Iron-Induced Expression of Bone Morphogenic Protein 6 in Intestinal Cells Is the Main Regulator of Hepatic Hepcidin Expression In Vivo

Transcription

1 GASTROENTEROLOGY 2010;138: Iron-Induced Expression of Bone Morphogenic Protein 6 in Intestinal Cells Is the Main Regulator of Hepatic Hepcidin Expression In Vivo STEPHANIE ARNDT,* ULRIKE MAEGDEFRAU,* CHRISTOPH DORN, KATHARINA SCHARDT,* CLAUS HELLERBRAND, and ANJA KATRIN BOSSERHOFF* *Institute of Pathology and Department of Internal Medicine I, University of Regensburg, Regensburg, Germany BACKGROUND & AIMS: Recent studies identified bone morphogenic protein 6 (BMP6) as a key regulator of hepatic hepcidin expression and iron metabolism, but the cellular source of BMP6 and the reason for its specific effect on hepatocytes are unknown. METHODS: BMP and hepcidin expression upon iron sensing were analyzed in vivo in BMP6 / and BMP6 / mice and ex vivo in tissue and in vitro in cells of the liver and the small intestine. RESULTS: BMP6 / mice developed severe hepatic iron accumulation and reduced hepcidin expression with increasing age. This phenotype could be triggered in younger BMP6 / mice by dietary or parenteral iron application. Furthermore, both treatments induced a marked up-regulation of BMP6 expression in the small intestine of BMP6 / mice. Ex vivo treatment of intestinal tissue of BMP6 / mice with iron sulfate or holotransferrin confirmed epithelial cells as an inducible source of BMP6. In contrast, iron overload did not promote a striking induction of BMP6 expression in hepatocytes or macrophages. Furthermore, iron-supplemented diet induced a compensatory up-regulation of BMP2, BMP4, and BMP9 in the small intestine of BMP6 / mice that was apparently not sufficient to assure iron homeostasis. As a potential explanation, analysis of hepatocytes revealed an expression pattern of BMP receptor subunits preferentially used by BMP6, and treatment of hepatocytes with different recombinant BMPs identified BMP6 as the most potent stimulator of hepcidin expression. CONCLUSIONS: Epithelial cells of the small intestine are the predominant cellular source of BMP6 upon iron sensing. Our findings reveal a previously unknown mechanism in which the small intestine controls iron homeostasis. Multiple pathways are known to regulate hepatic hepcidin expression and thus indirectly affect iron uptake and retention. 1,2 Hepcidin deficiency is the common pathogenic mechanism for both juvenile and adult forms of the genetic iron overload disorder hereditary hemochromatosis as a result of mutations in both alleles of any of the hepcidin (HAMP), hemochromatosis (HFE), hemojuvelin (HJV), or transferrin receptor 2 (TFR2) genes. 3 5 In patients with chronic anemia, intestinal iron absorption is impaired, and iron remains sequestered in reticuloendothelial cells, leading to hypoferremia and anemia. 6,7 Because both iron deficiency and iron overload have severe effects, cellular and systemic iron homeostasis are critically important. Recent advances in the field of iron metabolism have led to a better understanding of the pathways involved in iron homeostasis. Thus, several studies revealed the important role of the bone morphogenic protein (BMP) pathway in iron metabolism by up-regulating hepcidin expression in a process that involves translocation of Smad1/5/8 in complex with Smad4 into the nucleus In hepatocytes, the BMP receptor (BMPR) is adapted for iron regulation by its interaction with the coreceptor hemojuvelin. 8,9 Interestingly, stimulation of BMP signaling via hemojuvelin is paralleled by an increase of hepcidin expression. Support for the involvement of the BMPsignaling pathway in the regulation of hepcidin also comes from the observation that liver-specific disruption of the Smad4 gene in mice leads to reduced hepcidin expression and systemic iron loading similar as in hepcidin knockout mice. 10 Two recently published studies identified BMP6 as a key endogenous regulator of hepcidin expression and iron metabolism. 11,12 Although several studies began to unravel the complex regulatory mechanism of iron homeostasis, there remain several open questions. Thus, although hepcidin expression was shown to be induced by the BMP/Smad-signaling pathway, it is not yet known how iron regulates this pathway and whether BMP6 is the most relevant or only BMP molecule regarding regulation of hepatocellular hepcidin expression. Furthermore, the cellular source of BMP6 is unknown. Thus, it is not proven whether hepatocytes are the source of BMB6 upon sensing iron in vivo. In this study, we addressed these open questions and validated BMP6 as the most potent inductor of hepatocellular hepcidin expression compared with other BMPs. Moreover, we identified enterocytes of the small intestine Abbreviations used in this paper: BMP, bone morphogenic protein; MCD, methionine-choline-deficient diet by the AGA Institute /10/$36.00 doi: /j.gastro

