Building bridges with bio-iron: From bench to bedside


001

MACROPHAGE-HFE IS CRITICAL FOR LIPOPOLYSACCAHRIDE (LPS)-CONTROLLED HEPATIC HEPCIDIN ACTIVATION

Maja Vujic Spasic 1,2, Richard Sparla1,2, Jens Stolte3, Matthias W. Hentze2,3, Martina U. Muckenthaler1,2

1University Hospital of Heidelberg, Germany, 2Molecular Medicine Partnership Unit, 3European Molecular Biology Laboratory, Heidelberg, Germany

Systemic iron homeostasis is disrupted in the common, potentially fatal iron overload disorder hereditary hemochromatosis (HH). Mutations in the HFE/Hfe gene cause the most prevalent form of HH, hallmarked by inadequate expression of the Hepcidin, the key regulator of systemic iron homeostasis. To identify where Hfe acts to prevent HH, mice with tissue-specific Hfe ablation in hepatocytes, enterocytes and macrophages were generated. We uncovered that Hfe acts in hepatocytes to control systemic iron levels and Hepcidin expression, classifying genetic hemochromatosis as a liver disease.

While inappropriately low Hepcidin levels contribute to iron overload, increased Hepcidin expression plays an important role in the anaemia of inflammation (AI). Inflammation influences iron balance in the whole organism by reducing duodenal iron absorption and increasing macrophage iron retention resulting in low serum iron concentrations. Cytokine-mediated Hepcidin stimulation plays a critical role in this process. In Hfe-deficient mice the inflammatory response of Hepcidin is deregulated: while Hfe-/- mice mount a general inflammatory response following the injection of LPS (5μg) they fail to appropriately elevate Hepcidin mRNA expression and reduce serum iron levels.

It is however unclear how Hfe contributes to LPS-mediated Hepcidin induction in the liver and whether Hfe expression in hepatocytes, and/or liver macrophages, and/or other cell types is required. To answer this question, we injected constitutive, hepatocyte- and macrophage-specific Hfe mutant mice with LPS (5μg) for 8h. While LPS injections cause significant Hepcidin induction in Wt and hepatocyte-specific Hfe mutant mice, Hepcidin mRNA expression fails to be increased in macrophage-specific Hfe mutant mice.

This work demonstrates that Hfe expression in hepatocytes serves to maintain physiological Hepcidin expression and cellular/systemic iron homeostasis. By contrast, Hfe expression in macrophages is indispensable for the LPS-stimulated Hepcidin response. This study for the first time provides evidence for a role of Hfe in macrophages and innate immunity.

002

EFFECT OF ACUTE INFLAMMATION ON MOLECULES INVOLVED IN DUODENAL IRON UPTAKE

Abitha Sukumaran and Molly Jacob

Department of Biochemistry, Christian Medical College, Vellore-632002, Tamil Nadu, India

Background and hypothesis of study: Anaemia of inflammation is associated with decreased duodenal absorption of iron. This study was done to test the hypothesis that inflammation down-regulates the expression of molecules involved in duodenal iron uptake, leading to decreased iron absorption.

Experimental methods: Turpentine was injected into mice to produce inflammation. Blood, the duodenum and liver were taken from these animals at various time periods (ranging from 6h up to 7 days) after the injection. Serum interleukin-6 (IL-6) levels were measured to confirm the induction of inflammation. Expression of various proteins involved in duodenal iron uptake – divalent metal ion transporter 1 (DMT1), duodenal cytochrome b (dcytb), ferroportin, hephaestin, ferritin and ransferring receptor 1 (TfR1) – were determined by real-time polymerase chain reaction assays and western blotting. Hepcidin mRNA levels in the liver and serum iron levels were also measured.

Results: Serum IL-6 levels were significantly elevated in turpentine-treated mice, confirming the induction of inflammation in the animals. Expression of duodenal ferritin was significantly increased between 6 hours and 5 days, with associated decreases in the expression of DMT1 (over 24 hours to 5 days). Ferroportin expression was significantly lower initially (at 24 and 48 hours), but rose significantly between days 3 and 7. Serum iron levels were significantly lower initially (6-12 hours), followed by progressive increases that reached significant levels on day 5. Hepcidin mRNA in the liver was significantly induced over the period of the study. Many of the parameters measured showed significant correlations with one another.

Discussion: In this study, inflammation produced significant changes in the expression of several proteins involved in iron homeostasis. It caused induction of hepcidin which, we suggest, was responsible for the down-regulation seen in levels of duodenal ferroportin, leading to accumulation of iron (evidenced by increased ferritin levels) in the duodenum and falls in serum iron levels. We also postulate that duodenal accumulation of iron, in turn, resulted in down-regulation of DMT1. With subsidence of the inflammation, levels of duodenal ferroportin were found to increase. This appears to have led to increased export of iron from the duodenum and a consequent increase in serum iron levels by day 5.

Conclusion: Our data shows that inflammation altered the expression of many proteins involved in duodenal iron absorption, with associated falls in serum iron levels. We conclude that such changes, induced by inflammation, are likely to contribute to the pathogenesis of anaemia of inflammation.

Acknowledgements: Department of Biotechnology, India for financial support.

003

LUMINAL IRON: A RISK FACTOR FOR GUT-BORNE BACTERIAL INFECTIONS

Guus A.M. Kortman, Dorine W. Swinkels, Harold Tjalsma

Department of Laboratory Medicine of the Radboud University Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands

Background: The human gut is the natural habitat for a large and dynamic bacterial community. Major functions of this gut microbiota include important trophic effects on intestinal epithelia, on immune structure and function, and protection of the colonized host against invasion by foreign microbes. The provision of iron is an essential requirement for the growth of most bacterial species and in pathogenic bacteria, iron is frequently involved in the expression of virulence-associated properties. One of the eukaryotic mechanisms to constrain bacterial growth to the gut lumen is to tightly regulate the availability of free iron in the human body. Despite this protective mechanism, however, iron-rich diets/supplementation of iron have been associated with increased risk for bacterial infections in humans. To further gain insight in the role of iron on gut-borne bacterial infections, the effect of dietary (luminal) and systemic iron (host iron status) on the virulence of enteric pathogens and on the host immune responses will be investigated.

Methods:To mimic the luminal-epithelial-stromal architecture of the gastrointestinal tract in the laboratory, a transwell culture system with intestinal epithelial cells (Caco-2 cells) is used. This system enables to monitor pathogen and host responses individually and also host-pathogen interactions under variable conditions of iron loading.

Results:The first question to be answered is whether or not dietary iron supplementation can lead to a competitive advantage of pathogenic bacteria within the complex intestinal microbiome. To this purpose, several intestinal bacterial strains representing intestinal commensals and (opportunistic) pathogens were grown in defined medium with different concentrations of iron. These experiments clearly showed that enteric pathogens, like Salmonella typhimurium and Shigella flexerni, had a significant growth advantage in iron-rich medium, whereas commensals, like Lactobacillus plantarum have no iron-dependent growth characteristics.

Conclusions: These first findings implicate that pathogenic bacteria have the potential to outgrow the commensal population in an iron-rich intestinal lumen. This confirms the idea that dietary iron supplementation in anemic patients is not without risk as increased colonization of intestinal pathogens may lead to increased risk of gut-borne infections in immune-compromised individuals. Ongoing experiments focus on the effects of iron on bacterial translocation across the epithelial barrier and the capacity of the innate immune system to clear invading pathogens under high-iron conditions. Furthermore, we aim to investigate to which extent overgrowth of pathogenic bacteria occurs in vivo by mapping the human intestinal microbiome before and after oral iron supplementation.

004

RESTRICTION OF M. AVIUM INFECTION BY THE RHODAMINE-B LABELED HEXADENTATE CHELATOR CP777

Ana Rita Gomes, BSc1, Carla Queiróz2, MSc, Rui Appelberg1,3, PhD, MD, Maria Rangel, PhD 2,3, Maria Salomé Gomes, PhD 1,3

1IBMC – Instituto de Biologia Molecular e Celular, Universidade do Porto, 2REQUIMTE, Departamento de Química, Faculdade de Ciências, Universidade do Porto, 3ICBAS – Instituto de Ciências Biomédicas de Abel Salazar, Universidade do Porto

Iron availability has been found crucial to mycobacteria. Iron withholding being a critical host defense strategy towards invading pathogens, iron chelation therapy may be a functional adjunct therapy against mycobacterial infections. Nevertheless, the currently available iron chelators have poor impact on mycobacterial infections1. The relationship between iron availability and the growth of Mycobacterium avium is well established 1-2. Thus, we chose this intracellular pathogen as a model to identify new iron chelators that could inhibit mycobacterial growth more efficiently than those described previously. Formerly, we have identified a new hexadentate iron chelator, which we designated as CP777, able to inhibit the growth of M. avium inside macrophages (its natural host cell)3. In this work, we further explored the interesting anti-mycobacterial properties of this novel iron chelator on infection by M. avium in vivo, using a mouse model. At 40 mg/Kg body weight, CP777 significantly restricts the growth of M. avium both in the liver and in the spleen. This iron chelation therapy did not cause architectural changes in the principal organs affected by the intravenous mycobacterial infection, namely, liver and spleen. Notably, CP777 administration did not alter the hematocrit nor hemoglobin levels. In conclusion, our results confirm that iron deprivation, by the use of appropriate chelating compounds, restricts the growth of M. avium in vivo and that the new iron chelators4 offer a potential therapy to control mycobacterial infections.

005

INTERACTIONS BETWEEN DIETARY LIPIDS AND IRON METABOLISM

Umbreen Ahmed and Phillip S. Oates

Physiology, School of Biomedical, Biomolecular and Chemical Sciences, The University of Western Australia, Nedlands 6009, W. Australia

Background/Aims: The type and nature of dietary fats affect body iron status by altering the absorption and utilization of iron. Body iron status is regulated at the level of the intestinal enterocyte by the hepatocyte derived peptide, hepcidin. It is possible that fat also plays a role in the regulation of iron absorption, via modulating hepcidin expression/release. Fat can directly affect hepcidin expression and indirectly affect the regulators of hepcidin expression i.e haemochromatosis protein (HFE), hemojuvelin (HJV) transferrin receptor 2 (TFR2) endoplasmic reticulum (ER) stress regulators and proinflammatory cytokines, interleukein-6 (IL-6) and tumour necrosis factor-alpha (TNF-α)A previously validated model of non-alcoholic fatty liver (NAFL), was used to determine the possible interactions between hepatic lipid and iron metabolism. This model is not associated with obesity and the development of NAFL is early in its onset preceding oxidative stress, ER stress and inflammation.

Methods: Sprague Dawley rats were fed standard A (35% energy from fat) or high fat B (71% energy from fat) liquid diets with normal iron content (A/B groups) for 5 weeks. In addition rats were made iron deficient (A-/B- groups) and iron loaded (A+/B+ groups) by dietary and parenteral injections of iron dextran.

Results: Plasma osmolality, haemoglobin and MCHC were high in all A diet groups compared with all B diet groups. Hct was not affected by the dietary fats. Plasma iron and transferrin saturation were affected by an interaction between dietary fat and iron. These parameters were high in group A with normal dietary iron, compared with its respective B group (P< 0.05). Similarly this group also showed a four-fold increase in the mRNA expression of the HFE gene. Spleen iron was also high in the A+ group compared with all other groups. Hepatic iron, and mRNA expression of lipogenic (PPAR-γ, C/EBP-α), ER stress (CHOP, XBP1), IL-6, iron effector (TfR2, DMT1 IRE, DMT1 non IRE) genes were all increased in diet B groups. In contrast mRNA expression of TNFα) was increased in diet A groups. There was a trend for the iron regulatory gene HAMP to be higher in the high fat diet groups (P= 0.057).

Conclusion: The consumption of standard diets with high carbohydrate content leads to post prandial hyperosmolality, causing a fluid shift in different body compartments that may affect haematological parameters. HFE and iron effector gene expression is affected by the macronutrient composition of the diet and in turn may affect plasma iron parameters. High fat liquid diet with iron loading resulted in the release of spleen iron by hepcidin independent mechanism. Hepcidin expression can be affected by other nutritional dependent factors in addition to iron. Although failing to reach significance there was a trend towards higher hepcidin expression in rats fed the high fat diet, suggesting that fat can affects its expression possibly involving C/EBP-α), IL-6, and TfR2. The increased hepatic expression of TNFα seen with the standard diets was not associated with increased hepcidin expression suggesting its actions can be dissociated from IL-6 expression, possibly representing expression in different cell compartments.

006

ROLE OF HEMOPEXIN IN DUODENAL INORGANIC AND HEME IRON ABSORPTION

Veronica Fiorito, Simonetta Geninatti Crich, Lorenzo Silengo, Fiorella Altruda, Silvio Aime, Emanuela Tolosano

Molecular Biotechnology Center, University of Torino, Torino, Italy

Iron is an essential element for virtually all living organisms. It is required as a cofactor for a multitude of proteins of diverse biological function and it plays a central role in the formation of hemoglobin and myoglobin and in many vital biochemical pathways and enzyme systems. During evolution, nature has not developed a pathway for excretion of iron in humans, so the body concentration of iron can only be regulated by a fine equilibrium between absorption and losses of iron. Iron can be absorb from diet in two forms: inorganic iron and heme. Although inorganic iron absorption is item of study from many years, the recent identification of intestinal importers and exporters for heme and of new stimuli regulating iron uptake suggests that much more work has to be done to completely understand the mechanism of iron and heme absorption. Hemopexin is the plasma protein with the highest binding affinity for heme. As it provides protection against free heme-mediated oxidative stress, limits access by pathogens to heme, and contributes to iron homeostasis by recycling heme iron, we asked whether hemopexin is involved in physiologic heme and inorganic iron absorption. Despite of a normal expression of duodenal iron transporters and although two different techniques, i.e. 55Fe measurement in tied-off duodenal segments and 57Fe oral administration, revealed a normal ability to absorb inorganic iron, hemopexin-null mice showed increased iron deposits in duodenal enterocytes and higher H-ferritin levels. Iron overload was associated with increased duodenal expression and activity of heme oxygenase, and with lower 5-aminolevulinic synthase duodenal mRNA levels. Conversely, in hemopexin-null livers and skeletal muscles, hemopexin deficiency was associated to enhanced expression of this heme biosynthetic enzyme. These data suggest that heme trafficking in the duodenum of hemopexin-null mice is impaired and that iron deposits in the duodenum of these animals derives from an enhanced local heme catabolism, supporting the idea of a crucial role played by hemopexin in duodenal heme-iron absorption.

007

TREATING DISORDERS OF IRON MISDISTRIBUTION:FROM CONCEPT TO CLINICAL PRACTICE

Ioav Cabantchik, Joseph Sohn and William Breuer

The Alexander Silberman Institute of Life Sciences, The Hebrew University, Safra Campus at Givat Ram, Jerusalem 91904, Israel

Although chelation is useful for reducing multi-organ iron deposition found in iron overloaded patients, it is probably inappropriate for disorders associated with iron misdistribution within selected tissues/cells. In most of those disorders, a regional iron accumulation occurs in the absence of systemic IO and it is often accompanied by regional or systemic iron deficiency (1). Thus, iron raises to toxic levels in mitochondria of specific excitable cells in some forms of neurodegeneration with brain accumulation or following mutations in particular genes associated with mitochondrial functions such as abcb7 in x-linked sideroblastic anemia w/ ataxia and frataxin in Friedreich's ataxia (FRDA), often leaving the cytosol iron-depleted. In anemia of chronic disease (ACD) iron is withheld by macrophages, which become susceptible to infection by intracellular pathogens, while iron levels in extracellular fluids (e.g. plasma) are drastically reduced (causing anemia). In order to assess the therapeutic potential of redistributing iron in pathological conditions associated with metal misdistribution, we have recently shown that deferiprone (DFP), a membrane-permeant bidentate chelator in clinical use for treating systemic IO, can act as siderophore with iron relocating abilities in settings of regional iron accumulation. The modus operandi involves: a. the capture of labile iron accumulated in cell compartments and b. conveying the chelated iron either to other cell locations for metabolic integration or to extracellular transferrin for systemic reutilization (2). Using experimental models that phenotypically reproduce the retention of iron in macrophages (as in ACD) or cell iron misdistribution resulting from aberrant utilization of the metal by mitochondria (as in FRDA) we can: a. biophysically demonstrate the cell iron misdistribution and its biochemical and physiological consequences (3) and b. dissipate the metal misdistribution and thereby functionally correct the affected systems. We found that chelators endowed with demonstrable iron relocation abilities, confer upon cells protection from the presence of regionally accumulated iron and concomitantly mediate restoration of cellular (intra or extra) iron-deficient functions. These experimental findings were translated into clinical protocols for treating patients with FRDA (4) and others with ACD. Supported by the ISF and the Canadian Friends of HUJI.1. Kakhlon, et al. (10) Can. J. Physiol. Pharmacol. 88:187-96. 2. Sohn, Y.S. et al. (08) Blood 111:1690-9.; 3. Kakhlon, O. et al. (09). Blood. 112:5219-27; 4. Boddaert N, et al.(07). Blood. 110:401-8.

