol. 111, no. 10, pp. 5195–5204, 2008.

[17] B. Andriopoulos Jr., E. Corradini, Y. Xia, et al., “BMP6 isa key endogenous regulator of hepcidin expression and ironmetabolism,” Nature Genetics, vol. 41, no. 4, pp. 482–487,2009.

[18] D. Meynard, L. Kautz, V. Darnaud, F. Canonne-Hergaux, H.Coppin, and M.-P. Roth, “Lack of the bone morphogeneticprotein BMP6 induces massive iron overload,” Nature Genet-ics, vol. 41, no. 4, pp. 478–481, 2009.


[20] L. Lin, Y. P. Goldberg, and T. Ganz, “Competitive regulation ofhepcidin mRNAby soluble and cell-associated hemojuvelin,”Blood, vol. 106, no. 8, pp. 2884–2889, 2005.

[21] V. Niederkofler, R. Salie, and S. Arber, “Hemojuvelin isessential for dietary iron sensing, and its mutation leads tosevere iron overload,” Journal of Clinical Investigation, vol. 115,no. 8, pp. 2180–2186, 2005.

[22] L. Silvestri, A. Pagani, A. Nai, I. de Domenico, J. Kaplan, and C.Camaschella, “The serine protease matriptase-2 (TMPRSS6)inhibits hepcidin activation by cleaving membrane hemoju-velin,” Cell Metabolism, vol. 8, no. 6, pp. 502–511, 2008.

[23] A.-S. Zhang, F. Yang, K. Meyer, et al., “Neogenin-mediatedhemojuvelin shedding occurs after hemojuvelin traffics to theplasma membrane,” Journal of Biological Chemistry, vol. 283,no. 25, pp. 17494–17502, 2008.

[24] L. Kautz, D. Meynard, A. Monnier, et al., “Iron regulatesphosphorylation of Smad1/5/8 and gene expression of Bmp6,Smad7, Id1, and Atoh8 in the mouse liver,” Blood, vol. 112, no.4, pp. 1503–1509, 2008.

[25] M. B. Johnson, J. Chen, N. Murchison, F. A. Green, and C.A. Enns, “Transferrin receptor 2: evidence for ligand-inducedstabilization and redirection to a recycling pathway,” MolecularBiology of the Cell, vol. 18, no. 3, pp. 743–754, 2007.

[26] G. Ramey, J.-C. Deschemin, and S. Vaulont, “Cross-talkbetween the mitogen activated protein kinase and bonemorphogenetic protein/hemojuvelin pathways is required forthe induction of hepcidin by holotransferrin in primarymouse hepatocytes,” Haematologica, vol. 94, no. 6, pp. 765–772, 2009.

[27] T. Goswami and N. C. Andrews, “Hereditary hemochro-matosis protein, HFE, interaction with transferrin receptor2 suggests a molecular mechanism for mammalian ironsensing,” Journal of Biological Chemistry, vol. 281, no. 39, pp.28494–28498, 2006.

[28] A. Piperno, D. Girelli, E. Nemeth, et al., “Blunted hepcidinresponse to oral iron challenge in HFE-related hemochro-matosis,” Blood, vol. 110, no. 12, pp. 4096–4100, 2007.

[29] D. R. Ashby, D. P. Gale, M Busbridge, et al., “Erythropoietinadministration in humans causes a marked and prolongedreduction in circulating hepcidin,” Haematologica. In press.

[30] H. Huang, M. Constante, A. Layoun, and M. M. Santos, “Con-tribution of STAT3 and SMAD4 pathways to the regulationof hepcidin by opposing stimuli,” Blood, vol. 113, no. 15, pp.3593–3599, 2009.

[31] M. Vokurka, J. Krijt, K. Sulc, and E. Necas, “Hepcidin mRNAlevels in mouse liver respond to inhibition of erythropoiesis,”Physiological Research, vol. 55, no. 6, pp. 667–674, 2006.




[34] C. Peyssonnaux, A. S. Zinkernagel, R. A. Schuepbach, et al.,“Regulation of iron homeostasis by the hypoxia-inducibletranscription factors (HIFs),” Journal of Clinical Investigation,vol. 117, no. 7, pp. 1926–1932, 2007.

[35] G. Semrin, D. S. Fishman, A. Bousvaros, et al., “Impairedintestinal iron absorption in Crohn’s disease correlates withdisease activity and markers of inflammation,” InflammatoryBowel Diseases, vol. 12, no. 12, pp. 1101–1106, 2006.

[36] S. Sharma, E. Nemeth, Y.-H. Chen, et al., “Involvement ofhepcidin in the anemia of multiple myeloma,” Clinical CancerResearch, vol. 14, no. 11, pp. 3262–3267, 2008.

[37] E. Nemeth, E. V. Valore, M. Territo, G. Schiller, A. Lichtenstein,and T. Ganz, “Hepcidin, a putative mediator of anemia ofinflammation, is a type II acute-phase protein,” Blood, vol. 101,no. 7, pp. 2461–2463, 2003.

[38] M. V. Verga Falzacappa, M. Vujic Spasic, R. Kessler, J. Stolte, M.W. Hentze, and M. U. Muckenthaler, “STAT3 mediates hepatichepcidin expression and its inflammatory stimulation,” Blood,vol. 109, no. 1, pp. 353–358, 2007.

[39] E. Nemeth, S. Rivera, V. Gabayan, et al., “IL-6 mediateshypoferremia of inflammation by inducing the synthesis ofthe iron regulatory hormone hepcidin,” Journal of ClinicalInvestigation, vol. 113, no. 9, pp. 1271–1276, 2004.

[40] S. Rivera, V. Gabayan, and T. Ganz, “In chronic inflammation,there exists an IL-6 independent pathway for the induction ofhepcidin,” vol. 104, no. 11, 2004, Abstract no. 3205.

[41] C. Peyssonnaux, A. S. Zinkernagel, V. Datta, X. Lauth, R. S.Johnson, and V. Nizet, “TLR4-dependent hepcidin expressionby myeloid cells in response to bacterial pathogens,” Blood, vol.107, no. 9, pp. 3727–3732, 2006.

[42] S. Sazawal, R. E. Black, M. Ramsan, et al., “Effects of routineprophylactic supplementation with iron and folic acid onadmission to hospital and mortality in preschool childrenin a high malaria transmission setting: community-based,randomised, placebo-controlled trial,” The Lancet, vol. 367,no. 9505, pp. 133–143, 2006.

[43] M. A. Melis, M. Cau, R. Congiu, et al., “A mutationin the TMPRSS6 gene, encoding a transmembrane serineprotease that suppresses hepcidin production, in familial irondeficiency anemia refractory to oral iron,” Haematologica, vol.93, no. 10, pp. 1473–1479, 2008.

[44] K. E. Finberg, M. M. Heeney, D. R. Campagna, et al.,“Mutations in TMPRSS6 cause iron-refractory iron deficiencyanemia (IRIDA),” Nature Genetics, vol. 40, no. 5, pp. 569–571,2008.

[45] J. C. Chambers, W. Zhang, Y. Li, et al., “Genome-wideassociation study identifies variants in TMPRSS6 associatedwith hemoglobin levels,” Nature Genetics, vol. 41, no. 11, pp.1170–1172, 2009.

[46] B. Benyamin, M. A. R. Ferreira, G. Willemsen, et al., “Com-mon variants in TMPRSS6 are associated with iron statusand erythrocyte volume,” Nature Genetics, vol. 41, no. 11, pp.1173–1175, 2009.

[47] D. A. Weinstein, C. N. Roy, M. D. Fleming, M. F. Loda, J. I.Wolfsdorf, and N. C. Andrews, “Inappropriate expression ofhepcidin is associated with iron refractory anemia: implica-tions for the anemia of chronic disease,” Blood, vol. 100, no.10, pp. 3776–3781, 2002.

[48] A. Y. F. Chung, K. W. Leo, G. C. Wong, K. L. Chuah, J. W. Ren,and C. G. L. Lee, “Giant hepatocellular adenoma presenting

with chronic iron deficiency anemia,” American Journal ofGastroenterology, vol. 101, no. 9, pp. 2160–2162, 2006.


[50] E. Nemeth and T. Ganz, “Hepcidin and iron-loading anemias,”Haematologica, vol. 91, no. 6, pp. 727–732, 2006.

[51] R. Origa, R. Galanello, T. Ganz, et al., “Liver iron concentra-tions and urinary hepcidin in β-thalassemia,” Haematologica,vol. 92, no. 5, pp. 583–588, 2007.

[52] G. E. Cartwright, “The anemia of chronic disorders,” Seminarsin Hematology, vol. 3, no. 4, pp. 351–375, 1966.

[53] S. Rivera, L. Liu, E. Nemeth, V. Gabayan, O. E. Sorensen, and T.Ganz, “Hepcidin excess induces the sequestration of iron andexacerbates tumor-associated anemia,” Blood, vol. 105, no. 4,pp. 1797–1802, 2005.

[54] H. Li, M. J. Rose, L. Tran, et al., “Development of a methodfor the sensitive and quantitative determination of hepcidin inhuman serum using LC-MS/MS,” Journal of Pharmacologicaland Toxicological Methods, vol. 59, no. 3, pp. 171–180, 2009.

[55] E. H. J. M. Kemna, H. Tjalsma, V. N. Podust, and D. W.Swinkels, “Mass spectrometry-based hepcidin measurementsin serum and urine: analytical aspects and clinical implica-tions,” Clinical Chemistry, vol. 53, no. 4, pp. 620–628, 2007.

[56] L. M. Tussing-Humphreys, H. Liang, E. Nemeth, S. Freels,and C. A. Braunschweig, “Excess adiposity, inflammation, andiron-deficiency in female adolescents,” Journal of the AmericanDietetic Association, vol. 109, no. 2, pp. 297–302, 2009.

[57] L. M. Tussing-Humphreys, E. Nemeth, G. Fantuzzi, et al.,“Elevated systemic hepcidin and iron depletion in obesepremenopausal females,” Obesity. In press.

[58] B. J. Sasu, M. Hainu, T. C. Boone, et al., “Hepcidin,hepcidin antagonists and methods of use,” US patent no.US2008213277, Amgen Inc.Thousand Oaks, Calif, USA,September 2008.

[59] B. Young and J. Zaritsky, “Hepcidin for clinicians,” ClinicalJournal of the American Society of Nephrology, vol. 4, no. 8, pp.1384–1387, 2009.

[60] J. Zaritsky, B. Young, H. J. Wang, et al., “Hepcidin—a potentialnovel biomarker for iron status in chronic kidney disease,”Clinical Journal of the American Society of Nephrology, vol. 4,no. 6, pp. 1051–1056, 2009.

[61] D. R. Ashby, D. P. Gale, M. Busbridge, et al., “Plasma hepcidinlevels are elevated but responsive to erythropoietin therapy inrenal disease,” Kidney International, vol. 75, no. 9, pp. 976–981,2009.

[62] J. W. Adamson, “Hyporesponsiveness to erythropoiesis stim-ulating agents in chronic kidney disease: the many faces ofinflammation,” Advances in Chronic Kidney Disease, vol. 16,no. 2, pp. 76–82, 2009.

[63] J. Elliott, D. Mishler, and R. Agarwal, “Hyporesponsiveness toerythropoietin: causes and management,” Advances in ChronicKidney Disease, vol. 16, no. 2, pp. 94–100, 2009.

[64] M. Busbridge, C. Griffiths, D. R. Ashby, et al., “Developmentof a novel immunoassay for the iron regulatory peptidehepcidin,” British Journal of Biomedical Science, vol. 66, no. 3,pp. 150–157, 2009.

[65] A. Castagna, N. Campostrini, F. Zaninotto, and D. Girelli,“Hepcidin assay in serum by SELDI-TOF-MS and otherapproaches,” Journal of Proteomics. In press.

[66] S. S. Bansal, J. M. Halket, J. Fusova, et al., “Quantificationof hepcidin using matrix-assisted laser desorption/ ionizationtime-of-flight mass spectrometry,” Rapid Communications inMass Spectrometry, vol. 23, no. 11, pp. 1531–1542, 2009.


[67] J. J. C. Kroot, J. C. M. Hendriks, C. M. M. Laarakkers, et al.,“(Pre)analytical imprecision, between-subject variability, anddaily variations in serum and urine hepcidin: implications forclinical studies,” Analytical Biochemistry, vol. 389, no. 2, pp.124–129, 2009.

[68] M. F. Young, R. P. Glahn, M. Ariza-Nieto, et al., “Serumhepcidin is significantly associated with iron absorption fromfood and supplemental sources in healthy young women,”American Journal of Clinical Nutrition, vol. 89, no. 2, pp. 533–538, 2009.

[69] M. Ruivard, F. Laine, T. Ganz, et al., “Iron absorption in dys-metabolic iron overload syndrome is decreased and correlateswith increased plasma hepcidin,” Journal of Hepatology, vol.50, no. 6, pp. 1219–1225, 2009.

[70] P. B. Yu, C. C. Hong, C. Sachidanandan, et al., “Dorsomorphininhibits BMP signals required for embryogenesis and ironmetabolism,” Nature Chemical Biology, vol. 4, no. 1, pp. 33–41, 2008.

[71] J. L. Babitt, F. W. Huang, Y. Xia, Y. Sidis, N. C. Andrews, and H.Y. Lin, “Modulation of bone morphogenetic protein signalingin vivo regulates systemic iron balance,” Journal of ClinicalInvestigation, vol. 117, no. 7, pp. 1933–1939, 2007.

[72] N. Nishimoto, Y. Kanakura, K. Aozasa, et al., “Humanizedanti-interleukin-6 receptor antibody treatment of multicentricCastleman disease,” Blood, vol. 106, no. 8, pp. 2627–2632,2005.

[73] H. Kawabata, N. Tomosugi, J. Kanda, Y. Tanaka, K. Yoshizaki,and T. Uchiyama, “Anti-interleukin 6 receptor antibodytocilizumab reduces the level of serum hepcidin in patientswith multicentric Castleman’s disease,” Haematologica, vol. 92,no. 6, pp. 857–858, 2007.

[74] S. Klaus, M. Arend, P. Fourney, et al., “Induction of ery-thropoiesis and iron utilization by the HIF prolyl hydroxylaseinhibitor FG-4592,” Journal of the American Society of Nephrol-ogy, vol. 16, p. 49A, 2005.


Review Article

Ferroportin and Erythroid Cells: An Update

Luciano Cianetti, Marco Gabbianelli, and Nadia Maria Sposi

Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanita,Viale Regina Elena 299, 00161 Rome, Italy

Correspondence should be addressed to Nadia Maria Sposi, [email protected]

Received 31 December 2009; Revised 8 April 2010; Accepted 23 June 2010

Academic Editor: Elizabeta Nemeth

Copyright © 2010 Luciano Cianetti et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

In recent years there have been major advances in our knowledge of the regulation of iron metabolism that have had implicationsfor understanding the pathophysiology of some human disorders like beta-thalassemia and other iron overload diseases. However,little is known about the relationship among ineffective erythropoiesis, the role of iron-regulatory genes, and tissue irondistribution in beta-thalassemia. The principal aim of this paper is an update about the role of Ferroportin during humannormal and pathological erythroid differentiation. Particular attention will be given to beta-thalassemia and other diseases withiron overload. Recent discoveries indicate that there is a potential for therapeutic intervention in beta-thalassemia by means ofmanipulating iron metabolism.

1. Identification, Tissue Distribution,and Structural Features of Ferroportin

Ferroportin (FPN1, also known as Ireg 1 or MTP1), theproduct of SLC40A1 gene, was independently identified bythree groups, using different approaches [1–3]. FPN1 hasbeen reported to be expressed and to play a critical rolein several different tissues involved in mammalian ironhomeostasis, including duodenal enterocytes (iron uptakeand export into circulation), hepatocytes (storage), syncy-tiotrophoblasts (transfer to embryo) and reticuloendothelialmacrophages (iron recycling from senescent red blood cells)[4]. FPN1 appears to act as an iron exporter [2, 3] and tobe specifically regulated according to body iron requirementsin these tissues [2, 4–9]. The FPN1 gene is highly conservedduring evolution and encodes for a protein 571 aa inlength with a predicted mass of 62 KDa [1, 3]. Sequencedata showed that FPN1 is a multipass integral membraneprotein iron exporter and has at least nine transmembranealphahelices [1–3]. The locations of N- and C-termini havebeen largely debated in previous studies indicating for oneor both termini an extracellular [10–12] or an intracellularlocation [13–15] (Figure 1). Different results have also beenobtained for the membrane topology of FPN1 and the

number of its TM domains [2, 3, 13, 16] (Figure 1). Finally,the oligomeric state of FPN1 has also been debated forseveral years: the protein has been reported to be a monomer[12, 15, 17] as well as a dimer/multimer [14, 18]. A recentstudy by using recombinant expression of FPN1 in insectcells and a biophysical characterization of purified detergent-solubilized FPN1 showed that FPN1 protein is a monomer,having 12 transmembrane regions and N- and C-terminiboth cytosolic [19]. In the 5′-UTR of FPN1 mRNA a putativeiron responsive element (IRE) was found that could confer atranslational regulation by iron regulatory proteins (IRPs) ina manner similar to other 5′-UTR-IRE-regulated genes, thatis, ferritin, erythroid δ-aminolevulinate synthase (ALAS2)and mitochondrial aconitase [1, 20]. The 5′-UTR-FPN1-IREwas responsive to iron in HepG2 and CaCo2 cells [21]; invitro iron deprivation inhibited translational efficiency ofFPN1 mRNA [4, 6, 22]. However, the regulation of FPN1expression by iron is currently poorly understood and adirect proof of IRP-IRE control has not been provided yet.Both transcriptional and post-transcriptional mechanismshave been implicated in the regulation of FPN1 inducedby changes in cellular iron status [2, 23]. Some authorsdemonstrated that hepcidin, a major regulator of ironmetabolism, binds to FPN1 in tissue culture cells, resulting


in internalization and degradation of FPN1 and in decreasedexport of cellular iron [24]. The post-translational regulationof FPN1 by hepcidin may thus complete a homeostatic loop:iron regulates secretion of hepcidin, which then reducesexport of cellular iron [24].

2. Ferroportin and Iron Overload Disorders

Ferroportin disease, or type 4 hemochromatosis, or HFE4, isan autosomal dominant condition with heterozygous muta-tions in the FPN1 gene [23]. Hemochromatosis associatedwith mutation in FPN1 can result in two different types ofiron loading: one type is phenotypically indistinguishablefrom classical HFE hemochromatosis, in that the patientshave both an elevated transferrin saturation and serumferritin, while the other type termed “ferroportin disease”is associated with microcytic anemia, a raised serum ferritinand iron deposition in macrophages rather than hepatocytes[23]. FPN1 mutations may have three possible effects:causing misfolding of the protein and failure to reachthe cell surface (“loss of function”) [10] or producing amutant protein that is expressed at the cell surface but isnot inhibited by hepcidin (“loss of regulation”) [11], oraffecting iron transport ability [18]. Briefly it was shownthat A77D, V162del, G490D, and D157G mutations, that areassociated with typical pattern of disease in vivo, cause aloss of iron export function in vitro, but do not physicallyor functionally impede FPN1 protein coded by the wild-type allele [10, 11, 18] and, therefore, lead to diseaseby haploinsufficieny. These results are consistent with thescheme proposed by Montosi et al. [25] to explain the clinicalphenotype observed in patients with these mutations [10]:lower level of serum iron, resulting from iron sequestrationin macrophages, reduces availability to the bone marrowfor erythropoiesis, thus leading to a middle anemia in theearly stages of disease that was effectively observed in somepatients and that respond poorly to phlebotomy [10, 26].So iron overload may be a consequence of the erythronsignalling to the gut enterocyte to increase iron uptakefrom the diet to compensate for the anemia. According torecent progress in this field it is likely that the erythronsignalling is directly working through hepcidin-ferroportininteraction. By contrast the Y64N, N144D, Q248H, andC326Y mutations, which can be associated with greatertransferrin saturation and more prominent iron depositionin liver parenchyma in vivo, retain iron export functionin vitro [10, 11]. It was postulated that this group ofmutations may resist inhibition by hepcidin, so interferingwith its homeostatic negative feedback loop and resulting ina permanently “turned on” iron exporter [10, 11].

3. Overall View on Hematopoiesis andImportance of Iron Homeostasis duringErythroid Cells Differentiation

Iron homeostasis depends on a coordinated regulation ofmolecules involved in the import of this element and those

exporting it out of the cells. In some cell types, such as ery-throid cells, iron import mechanisms are highly expressed,thus allowing massive iron uptake [27, 28]. Excessive iron,however, may be toxic for these cells, particularly in view ofits capacity to generate superoxide radicals and H2O2, whichmay freely diffuse into the nucleus resulting in cell damage[29] and it seemed therefore of interest to investigate whethererythroid cells possess specific mechanisms for iron export.Within the hematopoietic differentiation, the maintenanceof iron homeostasis is essential for erythroid cells andmacrophages. Erythroid cells need to incorporate very highamounts of iron to support the continued synthesis of hemeand hemoglobin, while the macrophage cells play a key rolein iron storage and recycling [30–32]. Human erythropoiesisis a dynamic complex multistep process that involves differ-entiation of pluripotent hematopoietic stem cells (HSCs) andearly multipotent progenitors (MPPs) to generate committederythroid precursors, the erythroblasts, which then give riseto mature erythrocytes, that is, the red blood cells (RBCs)[33–36]. Briefly, the early erythroid progenitors (BFU-E,burst-forming units-erythroid) differentiate into late colony-forming units erythroid (CFU-E) and proerythroblasts fol-lowed by a progressive wave of erythroblast maturation inpolychromatic and orthochromatic erythroblasts coupledwith a gradual increase of erythroid-specific markers. As thehematopoietic process progresses from the early stages intoerythroid cell maturation, cells gradually lose their potentialfor cell proliferation and become mature enucleated cells.Mature erythrocytes are biconcave disks without mitochon-dria and other organelles but full of hemoglobin able tobind and deliver O2 [33, 37–39]. The hematopoietic differ-entiation is a highly complex system in which, from a poolof totipotent stem cells, originate all the cells of peripheralblood [40–43]. Developmentally, hematopoiesis in humansis characterized by three fundamental periods of activityprogressively involving the yolk sac, fetal liver and bonemarrow [44]. The survival, proliferation and differentiationof hematopoietic stem and progenitor cells are regulatedby a complex network of hematopoietic growth factorscollectively known as colony stimulating factors (CSFs),interleukins (ILs) or hemopoietins that are released fromaccessory cells such as fibroblasts, macrophages, lymphocytesand endothelial cells. Depending on their mechanism ofaction during hematopoietic differentiation, these factorscan be classified into three categories: the first categoryincludes growth factors that exert their action at the earlieststages of hematopoiesis, for example, the c-kit receptorligand (KL) or stem cell factor (SCF) [45], FLT-3 ligand(FL) [46, 47], the basic fibroblast growth factor (bFGF)[48, 49], and interleukin-6 (IL-6) [50]; in the second categoryare growth factors acting as multilineage, whose prototypesare the IL-3 and GM-CSF that are able to stimulateprimitive progenitors to proliferate and differentiate into allhematopoietic lineage [51]; and finally in the third categoryare included the growth factors acting as unilineage, that is,those that stimulate the differentiation and proliferation of asingle lineage and include erythropoietin (EPO) [52, 53], thegranulocytic growth factor (G-CSF) [54], monocytic growthfactor (M-CSF) [55] and thrombopoietin (TPO) [56]. These


G80S

A77DV72F

N144TN144DN144H

D157G

N174l

V162delQ182H

Exterior

G323V

C326SC326Y

G490DD270VQ248H

Y64N

Y64NV72F

Exterior

G323V

C326SC326Y

N

A77DG80S

N144TN144DN144H

D157GV162del

N174l Q182H Q248H

D270D

G490VC

1 2 3 4 5 6 7 8 9

1 2 3 4 5 6 7 8 9 10 11 12

C

N 57 1285

1348

2216

33 78 11 2133

17 57 15 57 33

16 11 3 1856 5

Figure 1: Membrane topology of FPN1. Topology of FPN1 protein is schematically represented, modified from two alternative modelsproposed by Devalia et al. (on the top) [16] and Liu et al. (on the bottom) [13]: the 9 or 12 predicted transmembrane helices (verticalgreen rectangles) are shown in relation to the lipid bilayer (horizontal violet rectangle). The positions of the mutations are marked as orangecircles. The N- and C-termini are denoted by N and C, respectively. The length of extra-membranous segments is indicated.

unilineage factors act on progenitors already moving towardan hematopoietic lineage and promote the production ofmature cells in the circulating blood, that is, erythro-cytes, neutrophils, eosinophils, monocytes/macrophages andmegakaryocytes. The circulating red cell mass is maintainedconstant by a homeostatic mechanism regulating erythro-poiesis, based on an erythropoietic stimulus which ensuresthat, under physiological conditions, the production of redblood cells equals their destruction. Moreover, in responseto hypoxia, hemorrhage or hemolysis, this stimulus causesincrease in the production of red blood cells [30]. Senescentor damaged erythrocytes are phagocytized by macrophagesof reticuloendothelial system that play a key role in recyclingiron from hemoglobin [30].

4. The Presence of Ferroportin inErythroid Cells

We reported for the first time the expression of FPN1 mRNAand protein in normal human erythroid cells at all stagesof differentiation [57]. This finding was very surprising,because the presence of a transporter, that in the other tissuesis appointed to export iron, was not expected for red bloodcells, in which iron is believed to be all retained into the cell,

committed for heme synthesis, and to exit only after deathof these cells by macrophage phagocytosis. The IRE elementin the 5′-UTR of FPN1 mRNA was demonstrated to befunctional in erythroid cells and able to mediate translationalmodulation by cellular iron levels [57]. Nonetheless, FPN1protein expression appeared to maintain a constant levelduring different steps of erythroid differentiation and afteriron treatments [57]. This paradox would have been difficultto explain, unless of supposing the existence of a FPN1transcript noncontaining the IRE element in erythroid cells.Actually previous studies already indicated the possibilityof an IRE-independent regulation of FPN1 in differenttissues and cell types, that is, iron deficiency was reportedto induce in mouse, human and rat duodenum, both invivo and in vitro, a significant increase of FPN1 mRNAexpression [3, 7, 35, 58]. A nonIRE FPN1 transcript hasbeen previously only described in polycythaemia mice asan aberrant mRNA resulting from a microdeletion in theFPN1 gene promoter [6, 59]. So, we also described for thefirst time the existence of two alternative FPN1 transcripts(variant II and III), other than the IRE-containing canonicalone (variant I), that did not contain the IRE element intheir 5′-UT region, did not respond to iron treatmentsand together accounted for more than half of total FPN1


mRNA present in erythroid cells [57] (Figure 2). Thesetranscripts arise from the usage of alternative upstreampromoters and differential splicing of 5′-UTR sequences.Interestingly, these transcripts were expressed mainly dur-ing the middle steps (4–11 days) of in vitro erythroiddifferentiation, corresponding to the maturation from lateerythroid progenitors to polychromatophilic erythroblasts(Figure 3). At these stages of erythroid differentiation TfR1,the receptor responsible for iron incoming in erythroidcells, is strongly and increasingly expressed [28]. Therefore,the nonIRE (variant II and III) FPN1 transcripts wereexpressed when erythroid progenitor/precursor cells needto accumulate iron into the cells [57]. We speculated thatexpression of the nonIRE FPN1 transcripts could producea constant level of the transporter, unresponsive to thevery high-iron levels present in maturing erythroid cell.In contrast, IRE-containing FPN1 transcripts were mainlyexpressed in undifferentiated erythroid progenitors and inmature terminal erythroblasts, suggesting a possible role atthese particular stages of erythroid differentiation [57]. Theexistence of multiple FPN1 alternative transcripts indicateda complex regulation of the FPN1 gene in erythroid cellsand the possibility that the control of FPN1 expression byiron conditions in different cell types might be complex. Wealso speculated that in erythroid cells the regulation of FPN1mRNA translation through the 5′-UTR IRE mechanismmight be silenced because in this cell type a high levelof iron uptake is needed to accumulate high amountsof iron required for optimal heme synthesis. A solutionfor this problem might be the utilization of an upstreamalternative promoter to produce mRNA species in which the5′-UTR IRE might be spliced out or made nonfunctional[57] (Figure 2). Our results showed that alternative FPN1transcripts are differentially expressed during erythroid dif-ferentiation, in particular indicating a sequential and specificactivation pathway, with an apparently mutual exclusionbetween variant I IRE and variant II/III not containingthe IRE transcripts [57]. These observations suggest thaterythroid precursor cells need FPN1 transcript without aIRE to evade translational control by IRP-IRE system inorder to export iron during the critical period when cellsare committed to proliferate and differentiate. Once theprecursor erythroid cells begin to produce hemoglobin,FPN1 without a IRE diminishes and FPN1 with a IREpredominates allowing erythroid cells to limit iron exportthrough the IRP-IRE system and synthesize heme withoutdeveloping microcytic anemia. Comparison between thesequence of our variant II mRNA and aberrant nonIREFPN1 transcript previously reported in polycythaemia mice[59] indicated a strong homology, thus strengthening ourhypothesis. Recently other authors have demonstrated thatalso mouse duodenal epithelial cells utilized an alternativeupstream promoter to express a FPN1 transcript, namedFPN1B, which lacks the IRE, is not repressed in irondefi-cient conditions and enables duodenal enterocytes to evadetranslational repression [60]. Enterocyte is a particular typeof cell because it must provide iron to satisfy systemiciron demands regardless of whether enterocyte itself is irondepleted [60]. The authors have so formulated a satisfactory

model of why FPN1B is significantly expressed in duodenum.According to this model in iron-replete conditions bothFPN1A and FPN1B transcripts are translated into FPN1protein, which traffics to the basolateral membrane totransport iron into the circulation [60]. When the ironstores are high, the liver produces hepcidin, which causesferroportin degradation and blocks iron absorption [24]. Onthe contrary in iron-deficient conditions, the liver ceases toproduce hepcidin and the degradation of FPN1 is eliminated[60]. So the iron deficiency activates the IRE/IRP systemwhich in turn blocks FPN1A translation via the IRE element.So it appears that FPN1B transcript has a key physiologic rolein duodenal cells: translation of FPN1B is not repressed byIRPs allowing sufficient iron export to satisfy the systemiciron demands [60]. They also demonstrated the presence ofFPN1B transcript in mouse bone marrow and in erythroidprecursor MEL and G1E cell lines [60] showing a regulationcomparable with our no-IRE FPN1 previously describedin human, thus supporting our hypothesis that ferroportinmay be subject to different regulation depending on celltype and its functions [57]. This mouse FPN1B transcriptwas homologous to our nonIRE variant II FPN1 transcriptobserved in human progenitor erythroid cells suggestingthat the utilization of an upstream alternative promoter toproduce mRNA species without IRE could be a physiologicand tissue-specific regulation mechanism conserved duringmammalian evolution. The identification of FPN1B revealshow FPN1 expression can bypass IRP-dependent repressionin intestinal iron uptake, even when cells throughout thebody are iron deficient [60]. Finally, in erythroid precursorcells, they hypothesized that FPN1B expression enhancesreal-time sensing of systemic iron status and facilitatesrestriction of erythropoiesis in response to low-systemic ironin order to not create microcythemia [60]. The existence ofFPN1 alternative transcripts unresponsive to regulation byIRE-IRP system in the duodenum and erythroid precursorcells suggests a cell type and tissue-specific control ofFPN1 expression by systemic iron status. Previous evidencefor differential effects of hepcidin in machrophages andintestinal epithelial cells can thus be explained [61].

5. Evidence for a Link between Erythropoiesisand Ferroportin-Hepcidin Way Regulation

In recent years there has been important advancement inour understanding of iron metabolism, mainly as a resultof the discovery of hepcidin [62–64], a key regulator ofwhole body iron homeostasis (for an exhaustive review see[65–67]). Increasing experimental evidence suggested that asingle molecule could be the “stores”, the “erythropoietic”and the “inflammation” regulator of iron absorption andrecycling [68, 69], and that hepcidin acted principally orsolely by binding to ferroportin, the only known cellulariron exporter, causing ferroportin to be phosphorylated,internalized, ubiquitylated, sorted [24] through the multi-vesicular body pathway and degraded in lysosomes [24, 70].The aim of this review is beyond a complete picture ofcurrent knowledge on the hepcidin regulation, therefore we


1 Kb

1 2 3 4 5 6 7 8

Ferroportin 1 gene

AAA

200 bp

P3 P2 P1IRE 1 2

VariantI (IRE)

II A

II B

III A

III B

III C

Exon 1 (IIIA/B)

Exon 1 (IIIA/B)

Exons 1 (IIIC) 2 (IIIC)

Exon 1 (IIA)

Exon 1 (IIB)

Exon 1

Exon 1’

Exon 1’

Exon 1’

Exon 1”

Exon 1’

Exon 2Predictedprotein

p571

p571

p571

p615

p596

p645

Figure 2: FPN1 gene structure and transcripts. Top: genomic organization and exon distribution of FPN1 (SLC40A1) locus, with (below)reported mRNA (GeneBank accession XM 047592). Middle: enlarged genomic region with exons 1-2 of FPN1 mRNA and upstream regions.Non-coding sequences are reported as open (white or coloured) boxes, coding sequences are indicated as black boxes and IRE element isindicated as a dashed box. Bent arrows below line indicate transcription start sites. P1, P2 and P3 indicate alternative promoter regions.Bottom: structure of clones obtained from 5′-RACE analysis. Alternative 5′ regions are indicated in different colours; sequences shared byall transcripts are in yellow. On the right is reported size (aa) of putative proteins coded by the respective transcripts.

will focus only on those aspects that we believe influenceerythropoiesis directly or indirectly. Different stimuli canmodulate hepcidin and act as positive or negative regulators.At the moment we know four major regulatory pathways(erythroid, iron store, inflammatory and hypoxia-mediatedregulation) that act through different signaling pathways tocontrol the production of hepcidin. It is obvious that thiscomplex network of interactions must be subjected to veryclose control in order to ensure that the iron erythropoieticdemand is met and, in turn, adequate concentrations ofiron in the circulation are always present (for a completereview see [66]). Under normal conditions iron store andinflammatory regulation activate hepcidin transcription inthe hepatocytes through the bone morphogenetic proteins(BMPs)/SMAD4 and signal transducer and activator oftranscription-3 (STAT-3) pathways, respectively [66, 71]. Thehemochromatosis protein HFE, transferrin receptor 2 (TfR2)and the membrane isoform of hemojuvelin (mHJV) areall positive modulators of hepcidin transcription and whendefective, lead to hemochromatosis (HH) in humans [66,72]. Oppositely, hypoxia, anemia, increased erythropoiesisand reduced iron stores all negatively regulate hepcidinexpression [66]. Emerging evidence suggests that ery-thropoiesis modulates hepcidin expression, with increasederythropoietic activity suppressing the action of hepcidin[73–77]. This in turn facilitates export of iron from thereticuloendothelial system and enterocytes, increasing theavailability of iron for erythropoiesis [73, 76]. Anemiaand hypoxia also suppress hepcidin expression, althoughrecent experiments indicate that functional erythropoiesisis required [73, 76, 77] for these conditions to regulatehepcidin expression. We have thus reached the issue thatmost concerns us: erythropoiesis and iron metabolism are

extremely intertwined in that alteration of one of the twomay have a major impact on the second (for completeknowledge on the topic see [66, 78–83]). That is the reasonwhy thalassemia intermedia and thalassemia major are thebest studied human models of hepcidin modulation by inef-fective erythropoiesis. Progressive iron overload is the mostsalient and ultimately fatal complication of beta-thalassemia.Iron deposition occurs in visceral organs (mainly in theheart, liver and endocrine glands), causing tissue damageand ultimately organ dysfunction and failure. Both transfu-sional iron overload and excess gastrointestinal absorptionare contributory. Paradoxically, excess gastrointestinal ironabsorption persists despite massive increases in total bodyiron load [68, 81, 83]. However, little is known about therelationship among ineffective erythropoiesis, the role ofiron-regulatory genes, and tissue iron distribution in beta-thalassemia. If iron were a dominant regulator, patientswith beta-thalassemia should express very high levels ofhepcidin in serum; in contrast, the levels are very low,suggesting that the ineffective erythropoiesis alone is able tosuppress the synthesis of hepcidin in spite of the presenceof a severe iron overload [66, 81]. Furthermore, serumfrom patients with thalassemia inhibited hepcidin mRNAexpression in the HepG2 cell line, which suggested thepresence of a humoral factor that downregulates hepcidin[84]. The nature of the erythropoietic regulator of hepcidinis still uncharacterized, but may include one or moreproteins during active erythropoiesis. Recent observationsin thalassemia patients has suggested that one of theseregulators could be the cytokine growth differentiationfactor-15 (GDF15) [66, 85]. GDF15 is a divergent memberof the transforming growth factor-beta superfamily that issecreted by erythroid precursors and other tissues. It has been


identified as an oxygen-regulated transcript responding tohypoxia and as a molecule involved in hepcidin regulation[85–91]. Serum from thalassemia patients suppressed hep-cidin mRNA expression in primary human hepatocytes anddepletion of GDF15 reversed the hepcidin suppression [66,85]. It was suggested that GDF15 overexpression arising froman expanded erythroid compartment contributed to ironoverload in thalassemia syndromes by inhibiting hepcidinexpression, possibly by antagonizing the BMP pathway [66,85]. Without going into a detailed analysis of the GDF15regulation mechanisms, we would like to recall the resultsobtained recently, that are in our view important to startreflecting on the existence of alternative ways that regulatehepcidin production. Very high levels of serum GDF15 werealso observed in patients with congenital dyserythropoieticanemia type 1 (CDA I) suggesting that GDF15 contributesto the inappropriate suppression of hepcidin with subse-quent secondary hemochromatosis in these patients [66,92]. Recently a very interesting study demonstrated thatexpression of both GDF15 mRNA and protein was stronglyand specifically responsive to intracellular iron depletion ina number of human cell lines and in vivo in humans [93].This upregulation is independent of IRP1, IRP2 and theHIF pathway suggesting the involvement of a novel iron-regulatory pathway [93]. This study showed that GDF15was induced by overexpression of wild-type ferroportin [93].This observation is very intriguing because it connects theiron-mediated regulation of GDF15 concentration to patho-physiological levels of iron: despite systemic iron overload,ineffective erythropoiesis and associated iron-fluxes in beta-thalassemia might generate an iron deficiency signal in arelevant molecular or cellular context and consequent stim-ulation of GDF15 expression in a particular erythroid com-partment [93]. Recent literatures provided at least two moremolecules potentially involved in the regulation of hepcidinby erythropoiesis, that is, the human twisted gastrulationfactor (TWSG1) [94] and the Oncostatin M (OsM) [95, 96].In contrast to GDF15, the highest-level expression of TWSG1was detected at early stages of erythroblast differentiationbefore hemoglobinization of the cells [94]. In human cells,TWSG1 suppressed hepcidin through a BMP-dependentmechanism [94]. In vivo studies on thalassemic mice showedthat TWSG1 expression was significantly increased in thespleen, bone marrow and liver [94]. So it was proposedthat TWSG1 might act with GDF15 to dysregulate ironhomeostasis in beta-thalassemia [94]. In contrast to GDF15and TWSG1, recent observations have showed that OsMcould induce hepcidin expression in human hepatoma celllines mainly through the JAK/STAT pathways [95]. Finally,results obtained by HuH7 hepatoma cells cocultered withprimary human erythroblasts or erythroleukemic UT7 cellspresented a 20- to 35-fold increase of hepcidin expression[96] and identified OsM responsible for increased levelsof hepcidin [96]. Furthermore, this study described thebiological involvement of OsM in iron metabolism “in vivo”through direct transcriptional regulation of hepcidin geneexpression and suggested a new OsM-hepcidin axis thatmight be critical in the development of hypoferremia ininflammation [96].