2 January 2010 BMP6 REGULATES HEPATIC HEPCIDIN EXPRESSION 373 but not hepatocytes or macrophages as the main cellular source of iron-induced BMP6 expression. These findings elucidate a previously unknown mechanism in which the small intestine controls iron homeostasis. In addition to iron absorption, we could show in this study that epithelial cells of the small intestine regulate hepatic hepcidin expression via BMP6 release. Materials and Methods Animals and Cells 129Sv/Ev wild-type (BMP6 / ) and BMP6 / mice on 129Sv/Ev background 13 were obtained from the Robertson Laboratory (Department of Molecular and Cellular Biology, Harvard University, Cambridge, MA). Primary hepatocytes, hepatic stellate cells, and Kupffer cells were isolated from the livers of BMP6 / mice and cultivated as described. 14 Primary murine macrophages were freshly isolated from the bone marrow. Briefly, cells were isolated by flushing sterile phosphate-buffered saline (PBS) through the femur of 2-month-old BMP6 / mice, and, subsequently, cells were centrifuged, and the cell pellet was resuspended in RPMI 1640 (Biochrom, Berlin, Germany) supplemented with 2% human AB serum (Cambrex, Walkersville, MD), penicillin (50 U/mL; GIBCO, Carlsbad, CA), streptomycin (50 g/ml; GIBCO), L-glutamine (2 mmol/l; GIBCO), and recombinant human Macrophage Colony Stimulating Factor (rhmcsf) (100 ng/ml; Cetus, Emmeryville, CA) and plated in bacteriologic grade dishes. Experimental Models and Diets Two-month-old male and female BMP6 / and BMP6 / mice were fed with a standard diet (control diet: 200 mg iron/kg; Ssniff, Soest, Germany), an ironsupplemented diet (7.5 g iron/kg consisting of Fe-citrate, -fumarate, -sulfate, and -gluconate; Ssniff; feeding interval: 3 weeks), an iron-deficient diet ( 6 8 mg iron/kg; Ssniff; feeding interval: 3 weeks), or a methionine-choline-deficient diet (MCD; MP Biomedicals, Illkirch, France; feeding interval: 6 weeks). After feeding the individual diets for the indicated time intervals, animals were killed, and the livers were photographed and histologically analyzed by Perl s Prussian blue staining. Furthermore, iron-dextran (0.2 g iron/kg body weight; iron (III)-hydroxid-complex; CosmoFer, Kirchzarten, Germany) or dextran (1% in PBS; Sigma Aldrich, Munich, Germany) as control were injected into the retro bulbar vein complex of 2- to 3-month-old BMP6 / or BMP6 / mice. Animals were killed 4 and 8 hours after injection, and liver tissue was harvested for subsequent analysis. The number of animals used for each experimental setting is indicated in the corresponding Figure legend. For the MCD model, only male mice have been used. All other models have been performed with equal numbers of male and female mice, and, here, no significant sex differences have been observed. Ex Vivo and In Vitro Experiments For ex vivo experiments, the small intestine and liver were immediately resected after death of mice, and, subsequently, both organs were dissected into equal slices (liver: 0.05 g, and small intestine: g wet weight) and incubated at room temperature in the dark on a shaker for 1 hour (RNA analysis) or 4 hours (protein analysis) with 50 mol/l FeSO 4 or 50 mol/l holo-transferrin (both from Sigma Aldrich) in 200 L PBS. For ex vivo coculture experiments, equal slices of small intestine (0.025 g wet weight) and liver tissue (0.05 g wet weight) were incubated together for 5 hours in a 50 mol/l FeSO 4 /PBS solution. As control, tissues were incubated in PBS. After 5 hours of incubation, tissues were separated and collected into fresh tubes, and total RNA was isolated from the liver slices. For in vitro experiments, primary hepatocytes, hepatic stellate cells, and Kupffer cells were incubated with 50 mol/l FeSO 4 or holo-transferrin for 1 hour. Subsequently, total cellular RNA was isolated and analyzed. Measurement of BMP6 in the Serum For serum extraction, whole blood was collected into endotoxin free, heparinized vials (Qiagen, Hilden, Germany) by cardiac puncture under anesthesia (6 8 mg/kg body weight xylazine/ mg/kg body weight ketamine) directly before death, and serum was separated by centrifugation at 4000 rpm for 10 minutes. BMP6 amount in the serum was measured by loading 5 L serum on a 10% SDS-PAGE and Western blot analysis as described below. Measurement of Iron Concentrations in Hepatic Tissue Hepatic iron concentration was determined by atomic absorption spectrophotometry as described. 15 Prussian Blue Staining Histologic paraffin sections of various tissues were stained with K 4 Fe(CN) 6 3H 2 O and counterstained with fast red based on standard procedures. Immunohistochemical Staining For immunohistochemistry, the tissues were deparaffinized, rehydrated, and subsequently heated in a pressure cooker for 5 minutes in Tris/EDTA buffer (ph 9.0). After blocking with H 2 O 2, tissue sections were incubated for 30 minutes with goat anti-bmp6 (S-20), sc antibody (1:50; Santa Cruz Biotechnology, Santa Cruz, CA) in Antibody Diluent with Background Reducing Components (DakoCytomation; Dako North America, Inc, Carpinteria, CA). The secondary antibody antigoat-peroxidase (1:50; Sigma Aldrich) was applied for 30 minutes at room temperature. Antibody binding was