008

AN EXPERIMENTAL CELLULAR MODEL FOR ASSESSING AGENTS WITH A POTENTIAL FOR INTERCELLULAR IRON REDISTRIBUTION

Yan Sung Sohn, William Breuer Anna-Maria Mitterstiller, Guenter Weiss and Ioav Cabantchik

The Institute of Life Sciences, The Hebrew University, Safra Campus at Givat Ram, Jerusalem 91904, and Medical University of Innsbruck, Dept. for Internal Medicine I, Clinical Immunology and Infectious Diseases, Anichstr. 35, A-6020 Innsbruck, Austria

Systemic iron deficiency concomitant with retention of macrophage accumulated iron is a characteristic feature of iron-refractory anemias associated with inflammation. The resulting iron misdistribution, which is also prone to exacerbation by parenteral iron supplementation, is mainly attributable to hepcidin-mediated down-regulation of the iron exporter ferroportin. We show here that the main iron accumulation/retention features are recapitulated in a macrophage cell subline that takes up excessive iron via endocytic loading of polymeric iron forms or damaged erythrocytes, resulting in increased labile iron pools and ensuing oxidative damage. These properties were found to be aggravated by hepcidin-evoked iron retention. We used that experimental model as a screening platform for agents that can rescue the iron-laden cells blocked in their export pathways but also convey the chelated iron to other co-cultured cells while sparing the growth of intracellular pathogenic Salmonella. As test of combined feasibility for rescuing iron-affected cells and selectively supporting other iron-starved cells we assessed a repertoire of chelators in clinical practice. We found that agents like the hydroxypyridin-on deferiprone optimally supported macrophage cell growth suppressed by iron overload while concomitantly inhibiting intracellular bacterial multiplication stimulated by intracellular iron and rendering the iron metabolically available to other. Thus, the capacity of chemical agents to mediate conservative iron relocation highlights the feasibility of using chelators with siderophore properties for safely relieving cells from iron burden without affecting their growth or those of other cells (1). Such treatment might also aid in correcting anemias associated with misdistribution of iron. (1) Sohn YS, Breuer W, Munnich A, Cabantchik ZI. Redistribution of accumulated cell iron: A modality of chelation with therapeutic implications. Blood.2008;113:1690-9. Supported by the Israel Sci. Found. (ISF) and a HUJI Appl. Science award via the Canadian Friends of the Hebrew University of Jerusalem.

009

IDENTIFICATION OF A MITOCHONDRIAL FLVCR ISOFORM

Deborah Chiabrando, Sonia Mercurio, Sharmila Fagoonee, Samuele Marro, Erika Messana, Emilia Turco, Lorenzo Silengo, Fiorella Altruda and Emanuela Tolosano

Department of Genetics, Biology and Biochemistry and Molecular Biotechnology Centre. University of Turin, Torino, Italy

Feline Leukemia Virus subgroup C Receptor (FLVCR) was originally identified and cloned as a cell-surface protein receptor for feline leukaemic virus subgroup C, causing pure red blood cell aplasia in cats. Recent studies have demonstrated that FLVCR is a heme exporter essential for erythropoiesis. The heme efflux via FLVCR was shown to be essential for erythroid differentiation in K562 cells as well as in CD34+ precursors cells. Moreover, Keel and co-authors have reported that Flvcr-null mice die in utero due to the failure of fetal erythropoiesis1. We have identified an alternative transcription start site giving rise to a novel FLVCR isoform (FLVCRb). Flvcr-b transcript completely lacks the first exon of the canonical isoform (FLVCRa) and code for a putative 6 transmembrane domain containing protein ubiquitously expressed. In vitro over-expression of FLVCRa and FLVCRb showed that the two proteins display different subcellular localization. As expected FLVCRa localizes at the cell membrane while FLVCRb localizes in the mitochondrial compartment. The mitochondrial localization of this novel isoform is further confirmed by the identification of a N-terminal mitochondrial sorting presequence. Because of FLVCRa is a heme exporter at the cell membrane, we hypothesized that FLVCRb could be the mitochondrial heme exporter. According to this hypothesis, FLVCRb expression increased following the stimulation of heme biosynthesis in vitro, in correlation with the increase in hemoglobin production. The ability of FLVCRb to bind and export heme out of the mitochondria is still under investigation. To gain insights into the specific roles of the two isoforms, we have generated Flvcr mutant mice different from those previously reported1. Keel and co-author generated a mouse model in which both FLVCRa and FLVCRb have been deleted. In our mouse model, FLVCRa has been specifically deleted and FLVCRb is still expressed (FLVCRa-null mice). Flvcr-a heterozygous mice were grossly normal, fertile and indistinguishable from their wild-type littermates. When Flvcr-a heterozygous mice were intercrossed, no Flvcr-a homozygous knock-out newborns were obtained, and the analysis of the embryos from timed Flvcr-a+/- intercrosses showed that the Flvcr-a-/- genotype was lethal between E14.5-E16.5. Flvcr-a-null embryos showed multifocal and extended hemorrhages, visible in the limbs, head and throughout the body wall, as well as subcutaneous edema. Alcian blue-alizarin red staining demonstrated skeletal abnormalities in limbs and head similar to that observed in Diamond-Blakfan anemia (DBA) patients. Interestingly, flow cytometric analyses of E14.5 fetal liver cells double-stained for Ter119 (erythroid-specific antigen) and CD71 (transferrin receptor) show normal erythropoiesis in Flvcr-a-null embryos, opposite to the previously reported Flvcr-null mice1.Taken together, these data demonstrated that FLVCRb is sufficient to support fetal erythropoiesis likely exporting heme from the mithocondrion for hemoglobin synthesis. The loss of FLVCRa leads to endothelial ruptures responsible for hemorrhages thus suggesting that FLVCRa is needed for detoxifying heme excess at these sites. 1.Keel SB et al. A heme export protein is required for red blood cell differentiation and iron homeostasis. Science 2008. Feb 8;319(5864):825-8.

010

MECHANISMS AND REGULATION OF IRON TRANSPORT ACROSS THE BLOOD TESTIS BARRIER

Yael Bardoogo-Leichtmann1, Britta Marohn2, Esther Meyron1

1Laboratory for Molecular Nutrition, Faculty of Biotechnology and Food Engineering. Technion, Israel, 2Medizinische Hochschule Hannover, Institut für Humangenetik, Hannover, Germany

The Blood Testis Barrier (BTB) protects the male reproduction system from blood borne pathogens and nutritional imbalances. It consists of the epithelial layer of Sertoli cells within the seminiferous tubules, which form tight junctions that separate the tubule into a basal and an adluminal compartment. The Sertoli cells secrete hormones and nutrients essential for sperm development and these secretions are regulated by the metabolic demands of the developing sperm. Germ cell development is highly organized and involves the interaction of groups of germ cells with Sertoli cells. More than ten male germ cell divisions and many steps of sperm development take place outside and across the Sertoli cell tight junction, which implies a high iron demand of these cells. We therefore hypothesized that iron has an important role in male fertility and that male fertility may be affected by systemic iron overload or deficiency.

We localized iron and iron metabolism proteins in the testis and found that ferric iron, ferritin and transferrin-receptor co-localized in the early spermatocytes, DMT1 appeared mainly in Sertoli cells and ferroportin was detected on the periphery of the seminiferous tubules and on blood capillaries of the testis. This surprising distribution of iron transport proteins stands in contrast to an early report where TfR was localized mainly to Sertoli cells. We also found that the BTB protects the developing sperm from iron-overload to some extent, and the detailed effect of nutritional iron overload on male fertility is presently under assessment. Taking these findings together, we suggest, that early spermatocytes accumulate iron to equip themselves prior to their entry through the BTB with stores of iron for instant supply during the multiple cell divisions of sperm development. We further hypothesize that the mature sperm cells get rid of their excess iron by the Sertoli phagocytosis of their cytoplasm. Excess iron in the Sertoli cells is exported out of the seminiferous tubule by ferroportin.

011

POLY-DIMENSIONAL INTERACTION ANALYSIS : GENERATING INSIGHT RE THE PROTEINS OF IRON METABOLISM

Ian C. Boulton1, Sana Hussain1, Vadim B. Vasilyev2, Robert W. Evans1

1Division of Biosciences : Brunel University : UXBRIDGE, 2Institute for Experimental Medicine, Russian Academy for Experimental Medicine, ul. Akademika Pavlova ST PETERSBURG

We have utilized solid phase peptide arrays, ELISA and surface plasmon resonance biosensors (SPR) to identify and characterize a number of novel interactions between proteins of iron metabolism including lactoferrin (hLf), transferrin (hTf), ceruloplasmin (Cp), ferroportin and myeloperoxidase (Mp). We are also characterizing the interactive properties of the beta amyloid precursor protein (App).

By these studies we have determined that Cp interacts with up to six basic peptides in the hLf N lobe and that this interaction is, to some degree, species specific, being demonstrable in human and certain non-human Lfs. We have also determined that Cp preferentially binds iron loaded (holo) hLf, suggesting that this interaction is influenced by ligand conformation. Cp also interacts with several peptides within the hTf sequence, although interaction could not be effectively established using intact proteins.

Further, we have established that CP interacts with two overlapping peptides in the human ferroportin sequence and that peptides form part of putative intracellular loop four of this protein suggesting that the current topology model for ferroportin may require revision. Ferroportin did not interact with either hTf or hTf in the conditions examined here. We have also established that human Mp interacts with both hLf and Cp and that Mp preferentially binds to holo hLf. In addition, we will present data relating to the affinities, stoichiometries and possible physiological significance of these associations.

012

COMPUTATIONAL STRUCTURE MODELS OF DIFERRIC AND APO TRANSFERRIN-TRANSFERRIN RECEPTOR2 COMPLEXES

Takaki Yamamura and Tetsuya Sakajiri

Faculty of Nutritional Sciences, The University of Morioka, Japan

Transferrin receptor 2 (TfR2), a new TfR1-like family member, is dominantly expressed in the liver, intestine, and cancer cell. The TfR2 molecule shows a sequence identity of 44% with TfR1. The main role of TfR2 is uptake of transferrin (Tf)-bound iron into the cell as well as that of TfR1. Diferric Tf (Fe2Tf) is bound to TfR (TfR1 or TfR2) at the cell surface (pH 7.4), followed by endocytosis and iron release due to acidification (pH 5.6) in the endosome. ApoTf remains bound to TfR in the endosome. The apoTf-TfR returns to the cell surface and apoTf is dissociated from TfR, allowing it to be recycled. Much data has been accumulated regarding Tf-TfR1 complexation, such as results from mutagenenesis and the crystal structure of TfR1. We have recently created 3D structures of human Fe2Tf- and apoTf-TfR1 complexes by computer modeling (protein J. (2009) 28:407-414). For TfR2, however, there is scarce physicochemical data. In this study, we firstly built a 3D structure model of TfR2 using the TfR1 crystal structure. Then, we molded structures of Fe2Tf- and apoTf-TfR2 complexes upon Fe2Tf- and apoTf-TfR1 complexes. The crucial ionic bonds that are present in common with both Fe2Tf- and apoTf-TfR1 complexes, i.e., Tf Arg50-Asp667 TfR1, Tf Arg352-Asp648 TfR1, Tf Glu367-Arg646 TfR1, and Tf Glu385- Arg651 TfR1 are completely conserved in Fe2Tf- and apoTf-TfR2 complexes, making pears of Tf Arg50-Asp669 TfR2, Tf Arg352-Asp680 TfR2, Tf Glu367-Arg678 TfR2, and Tf Glu385-Arg683 TfR2. However, pairwise interactions corresponding to two ionic bonds of Tf Lys148-Asp125 TfR1 and Tf Lys511-Glu759 TfR1 in the Fe2Tf-TfR1 complex are not made in the Fe2Tf-TfR2 complex, because of replacement of Asp125 and Glu759 in TfR1 with Ser138 and Gln800 in TfR2, respectively. Alternatively, the following two specific ionic bonds are found in both Fe2Tf- and apoTf-TfR2 complexes: Tf Glu56-Arg466 TfR2 and Tf Glu328-Arg689 TfR2 due to replacement of Gln443 and Thr657 in TfR1 with Arg466 and Arg689 in TfR2, respectively. In the apoTf-TfR1 complex, Tf His349 binds to TfR1 Trp641/Phe760 under the acidic condition (pH 5.6) in the endosome making the T-shaped interaction between the protonated imidazole ring of Tf His349 and benzene rings of both TfR1 Trp641 and Phe760. This pH depending interaction is completely conserved in TfR2 producing the interaction between Tf His349 and TfR2 Trp673/Phe801. However, the pairwise interactions corresponding to the ionic bond of Tf Glu141-Lys508 TfR1 in the apoTf-TfR1 and Tf Glu357-Arg629 TfR1 in both apoTf- and Fe2Tf-TfR1 complexes are not produced in TfR2. These missing bonds in the apoTf-TfR2 complex are compensated by the above-mentioned bonds of Tf Glu56-Arg466 TfR2 and Tf Glu328-Arg689 TfR2. A bond corresponding to Tf Lys148-Asp125 TfR1 in the Fe2Tf-TfR1 complex is not found in the Fe2Tf-TfR2 complex.In conclusion, structures of Fe2Tf- and apoTf-TfR2 complexes are similar to those for TfR1. Number of main ionic bonds in the apoTf-TfR2 complex is conserved compared to that for TfR1. However, one main ionic bond is missing in Fe2Tf-TfR2, resulting in 30-times lower affinity of Fe2Tf to TfR2 than to TfR1.

013

NEW BIOINFORMATIC TOOL FOR PREDICTION OF IRON-RESPONSIVE ELEMENTS

Mayka Sanchez1,2, Monica Campillos1, Ildefonso Cases2, Gerard Frigola-Quintana2, Matthias W. Hentze1

1European Molecular Biology Laboratory (EMBL), Meyerhofstrasse 1, 69117 Heidelberg, Germany, 2Institute of Predictive and Personalized Medicine of Cancer (IMPPC), Crta Can Ruti, Camí de les Escoles s/n, 08916 Badalona, Barcelona. Spain

The IRP/IRE regulatory system plays a crucial role in the post-transcriptional regulation of gene expression and its disruption results in human disease. Iron-responsive elements (IREs) are cis-acting regulatory motifs present in mRNAs that encode for proteins involved in iron metabolism. They function as binding sites for two related trans-acting factors, namely the iron regulatory proteins (IRP1 and IRP2). Among the cis-acting oligonucleotide patterns, the IRE is one of the best characterized. It is defined by a combination of RNA sequence and structure. However, currently available programs to predict IREs do not show a satisfactory level of sensitivity and fail to detect functional IREs. Here, we report an improved software for the prediction of IREs implemented as a user-friendly web-server tool. The SIREs (Search for iron-responsive elements) web-server uses a simple data input interface and provides structure analysis, predicted RNA folds, folding energy data and an overall quality flag based on properties of well characterized IREs. Results are reported in a tabular format and in a schematic visual representation that highlights important features of the IRE. The SIREs web-server is freely available on the web at http://ccbg.imppc.org/sires/index.html

014

A NEW MOUSE MODEL WITH GAIN OF IRP FUNCTION TO INVESTIGATE THE IN VIVO ROLE OF THE IRP/IRE SYSTEM IN NORMAL AND PATHOLOGICAL IRON METABOLISM

D. Casarrubea1, L. Viatte1, R. Eisenstein2, B. Galy1 and M.W. Hentze1

1European Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117, Heidelberg, Germany, 2University of Wisconsin, Dept. of Nutritional Sciences 1415 Linden Drive, Madison WI 53706

Iron Regulatory Proteins (IRP) −1 and −2 coordinately secure cellular iron homeostasis by binding to cis-regulatory iron-responsive elements (IRE) on mRNAs encoding proteins involved in iron uptake, handling, storage and export. In turn, the cellular labile iron pool regulates the IRE-binding activity of the two IRPs through distinct mechanisms.To explore the in vivo functions of the IRP/IRE system and to complement investigations with loss-of-function lines, we have developed a mouse model allowing conditional expression of a gain-of-function IRP1 mutant using Cre/Lox technology.We first generated a cDNA encoding a flag-tagged mutant version of IRP1 (IRP1*) that binds IREs in a constitutive manner due to the lack of critical residues required for Fe-S cluster assembly. This cDNA was inserted into the permissive Rosa26 locus by homologous recombination in murine embryonic stem (ES) cells together with a floxed β-geo stop cassette placed between the ubiquitous Rosa26 promoter and the IRP1* cDNA. This cassette prevents transcription of the IRP1* transgene; its excision upon Cre-mediated recombination enables expression of IRP1* in a conditional manner.Mouse line harbouring the targeted Rosa26 locus was obtained by injection of ES clones into blastocysts and was subsequently crossed with a pan-Cre-deletor strain to trigger IRP1* expression in the whole organism, regardless of developmental stages.IRP1* expression was detected by western blotting in different organs and in higher levels in the homozygous mice. IRP1* was shown to be functional by EMSA: it binds a radio-labeled ferritin-IRE sequence, the complex is supershifted by an anti-flag antibody and its formation is competed by molar excess of a cold wildtype IRE sequence but not by a mutant version. Ex vivo, bone marrow derived macrophages (BMDM) from IRP1*-pan-expressing mice showed altered sensitivity to iron concentration in the medium, reflected in altered expression levels of IRP target genes under standard culture conditions and in response to Fe loading and depletion.Of particular interest will be to analyze the in vivo effects of forced IRP overexpression on the IRP regulon and the systemic consequences on iron absorption, utilization, storage, and recycling. This study will establish the physiological importance of appropriate IRP regulation in different organs specialized in the control of body iron fluxes. Our new mouse line should also provide a valuable model to study diseases associated with abnormally high IRP activity as observed e.g. in the substantia nigra of patients with Parkinson's disease or in patients suffering from sideroblastic anemia linked to glutaredoxin 5 deficiency.