6. Role and Regulation of Ferroportin inErythroid Cells

The focus of this review is an update about the role ofFPN1 during human normal and pathological erythroiddifferentiation. There is still much work to do but we thinkthat the likely existence of alternative transcripts alteredexpression in all situations of ineffective erythropoiesis willgive answers to unresolved issues. In erythroid cells FPN1could be part of the signaling pathway through which theerythron communicates iron needs to expand the erythroidcompartment regardless of systemic iron level. Evidence ofa nonIRE FPN1 transcript in enterocytes of the duodenumsupports our belief in a setting far more complex andspecific to cell type. We also previously demonstrated theexistence of a FPN1 variant IIIA alternative transcript withthe potential to code for a longer protein with 44 additionalamino acids (p615), and a FPN1 variant IIIC alternativetranscript that potentially coded for a long protein with 74additional amino acids (p645), both N-terminal to and inframe with the canonical open reading frame [57] (Figure 2).Only FPN1 variant 1-IRE transcript has an IRE sequence inthe 5′UTR, whereas all the other transcript types do not.Unfortunately all these transcripts differ by 100-200 bp inlength and cannot be easily detected as distinct bands inNorthern analysis [57]. Interestingly, also these transcriptswere expressed mainly during the middle steps 4–11 days ofin vitro erythroid differentiation, corresponding to the matu-ration from late erythroid progenitors to polychromatophilicerythroblasts [57] (Figure 3). Therefore, the nonIRE (variantII and III) FPN1 transcripts were expressed when erythroidprogenitor/precursor cells need to accumulate iron in thecell [57]. At the moment we do not know yet whether thehypothetical isoforms p615 and p645 are actually present invivo because of the difficulty of obtaining specific antibodies.As mentioned earlier, to explain the surprising findingthat FPN1 protein expression was not responsive to ironconditions although about 50% of FPN1 is encoded by theIRE transcript, we speculated that in erythroid cells theregulation of FPN1 mRNA expression through the 5′-UTRIRE mechanism might be silenced because in this cell typeis needed to accumulate large amounts of iron for optimalheme synthesis [57] and a solution to this problem couldbe to use an upstream alternative promoters to producemRNA species in which the 5′-UTR IRE could be splicedout or made nonfunctional. Several studies support ourhypothesis: a recent work has shown an high frequency ofalternative first exons in erythroid genes suggesting a criticalrole in regulating gene function [97]. In the opinion ofthe authors the frequent presence of consensus translationinitiation sites among the alternative first exons suggeststhat many proteins have alternative N-terminal structureswhose expression can be coupled to promoter choice [97].So it seems that first exons and alternative promoters aremore widespread in the human genome than previouslyappreciated and that they may play a chief role in regulatingexpression of selected protein isoforms in a tissue-specificmanner [97]. Recently it was also demonstrated that manynon productive transcriptional initiation events occurred


in the vicinity of established promoters, some of whichmay produce mRNAs with altered translation efficiency,allowing transcripts to evolve to meet specific physiologicalneeds [98]. In conclusion we would like to emphasize thatthe presence of alternative ferroportin transcripts withoutan IRE in erythroid cells leaves open the possibility thatalterations in ferroportin mRNA splicing may be relevant inpathological conditions of altered erythroid differentiation[57, 59, 99]. So it would be interesting to investigate thepossibility of regulatory mutations in various iron disorders,in particular when type 4 hemochromatosis is present in theabsence of coding region mutations or in all cases of familialhyperferritinemia and in sporadic cases in the absence ofknown secondary causes (i.e., inflammation, malignancyinfection or dysmetabolism) where “ferroportin disease”should be suspected. We previously demonstrated that FPN1protein appeared to be localized at the level of the cytoplasmboth in vesicles and in the cytosol in erythroid cells, suggest-ing that FPN1 may be involved in the intracellular traffickingof iron between the cytosol and organelles [57]. In contrastto the clear detection of FPN1 at the basolateral membrane ofenterocytes, immunofluorescence studies with macrophagesrevealed a pronounced vesicular and mostly intracellularlocalization [100]. In particular confocal analysis revealedthe presence of some FPN1 microdomains at the plasmamembrane, likely suggesting a vesicular trafficking of theprotein between the cytosol and cell surface [100]. FPN1might be stored within the cell until it is needed for ironexport, at which point it might be recruited to the mem-brane [100]. Alternatively, FPN1 might mediate iron exportthrough the use of an intracellular vesicular compartment,in which FPN1 would act as an iron “concentrator” [100].Such a vesicular compartment could then be recruited tothe plasma membrane via exocytosis [100]. Furthermore,although proteins required for heme biosynthesis and Fe-Scluster assembly have been identified, we know little aboutintracellular iron trafficking, particularly to mitochondria.We do not exclude the possibility that FPN1 protein maybe involved in this pathway of iron metabolism. Someauthors have demonstrated that heme derived from humanor murine red blood cells or from an exogenous source ofheme led to marked transcriptional activation of the FPN1and HO1 genes [101]. Furthermore, the iron released fromheme catabolism subsequently stimulated the expressionof ferroportin mRNA and protein, indicating the existenceof a dual mechanism of ferroportin regulation in thiscell model, characterized by an early induction of genetranscription mediated primarily by heme, followed by apost-transcriptional regulation iron mediated [101]. So it istherefore tempting to speculate that similar regulatory mech-anisms could be involved in the transcriptional regulation ofFPN1 by heme in erythroid cells. Besides its function as pros-thetic group in heme proteins, heme itself can influence geneexpression at the level of transcription, protein synthesis,microRNA processing or post-translational modifications.Heme is a potent inducer of heme oxygenase 1 (HO1), a cyto-protective and anti-inflammatory molecule which catalyzesheme degradation. Heme inactivates the transcriptionalrepressor Bach1, thereby allowing the binding of Nrf2 to

Maf recognition elements (MAREs) present in the regulatoryregions HO1 [102, 103]. MAREs are also present in theenhancer of the H ferritin gene [102, 104] or in the ß globinLocus Control Region [102, 105]. Although HO1 expressionhas not been extensively studied in erythroid cells, it hasbeen shown that HO1 mRNA decreases following erythroiddifferentiation of Friend erythroleukemia cells, while mRNAscoding for the enzymes of the heme biosynthetic pathwayincrease [102, 106]. Furthermore, it was previously reportedthat heme mediated derepression of Maf recognition elementthrough direct binding to transcription repressor Bach1[107]; Nrf2 transcription factor regulated induction of theheme oxygenase-1 gene [108] and coordinately regulated agroup of oxidative stress-inducible genes in macrophages[109]; Bach1 was a sensor of cellular heme levels [110].A very recent study has showed that heme controlled thetranscription of the iron exporter FPN1 involving Bach1activity, Nrf2 nuclear accumulation and a highly conservedMARE/ARE enhancer element located at position −7007/ −7016 of the murine FPN1 promoter in macrophages [111].This suggest that iron recycling from heme involves asingle transcription control mechanism that regulates hemecatabolism, iron storage and detoxification as well as ironexport in a coordinated manner [111].

7. New Potential Therapeutic Approaches

It is increasingly evident that the iron metabolism, hemeand cellular erythropoiesis are inextricably linked, becauseiron metabolism [71, 112] and cellular heme (for exhaustivereview see [113]) are two of the most relevant key regulatorsof erythropoiesis. The complex regulation of erythropoiesissuggests the existence of several molecular targets that couldbe exploited therapeutically for treatment of RBC disorderslike thalassemias and anemias [33]. We must differentiatebetween primary iron overload, and iron overload thataccompanies ineffective erythropoiesis: in the latter case theadministration of hepcidin might be considered as a newpotential therapeutic approach to reduce iron overload inthalassemias and other forms of anemia associated withineffective erythropoiesis [33]. The reduced number orthe absence of mature erythroid cells in beta-thalassemiapatients is still very difficult to understand, and it has becomeone of the paradoxes among the most difficult to resolve(as noted Stefano Rivella in one of his very comprehensivereview [83]): when the body has greater need for red bloodcells instead it responds by decreasing their production. Themost probable hypothesis to explain this phenomenon mightrely on the existence of intrinsic and extrinsic mechanismsthat would affect the process of differentiation: for examplein cells where the synthesis of beta-globin gene is defectiveto the point that they ensure a stoichiometric between alphaand beta globin chains, a security mechanism can block theintrinsic maturation or, alternatively, an amount of heme inexcess can be an extrinsic signal to prevent the differentiationthat would lead to clusters of alpha globin chains productionof reactive oxygen species (ROS ) too toxic to survive [83].There is much experimental evidence that oxidative stressmay limit the process of differentiation. All this of course


Erythroblast progenitors

HSC

CFU-E

Pro-E Baso-E Poly-E Ortho-E

Colony formation(in vitro)

BFU-E

Progenitors Erythrocytes

Enucleation

Reticulocyte

FPN1-IRE mRNA

FPN1-nonIRE mRNA

FPN1-IRE mRNAErythropoiesis

Globin gene expression

Figure 3: Pathway of the erythropoiesis from progenitors to mature cells. Different stages are indicated: hematopoietic stem cell (HSC),burst-forming unit erythroid (BFU-E), colony-forming unit erythroid (CFU-E), proerythroblast (ProE), basophyilic (BasoE), polychromatic(PolyE) and orthochromatic erythroblast (OrthoE). Coloured bars indicate timing of FPN1 alternative transcript expression (bottom) andhemoglobin synthesis referred to stages of erythropoiesis (top).

worsens the anemic outline [83]. So the contribute of thesemechanisms to ineffective erythropoiesis might be differentin each patient according to level of beta-globin synthesisand other extrinsic factors such as iron overload [83]. At thispoint the question arises: is there a meeting point betweendifferent signaling pathways, although activated by differentsignals? Recent discoveries indicate that there is a potentialfor therapeutic intervention in beta-thalassemia by meansof manipulating iron metabolism [80, 83, 114]. A recentstudy suggested a link between EpoR/Jak/Stat signaling andiron metabolism, showing that in mice that completely lackStat5 activity the cell surface levels of TfR1 on erythroidcells were decreased more than 2-fold [115]. Another studysuggested a direct involvement of Epo in hepcidin regulationthrough the transcriptional factor C/EBP alpha [116]. Inaddition it has been shown a link between Jak 2 and FPN1:Jak2 phosphorylates FPN1 following binding of this proteinto hepcidin [117]. Phosphorylation of FPN1 then triggersits internalization and degradation [117]. Therefore Jak2might represent one of the major links at the interfacebetween erythropoiesis and iron metabolism suggestingthat use of Jak2 inhibitors, antioxidant, and analog of thehepcidin might be used to reduce ineffective erythropoiesisand abnormal iron absorption [83]. Administration ofsynthetic hepcidin or of agents that increase its expression,may be beneficial in controlling absorption of this metal[66]. Also GDF15 could be another potential therapeutictarget for beta-thalassemia syndromes [85]. A major goalof hemoglobinopathy research is to develop treatments thatcorrect the underlying molecular defects responsible forsickle cell disease and beta-thalassemia [114]. One approachto achieving this goal is the pharmacologic induction offetal hemoglobin (HbF) [114]. Although many of the eventscontrolling the activity of the beta-globin locus are known,

the details of those regulating normal human hemoglobinswitching and reactivation of HbF in adult hematopoieticcells remain to be elucidated. If the molecular events inhemoglobin switching or gamma-globin gene reactivationwere better understood and HbF could be fully reactivatedin adult cells, the insights obtained might lead to a cure forthese disorders. Agents that increase human HbF in patientsmay work at one or more levels: for example, hydroxyureaand 5-azacytidine kill dividing cells preferentially and mayincrease gamma-globin expression indirectly through thiseffect (for complete reviews see [114, 118]). Butyrate maywork both by HDAC inhibition and by increasing gamma-globin translation on ribosomes [114, 118]. The Stem CellFactor (SCF) induced an “in vitro” expansion of effectiveerythropoiesis and a reactivation of gamma-globin synthesisup to fetal levels, paving the way to its potential use in thetherapeutic treatment of this disease [119]. Recently it wasreported the ability of thalidomide to increase gamma-globingene expression and the proportion of HbF-containingcells in a human in vitro erythroid differentiation system[120] showing that thalidomide induced production ofROS that in turn caused p38 MAPK phosphorylation andglobally increased histone H4 acetylation [114, 120]. Allthese experiments present a body of evidence that suggests animportant role for intracellular signaling in HbF induction[114]. Finally recent publications have demonstrated theimportance of what has been termed the “integrated stressresponse” pathway in erythroid cells that is also activatedfrom a variety of stress stimuli, including viral infection, NO,heat shock, ROS, endoplasmic reticulum stress, ultravioletirradiation, proteosome inhibition, inadequate nutrientsand, in erythroid cells, limiting amounts of heme [114,121, 122]. In conclusion we are increasingly convinced ofthe importance to study the molecular mechanisms of iron


homeostasis dysregulation in thalassemia and in particularthe GDF15-BMP-Hepcidin-Ferroportin regulatory way inorder to understand its contribution to iron overload.

References

[1] S. Abboud and D. J. Haile, “A novel mammalian iron-regulated protein involved in intracellular iron metabolism,”Journal of Biological Chemistry, vol. 275, no. 26, pp. 19906–19912, 2000.

[2] A. Donovan, A. Brownlie, Y. Zhou et al., “Positional cloningof zebrafish ferroportin1 identifies a conserved vertebrateiron exporter,” Nature, vol. 403, no. 6771, pp. 776–781, 2000.

[3] A. T. McKie, P. Marciani, A. Rolfs et al., “A novel duodenaliron-regulated transporter, IREG1, implicated in the basolat-eral transfer of iron to the circulation,” Molecular Cell, vol. 5,no. 2, pp. 299–309, 2000.

[4] A. T. McKie and D. J. Barlow, “The SLC40 basolat-eral iron transporter family (IREG1/ferroportin/MTP1),”Pflugers Archiv European Journal of Physiology, vol. 447, no.5, pp. 801–806, 2004.

[5] A. Pietrangelo, “Non-HFE hemochromatosis,” Hepatology,vol. 39, no. 1, pp. 21–29, 2004.

[6] H. Mok, M. Mendoza, J. T. Prchal, P. Balogh, and A.Schumacher, “Dysregulation of ferroportin 1 interferes withspleen organogenesis in polycythaemia mice,” Development,vol. 131, no. 19, pp. 4871–4881, 2004.

[7] H. Zoller, R. O. Koch, I. Theurl et al., “Expression ofthe duodenal iron transporters divalent-metal transporter1 and ferroportin 1 in iron deficiency and iron overload,”Gastroenterology, vol. 120, no. 6, pp. 1412–1419, 2001.

[8] F. Yang, X. Wang, D. J. Haile, C. A. Piantadosi, and A. J. Ghio,“Iron increases expression of iron-export protein MTP1 inlung cells,” American Journal of Physiology, vol. 283, no. 5, pp.L932–L939, 2002.

[9] L. A. Martini, L. Tchack, and R. J. Wood, “Iron treatmentdownregulates DMT1 and IREG1 mRNA expression in Caco-2 cells,” Journal of Nutrition, vol. 132, no. 4, pp. 693–696,2002.

[10] L. M. Schimanski, H. Drakesmith, A. T. Merryweather-Clarke et al., “In vitro functional analysis of humanferroportin (FPN) and hemochromatosis-associated FPNmutations,” Blood, vol. 105, no. 10, pp. 4096–4102, 2005.

[11] H. Drakesmith, L. M. Schimanski, E. Ormerod et al.,“Resistance to hepcidin is conferred by hemochromatosis-associated mutations of ferroportin,” Blood, vol. 106, no. 3,pp. 1092–1097, 2005.

[12] A. S. Goncalves, F. Muzeau, R. Blaybel et al., “Wild-type andmutant ferroportins do not form oligomers in transfectedcells,” Biochemical Journal, vol. 396, no. 2, pp. 265–275, 2006.

[13] X.-B. Liu, F. Yang, and D. J. Haile, “Functional consequencesof ferroportin 1 mutations,” Blood Cells, Molecules, andDiseases, vol. 35, no. 1, pp. 33–46, 2005.

[14] I. De Domenico, D. M. Ward, G. Musci, and J. Kaplan,“Evidence for the multimeric structure of ferroportin,” Blood,vol. 109, no. 5, pp. 2205–2209, 2007.

[15] L. M. Schimanski, H. Drakesmith, C. Talbott et al., “Fer-roportin: lack of evidence for multimers,” Blood Cells,Molecules, and Diseases, vol. 40, no. 3, pp. 360–369, 2008.

[16] V. Devalia, K. Carter, A. P. Walker et al., “Autosomaldominant reticuloendothelial iron overload associated witha 3-base pair deletion in the ferroportin 1 gene (SLC11A3),”Blood, vol. 100, no. 2, pp. 695–697, 2002.

[17] E. Pignatti, L. Mascheroni, M. Sabelli, S. Barelli, S. Biffo, andA. Pietrangelo, “Ferroportin is a monomer in vivo in mice,”Blood Cells, Molecules, and Diseases, vol. 36, no. 1, pp. 26–32,2006.

[18] I. De Domenico, D. M. Ward, E. Nemeth et al., “Themolecular basis of ferroportin-linked hemochromatosis,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 102, no. 25, pp. 8955–8960, 2005.

[19] A. E. Rice, M. J. Mendez, C. A. Hokanson, D. C. Rees,and P. J. Bjorkman, “Investigation of the biophysical andcell biological properties of ferroportin, a multipass integralmembrane protein iron exporter,” Journal of MolecularBiology, vol. 386, no. 3, pp. 717–732, 2009.

[20] G. Cairo and A. Pietrangelo, “Iron regulatory proteins inpathobiology,” Biochemical Journal, vol. 352, no. 2, pp. 241–250, 2000.

[21] A. Lymboussaki, E. Pignatti, G. Montosi, C. Garuti, D. J.Haile, and A. Pietrangelo, “The role of the iron responsiveelement in the control of ferroportin1/IREG1/MTP1 geneexpression,” Journal of Hepatology, vol. 39, no. 5, pp. 710–715,2003.

[22] X. B. Liu, P. Hill, and D. J. Haile, “Role of the ferroportiniron-responsive element in iron and nitric oxide dependentgene regulation,” Blood Cells, Molecules & Diseases, vol. 29,no. 3, pp. 315–326, 2002.

[23] P. Brissot, M.-B. Troadec, E. Bardou-Jacquet et al., “Currentapproach to hemochromatosis,” Blood Reviews, vol. 22, no. 4,pp. 195–210, 2008.

[24] E. Nemeth, M. S. Tuttle, J. Powelson et al., “Hepcidinregulates cellular iron efflux by binding to ferroportin andinducing its internalization,” Science, vol. 306, no. 5704, pp.2090–2093, 2004.

[25] G. Montosi, A. Donovan, A. Totaro et al., “Autosomal-dominant hemochromatosis is associated with a mutationin the ferroportin (SLC11A3) gene,” Journal of ClinicalInvestigation, vol. 108, no. 4, pp. 619–623, 2001.

[26] A. Pietrangelo, “Hereditary hemochromatosis—a new lookat an old disease,” New England Journal of Medicine, vol. 350,no. 23, pp. 2383–2430, 2004.

[27] A. Pietrangelo, “Physiology of iron transport and thehemochromatosis gene,” American Journal of Physiology, vol.282, no. 3, pp. G403–G414, 2002.

[28] U. Testa, L. Conti, N. M. Sposi et al., “IFN-β selectively down-regulates transferrin receptor expression in human periph-eral blood macrophages by a post-translational mechanism,”Journal of Immunology, vol. 155, no. 1, pp. 427–435, 1995.

[29] G. Karthikeyan, L. K. Lewis, and M. A. Resnick, “Themitochondrial protein frataxin prevents nuclear damage,”Human Molecular Genetics, vol. 11, no. 11, pp. 1351–1362,2002.

[30] P. Ponka, “Tissue-specific regulation of iron metabolism andheme synthesis: distinct control mechanisms in erythroidcells,” Blood, vol. 89, no. 1, pp. 1–25, 1997.

[31] U. Testa, E. Pelosi, and C. Peschle, “The transferrin receptor,”Critical Reviews in Oncogenesis, vol. 4, no. 3, pp. 241–276,1993.

[32] U. Testa, “Recent developments in the understanding of ironmetabolism,” Hematology Journal, vol. 3, no. 2, pp. 63–89,2002.

[33] A. S. Tsiftsoglou, I. S. Vizirianakis, and J. Strouboulis,“Erythropoiesis: model systems, molecular regulators, anddevelopmental programs,” IUBMB Life, vol. 61, no. 8, pp.800–830, 2009.


[34] I. L. Weissman, “Stem cells: units of development, units ofregeneration, and units in evolution,” Cell, vol. 100, no. 1,pp. 157–168, 2000.

[35] S. H. Orkin and L. I. Zon, “Hematopoiesis: an evolvingparadigm for stem cell biology,” Cell, vol. 132, no. 4, pp. 631–644, 2008.

[36] J. Palis, “Ontogeny of erythropoiesis,” Current Opinion inHematology, vol. 15, no. 3, pp. 155–161, 2008.

[37] M. J. Koury, S. T. Sawyer, and S. J. Brandt, “New insights intoerythropoiesis,” Current Opinion in Hematology, vol. 9, no. 2,pp. 93–100, 2002.

[38] E. Ingley, P. A. Tilbrook, and S. P. Klinken, “New insights intothe regulation of erythroid cells,” IUBMB Life, vol. 56, no. 4,pp. 177–184, 2004.

[39] G. Stamatoyannopoulos, “Control of globin gene expressionduring development and erythroid differentiation,” Experi-mental Hematology, vol. 33, no. 3, pp. 259–271, 2005.

[40] D. W. Golde, “The stem cell,” Scientific American, vol. 265,no. 6, pp. 86–93, 1991.

[41] D. Metcalf, “The molecular control of cell division, differen-tiation commitment and maturation in haemopoietic cells,”Nature, vol. 339, no. 6219, pp. 27–30, 1989.

[42] S. H. Orkin, “Development of the hematopoietic system,”Current Opinion in Genetics and Development, vol. 6, no. 5,pp. 597–602, 1996.

[43] C. Smith, “Hematopoietic stem cells and hematopoiesis,”Cancer Control, vol. 10, no. 1, pp. 9–16, 2003.

[44] S. G. Emerson, S. Thomas, J. L. Ferrara, and J. L. Greenstein,“Developmental regulation of erythropoiesis by hematopoi-etic growth factors: analysis on populations of BFU-E frombone marrow, peripheral blood, and fetal liver,” Blood, vol.74, no. 1, pp. 49–55, 1989.

[45] I. D. Bernstein, R. G. Andrews, and K. M. Zsebo, “Recom-binant human stem cell factor enhances the formation ofcolonies by CD34+ and CD34+lin− cells, and the generationof colony-forming cell progeny from CD34+lin− cells cul-tured with interleukin-3, granulocyte colony-stimulating fac-tor, or granulocyte-macrophage colony-stimulating factor,”Blood, vol. 77, no. 11, pp. 2316–2321, 1991.

[46] S. D. Lyman, L. James, L. Johnson et al., “Cloning of thehuman homologue of the murine flt3 ligand: a growth factorearly hematopoietic progenitor cells,” Blood, vol. 83, no. 10,pp. 2795–2801, 1994.

[47] M. Gabbianelli, E. Pelosi, E. Montesoro et al., “Multi-leveleffects of flt3 ligand on human hematopoiesis: expansion ofputative stem cells and proliferation of granulomonocyticprogenitors/monocytic precursors,” Blood, vol. 86, no. 5, pp.1661–1670, 1995.

[48] M. Gabbianelli, M. Sargiacomo, E. Pelosi, U. Testa, G. Isacchi,and C. Peschle, ““Pure” human hematopoietic progenitors:permissive action of basic fibroblast growth factor,” Science,vol. 249, no. 4976, pp. 1561–1564, 1990.

[49] A. C. Berardi, A. Wang, J. Abraham, and D. T. Scadden, “Basicfibroblast growth factor mediates its effects on committedmyeloid progenitors by direct action and has no effect onhematopoietic stem cells,” Blood, vol. 86, no. 6, pp. 2123–2129, 1995.

[50] A. G. Leary, K. Ikebuchi, Y. Hirai et al., “Synergismbetween interleukin-6 and interleukin-3 in supporting pro-liferation of human hematopoietic stem cells: comparisonwith interleukin-1α,” Blood, vol. 71, no. 6, pp. 1759–1763,1988.

[51] D. Metcalf, “Hematopoietic regulators: redundancy or sub-tlety?” Blood, vol. 82, no. 12, pp. 3515–3523, 1993.

[52] S. B. Krantz, “Erythropoietin,” Blood, vol. 77, no. 3, pp. 419–434, 1991.

[53] W. Fried, “Erythropoietin,” Annual Review of Nutrition, vol.15, pp. 353–377, 1995.

[54] G. D. Demetri and J. D. Griffin, “Granulocyte colony-stimulating factor and its receptor,” Blood, vol. 78, no. 11, pp.2791–2808, 1991.

[55] C. J. Sherr, “Colony-stimulating factor-1 receptor,” Blood, vol.75, no. 1, pp. 1–12, 1990.

[56] K. Kaushansky, S. Lok, R. D. Holly et al., “Promotion ofmegakaryocyte progenitor expansion and differentiation bythe c-Mpl ligand thrombopoietin,” Nature, vol. 369, no. 6481,pp. 568–571, 1994.

[57] L. Cianetti, P. Segnalini, A. Calzolari et al., “Expressionof alternative transcripts of ferroportin-1 during humanerythroid differentiation,” Haematologica, vol. 90, no. 12, pp.1595–1606, 2005.

[58] D. M. Frazer, S. J. Wilkins, E. M. Becker et al., “Hep-cidin expression inversely correlates with the expressionof duodenal iron transporters and iron absorption inrats,” Gastroenterology, vol. 123, no. 3, pp. 835–844,2002.

[59] H. Mok, J. Jelinek, S. Pai et al., “Disruption of ferroportin1 regulation causes dynamic alterations in iron homeostasisand erythropoiesis in polycythaemia mice,” Development,vol. 131, no. 8, pp. 1859–1868, 2004.

[60] D.-L. Zhang, R. M. Hughes, H. Ollivierre-Wilson, M. C.Ghosh, and T. A. Rouault, “A ferroportin transcript thatlacks an iron-responsive element enables duodenal anderythroid precursor cells to evade translational repression,”Cell Metabolism, vol. 9, no. 5, pp. 461–473, 2009.

[61] T. Chaston, B. Chung, M. Mascarenhas et al., “Evidence fordifferential effects of hepcidin in macrophages and intestinalepithelial cells,” Gut, vol. 57, no. 3, pp. 374–382, 2008.

[62] A. Krause, S. Neitz, H.-J. Magert et al., “LEAP-1, a novelhighly disulfide-bonded human peptide, exhibits antimicro-bial activity,” FEBS Letters, vol. 480, no. 2-3, pp. 147–150,2000.


[64] C. Pigeon, G. Ilyin, B. Courselaud et al., “A new mouseliver specific gene, encoding a protein homologous to humanantimicrobial peptide hepcidin, is overexpressed during ironoverload,” Journal of Biological Chemistry, vol. 276, no. 11, pp.7811–7819, 2001.

[65] T. Ganz and E. Nemeth, “Regulation of iron acquisition andiron distribution in mammals,” Biochimica et Biophysica Acta,vol. 1763, no. 7, pp. 690–699, 2006.

[66] A. Piperno, R. Mariani, P. Trombini, and D. Girelli, “Hep-cidin modulation in human diseases: from research to clinic,”World Journal of Gastroenterology, vol. 15, no. 5, pp. 538–551,2009.

[67] P. L. Lee and E. Beutler, “Regulation of hepcidin and iron-overload disease,” Annual Review of Pathology, vol. 4, pp.489–515, 2009.

[68] R. E. Fleming and W. S. Sly, “Hepcidin: a putative iron-regulatory hormone relevant to hereditary hemochromatosisand the anemia of chronic disease,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 98, no. 15, pp. 8160–8162, 2001.


[69] G. Nicolas, C. Chauvet, L. Viatte et al., “The gene encodingthe iron regulatory peptide hepcidin is regulated by anemia,hypoxia, and inflammation,” Journal of Clinical Investigation,vol. 110, no. 7, pp. 1037–1044, 2002.

[70] T. Ganz, “Cellular iron: ferroportin is the only way out,” CellMetabolism, vol. 1, no. 3, pp. 155–157, 2005.

[71] N. C. Andrews, “Forging a field: the golden age of ironbiology,” Blood, vol. 112, no. 2, pp. 219–230, 2008.

[72] I. De Domenico, D. McVey Ward, and J. Kaplan, “Regulationof iron acquisition and storage: consequences for iron-linkeddisorders,” Nature Reviews Molecular Cell Biology, vol. 9, no.1, pp. 72–81, 2008.

[73] L. L. Dunn, Y. S. Rahmanto, and D. R. Richardson, “Ironuptake and metabolism in the new millennium,” Trends inCell Biology, vol. 17, no. 2, pp. 93–100, 2007.

[74] G. Dallalio, E. Law, and R. T. Means Jr., “Hepcidin inhibits invitro erythroid colony formation at reduced erythropoietinconcentrations,” Blood, vol. 107, no. 7, pp. 2702–2704,2006.

[75] A. Kattamis, I. Papassotiriou, D. Palaiologou et al., “Theeffects of erythropoetic activity and iron burden on hepcidinexpression in patients with thalassemia major,” Haematolog-ica, vol. 91, no. 6, pp. 809–812, 2006.

[76] M. Pak, M. A. Lopez, V. Gabayan, T. Ganz, and S. Rivera,“Suppression of hepcidin during anemia requires erythro-poietic activity,” Blood, vol. 108, no. 12, pp. 3730–3735, 2006.


[78] D. J. Weatherall and J. B. Clegg, “The thalassemia syn-dromes,” in The Thalassemia Syndromes, D. J. Weatherall andJ. B. Clegg, Eds., pp. 287–356, Blackwell Science, Oxford, UK,4th edition, 2001.

[79] D. J. Weatherall, “Phenotype-genotype relationships inmonogenic disease: lessons from the thalassaemias,” NatureReviews Genetics, vol. 2, no. 4, pp. 245–255, 2001.

[80] D. Rund and E. Rachmilewitz, “β-thalassemia,” New EnglandJournal of Medicine, vol. 353, no. 11, pp. 1135–1146, 2005.

[81] S. Gardenghi, M. F. Marongiu, P. Ramos et al., “Ineffectiveerythropoiesis in β-thalassemia is characterized by increasediron absorption mediated by down-regulation of hepcidinand up-regulation of ferroportin,” Blood, vol. 109, no. 11, pp.5027–5035, 2007.

[82] F. El Rassi, M. D. Cappellini, A. Inati, and A. Taher, “Beta-thalassemia intermedia: an overview,” Pediatric Annals, vol.37, no. 5, pp. 322–328, 2008.

[83] S. Rivella, “Ineffective erythropoiesis and thalassemias,”Current Opinion in Hematology, vol. 16, no. 3, pp. 187–194,2009.

[84] O. Weizer-Stern, K. Adamsky, N. Amariglio et al., “Down-regulation of hepcidin and haemojuvelin expression in thehepatocyte cell-line HepG2 induced by thalassaemic sera,”British Journal of Haematology, vol. 135, no. 1, pp. 129–138,2006.

[85] T. Tanno, N. V. Bhanu, P. A. Oneal et al., “High levelsof GDF15 in thalassemia suppress expression of the ironregulatory protein hepcidin,” Nature Medicine, vol. 13, no. 9,pp. 1096–1101, 2007.

[86] M. Bottner, M. Laaff, B. Schechinger, G. Rappold, K.Unsicker, and C. Suter-Crazzolara, “Characterization of therat, mouse, and human genes of growth/differentiationfactor-15/macrophage inhibiting cytokine-1 (GDF-15/MIC-1),” Gene, vol. 237, no. 1, pp. 105–111, 1999.

[87] R.-H. Wang, C. Li, X. Xu et al., “A role of SMAD4 iniron metabolism through the positive regulation of hepcidinexpression,” Cell Metabolism, vol. 2, no. 6, pp. 399–409,2005.

[88] J. L. Babitt, F. W. Huang, D. M. Wrighting et al., “Bonemorphogenetic protein signaling by hemojuvelin regulateshepcidin expression,” Nature Genetics, vol. 38, no. 5, pp. 531–539, 2006.

[89] J. Truksa, H. Peng, P. Lee, and E. Beutler, “Bone mor-phogenetic proteins 2, 4, and 9 stimulate murine hepcidin1 expression independently of Hfe, transferrin receptor 2(Tfr2), and IL-6,” Proceedings of the National Academy ofSciences of the United States of America, vol. 103, no. 27, pp.10289–10293, 2006.

[90] J. Massague, “TGF-β signal transduction,” Annual Review ofBiochemistry, vol. 67, pp. 753–791, 1998.

[91] M. De Caestecker, “The transforming growth factor-β super-family of receptors,” Cytokine and Growth Factor Reviews, vol.15, no. 1, pp. 1–11, 2004.

[92] H. Tamary, H. Shalev, G. Perez-Avraham et al., “Elevatedgrowth differentiation factor 15 expression in patients withcongenital dyserythropoietic anemia type I,” Blood, vol. 112,no. 13, pp. 5241–5244, 2008.

[93] S. Lakhal, N. P. Talbot, A. Crosby et al., “Regulation of growthdifferentiation factor 15 expression by intracellular iron,”Blood, vol. 113, no. 7, pp. 1555–1563, 2009.

[94] T. Tanno, P. Porayette, O. Sripichai et al., “Identification ofTWSG1 as a second novel erythroid regulator of hepcidinexpression in murine and human cells,” Blood, vol. 114, no.1, pp. 181–186, 2009.

[95] J. Kanda, T. Uchiyama, N. Tomosugi, M. Higuchi, T.Uchiyama, and H. Kawabata, “Oncostatin M and leukemiainhibitory factor increase hepcidin expression in hepatomacell lines,” International Journal of Hematology, vol. 90, no. 5,pp. 545–552, 2009.

[96] B. Chung, F. Verdier, P. Matak, J.-C. Deschemin, P. Mayeux,and S. Vaulont, “Oncostatin M is a potent inducer ofhepcidin, the iron regulatory hormone,” FASEB Journal, vol.24, no. 6, pp. 2093–2103, 2010.

[97] J. S. Tan, N. Mohandas, and J. G. Conboy, “High frequencyof alternative first exons in erythroid genes suggests a criticalrole in regulating gene function,” Blood, vol. 107, no. 6, pp.2557–2561, 2006.

[98] S. Buratowski, “Transcription: gene expression—where tostart?” Science, vol. 322, no. 5909, pp. 1804–1805, 2008.

[99] G. Cairo and S. Recalcati, “Iron-regulatory proteins: molec-ular biology and pathophysiological implications,” ExpertReviews in Molecular Medicine, vol. 9, no. 33, pp. 1–13,2007.

[100] F. Canonne-Hergaux, A. Donovan, C. Delaby, H.-J. Wang,and P. Gros, “Comparative studies of duodenal andmacrophage ferroportin proteins,” American Journal of Phys-iology, vol. 290, no. 1, pp. G156–G163, 2006.

[101] C. Delaby, N. Pilard, H. Puy, and F. Canonne-Hergaux,“Sequential regulation of ferroportin expression after ery-throphagocytosis in murine macrophages: early mRNAinduction by haem, followed by iron-dependent proteinexpression,” Biochemical Journal, vol. 411, no. 1, pp. 123–131,2008.

[102] A. Lolascon, L. De Falco, and C. Beaumont, “Molecularbasis of inherited microcytic anemia due to defects in ironacquisitionor heme synthesis,” Haematologica, vol. 94, no. 3,pp. 395–408, 2009.


[103] J. F. Reichard, G. T. Motz, and A. Puga, “Heme oxygenase-1 induction by NRF2 requires inactivation of the transcrip-tional repressor BACH1,” Nucleic Acids Research, vol. 35, no.21, pp. 7074–7086, 2007.

[104] K. Iwasaki, E. L. MacKenzie, K. Hailemariam, K. Sakamoto,and Y. Tsuji, “Hemin-mediated regulation of an antioxidant-responsive element of the human ferritin H gene and role ofref-1 during erythroid differentiation of K562 cells,” Molecu-lar and Cellular Biology, vol. 26, no. 7, pp. 2845–2856, 2006.

[105] T. Tahara, J. Sun, K. Nakanishi et al., “Heme positively regu-lates the expression of β-globin at the locus control region viathe transcriptional factor Bach1 in erythroid cells,” Journal ofBiological Chemistry, vol. 279, no. 7, pp. 5480–5487, 2004.

[106] H. Fujita, M. Yamamoto, T. Yamagami et al., “Sequentialactivation of genes for heme pathway enzymes duringerythroid differentiation of mouse Friend virus-transformederythroleukemia cells,” Biochimica et Biophysica Acta, vol.1090, no. 3, pp. 311–316, 1991.

[107] K. Ogawa, J. Sun, S. Taketani et al., “Heme mediatesderepression of Maf recognition element through directbinding to transcription repressor Bach1,” EMBO Journal,vol. 20, no. 11, pp. 2835–2843, 2001.

[108] J. Alam, D. Stewart, C. Touchard, S. Boinapally, A. M. K.Choi, and J. L. Cook, “Nrf2, a Cap’n’Collar transcriptionfactor, regulates induction of the heme oxygenase-1 gene,”Journal of Biological Chemistry, vol. 274, no. 37, pp. 26071–26078, 1999.

[109] T. Ishii, K. Itoh, S. Takahashi et al., “Transcription factor Nrf2coordinately regulates a group of oxidative stress-induciblegenes in macrophages,” Journal of Biological Chemistry, vol.275, no. 21, pp. 16023–16029, 2000.

[110] K. Igarashi, H. Hoshino, A. Muto et al., “Multivalent DNAbinding complex generated by small Maf and Bach1 as apossible biochemical basis for β-globin locus control regioncomplex,” Journal of Biological Chemistry, vol. 273, no. 19,pp. 11783–11790, 1998.

[111] S. Marro, D. Chiabrando, E. Messana, et al., “Heme controlsferroportin1 (FPN1) transcription involving Bach1, Nrf2and a MARE/ARE sequence motif at position -7007 of theFPN1 promoter,” Haematologica. In press.

[112] E. Nemeth, “Iron regulation and erythropoiesis,” CurrentOpinion in Hematology, vol. 15, no. 3, pp. 169–175, 2008.

[113] A. S. Tsiftsoglou, A. I. Tsamadou, and L. C. Papadopoulou,“Heme as key regulator of major mammalian cellularfunctions: molecular, cellular, and pharmacological aspects,”Pharmacology and Therapeutics, vol. 111, no. 2, pp. 327–345,2006.

[114] R. Mabaera, R. J. West, S. J. Conine et al., “A cell stresssignaling model of fetal hemoglobin induction: what doesn’tkill red blood cells may make them stronger,” ExperimentalHematology, vol. 36, no. 9, pp. 1057–1072, 2008.

[115] M. A. Kerenyi, F. Grebien, H. Gehart et al., “Stat5 regulatescellular iron uptake of erythroid cells via IRP-2 and TfR-1,”Blood, vol. 112, no. 9, pp. 3878–3888, 2008.

[116] J. P. Pinto, S. Ribeiro, H. Pontes et al., “Erythropoietinmediates hepcidin expression in hepatocytes through EPORsignaling and regulation of C/EBPα,” Blood, vol. 111, no. 12,pp. 5727–5733, 2008.

[117] I. De Domenico, E. Lo, D. M. Ward, and J. Kaplan,“Hepcidin-induced internalization of ferroportin requiresbinding and cooperative interaction with Jak2,” Proceedingsof the National Academy of Sciences of the United States ofAmerica, vol. 106, no. 10, pp. 3800–3805, 2009.

[118] A. Bank, “Regulation of human fetal hemoglobin: newplayers, new complexities,” Blood, vol. 107, no. 2, pp.435–443, 2006.

[119] M. Gabbianelli, O. Morsilli, A. Massa et al., “Effectiveerythropoiesis and HbF reactivation induced by kit ligand inβ-thalassemia,” Blood, vol. 111, no. 1, pp. 421–429, 2008.

[120] W. Aerbajinai, J. Zhu, Z. Gao, K. Chin, and G. P. Rodgers,“Thalidomide induces γ-globin gene expression throughincreased reactive oxygen species-mediated p38 MAPKsignaling and histone H4 acetylation in adult erythropoiesis,”Blood, vol. 110, no. 8, pp. 2864–2871, 2007.

[121] J.-J. Chen, “Regulation of protein synthesis by the heme-regulated eIF2α kinase: relevance to anemias,” Blood, vol.109, no. 7, pp. 2693–2699, 2007.

[122] R. C. Wek, H.-Y. Jiang, and T. G. Anthony, “Coping withstress: EIF2 kinases and translational control,” BiochemicalSociety Transactions, vol. 34, no. 1, pp. 7–11, 2006.


Review Article

Iron Loading and Overloading due to Ineffective Erythropoiesis

Toshihiko Tanno and Jeffery L. Miller

Molecular Medicine Branch, NIDDK, National Institutes of Health, Bethesda, MD 20892, USA

Correspondence should be addressed to Jeffery L. Miller, [email protected]

Received 11 December 2009; Accepted 18 February 2010


Copyright © 2010 T. Tanno and J. L. Miller. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Erythropoiesis describes the hematopoietic process of cell proliferation and differentiation that results in the production of maturecirculating erythrocytes. Adult humans produce 200 billion erythrocytes daily, and approximately 1 billion iron molecules areincorporated into the hemoglobin contained within each erythrocyte. Thus, iron usage for the hemoglobin production is aprimary regulator of plasma iron supply and demand. In many anemias, additional sources of iron from diet and tissue storesare needed to meet the erythroid demand. Among a subset of anemias that arise from ineffective erythropoiesis, iron absorptionand accumulation in the tissues increases to levels that are in excess of erythropoiesis demand even in the absence of transfusion.The mechanisms responsible for iron overloading due to ineffective erythropoiesis are not fully understood. Based upon data thatis currently available, it is proposed in this review that loading and overloading of iron can be regulated by distinct or combinedmechanisms associated with erythropoiesis. The concept of erythroid regulation of iron is broadened to include both physiologicaland pathological hepcidin suppression in cases of ineffective erythropoiesis.

1. Introduction

In the absence of blood transfusions, hereditary mutationsplay the primary role in most syndromes of iron over-load [1]. The genetic bases for several inherited forms ofhemochromatosis were identified and experimentally con-firmed as genes involved in iron regulation. These discoveriessubsequently led to major advances in understanding ironbiology [2]. Distinct from iron-regulating genes and theirproducts, a group of erythroid disorders that demonstrateineffective erythropoiesis also manifest a hemochromatosisphenotype. While tissue iron overload is a shared feature ofthese erythroid disorders, there is no evidence for a sharedgenetic mutation. Instead, the ineffective erythropoiesis itselfseems to cause iron accumulation and eventual overload.Iron loading in the liver and other tissues proceeds wellbeyond the levels needed to support erythropoiesis. Basedupon mapping of the human genome and the discovery ofhepcidin, new mechanisms are being explored for physio-logical and pathological regulation of iron associated withineffective erythropoiesis.

2. Understanding Ineffective Erythropoiesis

Erythrocytes serve a major function of oxygen transportand delivery throughout the body. Erythrocyte productionis appropriately driven by inadequate delivery of oxygento the tissues. The reduced tissue oxygen levels are sensedby peritubular cells in the renal cortex and outer medulla[3]. In response, those kidney cells express and secreteerythropoietin (EPO) [4]. Plasma EPO is transported to thebone marrow in order to promote the production of newerythrocytes. The new erythrocytes are produced throughthe process whereby erythroblasts respond to erythropoietinvia proliferation and differentiation over the course of severaldays. During erythropoiesis, large amounts of iron areneeded to produce hemoglobin. Importantly, transferrin-bound iron is endocytosed into erythroblasts after bindingto the plasma membrane receptors [5–8]. High levelsof membrane transferrin receptors are maintained duringeach cell cycle and at each stage of nucleated erythroblastdifferentiation [9]. As the iron demands are met, iron uptakeis decreased due to reduced transferrin receptor expression


during the terminal differentiation of the cells. Ultimately,hemoglobin production ceases, as does the demand for iron.In a concerted fashion, the transferrin receptor is releasedfrom the plasma membranes of reticulocytes as one of thefinal steps of erythroid differentiation [10].

Ineffective erythropoiesis describes a group of erythroiddisorders that produce fewer numbers of erythrocytes thanwould be expected to arise from the less mature erythroblastspresent in the marrow. As a result, there exists an imbalancebetween the amount of iron that is endocytosed by themarrow erythroblasts and the amount of iron releasedinto the circulation in erythrocytes [11, 12]. The conceptof ineffective erythropoiesis grew from ferrokinetic studies[13]. Classic ferrokinetic studies distinguished the patternsof iron utilization during ineffective erythropoiesis fromaplastic anemia, hemorrhage, or peripheral hemolysis [12].In addition to ferrokinetics, more recent research wasfocused upon identifying molecular and cellular mechanismsthat cause the underlying erythroid defects [14].