3 374 ARNDT ET AL GASTROENTEROLOGY Vol. 138, No. 1

4 January 2010 BMP6 REGULATES HEPATIC HEPCIDIN EXPRESSION 375 visualized using DAB solution (DakoCytomation; Dako North America, Inc), and counterstaining was performed with hemalaun solution (DakoCytomation; Dako North America, Inc). Evaluation of the staining was performed semiquantitatively by light microscopy (Axio- Vision; Zeiss, Munich, Germany). Western Blot Analysis Protein lysates were prepared as described, 16 and g of the lysates were separated on a 10% SDS- PAGE. Western blotting was performed applying antibodies against BMP6 (1:500; Santa Cruz Biotechnology; anti-goat BMP6 [S-20], sc27408) or -actin (1:5000; Sigma Aldrich) for loading control. An alkaline phosphate-coupled antibody (Chemicon International, Inc, Temecula, CA) was used as secondary antibody, and immunoreactions were visualized by NBT (nitro blue tetrazolium chloride)/bcip (5-bromo-4-chloro-3-indolyl phosphate, toluidine salt) (Zytomed Systems GmbH, Berlin, Germany). For quantitative analysis, the intensity of the BMP6 bands was determined applying densitometry, and values were calculated in relation to the intensity of the corresponding bands of -actin that has been used as loading control. RNA Analysis Isolation of total cellular RNA from cultured cells and tissues and reverse transcription were performed as described. 16 Quantitative real-time polymerase chain reaction (PCR) was performed with specific sets of primers (Supplementary Table 1) applying LightCycler technology (Roche Diagnostics, Mannheim, Germany) as described. 16 Statistical Analyses Statistical analyses were performed using SPSS version 10.0 (SPSS, Chicago, IL) and GraphPad Prism Software (GraphPad Software, Inc, San Diego, CA). Results are expressed as mean SEM (range) or percent. Comparison between groups was made using the Student paired t test. A P value.05 was considered significant. Results Age-Dependent Increase of Hepatic Iron Concentration and Reduced Hepcidin Expression in BMP6 / Mice on 129Sv/Ev Background In parallel with 2 other groups, 11,12 we analyzed the role of BMP6 in iron metabolism. We investigated BMP6-deficient (BMP6 / ) mice and wild-type control mice (BMP6 / ) on a 129Sv/Ev background, 13 whereas mice on a mixed 129Sv/C57 background or on a CD1 outbred background, respectively, were used in the studies of Meynard et al 11 and Andriopoulos et al. 12 Both groups used 1- or 2-month-old BMP6 / mice, respectively, and observed a phenotype resembling hereditary hemochromatosis, with reduced hepcidin expression and tissue iron overload. In contrast, we did not observe significantly increased hepatic iron staining (Figure 1A) or reduced hepcidin expression (Table 1) in 2- to 3-monthold BMP6 / mice compared with controls, clearly indicating strain specific differences. However, with increasing age, also BMP6 / mice on a 129Sv/Ev background developed a striking increase of hepatic iron accumulation and a remarkably reduced hepatic hepcidin expression compared with BMP6 / mice of the same age (Figure 1A and Table 1). In addition to hepatocytes, iron staining revealed a considerable iron accumulation in acinar cells of the pancreas, in epithelial cells of the renal tubules, in pigment epithelial cells of the eye, and in cardiomyocytes of BMP6 / mice on a 129Sv/Ev background at the age of 5 6 months (Supplementary Figure 1). Furthermore, an intensive iron accumulation was observed in mesenchymal cells of the testis and in macrophages in the lung, spleen, and bone marrow of BMP6 / mice compared with BMP6 / mice of the same age (Supplementary Figure 1). In contrast, we did not detect significant iron deposition in any of these tissues in younger BMP6 / mice (data not shown), whereas Meynard et al 11 and Andriopoulos et al 12 reported iron accumulation in extrahepatic tissue of BMP6 / mice already at the age of 2 4 Figure 1. Hepatic iron accumulation and hepcidin expression in BMP6 / mice on 129Sv/Ev background. (A) Iron accumulation in liver sections of BMP6 / mice with increasing age. BMP6 / mice older than 5 months develop a massive granular iron accumulation across all zones of the liver. In BMP6 / mice older than 8 months, a Scheuer grade 29 4 iron overload with focal deposition in portal macrophages was observed. BMP6 / mice did not show