015

POTENTIAL ROLE OF MINER 1, A NOVEL OUTER ORGANELAR MEMBRANE NEET PROTEIN WITH LABILE IRON SULFUR CLUSTER, IN CELLULAR TRAFFICKING OF IRON/IRON SULFUR CLUSTERS

Sagi Tamir, Y. Harir, A. Conlan, M. Shvartsman, M. Paddock, P. Jennings, R Mittler, I Cabantchik, Rachel Nechushtai

Hebrew University of Jerusalem, Israel

The NEET proteins, mitoNEET and Miner 1, that reside on the outer mitochondrial and ER membranes, are 2Fe-2S protein (ISP), have a unique structure composed of two protomers intertwined to form an homodimeric novel structure (among the ˜650 known ISPs) (1,2). The presence of His instead of Cys in their iron sulfur cluster (ISC)-binding site confer ISC lability. Here we assessed the Miner1 possible involvement in ISC trafficking by examining its ability to donate it's ISC and/or Fe to putative acceptor proteins and to cellular organelles, e.g. mitochondria. Soluble recombinant NEET proteins (33-108 aa) incubated with apo-acceptor proteins, e.g. apo-ferredoxin (apo-Fd) under reducing conditions, evoked a decrease in the optical signature for the NEET ISCs and a concomitant increase in the signature for ferredoxin ISC, 458nm and 423nm respectively, indicating ISC transfer from the NEET protein to apo-Fd. This was confirmed by native PAGE as colorless apo-Fd is converted to red holo-Fd (containing ISC) after incubation with a NEET protein (mitoNEET or Miner1). The ability of the NEET proteins to donate their ISC's or iron to cellular organelles, e.g. mitochondria, was assessed in digitonin-permeabilized HEK293 and/or H9c2 cells double-labeled with fluorescent labile iron sensors rhodamine-phenanthroline in mitochondria and calcein-green in cytosol (3). Fluorescence imaging microscopy of permeabilized cells showed that addition of NEET protein caused a time dependent quenching in mitochondrial RPA fluorescence, indicating the transfer of reduced iron to the mitochondrial matrix residing RPA probe. While the transfer in solution of ISC from the NEET protein to other ISP's was obtained only in reducing environments, the transfer to mitochondria occurred spontaneously. Moreover, the kinetics of ISC transfer to apo-ISP or chelators by Miner1 is much faster than that of mitoNEET. Miner 1 was shown as causative of Wolfram Syndrome 2 (4) and to be involved in longevity (5). Our results strongly suggest that Miner1 plays a functional role in the cellular trafficking of iron/ISC. Supported in part by ISF, NIH and EEC F6(LSHM-CT-2006-037296 Euroiron1). 1. Paddock et al. (2007) Proc Natl Acad Sci USA 104:14342-47; 2. Conlan et. al. (2009) J Mol. Biol 392: 143-153; 3. Shvartsman et al. (2007) Am J Physiol (Cell) 293:C1383-94, 4. Amr et. al. (2007) Am J Hum Genet 81: 673–683; 5. Chen et al. (2009) Genes Dev. 23:1183-1194.

016

COPPER-DEPENDENT REGULATION OF DMT1 IN HUH7 HEPATOMA CELLS

Solomis Solomou, Katayoun Pourvali, Paul Sharp

Nutritional Sciences Division, King's College London, UK

The close relationship between copper and iron metabolism has been recognised for many years. We have shown previously that copper and iron compete for uptake via DMT1 in Caco-2 cells. Furthermore, the expression of DMT1 and ferroportin are both influenced by copper status in intestinal epithelial cells. Less is known about the effects of copper on iron transporter expression in other cell types. Therefore in this study we have investigated the effects of copper on the protein and mRNA expression of the iron transporters DMT1 and ferroportin in human hepatoma cells. HuH7 hepatoma cells were treated with 50 μM copper chloride, or the copper chelator Triethylenetetramine dihydrochloride (TETA, 0.5mM) for 4 and 24 hours. Changes in whole cell levels of transport proteins were measured by western blotting and changes in mRNA expression assessed using Real Time PCR. Western blotting data were semi-quantified using ImageJ software for were analysed by one-way ANOVA and Tukey's post hoc test (significant at p<0.05).Following exposure to copper for 24h there was a significant decrease in DMT1 protein expression (-45%; p< 0.005) compared with control. TETA treatment resulted in a significant increase in DMT1 protein (+43%, p<0.05). There was no effect of copper on DMT1 mRNA expression. Ferroportin expression (protein & mRNA) was unaltered by either copper loading or deficiency.Taken together with our previous data, this work provide further evidence that copper status is an important factor in regulating iron homeostasis.

017

TMPRSS6 EXPRESSION IS MODULATED BY IRON AND INFLAMMATION

Delphine Meynard, Chia Chi Sun, Elena Corradini, Joddie L, Babitt, Herbert Y.Lin

Program in Membrane Biology, Nephrology Division, Center for Systems Biology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA

Iron deficiency is a leading cause of anemia, affecting over one-half billion people worldwide. Mutations in the TMPRSS6 gene are responsible for a familial anemia disorder, Iron-Refractory Iron Deficiency Anemia (IRIDA). TMPRSS6 gene encodes for matriptase-2, a type II transmembrane serine protease. These patients have inappropriately elevated levels of the iron regulatory hormone, hepcidin, suggesting that TMPRSS6 is involved in negatively regulating hepcidin expression. Hepcidin is regulated by the availability of iron via the BMP-SMAD signaling pathway and by inflammation via the IL-6/JAK/STAT3 signaling pathway. How TMPRSS6 is regulated is not known. Here, we investigated whether TMPRSS6 may also be co-regulated by iron, BMP-SMAD signaling, and inflammation to influence hepcidin expression. C57BL/6 mice fed with an iron-enriched diet for 2 weeks exhibited an increase in TMPRSS6 mRNA expression by 1.7 fold (p=0.0031). To address if the upregulation of TMPRSS6 is modulated by the BMP-SMAD pathway, we treated heptatoma-derived cells with BMP ligands and with the small molecule BMP inhibitor LDN-193189. Treatment with BMP6 (25ng/ml) increased in TMPRSS6 mRNA by 1.8 fold (p=0.05) compared to untreated controls. We observed similar upregulation of TMPRSS6 mRNA expression using other BMP ligands. In contrast, treatment with LDN-193189 led to a 2.5 fold decrease in TMPRSS6 mRNA expression. These results indicate that TMPRSS6 mRNA expression can be regulated by iron and by the BMP-SMAD pathway. Next, we assessed the effects of inflammation on TMPRSS6 mRNA expression in mice. We generated a mouse model of inflammatory anemia by injection of killed Brucella abortus bacteria. These mice express high levels of the inflammatory cytokine IL-6 mRNA within 6h of Brucella injection (86-fold (p=0.1)), and up to 11 days (96-fold (p=0.05)). Under these inflammatory conditions, TMPRSS6 mRNA was decreased by 1.6 fold (p=0.073) after 6h and by 1.7 fold (p=0.008) after 11 days. To determine whether IL-6 is responsible for these modulations of TMPRSS6 mRNA, Hep3B cells were treated with different amounts of IL-6. Treatment with 5ng/ml of IL-6 for 16h led to a decrease of TMPRSS6 mRNA by 1.4 fold, while 20ng/ml of IL-6 for the same time period decreased TMPRSS6 mRNA levels by 1.8 fold. In summary, we demonstrate that TMPRSS6 mRNA expression is increased by iron in vivo and by the BMP-SMAD pathway in vitro. In addition, we show that TMPRSS6 mRNA expression is decreased by inflammation and specifically by IL-6 both in vitro and in vivo. Our data suggest that iron, BMP-SMAD signaling, and

018

IS THE LACK OF BMP6 SUFFICIENT TO RESCUE THE ANEMIC PHENOTYPE OF MATRIPTASE-2 DEFICIENT MICE?

Anne Lenoir1, Jean-Christophe Deschemin1, Léon Kautz2, Marie-Paule Roth2, Carlos Lopez-Otin3, Sophie Vaulont1 and Gaël Nicolas1

1Institut Cochin, INSERM unité 1016, CNRS UMR 8104, Université Paris Descartes – France, 2INSERM unité 563 – France, 3Universidad de Oviedo - Spain

Hepcidin is the master regulator of systemic iron homeostasis. Hepcidin decreases iron availability by limiting dietary iron absorption and iron recycling from senescent red blood cells. A low hepcidin expression is responsible for iron overload (hemochromatosis) whereas a high hepcidin expression leads to iron deficiency and anemia. In the liver, iron-dependent stimulation of hepcidin expression is regulated through a pathway involving the Bone Morphogenetic Protein BMP6 and its cell surface receptor hemojuvelin. Accordingly, BMP6 deficient mice present with a severe hemochromatosis-like phenotype due to hepcidin down-regulation. Recently, the serine protease TMPRSS6/matriptase-2 has been proposed as an essential component of an iron-sensing pathway allowing hepcidin repression to permit increased iron mobilization to face iron deficiency. In humans, matriptase-2 mutations produces iron-refractory iron-deficiency anemia (IRIDA), characterized by inappropriately high levels of hepcidin and unresponsiveness to oral iron therapy. Recent in vitro studies have suggested that this protease down-regulated hepcidin expression by cleaving hemojuvelin.The goal of our study was to characterize in vivo the relationship between matriptase-2 and BMP6 and to analyze the role of BMP6 in the setting of anemia in mice deficient for matriptase-2. For this purpose, we intercrossed matriptase-2 and BMP6 deficient mice and analyzed the expression of Hamp1 encoding for functional hepcidin, plasma and hepatic iron level as well as hematological parameters in the double mutant mice compared to controls or single mutant mice. Results and conclusions of this analysis will be presented.

019

TMPRSS6 HAPLOINSUFFICIENT MICE HAVE INCREASED SUSCEPTIBILITY TO DEVELOP IRON DEFICIENCY

Antonella Nai, Alessia Pagani, Laura Silvestri and Clara Camaschella

Vita-Salute University and Division of Genetics and Cell Biology, San Raffaele Scientific Insitute, Milan, Italy

Matriptase-2 (MT-2), encoded by TMPRSS6, is a liver transmembrane serine-protease. By cleaving membrane-hemojuvelin (mHJV), MT-2 interrupts the BMP/SMAD signalling, downregulating hepcidin [1]. This function is essential for erythropoiesis: loss of MT-2 activity in humans [2] and mice [3][4] results in a severe iron-deficient-anemia. Oral iron administration is ineffective because of high levels of hepcidin. Genome-wide-association-studies (GWAS) have recently showed that heterozygous TMPRSS6 genetic variants are associated with decreased iron and haematological parameters [5]. Here we analyze the effect of the heterozygous deletion of the Tmprss6 gene in mice, confirming and complementing recent data [6]. Wild-type (wt) and Tmprss6+/- mice were analyzed at 3 and 8 weeks of age, the latter fed a normal or an iron-deficient diet. Blood samples were collected for blood cell counts and sTf saturation. After sacrifice animal livers were analyzed for quantification of hepcidin, TfR1, Id1, C-Reactive-Protein mRNA levels by qPCR. The inhibitory effect of MT-2 on hepcidin promoter and its ability to cleave mHJV were evaluated in vitro [1] using MT-2wt alone and in proportion 1:1 with the Mask mutant in order to simulate Tmprss6 haploinsufficiency. Tmprss6 +/- mice develop normally and are phenotypically indistinguishable from the wt littermates. However, at 3 weeks of age they have lower hemoglobin and red cell counts, and higher hepcidin levels as compared to wt mice. In adult mice hemoglobin normalizes, although MCV and sTf saturation persist lower than in wt littermates and hepcidin levels remain inappropriately high if related to sTf saturation. To assess the effect of iron restriction, 4 week-old animals of both genotypes were fed a normal or an iron deficient diet and analyzed at 8 weeks. The poor-iron diet caused a greater decrease of sTf saturation and a trend toward a more severe anemia in Tmprss6 +/- as compared to wt mice. However, as described for pregnant mice [6], hepcidin was strongly donwregulated in both genotypes, strenghening that iron deficiency overrides the SMAD-pathway activation of the Tmprss6 haploinsufficient mice. In vitro analyses demonstrate a dose-dependent activity of MT-2wt on hepcidin promoter inhibition and on HJV cleavage, supporting the in vivo observations. Taking together, these data indicate that a full MT-2 activity is crucial in maintaining normal hemoglobin levels, especially when iron demands are high for growth and erythropoiesis expansion or during iron restriction. Together with the results of GWAS our data suggest that TMPRSS6 heterozygous mutations may confers susceptibility to iron deficiency.

020

HEMOJUVELIN PROTEIN LEVELS IN MATRIPTASE-2-DEFICIENT (TMPRSS6-/-) MICE

Jan Krijt2, Yuzo Fujikura2, Emanuel Necas2, Andrew J. Ramsay1, Carlos Lopez-Otin1

2Institute of Pathological Physiology and Center for Experimental Hematology, Charles University in Prague, First Faculty of Medicine, Prague, Czech Republic, and 1Departamento de Bioquimica y Biologia Molecular, Facultad de Medicina, Instituto Universitario de Oncologia, Universidad de Oviedo, Oviedo, Spain

Hemojuvelin is a key component of the signaling pathway regulating hepcidin expression in response to body iron status. Matriptase-2, encoded by the TMPRSS6 gene, is a negative regulator of hepcidin expression. It has recently been demonstrated that, under in vitro conditions, matriptase-2 cleaves hemojuvelin protein. The matriptase-2/hemojuvelin interaction implies that, in vivo, matriptase-2 activity is modulated by iron status: Iron deficiency should increase matriptase-2 activity, resulting in decreased membrane hemojuvelin content, decreased Bmp6-dependent signaling, and decreased hepcidin expression. On the other hand, disruption of the Tmprss6 gene should result in increased membrane hemojuvelin content, leading to increased hepcidin expression and iron-refractory iron deficiency anemia.

To test this hypothesis in vivo, we measured liver membrane hemojuvelin levels in Tmprss6-/- mice by a commercially available anti-hemojuvelin antibody. To identify hemojuvelin-specific bands on immunoblots, hemojuvelin-mutant (Hjv-/-) mice were used as negative controls.

In immunoblots of liver samples from Hjv+/+ mice, two hemojuvelin-specific bands were detected at 35 and 20 kDa under reducing conditions. In Hjv-/+ mice, the intensity of the bands was decreased in comparison with Hjv+/+ mice; however, liver hepcidin mRNA content was similar in both groups. Unexpectedly, comparison of the hemojuvelin-specific bands in liver homogenates from Tmprss6+/+ and Tmprss6-/- mice showed decreased, rather than increased, hemojuvelin protein levels in Tmprss6-/- mice. The same results, i.e. decreased intensity of the hemojuvelin-specific bands in Tmprss6-/- mice, were obtained when liver membrane fraction was analyzed instead of whole liver homogenate, or when the samples were run under non-reducing conditions. The obtained data demonstrate that hepcidin expression is not modulated by changes in the amount of hemojuvelin protein at the hepatocyte membrane, since both Hjv-/+ and Hjv+/+ mice display the same level of liver hepcidin mRNA. In addition, the unexpected decrease of liver hemojuvelin protein in Tmprss6-/- mice suggests that the role of matriptase-2 in the regulation of hepcidin expression could be more complex, and not limited to the cleavage of membrane hemojuvelin.

021

THE ROLE OF HEMOJUVELIN IN THE REGULATION OF HEPCIDIN EXPRESSION

Yuzo Fujikura, Jan Krijt, Martin Vokurka, and Emanuel Necas

Institute of Pathological Physiology and Center for Experimental Hematology, First Faculty of Medicine, Charles University in Prague, Czech Republic

Hepatic hepcidin expression is regulated by three partially independent signaling pathways, which respond to iron status, erythropoietic activity and inflammation. The response to iron status is mediated by the Bmp6/hemojuvelin pathway. Hemojuvelin is a GPI-anchored membrane protein which serves as Bmp6 coreceptor. In our studies, we attempted to address the role of hemojuvelin in the control of hepcidin expression.