The erythroid response to tissue hypoxia is fundamentalfor understanding the pathophysiology of ineffective ery-thropoiesis. Tissue hypoxia is a common feature of anemias.Tissue hypoxia increases the expression of erythropoietin,and the erythropoietin drives the production of new erythro-cytes. In cases of anemia associated with ineffective erythro-poiesis, imbalance between erythrocyte supply and demandpersists despite increased tissue hypoxia and increasederythropoietin. As a result, erythropoietin levels remain high,and the marrow of patients with ineffective erythropoiesistypically becomes hypercellular [15–17]. Over the courseof time, the combination of tissue hypoxia, increased ery-thropoietin, and ineffective erythropoiesis creates a viciouscycle that may ultimately lead to a massive expansion oferythroblasts. Eventually, secondary bony pathologies [18]and extramedullary erythropoiesis [19] can also develop. Perthe topic of this review, pathological iron overload in theabsence of hereditary hemochromatosis is another hallmarkof the disease.

Several types of erythroid defects cause significant tissueiron overload in association with ineffective erythropoiesis.The major entities are summarized in Table 1. The tha-lassemia syndromes (thalassemia major and intermedia) rep-resent the most common causes of ineffective erythropoiesis.In the thalassemia syndromes, imbalances in the productionof alpha- and beta-globin chains result in increased apoptosisduring erythroblast maturation [14]. Iron loading, even inthe absence of transfused blood, is a well-recognized com-plication of the disease [20]. In sideroblastic anemias, globinchain production is intact, but erythropoiesis is characterizedby accumulation of iron in mitochondria that “ring” theerythroblast nucleus during maturation [21]. Intramedullaryapoptosis is a feature of acquired sideroblastic anemia [22].Some genes that regulate mitochondrial iron metabolismalso cause the inherited form of sideroblastic anemia [23].Since the inherited mitochondrial defects are not neces-sarily limited to hematopoietic cells, tissue iron loading inthose patients may not be derived solely from ineffectiveerythropoiesis [24]. However, the specialization of iron, iron-sulfur, and heme trafficking in erythroblasts strongly suggests

Table 1: Ineffective erythropoiesis associated with iron overload inabsence of transfusion.

Erythroid disorder Defect Reference

Thalassemia syndromesGlobin chainimbalance.

[67, 86]

Sideroblastic Anemia(inherited or acquired)

Iron accumulation inmitochondria.

[23, 70, 87, 88]

Dyserythropoietic Anemia(Types I and II)

Nuclear andmembrane defects.

[26, 27, 69]

erythroid involvement in the iron loading pathology. Asrecently reviewed by Sheftel et al., ringed sideroblasts areassociated with several erythroblast iron and mitochondrialdefects [25]. It remains to be determined whether separatedefects that result in sideroblastic anemia have equivalenteffects upon iron homeostasis in nonerythroid tissues. Athird group of disorders associated with ineffective erythro-poiesis and iron overloading are called dyserythropoieticanemias [26]. The dyserythropoietic defects are distinct fromothers in mutations of globin and mitochondrial genes[26, 27]. Tissue iron overload is not uncommon in twoof three dyserythropoietic subtypes. Pyruvate kinase defi-ciency results in defective glycolysis resulting in erythroblastapoptosis and peripheral blood hemolysis [28, 29]. Pyruvatekinase deficiency is being investigated as a separate cause ofineffective erythropoiesis, but iron overloading is less con-sistent among these patients [30]. Other erythroid disordersassociated with some degree of ineffective erythropoiesisinclude chronic pernicious anemia, hereditary spherocytosis,and sickle cell anemia [31, 32]. However, the associationbetween iron loading and ineffective erythropoiesis in thesedisorders is inconsistent [33].

3. Iron Loading and Overloading inIneffective Erythropoiesis

Adult humans produce approximately 200 billion erythro-cytes daily [34]. Each erythrocyte contains approximately300 million molecules of hemoglobin [35]. Each hemoglobinmolecule contains four heme molecules, and each hemecontains a single iron moiety. Therefore, to satisfy theproduction of erythrocytes, 2 E20 iron molecules (20 mg)are utilized daily for erythropoiesis even in the absence ofdisease. The robust demand for iron is met by transferrin-bound iron in the plasma [36]. Three major sources ofiron are utilized to maintain adequate levels of transferrin-bound iron: dietary iron, body iron stores, and recycled ironfrom senescent erythrocytes. In healthy adults, the majorityof transferrin-bound iron in the plasma is generated frommacrophage catabolism of mature erythrocytes in circulatingblood. In steady state, the iron that is recycled from oldererythrocytes is largely sufficient for the production of newerythrocytes. Phagocytosis of the older erythrocytes bymacrophages results in their catabolism and the breakdownof hemoglobin. The hemoglobin-salvaged iron is loadedon transferrin for transport to marrow erythroblasts. The


Hb synthesis

20 mg Fe/day

Hb catabolism

Figure 1: The hemoglobin iron cycle

membrane expression of ferroportin, an iron channel pro-tein, provides the key regulatory element of transferrinloading of iron from macrophages. Ferroportin levels onthe cell membrane are regulated by another protein namedhepcidin. Hepcidin acts through binding and internalizationof ferroportin from the surface of iron-exporting cells[37–39]. Therefore, hepcidin is a principal regulator ofthe hemoglobin iron cycle. Hepcidin expression is highlyregulated by a growing number of proposed mechanisms. Inthe steady state, broad variations in plasma levels of hepcidinare predicted among adults [40]. In iron deficient states,hepcidin expression is consistently suppressed. In ineffectiveerythropoiesis, hepcidin expression is less consistently sup-pressed [41].

In a variety of conditions, the hemoglobin iron cycle(Figure 1) becomes unbalanced due to a decreased supply ofiron from mature erythrocytes or an increased erythroblastdemand. In most cases of anemia, the imbalance is magnifiedby tissue hypoxia with increased erythropoietin production.If the supply of iron from aging erythrocytes is inadequate,transferrin-bound iron must be obtained from tissue storesand the diet [42]. Increased transferrin iron loading frommultiple sources is achieved when hepcidin is suppressed[43]. Erythroblast demand for iron continues until tissuedemands for oxygen are satisfied, and iron stores are replen-ished. As shown in Table 2, several erythroid disorders areassociated with imbalances between tissue hypoxia, erythro-poietin, and erythropoiesis. With the exception of aplasticanemia, the erythroblast demand for iron is increased. It ispresumed that hepcidin expression is reduced to meet theerythroblast demand in many, if not all of these disorders,but confirmatory studies are awaited. Despite the erythroiddemand for iron, extra-erythroid loading of iron is nota typical feature of anemia or polycythemia. Remarkably,ineffective erythropoiesis is unique in causing accumulationof iron in extra-erythroid tissues to levels that are wellbeyond the erythroid requirements.

4. How Does Ineffective Erythropoiesis CauseTissue Iron Overload?

Iron absorption is normally regulated by a combinationof iron stores, inflammation, hypoxia, and erythropoieticiron demand [44, 45]. Presumably, the mechanisms that

satisfy the iron appetite of immature erythroblasts are alsoactive in ineffective erythropoiesis. Those molecular andcellular mechanisms are a focus of curiosity- and clinical-driven hematology research [46]. Along with a physiolog-ical mechanism that provides a basis for the erythroidregulator of iron, it is proposed here that pathologicalmechanisms may be identified that are unique to ineffectiveerythropoiesis. The notion of a pathological iron regulatoris suggested by the unique accumulation of iron to toxiclevels in patients with ineffective erythropoiesis. As discussedbelow, the cytokine named GDF15 could serve the role of apathological erythroid signal. Since hepcidin plays a centralrole in that network, the discussion here is largely focusedupon erythroid-related variables, including GDF15, whichmay contribute to suppressed expression of hepcidin.

4.1. Iron, Iron Transport, and Iron Turnover. For manyyears, plasma iron parameters or the depletion of iron fromthe plasma compartment were actively investigated as themechanism of signaling between erythropoiesis and ironregulation [47, 48]. Recently, it was reported that hepcidinexpression correlates with transferrin saturation levels [49].While decreased transferrin saturation could provide amechanism for hepcidin suppression in iron deficient states,it seems less likely that the transferrin saturation levelsdetected in ineffective erythropoiesis (sometimes 100%)cause suppression of hepcidin. In addition to high transferrinsaturation levels, there is an overall increase of holo-transferrin removal from the plasma for iron delivery to theexpanded population of immature erythroblasts. Increasedplasma iron turnover is increased in humans with ineffectiveerythropoiesis [50]. In addition, the newly incorporatediron is recycled from marrow erythroblasts rather thancirculating erythrocytes. While increased plasma deliveryand erythroblast recycling of iron from the erythropoieticcompartment are interesting components of ineffective ery-thropoiesis, the significance of these features in suppressionof hepcidin remains uncertain. Transferrin metabolism,saturation kinetics, plasma iron turnover, and heme recyclingare all complex processes [48]. The topic remains unsettled.

4.2. Hypoxia and Erythropoietin in Ineffective Erythropoiesis.Tissue hypoxia directly inhibits hepcidin expression inhepatocytes. Hypoxia effects are generally independent ofiron stores [46]. Since patients with severe ineffectiveerythropoiesis are usually anemic, tissue hypoxia may playa role in iron regulation in this disorder. Comparativedata between hepatocyte cell lines and nonanemic animalsexposed to hypoxic conditions consistently demonstrated adown-regulation of hepcidin production [44]. The hypoxiainducible factor/von Hippel-Lindau (HIF/vHL) pathwaymediates responses to hypoxia and other cellular stressors.In normoxic, iron-sufficient conditions, an oxygen and iron-dependent prolyl hydroxylase modifies the HIF regulatorysubunit named HIF1α. In hypoxia or iron deficiency, HIF1αaccumulates, translocates to the nucleus, and associateswith HIF1β, a constitutively expressed HIF subunit. TheHIF heteroduplex binds promoter elements to modulate


Table 2: A comparison of erythroid pathologies.

Erythroid condition Tissue iron overload Tissue hypoxia Increased erythropoietin Increased erythropoiesis

Ineffective erythropoiesis Yes Yes Yes Yes

Hemolysis No Yes Yes Yes

Blood loss No Yes Yes Yes

Iron deficiency anemia No Yes Yes No

Aplastic anemia No Yes Yes No

Secondary polycythemia No Yes Yes Yes

Primary polycythemia No No No Yes

gene transcription [51]. Peyssonnaux and colleagues recentlydemonstrated that mice with liver-specific, conditionalinactivation of HIF1α maintained on an iron-deficientdiet develop inappropriately high levels of hepcidin [52].Hypoxia-related changes were not specifically reported, soadditional studies of this important area of research areneeded. In vitro, inhibition of the prolyl hydroxylasespromotes HIF1α stabilization, and also negatively regulateshepcidin transcription [53]. Despite the developing bodyof evidence in support of a direct role for hypoxia inhepcidin suppression, there is less consensus as to whetherhepcidin regulation is affected through hypoxia responseelements (HRE), on promoter regions of HIF/vHL pathway-dependent genes [52, 53]. Overall, it is becoming increasinglylikely that HIF/vHL participates in hepcidin gene regulationnetworks.

A second potential mechanism for hepcidin regulationby hypoxia involves a protein named hemojuvelin (HJV).HJV is a member of the repulsive guidance molecule (RGM)family of proteins that function as coreceptors for BoneMorphogenetic Protein (BMP) signaling. The membraneform of HJV binds to type I BMP receptors and stimulatesthe BMPs (such as BMP2, 4 or 9) signaling. The signalingenhances the phosphorylation of SMAD signaling pathwayand stimulates hepcidin transcription [54]. HJV proteinis cleaved and secreted as a soluble form (sHJV) that isprocessed by furin-like protease (a proprotein convertase)[55–57]. sHJV acts as a repressor of BMP signaling bycompeting with the membrane form of HJV [58, 59]. Thus,any stimulus that leads to increased sHJV production mayalso reduce hepcidin expression. The generation of sHJVappears to be increased by iron deficiency and hypoxia inassociation with the stabilization of HIF1α [57, 58, 60].Both stimuli may lead to reduced hepcidin production andincreased iron absorption.

Tissue hypoxia may also regulate hepcidin and iron load-ing indirectly by increasing the expression of the hormone,EPO. Increased EPO is associated with erythropoietic activ-ity, which is inversely correlated with hepcidin expressionin patients with thalassemia [61]. Hepcidin was not sup-pressed if erythropoiesis was inhibited by EPO neutralizingantibodies, chemotherapy, or irradiation of bone marrow[62, 63]. However, high doses of EPO directly down-regulatehepcidin expression in vitro through a mechanism involvingthe transcription factor core element binding protein a

(C/EBPa) at a cognate DNA binding site present in thehepcidin promoter [64]. While the relative contribution ofdirect versus indirect EPO effects upon hepcidin regulationis a matter of ongoing debate, increased expression of EPO iscentral to the increase in erythroid demand for iron in botheffective and ineffective erythropoiesis.

5. Molecules Released from Erythroblasts

It is assumed in this review that erythroblasts or the processof erythropoiesis in the bone marrow of humans includessome mechanism(s) for communicating the demand foriron to distant sites in the body. For many years, scien-tists searched for such a mechanism without satisfaction.Along with other approaches, the avenue of genomics-based research was recently utilized for the identificationof molecules released from erythroblasts that may serveas “signals” for iron regulation. Two candidate molecules(GDF15 and TWSG1) were identified by this approach. Inaddition to these molecules, soluble transferrin receptor hasbeen explored as a candidate iron regulator.

5.1. Growth Differentiation Factor 15 (GDF15). Thalassemicserum contains factors that suppress hepcidin expressionin hepatocytes or hepatocyte cell lines [65]. Based uponthe initial discovery that SMAD4 signal transduction isinvolved in hepcidin gene regulation [66], focus was placedupon signal transduction involving the transforming growthfactor-β (TGF-β) superfamily. Using a transcriptional pro-filing approach during erythropoiesis, a member of thatsuperfamily named growth differentiation factor 15 (GDF15)was discovered to be up-regulated in thalassemic serumand can suppress hepcidin expression in vitro [67]. Inter-estingly, GDF15 (also called, MIC-1, PLAB, PDF, PTGF-β, NRG-1, and NAG-1) can be regulated by p53 [68].Since intramedullary apoptosis is frequently associated withineffective erythropoiesis, GDF15 was explored as a leadcandidate molecule as a pathological erythroid regulator ofiron. In cultured human hepatocytes and hepatic cell lines,both recombinant GDF15 as well as GDF15 in the serumof thalassemia patients inhibited the expression of hepcidin.Like thalassemia syndromes, congenital dyserythropoieticanemia type I also showed high levels of serum GDF15and inappropriate suppression of hepcidin associated withiron-overload [69]. Moreover, the high elevation of GDF15


was identified in refractory anemia with ring siderob-lasts [70]. Interestingly, in vitro studies demonstrated thaterythropoiesis-specific production of GDF15 was dependentupon EPO. The production of GDF15 was also stimulated bya compound that reduces mitochondrial membrane poten-tial in erythroblasts [70]. These data support a concept thatapoptotic erythroblasts produce GDF15, and that GDF15contributes to extra-erythroid tissue iron loading due toineffective erythropoiesis.

It must be stressed that there is minimal evidence to datewhich suggests that GDF15 plays an important role in ironregulation outside the setting of ineffective erythropoiesis.Lakhal et al. demonstrated up-regulation of GDF15 inresponse to iron depletion using intracellular iron chelatorin cell lines and a robust amount of intravenous desferriox-amine (DFO) in humans. GDF15 up-regulation occurredindependently of HIF signaling, suggesting the involvementof a novel iron sensing pathway [71]. The study also reportedincreased levels of GDF15 in patients with iron deficiency.However, others reported conflicting results among blooddonors (see [72]; unpublished data). In separate studies,Kanda et al. performed an in vivo physiological study ofthe relationship between GDF15, serum hepcidin, and ery-thropoiesis in the clinical setting of stem cell transplantation(SCT) [73]. The pre- and post-SCT serum hepcidin levelswere monitored along with other factors that may affect hep-cidin expression. After SCT, serum hepcidin levels showed asignificant inverse correlation with markers of erythropoieticactivity, such as the sTfR and the reticulocyte counts, but notGDF15 levels. Ashby et al. also performed an in vivo studyfor physiological erythropoiesis focusing on the relationshipbetween plasma hepcidin, GDF15, and sTfR. Neither EPOadministration nor venesection caused significant changes inGDF15 or sTfR levels (see below) despite a clear suppressionof hepcidin [74]. Hence, it would surprise the authorsof this review if GDF15 is determined to be the majorerythroid regulator of iron in healthy adult humans. Instead,these results suggest that GDF15 contributes to hepcidinsuppression and iron overloading in the pathological settingof ineffective erythropoiesis as originally proposed [67].

5.2. Twisted Gastrulation (TWSG1). Erythroblast expressionof a second molecule named twisted gastrulation wasexplored as a potential erythroid regulator of hepcidin.Expression of the TWSG1 gene was discovered as part oferythroblast transcriptome analyses. Further interest inthis candidate molecule grew from the hypothesis thatTWSG1 may regulate hepcidin in a similar fashion to solublehemojuvelin (sHJV). Hemojuvelin has a key role in hepcidinregulation. Membrane HJV acts as a coreceptor for BMPs,whereas soluble HJV (sHJV) down-regulates hepcidin ina competitive way interfering with BMP signaling [57].The TWSG1 gene product is a small, secreted cysteine-rich protein that antagonizes the interaction of bonemorphogenetic proteins (BMP) with its receptor [75, 76].Like sHJV, TWSG1 could suppress hepcidin expressionby interfering with BMP signaling as a BMP antagonist.In vitro, TWSG1 protein interferes with BMP-mediatedhepcidin expression in human hepatocytes. Phosphorylation

studies further suggest that TWSG1 acts by inhibitingthe BMP-dependent activation of SMAD-mediated signaltransduction. TWSG1 secreted from erythroblasts may thuscontribute to iron loading by inhibiting BMP-mediatedhepcidin expression [77]. Unfortunately, assays are not yetavailable to measure the level of TWSG1 in human blood.Since murine hepatocytes produce BMPs and the major sitesof erythropoiesis in murine thalassemia are the spleen andliver, it was proposed that this cytokine may be more active inhepcidin suppression in the murine model system [78, 79].

5.3. Soluble Transferrin Receptor (sTfR). Approximately 80%of soluble TfR1 (sTfR) is generated during the maturationof erythroid cells [80]. Soluble TfR is a truncated form ofthe TfR on the surface of the cells [81]. The sTfR levelindicates erythropoietic activity and iron status of the organs[82]. sTfR is increased in thalassemia [67], congenital dysery-thropoietic anemia [69, 81], and sideroblastic anemia [83].All of these observations make sTfR a strong candidate forerythroid regulation of hepcidin and iron. However, it wasconcluded on the basis of a transgene expression model inmice that sTfR levels neither regulate hepcidin nor iron [84].

6. Summary

The phrase “ineffective erythropoiesis” collectively describesa group of erythroid defects that are marked by decreasederythrocyte production despite increased early erythro-poiesis. Ineffective erythropoiesis manifests a unique featureof non-transfusional iron overload in extra-erythroid tissues.This feature of secondary hemochromatosis distinguishesineffective erythropoiesis from other causes of anemia.Further, the excess iron accumulates in parenchymal cellsand negatively affects clinical outcome [85]. As such, theregulation of iron loading and overloading in ineffectiveerythropoiesis remains a fertile area of basic and clinicalresearch. Physiological mechanisms that regulate iron inthe context of hemoglobin production and catabolism arelikely involved in both effective (normal) and ineffectiveerythropoiesis. In this review, we proposed that additionalmechanisms or signals related to the erythroid pathol-ogy may contribute to iron overloading in other tissues.Increased expression of GDF15 from apoptotic erythroblastsis being explored in this context. Based upon the rapidpace of discovery within the field of iron biology, additionalmechanisms and insights regarding the special relationshipsbetween erythropoiesis and iron regulation are predicted.

References

[1] M. C. H. Janssen and D. W. Swinkels, “Hereditary haemochro-matosis,” Best Practice and Research: Clinical Gastroenterology,vol. 23, no. 2, pp. 171–183, 2009.

[2] N. C. Andrews, “Forging a field: the golden age of ironbiology,” Blood, vol. 112, no. 2, pp. 219–230, 2008.

[3] C. Lacombe, J.-L. Da Silva, P. Bruneval, et al., “Erythropoietin:sites of synthesis and regulation of secretion,” AmericanJournal of Kidney Diseases, vol. 18, no. 4, supplement 1, pp.14–19, 1991.


[4] G. L. Semenza, “Involvement of oxygen-sensing pathways inphysiologic and pathologic erythropoiesis,” Blood, vol. 114,no. 10, pp. 2015–2019, 2009.

[5] J. L. Miller, J. M. Njoroge, A. N. Gubin, and G. P. Rodgers,“Prospective identification of erythroid elements in culturedperipheral blood,” Experimental Hematology, vol. 27, no. 4, pp.624–629, 1999.

[6] D. Hemmaplardh and E. H. Morgan, “The role of endocytosisin transferrin uptake by reticulocytes and bone marrow cells,”British Journal of Haematology, vol. 36, no. 1, pp. 85–96, 1977.

[7] R. D. Klausner, G. Ashwell, J. van Renswoude, J. B. Harford,and K. R. Bridges, “Binding of apotransferrin to K562cells: explanation of the transferrin cycle,” Proceedings of theNational Academy of Sciences of the United States of America,vol. 80, no. 8, pp. 2263–2266, 1983.

[8] B. D. Grant and J. G. Donaldson, “Pathways and mechanismsof endocytic recycling,” Nature Reviews Molecular Cell Biology,vol. 10, no. 9, pp. 597–608, 2009.

[9] M. Schranzhofer, M. Schifrer, J. A. Cabrera, et al., “Remod-eling the regulation of iron metabolism during erythroiddifferentiation to ensure efficient heme biosynthesis,” Blood,vol. 107, no. 10, pp. 4159–4167, 2006.

[10] J. Ahn and R. M. Johnstone, “Origin of a soluble truncatedtransferrin receptor,” Blood, vol. 81, no. 9, pp. 2442–2451,1993.

[11] A. Xenakis and I. Marcu, “Radioisotope investigations insome hemolytic disorders of children,” Revue Roumaine deMedecine, vol. 15, no. 1, pp. 79–83, 1977.

[12] C. Ricketts and I. Cavill, “Ferrokinetics: methods and interpre-tation,” Clinical Nuclear Medicine, vol. 3, no. 4, pp. 159–164,1978.

[13] R. L. Huff, P. J. Elmlinger, J. F. Garcia, J. M. Oda, M. C.Cockrell, and J. H. Lawrence, “Ferrokinetics in normal personsand in patients having various erythropoietic disorders,” TheJournal of Clinical Investigation, vol. 30, no. 12:2, pp. 1512–1526, 1951.

[14] S. L. Schrier, “Pathophysiology of thalassemia,” Current Opin-ion in Hematology, vol. 9, no. 2, pp. 123–126, 2002.

[15] F. Centis, L. Tabellini, G. Lucarelli, et al., “The importance oferythroid expansion in determining the extent of apoptosisin erythroid precursors in patients with β-thalassemia major,”Blood, vol. 96, no. 10, pp. 3624–3629, 2000.

[16] G. Sansone, G. Masera, A. Cantu-Rajnoldi, and S. Terzoli, “Anunclassified case of congenital dyserythropoietic anaemia witha severe neonatal onset,” Acta Haematologica, vol. 88, no. 1, pp.41–45, 1992.

[17] “Ineffective erythropoiesis,” The Lancet, vol. 1, no. 7813, pp.1164–1165, 1973.

[18] R. Di Matteo, F. Liuzza, P. F. Manicone, et al., “Bone andmaxillofacial abnormalities in thalassemia: a review of theliterature,” Journal of Biological Regulators and HomeostaticAgents, vol. 22, no. 4, pp. 211–216, 2008.

[19] S. Aizawa, U. Kohdera, M. Hiramoto, et al., “Ineffectiveerythropoiesis in the spleen of a patient with pyruvate kinasedeficiency,” American Journal of Hematology, vol. 74, no. 1, pp.68–72, 2003.

[20] M. J. Pippard, S. T. Callender, G. T. Warner, and D. J.Weatherall, “Iron absorption and loading in β-thalassaemiaintermedia,” The Lancet, vol. 2, no. 8147, pp. 819–821, 1979.

[21] W. D. Bowman Jr., “Abnormal (“ringed”) sideroblasts invarious hematologic and non-hematologic disorders,” Blood,vol. 18, pp. 662–671, 1961.

[22] S. H. Merchant, N. J. Gonchoroff, and R. E. Hutchison,“Apoptotic index by annexin V flow cytometry: adjunct to

morphologic and cytogenetic diagnosis of myelodysplasticsyndromes,” Cytometry Part B, vol. 46, no. 1, pp. 28–32, 2001.

[23] C. Camaschella, “Recent advances in the understanding ofinherited sideroblastic anaemia,” British Journal of Haematol-ogy, vol. 143, no. 1, pp. 27–38, 2008.

[24] M. Fontenay, S. Cathelin, M. Amiot, E. Gyan, and E. Solary,“Mitochondria in hematopoiesis and hematological diseases,”Oncogene, vol. 25, no. 34, pp. 4757–4767, 2006.

[25] A. D. Sheftel, D. R. Richardson, J. Prchal, and P. Ponka,“Mitochondrial iron metabolism and sideroblastic anemia,”Acta Haematologica, vol. 122, no. 2-3, pp. 120–133, 2009.

[26] R. Renella and W. G. Wood, “The congenital dyserythropoieticanemias,” Hematology/Oncology Clinics of North America, vol.23, no. 2, pp. 283–306, 2009.

[27] K. Schwarz, A. Iolascon, F. Verissimo, et al., “Mutationsaffecting the secretory COPII coat component SEC23B causecongenital dyserythropoietic anemia type II,” Nature Genetics,vol. 41, no. 8, pp. 936–940, 2009.

[28] A. Zanella, M. B. Colombo, R. Miniero, L. Perroni, T. Meloni,and G. Sirchia, “Erythrocyte pyruvate kinase deficiency: 11new cases,” British Journal of Haematology, vol. 69, no. 3, pp.399–404, 1988.

[29] A. Zanella, P. Bianchi, and E. Fermo, “Pyruvate kinasedeficiency,” Haematologica, vol. 92, no. 6, pp. 721–723, 2007.

[30] A. Finkenstedt, P. Bianchi, I. Theurl, et al., “Regulation of ironmetabolism through GDF15 and hepcidin in pyruvate kinasedeficiency,” British Journal of Haematology, vol. 144, no. 5, pp.789–793, 2009.

[31] T. C. Iancu, “Iron overload,” Molecular Aspects of Medicine, vol.6, no. 1, pp. 1–100, 1983.

[32] C. J. Wu, L. Krishnamurti, J. L. Kutok, et al., “Evidence forineffective erythropoiesis in severe sickle cell disease,” Blood,vol. 106, no. 10, pp. 3639–3645, 2005.

[33] D. Samson, D. Halliday, and I. Chanarin, “Reversal of ineffec-tive erythropoiesis in pernicious anaemia following vitaminB12 therapy,” British Journal of Haematology, vol. 35, no. 2, pp.217–224, 1977.

[34] M. J. Koury and P. Ponka, “New insights into erythropoiesis:the roles of folate, vitamin B 12, and iron,” Annual Review ofNutrition, vol. 24, pp. 105–131, 2004.

[35] D. W. E. Smith, “The molecular biology of mammalianhemoglobin synthesis,” Annals of Clinical and LaboratoryScience, vol. 10, no. 2, pp. 116–122, 1980.

[36] L. T. Goodnough, B. Skikne, and C. Brugnara, “Erythropoi-etin, iron, and erythropoiesis,” Blood, vol. 96, no. 3, pp. 823–833, 2000.



[39] I. De Domenico, E. Lo, D. M. Ward, and J. Kaplan, “Hepcidin-induced internalization of ferroportin requires binding andcooperative interaction with Jak2,” Proceedings of the NationalAcademy of Sciences of the United States of America, vol. 106,no. 10, pp. 3800–3805, 2009.

[40] N. Grebenchtchikov, A. J. Geurts-Moespot, J. J. C. Kroot,et al., “High-sensitive radioimmunoassay for human serumhepcidin,” British Journal of Haematology, vol. 146, no. 3, pp.317–325, 2009.


[41] E. H. J. M. Kemna, A. E. R. Kartikasari, L. J. H. van Tits,P. Pickkers, H. Tjalsma, and D. W. Swinkels, “Regulation ofhepcidin: insights from biochemical analyses on human serumsamples,” Blood Cells, Molecules, and Diseases, vol. 40, no. 3,pp. 339–346, 2008.

[42] I. Theurl, E. Aigner, M. Theurl, et al., “Regulation of ironhomeostasis in anemia of chronic disease and iron deficiencyanemia: diagnostic and therapeutic implications,” Blood, vol.113, no. 21, pp. 5277–5286, 2009.

[43] M. Auerbach, D. Coyne, and H. Ballard, “Intravenous iron:from anathema to standard of care,” American Journal ofHematology, vol. 83, no. 7, pp. 580–588, 2008.

[44] G. Nicolas, C. Chauvet, L. Viatte, et al., “The gene encodingthe iron regulatory peptide hepcidin is regulated by anemia,hypoxia, and inflammation,” Journal of Clinical Investigation,vol. 110, no. 7, pp. 1037–1044, 2002.

[45] M. Munoz, I. Villar, and J. A. Garcıa-Erce, “An update on ironphysiology,” World Journal of Gastroenterology, vol. 15, no. 37,pp. 4617–4626, 2009.

[46] C. Finch, “Regulators of iron balance in humans,” Blood, vol.84, no. 6, pp. 1697–1702, 1994.

[47] T. H. Bothwell, A. V. Hurutado, D. M. Donohue, and C. A.Finch, “Erythrokinetics. IV. The plasma iron turnover as ameasure of erythropoiesis,” Blood, vol. 12, no. 5, pp. 409–427,1957.

[48] M. Cazzola, H. A. Huebers, M. H. Sayers, A. P. MacPhail, M.Eng, and C. A. Finch, “Transferrin saturation, plasma ironturnover, and transferrin uptake in normal humans,” Blood,vol. 66, no. 4, pp. 935–939, 1985.

[49] L. Lin, E. V. Valore, E. Nemeth, J. B. Goodnough, V. Gabayan,and T. Ganz, “Iron transferrin regulates hepcidin synthesisin primary hepatocyte culture through hemojuvelin andBMP2/4,” Blood, vol. 110, no. 6, pp. 2182–2189, 2007.

[50] G. Astaldi, G. R. Burgio, and A. Priolisi, “Evaluation of plasmairon turnover in Cooley’s anemia,” Annals of the New YorkAcademy of Sciences, vol. 165, no. 1, pp. 111–125, 1969.

[51] C. Peyssonnaux, V. Nizet, and R. S. Johnson, “Role of thehypoxia inducible factors in iron metabolism,” Cell Cycle, vol.7, no. 1, pp. 28–32, 2008.


[53] G. G. Braliou, M. V. Verga Falzacappa, G. Chachami,G. Casanovas, M. U. Muckenthaler, and G. Simos, “2-oxoglutarate-dependent oxygenases control hepcidin geneexpression,” Journal of Hepatology, vol. 48, no. 5, pp. 801–810,2008.


[55] D. Kuninger, R. Kuns-Hashimoto, R. Kuzmickas, and P.Rotwein, “Complex biosynthesis of the muscle-enriched ironregulator RGMc,” Journal of Cell Science, vol. 119, no. 16, pp.3273–3283, 2006.

[56] L. Lin, E. Nemeth, J. B. Goodnough, D. R. Thapa, V. Gabayan,and T. Ganz, “Soluble hemojuvelin is released by proproteinconvertase-mediated cleavage at a conserved polybasic RNRRsite,” Blood Cells, Molecules, and Diseases, vol. 40, no. 1, pp.122–131, 2008.

[57] L. Silvestri, A. Pagani, and C. Camaschella, “Furin-mediatedrelease of soluble hemojuvelin: a new link between hypoxiaand iron homeostasis,” Blood, vol. 111, no. 2, pp. 924–931,2008.

[58] L. Lin, Y. P. Goldberg, and T. Ganz, “Competitive regulation ofhepcidin mRNAby soluble and cell-associated hemojuvelin,”Blood, vol. 106, no. 8, pp. 2884–2889, 2005.

[59] J. L. Babitt, F. W. Huang, Y. Xia, Y. Sidis, N. C. Andrews, and H.Y. Lin, “Modulation of bone morphogenetic protein signalingin vivo regulates systemic iron balance,” Journal of ClinicalInvestigation, vol. 117, no. 7, pp. 1933–1939, 2007.

[60] M. Vujic Spasic, J. Kiss, T. Herrmann, et al., “Hfe acts inhepatocytes to prevent hemochromatosis,” Cell Metabolism,vol. 7, no. 2, pp. 173–178, 2008.

[61] A. Kattamis, I. Papassotiriou, D. Palaiologou, et al., “Theeffects of erythropoetic activity and iron burden on hepcidinexpression in patients with thalassemia major,” Haematologica,vol. 91, no. 6, pp. 809–812, 2006.



[64] J. P. Pinto, S. Ribeiro, H. Pontes, et al., “Erythropoietinmediates hepcidin expression in hepatocytes through EPORsignaling and regulation of C/EBPα,” Blood, vol. 111, no. 12,pp. 5727–5733, 2008.


[66] R.-H. Wang, C. Li, X. Xu, et al., “A role of SMAD4 iniron metabolism through the positive regulation of hepcidinexpression,” Cell Metabolism, vol. 2, no. 6, pp. 399–409, 2005.


[68] H. Yang, Z. Filipovic, D. Brown, S. N. Breit, and L. T. Vassilev,“Macrophage inhibitory cytokine-1: a novel biomarker for p53pathway activation,” Molecular Cancer Therapeutics, vol. 2, no.10, pp. 1023–1029, 2003.

[69] H. Tamary, H. Shalev, G. Perez-Avraham, et al., “Elevatedgrowth differentiation factor 15 expression in patients withcongenital dyserythropoietic anemia type I,” Blood, vol. 112,no. 13, pp. 5241–5244, 2008.

[70] J.-M. Ramirez, O. Schaad, S. Durual, et al., “Growth differen-tiation factor 15 production is necessary for normal erythroiddifferentiation and is increased in refractory anaemia withring-sideroblasts,” British Journal of Haematology, vol. 144, no.2, pp. 251–262, 2009.

[71] S. Lakhal, N. P. Talbot, A. Crosby, et al., “Regulation ofgrowth differentiation factor 15 expression by intracellulariron,” Blood, vol. 113, no. 7, pp. 1555–1563, 2009.

[72] A. E. Mast, T. M. Foster, H. L. Pinder, et al., “Behavioral,biochemical, and genetic analysis of iron metabolism in high-intensity blood donors,” Transfusion, vol. 48, no. 10, pp. 2197–2204, 2008.

[73] J. Kanda, C. Mizumoto, H. Kawabata, et al., “Serum hepcidinlevel and erythropoietic activity after hematopoietic stem celltransplantation,” Haematologica, vol. 93, no. 10, pp. 1550–1554, 2008.


[74] D. R. Ashby, D. P. Gale, M. Busbridge, et al., “Erythropoietinadministration in humans causes a marked and prolongedreduction in circulating hepcidin,” Haematologica, vol. 95, no.3, pp. 505–508, 2010.

[75] C. Chang, D. A. Holtzman, S. Chau, et al., “Twisted gastrula-tion can function as a BMP antagonist,” Nature, vol. 410, no.6827, pp. 483–487, 2001.

[76] J. J. Ross, O. Shimmi, P. Vilmos, et al., “Twisted is a conservedextracellular BMP antagonist,” Nature, vol. 410, no. 6827, pp.479–483, 2001.


[78] C. May, S. Rivella, A. Chadburn, and M. Sadelain, “Successfultreatment of murine β-thalassemia intermedia by transfer ofthe human β-globin gene,” Blood, vol. 99, no. 6, pp. 1902–1908, 2002.

[79] Y. Z. Ginzburg, A. C. Rybicki, S. M. Suzuka, et al., “Exogenousiron increases hemoglobin in β-thalassemic mice,” Experimen-tal Hematology, vol. 37, no. 2, pp. 172–183, 2009.

[80] S. R’zik, M. Loo, and Y. Beguin, “Reticulocyte transferrinreceptor (TfR) expression and contribution to soluble TfRlevels,” Haematologica, vol. 86, no. 3, pp. 244–251, 2001.

[81] G. Weiss, T. Houston, S. Kastner, K. Johrer, K. Grunewald,and J. H. Brock, “Regulation of cellular iron metabolismby erythropoietin: activation of iron-regulatory protein andupregulation of transferrin receptor expression in erythroidcells,” Blood, vol. 89, no. 2, pp. 680–687, 1997.

[82] M. Cazzola, Y. Beguin, G. Bergamaschi, et al., “Solubletransferrin receptor as a potential determinant of iron loadingin congenital anaemias due to ineffective erythropoiesis,”British Journal of Haematology, vol. 106, no. 3, pp. 752–755,1999.

[83] G. Metzgeroth, P. L. Rosee, C. Kuhn, et al., “The soluble trans-ferrin receptor in dysplastic erythropoiesis in myelodysplasticsyndrome,” European Journal of Haematology, vol. 79, no. 1,pp. 8–16, 2007.

[84] J. M. Flanagan, H. Peng, L. Wang, et al., “Soluble transferrinreceptor-1 levels in mice do not affect iron absorption,” ActaHaematologica, vol. 116, no. 4, pp. 249–254, 2006.

[85] M. Cazzola, M. G. Della Porta, and L. Malcovati, “Clini-cal relevance of anemia and transfusion iron overload inmyelodysplastic syndromes,” Hematology, pp. 166–175, 2008.

[86] S. Rivella, “Ineffective erythropoiesis and thalassemias,” Cur-rent Opinion in Hematology, vol. 16, no. 3, pp. 187–194, 2009.

[87] L. R. Solomon, R. S. Hillman, and C. A. Finch, “Serum ferritinin refractory anemias,” Acta Haematologica, vol. 66, no. 1, pp.1–5, 1981.

[88] P. L. Greenberg, N. S. Young, and N. Gattermann, “Myelodys-plastic syndromes,” Hematology, pp. 136–161, 2002.


Review Article

β-Thalassemia: HiJAKing Ineffective Erythropoiesis andIron Overload

Luca Melchiori, Sara Gardenghi, and Stefano Rivella

Weill Cornell Medical College, Department of Pediatrics, Division of Hematology-Oncology,515E 71st street, S702, New York, NY 10021, USA

Correspondence should be addressed to Stefano Rivella, [email protected]

Received 25 November 2009; Accepted 28 February 2010


Copyright © 2010 Luca Melchiori et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

β-thalassemia encompasses a group of monogenic diseases that have in common defective synthesis of β-globin. The defectsinvolved are extremely heterogeneous and give rise to a large phenotypic spectrum, with patients that are almost asymptomatic tocases in which regular blood transfusions are required to sustain life. As a result of the inefficient synthesis of β-globin, the patientssuffer from chronic anemia due to a process called ineffective erythropoiesis (IE). The sequelae of IE lead to extramedullaryhematopoiesis (EMH) with massive splenomegaly and dramatic iron overload, which in turn is responsible for many of thesecondary pathologies observed in thalassemic patients. The processes are intimately linked such that an ideal therapeutic approachshould address all of the complications. Although β-thalassemia is one of the first monogenic diseases to be described andrepresents a global health problem, only recently has the scientific community started to focus on the real molecular mechanismsthat underlie this disease, opening new and exciting therapeutic perspectives for thalassemic patients worldwide.

1. Introduction

The biochemical signature of β-thalassemia is a reducedsynthesis of the β-globin subunit of HbA (α2β2). Individualsinheriting two β-thalassemic alleles experience a profounddeficit in β-chain production, and this impairment leads toexcess production of α-globin. No compensatory regulatorymechanism exists where impaired synthesis of β-globinsubunit leads to an excess production of the α-globin.Therefore, in β-thalassemia, the excess α-globin chains formtetramers that accumulate and precipitate in the erythroidprogenitors, forming inclusion bodies that cause oxidativemembrane damage within the red blood cells and immaturedeveloping erythroblasts in the bone marrow. This leads topremature death of many late erythroid progenitors in thebone marrow and spleen (Hoffman et al., Hematology, BasicPrinciples and Practice). The profound anemia that resultsfrom production of only a few hypochromic and microcyticred blood cells leads to a dramatic increase in erythropoietin(EPO) levels, that ultimately drive an uncontrolled expan-sion of additional early erythroid progenitors inducing mas-sive EMH. These erythroid progenitors have an enhanced

proliferative and survival capacity, but eventually they failto differentiate, contributing to the process of IE [1]. Thestandard treatment for thalassemic patients is chronic bloodtransfusion to ameliorate the hemoglobin level. Chronictransfusion therapy is required to sustain life in patients withβ-thalassemia major and often becomes necessary in thosewith β-thalassemia intermedia who develop splenomegaly.The main side effect of transfusion therapy is the dramaticiron overload. Secondary iron overload is indeed one of themajor causes of morbidity in thalassemic patients. Excessiveiron in the circulation leads to abnormal accumulation inorgans such as liver, spleen and heart, leading ultimatelyto liver disease, cardiac dysfunction, arthropathy, gonadalinsufficiency, and other endocrine disorders (Hoffman et al.,Hematology, Basic Principles and Practice). β-thalassemiais also an iron loading anemia, meaning that thalassemicpatients have a dramatic increase in iron absorption fromthe gut due to their increased erythropoietic rate. Thisincreased iron absorption is mainly mediated by down-regulation of the ironregulatory hormone hepcidin [2–9],and together with the iron influx from chronic transfusionscontributes to the general setting of iron overload observed


in thalassemic patients. Current treatment for iron overloadincludes administration of iron chelators like desferoxamine,deferasirox, and deferiprone [10]. Those agents decrease theiron burden in the liver and heart, significantly increasingthe lifespan of thalassemic patients. Splenectomy is also oftennecessary to contain the iron burden when the transfusionrequirement becomes excessive (β-thalassemia major) or theanemia worsens (β-thalassemia intermedia). Splenectomy,however, introduces a new set of problems. Splenectomizedpatients have an increased risk of infections, of developingthrombotic events, and of pulmonary hypertension, all life-threatening conditions [11–14]. These observations suggest areconsideration of splenectomy in thalassemic patients. Theuse of agents able to reduce the spleen size thereby allowingpatients to avoid splenectomy could be an efficient optionin those who develop splenomegaly. All of the approachesdescribed above target secondary pathologies that resultfrom IE in β-thalassemia. Therefore, treatment of IE itselfcould ameliorate all secondary pathologies that stem from it.