pathologic changes. (B) Hepatic iron accumulation in sections of BMP6 / mice fed diets with different iron content or a methioninecholine-deficient (MCD) diet. Two- to 3-month-old mice were fed for 3 weeks with an iron-supplemented diet. An iron gradient was identified with maximum iron in periportal (zone 1) hepatocytes in BMP6 / animals, whereas no apparent accumulation was observed in BMP6 / mice. MCD-fed BMP6 / mice developed a massive and uniform distributed iron accumulation throughout the liver, whereas no iron accumulation was visible in the liver of BMP6 / mice. Prussian Blue staining does not reveal major differences regarding iron accumulation in BMP6 / and BMP6 / mice fed iron-balanced (standard) or iron-deficient diet for 3 weeks (n 4 11 mice per group in analogy to Table 2). (C) Expression of hepcidin in the liver after intravenous iron injection. mrna analysis by quantitative RT-PCR revealed no induction of hepcidin in the liver of BMP6 / mice after intravenous iron-dextran injection, whereas this treatment resulted in significantly increased hepcidin expression in BMP6 / mice (*P.05 compared with control). (D) BMP6 serum levels after intravenous iron injection or feeding a diet with elevated iron content. Parenteral application of iron-dextran or feeding an iron-supplemented diet induced BMP6 protein in the serum of BMP6 / mice as assessed by Western blot analysis. The 23-kilodalton band of mature BMP6 migrates in serum probes as a subtle doublet, which was most apparent in lower percentage polyacrylamide gels. This may have been caused by variations in glycosylation as described. 28

5 376 ARNDT ET AL GASTROENTEROLOGY Vol. 138, No. 1 Table 1. Hepatic Iron Concentration and Hepcidin Expression in BMP6 / and BMP6 / Mice of Different Ages Age, (mo) Liver iron (mg Fe/g dry weight) Hepatic hepcidin expression (relative to BMP6 / ) BMP6 / BMP6 / BMP6 / BMP6 / a a a a NOTE. Hepatic iron concentration was increased and hepatic hepcidin expression was strongly reduced with increasing age of BMP6 / mice. Results are expressed as mean SD (n 6 mice in the group of 2- to 3-month-old mice and n 3 or 4 in older mice). a P.05 compared with BMP6 / mice of the same age. months. Together, these finding indicate that the effects of BMP6 deficiency on body iron accumulation and hepatic hepcidin expression, respectively, reveal strain-specific differences, with BMP6 / mice on 129Sv/Ev background developing progressive iron overload with aging. Effects of Iron Overload and Iron Depletion on Hepatic Iron Levels and Hepcidin Expression in BMP6 / Mice on 129Sv/Ev Background To augment iron overload also in young (2 3 months old) BMP6 / mice on 129Sv/Ev background, we fed an iron-supplemented diet for 3 weeks. Furthermore, we applied a MCD diet (feeding interval, 6 weeks) that recently has been shown to enhance hepatic iron uptake. 17 Both diets induced a considerably enhanced hepatic iron overload in 2- to 3-month-old BMP6 / mice compared with BMP6 / mice of the same age (Figure 1B, Table 2). Noteworthy, in both models, hepatic hepcidin messenger RNA (mrna) expression increased significantly in BMP6 / mice compared with BMP6 / mice (Table 2). Moreover, also, short-term iron application by intravenous injection of iron-dextran (0.2 g/kg body weight for 4 or 8 hours) resulted in a significant induction of hepatic hepcidin over time in BMP6 / but not in BMP6 / mice (Figure 1C). Conversely, feeding an iron-deficient diet (for 3 weeks) led to a significant reduction of hepatic hepcidin expression and iron content in BMP6 / but not in BMP6 / mice (Figure 1B, Table 2). Still, it has to be noted that the considerably reduced hepcidin levels in BMP6 / mice under iron starvation are similar as those observed in BMP6 / mice receiving an iron-balanced (standard) diet. Thus, it may be speculated that hepcidin levels in BMP6 / mice are in the range of the lower threshold and, therefore, are only slightly affected by iron depletion. Importantly, dietary as well as short-term iron application led to marked increase of BMP6 serum levels in BMP6 / mice (Figure 1D). These findings indicate that BMP6 mediates both short-term response to iron and iron homeostasis. Strong BMP6 Expression in the Small Intestine Ex Vivo After Sensing Iron Of immediate interest were the cell types expressing BMP6 after sensing iron. Because liver, small intestine, and macrophages are known to be the central regulators of iron homeostasis, we focused on these tissues and cells, respectively. First, we analyzed the expression of BMP6 in bone marrow-derived macrophages and observed only a marginal expression of BMP6 in these cells compared with hepatic and small intestinal tissue (data not shown). Moreover, BMP6 expression in macrophages was not inducible by treatment with 50 mol/l FeSO 4 or 50 mol/l holo-transferrin (Figure 2A), suggesting that these cells are not responsible for increased BMP6 expression in response to iron sensing. Next, we assessed hepatocytes, hepatic stellate cells, and Kupffer cells isolated from murine livers because these hepatic cells have been described to express BMP6. 