Using Hjv-/- mice, we have demonstrated that erytropoietic activity decreases hepcidin expression even in the absence of hemojuvelin protein. Hepcidin mRNA levels were decreased by two orders of magnitude in Hjv-/- mice by repeated bleeding, or by treatment with erythropoietin. Prolonged feeding of diets with low iron content to Hjv-/- mice also decreased hepcidin expression. Thus, hemojuvelin protein is not necessary for hepcidin downregulation by erythropoietic activity, and does probably not form a part of the erythropoiesis-related signaling pathway.

It has been suggested that the response of hepcidin to hypoxia is mediated by soluble hemojuvelin, possibly released from skeletal muscle by the protease furin. However, repeated bleeding, resulting in a profound decrease in hematocrit and severe tissue hypoxia, did not change skeletal muscle furin mRNA, arguing against the hypoxia inducible factor (HIF)-dependent regulation of furin activity and soluble hemojuvelin release.

We have previously demonstrated that iron overload or erythropoietin treatment does not change hemojuvelin mRNA. To extend these studies to the protein level as well, we used tissue from Hjv+/+ and Hjv-/- mice to select and characterize a functional anti-Hjv commercial antibody. Analysis of liver homogenates detected Hjv-specific bands at 35 and 20 kDa. The Hjv-specific bands were also seen in homogenates from skeletal and heart muscle, but not from kidney or testis. The 35 and 20 kDa bands were not detected in mouse plasma. Erythropoietin treatment, iron treatment or repeated bleeding did not change liver membrane hemojuvelin protein. Interestingly, in Hjv-/+ mice, the amount of Hjv protein appeared to be slightly decreased compared to Hjv+/+ mice, while there was no difference in hepcidin mRNA content between these groups.

Overall, our results indicate that although a basal level of hemojuvelin protein is necessary for the maintenance of Bmp6-mediated signaling, hepcidin expression is not influenced by changes in membrane hemojuvelin protein levels. These results do not support the concept of that modulation of membrane hemojuvelin protein content would influence hepcidin expression. In addition, we have so far been unable to find evidence for the physiological role of soluble hemojuvelin.

022

EFFECTS OF ERYTHROPOIETIN ON HEPCIDIN-MEDIATED REGULATION OF IRON HOMEOSTASIS IN HUMANS AND RODENTS

Gaetano Cairo1, Valentina Ceresoli1, Victor Diaz2, Elena Gammella1, Max Gassmann2, Domenico Girelli3, Carsten Lundby2, Arianne Monge4, Stephane Moutereau5, Antonello Pietrangelo6, Stefania Recalcati1, Paul Robach7, Michele Samaja8, Paolo Santambrogio9

1Dept. Human Morphol. & Biomed. Sci. and 8Dept. Med., Surg. & Dentistry, Univ. Milan, Italy, 2Inst. Vet. Physiol. And 4Univ. Hosp., Univ. Zurich, Switzerland, 3Dept. Clin. Exp. Med. Univ. Verona, Italy, 5Hop. Henri-Mondor, Creteil, France, 6Dept. Int. Med., Univ. Modena, Italy, 7ENSA, Chamonix, France, 9S. Raffaele Sci. Inst. Milan, Italy

Hepcidin regulates iron homeostasis by harmonizing the needs of cells that consume iron (e.g. erythropoietic cells) with the activity of cells that provide iron (i.e. enterocytes, cells that release stored iron, and macrophages that recycle heme iron). Increased erythropoiesis is associated with a dramatic alteration in iron metabolism in order to satisfy the high demand for iron, a key component of functional haemoglobin. A high degree of erythropoietic activity suppresses hepcidin production, but the signals involved in this regulation are still poorly understood. The administration of erythropoietin (Epo), the main regulator of erythropoiesis, downregulates hepcidin mRNA in mouse liver, an effect possibly mediated by the enhanced erythropoietic activity. Indeed, several molecules have been proposed to play a role in transmitting the iron need of erythropoietic cells to the liver. On the other hand, it has been demonstrated that Epo directly downregulates hepcidin expression in hepatoma cell lines. In addition, increased erythropoiesis requires iron redistribution from storage sites, but the mechanisms and signals regulating iron mobilization are not entirely clear, as muscle iron loss and decreased myoglobin levels were found in humans in which erythropoiesis was triggered by exposure to high altitude but not in those treated with Epo.

In this study, we analyzed iron metabolism in two in vivo models: (a) healthy humans treated for very short times (i.e. before the occurrence of any change in circulating iron levels derived from increased iron consumption in the bone marrow), with different doses of rHuEpo; (b) mice exposed to high Epo levels. In humans, we found a strong and early (24 h) drop in serum hepcidin levels in the absence of significant changes in the levels of GDF15, soluble transferrin receptor, ceruloplasmin and in transferrin saturation. Short time Epo administration did not affect the expression of ferritin H and L subunits, transferrin receptor and ferroportin in muscle biopsies. In mice, liver hepcidin mRNA levels were significantly decreased at 24 h from Epo treatment (but not at 4 h) and dramatically suppressed after 10 days of repeated Epo administration. The increased iron demand for erythropoiesis led to an increase in ferroportin expression and a decrease in L ferritin content in tissues. Interestingly, the decrease in ferritin was two fold higher in the liver than in the skeletal muscle, thus indicating a different iron mobilization from these two tissues. Preliminary results in transgenic mice in which Epo overexpression leads to hematocrit values >0.8 (a model to understand the adaptive mechanisms of iron homeostasis to excessive erythrocytosis), indicate that iron metabolism adequately respond to changes in iron availability or iron demand despite permanently high levels of Epo, thus suggesting that Epo is not a direct modulator of hepcidin-regulated iron homeostasis.

These results give us a better understanding of how the hepcidin-ferroportin interaction functions as mechanistic link between the iron demand of erythroid cells and iron mobilisation from storage sites.

023

BMP6 GENOTYPE, GENDER, AND AGE INFLUENCE HEPCIDIN EXPRESSION AND IRON OVERLOAD

Léon Kautz, Céline Besson-Fournier, Chloé Latour, François Canonne-Hergaux, Marie-Paule Roth and Hélène Coppin

Centre de Physiopathologie de Toulouse Purpan, Inserm U563, Toulouse, France

We have previously shown that, at 7 weeks of age, Bmp6-deficient mice present with marked iron accumulation in liver parenchymal cells, reduced hepcidin expression compared with wild-type mice, and stabilisation of ferroportin at the membrane of enterocytes, hepatocytes and tissue macrophages. This definitely proved the critical role of Bmp6 inthe maintenance of iron homeostasis. Interestingly, although 7 w.o. Bmp6-deficient males and females had about the same amount of liver iron, males had a much stronger down-regulation of hepcidin mRNA than females. This prompted us to examine the phenotype of older Bmp6-deficient mice (8-10 months) and we observed that, in accordance with their lower hepcidin mRNA expression, males had accumulated considerably more iron, compared with their wild-type littermates, than did Bmp6-deficient females. Furthermore, although iron accumulation was strictly restricted to the liver in females, 8-10 month-old males had major iron accumulation in many other tissues such as the pancreas, the heart, and the kidney. The reasons for these important gender differences are currently being investigated. We also sought to determine whether heterozygous loss of Bmp6 function alters systemic iron homeostasis. When analyzed at 7 weeks of age, Bmp6+/– mice showed trends toward increased liver iron content compared with Bmp6+/+ controls. These differences became more evident with age, and were statistically significant at 6 months. These findings suggest that mice cannot fully compensate for heterozygous loss of Bmp6 to achieve normal systemic iron homeostasis. Genetically determined variations in the production of Bmp6 could therefore account for inter-individual differences in hepcidin production and possibly modulate the hemochromatosis phenotype. Bmp6+/– Hfe–/– mice are currently being produced to examine this possibility further.

024

THE LIVER-SPECIFIC MICRORNA-122 CONTROLS SYSTEMIC IRON HOMEOSTASIS

Mirco Castoldi1, Joacim Elmén2, Maja Vujci; Spasic1; Morten Lindow2, Judit Kiss6, Jens Stolte3, Richard Sparla1, Herman J Gröne4, Vladimir Benes3, Sakari Kauppinen5, Matthias W. Hentze3, Martina U. Muckenthaler1

1Department of Pediatric Hematology, Oncology and Immunology University of Heidelberg, Heidelberg, Germany, 2Santaris Pharma, Hørsholm, Denmark, 3EMBL, Heidelberg, Germany, 4DKFZ, Heidelberg, Germany, 5Santaris Pharma, Hørsholm, Denmark, 6University of Heidelberg, Germany

Systemic iron homeostasis is mainly controlled by the liver, which is critical for the synthesis of the peptide hormone hepcidin, the key regulator of duodenal iron absorption and macrophage iron release. Different positive and negative regulatory proteins for hepcidin expression have been identified. Here we show for the first time that a microRNA is important for maintaining tissue iron levels and hepcidin mRNA expression. Efficient and specific depletion of the liver-specific miR-122 by injection of a locked-nucleic-acid (LNA)-modified antimiR in wild-type mice causes systemic iron deficiency, hallmarked by reduced plasma and liver iron levels, mildly impaired hematopoiesis and increased extramedullary erythropoiesis in the spleen. We further show that miR-122 inhibition increases the mRNA expression of genes that control systemic iron levels (Hfe, hemojuvelin, Bmpr1a and hepcidin). Importantly, we identify Hfe and Hfe2 as direct miR-122 targets, explaining the observed phenotypes. Taken together, our data reveal a functional role of miR-122 in maintaining systemic iron homeostasis in mice.

025

COMPARISON OF HEPCIDIN MEASUREMENTS IN BLOOD USING TWO COMMERCIALLY AVAILABLE ELISAS AND LC MS/MS

Stephan Brincat, Sukhvinder S Bansal , Robert C Hider, Jolanta Malyszko, and Iain C Macdougall

King's College London, UK

Background:The reliable quantitation of hepcidin in biological matrices is widely sought after. Several immunoassays have been developed, and a number of these are commercially available. The concern with the immunoassays is that the antibodies also bind to truncated hepcidin (20 and 22 isoforms) as well as intact hepcidin-25. Recently mass spectrometric methods have been reported. In this study we have independently determined hepcidin concentrations using two ELISAs, as well as LC MS/MS in a population of chronic kidney disease patients (both dialysis and non-dialysis).

Methodology: We compared two commercially available ELISAs from DRG, Germany and Bachem, UK with a LC MS/MS assay developed in-house [1]. Hepcidin levels were measured on a total of 281 serum samples, and correlations between the assays were performed. All samples were analysed “blind” with regard to the demographics and clinical details of the patients.

Results:Preliminary results indicate that there is no correlation between the DRG ELISA and the Bachem ELISA or the LC MS/MS assay. However a good correlation was obtained between the Bachem ELISA and the LC MS/MS assay. Both the Bachem ELISA and the LC MS/MS assay correlated well with serum ferritin, in contrast to the DRG ELISA where no such correlation was observed.

Conclusion:Our data indicate some concerns over the validity of the DRG ELISA, but show a good correlation between the Bachem ELISA and the LC MS/MS assay. Further analysis with the three different methods are in progress.[1] Bansal SS, Abbate V, Bomford A, Halket JM, Macdougall IC, Thein SL, Hider RC: Quantitation of hepcidin in serum using ultra-high-pressure liquid chromatography and a linear ion trap mass spectrometer. Rapid Commun Mass Spectrom 2010;24:1251-1259.

026

HEPCIDIN IN CHRONIC HEART FAILURE PATIENTS WITH ANAEMIA

M Busbridge1, M Khan1, RS Chapman1. V Mangla2, J Arnold2

1Clinical Biochemistry, Imperial College Heathcare NHS, Trust, London, UK, 2Dept Gastroentrology, Ealing Hospital Trust, London, UK

Background: Anaemia of inflammation (AI) is the most prevalent type of anaemia seen in chronic heart failure (CHF) patients. Hepcidin a key iron regulatory peptide synthesised in the liver to suppress iron absorption and utilization. Hepcidin synthesis is suppressed by anaemia, hypoxia and erythropoiesis and induced by inflammation. Inflammatory cytokines, such as interleukin-6 (IL-6), increase the synthesis of hepcidin, resulting in AI. In this ongoing study, serum hepcidin was measured in newly diagnosed CHF patients with anaemia, in order to better understand the role of hepcidin in development of anaemia in CHF.

Methods: The aim of the study was to measure hepcidin, haemoglobin (Hb), iron studies (including soluble transferrin receptor (sTfR), erythropoietin (EPO) and IL-6 in stable CHF patients (n=27) compared to a healthy control group (HC) (n=22). Diagnosis of anaemia based upon WHO criteria (Hb <13g/dL in men and <12g/dL in women). Hepcidin was measured using a published immunoassay method. Hb, Iron, Ferritin and transferrin saturation (%TSAT) were measured using standard methods. EPO, sTfR and IL-6 were measured using commercially available immunoassays.

Results: Serum IL-6, EPO and sTfR levels were higher in CHF patients, compared to HC (16.9pg/mL (11.2-25.6) P=<0.0001), (19.3mIU/L (13.7-27.1) P=<0.0001) and (37nmol/L (29.2-46.9) P=<0.0001), respectively. There was no significant difference in serum hepcidin levels between CHF patients and HC, (44.4ng/mL (21.9-90.1) P=0.647). In CHF patients Hepcidin correlated with ferritin, EPO and sTfR (R=0.6, P=0.0013, R=-0.68, P=0.0001 and R=-0.39, P=0.04), respectively. But there was no significant correlation with IL-6 (R=0.16, P=0.43). Multiple regression analysis demonstrated that independant predictors of serum hepcidin included EPO and ferritin but not IL-6.Conclusion: Serum hepcidin levels in CHF patients with anaemia are regulated by iron storage and erythropoiesis and not by inflammatory cytokines such as IL-6. These results suggest that AI is a negligible cause of anaemia in CHF, and that the basis of the anaemia is probably multifactorial in nature.

027

HEPCIDIN IN ANCA-ASSOCIATED VASCULITIS

Petr Přikryl1, Pavel Konopásek2, Jana Vávrová1, Vladimír Tesař2, Emanuel Necas1, Martin Vokurka1

1Institute of Pathological Physiology, First Faculty of Medicine, Charles University, Prague, Czech Republic, 2Department of Nephrology of the First Faculty of Medicine and General Teaching Hospital, Prague, Czech Republic

Hepcidin is supposed to be involved in iron disturbances typical for anemia of chronic disease. This anemia accompanies inflammatory diseases and is thus called also anemia of inflammation (AI). The role of hepcidin was demonstrated by several experimental observations. Its expression is stimulated by inflammatory cytokines, mainly interleukin-6, and through its effects on target cells hepcidin decreases iron availability for erythropoiesis. However, apart from inflammation, hepcidin is also regulated by iron, erythropoiesis and hypoxia, and its concentration may be influenced not only by its production but also by its elimination. This makes the clinical ransferringn more difficult and complex. Limited availability of hepcidin measurement and different methods utilized so far caused the data to be variable and known only for several diseases and situations.ANCA-associated vasculitis (AAV) is a group od inflammatory diseases of vessels which can have significant multisystem manifestations, including the damage of kidneys and other internal organs. Hepcidin expression has not been so far studied in these diseases which combine inflammation and possible decrease of kidney functions. We used Hepcidin-25 human EIA kits (Bachem) for hepcidin ransferring in plasma and urine. We correlated the measured results with several clinical parameters including iron metabolism, kidney function, hemoglobin concentration and inflammation. Apart from that we tested other methods of hepcidin determination including MALDI.The serum concentration of hepcidin in patients with AAV was 92.93 ± 15.37 ng/mL vs. 15.43 ± 2.11 in healthy controls (p<0.01). The concentration of hepcidin correlated distinctly to ransferring saturation and weakly to serum ferritin and tended to decrease with the decline of glomerular filtration rate. No correlation was observed with CRP, serum iron, serum ransferring and hemoglobin concentrations. Supported by grants MSM 0021620806, and LC 06044 from the Ministry of Education of the Czech Republic, and NS10300-3 form the Ministry of Health of the Czech Republic.