2. Jak2 and Erythropoiesis

During the erythropoietic process, multipotent stem cellsin the bone marrow divide and differentiate giving rise toapproximately 2 million nonnucleated reticulocytes everysecond. This process is tightly regulated by an array of eventsthat include cytokine signaling and cell-cell interactions,mainly in the context of erythroblastic islands, a specializedniche for the maturation of erythroid progenitors [15].Since the pivotal role of erythrocytes is carrying oxygen totissues, the erythropoietic process must be able to respondquickly and effectively to changes in tissue oxygen tension.This is accomplished mainly by erythropoietin (EPO). EPOserves as the master regulator of erythropoiesis [16]. Itsignals through the erythropoietin receptor (EPOR) andcontrols virtually all stages of erythroid differentiation, fromcommitted common myeloid progenitors to the survival,proliferation, and maturation of more late-stage erythroidprogenitors such as proerythroblasts and basophilic ery-throblasts [17]. EPO exerts its effects by binding to EPOR,thereby activating the cytoplasmatic kinase Jak2. Activationof Jak2 involves auto and crossphosphorylation events thatultimately lead to activation of the signal transductorand activator of transcription Stat5 a and b and parallelsignaling pathways [18]. Once activated, Stat5 migrates tothe nucleus and activates genes crucial for proliferation,differentiation, and survival of erythroid progenitors. Thecrucial importance of the EPO-EPOR-JAK2-STAT5 axis hasbeen demonstrated by knock out studies in mice that showedhow lack of each one of these four molecules results in alethal phenotype (death due to severe anemia) during fetaldevelopment [19]. The binding of EPO to EPOR leads toactivation of another important signaling pathway duringerythropoiesis: the phosphoinositol-3-kinase (PI3K)—AKTpathway [20]. Studies in an apoptosis-resistant erythroidcell line indicate that activation of PI3K is necessary butnot sufficient to protect against apoptosis [21]. Moreover,mice that do not express the p85 α subunit of PI3K have

dramatically reduced erythropoiesis with reductions in thenumbers of CFU-E and BFU-E progenitors [22]. AKT, whichis the activated downstream of PI3K, transduces signals thatare necessary for the differentiation of erythroid progenitors[23] and is important in regulation of the activity of theFOXO3 transcription factor. The activity of FoxO3 is pivotalin the regulation of oxidative stress during erythropoiesis,as knockout mice exhibit a greater susceptibility to ROS-induced oxidative stress [24]. The absence of FoxO3 wasassociated with reduction in erythrocyte lifespan as well as anenhanced mitotic arrest in intermediate erythroid progenitorcells, resulting in a decreased rate of erythroid maturation.FoxO3-null erythrocytes also showed decreased expressionof ROS scavenging enzymes and evidence of oxidative dam-age. Jak2 is a member of the Jak tyrosine kinase family, andmediates the action of many other cytokines besides EPO.Some examples include growth hormone, prolactin, TPO,GM-CSF, interleukin 3, and interleukin 5. What makes theassociation between EPO and Jak2 unique is the fact that Jak2is the only kinase that is associated with EPOR and thereforethe only signal transductor of EPO. This important task isaccomplished by a complex 3D structure that involves kinase,pseudokinase, and regulatory domains. The 3D structureallows for a finely tuned regulation of its activity, from itsassociation with EPOR to its phosphorylation and activation[25]. When regulation fails, the consequences are deleterious.In 2005, five different research teams [26–30] described anactivating mutation (V617F) in the pseudokinase domain ofJak2 that is associated with ≥90% of cases of polycythemiavera and ∼50% of cases of essential thrombocythemia andchronic idiopathic myelofibrosis. In all of these studies, it wasclear how hyperactivation of Jak2 contributes to a massiveincrease in the erythropoietic activity and a huge expansionof the erythron with related splenomegaly, a phenotype thatis interestingly similar to what is observed in β-thalassemia.

2.1. Jak2, Ineffective Erythropoiesis and Splenomegaly in β-Thalassemia. In β-thalassemia, the erythropoietic process ismarkedly altered and is referred to as ineffective erythro-poiesis (IE). According to the traditional model, the lack orreduced synthesis of β-globin during IE induces the forma-tion of α-globin aggregates in erythroid progenitors. Theseaggregates precipitate and adhere to the membrane causingcellular damage, massive apoptosis of erythroid progenitorsin the bone marrow, and only limited production of redblood cells which are abnormal. These abnormal/damagedRBC are readily captured by the reticuloendothelial systemin the spleen and contribute to the splenomegaly observedin thalassemic patients. The hypoxia arising from the lackof production of normal RBC induces a dramatic increasein the levels of EPO, which in turn induces a bone marrowhyperplasia and bone deformities. This traditional view hasmainly focused on the apoptotic aspect of IE [31–33], whichis an important but not the only aspect of this process.Recent studies using both mouse models of β-thalassemiaand specimens from thalassemic patients showed that,together with apoptosis, a significant number of erythroidprogenitors undergo increased proliferation and decreaseddifferentiation in the spleen [1]. In this model, high levels


of EPO act as the driving force for the survival and prolif-eration of erythroid progenitors, albeit they fail to efficientlydifferentiate giving rise to only a few RBCs. This process hasbeen shown to be associated with the phosphorylated form ofJak2, leading to a higher number of proliferating thalassemicerythroid progenitors compared to normal conditions, ina sort of “physiological” gain of function (Figure 1). Thepersistent phosphorylation of Jak2 as a consequence of highEPO levels induces a massive extramedullary hematopoiesis(EMH), with early erythroid progenitors colonizing thenproliferating mainly in the spleen and liver. In this scenario,the spleen becomes a secondary erythropoietic niche, withits enlargement (splenomegaly) being due mainly to thecolonization and proliferation of erythroid progenitors fromthe bone marrow. In our study, we showed that erythroidprogenitors derived from the blood of thalassemic patientsexpress higher levels of cell cyclerelated mRNAs such asJak2, Ki67, Cyc A, BclXL, and EpoR, and that there area considerable number of erythroid progenitors in thespleen of thalassemic patients that are actively proliferating.Moreover, recent studies have shown that, at least in themouse, there are erythroid progenitors that develop inthe spleen, with different capacities for proliferation anddifferentiation than the ones found in the bone marrow.These erythroid progenitors display a higher sensitivity toEPO and are selectively responsive to BMP4. They showan improved capacity to respond to acute anemia and ahigher differentiation rate compared to their counterpartsin the bone marrow [34]. Another important study showshow the transcription factor ID1 is directly upregulated bythe Jak2-Stat5 pathway in erythroid cells [35]. Since highlevels of ID1 have been found to inhibit cell differentiation,its up-regulation due to a sustained activation of Jak2by EPO in β-thalassemia could explain the less matureerythroid progenitors observed. These findings shed lighton the mechanisms that underlie splenomegaly and raisethe intriguing question of whether there is also a basalphysiological level of erythropoietic activity in the humanspleen that is dramatically increased in stress conditionssuch as IE. Although these observations have been demon-strated in mouse models, further evaluation is necessaryin humans, especially considering the differences observedin the erythropoiesis in the spleen comparing these twospecies. Other mechanisms could be involved in the onsetof splenomegaly, like hepatic and portal vein obstruction,cirrhosis or congestive heart failure that often appear inthalassemic patients. All these conditions lead to an increasedblood flow to the spleen that in turn could be responsible forthe enlargement observed.

2.2. Possible Role of Jak2 Inhibitors in β-Thalassemia. Thediscovery of Jak2 as an important mediator of IE andsplenomegaly in β-thalassemia suggests that the use of smallorganic molecules to inhibit Jak2 could be beneficial inreducing IE and splenomegaly. The idea of treating anerythropoietic disorder with an agent that limits erythro-poiesis appears counterintuitive. However, the rationale forits use is related to the fact that IE in β-thalassemia resem-bles a leukemic blast expansion, with immature erythroid

progenitors that proliferate abnormally, fail to differentiate,and invade other organs to compromise their function [1].It is important to state that this is still a model that needsfurther confirmation, since mouse models of β-thalassemiaand leukemia have not been compared yet. Ideally, use ofa Jak2 inhibitor would target only the rapidly proliferat-ing erythroid progenitors in the spleen (CD71+ Ter119+early erythroid progenitors), blocking their expansion andtherefore allowing shrinkage of this organ by decreasing thepresence of red pulp. This in turn would contribute to animprovement of the spleen architecture and a reduction ofRBC sequestration, enhancing their lifespan. Jak2 inhibitorsare indeed able to induce a dramatic decrease in spleen size[1] in thalassemic animals with a limited effect on anemia.The drug would need to be carefully titrated in order toachieve an optimal plasma concentration that would allowthe inhibition of excessive proliferation or early progenitorswithout blocking erythropoiesis, as it would be expectedby completely inhibiting Jak2 [19]. This titration processwould also limit the potential for off-target effects on theimmune system by a Jak2 inhibitor, effects that howeverhave not been found relevant in other mouse models [36].However, it is important to point out that β-thalassemiamajor and β-thalassemia intermedia patients who developsplenomegaly require regular blood transfusions and oftenundergo splenectomy. It is appealing to speculate that β-thalassemia intermedia patients affected by splenomegalycould be treated temporarily with Jak2 inhibitors so asto reduce the spleen size and, in the presence of bloodtransfusions, to prevent further anemia. This suggests thateven patients affected by β-thalassemia major, who developsplenomegaly and EMH, may benefit from administrationof Jak2 inhibitors (Melchiori et al., in preparation). In thesesettings, the use of Jak2 inhibitors would be expected tolimit or reduce splenomegaly, thereby preventing or delayingthe need for splenectomy and indirectly improving themanagement of anemia and iron overload by reducing therate of blood transfusions. In conclusion, the use of Jak2inhibitors for β-thalassemia might be desirable, but it wouldrequire a careful optimization noting the potential for off-target immune suppression, as well as the anemia that wouldbe expected from continuous Jak2 inhibition.

2.3. Iron Overload as a Consequence of Ineffective Ery-thropoiesis. Erythropoiesis and iron metabolism are closelylinked processes. The large majority of iron in our body isutilized by the erythropoietic system to generate functionallyactive hemoglobin molecules, which are harbored in theRBC. It is estimated that 1800 mg of iron in our body arepresent in RBCs, 300 mg in the bone marrow, and 600 mgin the reticuloendothelial macrophages of the spleen [37],accounting for more than 60% of total body iron. Thismassive utilization of iron by the erythropoietic systemrequires the presence of finely tuned regulatory systemsthat allow storage, mobilization, and traffic of iron whilepreventing toxicity due to highly reactive iron ions. Itis also important to remember that our body lacks anefficient means of iron excretion. Therefore, its regulationoccurs primarily at the site of absorption in duodenal


Normal erythropoiesis

Progenitor erythroid cells

pJak2

RBC

RBCRetics

A new model for ineffective erythropoiesis

β-thalassemia major

(th3/th3)

β-thalassemia

intermedia (th3/+)β-thalassemia

Figure 1: In normal erythropoiesis, physiological levels of EPO induce the phosphorylation of Jak2 in normal erythroid progenitors andsustain the differentiation to mature RBC (top of the figure). In β-thalassemia, the high levels of EPO induce an uncontrolled proliferationof erythroid precursors, with a higher number of cells associated with the phosphorylated form of Jak2. In β-thalassemia intermedia, wherea certain amount of β-globin is still synthesized, there is a high production of reticulocytes that eventually mature in RBC. In β-thalassemiamajor, where there is a complete lack of β-globin, the orythroid progenitors continue to proliferate and fail to mature into reticulocytes andRBC or undergo apoptosis.

HampJAK2

inhibitors

Ineffectiveerythropoiesis

Ironoverload

Splenomegaly

Figure 2: Different approaches to target IE in β-thalassemia. Theadministration of hepcidin would allow to decrease the iron over-load in organs such the liver and heart and to redistribute the iron tohematopoietic organs, allowing a more efficient erythropoiesis andtherefore a decrease in splenomegaly. The administration of Jak2inhibitors would induce a decrease in the inefficient erythropoieticrate, therefore decreasing spleen size. The reduced erythropoiesiswould have as indirect effect the increase in serum hepcidin, thatin turn would decrease the iron absorption from the gut and theamelioration of iron overload.

enterocytes. Other important regulatory sites are the liver,where large quantities of iron can be stored in hepatocytesand Kuppfer cells, and the spleen, where macrophages recycleiron from senescent RBCs. All three compartments are of

pivotal importance for erythropoiesis since they control thebioavailability of iron to erythroid progenitors, and all ofthem respond to what has been found to be the masterregulator of iron metabolism, the peptide hormone hepcidin.Hepcidin [38, 39] is synthesized mainly in the liver, andwhen released into the blood stream, binds to ferroportin,the major cellular iron exporter expressed at high levels onthe surface of duodenal enterocytes, liver hepatocytes, andKuppfer cells, and splenic macrophages [40–43]. The bindingof hepcidin to ferroportin induces its internalization anddestruction in the cellular proteasome [44, 45]. Therefore,the main function of hepcidin is to reduce iron absorptionfrom enterocytes, and to limit iron export and traffickingfrom hepatocytes and splenic macrophages. The action ofhepcidin is critical in conditions of iron overload, when toomuch iron is potentially bioavailable and a drastic reductionin its absorption and mobilization is required. Hepcidinexpression is strictly regulated, with multiple pathways thatrespond to iron storage (storage regulator) [46], hypoxia(hypoxia regulator) [47, 48], inflammation (inflammatoryregulator) [49–52], and erythropoiesis (erythroid regulator)[51, 53, 54]. All these systems interact with one anotherresulting in a very complex and finely tuned regulation ofhepcidin expression. Since the intimate relationship betweenhepcidin and erythropoiesis exists, it is predictable thatpathologies that involve alteration in the erythropoietic rate,also involve alterations of iron metabolism and hepcidinexpression. In β-thalassemia, the process of IE is accompa-nied by a massive iron overload, due to an increased rateof iron absorption by the gastrointestinal (GI) tract andto frequent blood transfusions. In this setting, two majorsystems would contribute to hepcidin expression, the “storesregulator” and the “erythroid regulator”. Due to the highlevels of total body iron, the store regulator should act


by increasing hepcidin expression thereby avoiding furtheriron absorption. On the other hand, the erythroid regulatorwould decrease hepcidin expression in an attempt to com-pensate for anemia due to IE. Our recent study in mousemodels of β-thalassemia intermedia (th3/+) and major(th3/th3) showed that the erythroid regulator does dictatethe pattern of iron absorption and distribution relative tothe degree of IE (Gardenghi et al., [4]). In th3/+ mice, ironoverload and the degree of IE gradually become more severeas the animals age. Gardenghi and colleagues demonstratedhow dysregulation of iron absorption in young th3/+ miceis due mainly to a dramatic decrease of hepcidin expressionin the liver. However, as iron overload progressively increasesin older animals, hepcidin is upregulated, while ferroportinexpression is increased in the GI tract in order to maintainhigh levels of iron absorption to compensate for the anemiain IE. In th3/th3 mice, where IE is more pronounced, theincreased iron absorbed was not found in hematopoieticorgans such as the spleen and bone marrow, but rather in theliver and other nonhematopoietic organs [4]. This suggeststhat that iron is not utilized by erythroid progenitors, aswould be expected according to the model of IE describedabove where erythroid progenitors display an increasedproliferative activity but a decreased differentiation rate,resulting in a limited synthesis of hemoglobin and thereforelimited iron uptake. Instead, in th3/+ animals where thedegree of IE is lower and there is a substantial effectiveerythropoiesis, an increased iron content is found in thespleen and Kupffer cells of the liver [4]. This leads to theconclusion that the erythroid regulator overrides the storeregulator in th3/th3 mice, resulting in low levels of hepcidinexpression and further increasing the iron concentration inthe liver. In contrast, in states of relatively mild anemia, ironabsorption would be lower and the erythroid organs, spleenand bone marrow, would utilize part of the absorbed iron,as is observed in th3/+ animals. The idea that not all theiron absorbed in β-thalassemia is utilized for erythropoiesishas been confirmed by our recent data (Gardenghi et al.,submitted) in th3/+ animals kept on a low iron diet.These animals show a lower iron content compared tocounterparts fed a regular diet, but do not display a decreasein hemoglobin levels, suggesting again that an excessiveamount of iron is absorbed in β-thalassemia but is notutilized for the erythropoietic process.

2.4. Control of Hepcidin Expression by the Erythroid Regulator.New research data hints at the mechanisms by which theerythroid compartment controls hepcidin expression. In β-thalassemia, several factors have been identified and studiedas candidate hepcidin regulatory proteins including growthdifferentiation factor 15 (GDF15) [55], and human twistedgastrulation factor (TWSG1) [56]. Both of these two factorsare members of the TGFbeta superfamily, which controlsproliferation, differentiation, and apoptosis in numerouscells, and are secreted by erythroid precursors. TWSG1gene expression occurs early during erythroblast maturationcontrasting with the more sustained increase in GDF15expression in more mature hemoglobinized erythroblasts.GDF15, which is elevated in the sera of patients with

β-thalassemia, has the ability to down regulate the expressionof hepcidin in vitro, although the mechanism is still unclear.In fact sera from these patients also suppressed hepcidinexpression, albeit to a lesser degree, after immunoprecipi-tation of GDF15 [55]. Therefore, GDF15 may play a rolein hepcidin regulation when erythroid precursors undergocell death as occurs in the IE observed in β-thalassemiaand refractory anemia with ring sideroblasts (RARS) [57].TWSG1 has been shown to inhibit the upregulation ofhepcidin by bone morphogenic proteins 2 and 4 (BMP2,BMP4), mediated by Smad phosphorylation, in humanhepatocytes but was not BMP-mediated in murine hepa-tocytes [56]. Tanno and colleagues proposed that TWSG1might act with GDF15 to dysregulate iron homeostasis inβ-thalassemia. These represent new and exciting findingsthat can lead to novel discoveries and clinical applications.These factors will likely result in additional studies leading tothe potential characterization of novel pathways controllinghepcidin production, have the potential to be utilized asprognostic markers, and lead to novel clinically applicabletherapy in the future.

2.5. New Therapeutic Approaches to Limit Iron Absorption.The strong feedback of increased erythropoietic rate whichsuppresses hepcidin raises an important question aboutwhether agents that limit IE, such as Jak2 inhibitors, couldact indirectly as inducers of hepcidin expression. Our recentdata on th3/+ and transfused th3/th3 mice treated with aJak2 inhibitor suggest that this might be the case. Animalstreated with the drug showed increased levels of hepcidinin the liver compared to animals treated with a placebo.More importantly, the levels of hepcidin negatively correlatedwith the spleen weight in these animals, suggesting a strongconnection between spleen weight, erythropoietic activity,and hepcidin expression in β-thalassemia. Jak2 inhibitorscould have a more direct effect on iron metabolism. Thisis suggested by a recent study that shows how expression oftransferrin receptor 1 (TfR1) on erythroid cells depends onJak2-Stat5 signaling [58]. Perhaps the use of Jak2 inhibitorscould reduce the expression of TfR1 on erythroid cells, limit-ing iron uptake and potentially reducing the toxicity inducedby free iron ions. In this scenario, the therapeutic treatmentof β-thalassemia patients with Jak2 inhibitors could be usefulas it would target the two major complications of thispathology, IE with its related splenomegaly and the massiveiron overload. Therefore, not only the erythroid cells wouldbenefit from the lower iron load, but also liver parenchimalcells and cells in other tissues (i.e., heart) that are damagedin iron overload conditions. Another therapeutic approachto decrease iron overload might be to increase the expressionof hepcidin in thalassemic patients. Our recent study (Gar-denghi et al., submitted) showed that increasing the hepcidinlevels in th3/+ mice both by administration of exogenoussynthetic hepcidin and by overexpressing hepcidin reducesorgan iron overload resulting in a marked beneficial effecton hepatosplenomegaly and erythropoiesis. This reveals apotential role for hepcidin or hepcidin agonists in thetreatment of abnormal iron absorption in β-thalassemia andother related disorders.


3. Summary

Recent studies shed new light on the molecular mechanismsthat underlie IE in β-thalassemia. These findings suggestthat Jak2 plays a role in the onset of IE and splenomegaly,and show how the use of Jak2 inhibitors to limit theseprocesses could open exciting new therapeutic options forthalassemic patients. The large body of work on hepcidinis starting to reveal evidence of what has long been heldto be true, the existence of an “erythroid regulator”. Thediscovery of erythroid factors that regulate hepcidin produc-tion suggests possible new therapeutic targets to decreaseiron overload. Furthermore, the evidence that exogenoushepcidin administration can ameliorate organ iron loadwithout affecting anemia opens a new therapeutic fieldfor hepcidin agonists/mimetics involving the treatment ofdifferent kinds of hemochromatosis (Figure 2).

References

[1] I. V. Libani, E. C. Guy, L. Melchiori, et al., “Decreased differen-tiation of erythroid cells exacerbates ineffective erythropoiesisin β-thalassemia,” Blood, vol. 112, no. 3, pp. 875–885, 2008.

[2] K. Adamsky, O. Weizer, N. Amariglio, et al., “Decreasedhepcidin mRNA expression in thalassemic mice,” BritishJournal of Haematology, vol. 124, no. 1, pp. 123–124, 2004.

[3] L. Breda, S. Gardenghi, E. Guy, et al., “Exploring the roleof hepcidin, an antimicrobial and iron regulatory peptide, inincreased iron absorption in β-thalassemia,” Annals of the NewYork Academy of Sciences, vol. 1054, pp. 417–422, 2005.


[5] A. Kattamis, I. Papassotiriou, D. Palaiologou, et al., “Theeffects of erythropoetic activity and iron burden on hepcidinexpression in patients with thalassemia major,” Haematologica,vol. 91, no. 6, pp. 809–812, 2006.

[6] G. Papanikolaou, M. Tzilianos, J. I. Christakis, et al., “Hepcidinin iron overload disorders,” Blood, vol. 105, no. 10, pp. 4103–4105, 2005.


[8] O. Weizer-Stern, K. Adamsky, N. Amariglio, et al., “mRNAexpression of iron regulatory genes in β-thalassemia interme-dia and β-thalassemia major mouse models,” American Journalof Hematology, vol. 81, no. 7, pp. 479–483, 2006.

[9] S. Rivella, “Ineffective erythropoiesis and thalassemias,” Cur-rOpinHematol, vol. 16, no. 3, pp. 187–194, 2009.

[10] P. J. Giardina and R. W. Grady, “Chelation therapy in β-thalassemia: an optimistic update,” Seminars in Hematology,vol. 38, no. 4, pp. 360–366, 2001.

[11] S. E. Crary and G. R. Buchanan, “Vascular complications aftersplenectomy for hematologic disorders,” Blood, vol. 114, no.14, pp. 2861–2868, 2009.

[12] A. Eldor and E. A. Rachmilewitz, “The hypercoagulable statein thalassemia,” Blood, vol. 99, no. 1, pp. 36–43, 2002.

[13] A. Taher, G. Mehio, H. Isma’eel, and M. D. Cappellini, “Strokein thalassemia: a dilemma,” American Journal of Hematology,vol. 83, no. 4, p. 343, 2008.

[14] A. T. Taher, K. M. Musallam, W. Nasreddine, R. Hourani,A. Inati, and A. Beydoun, “Asymptomatic brain magneticresonance imaging abnormalities in splenectomized adultswith thalassemia intermedia,” Journal of Thrombosis andHaemostasis, vol. 8, no. 1, pp. 54–59, 2009.

[15] J. A. Chasis, “Erythroblastic islands: specialized microen-vironmental niches for erythropoiesis,” Current Opinion inHematology, vol. 13, no. 3, pp. 137–141, 2006.

[16] A. D’Andrea, G. Fasman, G. Wong, and H. Lodish, “Ery-thropoietin receptor: cloning strategy and structural features,”International Journal of Cell Cloning, vol. 8, supplement 1, pp.173–180, 1990.

[17] J. Fang, M. Menon, W. Kapelle, et al., “EPO modulation ofcell-cycle regulatory genes, and cell division, in primary bonemarrow erythroblasts,” Blood, vol. 110, no. 7, pp. 2361–2370,2007.

[18] D. M. Wojchowski, R. C. Gregory, C. P. Miller, A. K. Pandit,and T. J. Pircher, “Signal transduction in the erythropoietinreceptor system,” Experimental Cell Research, vol. 253, no. 1,pp. 143–156, 1999.

[19] H. Neubauer, A. Cumano, M. Muller, H. Wu, U. Huffstadt, andK. Pfeffer, “Jak2 deficiency defines an essential developmentalcheckpoint in definitive hematopoiesis,” Cell, vol. 93, no. 3, pp.397–409, 1998.

[20] T. D. Richmond, M. Chohan, and D. L. Barber, “Turning cellsred: signal transduction mediated by erythropoietin,” Trendsin Cell Biology, vol. 15, no. 3, pp. 146–155, 2005.

[21] H. Bao, S. M. Jacobs-Helber, A. E. Lawson, K. Penta, A.Wickrema, and S. T. Sawyer, “Protein kinase B (c-Akt),phosphatidylinositol 3-kinase, and STAT5 are activated byerythropoietin (EPO) in HCD57 erythroid cells but are con-stitutively active in an EPO-independent, apoptosis-resistantsubclone (HCD57-SREI cells),” Blood, vol. 93, no. 11, pp.3757–3773, 1999.

[22] H. Huddleston, B. Tan, F.-C. Yang, et al., “Functional p85αgene is required for normal murine fetal erythropoiesis,”Blood, vol. 102, no. 1, pp. 142–145, 2003.

[23] S. Ghaffari, C. Kitidis, W. Zhao, et al., “AKT induces erythroid-cell maturation of JAK2-deficient fetal liver progenitor cellsand is required for Epo regulation of erythroid-cell differen-tiation,” Blood, vol. 107, no. 5, pp. 1888–1891, 2006.

[24] D. Marinkovic, X. Zhang, S. Yalcin, et al., “Foxo3 is requiredfor the regulation of oxidative stress in erythropoiesis,” Journalof Clinical Investigation, vol. 117, no. 8, pp. 2133–2144, 2007.

[25] P. Saharinen, K. Takaluoma, and O. Silvennoinen, “Regulationof the Jak2 tyrosine kinase by its pseudokinase domain,”Molecular and Cellular Biology, vol. 20, no. 10, pp. 3387–3395,2000.

[26] E. J. Baxter, L. M. Scott, P. J. Campbell, et al., “Acquired muta-tion of the tyrosine kinase JAK2 in human myeloproliferativedisorders,” Lancet, vol. 365, no. 9464, pp. 1054–1061, 2005.

[27] C. James, V. Ugo, J.-P. Le Couedic, J. Staerk, et al., “A uniqueclonal JAK2 mutation leading to constitutive signalling causespolycythaemia vera,” Nature, vol. 434, no. 7037, pp. 1144–1148, 2005.

[28] R. Kralovics, F. Passamonti, A. S. Buser, et al., “A gain-of-function mutation of JAK2 in myeloproliferative disorders,”New England Journal of Medicine, vol. 352, no. 17, pp. 1779–1790, 2005.

[29] R. L. Levine, M. Wadleigh, J. Cools, et al., “Activating mutationin the tyrosine kinase JAK2 in polycythemia vera, essential


thrombocythemia, and myeloid metaplasia with myelofibro-sis,” Cancer Cell, vol. 7, no. 4, pp. 387–397, 2005.

[30] R. Zhao, S. Xing, Z. Li, et al., “Identification of an acquiredJAK2 mutation in polycythemia vera,” Journal of BiologicalChemistry, vol. 280, no. 24, pp. 22788–22792, 2005.

[31] F. Centis, L. Tabellini, G. Lucarelli, et al., “The importance oferythroid expansion in determining the extent of apoptosisin erythroid precursors in patients with β-thalassemia major,”Blood, vol. 96, no. 10, pp. 3624–3629, 2000.

[32] L. A. Mathias, T. C. Fisher, L. Zeng, et al., “Ineffectiveerythropoiesis in β-thalassemia major is due to apoptosisat the polychromatophilic normoblast stage,” ExperimentalHematology, vol. 28, no. 12, pp. 1343–1353, 2000.

[33] S. L. Schrier, “Pathophysiology of thalassemia,” Current Opin-ion in Hematology, vol. 9, no. 2, pp. 123–126, 2002.

[34] J. M. Perry, O. F. Harandi, and R. F. Paulson, “BMP4, SCF, andhypoxia cooperatively regulate the expansion of murine stresserythroid progenitors,” Blood, vol. 109, no. 10, pp. 4494–4502,2007.

[35] A. D. Wood, E. Chen, I. J. Donaldson, et al., “ID1 promotesexpansion and survival of primary erythroid cells and is atarget of JAK2V617F-STAT5 signaling,” Blood, vol. 114, no. 9,pp. 1820–1830, 2009.

[36] A. Pardanani, J. Hood, T. Lasho, et al., “TG101209, asmall molecule JAK2-selective kinase inhibitor potentlyinhibits myeloproliferative disorder-associated JAK2V617Fand MPLW515L/K mutations,” Leukemia, vol. 21, no. 8, pp.1658–1668, 2007.

[37] M. W. Hentze, M. U. Muckenthaler, and N. C. Andrews,“Balancing acts: molecular control of mammalian ironmetabolism,” Cell, vol. 117, no. 3, pp. 285–297, 2004.

[38] A. Krause, S. Neitz, H.-J. Magert, et al., “LEAP-1, a novelhighly disulfide-bonded human peptide, exhibits antimicro-bial activity,” FEBS Letters, vol. 480, no. 2-3, pp. 147–150, 2000.


[40] S. Abboud and D. J. Haile, “A novel mammalian iron-regulatedprotein involved in intracellular iron metabolism,” Journal ofBiological Chemistry, vol. 275, no. 26, pp. 19906–19912, 2000.

[41] F. Canonne-Hergaux, A. Donovan, C. Delaby, H.-J. Wang, andP. Gros, “Comparative studies of duodenal and macrophageferroportin proteins,” American Journal of Physiology, vol. 290,no. 1, pp. G156–G163, 2006.

[42] A. Donovan, A. Brownlie, Y. Zhou, et al., “Positional cloningof zebrafish ferroportin1 identifies a conserved vertebrate ironexporter,” Nature, vol. 403, no. 6771, pp. 776–781, 2000.


[44] I. De Domenico, D. M. Ward, C. Langelier, et al., “The molec-ular mechanism of hepcidin-mediated ferroportin down-regulation,” Molecular Biology of the Cell, vol. 18, no. 7, pp.2569–2578, 2007.


[46] C. Pigeon, G. Ilyin, B. Courselaud, et al., “A new mouseliver-specific gene, encoding a protein homologous to humanantimicrobial peptide hepcidin, is overexpressed during Iron

overload,” Journal of Biological Chemistry, vol. 276, no. 11, pp.7811–7819, 2001.


[48] D. Yoon, Y. D. Pastore, V. Divoky, et al., “Hypoxia-induciblefactor-1 deficiency results in dysregulated erythropoiesis sig-naling and iron homeostasis in mouse development,” Journalof Biological Chemistry, vol. 281, no. 35, pp. 25703–25711,2006.




[52] D. M. Wrighting and N. C. Andrews, “Interleukin-6 induceshepcidin expression through STAT3,” Blood, vol. 108, no. 9,pp. 3204–3209, 2006.





[57] J.-M. Ramirez, O. Schaad, S. Durual, et al., “Growth differen-tiation factor 15 production is necessary for normal erythroiddifferentiation and is increased in refractory anaemia withring-sideroblasts,” British Journal of Haematology, vol. 144, no.2, pp. 251–262, 2009.

[58] M. A. Kerenyi, F. Grebien, H. Gehart, et al., “Stat5 regulatescellular iron uptake of erythroid cells via IRP-2 and TfR-1,”Blood, vol. 112, no. 9, pp. 3878–3888, 2008.


Review Article

Iron Chelation Therapy in Myelodysplastic Syndromes

Emanuela Messa, Daniela Cilloni, and Giuseppe Saglio

Division of Hematology and Internal Medicine, Department of Clinical and Biological Sciences of the University of Turin,Regione Gonzole 10, 10043 Orbassano (To), Italy

Correspondence should be addressed to Giuseppe Saglio, [email protected]

Received 1 October 2009; Revised 18 January 2010; Accepted 20 April 2010

Academic Editor: Jeffery Lynn Miller

Copyright © 2010 Emanuela Messa et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Myelodysplastic syndromes (MDS) are a heterogeneous disorder of the hematopoietic stem cells, frequently characterized byanemia and transfusion dependency. In low-risk patients, transfusion dependency can be long lasting, leading to iron overload.Iron chelation therapy may be a therapeutic option for these patients, especially since the approval of oral iron chelators, whichare easier to use and better accepted by the patients. The usefulness of iron chelation in MDS patients is still under debate, mainlybecause of the lack of solid prospective clinical trials that should take place in the future. This review aims to summarize what iscurrently known about the incidence and clinical consequences of iron overload in MDS patients and the state-of the-art of ironchelation therapy in this setting. We also give an overview of clinical guidelines for chelation in MDS published to date and someperspectives for the future.

1. Introduction

Myelodysplastic syndromes (MDS) are clonal disorders ofthe hematopoietic stem cell and are mainly characterizedby bone marrow blasts up to 20%, one or more peripheralcytopenias and bone marrow dysplasia [1]. The prognosiscan be variable, with survival ranging from a few monthsto many years, and depends on the three features evaluatedby the International Prognostic Scoring System (IPSS) [2]:cytogenetic abnormalities, percentage of blasts in the bonemarrow, and number of peripheral cytopenias. Among sowide a spectrum of clinical features, clinicians must choosefrom different therapeutic options (reviewed in [3]), thatvary from supportive care or growth factor administration tochemotherapy or bone marrow transplantation in youngerand higher risk cases. New therapeutic options are nowavailable; some of them, addressed to patients with specificcytogenetic features, such as lenalidomide for patients with5q-. Other promising medications are the hypomethylatingagents, which may improve survival in higher risk subjects[3].

Lower-risk MDS patients often become transfusion-dependent during the course of the disease, which can bequite long, and this could contribute to increased cardiac

morbidity and mortality. It is well known that transfusiondependency is an important prognostic factor in MDS andportends worse prognosis [4]. Similarly, high ferritin level inrefractory anemia, but not in refractory cytopenia withoutanemia, is a negative prognostic factor for survival [5].However, it should be kept in mind that more aggressivedisease is frequently associated with a high transfusionrate, so significant transfusion dependency often becomes asurrogate marker of aggressive disease.

Blood transfusion therapy may lead to organ toxicitydue to the formation of nontransferrin bound iron (NTBI)and resulting oxidative stress, as is well documented intransfusion-dependent congenital anemias [6]. In low-riskMDS patients with longer life expectancy, preventing damagedue to iron overload is an important concern. In fact,several authors underline the importance of iron chelationas a prognostic variable for improving survival [7, 8]. Thedata, however, are mainly derived from retrospective studiesand need to be confirmed in prospective trials. Recently,Sanz et al. reported that iron overload (serum ferritin over1000 ng/mL) has a negative prognostic impact on leukemia-free survival [9], suggesting that reactive oxygen species(ROS), increased in iron overload, may be responsible for


DNA damage and disease progression in multiply transfusedpatients. Furthermore, other therapeutic indications haveto be considered for iron chelation in MDS patients:pretransplant high ferritin level has a negative prognosticsignificance for survival in acute myeloid leukemia (AML)and MDS patients undergoing bone marrow transplantation[10, 11]; so many reports underline the usefulness ofiron chelation for higher risk candidates for allografting.Moreover, Pullarkat et al. [12] recently have proposed thatiron chelation may lower infection risk, delay leukemictransformation, and improve hematologic parameters inpatients with higher IPSS scores who are not at thispoint generally considered to be candidates for chelationaccording to multiple guidelines [13–21]. This paper brieflysummarizes what it is currently known about iron overloadand iron chelation in MDS.

2. Iron Overload in MDS

Chronic transfusion therapy is the main cause of ironoverload in MDS patients. It is well known from patientsaffected by congenital anemias that multiple transfusionleads to the formation of NTBI, which includes the labile ironpool (LIP) [6, 22]. This is defined as the chelatable iron, ableto reach tissues and cells, and is thought to be responsiblefor tissue damage, fibrosis and organ failure, mainly affectingthe liver, heart and pancreas [6]. Several reports describeorgan damage possibly due to multiple transfusions in MDSpatients [23, 24]. However, comorbidities often exist, makingit difficult to discern how much organ damage is due totransfusion therapy as opposed to age-related comorbidities.Iron overload may have many clinical consequences in MDSpatients, which are briefly summarized below.

3. Iron Overload and Impact on Survival andLeukemic Evolution

Several retrospective studies suggest an important contri-bution of transfusion dependency in shortening survival inMDS patients [25]. Malcovati et al. [26] reported in a seriesof 467 MDS patients a high incidence of cardiac failure(accounting for 51% of non leukemic deaths) in transfusion-dependent MDS patients. It is not yet clear how significantis the contribution of iron overload from transfusions, andhow much is due to a higher disease severity in transfusion-dependent patients. However, a reduced survival rate hasbeen observed in refractory anemia (RA) and refractoryanemia with ringed sideroblasts (RARS) patients but notin refractory cytopenia with multilineage dysplasia (RCMD)patients [5]. More recently, the same group highlightedthe importance of transfusion dependency for prognosiswhether we consider leukemia-free or overall survival [27,28]. Similarly, Cermak et al. [29] reported a statisticallysignificant negative impact on survival of transfusi ondependency in MDS with erythroid dysplasia without excessblasts. The importance of iron overload on morbidity andmortality has been also shown by Takatoku et al. [30]. In thisretrospective study which included 292 patients affected by

MDS, myelofibrosis and aplastic anemia, 97% of deaths fromcardiac and hepatic failure were observed in patients with aserum ferritin level over 1000 g/L. More recently, Sanz et al.analyzed the impact of iron overload in a multicenter study ofthe Spanish group including 2994 patients [9]. A multivariateanalysis performed on 902 cases with complete data availabledemonstrated a strong association between not only highserum ferritin level and reduced survival (P < .0001), butalso leukemia-free survival (P < .0001) [7]. This is thefirst study evaluating the correlation between iron overloadand progression to leukemia, possibly suggesting a negativeinfluence of reactive oxygen species in inducing increasedgenomic instability in hematopoietic progenitors, as has beendemonstrated by in vitro and in vivo studies [31–33].

However, a strong and well-documented relationbetween shorter survival and iron overload is still lacking[34]. We need prospective studies able to correlatemore precisely iron overload parameters such as NTBImeasurement and appropriate magnetic imaging of the liverand heart to survival and leukemic progression.

4. Cardiac Iron and Transfusion Rate

Iron deposition in myocardial tissue is one of the mostconcerning potential complications in MDS patients receiv-ing chronic transfusion therapy. It is well known thatsecondary hemocromatosis in thalassemia patients and inother congenital anemias leads to cardiac iron deposition andheart failure [35]. Many studies in recent years addressingthis point in transfused MDS patients have been published.In this clinical setting we must consider that there isa shorter transfusional history compared with congenitalanemias, and that other factors besides iron depositioncould be involved in increased cardiac failure in transfusedpatients, such as anemia and MDS severity. Cardiac ironaccumulation in MDS can be variable and there are manystudies reporting conflicting results in this regard. In anautopsy series of transfusion-dependent patients, Buja andRoberts [36] found a direct correlation between the numberof red cell transfusions and the amount of iron deposition.More recently, studies based on MRI T2∗ imaging suggestthat cardiac iron accumulation is not a frequent feature ofMDS patients [37] and does not correlate with serum ferritinlevels or hepatic iron overload [38, 39] but only with the sizeof the chelatable iron pool [40].

5. Iron Overload and Risk of Infection

There are numerous observations suggesting an increasedrisk of infection in iron loaded patients, particularly thoseaffected by acute leukemia or allogenic stem cell trans-plantation (SCT) recipients [41–43]. In these conditions,there is a coexistence of many factors able to enhanceinfectious susceptibility, such as neutrophil dysfunction,severe cytopenias and an increase in labile plasma iron (LPI)due either to a high transfusion rate or to myelosuppressivetherapies. However, even if iron chelation can be useful inthis setting in preventing bacterial and fungal infections,


it is important to correctly choose the appropriate chela-tion agent: it has been suggested that deferoxamine canworsen fungal infections by acting as a siderophore [44]. Incontrast, it has been demonstrated that deferasirox can beeffective against mucormicosis infection in vivo [45]. So, assuggested recently by Pullarkat, iron chelators could havea possible new role in the future, lowering infection risksin MDS, expecially in high risk patients with more severeneutropenia [12].

6. Iron Overload in Transplantation

In recent years, different studies highlighted iron overloadas a negative prognostic indicator in patients undergoingstem cell transplantation (SCT). In a cohort of 590 patientswho underwent myeloablative SCT, Armand et al. [10]found a strong negative association between high serumferritin level and overall and disease-free survival. Thisassociation was seen in patients affected by acute myeloidleukemia or myelodysplastic syndrome. The five-year overallsurvival decreased from 54% for patients with serum ferritinunder 231 ng/mL to 27% for those with serum ferritinhigher than 2034 ng/mL. Additionally, in MDS patients aserum ferritin higher than 2034 ng/mL was associated witha significantly increased transplant-related mortality (hazardratio = 3,2), no correlation was found between serumferritin level and relapse rate, however. Similar results havealso been published by Pullarkat et al., who analyzed 190patients undergoing SCT: in a multivariate analysis highpretransplant serum ferritin (above 1000 ng/ml) stronglycorrelated with an increased risk of death (hazard ratio 2,28,P = .004), day 100 mortality (odds ratio in generalizedlinear models of 3,82, P = .013), bloodstream infection(odds ratio = 3, 11) and acute graft versus host disease(odds ratio = 1, 99) [11]. The 190 patients were followedfor a mean of 209 days: 27 died in the first 100 daysand 29 after this cut off. The results of Cox proportionalhazards model showed a higher risk of death in high-ferritingroup (P = .004). More recently, Mahindra et al. studied222 patients who underwent myeloablative allogeneic bonemarrow transplantation: pretransplantation serum ferritinabove 1910 μg/l was strongly associated with lower overalland relapse-free survival (P = .003 for both the variables)and with lower incidence of chronic and acute graft versushost disease (P = .019 and 0,10, resp.). No differences wereobserved in relapse mortality and incidence of hepatic veno-occlusive disease between patients stratified according theserum ferritin value of 1910 μg/l [46]. A higher morbidityrate in iron loaded transplant recipients has been shown byAu et al. [47]: high iron levels measured by T2∗ MRI mostlyin the liver, pancreas and pituitary gland were associated withabnormal pancreas and pituitary function in 40%–70% ofpatients studied.