20 However, neither treatment with iron sulfate nor with holo-transferrin led to a significant induction of BMP6 expression in these cells (Figure 2B). In addition, no significant induction of hepatic hepcidin expression was observed (Figure 2C). Comparable results were obtained ex vivo on mrna (Figure 3A and C) and protein level (Figure 3B) upon stimulation of slices of murine liver tissue with iron. Next, we analyzed whether an induction of hepatic BMP6 was detectable in vivo after sensing iron. Intravenous injection of iron-dextran (Figure 4A [4 hours] and B [8 hours]) or feeding iron-supplemented or iron-deficient diets, respectively, for 3 weeks (Figure 4C and D) resulted in minor alterations of hepatic BMP6 mrna expression and no significant change of protein expression ( fold; P.05), respectively. In accordance, immunohistochemical staining of liver sections did not reveal specific staining of BMP6 in hepatocytes of BMP6 / mice treated with iron-dextran or fed with iron-supplemented or iron-deficient diets (Figure 4E). In contrast and noteworthy, ex vivo treatment of tissue specimens from the small intestine with iron sulfate or holo-transferrin resulted in a significant increase of BMP6 mrna (Figure 3A) and protein (Figure 3B; fold and fold, respectively, P.05). Furthermore, we observed a 7- to 8-fold induction of BMP6 mrna in the small intestine after iron-dextran injection (Figure 4A) or feeding an iron-supplemented diet (Figure 4C). Densitomeric analysis of Western blots (representative blots are depicted in Figure 4B and D) confirmed a significant induction of BMP6 in the small intestine upon iron challenge also on the protein level (iron-dextran: fold and iron-supplemented diet: fold; P.05). In contrast, BMP6 protein went below the detection limit in the small intestine of mice fed with iron-deficient

6 January 2010 BMP6 REGULATES HEPATIC HEPCIDIN EXPRESSION 377 Table 2. Hepatic Iron Concentration and Hepcidin Expression in BMP6 / and BMP6 / Mice Fed Diets With Different Iron Content or a Methionine-Choline-Deficient Diet Liver iron (mg Fe/g dry weight) Hepatic hepcidin expression (relative to BMP6 / standard diet) Diet BMP6 / BMP6 / BMP6 / BMP6 / Standard a a Iron deficient a Iron supplemented a a MCD a a NOTE. Two- to 3-month-old BMP6 / and BMP6 / mice were fed with iron-balanced diet (control: feeding interval, 6 weeks; 5 or 6 mice/genotype), iron-deficient diet (feeding interval, 3 weeks; n 8), iron-supplemented diet (feeding interval, 3 weeks; n 11), or methionine-choline-deficient diet (feeding interval, 6 weeks; n 4). a P.05 compared with BMP6 / mice fed the same diet. diet, whereas BMP6 mrna expression was not significantly affected (Figure 4C and D). These findings indicate that very low intestinal BMP6 levels are only marginally repressed by iron deficiency, whereas iron sensing leads to a marked increase of BMP6 predominantly in the small intestine. To identify the cell type expressing BMP6 in the small intestine, we performed immunohistochemical analysis and revealed distinct induction of BMP6 expression in sporadic enterocytes of small intestinal villus tips of BMP6 / mice injected with iron-dextran or fed with an iron-supplemented diet (Figure 4F). Goblet cells and cells of the crypt revealed nonspecific immunohistochemical signals as recognized by comparison with tissue sections from BMP6 / mice, which were included as negative control. BMP Expression in the Small Intestine After Feeding an Iron-Supplemented Diet and Effects of Different BMPs on Hepcidin Expression of Hepatocytes in Vitro Next, we analyzed whether BMP6 is actually the main endogenous regulator of hepcidin expression because previous studies suggested other BMPs, such as BMP2, BMP4, or BMP9, as potential hepcidin effectors. 8,21,22 Feeding an iron-supplemented diet did not change expression levels of BMP2, 4, and 9 in the small intestine of BMP6 / mice, but analysis of BMP6 / mice Figure 2. BMP6 and hepcidin expression after sensing iron in vitro in macrophages and hepatic cells. BMP6 mrna expression was analyzed in macrophages (A) and hepatic cells (B) of BMP6 / mice after treatment with FeSO 4 or holotransferrin (50 mol/l for 1 hour) by quantitative RT-PCR. Neither macrophages nor hepatocytes (HEP), hepatic stellate cells (HSC), or Kupffer cells (KC) revealed a significant induction of BMP6 expression after sensing iron. (C) Furthermore, treatment with FeSO 4 or holo-transferrin did not induce hepcidin mrna expression of hepatic cells in vitro.