028

A MODEL FOR ERYTHROPHAGOCYTOSIS USING IN VIVO AGED RED BLOOD CELLS

Yehonatan Gottlieb1, Orit Topaz1, Eitan Fibach2, Esther Meyron-Holtz1

1Laboratory for Molecular Nutrition, Faculty of Biotechnology and Food Engineering. Technion, Israel, 2Hematology Department, Hadassah - Hebrew University Medical Center, Jerusalem, Israel

Most of the body iron is found in hemoglobin of the red blood cells and approx. 1% of this iron is recycled daily from senescent red blood cells (sRBCs). The lifespan of human and mouse RBCs is around 120 and 45 days respectively and thereafter they are removed from the circulation by macrophages through a process named erythrophagocytosis (EPC). EPC is performed mainly by splenic red pulp macrophages that express the transcription factor SpiC. The mechanism by which macrophages recognize sRBCs has been discussed widely, but not fully elucidated, yet a variety of cellular changes have been reported in sRBCs, including: decreased enzymatic activities, oxidative damage and exposure of phosphatidylserine on the cell surface. Further, the process of EPC by macrophages is still poorly understood. Using serial blood hypertransfusions for six weeks in mice, we created in vivo a RBC population of 50% sRBCs that are at least 6 weeks old. To characterize these sRBCs, biotin-labeled RBCs were used for transfusion, and the mouse RBCs were analyzed by flow cytometry after 6 weeks - comparing the biotinylated RBCs (i.e., older than 6 weeks) to the non-biotinylated RBCs (younger than 6 weeks). Analysis of several known aging parameters, such as ROS levels, phosphatidylserine externalization, esterase activity and membrane CD47 levels, indicated that the enriched sRBCs in the hypertransfused mice had the same characteristics as the naturally aged RBCs (biotinylated RBC in non-transfused mice). We are currently using isolated splenic macrophages to study the EPC of the in vivo generated sRBCs. We believe that the use of an authentic in vivo-generated enriched population of sRBCs will be helpful in getting a better understanding of the EPC mechanism.

029

HEPARIN: A POTENT INHIBITOR OF HEPCIDIN EXPRESSION IN VITRO AND IN VIVO

Maura Poli, Domenico Girelli, Natascia Campostrini, Dario Finazzi, Sara Luscieti, Federica Maccarinelli, Antonella Nai, Paolo Arosio

Hepcidin is a major regulator of iron homeostasis, and its expression in liver is regulated by iron, inflammation and erythropoietic activity with mechanisms that involve bone morphogenetic proteins (BMPs) binding their receptors and co-receptors (Nemeth and Ganz 2009). BMPs were originally isolated from heparin columns and in fact are known to bind heparin with high affinity (Wozney, et al 1988). Here we show that exogenous heparin added to hepatic HepG2 cells strongly inhibits hepcidin and BMP/SMAD signalling in a dose dependent manner. Unfractionated heparin is more effective than low molecular weight heparin, while the pentasaccharide Fondaparinux has little activity. Moreover the inhibition reaches its maximum after 4 h and after washing the cells from heparin the level of hepcidin mRNA slowly increases in 12 h. Cells can be grown in heparin for a week or longer, in conditions that maintain the level of hepcidin transcript below 10% of the basal without evident effects on cell viability or proliferation. Heparin strongly inhibits hepcidin stimulation by BMP6 and IL-6, but is much less effective in inhibiting stimulation by BMP2. Since pharmacological concentrations of heparin are sufficient to inhibit about hundred fold hepcidin expression in HepC2 cells, mice were treated for 5-7 days with subcutaneous injection of heparin. This caused a strong reduction of hepcidin mRNA in the liver and also of the level of phosphorylated SMAD1/5/8. Moreover, we observed a strong reduction of serum hepcidin in five patients treated with heparin to prevent deep vein thrombosis and this was accompanied by an increase of serum iron and a reduction of C-reactive protein levels. The data show that heparin is a potent inhibitor of hepcidin expression in liver and in hepatic cells which acts by sequestering BMP6. This unrecognised role of heparin in regulating iron homeostasis indicates novel approaches to the treatments of anemia of inflammation.

030

URINARY HEPCIDIN: A MARKER OF RENAL TUBULAR DYSFUNCTION

Hilde P Peters, MD1, Coby M M Laarakkers2, Rosalinde Masereeuw3, Dorine W. Swinkels, MD, PhD2 and Jack F. Wetzels, MD1

1Nephrology, Radboud University Nijmegen Medical Center, 2Laboratory of genetic, endocrine and metabolic diseases, Radboud University Nijmegen Medical Center and 3Toxicology and Pharmacology, Radboud University Nijmegen Medical Center, Nijmegen, Netherlands

Introduction: Urinary hepcidin is elevated in various renal diseases. Recent data suggest that urinary hepdicin may serve as a biomarker for kidney disease. However, it is unknown which processes - filtration, reabsorption, local production, and/or degradation - contribute to elevated urinary hepcidin levels. We aimed to assess whether tubular dysfunction affects urinary hepcidin levels, through comparison with beta 2-microglobulin (B2m) excretion.

B2m was measured by ELISA. Gelofusine, an inhibitor of tubular protein reabsorption, was administered to 2 healthy volunteers. Urinary hepcidin was measured in megalin deficient mice and wild type controls.

Results: We studied 21 patients with glomerulopathies, median serum creatinine was 116 (range 70-301) mol/L, median estimated GFR 52 (13-84) ml/min/1.73m2, median proteinuria 7.1 (1.7-19.1) g/d, and fractional excretion of B2m 1.3 (0.1-71)%. Fractional excretion of hepcidin was 2.8 (0.6-43) versus 1.9 (range 0.2-5.8)% in 24 controls (p=0.04). Fractional excretion of hepcidin correlated strongly with fractional excretion of B2m (r=0.85, p<0.01). During Gelofusine infusion a rapid increase in fractional excretion of both B2m and hepcidin was observed; 150 (0.1 16.5%), and 7-fold (3.0 15.4%), respectively. Urinary hepcidin in megalin deficient mice was increased 7-fold B2m is a low molecular weight protein and is filtered by the glomerulus and, in the absence of renal disease, almost completely reabsorbed by the proximal renal tubules.

Method: Serum and urine levels of hepcidin-25 were determined by a MALDI-TOF Mass Spectrometry based assay in healthy controls and in patients with varying degrees of tubular dysfunction due to glomerular disease. Urinary compared to wild type mice (n=5, p<0.01).

Conclusion: Hepcidin-25 is reabsorbed by the renal tubules, a process mediated by megalin. Increased levels of urinary hepcidin may thus largely reflect tubular dysfunction. This should be considered when evaluating urinary hepcidin as a biomarker in patients with kidney disease.

031

A TIME COURSE OF HEPCIDIN RESPONSE TO IRON CHALLENGE IN HFE AND TFR2 HAEMOCHROMATOSIS PATIENTS

Natascia Campostrini1, Domenico Girelli1, Paola Trombini2, Fabiana Busti1, Michela Corbella1, Marco Sandri1, Sara Pelucchi2, Mark Westerman3, Tomas Ganz3,4, Elizabeta Nemeth3,4, Alberto Piperno2 and Clara Camaschella5

1Department of Medicine, University of Verona, Verona, Italy, 2Department of Clinical Medicine, Prevention and Biotechnologies, Milano Bicocca University, Milan, Italy, 3Intrinsic LifeSciences, LLC, La Jolla, CA, 4Departments of Medicine and Pathology, David Geffen School of Medicine, University of California, Los Angeles, CA, USA, 5Vita-Salute University and Division of Genetics and Cell Biology San Raffaele Scientific Institute, Milan, Italy

Inadequate hepcidin production leads to iron overload in nearly all types of haemochromatosis. We explored the acute response of hepcidin to iron challenge in 25 patients with HFE-haemochromatosis, in 2 with TfR2-haemochromatosis and in 13 controls. Sixteen patients (10 C282Y/C282Y, 6 C282Y/H63D compounds heterozygotes) were iron overloaded and 9 (6 C282Y/C282Y, 3 C282Y/H63D) had been iron-depleted. We analyzed serum iron, transferrin saturation, and serum hepcidin by both ELISA and mass-spectrometry at baseline, 4, 8, 12 and 24 hours after a single 65-mg dose of oral iron. Serum iron and transferrin saturation significantly increased at 4 and returned to baseline at 8-12 hours in all groups, except in the iron-depleted patients who showed the highest and longest increase of both parameters. Hepcidin increased significantly at 4 and returned to baseline at 24 hours in controls and in C282Y/H63D compound heterozygotes at diagnosis. Hepcidin response was smaller in C282Y-homozygotes than in controls (p=0.023), barely detectable in iron-depleted HFE- and absent in TfR2-patients. Our results indicate that TfR2 is essential for hepcidin response to acute oral iron loading and HFE contributes to it. In iron-depleted HFE-patients, both the low hepcidin baseline and the weak response to iron contribute to hyperabsorption of iron.

032

INTRA AND EXTRACELLULAR LABILE IRON POOLS: FACTS AND FANCIES

William Breuer and Z. Ioav Cabantchik

The Alexander Silberman Institute of Life Sciences. The Hebrew University of Jerusalem, Givat Ram, Jerusalem 91904, Israel

The pathophysiology of iron has been classically associated with its labile properties. In cells, only a small fraction of redox-active iron is labile, i.e. amenable to chelation or exchangeable, and that is the transitory iron that is at the cross-roads of cellular metabolism. The labile cell iron-LCI, is segregated among pools in various chemical forms that can be exchanged between cell components with the aid of specific ligands/siderophores. LCI constitutes a regulated pool of cell metal, that is controlled by regulatory proteins (IRP) that sense its levels and coordinately control the expression of the iron uptake machinery vis a vis the partners involved in iron utilization/storage and in some cells also export. That cell machinery functions to prevent cell iron from rising to toxic levels and engaging in the promotion of ROS formation and ensuing oxidative damage. LCI can be: a. of cellular origin, resulting from mutations in genes that affect cell iron management and cause it to rise excessively in particular cell compartments (often depleting others) as exemplified in Friedreich ataxia and sideroblastic anemia or b. of systemic origin, resulting from mutations in genes affecting body iron absorption/recycling causing plasma iron to rise, surpass the transferrin (Tf) binding capacity and generate forms not bound to Tf (NTBI) that cause organ iron overload (IO) as in hereditary hemochromatosis or thalassemia intermedia. However, NTBI appears also after repeated blood transfusions that cause IO in the liver, endocrine glands and heart, as in thalassemia major and myelodysplasia. Since plasma NTBI is chemically heterogeneous (bound to organic acids, to proteins directly or via ligands) and of composition that varies with the degree of IO, origin and history of transfusions/phlebotomy and chelation, it is not clear which forms permeate into cells excessively raising LCI to toxic levels As most of the plasma NTBI is protein adsorbed, its uptake into macrophages, cardiomyocytes and endocriynocytes is limited to endocytosis, and only a subfraction comprising labile forms is taken up by resident membrane transporters (opportunistically!) The labile iron forms LPI and LCI are important diagnostically and pharmacologically as they are the direct targets of chelators and as such they can be used as indicators of both iron overload and of treatment efficacy. However, LCI can also serve as indicator of iron deficiency. In this presentation we shall evaluate the advantages and limitations of present methodologies aimed at defining these elusive parameters of increasing clinical relevance

033

BOTH DEGRADATION AND BILIARY EXCRETION ACCOUNT FOR HEMOPEXIN-MEDIATED HEME DETOXIFICATION IN HEPATOCYTES

Francesca Vinchi, Lorenzo Silengo, Fiorella Altruda and Emanuela Tolosano

Molecular Biotechnology Center, Department of Genetics, Biology and Biochemistry, University of Turin, Italy

Intravascular hemolysis is associated to several pathological conditions, including hemoglobinopathies, bacterial infections and malaria, and results in the release of potentially toxic free heme into plasma, where it is rapidly bound by the plasma scavenger Hemopexin. In order to define the contribution of Hemopexin to heme recovery and detoxification and to heme-iron recycling, we analyzed the response of heme-overloaded Hemopexin-null mice compared to Wild-type animals and we focused mainly on isolated macrophages and hepatocytes, the cell types known to be involved in heme handling.Even though plasma heme clearance was quite similar between Wild-type and Hemopexin-null mice, heme was efficiently recovered by the liver only in Wild-type mice; conversely, heme uptake was strongly reduced in the liver of Hemopexin-null mice. To investigate how heme is inactivated in the liver, we evaluated the possibility that it may be degraded by Heme-oxygenase or excreted into the bile. HO activity and biliary excretion of bilirubin were strongly increased in Wild-type mice, soon after heme injection, while their induction was delayed in Hemopexin-null mice. Interestingly, biliary excretion of heme was significantly higher in Wild-type mice compared to Hemopexin-null animals. Heme excretion occurred with the same kinetics of bilirubin but we found an inverse correlation between biliary excretion of bilirubin and heme: mice that excreted more bilirubin showed a reduced excretion of heme and viceversa, suggesting that an interplay between these detoxification mechanisms really exists. We performed a “rescue experiment” by injecting Hemopexin in knock-out mice before heme overload and we were able to fully recover hepatic heme uptake, as well as biliary excretion of heme and bilirubin in Hemopexin-null mice, further demonstrating the high detoxifying potential of this protein.Then, we asked what happens to heme excess in Hemopexin-deficient mice and we demonstrated that heme is recovered mainly by hepatic and splenic macrophages and, at later time, by extra-hepatic tissues, as kidney and duodenum. We found that HO-1 Mrna and protein induction were significantly higher in Hemopexin-null macrophages compared to Wild-type ones. The same occurred for L-Ferritin and Ferroportin. In vivo data were confirmed by using heme-treated Raw 264.7 macrophages. In these cells heme recovery was significantly higher after treatment with heme-Albumin than heme-Hemopexin, suggesting that Hemopexin prevents heme uptake by macrophages. Accordingly, HO-1 expression, as well as Ferritins and Ferroportin, were strongly induced by heme-Albumin, but not by heme-Hemopexin. In this study we provide evidence that the specific function of Hemopexin is to mediate heme uptake in hepatocytes, thus promoting its detoxification. Furthermore, these data suggest that, besides the canonical pathway of heme degradation, a mechanism of excretion in bile gives its contribution to heme detoxification, thus providing a way for iron to leave the body.Finally, these findings may have relevance for pathologic conditions associated to massive hemolysis, when circulating Hemopexin is saturated: the administration of exogenous Hemopexin may be used to promote heme detoxification by hepatocytes and to counteract the deleterious effects of heme, mainly preventing macrophage overload and activation.

034

THERAPEUTIC ERYTHROCYTAPHERESIS VERSUS PHLEBOTOMY IN THE TREATMENT OF HEMOCHROMATOSIS PATIENTS. RESULTS OF THE COST ANALYSIS FROM A RANDOMISED, FEASIBILITY AND PROOF OF PRINCIPLE TRIAL

E. Rombout-Sestrienkova1, B.A.B. Essers2, P.A.H. van Noord1, M.C. H. Janssen3, C.Th. B. M. van Deursen4, L.B. Bos5, G.H. Koek2

1Sanquin Blood Bank, Maastricht, The Netherlands, 2Maastricht University Medical Centre, The Netherlands, 3University Hospital St Radboud Nijmegen, The Netherlands, 4Atrium Medical Centre, Heerlen/Brunssum,The Netherlands, 5Orbis Medical Centre, Sittard, The Netherlands, 2Maastricht University Medical Centre, The Netherlands

Background: Phlebotomy (P) is currently the standard therapy for patients with hereditary hemochromatosis (HH). Per procedure 500 ml whole blood (250 ml erythrocytes) is removed until the serum ferritin level (SF) is 50 μg/L. Depending on the initial SF levels, this requires 20-100 P's over a period of 6 to 24 months. Recently therapeutic erythrocytapheresis (TE) has become a new therapeutic modality, with which. up to 1000 ml erythrocytes per procedure can be removed. The aim of our study was to compare TE and P in reducing the number of treatment procedures, the duration of therapy and costs.

Methods: A prospective randomized clinical trial among naïve HH patients (C282Y homozygous). Statistical analysis: Univariate analysis by two-sided t-test. Adjustments for observed differences in initial ferritin level were performed by logistic regression analysis. Uncertainty intervals (2.5 th and 97.5 th percentiles) for the mean differential costs were obtained by the bootstrap method. Cost-analysis in Euro for 2009: costs related to the treatment and loss of productivity were collected. Direct costs of both procedures contained personnel, material and equipment costs, calculated by the Sanquin Blood Bank. Costs resulting from productivity losses were calculated according to the friction cost method. Overhead costs related to the Blood bank were allocated to the direct costs as an overall percentage of 50%. No discounting was applied since all costs occurred within one year. Results: 38 patients (19 in each treatment) were included. The baseline characteristics were similar among the treatment groups except for the baseline SF levels which were significantly lower in the TE group (mean 1104 microgram/l; SD 677 versus mean 1666 microgram/l; SD 912 in the P group). Univariate t-test showed a reduction of the number of treatments with 66 % (9 vs 27; TE vs P, p< = 0.05) in the TE group. Multivariate logistic regression analysis showed a decline in the reduction factor for the number of treatments to 2.17 (CI 95% 1.23–3.80) when adjusted for the difference in baseline serum ferritin levels and a reduction factor of 2.60 (CI 95% 1.1-6.14) when also body weight was include in the model.A similar but smaller reduction factor (1.7 CI 95% 1.65–1.76) was found for treatment duration in weeks; TE required on average 20 weeks (range 4–20) versus 34 weeks (range 12–79) in the P group. Costs: Although the costs for single procedure are higher for TE (€ 251 vs € 71) the treatment costs are comparable and not significantly different (P: €1.898 versus €2.263 for TE). The total treatment costs for P €4.438 € (SD: €599) were higher compared to TE €3005 (SD: €444), although the difference is not statistically significant. The total costs consist of the treatment costs for P and TE and costs resulting from productivity loss.