During the SCT procedure and follow up, serum fer-ritin and NTBI levels are markedly increased by severalmechanisms (reviewed in [48]), increasing the risk of infec-tions, mucositis, chronic liver disease, sinusoidal obstructionsyndrome and idiopathic pneumonia. Therefore, reducing

iron burden after SCT may be helpful in lowering the riskof infection [49] and organ toxicity, and can be achievedby phlebotomy in patients with a normal hemoglobin.Prospective clinical trials to evaluate the effectiveness andsafety of iron chelators in this setting are ongoing [48].

7. Iron Chelation in MDS

7.1. Iron Chelators Commercially Available. Currently, threeiron chelators are commercially available: deferoxamine(DFO), deferasirox (DFR) and deferiprone (DFP). DFO wasintroduced in the 1970s and had a profound impact on thesurvival of thalassemia patients [50]. In such patients thedrug reduced organ dysfunctions and mortality, restoring alife expectancy similar to normal individuals with a survivalrate strictly correlating to days of deferoxiamine infusions[51]. Due to its short half life, DFO is administered bysubcutaneous infusion 5 to 7 nights/week, though it may bedifficult to use in MDS patients due to thrombocytopenia.More recently, two oral agents have been introduced: thethree-times daily agent deferiprone and the once-dailydeferasirox [52–54]. DFP is effective in reducing hepaticand cardiac iron content in thalassemia patients, but itsclinical use is partially limited by the risk of occurrenceof agranulocytosis. This concerning adverse effect is rare,but potentially harmful in thalassemia and also in MDS,where pancytopenia is a common clinical feature. Thelatter agent, deferasirox, is administered once daily dueto a long half life, and has been recently released on themarket for the treatment of secondary iron overload intransfusion-dependent anemias. Several clinical studiesevaluating the efficacy and safety of deferasirox in manytransfusion-dependent congenital anemias have beenpublished [55–58]. Moreover, its usefulness has also beentested in a cohort of MDS patients, yielding good data onefficacy and safety in this older population [59, 60]. Itsside effects are generally mild, consisting mainly of nausea,diarrhea and a self-limiting serum creatinine increase, thusmaking this agent possibly the most suitable for chelationtherapy in the MDS population [61]. We should also keepin mind the recent guidance from FDA regarding theuse of deferasirox in MDS patients where a greater riskfor adverse events such as kidney failure, gastrointestinalhemorrhage and deaths is reported for myelodysplasticpatients compared to others without this condition (http://www.fda.gov/Drugs/DrugSafety/PostmarketDrugSafetyInfo-rmationforPatientsandProviders/DrugSafety).

7.2. Impact of Iron Chelation on Survival in MDS. Two retro-spective studies reported a positive impact of iron chelationon the survival of MDS patients: Leitch et al. analyzed apopulation of 178 MDS patients from the University ofBritish Columbia database, revealing a positive associationbetween iron chelation and survival in a subcohort of28 matched patients [7]. In a multivariate analysis ironchelation therapy (based on deferoxamine administration bysubcutaneous infusion for at least 12 hours per day and 5days per week) was a significant factor for overall survival


(P = .01, hazard ratio 0,29). Similar results have beenobtained also in a partially prospective analysis performedon a population of 170 MDS patients from 18 hematologicalcenters in France. Rose et al. analyzed survival data ina cohort of MDS patients referred for transfusions in amonth period in all the French centers involved. Data onsurvival were collected prospectively whereas transfusionand previous clinical history retrospectively. Among thecohort of iron chelated patients, 19 received deferoxaminetreatment by intermittent bolus, 57 DFO by infusion formore than 3 days per week or, alternatively, DFP or DFR.The authors observed an improvement in overall survivalin patients who underwent iron chelation therapy using amultivariate analysis: overall survival was 115 months inthe chelated group versus 51 months in the nonchelatedpatients with a statistically significant difference (P = .0001)[8]. Although these data are promising, ad hoc prospective,controlled studies are needed in order to clarify the impactof iron chelation on overall survival and leukemic evolutionin MDS.

7.3. Guidelines for Iron Chelation Therapy in MDS. Severalinternational guidelines, based mostly on the opinions ofexpert panels, have been published in recent years, and theirmain features are summarized in Table 1 and reviewed byGattermann [13]. Concerning the threshold of the number ofred blood cell (RBC) transfusion for starting iron chelation,the Italian guidelines recommend chelation for patientswho reached an average of 50 RBC units, but only if theyhave an expected lifespan of 6 months [14], while the UKexpert panel considers candidates for iron chelation MDSpatients who have received 25 units of blood and whowill receive ongoing transfusions [15]. A similar number ofprevious transfusions is included in the NCCN (NationalComprehensive Cancer network) guidelines, where ironchelation is suggested after the receipt of 20–30 units of RBC[16]. The Japanese guidelines considered a higher thresholdof RBC units transfused; however, it should be noted that theunit in this country is smaller [17].

Most guidelines suggest as an indicator of iron overloada cut-off of serum ferritin level: even if ferritin is notsuch a reliable indicator of iron overload, it is easilyobtained in clinical practice. The ferritin values indicatedvary slighty in the guidelines (see Table 1), but the majorityconsidered 1,000 g/L as a reasonable threshold for startingiron chelation. Serum ferritin evaluation is recommendedduring iron chelation to establish efficacy of the treatment atleast every three months in transfusion-dependent patients.A useful and more reliable tool in monitoring iron overloadand iron chelation therapy is liver magnetic resonance but itis not available in all centers. Monitoring of organ functions(mainly cardiac, hepatic and endocrine) with appropriatetechniques is frequently indicated, other parameters such asROS, NTBI and LPI are under investigation [61]. Durationof chelation has to be individually established and thetherapy has to be maintioned as long as transfusion therapycontinues or as long as iron overload remains clinicallyrelevant [18].

In summary, the majority of authors are in agreementthat iron chelation should be offered to lower-risk MDSpatients with a life expectancy more than one year anda serum ferritin above the mentioned threshold. Patientswho are candidates for allogenic transplantation or othereffective therapies can also be considered candidates forchelation.

7.4. Improvement in Hematopoiesis with Chelation. In somepatients with MDS receiving iron chelation, improvedhematopoiesis has been noted. There are some data regardingthis effect with deferoxamine and deferiprone. In 1996Jensen et al. described a hemoglobin improvement in 11MDS patients treated with deferoxamine for up to 60months, but in some cases there was even a trilineageresponse [62, 63]. It has also been reported that patientswith myelofibrosis have a similar response to deferiprone[64]. In all of these reports, the hematological responsewas associated with a sharp decrease in iron burden andthe effect was fairly long lasting, in the order of one year.The effect of deferasirox therapy seems to differ: severalreports and our own experience have demonstrated anunexpected improvement in hemoglobin level and reductionin transfusion requirements even after only a few monthsof chelation [65–67]. Three out of five patients described inthese reports became transfusion independent. A hemato-logic improvement by World Health Organization (WHO)criteria has also been recently described in clinical trialswith deferasirox in low and int-1 MDS patients [68]. Thehemopoiesis improvement has not yet been reported in anypatients affected by thalassemia or other hereditary anemias,but it seems not restricted to MDS patients, as severalcases of myelofibrotic patients have been observed. It is notyet clear which mechanism could lead to the hemoglobinimprovement and which patients could benefit from thiseffect. Removing excess iron from erythroid precursorscould improve erythropoiesis, similar to what occurred ina case of sideroblastic congenital anemia due to GLRX5deficiency described recently by Camaschella et al. wherean efficient iron chelation obtained by DFO subcutaneousinfusion led the patients to transfusion independence due toa rebalancing in iron-responsive protein 1 and 2 functionsand intracellular iron distribution [69]. Also observed was areduced growth of erythroid colonies in vitro in iron-loadedpatients: the study performed by Hartmann et al. [70] on 42peripheral blood samples of MDS patients demonstrated astatistically significant (P < .004) reduced erythroid colonyformation in patients with high serum ferritin levels thussupporting the hypothesis that iron chelation could inducean improvement in hemopoiesis, perhaps by reducing NTBI-mediated toxicity or, alternatively, the phenomenon could beexplained by the more severity of bone marrow disease intransfusion-dependent patients. It should be kept in mind,however, that changes in hemopoiesis have been seen onlyin a portion of patients, while NTBI detoxification occursin all chelated subjects according to the drug used and itstype of administration. Our hypothesis to account for thisdiscrepancy is that deferasirox could interfere directly with


Ta

ble

1:O

verv

iew

ofth

epu

blis

hed

guid

elin

esfo

rir

onch

elat

ion

ther

apy

inM

DS

pati

ents

.RB

C(r

edbl

ood

cell)

,IP

SS(I

nte

rnat

ion

alP

rogn

osti

cSc

orin

gSy

stem

).

Gu

idel

ines

Ital

y20

02U

K20

03N

agas

aki

Con

sen

sus

2005

Can

ada

2008

Spai

n20

08Ja

pan

2008

MD

SFo

un

dati

on20

08A

ust

ria

2008

Isra

el20

08N

CC

N20

09

Tran

sfu

sion

stat

us

Rec

eive

d≥5

0R

BC

un

its

Rec

eive

d∼2

5R

BC

un

its

Tran

sfu

sion

depe

nde

nt

Tran

sfu

sion

depe

nde

nt

Tran

sfu

sion

depe

nde

nt

Rec

eive

d>

40Ja

pan

ese

un

its

Rec

eive

d2

RB

Cu

nit

s/m

onth

for≥1

year

Tran

sfu

sion

depe

nde

nt

Rec

eive

d20

–25

RB

Cu

nit

sor

mor

eth

an2

RB

Cu

nit

s/m

onth

for

6m

onth

s

Rec

eive

d20

–30

RB

Cu

nit

sto

con

tin

ue

tran

sfu

sion

s

Seru

mfe

rrit

inle

vel(

ng/

mL)

——

>10

00–2

000

>10

00>

1000

>10

00>

1000

>20

00>

1000

>25

00

Pati

ents

Life

-exp

ecta

ncy

>6

mon

ths

Lon

g-te

rmsu

rviv

allik

ely

(eg,

pure

side

robl

asti

can

emia

,del

(5q)

)

Low

orIn

t-1

IPSS

Ref

ract

ory

anem

ia(R

A),

RA

wit

hri

nge

dsi

dero

blas

ts(R

AR

S),d

el(5

q)

Low

orIn

t-1

IPSS

,RA

,RA

RS,

del(

5q)

(In

t-2

orH

igh

IPSS

)

IPSS

Low

orIn

t-1,

Low

risk

(Spa

nis

hpr

ogn

osti

cin

dex)

Life

-exp

ecta

ncy

>1

year

Life

-ex

pec

tan

cy>

1ye

ar

Life

exp

ecta

ncy

>2

year

s

Life

-ex

pec

tan

cy>

1ye

ar

Low

orIn

t-1

IPSS


oncogenes in the MDS blast cell. Nuclear factor-kappaB (NF-κB) is a key regulator of many cellular processes and itsimpaired activity has been described in different myeloidmalignancies including MDS and could be a good potentialtarget of deferasirox activity. Preliminary in vitro analysisdemonstrated a deferasirox-dependent reduction in bothnuclear localization and activity of NF-κB in MDS cells andleukemia cell lines [71]. NF-κB effects were not detected incells incubated with DFO or DFP. The NF-κB inhibitionby deferasirox could prove to be an important therapeuticoption in higher risk MDS patients, targeting blast cells inwhich increased NF-κB activity has been extensively demon-strated, and acting as a possible enhancer of chemosensitivityof the neoplastic clone. Our promising hypothesis needsobviously to be validated in vivo by prospective clinicaltrials.

8. Conclusion

The usefulness and clinical indications of iron chelationtherapy in myelodysplastic syndrome patients are still underdebate within the scientific community [12, 72, 73]. Manyauthors underline the deleterious effect of iron deposition intissues due to secondary hemochromatosis, similar to whatoccurs in the well-studied setting of congenital anemias.These data are mainly derived from clinical studies showingthat transfusion dependence and high serum ferritin levelsare independent prognostic variables for reduced overall anddisease-free survival [4, 5]. High serum ferritin level wasalso recently identified as a prognostic factor for shortertime to progression to leukemia [9]. There are also datashowing an impact of iron chelation on survival, thoughthese data are from retrospective studies [7, 8]. The mainconcern of experts not in favour of iron chelation is thatthere is insufficient evidence about the contribution of ironoverload to mortality rate and cardiac dysfunction in MDSpatients, and about a positive impact of iron chelation onimproving organ dysfunction, survival and quality of life.Strong prospective studies addressing these points are neededin the coming years, especially to address a putative survivaleffect. It is not yet clear which MDS patients could benefitmost from iron chelation therapy: international guidelinesare not completely in agreement on serum ferritin thresholdlevel or number of RBC units transfused for starting ironchelation [13]. Candidates for this therapy are mainlylower-risk patients with a better prognosis and a longerlife expectancy, but patients who are candidates for SCTor chemotherapy should also be included. More recently,Pullarkat proposed to change our thinking about restrictingchelation to patients with lower risk MDS, consideringrecent data about leukemic progression and incidence ofinfection in iron loaded patients [12]. Finally, the data abouthemopoiesis improvement and NF-κB activity inhibitionhave the potential to provide new insight into iron chelationand its beneficial effects, all suggesting that higher risk MDSpatients may also be good candidates for this therapy [66–68, 71]. In our opinion, we should change our thinking aboutiron chelation in MDS, expanding its indications from a

mere supportive therapy for preventing organ damage due toRBC transfusions, to an active measure against progressionto leukemia.

References

[1] C. Aul, D. T. Bowen, and Y. Yoshida, “Pathogenesis, etiologyand epidemiology of myelodysplastic syndromes,” Haemato-logica, vol. 83, no. 1, pp. 71–86, 1998.

[2] P. Greenberg, C. Cox, M. M. LeBeau et al., “Internationalscoring system for evaluating prognosis in myelodysplasticsyndromes,” Blood, vol. 89, no. 6, pp. 2079–2088, 1997.

[3] A. F. List, “Treatment strategies to optimize clinical benefit inthe patient with myelodysplastic syndromes,” Cancer Control,vol. 15, no. 4, pp. 29–39, 2008.

[4] M. Cazzola and L. Malcovati, “Myelodysplastic syndromes—coping with ineffective hematopoiesis,” The New EnglandJournal of Medicine, vol. 352, no. 6, pp. 536–538, 2005.

[5] L. Malcovati, M. G. Della Porta, and M. Cazzola, “Predictingsurvival and leukemic evolution in patients with myelodys-plastic syndrome,” Haematologica, vol. 91, no. 12, pp. 1588–1590, 2006.

[6] Z. I. Cabantchik, W. Breuer, G. Zanninelli, and P. Cianciulli,“LPI-labile plasma iron in iron overload,” Best Practice andResearch: Clinical Haematology, vol. 18, no. 2, pp. 277–287,2005.

[7] H. A. Leitch, C. S. Leger, T. A. Goodman, et al., “Improvedsurvival in patients with myelodysplastic syndrome receivingiron chelation therapy,” Clinical Leukemia, vol. 2, pp. 205–211,2008.

[8] C. Rose, S. Brechignac, D. Vassilief, et al., “Positive impact ofiron chelation therapy (CT) on survival in regularly transfusedMDS patients. A prospective analysis by the GFM,” Blood, vol.110, article 249a, 2007.

[9] G. Sanz, B. Nomdedeu, E. Such, et al., “Independent impactof iron overload and transfusion dependency on survivaland leukemic evolution in patients with myelodysplasticsyndrome,” Blood, vol. 112, article 640a, 2008.

[10] P. Armand, H. T. Kim, C. S. Cutler et al., “Prognostic impactof elevated pretransplantation serum ferritin in patientsundergoing myeloablative stem cell transplantation,” Blood,vol. 109, no. 10, pp. 4586–4588, 2007.

[11] V. Pullarkat, S. Blanchard, B. Tegtmeier et al., “Iron overloadadversely affects outcome of allogeneic hematopoietic celltransplantation,” Bone Marrow Transplantation, vol. 42, no.12, pp. 799–805, 2008.

[12] V. Pullarkat, “Objectives of iron chelation therapy inmyelodysplastic syndromes: more than meets the eye?” Blood,vol. 114, no. 26, pp. 5251–5255, 2009.

[13] N. Gattermann, “Guidelines on iron chelation therapy inpatients with myelodysplastic syndromes and transfusionaliron overload,” Leukemia Research, vol. 31, no. 3, pp. S10–S15,2007.

[14] E. P. Alessandrino, S. Amadori, G. Barosi et al., “Evidence-and consensus-based practice guidelines for the therapy ofprimary myelodysplastic syndromes. A statement from theitalian society of hematology,” Haematologica, vol. 87, no. 12,pp. 1286–1306, 2002.

[15] D. Bowen, D. Culligan, S. Jowitt et al., “Guidelines for thediagnosis and therapy of adult myelodysplastic syndromes,”British Journal of Haematology, vol. 120, no. 2, pp. 187–200,2003.


[16] National Comprehensive Cancer Network, “NCCN Clini-cal Practice Guidelines. Myelodysplastic Syndromes,” 2009,http://www.nccn.org/.

[17] T. Suzuki, M. Tomonaga, Y. Miyazaki et al., “Japaneseepidemiological survey with consensus statement on Japaneseguidelines for treatment of iron overload in bone marrowfailure syndromes,” International Journal of Hematology, vol.88, no. 1, pp. 30–35, 2008.

[18] J. M. Bennett, “Consensus statement on iron overload inmyelodysplastic syndromes,” American Journal of Hematology,vol. 83, no. 11, pp. 858–861, 2008.

[19] R. A. Wells, B. Leber, R. Buckstein et al., “Iron overload inmyelodysplastic syndromes: a Canadian consensus guideline,”Leukemia Research, vol. 32, no. 9, pp. 1338–1353, 2008.

[20] P. Valent, O. Krieger, R. Stauder et al., “Iron overload inmyelodysplastic syndromes (MDS)—diagnosis, management,and response criteria: a proposal of the Austrian MDSplatform,” European Journal of Clinical Investigation, vol. 38,no. 3, pp. 143–149, 2008.

[21] M. Mittelman, G. Lugassy, D. Merkel et al., “Iron chelationtherapy in patients with myelodysplastic syndromes: consen-sus conference guidelines,” Israel Medical Association Journal,vol. 10, no. 5, pp. 374–376, 2008.

[22] M. Kruszewski, “The role of labile iron pool in cardiovasculardiseases,” Acta Biochimica Polonica, vol. 51, no. 2, pp. 471–480,2004.

[23] A. I. Schafer, R. G. Cheron, R. Dluhy, et al., “Clinical conse-quences of acquired transfusional iron overload in adults,” TheNew England Journal of Medicine, vol. 304, no. 6, pp. 319–324,1981.

[24] L. M. Buja and W. C. Roberts, “Iron in the heart. Etiology andclinical significance,” The American Journal of Medicine, vol.51, no. 2, pp. 209–221, 1971.

[25] P. Mathew, A. Tefferi, G. W. Dewald et al., “The 5q-syndrome:a single-institution study of 43 consecutive patients,” Blood,vol. 81, no. 4, pp. 1040–1045, 1993.

[26] L. Malcovati, M. G. Della Porta, C. Pascutto et al., “Prognosticfactors and life expectancy in myelodysplastic syndromesclassified according to WHO criteria: a basis for clinicaldecision making,” Journal of Clinical Oncology, vol. 23, no. 30,pp. 7594–7603, 2005.

[27] L. Malcovati, “Impact of transfusion dependency and sec-ondary iron overload on the survival of patients withmyelodysplastic syndromes,” Leukemia Research, vol. 31, sup-plement 3, pp. S2–S6, 2007.

[28] L. Malcovati, U. Germing, A. Kuendgen et al., “Time-dependent prognostic scoring system for predicting survivaland leukemic evolution in myelodysplastic syndromes,” Jour-nal of Clinical Oncology, vol. 25, no. 23, pp. 3503–3510, 2007.

[29] J. Cermak, P. Kacirkova, D. Mikulenkova, and K. Michalova,“Impact of transfusion dependency on survival in patientswith early myelodysplastic syndrome without excess of blasts,”Leukemia Research, vol. 33, no. 11, pp. 1469–1474, 2009.

[30] M. Takatoku, T. Uchiyama, S. Okamoto, et al., “Retrospectivenationwide survey of Japanese patients with transfusionde-pendent MDS and aplastic anemia highlights the negativeimpact of iron overload onMorbiddity/mortality,” EuropeanJournal of Haematology, vol. 78, pp. 487–494, 2007.

[31] K. Naka, T. Muraguchi, T. Hosh II, and A. Hirao, “Regulationof reactive oxygen species and genomic stability in hematopoi-etic stem cells,” Antioxidants and Redox Signaling, vol. 10, no.11, pp. 1883–1894, 2008.

[32] M. Koptyra, R. Falinski, M. O. Nowicki et al., “BCR/ABLkinase induces self-mutagenesis via reactive oxygen species toencode imatinib resistance,” Blood, vol. 108, no. 1, pp. 319–327, 2006.

[33] F. V. Rassool, T. J. Gaymes, N. Omidvar et al., “Reactive oxygenspecies, DNA damage, and error-prone repair: a model forgenomic instability with progression in myeloid leukemia?”Cancer Research, vol. 67, no. 18, pp. 8762–8771, 2007.

[34] C. E. Chee, D. P. Steensma, W. Wu, C. A. Hanson, and A.Tefferi, “Neither serum ferritin nor the number of red bloodcell transfusions affect overall survival in refractory anemiawith ringed sideroblasts,” American Journal of Hematology, vol.83, no. 8, pp. 611–613, 2008.

[35] M. G. Zurlo, P. De Stefano, C. Borgna-Pignatti et al., “Survivaland causes of death in thalassaemia major,” Lancet, vol. 2, no.8653, pp. 27–30, 1989.

[36] L. M. Buja and W. C. Roberts, “Iron in the heart. Etiology andclinical significance,” The American Journal of Medicine, vol.51, no. 2, pp. 209–221, 1971.

[37] E. Konen, H. Ghoti, O. Goitein et al., “No evidence formyocardial iron overload in multitransfused patients withmyelodysplastic syndrome using cardiac magnetic resonanceT2∗ technique,” American Journal of Hematology, vol. 82, no.11, pp. 1013–1016, 2007.

[38] A. A. Di Tucci, G. Matta, S. Deplano et al., “Myocardial ironoverload assessment by T2∗ magnetic resonance imaging inadult transfusion dependent patients with acquired anemias,”Haematologica, vol. 93, no. 9, pp. 1385–1388, 2008.

[39] J. Chacko, D. J. Pennell, M. A. Tanner et al., “Myocardialiron loading by magnetic resonance imaging T2∗ in goodprognostic myelodysplastic syndrome patients on long-termblood transfusions,” British Journal of Haematology, vol. 138,no. 5, pp. 587–593, 2007.

[40] P. D. Jensen, F. T. Jensen, T. Christensen, H. Eiskjaer, U.Baandrup, and J. L. Nielsen, “Evaluation of myocardial ironby magnetic resonance imaging during iron chelation therapywith deferioxamine: indication of close relation betweenmyocardial iron content and chelatable iron pool,” Blood, vol.101, pp. 4632–4639, 2003.

[41] J. J. Bullen, H. J. Rogers, P. B. Spalding, and C. G. Ward,“Natural resistance, iron and infection: a challenge for clinicalmedicine,” Journal of Medical Microbiology, vol. 55, no. 3, pp.251–258, 2006.

[42] D. P. Kontoyiannis, G. Chamilos, R. E. Lewis et al., “Increasedbone marrow iron stores is an independent risk factor forinvasive aspergillosis in patients with high-risk hematologicmalignancies and recipients of allogeneic hematopoietic stemcell transplantation,” Cancer, vol. 110, no. 6, pp. 1303–1306,2007.

[43] A. Altes, A. F. Remacha, P. Sarda et al., “Early clinical impactof iron overload in stem cell transplantation. A prospectivestudy,” Annals of Hematology, vol. 86, no. 6, pp. 443–447, 2007.

[44] J. R. Boelaert, M. De Locht, J. van Cutsem et al., “Mucormy-cosis during deferoxamine therapy is a siderophore-mediatedinfection: in vitro and in vivo animal studies,” Journal ofClinical Investigation, vol. 91, no. 5, pp. 1979–1986, 1993.

[45] A. S. Ibrahim, T. Gebermariam, Y. Fu et al., “The iron chelatordeferasirox protects mice from mucormycosis through ironstarvation,” Journal of Clinical Investigation, vol. 117, no. 9, pp.2649–2657, 2007.

[46] A. Mahindra, R. Sobecks, L. Rybicki, et al., “Elevated pretrans-plant ferritin is associated with a lower incidence of chronic


graft-versus-host disease and inferior survival after myeloabla-tive allogenic hematopoietic stem cell transplantation,” BritishJournal of Haematology, vol. 146, pp. 310–316, 2009.

[47] W. Y. Au, W. M. Lam, W. C. Chu et al., “A magneticresonance imaging study of iron overload in hemopoieticstem cell transplant recipients with increased ferritin levels,”Transplantation Proceedings, vol. 39, no. 10, pp. 3369–3374,2007.

[48] T. de Witte, “The role of iron in patients after bone marrowtransplantation,” Blood Reviews, vol. 22, no. 2, pp. S22–S28,2008.

[49] L. von Bonsdorff, L. Sahlstedt, F. Ebeling, T. Ruutu, and J.Parkkinen, “Apotransferrin administration prevents growth ofStaphylococcus epidermidis in serum of stem cell transplantpatients by binding of free iron,” FEMS Immunology andMedical Microbiology, vol. 37, no. 1, pp. 45–51, 2003.

[50] C. Borgna-Pignatti, S. Rugolotto, P. De Stefano et al., “Survivaland complications in patients with thalassemia major treatedwith transfusion and deferoxamine,” Haematologica, vol. 89,no. 10, pp. 1187–1193, 2004.

[51] V. Gabutti and A. Piga, “Results of long-term iron-chelatingtherapy,” Acta Haematologica, vol. 95, no. 1, pp. 26–36, 1996.

[52] E. J. Neufeld, “Oral chelators deferasirox and deferiprone fortransfusional iron overload in thalassemia major: new data,new questions,” Blood, vol. 107, no. 9, pp. 3436–3441, 2006.

[53] M. D. Cappellini and P. Pattoneri, “Oral iron chelators,”Annual Review of Medicine, vol. 60, pp. 25–38, 2009.

[54] G. J. Kontoghiorghes, K. Pattichi, M. Hadjigavriel, and A.Kolnagou, “Transfusional iron overload and chelation therapywith deferoxamine and deferiprone (L1),” Transfusion Science,vol. 23, no. 3, pp. 211–223, 2000.

[55] M. D. Cappellini, “Long-term efficacy and safety ofdeferasirox,” Blood Reviews, vol. 22, no. 2, pp. S35–S41,2008.

[56] M. D. Cappelini, A. Choen, A. Piga, et al., “A phase 3 studyof deferasirox (ICL670), a once-daily oral chelator, in patientswith-thalassemia,” Blood, vol. 107, pp. 3455–3462, 2006.

[57] A. Piga, R. Galanello, G. L. Forni et al., “Randomized phase IItrial of deferasirox (Exjade�, ICL670), a once-daily, orally-administered iron chelator, in comparison to deferoxaminein thalassemia patients with transfusional iron overload,”Haematologica, vol. 91, no. 7, pp. 873–880, 2006.

[58] E. Vichinsky, O. Onyekwere, J. Porter et al., “A randomisedcomparison of deferasirox versus deferoxamine for the treat-ment of transfusional iron overload in sickle cell disease,”British Journal of Haematology, vol. 136, no. 3, pp. 501–508,2007.

[59] J. Porter, R. Galanello, G. Saglio et al., “Relative responseof patients with myelodysplastic syndromes and othertransfusion-dependent anaemias to deferasirox (ICL670): a 1-yr prospective study,” European Journal of Haematology, vol.80, no. 2, pp. 168–176, 2008.

[60] N. Gattermann, M. Schmid, M. Della Porta, et al., “Efficacyand safety of deferasirox (Exjade) during 1 year of treatmentin transfusion-dependent patients with myelodysplastic syn-dromes: results from EPIC trial,” Blood, vol. 112, article 633a,2008.

[61] H. A. Leitch and L. M. Vickars, “Supportive care and chelationtherapy in MDS: are we saving lives or just lowering iron?”Hematology, pp. 664–672, 2009.

[62] P. D. Jensen, I. M. Jensen, and J. Ellegaard, “Desferrioxaminetreatment reduces blood transfusion requirements in patients

with myelodysplastic syndrome,” British Journal of Haematol-ogy, vol. 80, no. 1, pp. 121–124, 1992.

[63] P. D. Jensen, L. Heickendorff, B. Pedersen et al., “The effectof iron chelation on haemopoiesis in MDS patients withtransfusional iron overload,” British Journal of Haematology,vol. 94, no. 2, pp. 288–299, 1996.

[64] M. E. P. Smeets, S. Vreugdenhil, and R. S. G. Holdrinet,“Improvement of erythropoiesis during treatment withdeferiprone in a patient with myelofibrosis and transfusionalhemosiderosis,” American Journal of Hematology, vol. 51, no.3, pp. 243–244, 1996.

[65] A. A. Di Tucci, R. Murru, D. Alberti, B. Rabault, S. Deplano,and E. Angelucci, “Correction of anemia in a transfusion-dependent patient with primary myelofibrosis receiving ironchelation therapy with deferasirox (Exjade�, ICL670),” Euro-pean Journal of Haematology, vol. 78, no. 6, pp. 540–542, 2007.

[66] E. Messa, D. Cilloni, F. Messa, F. Arruga, A. Roetto, and G.Saglio, “Deferasirox treatment improved the hemoglobin leveland decreased transfusion requirements in four patients withthe myelodysplastic syndrome and primary myelofibrosis,”Acta Haematologica, vol. 120, no. 2, pp. 70–74, 2008.

[67] S. Capalbo, G. Spinosa, M. G. Franzese, and G. Palumbo,“Early deferasirox treatment in a patient with myelodysplasticsyndrome results in a long-term reduction in transfusionrequirements,” Acta Haematologica, vol. 121, no. 1, pp. 19–20,2009.

[68] A. F. List, M. R. Baer, D. Steensma, et al., “Iron chelation withdeferasirox (Exjade) improves iron burden in patients withmyelodysplastic syndromes (MDS),” Blood, vol. 112, article634a, 2008.

[69] C. Camaschella, A. Campanella, L. De Falco et al., “Thehuman counterpart of zebrafish shiraz shows sideroblastic-likemicrocytic anemia and iron overload,” Blood, vol. 110, no. 4,pp. 1353–1358, 2007.

[70] J. Hartmann, U. Sinzig, G. Wulf, et al., “Evidence for asuppression of the colony forming capacity of erythroidprogenitors by iron overload in patients with MDS,” Blood,vol. 112, article 2694a, 2008.

[71] E. Messa, S. Carturan, C. Maffe, et al., “Deferasirox is apowerful NF-κB inhibitor in myelodysplastic cells and inleukemia cell lines acting independently from cell iron depri-vation by chelation and reactive oxygen species scavenging,”Haematologica. In press.

[72] D. P. Steensma, “Myelodysplasia paranoia: iron as the newradon,” Leukemia Research, vol. 33, no. 9, pp. 1158–1163, 2009.

[73] A. Tefferi and R. M. Stone, “Iron chelation therapy inmyelodysplastic syndrome—Cui bono?” Leukemia, vol. 23, no.8, p. 1373, 2009.


Case Report

Red Blood Cell Transfusion IndependenceFollowing the Initiation of Iron Chelation Therapy inMyelodysplastic Syndrome

Maha A. Badawi,1 Linda M. Vickars,2 Jocelyn M. Chase,1 and Heather A. Leitch2

1 Department of Medicine, St. Paul’s Hospital, The University of British Columbia, Vancouver, BC, Canada V6T1Z42 Department of Hematology, St. Paul’s Hospital, The University of British Columbia, Vancouver, BC, Canada V6Z2A5

Correspondence should be addressed to Heather A. Leitch, [email protected]

Received 2 November 2009; Revised 11 January 2010; Accepted 18 January 2010


Copyright © 2010 Maha A. Badawi et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Iron chelation therapy is often used to treat iron overload in patients requiring transfusion of red blood cells (RBC). A 76-year-old man with MDS type refractory cytopenia with multilineage dysplasia, intermediate-1 IPSS risk, was referred when hebecame transfusion dependent. He declined infusional chelation but subsequently accepted oral therapy. Following the initiationof chelation, RBC transfusion requirement ceased and he remained transfusion independent over 40 months later. Over the sametime course, ferritin levels decreased but did not normalize. There have been eighteen other MDS patients reported showingimprovement in hemoglobin level with iron chelation; nine became transfusion independent, nine had decreased transfusionrequirements, and some showed improved trilineage myelopoiesis. The clinical features of these patients are summarized andpossible mechanisms for such an effect of iron chelation on cytopenias are discussed.

1. Introduction

The myelodysplastic syndromes (MDS) are characterized byineffective hematopoiesis, cytopenias, and a risk of transfor-mation to acute myeloid leukemia (AML); survival and AMLrisk are predicted by the International Prognostic ScoringSystem (IPSS) [1]. Because the median age of the MDS onsetis in the seventh decade, most patients are ineligible forpotentially curative hematopoietic stem cell transplantation[2]. Although other treatments are now available [3–7],the standard treatment for many MDS patients remainssupportive care.

Most MDS patients eventually become red blood cell(RBC) transfusion dependent, risking iron overload [8],which may lead to cardiac, hepatic, and endocrine dysfunc-tion. Recent studies suggest an adverse effect of RBC trans-fusion dependence on survival, predominantly in lower-riskMDS [9]. This effect was sufficiently significant that RBCtransfusion dependence was incorporated into the World

Health Organization Prognostic Scoring System (WPSS) forMDS [10].

While the benefits of iron chelation therapy are betterestablished in thalassemia [11], recent retrospective studiesin lower-risk MDS suggest a possible improvement insurvival in transfusion dependent patients who receivedchelation [12]. Guidelines in MDS recommend chelationwith an otherwise reasonable life expectancy and evidenceof iron overload: elevated serum ferritin, iron related organdysfunction, or chronic RBC transfusions [13, 14]. Wepresent the clinical course of a RBC transfusion dependentMDS patient who became transfusion independent shortlyafter starting chelation and has remained transfusion inde-pendent for over three years. We review the literature onthe abrogation of cytopenias in acquired anemias followingchelation.

This paper was prepared in accordance with the require-ments of the St. Paul’s Hospital Institutional Research EthicsBoard.


2. Case Report

A 76-year-old man was referred in June 2004. He wasdiagnosed with MDS in 1997 during a work-up of abnormalblood counts: white blood cells (WBC) 2.4 (normal 4.0–11.0)× 109/L, neutrophils 0.7 (2.0–8.0) × 109/L, hemoglobin(Hb) 133 (135–180) G/L, and platelets 108 (150–400)× 109/L. The following laboratory parameters were nor-mal: creatinine, bilirubin, thyroid stimulating hormone,reticulocyte count, serum B12 level, red blood cell folate;and serum protein electropheresis. Bone marrow aspirationand biopsy showed refractory anemia (RA) by the French-American-British (FAB) classification [15] and cytogeneticanalysis revealed trisomy 8 and loss of chromosome Y.Stem cell culture showed no erythropoietin independentcolony growth, serum erythropoietin level was 148.3 (normal3.3–16.6) IU/mL and IPSS score was intermediate-1. Heremained transfusion independent until one month prior toreferral, when the hemoglobin was 60 G/L, prompting theinitiation of RBC transfusion support.

History and physical examination were otherwise unre-markable. WBC count at referral was 3.4 × 109/L, Hb(transfused) 86 G/L, mean cellular volume (MCV) 121 fl(80–100), and platelets 44 × 109/L. Serum ferritin was 1293(15–370) ug/L with no prior ferritin levels available. Bonemarrow aspiration and biopsy confirmed RA/refractorycytopenia with multilineage dysplasia (RCMD) by WorldHealth Organization (WHO) criteria [16]. Marrow blastcount was 2%.

Over a 30-month period, he required transfusion of3 RBC units every 4 weeks to maintain the hemoglobinabove 90 G/L and he complained of fatigue and functionallimitation; he received approximately 90 RBC units intotal. In January 2005, the ferritin was 2197 ug/L buthe declined deferoxamine; however, in September 2006,he agreed to start deferasirox. Bone marrow aspirationand biopsy showed unchanged RCMD and karyotype.Deferasirox was started at 20 mg/kg/day. He required severaldose interruptions and adjustments for renal insufficiency(peak creatinine 141 umol/L, normal to 100 umol/L) and thedose of deferasirox was titrated between 5–30 mg/kg/day. Hereceived no other treatment for anemia.

Two months after starting chelation, the hemoglobinincreased to 109 G/L and he has not required transfusionsince. Mean hemoglobin over 24 months was 122 (range96–144) G/L. Hemoglobin and ferritin levels are shownin Figure 1. The patient reports excellent energy and asignificantly improved quality of life.

In May 2008, he was assessed for skin nodules andreported having similar nodules that appeared and regressedspontaneously for at least two years. A biopsy revealedleukemia cutis (LC). Despite this, he remained clinicallywell and transfusion independent for 17 months since thediagnosis of LC, over 41 months since the initial appearanceof nodules, and 40 months since the initiation of chelation.

Characteristics of ten MDS patients, including ours,achieving transfusion independence with chelation are sum-marized in Table 1 [17–19]. Nine other patients with signifi-cant improvement in hemoglobin with chelation have been

0

1000

2000

3000

4000

5000

6000

0

20

40

60

80

100

120

140

160

May

.04

Sep.

04

Jan

.05

May

.05

Sep.

05

Jan

.06

May

.06

Sep.

06

Jan

.07

May

.07

Sep.

07

Jan

.08

May

.08

Sep.

08

FerritinHgb

Figure 1: Hemoglobin and serum ferritin levels for a patientwith MDS receiving iron chelation therapy. The solid black arrowrepresents the date at which chelation was initiated and the dashedarrow represents the date of his last red blood cell transfusion. Thegrey bar indicates the period during which transfusion requirementwas 3 red blood cell units every 4 weeks.

reported [19, 20]. Several features of these latter patientswere not reported; however, eight received deferoxamine andone deferasirox, and the median time to improvement inRBC transfusion requirement was 14.4 (3–24) months. Noneof these patients were reported to have received any MDStreatment other than chelation.

3. Discussion

It is well established that chelation extends the survival oftransfusion dependent patients with thalassemia by mitigat-ing iron toxicity [21–24]. Recent retrospective data suggesta possible association between chelation and improvedsurvival in MDS [12, 25]. The first report of decreasedtransfusion requirements with chelation was in 1990 [26].Since then, nineteen MDS patients, including ours, arereported who had an improvement in hemoglobin ordecreased transfusion requierements.

Our patient was transfusion independent within twomonths of starting chelation. The ferritin level decreasedfrom 5271 to 1225 ug/L but remains elevated. Once trans-ferrin is saturated, non-transferrin bound iron (NTBI) maybe detected [27], correlating with the presence of potentiallycytotoxic reactive oxygen species (ROS) [28]. Whetheroxidative stress was present in our patient is unknown asfew transferrin saturations were recorded and NTBI andROS measurement are not readily available. However, theelevated ferritin over a long course despite chelation whiletransfusion independent may indicate a significant iron load,


Table 1: Clinical characteristics of 10 MDS patients achieving red blood cell transfusion independence with iron chelation therapy.

Clinical Feature (units) n

Age at MDS diagnosis (years) Median 58 (range 18–74)

Gender M : F 5 : 5

MDS subtype (FAB or WHO)

RA 5

RARS 2

RCMD 1

RAEB 2

IPSS score

Low 1

Intermediate-1 5

Intermediate-1 or 2 1

High 1

Not available 2

Iron chelation agent

Deferoxamine 7

Deferasirox 3

Time to RBC transfusion independence (months) Median 17.5 (range 1–24)

Duration of RBC transfusion independence (months) Median 13 (range 3–40)

Abbreviations: F: female; FAB: French-American-British; IPSS: International Prognostic ScoringSystem; M: male; n: number; RA: refractory anemia; RARS: refractory anemia with ringed sideroblasts,RAEB: refractory anemia with excess blasts; RBC: red blood cell; RCMD: refractory cytopenia with multilineage dysplasia; and WHO: World HealthOrganization.

potentially leading to marrow toxicity and suppression ofhematopoiesis.

A patient with primary myelofibrosis (PMF) wasreported whose hemoglobin increased from 76 ± 10 G/L to100 G/L after starting chelation [19]; it returned to baseline(80 G/L) when chelation was interrupted, and increasedagain to 100 G/L when chelation resumed. A second PMFpatient with baseline hemoglobin 60 G/L requiring 2 RBCunits every two weeks achieved long-term transfusion inde-pendence one month after beginning deferiprone [29]. Athird PMF patient with baseline hemoglobin 50–60 G/Lrequiring 2-3 RBC units per month became transfusionindependent two months after starting deferasirox, an effectwhich persisted two years after chelation was stoppedfor improvement in ferritin (953 ug/L) and transferrinsaturation (45%) [30]. These patients received no othertreatment for PMF. A patient with aplastic anemia (initial Hb45 G/L, neutrophil count 0.3× 109/L, and platelet count 3×109/L) had trilineage recovery and became RBC transfusionindependent after four years of deferoxamine [31]; thispatient received low-dose erythropoietin following an initialimprovement in blood counts.

An improvement in MCV, platelet and white blood cellcounts was also noted [18, 20]. In a report of six patients,two with pancytopenia had significant increases in WBC,neutrophil, and platelet counts (P ≤ .001) [20], seen within 3months, maximized by 18 months, and in some patients, theeffect persisted after chelation was discontinued. All of themhad an elevated MCV prior to chelation, which decreased infive and normalized in two, suggesting possible improvementin erythropoiesis outside the MDS clone. In a report of

eleven patients, the neutrophil count increased in eight ofnine, and the platelet count in seven of eleven [18]. In ourpatient, recent WBC counts range between 3.1–4.3 × 109/Land platelets consistently clump; the MCV is unchanged at120 fl.