7 378 ARNDT ET AL GASTROENTEROLOGY Vol. 138, No. 1 Figure 3. BMP6 and hepcidin expression in hepatic and small intestinal tissue after sensing iron ex vivo. Ex vivo assessment of BMP6 mrna (A) and protein (B) expression in tissue slices of the liver and the small intestine after treatment with FeSO 4 or holotransferrin (50 mol/l; 1 hour for RNA and 4 hours for protein analysis). Only intestinal tissue revealed a marked increase of BMP6 expression after sensing iron (*P.05 compared with PBS control). (C) Hepatic hepcidin mrna expression was not altered by treatment with FeSO 4 or holo-transferrin. revealed a significant induction (Figure 5A). However, this compensatory up-regulation of BMP2, 4, and 9 in the small intestine was apparently not sufficient to assure the reduced hepcidin expression and the iron overload phenotype of BMP6 / mice presented in this study. In accordance, treatment of primary hepatocytes isolated from BMP6 / mice with different recombinant BMPs (100 ng/ml BMP2, 4, 6, or 9 for 12 hours) revealed BMP6 as the most potent stimulator of hepcidin expression (approximately 90-fold induction; Figure 5B). Furthermore, we performed coculture experiments with tissue samples from the small intestine and the liver of wild-type and BMP6 / mice, respectively. Incubation of cocultures of small intestine from BMP6 / and liver from BMP6 / mice with iron sulfate (5 hours) resulted in a marked up-regulation of hepatic hepcidin mrna expression (Figure 5C). In contrast, iron sulfate did not alter hepatic hepcidin expression in cocultures of small intestine from BMP6 / and liver from BMP6 / mice (Figure 5C). These data confirm that BMP6 is critical for iron homeostasis and that it is functionally nonredundant with other members of the BMP family. To elucidate the molecular background of this specificity, we analyzed the expression pattern of BMP receptor subunits in primary hepatocytes and observed strong expression of the receptor subunits BMPRIA and ActRIIA (Figure 5D) both primarily used by BMP6 23,24 in BMP6 / and BMP6 / mice. In contrast, expression of BMPRIB, ActRIIB, or AMHR2, which serve as receptors for other BMPs except BMP6, was weak or undetectable (Figure 5D). This expression pattern of BMP receptors may explain the critical role of BMP6 in regulating hepatocellular hepcidin expression. Discussion Hepcidin can be induced by several BMPs in vitro, including BMP2, BMP4, and BMP9, 9,21 but, recently, Meynard et al 11 and Andriopoulos et al 12 revealed that BMP6 plays a central role in the regulation of hepatic hepcidin expression and iron homeostasis. Both groups observed that 2-month-old BMP6 / mice already show strong iron accumulation and reduced hepcidin levels in the liver. However, and interestingly, this affection of iron metabolisms develops later in the present study investigating BMP6 / mice on a 129Sv/Ev background. We suggest that this difference is explained by strainspecific responsiveness to iron overload. 20,25,26 Younger BMP6 / mice on 129Sv/Ev background developed severe hepatic iron overload upon single parenteral application of iron-dextran as well as chronic iron overload by feeding iron-supplemented diet for several weeks. Thus, these findings are novel in indicating that BMP6 deficiency affects both short-term response to iron

8 January 2010 BMP6 REGULATES HEPATIC HEPCIDIN EXPRESSION 379 Figure 4. BMP6 expression in the liver and the small intestine after sensing iron in vivo. BMP6 mrna (A and C) and protein (B and D) expression were analyzed in the liver and the small intestine after (a single) parenteral application of iron-dextran or dextran as control (n 5 mice/group). Furthermore, mice were fed an iron-balanced (control; 5 or 6 mice/group), an iron-supplemented (11 mice/group), or an iron-deficient (8 mice/group) diet for 3 weeks. Iron sensing resulted in a significant induction of BMP6 expression in the small intestine. Immunohistologic analysis of the liver (E) and the small intestine (F) after iron exposure. (E) Liver sections of 2- to 3-month-old BMP6 / mice revealed no specific BMP6 staining after parenteral application of iron-dextran or feeding iron-supplemented and irondeficient diets. The visible staining around the hepatocytes is nonspecific and because of iron overload as determined in the BMP6 / mice. (F) Immunohistochemical analysis of the small intestine revealed specific BMP6 staining in enterocytes of 2- to 3-month-old BMP6 / mice injected with iron-dextran or fed with an iron-supplemented diet.