Conclusions: With TE, a 54%–66% reduction in the total number of therapeutic procedures, an 11%–42% reduction in treatment duration and a 32% reduction in total costs is achieved. From a societal perspective TE might not only be an effective but also a cost-saving treatment.

035

QUANTITATIVE MAGNETIC DETERMINATION OF FERRITIN IRON IN DIFFERENT ORGANS IN AN IRON OVERLOAD HEREDITARY HEMOCHROMATOSIS MOUSE MODEL

L. Gutiérrez1, M. Vujic Spasic2, M.U. Muckenthaler2 and F.J. Lázaro3

1Departamento de Biomateriales y Materiales Bioinspirados, Instituto de Ciencia de Materiales de Madrid (CSIC), Cantoblanco, 28049 Madrid, Spain, 2University Hospital of Heidelberg, Germany, 3Departamento de Ciencia y Tecnología de Materiales y Fluidos, Universidad de Zaragoza, 50018 Zaragoza, Spain

Iron overload conditions include genetic disorders (e.g. haemochromatosis) or regional iron overload (e.g. viral hepatitis, metabolic syndrome, or neurodegeneration). Hereditary hemochromatosis (HH) is a prevalent, potentially fatal disorder of iron metabolism hallmarked by intestinal hyperabsorption of iron, hyperferremia, and liver iron overload. Impaired regulation of hepcidin expression in response to iron loading appears to be the central pathogenic mechanism for HH. Once iron enters the cell, the portion that is not immediately needed is predominantly stored in ferritin. Degradation of ferritin and concomitant iron release helps to mobilize iron for cellular utilization. As the stores enlarge an increasing proportion is held as insoluble, stainable haemosiderin. We quantify ferritin iron in tissues by a recently developed method that includes measuring the AC (alternating current) magnetic susceptibility followed by analysis of the temperature dependence of its out-of-phase component, using an iron loaded ferritin standard. Additionally, the in-phase component of the susceptibility informs on other iron species as heme-type, or other mineralised iron-containing deposits larger than the ferritin crystallite cores. We investigate chemical iron speciation in health and iron-overload conditions as Hfe-HH. We used age- and sex-matched Hfe-/- mice, raised on a C57BL/6J genetic background. Initially liver, spleen, heart, duodenum and kidney of Hfe-/- mutant mice have been magnetically characterised, in addition to atomic emission spectroscopy, and also with Prussian blue staining to determine the non-heme iron. The magnetic data revealed that i) ferritin iron is the predominant form of iron in these tissues, ii) the livers of Hfe-/- mice are highly iron overloaded with ferritin iron at concentration of 2.5 mgFe/g dry tissue, in contrast to the levels in control, Wt mice (0.25 mgFe/g dry tissue), and iii) the spleens of Hfe-/- mice are iron spared in comparison to control, Wt mice (2.5 mgFe/g in Hfe-/- and 8 mgFe/g dry tissue in Wt mice, respectively). Although in Hfe-HH the liver is the most affected, over time iron gets deposited in other tissues (heart, kidney and brain) eventually leading to organ failure disorders. Using the magnetic method, we demonstrate here that heart, duodenum and kidney of Hfe-/- mice contain significant amount of ferritin iron (0.6, 0.5 and 0.3 mgFe/g dry tissue respectively), hence at levels significantly less then in liver and spleen of Hfe-/- mice. With magnetic methods we aim to provide evidences on the iron-speciation in other tissues (brain, pancreas, lungs, etc.) affected by iron deregulation in Hfe-HH conditions.

036

CD8+ T LYMPHOCYTES, THE MHC-CLASS I REGION AND THE CLINICAL EXPRESSION IN HH: STRATEGIES TO FIND NEW GENETIC MODIFIERS

Graça Porto1,2,3, Sandra Morais1, Mónica Costa2, Eugénia Cruz1,2

1CHP – Santo António Hospital, Porto, Portugal, 2IBMC, Institute for Molecular and Cell Biology, Porto, Portugal 3ICBAS, Abel Salazar Institute for the Biomedical Sciences, Porto University, Portugal

The numbers of CD8+ T lymphocytes (CD8N) are known to be genetically determined in association with genes at the MHC-class I region both in patients with Hereditary Hemochromatosis (HH) (Cruz et al.Tissue Antigens,64:25-34,2004) and in normal controls (Ferreira et al.Am J Hum Genet,86:88-92,2010). A question still remains if the allele variability in HLA genes in combination with the mutated HFE is sufficient to explain the patients' phenotypic heterogeneity or if there is another genetic trait in this chromosomal region implicated in the setting of CD8N. Evidence that cells from C282Y homozygous patients have reduced MHC-class I expression triggered by an UPR (De Almeida et al.J Immunol,178:3612-9,2007) supports an interaction between the C282Y mutation and the MHC for antigen presentation and lymphocyte activation. Nevertheless, the remarkable linkage disequilibrium and the high density of immune response related genes at the chromosomal region between HLA and HFE, makes the alternative hypothesis of a novel putative trait involved in the setting of CD8N highly attractive. We have been studying for sometime the contribution of MHC-class I genes in the homeostatic regulation of CD8N using HH as a model. In a previous study we identified an extremely conserved region of 500 Kb including the genes PGBD1,ZNF193,ZNF165 where two distinct haplotypes were associated with the inheritance of “low” or “high” CD8N (respectively AAT and GGG) and predicted the development of a severe or mild clinical expression respectively (Cruz et al.BMC Med Genet,9:97,2008). These haplotypes do not explain, however, all the variability in CD8N. 33,3% of AAT homozygous patients have high CD8N, showing that other markers in the region are still necessary to clarify the question. In a more recent study we described, for the first time, the inheritance of CD8N in the context of “extended” haplotypes defined between HFE and HLA-B (markers: PGBD1,ZNF193,ZNF165,HLA-A,HLA-B) by segregation analysis in HH families. From a total of 71 unrelated C282Y homozygous HH patients and 61 family members we defined 147 extended haplotypes for correlation with the CD8N. We confirmed the strong association of the most common A3B7-AAT ancestral haplotype with the inheritance of low CD8N. Other conserved haplotypes (A2B7-AAT;A3B40-AAT) were also associated with low CD8N. In contrast, a greater allelic and haplotypic variability was found in chromosomes associated with the high CD8N, supporting the hypothesis of a divergent recombination history and pointing to a possible localization of the putative trait associated with CD8N centromeric to the HLA-A locus.

037

ALTERNATIVE SPLICING REGULATES HFE GENE EXPRESSION IN HUMAN TISSUES

Rute Martins*, Bruno Silva*, Daniela Proença, Paula Faustino

Departamento de Genética, Instituto Nacional de Saúde Dr. Ricardo Jorge, Lisboa, Portugal *These authors contributed equally for this work

Background/aims: Hereditary hemochromatosis (HH) is a disorder characterized by massive iron overload, mainly due to mutations in HFE gene. This gene codes for a protein that associates with beta-2 microglobulin (b2M) for trafficking to the cell surface. However, the pathophysiology of HFE-derived HH and function of the full-length HFE protein in iron homeostasis remain uncertain. Also, the role of alternative splicing in HFE gene expression regulation and possible function of the correspondent protein isoforms remain elusive

Methods:Total RNA from eight human tissues was used to synthesise cDNA and therefore identify alternative spliced HFE transcripts by RT-PCR, cloning and sequencing. Total HFE transcripts, as well as two alternative splicing transcripts were studied using a real-time quantitative PCR methodology. By immunofluorescence and immunoprecipitation assays we have also analysed the intracellular localization, trafficking and protein association of GFP-tagged HFE protein variants in HepG2 cells.

Results:We have identified several alternative spliced HFE transcripts resulting from exon skipping or total and partial intron inclusion. These varies both in level- and tissue-specificity. Concerning the exon 2 skipping and intron 4 inclusion transcripts, the liver presents the lowest relative level, while duodenum presents one of the highest relative amounts. At the protein level, while HFE full-length associates with b2M and TfR1, the protein resulting from the exon 2 skipping transcript is unable to bind these proteins and present itself at the cellular membrane. Furthermore, it co-localizes with calnexin, revealing an ER retention and possible degradation. Conversely, the inclusion of intron 4 transcript gives rise to a truncated, soluble protein (sHFE) that is not able to associate with TfR1 at the cell surface under normal iron status but that is mostly secreted by cells to the medium in association with b2M. Presently we are analysing if the level of the sHFE mRNA varies in HepG2 and HuTu80 cells under increasing iron concentrations.

Conclusions:Since the precise function of HFE is still unknown, through this study we might find a clue to the contribution of HFE splice variants to the maintenance of iron homeostasis. sHFE may be secreted into the bloodstream and act in remote tissues being an agonist or antagonist of full-length HFE; acting in liver as an hepcidin expression activator and/or regulating dietary iron absorption in duodenum. Therefore we look forward to a sHFE being useful as a therapeutic agent in treatment of iron overload disorders. Partially funded by FCT: PTDC/SAU-GMG/64494/2006; CIGMH; SFRH/BD/21340/2005 and SFRH/BD/60718/2009

038

HFE-DEPENDENT IRON TRANSPORT ACROSS PLACENTA

Sara Balesaria, Rumeza Hanif, Kishor Raja, Harry McArdle, Kaila Srai

University College London, UK

Iron transfer from mother to foetus is a regulated process involving uptake of iron from the maternal circulation, transport across the placenta and subsequent transfer to the foetus.HFE functions as an upstream regulator of liver hepcidin which has been demonstrated to be a negative regulator of intestinal dietary iron absorption and efflux of recycled iron from macrophages. Hepcidin has also been proposed to be a negative regulator of iron efflux from placenta, however it is not known if this is of maternal or foetal origin.Furthermore, HFE has been demonstrated to be present in the placenta, but its role in the dynamics of maternal-foetal iron transfer, independent of hepcidin is unknown. In this study we investigated the effects of HFE and dietary iron levels on transfer of iron from mother to foetus in order to determine the importance of maternal and foetal HFE status.HFE knockout (KO), wild type (WT), and heterozygote (Het) dams were fed 25, 50 and 150ppm iron diets and mated with Het males to produce pups of all genotypes. Dams and pups were sacrificed; pup total iron and non-haem liver iron levels of pups and dams were determined. mRNA levels of various iron transporter genes were determined in placental tissue by real-time PCR.Maternal liver iron levels were dependent on both dietary iron intake and HFE status with KO dams having highest liver iron levels of all genotypes. Loss of maternal HFE resulted in increased foetal liver iron storage but only when dietary iron levels were adequate. Loss of HFE in the foetus resulted in increased iron transport across placenta indicating that the ability to regulate iron transport is lost in these animals. In all cases, increasing levels of iron in the maternal diet resulted in enhanced total iron in the foetus in addition to elevated liver iron storage. Expression of iron transport genes in the placenta differed only when maternal dietary iron was high, this difference being genotype specific. It appears that pups derived from KO dams increase expression of iron transporter genes and this is also true for KO pups. Taken together our data shows that maternal genotype plays a role in regulating placental iron transfer across placenta. Furthermore foetal genotype also appears to play a role in regulating iron transport, possibly due to hepcidin regulation in foetal liver. It also appears that maternal iron ingestion is essential in maintaining adequate iron stores during pregnancy.

039

EARLY NON-IRE FPN1 MRNA INDUCTION BY HEME IN ERYTHROID CELLS

Luciano Cianetti, Gianna Vespa, Cristina Nodale, Patrizia Segnalini, Marco Gabbianelli, Nadia Maria Sposi

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

Background: Ferroportin (FPN1) has been reported to play a critical role in iron homeostasis, and the regulation of its expression has been shown to relay on transcriptional, post-transcriptional and post-translational mechanisms in response to various stimuli in different cell types. We previously demonstrated the presence of FPN1 mRNA and protein in human erythroid differentiating cells and the existence of two alternative FPN1 transcripts (variant II and III), other than the IRE-containing canonical one, that do not contain the IRE element in their 5′-UT region, do not respond to iron treatment and together account for more than half of the total FPN1 mRNA present in erythroid cells. These transcripts are expressed mainly during the middle steps corresponding to 4-11 days of “in vitro” erythroid differentiation, i. E. the time in which erythroid progenitor/precursor cells need to accumulate iron within the cells. So we speculate that heme may be involved in the transcriptional regulation of FPN1 by choosing non-IRE FPN1 alternative promoters during distinct stages of erythroid cell maturation. In support to this hypothesis some authors have demonstrated that heme derived from human or murine red blood cells or from an exogenous source of heme leads to marked transcriptional activation of the FPN1 and HO1 genes.Aims. The aim of this study was to gain insight into the regulatory mechanisms that control the transcriptional response of FPN1 to heme in erythroid cells.

Methods. As an experimental cellular model we used human erythroleukemic K562 cells treated for various times with hemin and/or heme synthesis inhibitor (succinylacetone-SIGMA). Gamma-globin was used as erythroid differentiation marker. The expression of non-IRE FPN1 alternative transcripts were analyzed by real-time PCR. The effect of treatment with hemin was evaluated by luciferase assay in K562 cells using constructs carrying the different promoter sequences specific for FPN1 alternative transcripts. FPN1 protein from total cellular extracts and in cytosolic and membrane fractions were analysed by Western blot. Possible regulative elements located in genomic DNA upstream of the FPN1 transcription start site were searched for by bioinformatic analysis.

Results:. We obtained the following results: a) modulation of FPN1 protein was hemin concentration dependent; b) we observed the FPN1 induction as soon as thirty minutes after treatment; c) FPN1 protein progressively decreased in presence of an heme synthesis inhibitor; d) finally we observed an early induction of non-IRE FPN1 alternative transcripts with an apparent sequential and specific activation and mutual exclusion between the IRE and non-IRE transcripts especially in the early times of exposition; e) only the non-IRE FPN1 transcript is translationally regulated by heme; f) sequence analysis of the non-IRE FPN1 gene promoter revealed the presence of a putative Maf recognition element (MARE).

Conclusions. In conclusion our results showed that in erythroid cells FPN1 gene was regulated by heme levels through non-IRE FPN1 alternative transcript induction. This regulation is of particular interest with regard to disorders such as thalassemias where the ineffective erythropoiesis and hemolysis cause severe anaemia.

040

RELATIONSHIPS BETWEEN HLA-TYPE, T-LYMPHOCYTE SUBSETS, MHC-I LINKED HAPLOTYPES AND THE EXPRESSION OF IRON OVERLOAD IN C282Y HOMOZYGOUS INDIVIDUALS FOUND BY POPULATION SCREENING

Ketil Thorstensen1, Mona Kvitland1, Wenche Irgens1, Arne Åsberg1, Berit Borch-Iohnsen2, Torolf Moen3, Kristian Hveem4,

1Dept. of Medical Biochemistry, St. Olav Hospital, Trondheim, Norway, 2Institute of Basic Medical Sciences, Department of Nutrition, University of Oslo, Norway, 3Dept. of Laboratory Medicine, Children's and Women's Health, Faculty of Medicine, Norwegian University of Science and Technology Trondheim, Norway, 4HUNT Research Centre, Dept. of Public Health and General Practice, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway

Background: A common feature of C282Y homozygotes identified through population screening is the relatively low iron load as determined by their serum ferritin levels (Scand. J. Gastroenterol. 2001; 36, 1108). The proportion of C282Y homozygotes that will progress to life-threatening complications of the disease is unknown, and it is unclear which patients need closest follow up. Thus, the finding of additional, modifying factors that could help predict the course of the disease may lead to a more targeted treatment. Previous work on T-lymphocyte subsets (J. Hepatol. 1998; 28, 1) has indicated that abnormally high CD4+/CD8+ ratios due to low peripheral blood CD8+ counts are associated with more severe forms of iron overload and with faster re-entry of iron after treatment. Recently, two conserved haplotypes defined by the SNP markers PGBD1, ZNF193 and ZNF165 were found associated with the clinical expression of hemochromatosis (BMC Med. Gen. 2008; 9, 97). The major haplotype, designated A-A-T, was characterized by low CD8+ T-lymphocyte numbers and severe iron overload, whereas the minor haplotype, designated G-G-G, was associated with high CD8+ numbers and lower iron stores. Furthermore, HLA-type has previously been reported to modify the degree of iron accumulation (Eur. J. Haematol. 2007; 79, 429; ibid 2009; 84, 145). We have examined these factors in C282Y homozygotes found through a hemochromatosis screening program.