The mechanisms by which chelation may improvecytopenias are unclear; however, iron was recently shown tohave a suppressive effect on erythroid progenitors in vitro[32]. Erythroid colony assays on 42 MDS patients showed,in patients with ferritin 250 ug/L or more, that BFU-E werea mean of 2.35 (range 0–27) colonies per culture, comparedto 10.1 (0–76) in patients with normal ferritin (P < .004);whether this is an effect of iron or due to other factors awaitsfurther study.

Although chelation may exert its protective effect byreversing the deposition of iron [23, 33], oxidative stress fromiron overload may damage lipids, proteins, and nucleic acids[27, 28, 34–37], and it would be interesting to determinewhether the protective effect of chelation on BFU-E might befrom oxidative stress alleviation. A study of 15 patients withlower-risk MDS showed a decrease in RBC ROS followingthree months of chelation [38] and a relationship betweenferritin and ROS content of CD34+ cells in MDS patients wasestablished [39]. In thalassemia, chelation reduced oxidationin RBC and increased half-life from 12.1 ± 2.4 to 16.4 ± 4.3days [40]. In the US03 trial of deferasirox in MDS, hema-tologic improvement was seen in 5 of 53 patients (9.4%)[41] and LPI, an indicator of oxidative stress, normalizedover 12 months of chelation; whether this accounts for themitigation of cytopenias remains to be determined. Finally,there are reports of increased erythropoietin levels with


chelation in normal volunteers and this could contribute toan improvement in hemoglobin in MDS [3, 42].

It has been suggested that the transcription factor NF-κBmay be important in modifying myelopoiesis with chelation.In mononuclear cells of MDS patients [43], deferasiroxinduced a significant reduction in NF-κB activity, but theopposite effect was seen with deferoxamine and deferiproneand no difference was noted in patients with or without ironoverload. Although these findings might explain an effectof deferasirox on cytopenias, the effect of deferoxamine anddeferiprone is not accounted for [18].

Sloand et al. showed improvement in erythropoiesiswithin the MDS clone in patients with trisomy 8 respondingto immunosuppressive therapy [44]. While our patient has+8, no therapy other than chelation was administered;however, the MCV remains elevated, possibly indicating asignificant contribution to erythropoiesis by the MDS clone.In the Jensen study, two of eleven patients had +8; in the first,the +8 clone decreased from 60% to 10% with chelation, andthe second had persistence of +8 and clonal evolution to adeletion of 5q as well. Thus, immunomodulation resultingin improvement of erythropoiesis cannot be invoked as apredominant mechanism for transfusion independence inthese patients.

4. Conclusions

In summary, a number of patients with acquired anemiashave been reported in whom an improvement in cytopeniaswas seen following the initiation of iron chelation therapy,clinically manifested as a decrease in RBC transfusionrequirements or even transfusion independence. This mayoccur in up to 9% of MDS [41] and possible mechanismsinclude reducing oxidative stress, altering intracellular levelsof NF-κB; increasing erythropoietin levels, or other mech-anisms yet to be elucidated. In future trials of chelation,consideration could be given to including measures ofthese parameters, and conversely, trials of medicationsknown to induce transfusion independence in MDS suchas immunomodulatory, demethylating, or erythropoiesisstimulating agents could compare these in responders andnonresponders. Patients with iron overload considered forchelation should be assessed and monitored by a physicianexperienced with chelation medications.

Conflict of Interest

HL and LV have received honoraria and research fundingfrom Novartis Canada. All data collection and manuscriptpreparation were performed independent of financial sup-port. MB and JC have no conflict of interest to disclose.

References

[1] P. Greenberg, C. Cox, M. M. LeBeau, et al., “Internationalscoring system for evaluating prognosis in myelodysplasticsyndromes,” Blood, vol. 89, no. 6, pp. 2079–2088, 1997.

[2] C. A. Schiffer, “Clinical issues in the management of patientswith myelodysplasia,” Hematology, pp. 205–210, 2006.

[3] E. Hellstrom-Lindberg, N. Gulbrandsen, G. Lindberg, et al.,“A validated decision model for treating the anaemia ofmyelodysplastic syndromes with erythropoietin + granulocytecolony-stimulating factor: significant effects on quality of life,”British Journal of Haematology, vol. 120, no. 6, pp. 1037–1046,2003.

[4] H. Kantarjlan, J.-P. J. Issa, C. S. Rosenfeld, et al., “Decitabineimproves patient outcomes in myelodysplastic syndromes:results of a phase III randomized study,” Cancer, vol. 106, no.8, pp. 1794–1803, 2006.

[5] A. List, S. Kurtin, D. J. Roe, et al., “Efficacy of lenalidomide inmyelodysplastic syndromes,” New England Journal of Medicine,vol. 352, no. 6, pp. 549–557, 2005.

[6] A. Raza, P. Meyer, D. Dutt, et al., “Thalidomide producestransfusion independence in long-standing refractory anemiasof patients with myelodysplastic syndromes,” Blood, vol. 98,no. 4, pp. 958–965, 2001.

[7] L. R. Silverman, E. P. Demakos, B. L. Peterson, et al.,“Randomized controlled trial of azacitidine in patients withthe myelodysplastic syndrome: a study of the cancer andleukemia group B,” Journal of Clinical Oncology, vol. 20, no.10, pp. 2429–2440, 2002.

[8] J. B. Porter, “Practical management of iron overload,” BritishJournal of Haematology, vol. 115, no. 2, pp. 239–252, 2001.

[9] L. Malcovati, M. G. Della Porta, C. Pascutto, et al., “Prognosticfactors and life expectancy in myelodysplastic syndromesclassified according to WHO criteria: a basis for clinicaldecision making,” Journal of Clinical Oncology, vol. 23, no. 30,pp. 7594–7603, 2005.

[10] L. Malcovati, U. Germing, A. Kuendgen, et al., “Time-dependent prognostic scoring system for predicting survivaland leukemic evolution in myelodysplastic syndromes,” Jour-nal of Clinical Oncology, vol. 25, no. 23, pp. 3503–3510, 2007.

[11] N. F. Olivieri, D. G. Nathan, J. H. Macmillan, et al., “Survivalin medically treated patients with homozygous β-thalassemia,”New England Journal of Medicine, vol. 331, no. 9, pp. 574–578,1994.

[12] H. A. Leitch, C. S. Leger, T. A. Goodman, et al., “Improvedsurvival in patients with myelodysplastic syndrome receivingiron chelation therapy,” Clinical Leukemia, vol. 2, no. 3, pp.205–211, 2008.

[13] N. Gattermann, “Overview of guidelines on iron chelationtherapy in patients with myelodysplastic syndromes and trans-fusional iron overload,” International Journal of Hematology,vol. 88, no. 1, pp. 24–29, 2008.

[14] J. M. Bennett, “Consensus statement on iron overload inmyelodysplastic syndromes,” American Journal of Hematology,vol. 83, no. 11, pp. 858–861, 2008.

[15] J. M. Bennett, D. Catovsky, M. T. Daniel, et al., “Proposals forthe classification of the myelodysplastic syndromes,” BritishJournal of Haematology, vol. 51, no. 2, pp. 189–199, 1982.

[16] J. W. Vardiman, N. L. Harris, and R. D. Brunning, “The WorldHealth Organization (WHO) classification of the myeloidneoplasms,” Blood, vol. 100, no. 7, pp. 2292–2302, 2002.

[17] J. Del Rıo Garma, L. C. Fernandez, and B. F. J. Fonrodona,“Desferrioxamine in the treatment of myelodysplastic syn-dromes,” Haematologica, vol. 82, no. 5, pp. 639–640, 1997.

[18] P. D. Jensen, L. Heickendorff, B. Pedersen, et al., “The effectof iron chelation on haemopoiesis in MDS patients withtransfusional iron overload,” British Journal of Haematology,vol. 94, no. 2, pp. 288–299, 1996.

[19] E. Messa, D. Cilloni, F. Messa, F. Arruga, A. Roetto, and G.Saglio, “Deferasirox treatment improved the hemoglobin leveland decreased transfusion requirements in four patients with


the myelodysplastic syndrome and primary myelofibrosis,”Acta Haematologica, vol. 120, no. 2, pp. 70–74, 2008.

[20] P. D. Jensen, I. M. Jensen, and J. Ellegaard, “Desferrioxaminetreatment reduces blood transfusion requirements in patientswith myelodysplastic syndrome,” British Journal of Haematol-ogy, vol. 80, no. 1, pp. 121–124, 1992.

[21] E. Angelucci, B. Turlin, D. Canatan, et al., “Iron chelationtherapy with deferasirox (Exjade, ICL670) or deferoxaimine iseffective in reducing iron overload in patients with advancedfibrosis and cirrhosis,” Blood, vol. 106, no. 11, p. 757a, 2005.

[22] G. M. Brittenham, P. M. Griffith, A. W. Nienhuis, et al.,“Efficacy of deferoxamine in preventing complications of ironoverload in patients with thalassemia major,” New EnglandJournal of Medicine, vol. 331, no. 9, pp. 567–573, 1994.

[23] B. A. Davis and J. B. Porter, “Long-term outcome of continu-ous 24-hour deferoxamine infusion via indwelling intravenouscatheters in high-risk β-thalassemia,” Blood, vol. 95, no. 4, pp.1229–1236, 2000.

[24] N. F. Olivieri and G. M. Brittenham, “Iron-chelating therapyand the treatment of thalassemia,” Blood, vol. 89, no. 3, pp.739–761, 1997.

[25] C. Rose, S. Brechignac, D. Vassilief, et al., “Positive impact ofiron chelation therapy (CT) on survival in regularly transfusedMDS patients. A prospective analysis by the GFM,” Blood, vol.110, no. 11, pp. 80–81a, 2007.

[26] J. H. Marsh, M. Hundert, and P. Schulman, “Deferoxamine-induced restoration of haematopoiesis in myelofibrosis sec-ondary to myelodysplasia,” British Journal of Haematology, vol.76, no. 1, pp. 148–149, 1990.

[27] W. Breuer, M. Shvartsman, and Z. I. Cabantchik, “Intracellularlabile iron,” International Journal of Biochemistry and CellBiology, vol. 40, no. 3, pp. 350–354, 2008.

[28] R. S. Britton, K. L. Leicester, and B. R. Bacon, “Iron toxicityand chelation therapy,” International Journal of Hematology,vol. 76, no. 3, pp. 219–228, 2002.

[29] M. E. Smeets, G. Vreugdenhil, and R. S. Holdrinet, “Improve-ment of erythropoiesis during treatment with deferiprone ina patient with myelofibrosis and transfusional hemosiderosis,”American Journal of Hematology, vol. 51, no. 3, pp. 243–244,1996.

[30] A. A. Di Tucci, R. Murru, D. Alberti, B. Rabault, S. Deplano,and E. Angelucci, “Correction of anemia in a transfusion-dependent patient with primary myelofibrosis receiving ironchelation therapy with deferasirox (Exjade, ICL670),” Euro-pean Journal of Haematology, vol. 78, no. 6, pp. 540–542, 2007.

[31] S.-J. Park and C.-W. Han, “Complete hematopoietic recoveryafter continuous iron chelation therapy in a patient with severeaplastic anemia with secondary hemochromatosis,” Journal ofKorean Medical Science, vol. 23, no. 2, pp. 320–323, 2008.

[32] J. Hartmann, U. Sinzig, G. Wulf, et al., “Evidence for asuppression of the colony forming capacity of erythroidprogenitors by iron overload in patients with MDS,” Blood,vol. 11, p. 932a, 2008.

[33] L. J. Anderson, M. A. Westwood, S. Holden, et al., “Myocardialiron clearance during reversal of siderotic cardiomyopathywith intravenous desferrioxamine: a prospective study usingT2∗ cardiovascular magnetic resonance,” British Journal ofHaematology, vol. 127, no. 3, pp. 348–355, 2004.

[34] G. J. Anderson, “Mechanisms of iron loading and toxicity,”American Journal of Hematology, vol. 82, no. 12, supplement,pp. 1128–1131, 2007.

[35] J. Amer and E. Fibach, “Oxidative status of platelets in normaland thalassemic blood,” Thrombosis and Haemostasis, vol. 92,no. 5, pp. 1052–1059, 2004.

[36] J. Amer, A. Goldfarb, and E. Fibach, “Flow cytometric analysisof the oxidative status of normal and thalassemic red bloodcells,” Cytometry A, vol. 60, no. 1, pp. 73–80, 2004.

[37] B. P. Esposito, W. Breuer, P. Sirankapracha, P. Pootrakul, C.Hershko, and Z. I. Cabantchik, “Labile plasma iron in ironoverload: redox activity and susceptibility to chelation,” Blood,vol. 102, no. 7, pp. 2670–2677, 2003.

[38] E. Rachmilewitz, D. Merkel, H. Ghoti, et al., “Improvementof oxidative stress parameters in MDS patients with ironoverload treated with deferasirox,” Blood, vol. 112, no. 11, pp.A924–A925, 2008.

[39] L. S. A. Chan, R. Buckstein, M. D. Reis, et al., “Iron overloadand haematopoiesis in MDS: does blood transfusion promoteprogression to AML?” Blood, vol. 112, no. 11, p. 928a, 2008.

[40] N. Szuber, J. L. Buss, S. Soe-Lin, et al., “Alternative treatmentparadigm for thalassemia using iron chelators,” ExperimentalHematology, vol. 36, no. 7, pp. 773–785, 2008.

[41] A. F. List, M. R. Baer, D. Steensma, et al., “Deferasirox(ICL670); Exjade) reduces serum ferritin (SF) and labileplasma iron (LPI) in patients with myelodysplastic syndromes(MDS),” Blood, vol. 112, no. 11, p. 523a, 2008.

[42] X. Ren, K. L. Dorrington, P. H. Maxwell, and P. A. Robbins,“Effects of desferrioxamine on serum erythropoietin andventilatory sensitivity to hypoxia in humans,” Journal ofApplied Physiology, vol. 89, no. 2, pp. 680–686, 2000.

[43] E. Messa, I. Defilippi, A. Roetto, et al., “Deferasirox is theonly iron chelator acting as a potent NFKB inhibitor inmyelodysplastic syndromes,” Blood, vol. 112, no. 11, p. 923a,2008.

[44] E. M. Sloand, L. Mainwaring, M. Fuhrer, et al., “Preferentialsuppression of trisomy 8 compared with normal hematopoi-etic cell growth by autologous lymphocytes in patients withtrisomy 8 myelodysplastic syndrome,” Blood, vol. 106, no. 3,pp. 841–851, 2005.


Review Article

Iron Overload in Sickle Cell Disease

Radha Raghupathy,1 Deepa Manwani,2 and Jane A. Little1

1 Department of Hematology, Albert Einstein College of Medicine and Montefiore Medical Center, Bronx, NY 10461, USA2 Division of Pediatric Hematology/Oncology, Department of Pediatrics, Albert Einstein College of Medicine and Children’sHospital at Montefiore, Bronx, NY 10461, USA

Correspondence should be addressed to Jane A. Little, [email protected]

Received 25 November 2009; Accepted 16 February 2010


Copyright © 2010 Radha Raghupathy et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

In sickle cell disease transfusions improve blood flow by reducing the proportion of red cells capable of forming sickle hemoglobinpolymer. This limits hemolysis and the endothelial damage that result from high proportions of sickle polymer-containing redcells. Additionally, transfusions are used to increase blood oxygen carrying capacity in sickle cell patients with severe chronicanemia or with severe anemic episodes. Transfusion is well-defined as prophylaxis (stroke) and as therapy (acute chest syndromeand stroke) for major complications of sickle cell disease and has been instituted, based on less conclusive data, for a range ofadditional complications, such as priapism, vaso-occlusive crises, leg ulcers, pulmonary hypertension, and during complicatedpregnancies. The major and unavoidable complication of transfusions in sickle cell disease is iron overload. This paper provides anoverview of normal iron metabolism, iron overload in transfused patients with sickle cell disease, patterns of end organ damage,diagnosis, treatment, and prevention of iron overload.

1. Introduction

The human body has no effective physiological mechanismfor excreting excess iron. Therefore, in conditions such assickle cell disease (SCD), where transfusions are frequentlyindicated, exogenous iron can accumulate, circulate as non-transferrin bound iron (NTBI), enter tissues, form reactiveoxygen species (ROS), and result in end organ damage. How-ever, patients with SCD, compared with thalassemic patients,despite a similar transfusion load, may be relatively protectedfrom iron mediated cardiac and endocrine gland toxicity[1]. In thalassemia, ineffective erythropoiesis contributes toiron overload directly and by regulating other downstreampathways. The unfolding pathophysiology of transfusionaliron toxicity in SCD, less well studied than in thalassemia,will be discussed.

2. Normal Iron Metabolism

Iron homeostasis in humans is maintained by the strictregulation of absorption based on body needs. 1 mg (10%of total dietary iron) is absorbed daily, predominantly inthe duodenum, and an equal amount is lost through feces,

urine, and sweat [2]. In normal physiological conditions,iron deficiency and anemia increase iron absorption, whileiron overload decreases it [3].

Nonheme iron absorption is relatively well characterized.Ferric (Fe3+) is reduced to ferrous (Fe2+) iron in the duodenalenterocyte by a ferric reductase (DcytB). Fe2+ is transportedinto the cell by the Divalent Metal Transporter (DMT-1),located at the apical brush border. In the absence of ironoverload, some absorbed iron is stored in the enterocyteas ferritin and the rest is transported across the basolateralmembrane by ferroportin, with the aid of the ferroxidasehephaestin. In the circulation, iron is bound to transferrinand transported to the liver and bone marrow. In the liver,transferrin receptors 1 and 2 mediate the endocytosis of iron,which is then stored as ferritin and released by a ferroportin-mediated mechanism when bodily needs increase. In theerythroid precursors, transferrin bound iron is taken up bytransferrin receptor 1 and utilized for erythropoiesis. Duringred cell senescence, iron is released into macrophages in thereticuloendothelial system (RES) and is stored as ferritin andhemosiderin; again, egress of iron from the macrophage isferroportin dependent [4](Figure 1).


Dietaryiron

Intestinallumen

Enterocytes

DMT1Dcytb

Hephaestin

Ferroportin

Hepcidin

Hemojuvelin

Hepcidin

Hepatocyte

Macrophage

Erythrocyte

Ferroportin

HepcidinTransferrin-bound ironTransferrin

Transferrinreceptor2

Transferrinreceptor1

Ferroportin

HFE

Figure 1: Iron absorption and transport [4]. Reproduced with permission from MMS, and author. Copyright c© (2005) MassachusettsMedical Society. All rights reserved.

The presence of ferroportin on the cell membrane isregulated by hepcidin, a 25-amino acid peptide synthesizedby the liver that is the principal hormone involved inregulating iron absorption [5]. Hepcidin acts by bindingto the ferroportin transporter, triggering its internalizationand degradation, thereby diminishing net circulating ironby reducing iron absorption in the gut and increasing ironsequestration in the RES. Anemia, hypoxia, and erythro-poiesis decrease hepcidin gene expression, thereby stabilizingferroportin and increasing circulating iron available for ery-thropoiesis [6]. In contrast, acute and chronic inflammationincrease hepcidin expression and ferroportin degradation[7]. The paradoxical iron restriction seen in the anemiaof chronic inflammation is associated with increased RESiron and results from a high-hepcidin state. Hemojuvelin,also expressed in the liver, is believed to positively regulatehepcidin production [8]. Matriptase 2, a recently identifiedserine protease, appears to be a sensor of iron deficiency andinhibitor of hepcidin [9]. These peptides help regulate ironabsorption and maintain homeostasis.

Heme iron absorption is less clearly characterized. HemeCarrier Protein 1 (HCP-1), believed to facilitate hemeiron uptake, has been recently identified on the duodenalenterocyte brush border [10]. Heme iron taken up by thistransporter is broken down by a heme oxygenase in theenterocyte into iron and protoporphyrin [11]. It is unclearwhether heme is completely degraded into iron in the

enterocyte and absorbed via ferroportin, or if intact heme isalso transported via other mechanisms. The Feline LeukemiaVirus subgroup C Receptor (FLVCR) appears to play sucha role in transporting heme from erythroid precursors[12, 13].

3. Indications for Transfusion in SCD

Transfusion is a frequently employed therapy in SCD, butits best-validated uses have been in preoperative prophylaxis,treatment of acute chest syndrome (ACS) and prophylaxis,and treatment of stroke [14–16].

Transfusions first demonstrated their effectiveness inreducing recurrent strokes in SCD [14, 17]. Subsequently,transfusions have also proved to be effective prophylaxisagainst first stroke in high risk patients. The Stroke Pre-vention Trial in SCD (STOP) randomized 130 high-riskchildren with SCD to either transfusion therapy (to maintainHbS <30%) or observation [18]. These high risk childrenhad an increased blood flow in the Internal Carotid orMiddle Cerebral Artery by Transcranial Doppler (TCD).This study showed a 92% reduction in incidence of firststroke in transfused high-risk patients. A follow-up study,STOP2, randomized 72 children whose TCD had normalizedafter 30 months of transfusion therapy to either ongoing ordiscontinued transfusions. The study was halted prematurelywhen a significant increase in abnormal TCD velocity and


stroke risk was seen in the group in which transfusiontherapy had been halted [19]. The optimal duration oftransfusion in SCD patients at high risk for primary orrecurrent stroke is still undefined, resulting in long-termtransfusion management of a young population.

In addition to preventing strokes, transfusion is ben-eficial in other complications of SCD such as ACS [20,21]. Vichinsky et al. showed that transfusion improvesoxygenation in ACS [22]. Subset analysis from the STOPtrial also showed a significant reduction in the frequencyof ACS and painful crises in the transfused group [23]. Inaddition, preemptive transfusion therapy was effective inpreventing ACS in a small number of SCD patients withelevated plasma secretory phospholipase A2 levels, an earlysign of ACS [24]. In a study of pregnant women (n = 72)with SCD randomized to receive prophylactic transfusionsor transfusions for medical and obstetric emergencies only,a significant reduction in pain crises was observed in thearm that received prophylactic transfusions [25]. Finally,preoperative simple transfusion to maintain a hemoglobinof 10 gm/dL reduces peri- and postoperative complicationsin patients with SCD [26–28].

4. Mechanism of Transfusion MediatedIron Overload

While transfusion may improve disease complications, ironoverload is a dreaded and inevitable consequence of ongoingtransfusion therapy. Chronically transfused iron overloadedpatients with SCD have significantly higher mortality thanless transfused counterparts without iron overload, as wellas age and race-matched normal controls [29, 30]. Thisfinding may reflect disease severity rather than iron burden;nonetheless, patients with SCD are far less likely to bescreened for end organ damage than patients with tha-lassemia, despite a similar transfusion history [31]. Thereforeknowledge of iron toxicity in SCD is of paramount impor-tance.

Transfusion of packed red blood cells (RBCs) provides1 mg per mL transfused of additional elemental iron. Long-term transfusion therapy of, for instance, 20-units RBCs/yearis associated with significant iron overload (20 units ×220 mL per unit × 1 mg per mL = 4400 mgm exogenousiron/year) [32]. With repeated transfusions, serum trans-ferrin becomes saturated and the excess circulating iron istransported as NTBI [33]. NTBI enters cells in a dysregulatedfashion; a subset of NTBI, called Labile Plasma Iron (LPI),may cause end organ damage secondary to its high redoxpotential [34].

5. End Organ Toxicity due toIron Overload in SCD

Saturation of transferrin by excess circulating iron resultsin increased NTBI and LPI [35]. NTBI can appear in theabsence of transferrin saturation, the mechanism of whichis not yet clear [36]. NTBI and LPI tend to enter tissuesmore readily and result in formation of reactive oxygen

Repeated transfusions resulting inincreased circulating iron

Increased saturation of transferrin

Increased NTBI and LPI

NTBI and LPI enter tissues and form ROS causinglipid peroxidation and end organ damage

Figure 2: Mechanism of end organ damage in iron overload.

species (ROS) such as the hydroxyl radical by the HaberWeiss reaction [37, 38]. Excess iron tends to deposit in thehepatic parenchyma, in endocrine organs, and in cardiacmyocytes, causing end organ damage by ROS-mediated lipidperoxidation [39, 40] (Figure 2).

In patients with β thalassemia major (TM), long-termtransfusion causes iron overload that results in cardiacdamage, liver fibrosis, gonadal dysfunction, and growthretardation. Cardiac iron overload still remains the maincause of mortality in TM despite chelation therapy [41].

Although large studies looking at long-term transfusionand iron overload in SCD are lacking, available data suggestthat SCD patients are relatively protected from iron-inducedcardiac and endocrine organ damage as compared with TMpatients. In a study by Wood et al. comparing 19 patientswith TM and 17 patients with SCD (matched for age, sex,transfusion duration, chelation therapy, and hepatic ironcontent), cardiac iron overload, measured by T2∗MRI, andcardiac dysfunction were significantly more prevalent in thegroup with TM [42]. In another study, cardiac disease andgonadal dysfunction seen in TM were not present in SCDpatients with similar serum ferritin levels and equivalent liveriron content. Hepatic fibrosis was detectable pathologically,albeit at 39% in SCD (n = 43) compared with 81% in TM(n = 30). TM patients had a significantly longer durationof transfusions and higher incidence of viral hepatitis in thisstudy [43].

Studies comparing the sequelae of iron overload in TMversus SCD are of interest, but one must be cautious inapplying conclusions from these studies given the manydifferences between TM and SCD patients clinically. Thesereports highlight the need for independent study of thecontribution of iron overload to the morbidity and mortalityin SCD.

6. Potential Factors ModifyingIron Toxicity in SCD

The possible difference in iron mediated toxicity betweenTM and SCD underscores the complexity of iron regulation.Damage from iron overload is not merely a function of theabsolute amount of excess iron present. Modifying factorsinclude, but are not limited to, relative distribution of ironloading in RES versus parenchymal cells, levels of NTBI


and LPI, coexisting hereditary iron loading defects, and theimpact of ineffective erythropoiesis.

The levels of NTBI and LPI, the most redox reactiveand toxic iron species, have been compared between SCDand TM. In a small study, NTBI levels in SCD were lessthan half of those seen in TM, despite higher ferritin levelsand hepatic iron concentration in SCD. However, patientswith thalassemia in this study had received 4 times moretransfusion than those with SCD [44]. Koren studied patientswith SCD including HbSS and Sβ◦ thalassemia (n = 36) andβ Thalassemia, including TM and Thalassemia intermedia(n = 43). The two groups overall were matched fortransfusion load (SCD: 104 ± 77 and β Thalassemia 179 ±103 cc/kg/yr). In subset analyses, however, TM patients hada significantly greater transfusion burden than patients withHbSS, which may account for significantly lower NTBI andLPI observed in SCD patients in this study. While only 1patient with SCD developed cardiomyopathy, 32 patientswith TM had evidence of either cardiac or gonadal damage[45].

Coexistent hereditary iron overload conditions have beenstudied in the context of SCD and the iron overload phe-notype. Currently defined hemochromatosis polymorphismsof European origin are less frequent in African Americansand do not seem to contribute to exacerbated iron loading inAfrican Americans with SCD [46]. Nonetheless, other causesof primary and dietary iron overload are well describedin populations of African descent, even if the underlyinggenetics are not, and these may contribute to phenotypicdiversity in SCD and iron overload [47, 48].

The role of hepcidin in the pathophysiology of ironoverload in SCD remains controversial. Patients with SCDsuffer from recurrent infections and ongoing endothelialdamage by reperfusion injury that result in a chronic-inflammatory-like state [49]. This is evidenced by increasedCRP, IFNγ, IL-1, and IL-6 in steady state and IL-6 and TNF-αincrease during crises [50, 51]. By analogy with the anemia ofchronic inflammation (in which cytokines increase hepcidinlevels), one may speculate that hepcidin levels would beelevated in SCD. However, hepcidin levels in nontransfusedsteady-state SCD (n = 40) are not elevated when comparedwith normal controls (n = 30), even when SCD patients withinfections and end organ damage were assessed as a subgroup[52]. Another study, comparing hepcidin in 16 patients withsickle syndromes (HbSS, HbSβ◦ thalassemia, HbSC) againstrace matched controls heterozygous for HbS or HbC, alsofound no difference among the groups. In fact, in 5 patientswith SCD in this study, hepcidin levels were below thelower limit of normal, which was attributed to increasederythropoietic activity [53]. Coexistent iron deficiency canalso affect hepcidin estimation. Iron deficiency anemia hasbeen described in the pediatric population with SCD, bothdue to nutritional status and intravascular hemolysis withurinary iron losses. In a study of 70 children with HbSS andHbSC, 9% of nontransfused children were found to have irondeficiency [54].

Although steady-state SCD patients do not exhibitincreased hepcidin levels, transfusion acutely up regulateshepcidin mRNA expression [55]. It is unclear whether this

increased gene expression results in increased circulatinghepcidin and decreased unbound iron immediately posttransfusion. Nor is it known whether this transient increasein hepcidin expression is of clinical relevance.

In contrast, hepcidin is significantly downregulated inpatients with ineffective erythropoiesis such as thalassemia,thereby worsening unbound iron toxicity in this population[56–58]. Depressed urinary hepcidin levels relative to ironburden have been associated with augmented but ineffectiveerythropoiesis (thalassemia, congenital dyserythropoieticanemia and sideroblastic anemia). This is in distinction toconditions with augmented but effective erythropoiesis, asin hemolysis (hereditary spherocytosis, and autoimmunehemolytic anemias), in which urinary hepcidin levels rel-ative to iron burden are not significantly depressed [59].Hepcidin suppression in thalassemia may be mediated inpart by growth differentiation factor 15 (GDF-15). GDF-15, a member of the TGF-β superfamily, is produced byerythroid progenitors and reaches measurable serum levelsin the presence of ineffective erythropoiesis. GDF-15 levelsare elevated in TM, but not in SCD. Serum from TMpatients suppresses hepcidin expression in vitro, unlessGDF-15 is blocked [60]. Twisted gastrulation (TWSG1), acytokine produced during early stages of erythropoiesis, alsodownregulates hepcidin and is highly expressed in the spleen,bone marrow, and liver of thalassemic mice. The role ofTWSG1 in SCD has not been studied yet [61]. Hepcidin andother cytokines (such as GDF 15 and TWSG1) are likely tobe key to observed iron regulation in SCD.

7. Measurement of Iron Overload

The gold standard for assessing liver iron stores in theabsence of cirrhosis is the hepatic iron content (HIC),determined by liver biopsy and quantitation with atomicabsorption spectrophotometry [62]. The normal HIC isbetween 0.4 and 2.2 mg/g of liver dry weight. Based ondata from hereditary hemochromatosis, <7 mg/g is notassociated with obvious hepatic pathology while >15 mg/gis consistently associated with liver fibrosis [63]. The use ofbiopsy-measured HIC as a marker of iron overload is limitedby the small but finite risk of complications of liver biopsy,lack of reproducibility of quantitative assays, and samplingerror [64].

Noninvasive methods including blood tests (ferritinand iron saturation) and imaging techniques (MRI-basedtechniques) have been evaluated as predictors of HIC.Ferritin has been shown to correlate with HIC in TM,but the correlation in SCD is less clear [65]. In a cross-sectional study of 27 children with SCD who had receivedchronic transfusion therapy without chelation, transfusionvolume correlated with hepatic iron content (HIC) but, whenadjusted for transfusion volume, serum markers did not[66]. In another study of 20 patients with SCD undergoingchronic transfusion therapy with iron chelation, HIC showeda positive correlation with the duration of transfusion andliver fibrosis but not with serum markers [67]. Analyses ofchronically transfused SCD patients without viral hepatitisfrom the STOP and STOP2 trials, most of whom were on


Table 1: Tests to estimate iron load.

TEST Application

Serum ferritinRelatively unreliable in SCD, especially in therange of 1500–3000 ng/mL. In this range, levelsof ferritin do not linearly correlate with HIC

SQUIDReliable predictor of HIC, but expensive andavailable in few institutions worldwide, mostlyfor research purposes

T2∗ MRIWell-validated predictor of HIC and cardiaccomplications from iron overload

Liver biopsyGold standard. Accurate estimation of ironoverload except in fibrosis. Invasive but <1%risk of complications. Sampling error possible

chelation therapy, showed that a ferritin level <1500 ng/mLwas correlated with low transfusion burden and low mea-sured HIC, while a ferritin >3000 ng/mL was consistentlypredictive of HIC >10 mg/g. The intermediate levels did notcorrelate linearly with iron overload [68]. These data suggestthat serum ferritin may not be an accurate predictor of liveriron stores, especially in the range of 1500 to 3000 ng/mL.

Imaging techniques for noninvasive determination ofHIC have been developed and validated. The Superconduct-ing Quantum Interference Device (SQUID) quantitativelydetermines HIC by magnetic measurement, and in casesof hereditary and transfusional iron overload has shownsignificant correlation with HIC as measured by biopsy [69].However, this device is expensive and not readily available.T2∗ MRI has been validated as a reliable noninvasive meansto assess iron stores in the liver and heart [70]. Increasingiron content in the liver and heart reduces relaxation timesas measured by T1, T2, and T2∗ MRI. T2∗ values of<20 milliseconds in nonfibrotic livers correlates (r = 0.93)with increased HIC by biopsy [71]. Similarly, cardiac T2∗

values of <20 ms correlate with a decline in left ventricularejection fraction values, increase in cardiac arrhythmias, andneed for cardiac medications [72]. The reciprocal of thisrelaxation time, R2 and R2∗, has also demonstrated gooddirect correlation with HIC in patients with SCD [73]. Fromthese data it appears that MRI is a good noninvasive tool toassess iron overload (Table 1).

8. Therapy for TransfusionalIron Overload in SCD

8.1. Non Pharmacological Therapy. Decreasing transfusionrequirement by substituting erythracytapharesis for simpletransfusion, transfusing younger red cells, and performingsplenectomy in patients with hypersplenism have beenevaluated as nonpharmacological techniques for preventingand treating iron overload. In erythrocytapharesis, patientRBCs are removed as transfused blood is being infused. Astudy was performed using erythrocytapharesis to a goal HbSof 50% in patients with history of stroke, recurrent VOCor priapism, comparing them to historic controls treatedwith simple transfusion or modified simple transfusions.The degree of iron overload, as assessed by serum ferritin,

was significantly reduced in 14 SCD patients treated byerythrocytapharesis compare to those treated with simpletransfusion, in spite of an increased absolute requirement fordonor units [74]. This approach is hampered by practicalresource limitations at donor centers and by complicationsof permanent vascular access in SCD [75, 76]. Furthermore,increased exposure to donor units also raises the risk ofalloimmunization to further deter its wider acceptance in thecommunity [77].

Transfusion of young erythrocytes called neocytes sig-nificantly reduces transfusion requirements and prolongstransfusion interval by 15–25%. However, this occurs at theexpense of increased donor exposure and costs [78], and thistechnique is not widely used.

Retrospective analysis of 34 pediatric patients withSCD (HbSS, HbSC, HbSβ◦ thal) for 6 months before and12 months after splenectomy for splenic sequestration orhypersplenism showed a significant reduction in transfusionrequirements postsplenectomy [79]. Splenectomy, however,is associated with long-term risk of infection by encapsulatedorganisms and likely benefits few adult patients with SCDsince autosplenectomy commonly leaves them without afunctioning spleen.

8.2. Pharmacological Therapy. Chelation therapy is routinelyemployed to prevent and treat iron overload in chronicallytransfused SCD patients. Chelators work by targeting theunbound iron in the blood, including NTBI and LPI thatcause tissue injury [80]. Different parameters have been usedto estimate iron load and determine need for chelation. Theseinclude HIC by liver biopsy, serum markers such as ferritin,and transfusional iron load (TIL).

8.2.1. Indications to Commence Chelation Therapy. Tradi-tionally, determining HIC by liver biopsy is consideredthe gold standard to predict iron overload and determineneed for chelation therapy. Based on data from hereditaryhemochromatosis and TM showing that elevated HIC over7 mg/g liver dry weight is a risk factor for hepatic fibrosis,this value has been used as a guide to start chelation [63, 81].In a study of 12 pediatric patients with SCD who previouslyreceived an average of 15 transfusions over 21 months, meanHIC was 9.4 ± 1.2 mg/g dry weight, and 4 of 12 patients hadliver fibrosis, validating that a cut off of 7 mg/g could be usedas a guide for SCD as well [82]. Vichinsky et al. described 43adult patients with SCD who were previously transfused fora mean of 6 years, resulting in elevated ferritin levels at 2916± 233 ng/mL, and increased HIC, at 14.33 ± 1.4 mg/g dryweight. Only 2% of these SCD patients had viral hepatitis,but 39% of them had an elevated fibrosis score at liver biopsysuggesting that adults with SCD have a similar response toiron overload [43].

As mentioned previously, noninvasive estimates of HICare less predictive of the need for liver biopsy and possibletherapy for iron overload. Serum ferritin of >1000 ng/mL,used as a guide in patients with thalassemia, is not validatedin SCD. While in SCD ferritin levels of over 3000 ng/mLare associated with HIC >10 mg/g, values between 1500and 3000 ng/mL are not predictive of an elevated HIC


Table 2: Indications for chelation therapy in SCD.

Test Adult patients Pediatric patients

Transfusional iron load 20 to 30 units >100 mg/kg

Serum ferritin >3000 ng/mL, 1500–3000 ng/mL: equivocal >3000 ng/mL, 1500–3000 ng/mL: equivocal

HIC >7–9 mg/g dry weight >7–9 mg/g dry weight

[67]. A better parameter for noninvasive estimation istransfusional iron load (TIL). TIL of >100 mg/kg has beenclosely correlated with high liver iron stores and liver fibrosisin the pediatric population and is an indication to startchelation therapy [65, 67] (Table 2). There are no systematicstudies in adults with SCD. 30 units of RBCs in a 70 Kgadult would result in 94 mg/Kg iron load, and so findingsin children with SCD are congruent with recommendationsfor secondary iron overload in myelodysplastic syndromein adults, where chelation is suggested after transfusionof 20–30 units of packed RBCs, or a ferritin of>2500 ng/mL [83] (see National Cancer ComprehensiveNetwork Clinical Practice Guidelines in Oncology v.2.2010)(http://www.nccn.org/professionals/physician gls/PDF/mds.pdf).

8.2.2. Chelating Agents. Deferoxamine (DFO), the first ironchelator introduced in the 1970s, is derived from the microbeStreptomyces pilosus. Due to poor oral absorption it isadministered parenterally. In the liver it is converted to activemetabolites that chelate-free iron and eliminate it throughurine and feces. Only 10% of the administered drug isavailable for chelation [84]. Although initially given intra-muscularly, therapy was later optimized as subcutaneousinfusions of 30–50 mg/kg for 8–12 hours every night for 5–7nights per week [41]. Dose adjustment is performed basedon ferritin to DFO ratio. With good compliance the drugprevents and reverses cardiac dysfunction and significantlyimproves survival [85]. Adverse effects include reversiblesensorineural hearing loss, retinal damage, and growthretardation [86]. Yearly audiologic and ophthalmologicevaluations are advised since early effects are reversible. Rarecases of renal failure and interstitial pneumonitis have alsobeen reported.

However, continuous parenteral administration makesDFO less attractive, and compliance is poor [87, 88].Children and adults with thalassemia have been treated withsubcutaneous DFO injections every twelve hours as a moreacceptable alternative to continuous infusions, though theamount of drug that can be administered by bolus injectionsis limited [89, 90].

Oral alternatives include Deferasirox (DFS) andDeferiprone (DFP). DFS is an FDA approved oral ironchelator. In a randomized open label multicenter phaseII study, one year of DFS therapy was compared againstDFO in chronically transfused SCD and both drugs werefound to be equally efficacious. Side effect profile of DFSwas similar to DFO including >10% headache, skin rash,and gastrointestinal toxicity as well as less common hearingand visual side effects [91]. Renal toxicity has been describedin DFS, and the target dose for efficacy is approximately

25–30 mg/Kg. Compliance is variable because of taste, andDFS, like DFO, is expensive.

Deferiprone is an oral iron chelator licensed for usein Europe. Safety of this drug was demonstrated in amulticenter prospective study but long-term data regardingefficacy is lacking. It appears to have the greatest efficacyin chelating cardiac iron but, because of early controversialstudies in North America, is not licensed for use by the FDA[92, 93].

The efficacy and safety of combination therapy usingdifferent chelators is also being studied.

9. Conclusion

Iron overload is a feared complication of long-term transfu-sion in SCD. The indications for transfusion in SCD continueto broaden and minimizing unnecessary transfusions inpatients with SCD should be strongly emphasized in theclinical setting to reduce complications of iron overload.Optimum use of nontransfusion therapy such as Hydrox-yurea may mitigate some, but not all need for transfusiontherapy. Whether nonmyeloablative bone marrow trans-plantation replaces long-term transfusion therapy in somesituations remains to be seen. SCD patients may be relativelyprotected from end organ damage due to iron toxicity, forreasons incompletely understood. Hepcidin is likely centralto differences observed between iron toxicity in SCD and inthalassemia.

In SCD, a condition with a strong inflammatory com-ponent, ferritin, which is an acute-phase reactant, maynot accurately predict body iron stores. Based on currentlyavailable data (often extrapolated from other diseases),chelation therapy should be considered when:

(1) adult patients with SCD have received 20–30 units ofRBC transfusions,

(2) pediatric patients with SCD are approaching a TIL ≥100 mg/Kg, and/or

(3) HIC in any age group exceeds 7–9 mg/g

(a) Excessive HIC is likely when the serum ferritinis >3000 ng/mL, less clear at 1500–3000 ng/mL.

Oral and parenteral chelators with a range of sideeffects and costs are now available that can be tailoredto individual patients; nonetheless, simple and inexpensivechelation remains an elusive goal.

Challenges remain in the prevention and managementof iron overload in SCD. Increased federal research supportwould benefit clinical investigation into practical matterssuch as clarification of the indications for transfusion therapy


in adults with SCD, improved catheter technology to allowsafe exchange transfusions, and optimal chelation therapy.Important unaddressed pathophysiologic questions includethe etiology of nonhepatic risks of iron overload in adults(such as pulmonary hypertension, the risk for which hasbeen correlated with ferritin levels), and a more completeelucidation of the role that perturbed iron absorption,transport, and storage play in damage from iron overload inSCD.