9 380 ARNDT ET AL GASTROENTEROLOGY Vol. 138, No. 1 Figure 5. Intestinal BMP expression after sensing iron in vivo, effect of different BMPs on hepatocellular hepcidin expression, and analysis of hepatocellular BMP-receptor expression in vitro. (A) Feeding an iron-supplemented diet increased expression levels of BMP2, BMP4, and BMP9 in the small intestine of BMP6 / but not BMP6 / mice as determined by quantitative RT-PCR (*P.05 compared with BMP6 / mice). (B) In accordance, treatment of primary hepatocytes isolated from BMP6 / mice with different recombinant BMPs (100 ng/ml BMP2, 4, 6, or 9 for 12 hours) revealed BMP6 as the most potent stimulator of hepcidin mrna expression (*P.05 compared with control). (C) Ex vivo coculture experiments with tissue samples from the small intestine and the liver of BMP6 / and BMP6 / mice. Incubation of cocultures of small intestine from BMP6 / and liver from BMP6 / mice with FeSO 4 (5 hours) resulted in a strong induction of hepatic hepcidin mrna expression (P.05 compared with PBS). (D) The expression of different BMP-receptor subunits in primary hepatocytes isolated from BMP6 / and BMP6 / mice was analyzed by quantitative RT-PCR. BMPRIA (Alk-3), BMPR2, and ActRIIA are strongly expressed in hepatocytes. BMPRIB (Alk-6), ActRIIB, AMHR2, and ActRIA (Alk-2) are only weakly expressed in hepatocytes. and iron homeostasis. Moreover, we observed a striking increase of BMP6 in the serum of wild-type mice after short-term as well as chronic iron overload and newly identified intestinal epithelial cells of the small intestine as the cellular source of the increased BMP6 expression in response to sensing iron. BMP6 / mice revealed an increased expression of other BMP molecules in place of BMP6 in the small intestine upon iron overload; however, this compensatory up-regulation is not sufficient to balance the loss of BMP6 regarding hepatic hepcidin expression and iron accumulation. In vitro and ex vivo experiments confirmed BMP6 as the most potent inductor of hepatocellular hepcidin expression compared with other BMPs. As potential explanation, analysis of hepatocytes revealed strong expression of a pattern of receptor subunits preferentially used by BMP6 but not other BMPs. 23,24 Furthermore, the observed second band in Western blot analysis of BMP6 expression in the serum may be indicative of glycosylation, and, notably, glycosylation plays an important role in type I receptor specificity of BMP6. 27,28 Together, these findings further confirm BMP6 as the main upstream regulator of hepatic hepcidin expression. Figure 6 depicts a working model for hepcidin regulation after iron exposure. Iron absorption in enterocytes leads to the activation of BMP6 expression and, subsequently, to the delivery of BMP6 to the liver via portal circulation. BMP6 binds to the type I and II BMP receptors BMPRIA and ActRIIA and to the coreceptor hemojuvelin on hepatocytes and activates the Smad-signaling pathway through phosphorylation of Smad1/5/8 and complex formation with Smad4. 8,10 This complex translocates into the nucleus and stimulates the expression of hepcidin. In summary, our study adds new and complementary information to recently published data in the field of

10 January 2010 BMP6 REGULATES HEPATIC HEPCIDIN EXPRESSION 381 Figure 6. Working model for hepcidin regulation after iron exposure. Communication between the small intestine and the liver regulates hepcidin expression. Iron absorption in enterocytes leads to the activation of BMP6 expression and, subsequently, to the delivery of BMP6 to the liver via portal circulation. BMP6 binds to type I and II BMP receptors BMPRIA and ActRIIA and to the coreceptor hemojuvelin (HJV) on hepatocytes and activates the Smad signaling pathway through phosphorylation of Smad1/ 5/8 and complex formation with Smad4. This complex translocates into the nucleus and stimulates the expression of hepcidin. 