Methods: A total of 529 screening-detected individuals were invited to participate in the study, and 333 subjects, of which 182 (103 men , 79 women) were C282Y homozygous, responded positively. HLA-type and T-lymphocyte subsets (CD4+, CD8+, total T-lymphocytes) was determined in all C282Y homozygous individuals, whereas the three SNPs PGBD1 (rs1997660), ZNF193 (rs7206) and ZNF165 (rs203878) have been determined in 161 individuals (89 men, 72 women). The newly obtained data were compared with data obtained in the original screening study.

Results: C282Y homozygous men, but not women, carrying two HLA-A3 alleles had significantly increased iron levels, as represented by serum ferritin at the time of diagnosis, compared to those carrying one or no A3 allele. The major SNP haplotype (A-A-T) showed a tendency (not statistically significant) of higher ferritin levels compared to the minor haplotype (G-G-G). In addition, the A-A-T haplotype was associated with HLA-type A3,B7 or A3,B14, whereas the minor haplotype (G-G-G) was found predominantly in HLA-A1,B8 individuals. The frequencies of the minor alleles were PGBD1-G 16.5%, ZNF193-G 23.0% and ZNF165-G 22.7%, yielding homozygote frequencies of 2.7%, 5.3% and 5.1%, respectively, assuming Hardy-Weinberg equilibrium.

041

IRON AND DISEASE IN DIABETES: PROTEIN GLYCATION AND THE PATHOLOGIES OF ADVANCED DIABETES

White, D. Collard, K

University of Plymouth, UK

Disordered iron homeostasis has been associated with risk for type 2 diabetes, although the specific mechanisms involved are not clear. Haemochromatosis – a genetic iron overload condition – commonly results in diabetes, and is suggested to be the result of oxidative stress to β-cells following pancreatic iron deposition. In established diabetes, studies into the effects of iron depletion have found improvements in insulin secretion and sensitivity, coronary artery responses, and endothelial dysfunction. Significant improvements in kidney function have been demonstrated in microalbuminuric type 2 diabetes patients, following reduction of dietary iron. High plasma glucose levels, as seen in diabetes, cause increase in rate of protein glycation reactions. This can be the result of either the Amadori reaction between acyclic monosaccharides and protein amine groups, such as measured on haemoglobin HbA1c, or by autoxidation of cyclic monosaccharides, which may be catalysed by trace amounts of transition metals. Continuing reactions, also vulnerable to catalysis by iron, and contributed to by dicarbonyl products of glucose and lipid metabolism, lead to the formation of irreversible advanced glycation end products (AGE), and their degraded species. Over time, AGE in the extracellular matrix develop cross-links, causing thickening and stiffening of tissue. Pro-inflammatory responses further contribute to oxidative stress and subsequent irreversible damage in a range of tissues such as blood vessels, kidney basement membrane, and the heart, developing into the endothelial and vascular complications associated with advanced diabetes.Evidence is accumulating that glycation of a number of iron-binding and antioxidant proteins impairs their function, and therefore the potential exists for a vicious circle whereby redox available iron may contribute to glycoxidation damage and further reduction in iron-binding antioxidant capacity. Whilst glucose toxicity clearly drives AGE development, iron may act to amplify effects and consolidate damage. Chelatable loosely-bound plasma iron, iron; oxidant/antioxidant status; protein glycation; nephropathy; cardiopathy; and retinopathy will be assessed in diabetes; impaired glucose tolerance; and healthy controls. Associations between iron and the pathologies of advanced diabetes will be explored in this study.

042

EXPRESSION OF IRON TRANSPORT MOLECULES IN DUODENAL MUCOSA OF PATIENTS WITH CHRONIC LIVER DISEASE

Marketa Dostalikova1, Karolina Kratka2, Kamila Balusikova1, Jitka Chmelikova1, Jitka Neubauerova1, Vaclav Hejda3, Jan Hnanicek4, Jiri Horak2, Jan Kovar1

1Department of Cell and Molecular Biology & Center for Research of Diabetes, Metabolism and Nutrition, Third Faculty of Medicine, Charles University in Prague, Prague, Czech Republic, 2Department of Internal Medicine I, Third Faculty of Medicine, Charles University, Prague, Czech Republic, 3The University Hospital in Pilsen, Czech Republic, 4Department of Internal Medicine II, Third Faculty of Medicine, Charles University, Prague, Czech Republic

Iron and its uptake and transport within the body are very closely related to several serious diseases resulting without treatment in very damaging sequels. Therefore, in the present study we examined the gene expression of duodenal iron transport molecules and hepcidin in patients with hemochromatosis (treated and untreated and with different genotypes which represent risk for I. type hemochromatosis), in patients with alcoholic liver disease and iron deficiency anemia. Gene expressions of DMT1, ferroportin, Dcytb, hephaestin, HFE and transferrin receptor 1 were measured in duodenal biopsies using real-time PCR and Western blot analysis. Hepcidin levels were measured employed ELISA assesment. Major mutations in HFE gene responsible for I. type hemochromatosis, i.e. C282Y, H63D and S65C, were analyzed by PCR-RFLP method. On mRNA level, all analyzed molecules were not significantly different in hemochromatic patients with different genotypes or in untreated hemochromatic patients compared with controls. DMT1, ferroportin, and transferrin receptor 1 levels were significantly increased in post-phlebotomized hemochromatic patients and also in patients with alcoholic liver disease and normal iron indices relative to controls. All tested molecules were significantly increased in patients with iron deficiency anemia compared with controls. Protein expression of analyzed molecules was not significantly different in any of the examined groups when compared with controls. Spearman rank correlations showed that DMT1 and Ferroportin, Dcytb and hephaestin mRNA and ferroportin and Dcytb proteins are positively correlated to each other regardless of the underlying disease. Serum ferritin was negatively correlated to Dcytb, hephaestin, DMT1, ferroportin, and transferrin receptor 1 mRNA but not to proteins in studied groups of patients. Serum hepcidin was significantly decreased in patients with alcoholic liver disease when compared to controls. The decrease was observed in patients with iron deficiency anemia as well, but this was not statistically significant. Our data show that untreated hemochromatic patients do not have increased iron transport molecule mRNAs. Post-phlebotomized hemochromatic patients had increased DMT1 and ferroportin levels in response to increased erythropoesis after phlebotomy. Observed positive correlation among Dcytb, hephaestin, DMT1, ferroportin, and transferrin receptor 1 mRNA indicates the presence of coordinated regulation of these genes. On the other hand, the changes on mRNA level in patients with iron deficiency anemia are probably a part of physiological compensation of iron deficiency in organism.

This study is supported by research program of 3rd Faculty of Medicine, Charles University in Prague MSM0021620814 and SVV grant No. 260705.

043

IS ACTIVATION OF HAEM OXYGENASE – 1 INVOLVED IN IRON ACCUMULATION IN ALCOHOLIC LIVER DISEASE?

Joe Varghese, Subhosmito Chakraborthy, Molly Jacob

Department of Biochemistry, Christian Medical College, Vellore, INDIA

Introduction: The progression of reversible alcoholic steatohepatitis to irreversible cirrhosis of the liver is often associated with accumulation of iron. However, the mechanisms by which iron accumulates in the liver in chronic alcoholism are not clearly known. Alcohol ingestion is known to induce a state of oxidative stress in the liver secondary to up-regulation of cytochrome P450 2E1 (CYP2E1). Oxidative stress is known to activate haem oxygenase 1 (HO-1). We hypothesized that alcohol-induced oxidative stress may lead to induction of HO-1 leading to increased haem degradation and subsequent release of ferrous iron (Fe2+), contributing to iron overload in this condition. This study was designed to test the above hypothesis.

Methodology:Male Swiss albino mice (weighing 28–30g) were pair-fed with commercially available Lieber DeCarli alcohol and control diets for 1, 2, 4 and 8 weeks. Alcohol-fed mice received 20% of total calories in the form of alcohol. Daily alcohol consumption by alcohol-fed mice and weights of mice were monitored. They were sacrificed at the end of the study periods and the liver was retrieved. Microsomes isolated from liver tissue were used for western blots for CYP2E1 and HO-1. The activities of CYP2E1 and HO-1 were estimated by spectrophotometric assays. Levels of serum and liver iron are currently being measured. Data from alcohol-fed mice and pair-fed controls were compared.

Results and discussion:All alcohol-fed mice consumed between 12 to 15 grams of alcohol per kg per day. There were no significant differences in weight gain/loss between the alcohol-fed mice and their pair-fed controls. CYP2E1 enzyme activity was found to be significantly increased in alcohol-fed mice at 1, 2, 4 and 8 weeks, with associated significant increases in protein levels of the enzyme at 4 and 8 weeks. No significant induction of HO-1 (either in protein content or enzyme activity) was seen following 1, 2, and 4 weeks of alcohol feeding. Mild induction was seen at 8 weeks; however this increase was not statistically significant. Results on iron levels in serum and blood are awaited. These will be correlated with the other data obtained.

Conclusions: The preliminary results of this study show that alcohol significantly induced CYP2E1 in the liver of alcohol-fed mice but failed to significantly induce HO-1 up to 8 weeks of alcohol feeding, when alcohol constituted 20% of the calorie intake. Further work in this study will be required to confirm these findings. Acknowledgment: Department of Biotechnology, Government of India

044

THE FERROPORTIN (SLC40A1) GENE IS SIGNIFICANTLY ASSOCIATED WITH HIGHER RISK OF ALZHEIMER'S DISEASE IN A PORTUGUESE POPULATION

Ângela C. Crespo1,2, Liliana Marques1, Erica Marcelino3, Carolina Maruta4, Sónia Costa5, Ângela Timóteo5, Frederico Simões Couto3, Ana Herrero5, Ana Verdelho3, Manuela Guerreiro4, Cristina Costa5, Alexandre de Mendonça3, Madalena Martins2,3 and Luciana Costa3

1Molecular and Cellular Immunology, Dept. Health Promotion and Chronic Diseases, National Institute of Health Dr. Ricardo Jorge, Lisbon, Portugal, 2Instituto Gulbenkian de Ciência, Oeiras, Portugal, 3Neurological Clinical Research Unit, Instituto de Medicina Molecular, Lisbon, Portugal, 4Language Studies Laboratory, Instituto de Medicina Molecular, Lisbon, Portugal, 5Neurology Service, Fernando Fonseca Hospital, Amadora, Portugal

Alzheimer's disease (AD) is the most frequent neurodegenerative disorder worldwide. The distinction between normal aging and AD is a relevant step to combat this disease efficiently. Thus, the identification of biomarkers and genetic factors underlying AD pathology is extremely important. Oxidative injury in the brain, mediated by the imbalance of redox-active metals as iron (Fe) and copper (Cu) has been recognized to contribute to the pathology of AD. In this context, we further investigated this hypothesis by: (I) comparing serum biochemical markers of Fe/Cu metabolism in a sample of 73 AD patients and 60 healthy controls; (II) testing in the same sample a set of Fe/Cu metabolism-related genes and APOE for association with AD. Genetic analysis was performed through high density SNP genotyping of the candidate genes CP, CYBRD1, HAMP, HFE, IREB1, IREB2, SLC11A2, SLC40A1, TF, TFR2, FB19, CALR, and APOE. Biochemical analysis was assessed for: serum Fe, total Fe binding capacity, transferrin levels, transferrin saturation, serum ferritin, serum ceruloplasmin and its oxidase activity. Significant differences were found between female AD patients and controls for serum Fe concentration (71.02 ± 22.34 μg/dL and 86.38 ± 20.80 μg/dL, respectively, p=0.001) and transferrin saturation (22.29 ± 7.89 % and 26.21 ± 6.24 %, respectively, p=0.007). A significant association with AD was found for TF – transferrin gene (p=0.0082) and for the first time for SLC40A1 – ferroportin (Fpn) gene (p=0.0355). APOE&#949;4 was also significantly associated with AD (p=0.0004), in agreement with previous studies. We hypothesize that the lower serum Fe concentration observed in AD patients can be due to impaired Fe excretion from cells, since Fpn is the only known Fe exporter in mammalian cells. The intracellular accumulation of Fe, particularly in the brain where Fpn is also expressed, would lead to a rise in oxidative damage, contributing to the AD physiopathology. Further research is demanded in a greater sample to confirm the results obtained in this pilot study. Noteworthy, an integrative approach was followed to deal with heterogeneity in this complex disorder, and new directions were raised related to the study of the involvement of Fe metabolism in AD.

045

DOES HAEMOCHROMATOSIS AFFECT THE NEUROMUSCULAR SYSTEM? A CLINICAL PERSPECTIVE

S.F. Wouthuis, MD1, C.Th.B.M. van Deursen, MD, PhD3, M.P. te Lintelo, MD1, C.A.M. Rozeman, MD, PhD2, R. Beekman MD, PhD1

1Departments of Neurology, Atrium Medical Centre, Heerlen, The Netherlands, 2Neurophysiology, Atrium Medical Centre, Heerlen, The Netherlands , 3Internal Medicine and Gastroenterology, Atrium Medical Centre, Heerlen, The Netherlands

Objectives: Involvement of peripheral nerves and skeletal muscles has been reported in the course of hereditary haemochromatosis (HH), but a systematic study is lacking. However, patients with HH report symptoms suggesting involvement of the neuromuscular system: e.g., exercise intolerance, sensory complaints, muscle pain and weakness. The aim of this study was to determine the frequency of these signs and symptoms in HH and possible underlying causes.

Patients and methods: 46 Patients with definite HH from the department of internal medicine of a large general teaching hospital were systematically examined for symptoms and signs of a neuromuscular disorder (NMD) by means of a structured interview and physical exam. After reviewing these data an expert panel reached consensus about the presence of a possible neuropathy or myopathy and made recommendations for ancillary investigations (nerve conduction studies, electromyography, thermal threshold tests, laboratory tests). Only patients having a possible NMD of unknown origin were further analysed. After a second meeting consensus was reached about the final diagnosis. Patients who had a neuropathy or myopathy of which the origin was still unclear were referred to an independent neurologist for further evaluation.

Results: We included 30 men and 16 women with a mean disease duration of 7.2 years. Muscle pain was reported during rest in 39.1%, during exercise in 32.6% and after exercise in 41.3%. Muscle cramps were reported during rest in 32.6%, during exercise in 15.2% and after exercise in 15.2%. Exercise intolerance due to muscle pain was reported in 19.6%, to muscle cramps in 13.0%, to muscle weakness in 17.4% and to heaviness of the limbs in 17.4%. Clinical exam was normal in 24 patients. After the first round, in 33 patients (72%) a NMD was considered. Ultimately after additional investigations, 25 patients had no myopathy of neuropathy, 5 an axonal sensory motor polyneuropathy of which the cause was found (diabetes in 2, combination of diabetes and chemotherapy in 1, Charcot Marie Tooth type 2 in 1, Morbus Sjögren in 1), 9 patients had an idiopathic axonal sensory motor polyneuropathy, 3 an idiopathic small fiber polyneuropathy and 4 a carpal tunnel syndrome. There were no cases of proven myopathy. No significant differences in patient characteristics or haemochromatosis related data were found between patients diagnosed with polyneuropathy and those without.

Conclusion: We conclude that an idiopathic polyneuropathy was diagnosed in a relative large number of patients with HH (26%), which may suggest a relationship between these two disorders.

046

LACK OF HEPCIDIN SURGE IN RESPONSE TO ORAL IRON CHALLENGE IS THE CAUSE OF IRON OVERLOAD IN HFE-HEMOCHROMATOSIS

M Busbridge1,O Willson1, RS Chapman1, A Sangwaiya2, J Arnold2, M Thursz3

1Clinical Biochemistry, Imperial College Healthcare NHS Trust, London, UK, 2Dept Gastroenterology, Ealing Hospital Trust, London, UK, 3Hepatology, St Marys Hospital Trust, London, UK

Background: Systemic iron overload in humans is predominantly associated with a variety of genetic and acquired conditions. Of these conditions, HFE hemochromatosis (HFE-HC) is the most common and well defined inherited cause for iron related morbidity and mortality. The majority of patients with HFE-HC are homozygote for the C282Y polymorphism and without therapeutic intervention, there is a risk that iron overload will occur. Iron overload in most genetic types of hemochromatosis, can be explained by the insufficiency of the hepatic peptide hepcidin, the key regulator of systemic iron homeostasis.

Methods: The aim of the study was to measure hepcidin response to an oral iron challenge, (200mg ferrous sulphate), in HFE-HC (C282Y/C282Y) patients (n=9), compared to healthy controls (n=14) over a 4 hour period, with hourly bloods being taken. All C282Y/C282Y patients were iron depleted and studied at a time distant to phlebotomy. Hepcidin was measured using a published immunoassay method. Iron, Ferritin and transferrin saturation (%TSAT) were measured using standard methods. The area under the curve (AUC) was calculated and compared between control and groups.