Acknowledgment

The authors would like to thank Dr. Lawrence Cytryn and Dr.Oswaldo Castro for reviewing the manuscript and providingvaluable suggestions.

References

[1] P. B. Walter, P. Harmatz, and E. Vichinsky, “Iron metabolismand iron chelation in sickle cell disease,” Acta Haematologica,vol. 122, no. 2-3, pp. 174–183, 2009.

[2] R. Dubach, C. V. Moore, and S. Callender, “Studies in irontransportation and metabolism. IX. The excretion of iron asmeasured by the isotope technique,” The Journal of Laboratoryand Clinical Medicine, vol. 45, no. 4, pp. 599–615, 1955.

[3] N. C. Andrews, “Iron homeostasis: insights from genetics andanimal models,” Nature Reviews Genetics, vol. 1, no. 3, pp.208–217, 2000.

[4] R. E. Fleming and B. R. Bacon, “Orchestration of ironhomeostasis,” New England Journal of Medicine, vol. 352, no.17, pp. 1741–1744, 2005.

[5] M. W. Hentze, M. U. Muckenthaler, and N. C. Andrews,“Balancing acts: molecular control of mammalian ironmetabolism,” Cell, vol. 117, no. 3, pp. 285–297, 2004.

[6] E. Nemeth, “Iron regulation and erythropoiesis,” CurrentOpinion in Hematology, vol. 15, no. 3, pp. 169–175, 2008.



[9] X. Du, E. She, T. Gelbart, et al., “The serine protease TMPRSS6is required to sense iron deficiency,” Science, vol. 320, no. 5879,pp. 1088–1092, 2008.

[10] M. Shayeghi, G. O. Latunde-Dada, J. S. Oakhill, et al.,“Identification of an intestinal heme transporter,” Cell, vol.122, no. 5, pp. 789–801, 2005.

[11] S. B. Raffin, C. H. Woo, and K. T. Roost, “Intestinalabsorption of hemoglobin iron heme cleavage by mucosalheme oxygenase,” Journal of Clinical Investigation, vol. 54, no.6, pp. 1344–1352, 1974.



[14] M. O. Russell, H. I. Goldberg, and A. Hodson, “Effect oftransfusion therapy on arteriographic abnormalities and onrecurrence of stroke in sickle cell disease,” Blood, vol. 63, no. 1,pp. 162–169, 1984.

[15] M. J. Telen, “Principles and problems of transfusion in sicklecell disease,” Seminars in Hematology, vol. 38, no. 4, pp. 315–323, 2001.

[16] S. K. Ballas, “Sickle cell disease: current clinical management,”Seminars in Hematology, vol. 38, no. 4, pp. 307–314, 2001.

[17] C. H. Pegelow, R. J. Adams, V. McKie, et al., “Risk ofrecurrent stroke in patients with sickle cell disease treated witherythrocyte transfusions,” Journal of Pediatrics, vol. 126, no. 6,pp. 896–899, 1995.

[18] R. J. Adams, V. C. Mckie, L. Hsu, et al., “Prevention of a firststroke by transfusions in children with sickle cell anemia andabnormal results on transcranial Doppler ultrasonography,”New England Journal of Medicine, vol. 339, no. 1, pp. 5–11,1998.

[19] R. J. Adams and D. Brambilla, “Discontinuing prophylactictransfusions used to prevent stroke in sickle cell disease,” NewEngland Journal of Medicine, vol. 353, no. 26, pp. 2769–2778,2005.

[20] P. S. Swerdlow, “Red cell exchange in sickle cell disease,”Hematology, pp. 48–53, 2006.

[21] U. Emre, S. T. Miller, M. Gutierez, P. Steiner, S. P. Rao, and M.Rao, “Effect of transfusion in acute chest syndrome of sicklecell disease,” Journal of Pediatrics, vol. 127, no. 6, pp. 901–904,1995.

[22] E. P. Vichinsky, L. D. Neumayr, A. N. Earles, et al., “Causes andoutcomes of the acute chest syndrome in sickle cell disease,”New England Journal of Medicine, vol. 342, no. 25, pp. 1855–1865, 2000.

[23] S. T. Miller, E. Wright, M. Abboud, et al., “Impact of chronictransfusion on incidence of pain and acute chest syndromeduring the Stroke Prevention Trial (STOP) in sickle-cellanemia,” Journal of Pediatrics, vol. 139, no. 6, pp. 785–789,2001.

[24] L. A. Styles, M. Abboud, S. Larkin, M. Lo, and F. A. Kuypers,“Transfusion prevents acute chest syndrome predicted byelevated secretory phospholipase A2,” British Journal ofHaematology, vol. 136, no. 2, pp. 343–344, 2007.

[25] M. Koshy, L. Burd, D. Wallace, A. Moawad, and J. Baron,“Prophylactic red-cell transfusions in pregnant patients withsickle cell disease. A randomized cooperative study,” NewEngland Journal of Medicine, vol. 319, no. 22, pp. 1447–1452,1988.

[26] E. P. Vichinsky, C. M. Haberkern, L. Neumayr, et al., “A com-parison of conservative and aggressive transfusion regimensin the perioperative management of sickle cell disease,” NewEngland Journal of Medicine, vol. 333, no. 4, pp. 206–213, 1995.

[27] M. Koshy, S. J. Weiner, S. T. Miller, et al., “Surgery andanesthesia in sickle cell disease,” Blood, vol. 86, no. 10, pp.3676–3684, 1995.

[28] R. Ware, H. C. Filston, W. H. Schultz, and T. R. Kin-ney, “Elective cholecystectomy in children with sickle hemo-globinopathies. Successful outcome using a preoperativetransfusion regimen,” Annals of Surgery, vol. 208, no. 1, pp.17–22, 1988.

[29] E. B. Fung, P. Harmatz, M. Milet, et al., “Morbidity andmortality in chronically transfused subjects with Thalassemiaand Sickle Cell Disease: a report from the multi-center studyof iron overload,” American Journal of Hematology, vol. 82, no.4, pp. 255–265, 2007.


[30] S. K. Ballas, “Iron overload is a determinant of morbidity andmortality in adult patients with sickle cell disease,” Seminars inHematology, vol. 38, no. 1, supplement 1, pp. 30–36, 2001.

[31] E. B. Fung, P. R. Harmatz, M. Milet, et al., “Disparity inthe management of iron overload between patients withsickle cell disease and thalassemia who received transfusions,”Transfusion, vol. 48, no. 9, pp. 1971–1980, 2008.

[32] N. C. Andrews, “Disorders of iron metabolism,” New EnglandJournal of Medicine, vol. 341, no. 26, pp. 1986–1995, 1999.

[33] W. C. Wang, N. Ahmed, and M. Hanna, “Non-transferrin-bound iron in long-term transfusion in children with congen-ital anemias,” Journal of Pediatrics, vol. 108, no. 4, pp. 552–557,1986.

[34] B. P. Esposito, W. Breuer, P. Sirankapracha, P. Pootrakul, C.Hershko, and Z. I. Cabantchik, “Labile plasma iron in ironoverload: redox activity and susceptibility to chelation,” Blood,vol. 102, no. 7, pp. 2670–2677, 2003.

[35] C. Hershko, G. Graham, G. W. Bates, and E. A. Rachmilewitz,“Non-specific serum iron in thalassaemia: an abnormalserum iron fraction of potential toxicity,” British Journal ofHaematology, vol. 40, no. 2, pp. 255–263, 1978.

[36] W. Breuer, C. Hershko, and Z. I. Cabantchik, “The impor-tance of non-transferrin bound iron in disorders of ironmetabolism,” Transfusion Science, vol. 23, no. 3, pp. 185–192,2000.

[37] G. Link, A. Pinson, and C. Hershko, “Heart cells in culture: amodel of myocardial iron overload and chelation,” The Journalof Laboratory and Clinical Medicine, vol. 106, no. 2, pp. 147–153, 1985.

[38] Z. I. Cabantchik, W. Breuer, G. Zanninelli, and P. Cianciulli,“LPI-labile plasma iron in iron overload,” Best Practice andResearch: Clinical Haematology, vol. 18, no. 2, pp. 277–287,2005.

[39] J. G. Goddard and G. D. Sweeney, “Ferric nitrilotriacetate:a potent stimulant of in vivo lipid peroxidation in mice,”Biochemical Pharmacology, vol. 32, no. 24, pp. 3879–3882,1983.

[40] C. Hershko, G. Link, A. M. Konijn, and Z. I. Cabantchik,“Objectives and mechanism of iron chelation therapy,” Annalsof the New York Academy of Sciences, vol. 1054, pp. 124–135,2005.

[41] N. F. Olivieri and G. M. Brittenham, “Iron-chelating therapyand the treatment of thalassemia,” Blood, vol. 89, no. 3, pp.739–761, 1997.

[42] J. C. Wood, J. M. Tyszka, S. Carson, M. D. Nelson, and T.D. Coates, “Myocardial iron loading in transfusion-dependentthalassemia and sickle cell disease,” Blood, vol. 103, no. 5, pp.1934–1936, 2004.

[43] E. Vichinsky, E. Butensky, E. Fung, et al., “Comparison oforgan dysfunction in transfused patients with SCD or βthalassemia,” American Journal of Hematology, vol. 80, no. 1,pp. 70–74, 2005.

[44] P. B. Walter, E. B. Fung, D. W. Killilea, et al., “Oxidativestress and inflammation in iron-overloaded patients with β-thalassaemia or sickle cell disease,” British Journal of Haema-tology, vol. 135, no. 2, pp. 254–263, 2006.

[45] A. Koren, D. Fink, O. Admoni, Y. Tennenbaum-Rakover, andC. Levin, “Non-transferrin bound labile plasma iron and ironoverload in sickle cell disease: a comparative study betweensickle cell disease and β thalassemic patients,” EuropeanJournal of Haematology, vol. 84, no. 1, pp. 72–78, 2010.

[46] M. R. Jeng, P. Adams-Graves, T. A. Howard, M. R. Whorton,C.-S. Li, and R. E. Ware, “Identification of hemochromatosisgene polymorphisms in chronically transfused patients withsickle cell diseases,” American Journal of Hematology, vol. 74,no. 4, pp. 243–248, 2003.

[47] V. R. Gordeuk, R. D. Boyd, and G. M. Brittenham, “Dietaryiron overload persists in rural sub-Saharan Africa,” Lancet, vol.1, no. 8493, pp. 1310–1313, 1986.

[48] R. K. Wurapa, V. R. Gordeuk, G. M. Brittenham, A. Khiyami,G. P. Schechter, and C. Q. Edwards, “Primary iron overloadin African Americans,” American Journal of Medicine, vol. 101,no. 1, pp. 9–18, 1996.

[49] R. P. Hebbel, R. Osarogiagbon, and D. Kaul, “The endothelialbiology of sickle cell disease: inflammation and a chronicvasculopathy,” Microcirculation, vol. 11, no. 2, pp. 129–151,2004.

[50] M. L. Jison, P. J. Munson, J. J. Barb, et al., “Blood mononu-clear cell gene expression profiles characterize the oxidant,hemolytic, and inflammatory stress of sickle cell disease,”Blood, vol. 104, no. 1, pp. 270–280, 2004.

[51] J. D. Belcher, P. H. Marker, J. P. Weber, R. P. Hebbel, andG. M. Vercellotti, “Activated monocytes in sickle cell disease:potential role in the activation of vascular endothelium andvaso-occlusion,” Blood, vol. 96, no. 7, pp. 2451–2459, 2000.

[52] C. Ezeh, C. C. Ugochukwu, J. Weinstein, and I. Okpala,“Hepcidin, haemoglobin and ferritin levels in sickle cellanaemia,” European Journal of Haematology, vol. 74, no. 1, pp.86–88, 2005.

[53] J. J. C. Kroot, C. M. M. Laarakkers, E. H. J. M. Kemna, B. J.Biemond, and D. W. Swinkels, “Regulation of serum hepcidinlevels in sickle cell disease,” Haematologica, vol. 94, no. 6, pp.885–887, 2009.

[54] E. Vichinsky, K. Kleman, S. Embury, and B. Lubin, “Thediagnosis of iron deficiency anemia in sickle cell disease,”Blood, vol. 58, no. 5, pp. 963–968, 1981.

[55] Z. A. Jenkins, W. Hagar, C. L. Bowlus, et al., “Iron homeostasisduring transfusional iron overload in β-thalassemia and sicklecell disease: changes in iron regulatory protein, hepcidin, andferritin expression,” Pediatric Hematology and Oncology, vol.24, no. 4, pp. 237–243, 2007.

[56] G. Papanikolaou, M. Tzilianos, J. I. Christakis, et al., “Hepcidinin iron overload disorders,” Blood, vol. 105, no. 10, pp. 4103–4105, 2005.

[57] R. Origa, R. Galanello, T. Ganz, et al., “Liver iron concentra-tions and urinary hepcidin in β-thalassemia,” Haematologica,vol. 92, no. 5, pp. 583–588, 2007.






[62] E. Angelucci, G. M. Brittenham, C. E. McLaren, et al., “Hepaticiron concentration and total body iron stores in thalassemiamajor,” New England Journal of Medicine, vol. 343, no. 5, pp.327–331, 2000.

[63] M. L. Bassett, J. W. Halliday, and L. W. Powell, “Valueof hepatic iron measurements in early hemochromatosisand determination of the critical iron level associated withfibrosis,” Hepatology, vol. 6, no. 1, pp. 24–29, 1986.

[64] E. Angelucci, D. Baronciani, G. Lucarelli, et al., “Needleliver biopsy in thalassaemia: analyses of diagnostic accuracyand safety in 1184 consecutive biopsies,” British Journal ofHaematology, vol. 89, no. 4, pp. 757–761, 1995.

[65] P. T. Telfer, E. Prestcott, S. Holden, M. Walker, A. V. Hoffbrand,and B. Wonke, “Hepatic iron concentration combined withlong-term monitoring of serum ferritin to predict complica-tions of iron overload in thalassaemia major,” British Journalof Haematology, vol. 110, no. 4, pp. 971–977, 2000.

[66] K. Brown, C. Subramony, W. May, et al., “Hepatic iron over-load in children with sickle cell anemia on chronic transfusiontherapy,” Journal of Pediatric Hematology/Oncology, vol. 31, no.5, pp. 309–312, 2009.

[67] P. Harmatz, E. Butensky, K. Quirolo, et al., “Severity of ironoverload in patients with sickle cell disease receiving chronicred blood cell transfusion therapy,” Blood, vol. 96, no. 1, pp.76–79, 2000.

[68] T. V. Adamkiewicz, M. R. Abboud, C. Paley, et al., “Serumferritin level changes in children with sickle cell disease onchronic blood transfusion are nonlinear and are associatedwith iron load and liver injury,” Blood, vol. 114, no. 21, pp.4632–4638, 2009.

[69] G. M. Brittenham, D. E. Farrell, and J. W. Harris, “Magnetic-susceptibility measurement of human iron stores,” New Eng-land Journal of Medicine, vol. 307, no. 27, pp. 1671–1675, 1982.

[70] E. Voskaridou, M. Douskou, E. Terpos, et al., “Magneticresonance imaging in the evaluation of iron overload inpatients with beta thalassaemia and sickle cell disease,” BritishJournal of Haematology, vol. 126, no. 5, pp. 736–742, 2004.

[71] E. Voskaridou, M. Douskou, E. Terpos, et al., “Deferiproneas an oral iron chelator in sickle cell disease,” Annals ofHematology, vol. 84, no. 7, pp. 434–440, 2005.

[72] L. J. Anderson, S. Holden, B. Davis, et al., “CardiovascularT2-star (T2∗) magnetic resonance for the early diagnosis ofmyocardial iron overload,” European Heart Journal, vol. 22, no.23, pp. 2171–2179, 2001.

[73] J. C. Wood, C. Enriquez, N. Ghugre, et al., “MRI R2 andR2∗ mapping accurately estimates hepatic iron concentrationin transfusion-dependent thalassemia and sickle cell diseasepatients,” Blood, vol. 106, no. 4, pp. 1460–1465, 2005.

[74] H. C. Kim, N. P. Dugan, J. H. Silber, et al., “Erythrocytaphere-sis therapy to reduce iron overload in chronically transfusedpatients with sickle cell disease,” Blood, vol. 83, no. 4, pp. 1136–1142, 1994.

[75] S. C. Wagner, D. J. Eschelman, C. F. Gonsalves, J. Bonn, and K.L. Sullivan, “Infectious complications of implantable venousaccess devices in patients with sickle cell disease,” Journal ofVascular and Interventional Radiology, vol. 15, no. 4, pp. 375–378, 2004.

[76] C. E. McCready, H. A. Doughty, and T. C. Pearson, “Experi-ence with the Port-A-Cath in sickle cell disease,” Clinical andLaboratory Haematology, vol. 18, no. 2, pp. 79–82, 1996.

[77] B. Aygun, S. Padmanabhan, C. Paley, and V. Chandrasekaran,“Clinical significance of RBC alloantibodies and autoanti-bodies in sickle cell patients who received transfusions,”Transfusion, vol. 42, no. 1, pp. 37–43, 2002.

[78] A. F. Collins, C. Goncalves-Dias, S. Haddad, et al., “Compar-ison of a transfusion preparation of newly formed red cellsand standard washed red cell transfusions in patients withhomozygous β- thalassemia,” Transfusion, vol. 34, no. 6, pp.517–520, 1994.

[79] R. N. Haricharan, J. M. Roberts, T. L. Morgan, et al.,“Splenectomy reduces packed red cell transfusion requirementin children with sickle cell disease,” Journal of Pediatric Surgery,vol. 43, no. 6, pp. 1052–1056, 2008.

[80] P. Pootrakul, W. Breuer, M. Sametband, P. Sirankapracha, C.Hershko, and Z. I. Cabantchik, “Labile plasma iron (LPI)as an indicator of chelatable plasma redox activity in iron-overloaded β-thalassemia/HbE patients treated with an oralchelator,” Blood, vol. 104, no. 5, pp. 1504–1510, 2004.

[81] R. A. Risdon, M. Barry, and D. M. Flynn, “Transfusional ironoverload: the relationship between tissue iron concentrationand hepatic fibrosis in thalassaemia,” Journal of Pathology, vol.116, no. 2, pp. 83–95, 1975.

[82] N. F. Olivieri, “Progression of iron overload in sickle celldisease,” Seminars in Hematology, vol. 38, no. 1, supplement1, pp. 57–62, 2001.

[83] S. Mahesh, Y. Ginzburg, and A. Verma, “Iron overload inmyelodysplastic syndromes,” Leukemia and Lymphoma, vol.49, no. 3, pp. 427–438, 2008.

[84] J. B. Porter, “Practical management of iron overload,” BritishJournal of Haematology, vol. 115, no. 2, pp. 239–252, 2001.

[85] G. M. Brittenham, P. M. Griffith, A. W. Nienhuis, et al.,“Efficacy of deferoxamine in preventing complications of ironoverload in patients with thalassemia major,” New EnglandJournal of Medicine, vol. 331, no. 9, pp. 567–573, 1994.

[86] N. F. Olivieri, J. R. Buncic, and E. Chew, “Visual andauditory neurotoxicity in patients receiving subcutaneousdeferoxamine infusions,” New England Journal of Medicine,vol. 314, no. 14, pp. 869–873, 1986.

[87] G. Kidson-Gerber and R. Lindeman, “Adherence to desfer-rioxamine and deferiprone and the impact of deferiproneco-prescription in thalassaemia major patients. Does theaddition of deferiprone improve adherence?” British Journalof Haematology, vol. 142, no. 4, pp. 679–680, 2008.

[88] M. D. Cappellini, “Overcoming the challenge of patient com-pliance with iron chelation therapy,” Seminars in Hematology,vol. 42, no. 2, supplement 1, pp. S19–S21, 2005.

[89] M. Franchini, G. Gandini, D. Veneri, and G. Aprili, “Safety andefficacy of subcutaneous bolus injection of deferoxamine inadult patients with iron overload: an update,” Blood, vol. 103,no. 2, pp. 747–748, 2004.

[90] N. Yarali, T. Fisgin, F. Duru, et al., “Subcutaneous bolusinjection of deferoxamine is an alternative method to subcu-taneous continuous infusion,” Journal of Pediatric Hematol-ogy/Oncology, vol. 28, no. 1, pp. 11–16, 2006.

[91] E. Vichinsky, O. Onyekwere, J. Porter, et al., “A randomisedcomparison of deferasirox versus deferoxamine for the treat-ment of transfusional iron overload in sickle cell disease,”British Journal of Haematology, vol. 136, no. 3, pp. 501–508,2007.

[92] A. R. Cohen, R. Galanello, A. Piga, V. De Sanctis, and F. Tricta,“Safety and effectiveness of long-term therapy with the oraliron chelator deferiprone,” Blood, vol. 102, no. 5, pp. 1583–1587, 2003.

[93] D. J. Roberts, S. J. Brunskill, C. Doree, S. Williams, J. Howard,and C. J. Hyde, “Oral deferiprone for iron chelation in peoplewith thalassaemia,” Cochrane Database of Systematic Reviews,no. 3, Article ID CD004839, 2007.


Review Article

Diamond Blackfan Anemia at the Crossroad between RibosomeBiogenesis and Heme Metabolism

Deborah Chiabrando and Emanuela Tolosano

Molecular Biotechnology Center, University of Torino, Via Nizza 52, 10126 Torino, Italy

Correspondence should be addressed to Emanuela Tolosano, [email protected]

Received 20 November 2009; Revised 22 January 2010; Accepted 16 February 2010

Academic Editor: Jeffery Lynn Miller

Copyright © 2010 D. Chiabrando and E. Tolosano. This is an open access article distributed under the Creative CommonsAttribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

Diamond-Blackfan anemia (DBA) is a rare, pure red-cell aplasia that presents during infancy. Approximately 40% of cases areassociated with other congenital defects, particularly malformations of the upper limb or craniofacial region. Mutations in thegene coding for the ribosomal protein RPS19 have been identified in 25% of patients with DBA, with resulting impairment of 18SrRNA processing and 40S ribosomal subunit formation. Moreover, mutations in other ribosomal protein coding genes account forabout 25% of other DBA cases. Recently, the analysis of mice from which the gene coding for the heme exporter Feline LeukemiaVirus subgroup C Receptor (FLVCR1) is deleted suggested that this gene may be involved in the pathogenesis of DBA. FLVCR1-null mice show a phenotype resembling that of DBA patients, including erythroid failure and malformations. Interestingly, someDBA patients have disease linkage to chromosome 1q31, where FLVCR1 is mapped. Moreover, it has been reported that cellsfrom DBA patients express alternatively spliced isoforms of FLVCR1 which encode non-functional proteins. Herein, we review theknown roles of RPS19 and FLVCR1 in ribosome function and heme metabolism respectively, and discuss how the deficiency of aribosomal protein or of a heme exporter may result in the same phenotype.

1. Introduction

Diamond Blackfan anemia (DBA; OMIM #205900) is arare, congenital, pure red-cell aplasia that presents duringinfancy, usually within the first year of life. Its main clinicalfeatures are normochromic and macrocytic anemia, reticu-locytopenia, and nearly complete absence of red blood cellprecursors in the bone marrow. The erythroid hypoplasia isdue to impaired proliferation and differentiation of red bloodcell progenitors in the bone marrow. Other hematopoieticlineages are usually normal. High serum levels of folicacid, vitamin B12, erythropoietin, elevated fetal hemoglobin,and erythrocyte adenosine deaminase further support DBAdiagnosis [1].

DBA is a clinically heterogeneous disorder. In addition toerythroid failure, it is also characterized by congenital mal-formations and cancer predisposition. Growth retardationand a wide variety of congenital anomalies have been seen inmore than 50% of DBA cases. Short stature is constitutionalin most patients. Thumbs, upper limbs, and hands and

craniofacial, urogenital, and cardiovascular anomalies arealso common. Although not yet statistically validated, DBApatients have an increased risk of cancer development.Both hematopoietic malignancy (acute myeloid leukemia,myelodysplastic syndrome, Hodgkin and non/Hodgkin lym-phomas, and acute lymphoblastic leukemia) and non-hematopoietic tumors (osteogenic sarcoma, breast can-cer, hepatocellular carcinoma, melanoma, fibrohistiocytoma,gastric cancer, and colon cancer) have been described insome DBA patients [2, 3].

DBA is an autosomal dominant disorder with a diseaseincidence of 5–10 cases per million live births in Europe.The molecular genetics of DBA is evident in about half ofthe patients. All the DBA mutations identified so far, bothin sporadic and familial cases, have been found in genescoding for ribosomal proteins, and DBA is now consideredthe prototype of “ribosome-based disorders”. The first DBAgene to be identified is Ribosomal Protein S19 (RPS19),located on chromosome 19q13.2 and mutated in 25% ofDBA patients. Recently, mutations in several other ribosomal


proteins, RPS24, RPS17, RPL11, RPL5, RPS7, and RPL35a,have been identified in approximately 20% of DBA patients[4]. In the remaining 55% of DBA patients, no mutationshave been reported suggesting the existence of other genesinvolved in the pathogenesis of DBA.

Recently, it has been demonstrated that Feline LeukemiaVirus subgroup C Receptor (FLVCR1) deficient embryosdisplay a phenotype very close to DBA patients, includingbone marrow failure and congenital malformations [5].FLVCR1 is a heme exporter [6, 7] suggesting that alteredheme homeostasis could play a fundamental role in thepathogenesis of DBA. Although no mutations in FLVCR1have been found in DBA patients [8], it will be interestingto understand how the deficiency in a heme exportercould recapitulate the human disease and if a link betweenFLVCR1 and the ribosome biosynthetic pathway exists. Inthis paper, the possible molecular mechanisms underlyingthe pathogenesis of DBA are discussed.

2. Diamond-Blackfan Anemia as a Disorder ofRibosome Biogenesis

2.1. Ribosomal Protein S19. RPS19 is one of 33 ribosomalproteins that, together with 18S rRNA, constitute the 40Sribosomal subunit. Immunoelectron microscope studiesshowed that it localizes to the external surface of the 40Ssubunit, in close proximity to RPS3A, RPS13, RPS16, andRPS24, a region that interacts with the eukaryotic initiationfactor eIF-2 [9].

RPS19 is a highly conserved, ubiquitously expressed pro-tein. In red blood cells, RPS19 expression strongly decreasesduring terminal erythroid differentiation [10]. RPS19 islocalized predominantly in the nucleus, in particular into thenucleolus where ribosome synthesis takes place, and in thecytoplasm as a ribosomal component [11].

Different kinds of mutation in RPS19 have been discov-ered in DBA patients: nonsense and missense mutations,deletions, insertions, splice defects, and larger rearrange-ments. RPS19 mutations could lead to reduction of mRNAand/or protein levels, loss of nucleolar localization, orimpairment of ribosomal association [12]. All the mutationsidentified so far in ribosomal proteins have been found inheterozygosity. How the haploinsufficiency of a ribosomalprotein may lead to erythroid failure, malformations, andcancer is still not completely understood and the multiplemolecular mechanisms in which ribosomal proteins areinvolved are only just becoming clearer.

2.2. In Vitro and In Vivo Models of RPS19 Deficiency. Theerythroid hypoplasia characterizing DBA is due to a defectof erythroid progenitor differentiation. CD34+ bone marrowcells derived from DBA patients and cultured in vitro showa reduction in proliferation rates and colony formationcapacity associated with increased apoptosis [13, 14].

The fact that RPS19 has a fundamental role duringerythropoiesis was first evident when RPS19 mutations wereidentified in DBA patients. This role was further confirmedby in vitro studies in which the effect of RPS19 silencing

and overexpression was analyzed. In both primary culture(CD34+ umbilical cord blood and bone marrow cells)and erythroid-like cell lines (TF-1 and U-7 cells) RPS19silencing mimics the DBA phenotype: impaired erythroiddifferentiation and proliferation of erythroid progenitors,cell growth arrest at G0/G1, and apoptosis [13, 15]. Similarresults were obtained by overexpressing mutated forms ofRPS19, analogous to those found in DBA patients, in K562cells and in human CD34+ bone marrow cells [16]. On thecontrary, RPS19 overexpression into CD34+ bone marrowcells from RPS19-deficient DBA patients increases the ery-throid colony formation capacity and improves erythroidprogenitors proliferation in vitro [17–19]. Furthermore, theloss of different ribosomal proteins causes different cell cycledefects; primary fibroblast from DBA patients with RPS19mutations is characterized by cell cycle arrest at the G1 phasewhile RPS24 mutation impairs the progression through the Sphase [20].

Generation of a mouse model of RPS19 deficiency hasbeen attempted by Matsson and colleagues but the deletionof RPS19 in mice is lethal before implantation; no RPS19deficient embryo may be recovered as early as the blasto-cyst stage. Moreover, RPS19 heterozygous mice are viablebut indistinguishable from wild-type mice: hematologicalparameters, erythroid proliferation, and differentiation arenormal as well as organ development and morphology. Nodifference may be found in DBA markers such as erythrocyteadenosine deaminase and globin isoforms. Contrary tothe occurrence in human patients, heterozygous mice fullycompensate the loss of one RPS19 allele with normal proteinlevels [21, 22].

On the other hand, zebrafish appears to be a goodanimal model to study the consequences of RPS19 defi-ciency. Knocking-down RPS19 in zebrafish, using antisensemorpholinos, perfectly recapitulates the DBA phenotype.Both the primitive and definitive waves of erythropoiesisare impaired: an increased number of red cell progenitorsremains in the intermediate cell mass and primitive cellsare not replaced by adult ones. Also erythroid markers ofdifferentiation, such as gata-1 and c-myb, are expressed atlower levels compared to controls. Myeloid and lymphoidlineages appear normal. Morphants also show delayed devel-opment and morphological defects that are evident duringsomatogenesis: reduced forebrain and eye field, smaller headand eyes with head cartilage not formed properly [23–25]. Similar phenotypes are observed when the expressionof other ribosomal proteins is lost. The knock-down ofRPL35, RPL35a, and RPLP2 in zebrafish leads to erythroidfailure and malformations [24], while the loss of RPL11 onlyleads to alteration of embryonic development [26]. Thesedata highlight an important role for ribosomal proteinsnot only during erythropoiesis but also for proper organdevelopment.

The fundamental role of ribosome integrity for ery-thropoiesis and development is further confirmed by theobservation that a heterogeneous group of disorders, referredto as “inherited bone marrow failure syndromes”, is alsocharacterized by bone marrow failure, congenital malfor-mations, and cancer predisposition. In addition to DBA,


this group of disorders include dyskeratosis congenita (DC),cartilage-hair hypoplasia (CHH), and Shwachman-Diamondsyndrome (SDS). In each case, a defect in proteins thatare involved in ribosome synthesis has been identified. DCis due to mutations in dyskerin gene (DKC1), a puta-tive pseudouridyl synthase involved in rRNA modification.Mutations in RMPR, which encode the RNA componentof ribonuclease mitochondrial RNA processing complex(RNase MRP), involved in the cleavage of rRNA precursorsand other functions, have been identified in CHH. SDS is dueto mutations in SBDS, which encode for a protein involved inribosome synthesis [1].

2.3. Role of RPS19 in Ribosome Biogenesis. The mammalianribosome is composed of a large 60S and a small 40Ssubunit consisting of 4 rRNAs and 80 ribosomal proteins.Ribosome biogenesis is a very complex process. First, apolycistronic pre-rRNA transcript is processed into the 18SrRNA (component of the 40S subunit) and the 5.8S and25S/28S rRNAs (component of the 60S subunit). Then,rRNAs associate with the ribosomal proteins and the 5SrRNA, which is independently transcribed, and form an early90S particle that is subsequently processed into 66S and43S preribosomes. Following assembly in the nucleolus, thepreribosomes are exported into the cytoplasm through thenuclear pore complex as independent entities and additionalmaturation steps are necessary to achieve translationalcompetence. RPS19 is one of the ribosomal proteins that,together with 18S rRNA, constitute the small 40S subunit[27].

Recent studies demonstrated that RPS19 not only has astructural role, as initially thought, but also has additionalfunctions. The analysis of the RPS19 interactome, by a highthroughput proteomic approach, suggests that RPS19 mayplay an important role in rRNA processing, metabolism,and translation. One hundred and fifty-nine new proteinsinteracting with RPS19 have been identified: NT-Pases,hydrolases/helicases, isomerases, kinases, splicing factors,structural constituents of ribosomes, transcription factors,transferases, transporters, and DNA/RNA-binding protein. Ithas been demonstrated that RPS19 not only interacts withother components of the 40S ribosome but also with thepreribosome 90S. Moreover, RPS19 interacts with proteinsimportant for the transport of the small subunit from thenucleus to the cytoplasm and also with proteins involved inthe pseudouridylation of rRNA, a process involved in earlystages of ribosome biogenesis [28].

Interference with RPS19 expression in vitro leads to areduced maturation of the 40S ribosome subunit, increasedexpression of the 60S subunit, reduced 18S rRNA synthesis,and accumulation of a novel 21S pre-rRNA [1, 29]. Thesame results were obtained analyzing bone marrow CD34+cells and fibroblasts from DBA patients [13, 30]. These datasuggest that RPS19 might be required for a specific step in18S rRNA processing from the 21S precursor, highlightingthat RPS19 actively participates in ribosome biogenesis.

Recent works showed that RPS19 haploinsufficiencyleads to the posttranscriptional downregulation of othersmall subunit ribosomal proteins such as RPS20, RPS21 and

RPS24 compared to large subunit ribosomal proteins. Thishas been demonstrated in TF1 cells interfered for RPS19and in lymphoblastoid and fibroblast cell lines derived fromRPS19 deficient DBA patients [20, 31] suggesting the impor-tance of the correct subunits amount and stoichiometry forribosome assembly.

2.4. Role of p53 Activation in the Pathogenesis of DBA.Recent data suggest an important role of p53 family inthe pathogenesis of DBA. The knock-down of RPS19 inzebrafish leads to erythroid and developmental defects whichare associated with over-expression of p53 and ΔNp63 [23].The p53 family plays a crucial role in cell proliferationand differentiation during development and p53 activationpromotes cell growth arrest and apoptosis, while ΔNp63over-expression supports proliferation. It has been hypoth-esized that an imbalance between p53 and ΔNp63 may affectthe ability of red cell precursors to differentiate. Moreover,ΔNp63 has a crucial role in specification of nonneuralectoderm during gastrulation, thus likely contributing to thecraniofacial defects seen in morphants. The phenotype ofRPS19 deficient embryos is alleviated when the expressionof p53 and ΔNp63 is down-regulated [23].

Similar results have been observed when the expressionof RPL11 is knocked-down in zebrafish. Also the loss ofRPL11 leads to developmental defects associated with theactivation of p53 pathway [26].

Interestingly, the activation of p53 has been observed alsowhen the expression of other ribosomal proteins not directlyinvolved in the pathogenesis of DBA, like RPS8, RPS11,RPS18, and RPS6, is lost [23, 32], or when the expression ofother factors involved in ribosome biogenesis like Bop1 [33],bap28 [34], or TCOF is impaired [35]. These data suggestthat impairment of ribosome biogenesis in general leads tothe activation of p53 pathway.

Disruption of the nucleolus, the region where ribosomebiogenesis takes place, also leads to p53 activation [36].Therefore it has been hypothesized that ribosome biogenesisimpairment due to mutations in RPS19 could lead to nucleo-lar stress, p53 activation, erythroid failure, and malformationtypical of DBA.

2.5. Role of RPS19 during Erythropoiesis. How the hap-loinsufficiency of a ribosomal protein could lead to ery-throid impairment is still an open question. It has beenhypothesized that immature erythroid progenitors couldbe particularly sensitive to ribosomal proteins deficiencywith respect to other cell types because of their increasedproliferation rates. High rates of RNA synthesis whichexceeded the cell proliferation rates have been shown inprimary erythroid culture [37]. So these cells could be moresensitive to ribosome biogenesis impairment due to RPS19loss than other cell types. This also correlates with the highexpression of RPS19 in the early phases of erythropoiesis[10, 38].

Furthermore, Sieff and coworker showed for the firsttime that loss of RPS19 specifically affects erythropoiesisas the knock-down of RPS19 in primary murine fetalliver erythroid cells results in the downregulation of key


erythroid signaling proteins [37]. Loss of RPS19 decreasesthe expression of MYB, a transcription factor criticallyimportant for erythropoiesis, and of its transcriptional targetKIT. The downregulation of MYB has also been shown usingglobal gene expression analysis on bone marrow progenitorsfrom DBA patients [37].

So impairment of ribosome biogenesis during the earlystages of erythropoiesis leads to the deficiency of criticallyimportant erythroid transcription factors.

3. Diamond Blackfan Anemia as a PutativeDisorder of Heme Metabolism

3.1. Feline Leukemia Virus Subgroup C Receptor. FelineLeukemia Virus subgroup C Receptor (FLVCR1) is a hemeexporter belonging to the major facilitator superfamily(MFS) of transporters [39]. Overexpression of FLVCR1 inthe rat renal epithelial cell line NRK causes a slight decreasein intracellular heme concentration while the impairment ofFLVCR1 expression in feline embryonic fibroblasts leads toincreased intracellular heme levels. NRK cells overexpressingFLVCR1 show an increase in heme export measured usingboth the fluorescent heme analog zinc mesoporphyrin andFe59-hemin [7].

FLVCR1 is ubiquitously expressed [5]. High levels ofFLVCR1 expression have been detected in Caco-2 and HepG2cell lines, cancer cell lines, peripheral blood CD34+ stemcell progenitors, and hematopoietic cell lines with erythroidfeatures [7].

3.2. Role of FLVCR1 during Erythropoiesis. FLVCR1 has beeninitially identified as the receptor for Feline Leukemia Virussubgroup C (FeLV C) causing a severe erythroid aplasia incats similar to the anemia seen in DBA patients, suggestingthat FLVCR1 plays a fundamental role during erythropoiesis[40].

FLVCR1 expression is high in early erythroid precursorsand decreases during erythroid differentiation, similar toRPS19 expression. Cell lines with undifferentiated erythroidfeatures like K562 or HEL-DR express high levels of FLVCR1while it is expressed at low levels in HEL-D, a more maturecell line with spontaneous hemoglobinization. In the sameway, FLVCR1 is expressed at high levels in CFU-E frommobilized peripheral blood CD34+ stem progenitor cells andits expression decreases during erythroid differentiation invitro. Impairment of FLVCR1 function, both in K562 cellsand in lineage-depleted human umbilical cord blood cells,results in a decreased erythroid differentiation and increasedapoptosis [7]. In vivo, the loss of FLVCR1 in mice resultsin embryonic lethality due to the impairment of definitiveerythropoiesis. Also when FLVCR1 expression is deletedpostnatally, mice develop severe anemia [5].

3.3. Erythropoiesis and Heme Homeostasis. Heme is a com-plex of iron and protophorpyrin IX fundamental for cell biol-ogy. Heme is the prosthetic group of many essential proteinsincluding hemoglobin, myoglobin, and cytochromes. How-ever, high concentrations of intracellular free heme are toxic,

causing cell oxidative damage through lipid peroxidation. Sothe balance between heme biosynthesis, heme utilization forhemoprotein production, and heme catabolism is controlledat multiple levels [41].

Moreover, heme is not only a structural component ofhemoproteins but it is also involved in many intracellularpathways. Intracellular heme regulates both the transcriptionand translation of globin chains through interaction withthe transcriptional repressor Bach1 and the heme-regulatedeIF2α kinase (HRI), respectively (Figure 1(a)).

Heme positively regulates the transcription of globinmRNAs through binding to the transcriptional repressorBtb and Cnc Homology 1 (Bach1). Under basal condition,Bach1 heterodimerizes with small Maf transcription factorsand binds to MARE (Maf recognition elements) sequencesinhibiting the transcription of target mRNAs. Heme bind-ing to Bach1 decreases its affinity to small Maf proteinsand promotes Bach1 nuclear export; small Maf proteinsheterodimerize with Nuclear Factor Erythroid 2-like (Nrf2)proteins activating the transcription of target genes [42].So, when intracellular heme concentration increases, thetranscription of globin mRNAs increases.

HRI is a kinase able to phosphorylate the translationinitiation factor eIF2α in a heme-dependent manner. Phos-phorylation of eIF2 leads to inhibition of protein synthesis.During erythropoiesis when intracellular heme increases,heme binding to HRI keeps HRI in an inactive state, thusallowing protein synthesis. During this stage of erythroiddifferentiation, the main proteins to be synthesized are α-and β-globin chains and so hemoglobin is formed [43]. Theregulation of HRI activity by intracellular heme allows forcoordination of the synthesis of globin chains with hemeconcentration. Imbalance between heme and globin chainsis deleterious to red blood cells and their progenitors asintracellular free heme promotes oxidative cell damage andapoptosis. It has been hypothesized that FLVCR1 couldrepresent an additional control step in this process [7].Because FLVCR1 expression is high during the earlier stagesof erythropoiesis, it has been suggested that FLVCR1 exportsexcess heme out of erythroid progenitors when globinsynthesis is not yet initiated. Then, as erythroid differen-tiation proceeds, globin chains synthesis initiates, FLVCR1expression decreases, and intracellular heme concentrationincreases allowing hemoglobin formation (Figure 1(b)).According to this model, the loss of FLVCR1 expressionleads to increased intracellular free heme, cytotoxicity, andimpaired differentiation (Figure 1(c)).

3.4. Role of FLVCR1 during Development. In addition tothe erythroid defect, FLVCR1-null embryos display defectivegrowth and developmental anomalies resembling thosefound in DBA [5]: abnormal limb, hand, and digit matura-tion, flattened faces, and hypertelorism. These data suggestthat FLVCR1 expression is also fundamental for proper organdevelopment.