Results: The basal serum hepcidin level in patients with C282Y/C282Y were low compared to controls (P=.0002). Serum hepcidin response was significant at 4 hours post iron challenge (P=.0085) returning to baseline only at 24 hours. Whereas this response was absent in patients with C282Y/C282Y (P=.294). The hepcidin response was significantly lower in C282Y/C282Y compared to controls (AUC: P=.0127).

Conclusion: This rapid hepcidin response to an oral iron challenge suggests that there may be both transcriptional and post transcriptional regulation of serum hepcidin levels. This raises the possibility that cleavage of the active 25-mer from prohepcidin may be an important regulatory step. These findings also suggest that a lack of ‘pulsatile release’ of hepcidin in response to iron challenge is one of the mechanisms of iron overload despite normal diet in HFE-HC patients with the C282Y/C282Y mutation.

047

EXPOSURE TO ETHANOL MODULATES THE EXPRESSION OF MOLECULES INVOLVED IN NON-TRANSFERRIN IRON TRANSPORT IN CACO-2 AND HEP-G2 CELLS

Kamila Balusikova, Jitka Neubauerova, Marketa Dostalikova, Jan Kovar

Department of Cell and Molecular Biology & Center for Research of Diabetes, Metabolism and Nutrition, Third Faculty of Medicine, Charles University in Prague, Prague, Czech Republic

The detrimental effect associated with alcohol consumption, and very often seen in alcoholic patients, is hepatic iron overload. Although chronical iron overload is a significant risk factor for associated liver failure, the mechanism, by which ethanol causes iron overload, is not still fully understood. We studied the effect of short-term and long-term exposure to ethanol on the expression of molecules involved in non-transferrin iron transport and molecules involved in the regulation of iron homeostasis (DMT1, Dcytb, ferroportin, hephaestin, ceruloplasmin, hepcidin, HFE, TfR1, and TfR2) in two specific functional types of human cells. We used tumor transformed Caco-2 and HEP-G2 cell lines as an in vitro model, where Caco-2 cells represent a model of human intestinal cell line and HEP-G2 represent a type of liver cells. The expression was tested on mRNA level by real-time PCR and on protein level by western blot analysis. We demonstrated that the mRNA level, but not protein level, of ferrireductase Dcytb was significantly affected by long-term exposure to ethanol in a cell type specific manner. The level of Dcytb mRNA was almost 10-fold lower in Caco-2 cells after cultivation with ethanol when compared with control cells. On the other hand, the expression of Dcytb in HEP-G2 was almost 30-fold higher on mRNA level when compared with control. In both cell lines, significantly decreased expression of ferroportin protein was detected after the long-term ethanol treatment. A significant increase of hepcidin expression on mRNA level (more than 2-fold) was detected only in HEP-G2 cells. Additionally, long-term exposure to ethanol induced significant increase of HIF-1alpha; protein level and decrease of ADH mRNA level in HEP-G2, but not in Caco-2 cells. In Caco-2 cells, the ADH mRNA was not detected. Long-term cultivation with acetaldehyde significantly increased Dcytb mRNA level (approximately 2-fold) and hepcidin mRNA level (approximately 3-fold) in HEP-G2 cells. It also dramatically changed the mRNA level of hephaestin in Caco-2 cells. Its expression was more than 30-fold lower when compared with control. Changes in the expression of tested molecules under the treatment with ethanol and acetaldehyde in both cell types can indicate that ethanol affects the molecule expression rather per se than via its metabolite. Moreover, we suggest that in ethanol effect on the expression of molecules involved in non-transferrin iron transport can be incorporated some other molecules involved in the regulation of gene expression and it needs some further testing.This project is supported by research program MSM0021620814

048

ETHANOL EFFECT ON CELL DIFFERENTIATION IN THE HUMAN HEPATOMA HEPARG CELL LINE: INTERRELATION WITH IRON METABOLISM

Thi Hong Tuoi DO1,2, François Gaboriau1,2, Martine Ropert1,2,3, Romain Moirand1,2,3, Isabelle Cannie1,2, Pierre Brissot1,2,3, Olivier Loreal1,2, Gérard Lescoat1,2*

1Inserm, UMR 991, ≪ Foie, Métabolismes et Cancer ≫, F-35033 Rennes, France, 2Université de Rennes 1, F-35043, Rennes, France, 3CHU Rennes F-35033, Rennes, France

Introduction: Ethanol and iron play a role in cell differentiation. Moreover, metabolism of both ethanol and iron depends on the differentiation state. However, so far no study has analyzed the ethanol effect on the differentiation of hepatocytes which are the main target of alcohol and iron store. Thus, the purpose of this study was to evaluate the ethanol effect on hepatocyte differentiation in relation with iron metabolism.

Methods: The human hepatoma HepaRG cell line which evolves from proliferation to confluence followed by differentiation in hepatocytes (HC) surrounded by epithelioid biliary like cells (BC) was used. This cell line expresses some enzymes of ethanol metabolism (ADH2, CYP 2E1). 24h after cell seeding, different treatments were performed for 14 or 20 days and renewed each 24h because of ethanol evaporation. Cell viability was determined by MTT test, total proteins of the cultures by Bradford method, LDH leakage by "Cytotoxicity detection LDH" kit, lipid peroxidation by TBARS test. AST/ALT release, total iron, soluble transferrin receptor (sTfR), ferritin and phosphatase alkaline were quantified by biochemical analysis. Hepatocyte and biliary cells were identified and measured by immunohistochemical analysis. Gene expression was analyzed by quantitative RT-PCR. Cellular iron was identified by Perls coloration and iron uptake was measured by 55Fe(III) incorporation.

Results: Ethanol decreases in a time and dose-dependent manner cell viability and increases the expression of enzymes of its metabolism: ADH2 and CYP 2E1. This effect is associated to cytotoxicity: LDH, AST, ALT leakage and MDA production. Ethanol favours differentiation, as shown by a reduction of alcaline phosphatase activity, an increase of hepatocyte specific markers (albumin, transferrin, aldolase B, CYP 3A4) and a modification of the HC and BC percentages. Ethanol increases also the cellular iron content which was correlated to sTfR decrease and ferritin increase. An overexpression of genes implicated in transmembrane iron transport is also observed: TfR1 and DMT1 implicated in its ingress; ferroportin and ceruloplasmin implicated in its egress. Moreover, the addition of exogenous iron amplifies ethanol hepatotoxicity and modulates its metabolism and its effect on differentiation. On the opposite, ethanol effects are attenuated by iron chelators.

Conclusion: This study demonstrates that ethanol modulates hepatocyte differentiation. This effect is associated to a dysfunction of iron metabolism and appears to be modulable by the cellular iron level. These results suggest that iron metabolism manipulation by chelators could have an interest in the treatment of alcoholic liver diseases.

049

PERIPHERAL BLOOD CD8+ BUT NOT CD4+ T-LYMPHOCYTE NUMBERS ARE ASSOCIATED WITH INFLAMMATION AND FIBROSIS IN NON-ALCOHOLIC STEATO-HEPATITIS

Mónica Costa1, José Fraga2, Jorge Vieira3, Helena Alves4, Cristina Vieira3, Rosa Lacerda5, Graça Porto 1,5,6 and Eugénia Cruz 1,6

1Iron Genes and the Immune System (IRIS) Group, Institute for Molecular and Cell Biology (IBMC), Porto,Portugal, 2Gastroenterology Unit of Centro Hospitalar V.N.Gaia/Espinho, V.N.Gaia, Portugal, 3Molecular Evolution Group, IBMC,Porto,Portugal, 4Centro de Histocompatibilidade do Norte, Porto, Portugal, 5Abel Salazar Institute for Biomedical Sciences(ICBAS), Porto, Portugal, 6Centro Hospitalar do Porto(CHP)-Santo António Hospital, Porto, Portugal

Non-alcoholic steatohepatitis (NASH) is a common disease characterized by steatosis and inflammation in the liver, with or without fibrosis and siderosis (Varela-Rey M et al. 2009). Mutations of the HFE gene, associated with hereditary hemochromatosis(HH), were implicated in NASH, due to the hepatic iron accumulation described in some of these patients (Wallace DF et al. 2009). Other genes were also implicated in NASH, however results still controversial. CD8+ T cells are known to be modifiers of severity of disease in HH and it is well accepted that these numbers are regulated by genes located at the MHC class-I region (Cruz et al. 2006). Here, we studied 8 MHC Class-I genetic markers (HLA-A, B and C; PGBD1, ZNF193, HFE mutations, rs724078 and rs4713207) and CD8+T cell numbers in relation to the severity of NASH (evaluated by the degree of steatosis, inflammation, fibrosis and siderosis in liver biopsies) in 59 NASH patients. Controls (n=264) were used for comparing frequencies of all genes except for SNPs where a subgroup (n=56) was used. In general, no statistically significant differences were observed in genetic markers between patients and controls. However, an increased frequency of the haplotypes A29-B44 (0.052) and A33-B14 (0.060) were found in NASH patients in comparison with controls (0.027 and 0.029, respectively). Moreover, a remarkable conservation was observed for two particular extended haplotypes defined by 8 markers. These haplotypes carrying the HLA-A alleles A*29 and A*33. Liver inflammation was strongly correlated with fibrosis (Rsquared=0.33, p<0.000001), but not with steatosis or siderosis. Moreover, a negative association was found between CD8+T-cells and both inflammation (p=0.007, Student T-test) and fibrosis (p=0.037, Student T-test). This was not observed for CD4+ T-cells. Patients carrying the extended haplotypes with the HLA-A*33 were associated with more fibrosis (6 out 7 NASH patients with fibrosis) and low CD8+ T-cells (0.38±0.20 x10∧6/ml), while patients carrying the extended haplotype with the HLA-A*29 were associated with less fibrosis (1 out of 7 NASH patients with fibrosis) and high CD8+ T-cells (0.50±0.16 x10∧6/ml). Results show that inflammation and fibrosis in NASH were associated with CD8+T lymphocytes and that MHC Class-I genes may be involved in these associations.

050

IRON CHELATION IN OESOPHAGEAL ADENOCARCINOMA – THE IN VITRO EFFECTS OF DEFERASIROX AND DEFEROXAMINE AS ANTI-NEOPLASTIC AGENTS

Samuel J Ford, Thomas Roe, Tariq Iqbal, Derek Alderson and Chris Tselepis

Cancer Research UK Centre for Cancer Studies, University of Birmingham, Edgbaston, Birmingham, B152TT. UK

Background: The malignant progression of Barrett's metaplasia to oesophageal adenocarcinoma is associated with altered expression and function of the pertinent cellular iron transport proteins. These alterations result in increased cellular iron loading which is likely to drive cellular proliferation; a key hallmark of cancer. This is further supported by a wealth of data demonstrating that iron chelators cause cell cycle arrest and apoptosis. However, to date whether the clinically established licensed iron chelators deferasirox and deferoxamine (DFO) might be of use as anti-neoplastic agents in the treatment of oesophageal cancer has yet to be determined.

Methods: Oesophageal adenocarcinoma cell lines OE33 and OE19 were cultured with increasing doses of deferasirox and DFO to determine IC50 values. The same cell lines were than cultured with iron or haem in the presence or absence of deferasirox or DFO. The effect on cellular viability, proliferation and migration was assessed by phenotype employing an MTT, BrdU and scratch assay respectively. The effect of the iron chelators on cellular iron concentration was determined by ferrozine assay.

Results: Both deferasirox and DFO appear to have an anti-neoplastic effect on oesophageal adenocarcinoma cells in vitro with IC50 values of 20μM and 10μM respectively. MTT and BrdU assays demonstrated a significant decrease in viability and proliferation in OE33 and OE19 lines cultured in supplemental iron with or without an iron chelator (all p-values <0.0001). A similar observation was made with the addition of a chelator to cells cultured with haem iron. Cellular migration was impeded by the addition of either deferasirox or DFO to culture media. Cellular iron concentration was significantly reduced, in the majority of cases, by the addition of either deferasirox or DFO to culture media for both OE33 and OE19 cell lines cultured with supplementary iron (p=0.05, p=0.0028, p=0.076, p=0.065 respectively). Again a similar effect was noted when the chelators were added to cultured cells with haem iron, although the effect of DFO was slightly blunted compared to deferasirox.

Conclusions: Oesophageal adenocarcinoma cell lines appear to be sensitive to iron deprivation mediated by iron chelation. The clinically established iron chelators, deferasirox and DFO, have anti-neoplastic properties in vitro. DFO was slightly less effective than deferasirox in chelating intracellular iron derived from culture with haem loaded media; an effect possibly mediated by the lipophilic nature of deferasirox. Iron chelators may have a future role in the chemotherapeutic treatment of oesophageal adenocarcinoma.

051

THE EXPRESSION OF HAEM IMPORT PROTEINS LRP1 AND HCP1 AND EXPORT PROTEINS FLVCR AND BCRP IN THE PROGRESSION OF BARRETT'S METAPLASIA TO OESOPHAGEAL ADENOCARCINOMA

Samuel J Ford, Thomas Roe, Tariq Iqbal, Derek Alderson and Chris Tselepis

Cancer Research UK Centre for Cancer Studies, University of Birmingham, Edgbaston, Birmingham, B152TT. UK

Aims: The incidence of oesophageal adenocarcinoma is currently rising faster than any other malignancy in the Western world. The strongest known risk factor for this disease is Barrett's metaplasia. In the progression of Barrett's metaplasia to adenocarcinoma we have previously reported an over expression of cellular iron import proteins and loss of export function resulting in increased cellular iron loading. This resulting increased cellular iron load was associated with a more aggressive cellular phenotype in vitro. However, to date what has not been addressed is the expression profile of the pertinent cellular haem transport proteins in the malignant progression of this disease and what effect haem has on oesophageal cell fate. Methods: mRNA and protein expression of the haem import proteins (LRP1 and HCP1) and export proteins (BCRP and FLVCR) was determined by qRT-PCR and western blotting respectively, in a series of matched samples of squamous oesophagus, benign and dysplastic Barrett's metaplasia and adenocarcinoma. Immunolocalisation was assessed using immunohistochemistry.

Results: No significant change in mRNA or protein expression between normal squamous mucosa and benign Barrett's for any of the haem transport proteins was noted. However, mRNA and protein levels were significantly up-regulated for all haem transport proteins when comparing dysplastic Barrett's to adenocarcinoma (LRP1 p=0.049/p=0.001, HCP1 p=0.018/p=0.030, BCRP p=0.007/p=0.001 and FLVCR p=0.032/p=0.001). Immunolocalisation demonstrated LRP1, HCP1 and BCRP to be membranous in location with FLVCR exclusively nuclear. Semi-quantitative analysis demonstrated a significant up-regulation in immunoreactivity between dysplastic Barrett's and adenocarcinoma (LRP1 p=0.003, HCP1 p=0.001, BCRP p=0.001 and FLVCR p=0.001). Conclusions: The over expression of LRP1 and HCP1 in the progression of Barrett's to adenocarcinoma is strongly suggestive of an involvement of cellular haem import in contributing to cellular iron levels. The concomitant over expression of the haem export proteins may represent a mechanism of protecting cancer cells from the toxic effects of haem overload or may be important in effluxing other cytotoxic substances such as chemotherapeutic agents.

052

A POTENTIAL ROLE FOR IRON AND HAEM IN BREAST CANCER

Thomas Roe, Chris Tselepis,

University of Birmingham CRUK Institute for Cancer Studies, UK

Introduction: There is an overwhelming body of evidence implicating iron in gastrointestinal carcinogenesis. However, what remains unclear is whether this association is confined to gastrointestinal malignancies or whether iron represents a generic carcinogen in all cancers including breast.

Methods: To assess the importance of iron in breast cancer, expression of the pertinent cellular iron import (TfR1, HCP-1, LRP1, DMT-1 and Dcytb), export (FPN, BCRP, and FLVCR) and storage (ferritin) proteins were determined by immunohistochemistry on normal breast, ductal carcinoma in situ (DCIS) and breast cancer sections. Breast cell lines were also cultured with iron and haem to assess the effects on intracellular iron levels and cell phenotype including viability, proliferation, migration and anchorage independent growth.

Results: Immunolocalisation studies showed a statistically significant over expression in both the cellular iron import machinery (Dcytb, DMT-1, and TfR1) and export machinery (FPN, BCRP) in breast cancer compared to normal breast tissue (p<0.015 for all). No expression differences were observed between DCIS and normal breast, but there was a statistical over expression for hepcidin when comparing DCIS with breast cancer (p<0.05). Cell culture studies demonstrated that culturing breast cancer lines (ZR751, T47D and MCF7) with iron or haem resulted in increased cellular iron accumulation associated with statistically significant increases in cellular viability, proliferation, migration and anchorage independent growth. These responses were not observed in the benign HB2 cell line.

Conclusions: Breast cancer development is associated with a modulation in the expression of iron transport proteins which likely reflects the increased demand of cancer cells for iron. This increased iron load is likely to be pertinent to driving the tumourigenic phenotype. This suggests iron chelators as a potential treatment for breast cancer.

Ancillary