It has been observed that hemoglobin, heme, andbilirubin have an inhibitory effect on cartilage metabolismand growth in vitro although the molecular mechanism isnot clear [44–46]. FLVCR1 is expressed at high levels during


Hemoglobin

Globinschains

43S pre-initiationcomplex

40SMet-tRNA

P-eIF2aeIF2

Globins mRNAs

Mitochondria

HemeBiosynthesis

Heme

HRI

Active Inactive

TfR1

Tf-Fe

FLVCR

Bach1

Bach1

Bach1

NucleusM

af

Maf

Nrf2

Nrf2

Globins mRNAs

Globins mRNAs

Erythroid cell

(a)

Nucleusextrusion

Maturered blood cell

Lateerythroid progenitors

Earlyerythroid progenitors

Hematopoieticstem cell

FLVCR expression

RPS19 expression

FLVCR

Tf-Fe

HemeRibosomebiogenesis

Hemebiosynthesis

B

TfR1

Hemoglobin

(b)

Tf-FeTf-Fe

RPS19 deficiency

Hemebiosynthesis

Impaired

ribosomebiogenesis

Hemebiosynthesis

40S

RPS19

Hemeoverload

No proteinsynthesis

Oxidative damageand apoptosis

P53activation

Late erythroid progenitors

TfR1

Tf-FeTf-Fe

Oxidative damageand apoptosis

Early erythroid progenitors

Hemeoverload

TfR1

FLVCR deficiency

FLVCR

(c)

Figure 1: (a) Multiple roles of heme during erythroid cell differentiation. (b) Role of FLVCR1 during erythroid differentiation. (c) A modelto explain how both FLVCR1, and RPS19 deficiencies may result in the failure of erythroid differentiation.


development in the yolk sac and placenta [5]. It is possiblethat beyond a role in erythropoiesis, FLVCR1 could be alsoinvolved in maternal-fetus heme exchange. If so, we canhypothesize that the lack of FLVCR1 during developmentmay determine a heme overload condition in the embryoand, consequently, an inhibition of cartilage development,resulting in malformations.

3.5. Links between FLVCR1 and DBA. It is now clear thatDBA is a clinically heterogeneous disorder and no cleargenotype-phenotype correlation has been found. Congenitalmalformation is constitutional in most patients, but the typeand degree of severity of these anomalies is also differentamong DBA patients affected by the same mutation. In thesame way there is no correlation between genotype andcancer predisposition or response to therapies. Mutationsin ribosomal proteins account for about half of DBA cases,but the molecular genetics of the remaining cases is stillunknown, suggesting the existence of other DBA genesor unidentified modifiers that modulate the phenotype ofDBA.

Interestingly, FLVCR1-null mice display a phenotypevery close to DBA. Impairment of erythropoiesis occurs atthe same stage as that seen in DBA patients, and also thekind of malformations seen in FLVCR1-null mice are verysimilar to the congenital abnormalities described in DBApatients.

Human FLVCR1 has been mapped by radiation hybridmapping to chromosome 1q31.3. Apart from the transcrip-tion factor ATF3, no other genes have been mapped to thischromosome region. It is intriguing that rearrangement ofthe distal region on chromosome 1q has been described in apatient with DBA. Mutations in FLVCR1 have been searchedin a small group of multiplex DBA families with diseaselinkage to 1q31 but no mutations have been identified [8].

Alternatively spliced isoforms of FLVCR1 have beenidentified in immature bone marrow erythroid cells of someDBA patients negative for RPS19 gene mutations. Fourdifferent FLVCR1 isoforms result from the deletion of exon2 or exon 3 alone or in combination with exon 6 [47].

Splicing of FLVCR1 exon 3 or exon 6 causes in-frame mutations that encode potential proteins with majordeletions. Splicing of exon 2 causes a frame-shift mutationresulting in a premature stop codon and potentially encodesa truncated isoform of the protein.

In vitro, overexpression of deleted exon 3 or exon6 FLVCR1 isoforms leads to phenotypes that are weaklysusceptible to FeLV-C infection, indicating that FLVCR1function is impaired. These isoforms are expressed atlower levels than normal FLVCR1 and are mislocalized:while canonical FLVCR1 is expressed at the cell membrane,alternative spliced FLVCR1 isoforms localize intracellularly.Moreover, immature erythroid cells from DBA patients showa decreased expression of FLVCR1 mRNA and an enhance-ment of alternative spliced transcripts. It is interesting thatRPS19 interference in K562 cells results in an increaseof FLVCR1 alternative splicing, establishing the first linkbetween RPS19 and FLVCR1 in the pathogenesis of DBA[47].

4. Conclusions

DBA is a clinically heterogeneous disorder characterized byhypoplastic anemia associated with congenital malforma-tions, and cancer predisposition. The molecular mechanismsunderlying the pathogenesis of DBA are still not completelyunderstood. Mutations in RPS19 cause an impairment ofribosome biogenesis and protein synthesis but how this leadsto erythroid failure, congenital malformations and canceris unknown. It has been hypothesized that impairment ofribosome biogenesis leads to nucleolar disorganization andactivation of the p53 family.

It is intriguing to note the similarity between thephenotype of FLVCR1-null embryos and DBA. It has beenhypothesized that the loss of both FLVCR1 and RPS19could converge on the same pathway. Loss of FLVCR1causes increased intracellular heme levels during a stage oferythroid differentiation in which globin chains synthesisis not yet initiated. The haploinsufficiency of RPS19 leadsto impaired ribosome biogenesis and therefore proteinsynthesis. As globin chains are the major proteins synthesizedduring erythroid differentiation, there is an increase inintracellular free heme. In both cases, high intracellularheme concentration would result in the block of erythroiddifferentiation and apoptosis (Figure 1(c)). This hypothesisclearly explains the erythroid phenotype but not how RPS19or FLVCR1 deficiency leads to congenital malformations. Asmentioned earlier, heme is a potent inhibitor of cartilagegrowth and metabolism. Thus, even in this case it is possibleto speculate that heme overload resulting from FLVCR1 genefunction impairment and reduced protein synthesis due toribosomal protein genes mutations may generate the samephenotype, that is, skeletal malformation. Interestingly, it hasbeen reported that perichondrial zone expresses high levelsof genes involved in protein synthesis including RPL35a andRPS6, as well as the hemoglobin beta2 gene.

Finally, the work of Rey and colleagues suggested a directrelationship between RPS19 and FLVCR1 [47]. Interferencewith RPS19 results in a decreased expression of FLVCR1and in an increased level of alternative spliced isoforms ofFLVCR1, with compromised function. It could be speculatedthat RPS19 is able to directly or indirectly regulate theexpression of FLVCR1 but the molecular details of thisrelationship are yet to be understood. Moreover, nothing isknown about the proteins interacting with FLVCR1 and sowe cannot exclude the fact that RPS19 and FLVCR1 mayshare common downstream interactors.

The availability of cellular and animal models forstudying ribosomal proteins and FLVCR1 gene functionconstitutes the basis for future work aimed at elucidating themolecular pathogenetic mechanism of DBA.

Acknowledgments

The authors wish to thank Fiorella Altruda for helpful dis-cussion and Radhika Srinivasan for editing of the paper. Thiswork was supported by the Italian Ministry of University andResearch to E.T, by Regione Piemonte “Progetto Ricerca San-itaria Finalizzata 2008”, and by Telethon Grant GGP04181.


References

[1] J. Flygare and S. Karlsson, “Diamond-Blackfan anemia: ery-thropoiesis lost in translation,” Blood, vol. 109, no. 8, pp. 3152–3154, 2007.

[2] A. Vlachos, S. Ball, N. Dahl, et al., “Diagnosing and treatingDiamond Blackfan anaemia: results of an international clinicalconsensus conference,” British Journal of Haematology, vol.142, no. 6, pp. 859–876, 2008.

[3] J. M. Lipton, “Diamond Blackfan anemia: new paradigmsfor a “Not So Pure” inherited red cell aplasia,” Seminars inHematology, vol. 43, no. 3, pp. 167–177, 2006.

[4] I. Boria, P. Quarello, F. Avondo, et al., “A new database forribosomal protein genes which are mutated in Diamond-Blackfan anemia,” Human Mutation, vol. 29, no. 11, pp. E263–E270, 2008.


[6] C. S. Tailor, B. J. Willett, and D. Kabat, “A putative cellsurface receptor for anemia-inducing feline leukemia virussubgroup C is a member of a transporter superfamily,” Journalof Virology, vol. 73, no. 8, pp. 6500–6505, 1999.


[8] J. G. Quigley, H. Gazda, Z. Yang, S. Ball, C. A. Sieff, and J. L.Abkowitz, “Investigation of a putative role for FLVCR, a cyto-plasmic heme exporter, in Diamond-Blackfan anemia,” BloodCells, Molecules, and Diseases, vol. 35, no. 2, pp. 189–192, 2005.

[9] G. Lutsch, J. Stahl, H.-J. Kargel, F. Noll, and H. Bielka,“Immunoelectron microscopic studies on the location ofribosomal proteins on the surface of the 40S ribosomalsubunit from rat liver,” European Journal of Cell Biology, vol.51, no. 1, pp. 140–150, 1990.

[10] L. Da Costa, G. Narla, T.-N. Willig, et al., “Ribosomal proteinS19 expression during erythroid differentiation,” Blood, vol.101, no. 1, pp. 318–324, 2003.

[11] L. Da Costa, G. Tchernia, P. Gascard, et al., “Nucleolar local-ization of RPS19 protein in normal cells and mislocalizationdue to mutations in the nucleolar localization signals in 2Diamond-Blackfan anemia patients: potential insights intopathophysiology,” Blood, vol. 101, no. 12, pp. 5039–5045, 2003.

[12] M. F. Campagnoli, U. Ramenghi, M. Armiraglio, et al.,“RPS19 mutations in patients with Diamond-Blackfananemia,” Human Mutation, vol. 29, no. 7, pp. 911–920, 2008.

[13] J. Flygare, T. Kiefer, K. Miyake, et al., “Deficiency of ribosomalprotein S19 in CD34+ cells generated by siRNA blockserythroid development and mimics defects seen in Diamond-Blackfan anemia,” Blood, vol. 105, no. 12, pp. 4627–4634, 2005.

[14] K. Miyake, T. Utsugisawa, J. Flygare, et al., “Ribosomalprotein S19 deficiency leads to reduced proliferation andincreased apoptosis but does not affect terminal erythroiddifferentiation in a cell line model of Diamond-Blackfananemia,” Stem Cells, vol. 26, no. 2, pp. 323–329, 2008.

[15] K. Miyake, J. Flygare, T. Kiefer, et al., “Development ofcellular models for ribosomal protein S19 (RPS19)-deficientDiamond-Blackfan anemia using inducible expression ofsiRNA against RPS19,” Molecular Therapy, vol. 11, no. 4, pp.627–637, 2005.

[16] M. Kuramitsu, I. Hamaguchi, M. Takuo, et al., “DeficientRPS19 protein production induces cell cycle arrest inerythroid progenitor cells,” British Journal of Haematology,vol. 140, no. 3, pp. 348–359, 2008.

[17] J. Flygare, K. Olsson, J. Richter, and S. Karlsson, “Genetherapy of Diamond Blackfan anemia CD34(+) cells leads toimproved erythroid development and engraftment followingtransplantation,” Experimental Hematology, vol. 36, no. 11,pp. 1428–1435, 2008.

[18] I. Hamaguchi, J. Flygare, H. Nishiura, et al., “Proliferationdeficiency of multipotent hematopoietic progenitors inribosomal protein S19 (RPS19)—deficient Diamond-Blackfan anemia improves following RPS19 gene transfer,”Molecular Therapy, vol. 7, no. 5 I, pp. 613–622, 2003.

[19] I. Hamaguchi, A. Ooka, A. Brun, J. Richter, N. Dahl, and S.Karlsson, “Gene transfer improves erythroid development inribosomal protein S19-deficient Diamond-Blackfan anemia,”Blood, vol. 100, no. 8, pp. 2724–2731, 2002.

[20] J. Badhai, A.-S. Frojmark, E. J. Davey, J. Schuster, and N. Dahl,“Ribosomal protein S19 and S24 insufficiency cause distinctcell cycle defects in Diamond-Blackfan anemia,” Biochimica etBiophysica Acta, vol. 1792, no. 10, pp. 1036–1042, 2009.

[21] H. Matsson, E. J. Davey, N. Draptchinskaia, et al., “Targeteddisruption of the ribosomal protein S19 gene is lethal prior toimplantation,” Molecular and Cellular Biology, vol. 24, no. 9,pp. 4032–4037, 2004.

[22] H. Matsson, E. J. Davey, A. S. Frojmark, et al., “Erythropoiesisin the Rps19 disrupted mouse: analysis of erythropoietinresponse and biochemical markers for Diamond-Blackfananemia,” Blood Cells, Molecules, and Diseases, vol. 36, no. 2,pp. 259–264, 2006.

[23] N. Danilova, K. M. Sakamoto, and S. Lin, “Ribosomal proteinS19 deficiency in zebrafish leads to developmental abnormal-ities and defective erythropoiesis through activation of p53protein family,” Blood, vol. 112, no. 13, pp. 5228–5237, 2008.

[24] T. Uechi, Y. Nakajima, A. Chakraborty, H. Torihara, S. Higa,and N. Kenmochi, “Deficiency of ribosomal protein S19during early embryogenesis leads to reduction of erythrocytesin a zebrafish model of Diamond-Blackfan anemia,” HumanMolecular Genetics, vol. 17, no. 20, pp. 3204–3211, 2008.

[25] T. Uechi, Y. Nakajima, A. Nakao, et al., “Ribosomal proteingene knockdown causes developmental defects in zebrafish,”PLoS ONE, vol. 1, no. 1, article e37, 2006.

[26] A. Chakraborty, T. Uechi, S. Higa, H. Torihara, and N.Kenmochi, “Loss of ribosomal protein L11 affects zebrafishembryonic development through a p53-dependent apoptoticresponse,” PLoS ONE, vol. 4, no. 1, article e4152, 2009.

[27] I. Zemp and U. Kutay, “Nuclear export and cytoplasmicmaturation of ribosomal subunits,” FEBS Letters, vol. 581, no.15, pp. 2783–2793, 2007.

[28] S. Orru, A. Aspesi, M. Armiraglio, et al., “Analysis of theribosomal protein S19 interactome,” Molecular and CellularProteomics, vol. 6, no. 3, pp. 382–393, 2007.

[29] R. A. Idol, S. Robledo, H.-Y. Du, et al., “Cells depleted forRPS19, a protein associated with Diamond Blackfan anemia,show defects in 18S ribosomal RNA synthesis and smallribosomal subunit production,” Blood Cells, Molecules, andDiseases, vol. 39, no. 1, pp. 35–43, 2007.

[30] V. Choesmel, S. Fribourg, A.-H. Aguissa-Toure, et al.,“Mutation of ribosomal protein RPS24 in Diamond-Blackfananemia results in a ribosome biogenesis disorder,” HumanMolecular Genetics, vol. 17, no. 9, pp. 1253–1263, 2008.

[31] J. Badhai, A.-S. Frojmark, H. R. Razzaghian, E. Davey, J.Schuster, and N. Dahl, “Posttranscriptional down-regulationof small ribosomal subunit proteins correlates with reductionof 18S rRNA in RPS19 deficiency,” FEBS Letters, vol. 583, no.12, pp. 2049–2053, 2009.


[32] L. Panic, J. Montagne, M. Cokaric, and S. Volarevic, “S6-haploinsufficiency activates the p53 tumor suppressor,” CellCycle, vol. 6, no. 1, pp. 20–24, 2007.

[33] D. G. Pestov, Z. Strezoska, and L. F. Lau, “Evidence of p53-dependent cross-talk between ribosome biogenesis and the cellcycle: effects of nucleolar protein Bop1 on G(1)/S transition,”Molecular Cell Biology, vol. 21, pp. 4246–4255, 2001.

[34] M. Azuma, R. Toyama, E. Laver, and I. B. Dawid,“Perturbation of rRNA synthesis in the bap28 mutationleads to apoptosis mediated by p53 in the zebrafish centralnervous system,” The Journal of Biological Chemistry, vol. 281,no. 19, pp. 13309–13316, 2006.

[35] N. C. Jones, M. L. Lynn, K. Gaudenz, et al., “Prevention ofthe neurocristopathy Treacher Collins syndrome throughinhibition of p53 function,” Nature Medicine, vol. 14, no. 2,pp. 125–133, 2008.

[36] C. P. Rubbi and J. Milner, “Disruption of the nucleolus medi-ates stabilization of p53 in response to DNA damage and otherstresses,” EMBO Journal, vol. 22, no. 22, pp. 6068–6077, 2003.

[37] C. A. Sieff, J. Yang, L.B. Merida-Long, and H. F. Lodish,“Pathogenesis of the erythroid failure in Diamond Blackfananaemia,” British Journal of Haematology, vol. 148, no. 4, pp.611–622, 2010.

[38] B. L. Ebert, M. M. Lee, J. L. Pretz, et al., “An RNA interferencemodel of RPS19 deficiency in Diamond-Blackfan anemiarecapitulates defective hematopoiesis and rescue by dexam-ethasone: identification of dexamethasone-responsive genesby microarray,” Blood, vol. 105, no. 12, pp. 4620–4626, 2005.

[39] S. S. Pao, I. T. Paulsen, and M. H. Saier Jr., “Major facilitatorsuperfamily,” Microbiology and Molecular Biology Reviews, vol.62, no. 1, pp. 1–34, 1998.

[40] J. G. Quigley, C. C. Burns, M. M. Anderson, et al., “Cloningof the cellular receptor for feline leukemia virus subgroup C(FeLV-C), a retrovirus that induces red cell aplasia,” Blood,vol. 95, no. 3, pp. 1093–1099, 2000.

[41] E. Tolosano, S. Fagoonee, N. Morello, F. Vinchi, and V. Fiorito,“Heme scavenging and the other facets of hemopexin,” Antiox-idants & Redox Signaling, vol. 12, no. 2, pp. 305–320, 2010.

[42] K. Igarashi and J. Sun, “The heme-Bach1 pathway inthe regulation of oxidative stress response and erythroiddifferentiation,” Antioxidants & Redox Signaling, vol. 8, no.1-2, pp. 107–118, 2006.

[43] J.-J. Chen and I. M. London, “Regulation of protein synthesisby heme-regulated eIF-2α kinase,” Trends in BiochemicalSciences, vol. 20, no. 3, pp. 105–108, 1995.

[44] R. Vassilopoulou-Sellin, C. O. Oyedeji, P. L. Foster, M. M.Thompson, and N. A. Saman, “Hemoglobin as a directinhibitor of cartilage growth in vitro,” Hormone and MetabolicResearch, vol. 21, no. 1, pp. 11–14, 1989.

[45] R. Vassilopoulou-Sellin, C. O. Oyedeji, and N. A. Samaan,“Bilirubin inhibits cartilage metabolism and growth in vitro,”Metabolism, vol. 38, no. 8, pp. 759–762, 1989.

[46] R. Vassilopoulou-Sellin, M. M. Thompson, C. O. Oyedeji,and N. A. Samaan, “Heme inhibits cartilage metabolism andgrowth in vitro,” Metabolism, vol. 38, no. 1, pp. 52–56, 1989.

[47] M. A. Rey, S. P. Duffy, J. K. Brown, et al., “Enhanced alternativesplicing of the FLVCR1 gene in Diamond Blackfan anemiadisrupts FLVCR1 expression and function that are critical forerythropoiesis,” Haematologica, vol. 93, no. 11, pp. 1617–1626,2008.


Review Article

Unexplained Aspects of Anemia of Inflammation

Elizabeth A. Price1, 2 and Stanley L. Schrier1

1 Department of Medicine (Hematology), Stanford University, Stanford, CA 94305-3434, USA2 Division of Hematology, Stanford Comprehensive Cancer Center, 875 Blake Wilbur Drive, Stanford, CA 94305, USA

Correspondence should be addressed to Elizabeth A. Price, [email protected]

Received 24 October 2009; Accepted 16 February 2010


Copyright © 2010 E. A. Price and S. L. Schrier. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Anemia of inflammation (AI), also known as anemia of chronic inflammation or anemia of chronic disease was described over50 years ago as anemia in association with clinically overt inflammatory disease, and the findings of low plasma iron, decreasedbone marrow sideroblasts and increased reticuloendothelial iron. Pathogenic features underlying AI include a mild shorteningof red cell survival, impaired erythropoietin production, blunted responsiveness of the marrow to erythropoietin, and impairediron metabolism mediated by inflammatory cytokines and the iron regulatory peptide, hepcidin. Despite marked recent advancesin understanding AI, gaps remain, including understanding of the pathogenesis of AI associated with “noninflammatory” ormildly inflammatory diseases, the challenge of excluding iron deficiency anemia in the context of concomitant inflammation, andunderstanding more precisely the contributory role of hepcidin in the development of AI in human inflammatory diseases.

1. Introduction

Anemia of inflammation (AI), also known as anemiaof chronic inflammation or anemia of chronic disease,was described over 50 years ago as a normo to micro-cytic anemia, typically of mild severity, characterized bio-chemically by a low plasma iron, decreased total iron-binding capacity, decreased transferrin saturation, and,on bone marrow examination, decreased sideroblasts andincreased reticuloendothelial iron [1]. Despite the hetero-geneity of underlying infectious, malignant and inflam-matory diseases described in conjunction with this typeof anemia, there was a remarkable degree of similar-ity in terms of the severity of the anemia, correlatingdirectly with the degree of inflammation present, and itsstability over time, features which remain characteristicof AI today [1]. Despite advances in understanding theunderlying mechanisms leading to AI, several key clinicalissues remain, including understanding of mechanismsof AI in “noninflammatory” diseases, optimal methodsfor distinguishing AI from iron deficiency anemia, andunderstanding the contributory role of various pathologicmechanisms in individual human diseases that lead toAI.

2. Pathophysiology of AI

Pathogenic features underlying AI include a mild shorteningof red cell survival, impaired erythropoietin production,blunted responsiveness of the marrow to erythropoietin,and impaired iron metabolism [2]. Inflammatory cytokines,including interleukin-1 (IL-1), tumor necrosis factor-α(TNF-α), and interferon-γ (IFN-γ), affect the differentiationand proliferation of erythroid progenitor colony formationin vitro [2–4]; the proapoptotic effects of which have beenreversed by treatment with the TNF-inhibitor infliximabin patients with rheumatoid arthritis [5, 6]. Cytokinesand acute-phase reactants may also cause sequestration ofiron in reticulo-endothelial cells (reviewed in [4]); how-ever, hepcidin, a key regulator of iron homeostasis, hasrecently emerged as a central player in the disordered ironmetabolism seen in AI. The hepcidin gene is located onchromosome 19 and encodes an 84 aa prepropeptide whichundergoes cleavage to produce first the circulating 60 aaprohepcidin, followed by cleavage to produce the 25 aa formof hepcidin [7]. This 25 aa peptide acts via binding to theonly known mammalian cellular iron exporter, ferroportin,leading to its internalization and degradation, with conse-quent loss of iron export from the duodenal enterocytes


and reticulo-endothelial cells [8]. Hepcidin synthesis issuppressed by erythropoiesis [9] and iron deficiency andupregulated by iron overload and inflammation. Infusionof IL-6 [10] or lipopolysaccharide (LPS) [11] into healthyvolunteers leads to a rapid increase in urinary hepcidin, withconcomitant decrease in serum iron levels. IL-6 upregulatessynthesis of hepcidin via signal transducer and activatorof transcription-3 (STAT-3) signaling [12]. Hepcidin levelshave been found to be elevated in patients with severeinflammation when compared to healthy controls [13]. Ahepcidin transgenic mouse model replicates many of thepathogenic features of AI, including a mild, hypochromicanemia, sequestration of iron in the spleen, and impairedmarrow response to erythropoietin [14]. Thus, in vitro andin animal and human models, hepcidin has been shown toeffect most if not all of the iron-related changes which arepathognomic features of AI.

3. AI in “Noninflammatory” States

Diagnostic features of AI in the appropriate clinical contextinclude characteristic iron indices associated with reducedbone marrow sideroblasts and increased reticuloendothelialiron [1]. In current practice, the diagnostic criteria of AIare frequently modified to require only the presence ofanemia, the absence of overt other etiologies, and associatedperipheral blood findings of disordered iron metabolism,such as a low serum iron with normal or elevated serumferritin [15]. This has led to widening of the spectrum ofclinical states associated with AI, including aging, diseasesassociated with aging such as heart failure, and hospitaliza-tion [16]. It is important to note, however, that there arecurrently no uniform peripheral blood laboratory criteria forthe diagnosis of AI.

3.1. Anemia of the Elderly. Abnormal iron indices suggestiveof AI are relatively common findings in elderly patients. Inthe U.S. National NHANES III study, 20% of elderly adultswith anemia were characterized as having AI as the sole causeof their anemia, defined as a serum iron <60 mcg/dL andno evidence of iron deficiency. Iron deficiency was in turndefined as having at least two of the following: transferrinsaturation less than 15%, serum ferritin less than 12 ng/mL,or erythrocyte protoporphyrin concentration greater than1.24 uM [17]. Aging is also associated with increased inflam-mation and a mild rise in inflammatory markers, includingIL-6 [18, 19], and elderly patients with anemia have beenfound to have higher levels of inflammatory markers thanelderly patients without anemia [20]. However, the degree ofinflammation present is mild; in one study the differences inmedian levels of IL-6 and C-reactive protein (CRP) betweenanemic and nonanemic elderly patients were relatively small(2.5 versus 1.4 pg/mL and 4.9 versus 2.7 mg/L, resp.) [21].

Can AI-type pathophysiology exist in the absence ofassociated clinically evident inflammatory disease? Thisquestion has been studied in elderly patients. In an earlyreport, 5 elderly patients without underlying inflammatorydisease presented with a microcytic, hypochromic anemia,

low serum iron levels, and increased bone marrow iron [22].In ferrokinetic studies, these patients were found to have lowiron absorption, normal iron utilization (incorporation ofinfused iron into red cells), and decreased reutilization ofiron (incorporation of iron from infused hemoglobin intored cells). These findings of low serum iron, low iron bindingcapacity, high serum ferritin, and increased bone marrowiron in elderly patients in the absence of malignant orinflammatory diseases were termed “primary defective iron-reutilization syndrome” [23]. Interestingly, treatment withtestosterone enanthate led to improvements in hemoglobin,red cell mass, and iron reutilization. In a follow-up study, twosimilar patients were described [23]. Treatment with danazolin both patients resulted in an increases in hemoglobin,serum iron, percent transferrin saturation, and decreasedferritin, suggesting that treatment led to release of previ-ously sequestered reticulo-endothelial iron. These results areintriguing and suggest both that disordered iron metabolismcan develop in the absence of overt inflammatory disease,and also that androgen therapy, at least in this smallpatient sample, corrected the underlying defect. Androgenshave been shown to increase erythropoietin levels [24] andhematopoietic stem cell cycling [25]; however it seems morelikely that an unrelated mechanism led to improvement inthe anemia and iron homeostasis in these cases. It would alsobe interesting to assess whether milder biochemical evidenceof disordered iron metabolism, that is, a mild normocyticnormochromic anemia with less dramatically abnormal ironindices, had similar ferrokinetic findings. It is possible thatthe use of more sensitive markers of inflammation such asthe high sensitivity CRP would isolate a subset of patientswithout overt inflammatory disease who nonetheless hadfindings consistent with AI.

3.2. Anemia of Heart Failure. AI as defined by iron indicesis also common in patients with congestive heart failure.In one study, 85 of 148 (57%) anemic outpatients withcongestive heart failure had AI [26]. Those with AI hadhigher levels of inflammatory markers, including IL-6 andTNF-α, compared to levels found in previously reportedhealthy subjects, supporting the diagnostic criteria. However,in another study in anemic hospitalized patients with moreadvanced heart failure, 27 of 37 (73%) had iron deficiencyas diagnosed by the gold standard of absent iron on bonemarrow aspirate, despite the group having a mean ferritinof 75 ± 59.1 ng/mL [27]. Thus, using a serum ferritin cutofflevel of >50 ng/mL may be too restrictive in excluding irondeficiency anemia in patients with heart failure. The findingsfrom this study highlight the dilemma of distinguishingiron deficiency anemia from AI, which remains a frequentclinical challenge despite the availability of tests such as thesoluble transferrin receptor (sTfR), a marker of increasederythropoiesis and iron deficiency [28].

3.3. Identifying Iron Deficiency in Patients with AI. Recentwork suggests that measuring hepcidin levels may havefuture utility in distinguishing the two disorders. In a ratmodel of chronic inflammation and iron deficiency anemia,


hepatic hepcidin mRNA was suppressed in controls withiron deficiency alone, elevated in rats with inflammationalone, and suppressed in rats with combined inflammationand iron deficiency [29]. Serum hepcidin levels followeda similar but less distinct pattern. Rats with inflammationhad significantly less uptake of orally administered ironcompared with both control rats and rats with combinedinflammation and iron deficiency. In the same study, humanpatients were designated as having anemia of inflammationin combination with iron deficiency if they had anemia, anassociated inflammatory disease, and a transferrin satura-tion <16%, serum ferritin of <100 mcg/mL and sTfR/logferritin >2. Serum hepcidin levels were significantly higherin patients with inflammation alone than in those witheither iron deficiency alone or combined iron deficiencyand inflammation. Consistent with these results, in humanduodenal biopsy samples, mRNA expression of ferroportinwas increased in those with combined inflammation and irondeficiency compared to those with inflammation alone. Inanother study utilizing a mouse model of sepsis and pro-longed organ dysfunction, inflammation-induced increasein hepatic hepcidin mRNA was suppressed by phlebotomy,and, to a lesser extent, by erythropoietin administration [30].These results suggest that severe iron deficiency exerts adominant effect over inflammation with regard to regulationof hepcidin in vivo, and furthermore, that hepcidin levels,if validated for clinical use, may be of utility in confirmingor excluding iron deficiency in the presence of concomitantinflammation.

3.4. Anemia of Chronic Kidney Disease (CKD). Anemiaof CKD is another “noninflammatory” disease in whichinflammation may actually play a prominent role in thedevelopment of anemia. The pathogenesis of anemia in CKDis multifactorial and predominately driven by impaired ery-thropoietin production; however inflammation negativelyimpacts erythropoiesis and is associated with resistance toerythropoietic therapy [31, 32]. Several recent studies havemeasured hepcidin levels in patients with CKD in an effort toclarify the contributory role of hepcidin and disordered ironmetabolism in anemia in this disease. Ashby et al. measuredplasma hepcidin (by immunoassay) in healthy controls andpatients with chronic kidney disease both off (44 patients)and on (94 patients) hemodialysis [33]. Although hepcidinlevels did not correlate strongly with ferritin levels inpatients on hemodialysis (felt by the authors to be relatedto intravenous iron supplementation in this group), in thosenot on hemodialysis, hepcidin levels did correlate stronglywith ferritin levels, consistent with upregulation of hepcidinin relation to increased iron. In addition, hepcidin levelswere negatively correlated with erythropoietin dose, and fellslightly in patients who were initiated on erythropoietintherapy. There was no correlation between hepcidin levelsand markers of inflammation (CRP or IL-6) in eithergroup. Interestingly, there was a negative correlation betweenhepcidin and worsening renal function. Plasma growthdifferentiation factor-15 (GDF15) levels, which are elevatedin thalassemia major and inhibit hepdcidin synthesis [34],

did not appear to negatively regulate hepcidin in this setting,consistent with results in normal volunteers treated withlow-dose recombinant human erythropoietin [35]. Weisset al. measured serum hepcidin as well as iron indices in20 patients on hemodialysis and found that hepcidin levelscorrelated positively with serum iron, ferritin, transferrinsaturation, and negatively with reticulocyte count, but notwith markers of inflammation, including serum IL-6, trans-forming growth factor-β (TGF-β), or CRP [36]. However,the patients had relatively low levels of inflammation, withmean CRP levels of only 0.7 mg/dL. Similarly, Tomosugi et al.measured hepcidin levels using a semiquantitative surface-enhanced laser desorption ionization time of flight massspectrometry (SELDI-TOF MS) assay and found that whileserum hepcidin-25 levels correlated well with IL-6 levelsin patients with acute infection, there was no correlationbetween hepcidin and IL-6 in patients receiving dialysiswithout additional associated inflammatory disease [37].Zaritsky et al. found in adults with chronic kidney diseasestages 2 to 4, ferritin, soluble transferrin receptor levels andglomerular filtration rate (GFR) were directly correlated withserum hepcidin levels. When pediatric patients with CKDstages 2–5 were included in a multivariate analysis, highresolution CRP also correlated with serum hepcidin levels[38]. Thus, these studies, taken together, suggest that inpatients with CKD without additional inflammatory disease,hepcidin is regulated predominately by iron status, and to alesser extent by erythropoiesis, and even less by inflammationand thus may not play a prominent contributory role in theanemia associated with CKD.

4. Hepcidin Levels inInflammatory Bowel Disease

Measurements of hepcidin levels in inflammatory bowel dis-ease (IBD) have also yielded interesting results. In one study,six pediatric patients with active Crohn’s disease (definedas an IL-6 level >5 pg/mL) had poor absorption of a singledose of oral iron and high urinary hepcidin levels (mean295.3 ng/mg creatinine) compared to ten patients withrelatively inactive Crohn’s disease (with IL-6 levels ≤5 pg/mLand mean urinary hepcidin levels 31.22 ng/mg creatinine).Thus, patients with active disease had a markedly higherhepcidin levels than those with quiescent disease, consistentwith the hepcidin model of AI. Interestingly, hemoglobinlevels were similar in both groups (mean 11.7 and 11.6 g/dLin active and inactive disease, resp.). Although this waspossibly due to the small number of patients studied, theabsence of a difference in hemoglobin level despite markedlydifferent hepcidin levels also raises the possibility that ele-vated hepcidin does not drive the anemia in this disease, evenwhen active. Surprisingly, in another study in 61 patientswith inflammatory bowel disease (mean age 44 years), serumhepcidin levels, as measured by radioimmunoassay, weresignificantly lower in those with IBD compared to healthycontrols. In this study, iron deficiency anemia was definedas the presence of a low serum iron (<10.6 umol/L) and lowserum ferritin (<15 ug/L). AI was defined as anemia with low


serum iron and normal ferritin level. Although the directlycontrasting results between these two studies may potentiallybe explained by differences in urine versus serum hepcidinassays, an alternative explanation is that patients with IBDwith a serum ferritin above 15 may still be iron deficientand have suppressed hepcidin synthesis. Regardless, thesecontrasting results highlight the need for further studies tomore accurately delineate the contributory role of hepcidinin the development of AI in IBD.

In summary, despite marked recent advances in under-standing AI, gaps remain, including understanding thepathogenesis of AI associated with “noninflammatory” ormildly inflammatory diseases, the challenge of excluding irondeficiency anemia in the context of concomitant inflam-mation, and understanding more precisely the contributoryrole of hepcidin in the development of AI in humaninflammatory diseases.

Acknowledgment

This work was supported by NIH Grant 5 R01 AG029124-03.

References

[1] G. E. Cartwright, “The anemia of chronic disorders,” Seminarsin Hematology, vol. 3, no. 4, pp. 351–375, 1966.

[2] R. T. Means Jr., “Advances in the anemia of chronic disease,”International Journal of Hematology, vol. 70, no. 1, pp. 7–12,1999.

[3] G. Weiss and L. T. Goodnough, “Anemia of chronic disease,”The New England Journal of Medicine, vol. 352, no. 10, pp.1011–1023, 2005.

[4] G. Weiss, “Iron metabolism in the anemia of chronic disease,”Biochimica et Biophysica Acta, vol. 1790, no. 7, pp. 682–693,2009.

[5] H. A. Papadaki, H. D. Kritikos, V. Valatas, D. T. Boumpas, andG. D. Eliopoulos, “Anemia of chronic disease in rheumatoidarthritis is associated with increased apoptosis of bone marrowerythroid cells: improvement following anti-tumor necrosisfactor-α antibody therapy,” Blood, vol. 100, no. 2, pp. 474–482,2002.

[6] H. A. Papadaki, H. D. Kritikos, C. Gemetzi, et al., “Bonemarrow progenitor cell reserve and function and stromal cellfunction are defective in rheumatoid arthritis: evidence for atumor necrosis factor alpha-mediated effect,” Blood, vol. 99,no. 5, pp. 1610–1619, 2002.

[7] P. L. Lee and E. Beutler, “Regulation of hepcidin and iron-overload disease,” Annual Review of Pathology, vol. 4, pp. 489–515, 2009.




[11] E. Kemna, P. Pickkers, E. Nemeth, H. van der Hoeven, and D.Swinkels, “Time-course analysis of hepcidin, serum iron, andplasma cytokine levels in humans injected with LPS,” Blood,vol. 106, no. 5, pp. 1864–1866, 2005.

[12] D. M. Wrighting and N. C. Andrews, “Interleukin-6 induceshepcidin expression through STAT3,” Blood, vol. 108, no. 9,pp. 3204–3209, 2006.


[14] C. N. Roy, H. H. Mak, I. Akpan, G. Losyev, D. Zurakowski, andN. C. Andrews, “Hepcidin antimicrobial peptide transgenicmice exhibit features of the anemia of inflammation,” Blood,vol. 109, no. 9, pp. 4038–4044, 2007.

[15] P. Agnihotri, M. Telfer, Z. Butt, et al., “Chronic anemia andfatigue in elderly patients: results of a randomized, double-blind, placebo-controlled, crossover exploratory study withepoetin alfa,” Journal of the American Geriatrics Society, vol.55, no. 10, pp. 1557–1565, 2007.

[16] J. M. Cash and D. A. Sears, “The anemia of chronic disease:spectrum of associated diseases in a series of unselectedhospitalized patients,” American Journal of Medicine, vol. 87,no. 6, pp. 638–644, 1989.

[17] J. M. Guralnik, R. S. Eisenstaedt, L. Ferrucci, H. G. Klein, andR. C. Woodman, “Prevalence of anemia in persons 65 yearsand older in the United States: evidence for a high rate ofunexplained anemia,” Blood, vol. 104, no. 8, pp. 2263–2268,2004.

[18] J. Wei, H. Xu, J. L. Davies, and G. P. Hemmings, “Increase ofplasma IL-6 concentration with age in healthy subjects,” LifeSciences, vol. 51, no. 25, pp. 1953–1956, 1992.

[19] M. Maggio, J. M. Guralnik, D. L. Longo, and L. Ferrucci,“Interleukin-6 in aging and chronic disease: a magnificentpathway,” Journals of Gerontology A, vol. 61, no. 6, pp. 575–584, 2006.

[20] L. Ferrucci, J. M. Guralnik, R. C. Woodman, et al., “Proinflam-matory state and circulating erythropoietin in persons withand without anemia,” American Journal of Medicine, vol. 118,no. 11, pp. 1288.e11–1288.e19, 2005.

[21] L. Ferrucci, J. M. Guralnik, S. Bandinelli, et al., “Unexplainedanaemia in older persons is characterised by low erythro-poietin and low levels of pro-inflammatory markers,” BritishJournal of Haematology, vol. 136, no. 6, pp. 849–855, 2007.

[22] F. I. Haurani and D. Green, “Primary defective iron reutiliza-tion. Response to testosterone therapy,” The American Journalof Medicine, vol. 42, no. 1, pp. 151–158, 1967.

[23] E. C. Besa, P. W. Kim, and F. I. Haurani, “Treatment of primarydefective iron-reutilization syndrome: revisited,” Annals ofHematology, vol. 79, no. 8, pp. 465–468, 2000.

[24] R. Alexanian, “Erythropoietin and erythropoiesis in anemicman following androgens,” Blood, vol. 33, no. 4, pp. 564–572,1969.

[25] J. W. Byron, “Effect of steroids on the cycling of haemopoieticstem cells,” Nature, vol. 228, no. 5277, p. 1204, 1970.

[26] C. Opasich, M. Cazzola, L. Scelsi, et al., “Blunted erythropoi-etin production and defective iron supply for erythropoiesisas major causes of anaemia in patients with chronic heartfailure,” European Heart Journal, vol. 26, no. 21, pp. 2232–2237, 2005.

[27] J. N. Nanas, C. Matsouka, D. Karageorgopoulos, et al.,“Etiology of anemia in patients with advanced heart failure,”Journal of the American College of Cardiology, vol. 48, no. 12,pp. 2485–2489, 2006.


[28] C. Brugnara, “Iron deficiency and erythropoiesis: new diag-nostic approaches,” Clinical Chemistry, vol. 49, no. 10, pp.1573–1578, 2003.

[29] I. Theurl, E. Aigner, M. Theurl, et al., “Regulation of ironhomeostasis in anemia of chronic disease and iron deficiencyanemia: diagnostic and therapeutic implications,” Blood, vol.113, no. 21, pp. 5277–5286, 2009.

[30] S. Lasocki, S. Millot, V. Andrieu, et al., “Phlebotomies orerythropoietin injections allow mobilization of iron stores in amouse model mimicking intensive care anemia,” Critical CareMedicine, vol. 36, no. 8, pp. 2388–2394, 2008.

[31] P. Barany, J. C. Divino Filho, and J. Bergstrom, “HighC-reactive protein is a strong predictor of resistance toerythropoietin in hemodialysis patients,” American Journal ofKidney Diseases, vol. 29, no. 4, pp. 565–568, 1997.

[32] A. C. Cooper, A. Mikhail, M. W. Lethbridge, D. M. Kemeny,and I. C. MacDougall, “Increased expression of erythropoiesisinhibiting cytokines (IFN-gamma, TNF-alpha, IL-10, and IL-13) by T cells in patients exhibiting a poor response toerythropoietin therapy,” Journal of the American Society ofNephrology, vol. 14, no. 7, pp. 1776–1784, 2003.

[33] D. R. Ashby, D. P. Gale, M. Busbridge, et al., “Plasma hepcidinlevels are elevated but responsive to erythropoietin therapy inrenal disease,” Kidney International, vol. 75, no. 9, pp. 976–981,2009.


[35] P. Robach, S. Recalcati, D. Girelli, et al., “Alterations ofsystemic and muscle iron metabolism in human subjectstreated with low-dose recombinant erythropoietin,” Blood,vol. 113, no. 26, pp. 6707–6715, 2009.

[36] G. Weiss, I. Theurl, S. Eder, et al., “Serum hepcidin concentra-tion in chronic haemodialysis patients: associations and effectsof dialysis, iron and erythropoietin therapy,” European Journalof Clinical Investigation, vol. 39, no. 10, pp. 883–890, 2009.

[37] N. Tomosugi, H. Kawabata, R. Wakatabe, et al., “Detection ofserum hepcidin in renal failure and inflammation by usingProteinChip System,” Blood, vol. 108, no. 4, pp. 1381–1387,2006.

[38] J. Zaritsky, B. Young, H.-J. Wang, et al., “Hepcidin—apotential novel biomarker for iron status in chronic kidneydisease,” Clinical Journal of the American Society of Nephrology,vol. 4, no. 6, pp. 1051–1056, 2009.

downloads.hindawi.comdownloads.hindawi.com/journals/specialissues/852308.pdf · 2019-08-07 ·...

Documents

Transcript of downloads.hindawi.comdownloads.hindawi.com/journals/specialissues/852308.pdf · 2019-08-07 ·...