8,10 hepcidin regulation by BMP6. Most importantly, we discovered that BMP6, the central regulator of hepatic hepcidin expression, is predominantly produced in the small intestine and not in the liver or by macrophages upon iron loading. Thus, this study reveals a previously unrecognized mechanism in which intestinal epithelial cells of the small intestine play a central role in iron homeostasis other than their role in iron absorption, ultimately regulating absorption by sensing iron status. Supplementary Data Note: To access the supplementary material accompanying this article, visit the online version of Gastroenterology at and at doi: /j.gastro References 1. Lee P, Peng H, Gelbart T, et al. Regulation of hepcidin transcription by interleukin-1 and interleukin-6. Proc Natl Acad SciUSA 2005;102: Nemeth E, Rivera S, Gabayan V, et al. IL-6 mediates hypoferremia of inflammation by inducing the synthesis of the iron regulatory hormone hepcidin. J Clin Invest 2004;113: Hentze MW, Muckenthaler MU, Andrews NC. Balancing acts: molecular control of mammalian iron metabolism. Cell 2004; 117: Pietrangelo A. Hereditary hemochromatosis. Biochim Biophys Acta 2006;1763: Roetto A, Papanikolaou G, Politou M, et al. Mutant antimicrobial peptide hepcidin is associated with severe juvenile hemochromatosis. Nat Genet 2003;33: Weiss G, Gordeuk VR. Benefits and risks of iron therapy for chronic anaemias. Eur J Clin Invest 2005;35(Suppl 3): Weiss G, Goodnough LT. Anemia of chronic disease. N Engl J Med 2005;352: Babitt JL, Huang FW, Wrighting DM, et al. Bone morphogenetic protein signaling by hemojuvelin regulates hepcidin expression. Nat Genet 2006;38: Babitt JL, Huang FW, Xia Y, et al. Modulation of bone morphogenetic protein signaling in vivo regulates systemic iron balance. J Clin Invest 2007;117: Wang RH, Li C, Xu X, et al. A role of SMAD4 in iron metabolism through the positive regulation of hepcidin expression. Cell Metab 2005;2: Meynard D, Kautz L, Darnaud V, et al. Lack of the bone morphogenetic protein BMP6 induces massive iron overload. Nat Genet 2009;41: Andriopoulos B Jr, Corradini E, Xia Y, et al. BMP6 is a key endogenous regulator of hepcidin expression and iron metabolism. Nat Genet 2009;41: Solloway MJ, Dudley AT, Bikoff EK, et al. Mice lacking Bmp6 function. Dev Genet 1998;22: Muhlbauer M, Fleck M, Schutz C, et al. PD-L1 is induced in hepatocytes by viral infection and by interferon- and - and mediates T-cell apoptosis. J Hepatol 2006;45: Hellerbrand C, Bosserhoff AK, Seegers S, et al. Mutation analysis of the HFE gene in German hemochromatosis patients and controls using automated SSCP-based capillary electrophoresis and a new PCR-ELISA technique. Scand J Gastroenterol 2001;36: Arndt S, Bosserhoff AK. Reduced expression of TANGO in colon and hepatocellular carcinomas. Oncol Rep 2007;18: Kirsch R, Sijtsema HP, Tlali M, et al. Effects of iron overload in a rat nutritional model of non-alcoholic fatty liver disease. Liver Int 2006;26: Edison ES, Bajel A, Chandy M. Iron homeostasis: new players, newer insights. Eur J Haematol 2008;81: Batts KP. Iron overload syndromes and the liver. Mod Pathol 2007;20(Suppl 1):S31 S Kautz L, Meynard D, Monnier A, et al. Iron regulates phosphorylation of Smad1/5/8 and gene expression of Bmp6, Smad7, Id1, and Atoh8 in the mouse liver. Blood 2008;112: Truksa J, Peng H, Lee P, et al. Bone morphogenetic proteins 2, 4, and 9 stimulate murine hepcidin 1 expression independently of Hfe, transferrin receptor 2 (Tfr2), and IL-6. Proc Natl Acad Sci U S A 2006;103: Yu PB, Hong CC, Sachidanandan C, et al. Dorsomorphin inhibits BMP signals required for embryogenesis and iron metabolism. Nat Chem Biol 2008;4: Aoki H, Fujii M, Imamura T, et al. Synergistic effects of different bone morphogenetic protein type I receptors on alkaline phosphatase induction. J Cell Sci 2001;114: Derynck R, Zhang YE. Smad-dependent and Smad-independent pathways in TGF- family signalling. Nature 2003;425: Rodriguez A, Hilvo M, Kytomaki L, et al. Effects of iron loading on muscle: genome-wide mrna expression profiling in the mouse. BMC Genomics 2007;8: Courselaud B, Troadec MB, Fruchon S, et al. Strain and gender modulate hepatic hepcidin 1 and 2 mrna expression in mice. Blood Cells Mol Dis 2004;32: Saremba S, Nickel J, Seher A, et al. Type I receptor binding of bone morphogenetic protein 6 is dependent on N-glycosylation of the ligand. FEBS J 2008;275: Gitelman SE, Kobrin MS, Ye JQ, et al. Recombinant Vgr-1/BMP- 6-expressing tumors induce fibrosis and endochondral bone formation in vivo. J Cell Biol 1994;126: Scheuer PJ, Williams R, Muir AR. Hepatic pathology in relatives of patients with haemochromatosis. J Pathol Bacteriol 1962;84:

11 382 ARNDT ET AL GASTROENTEROLOGY Vol. 138, No. 1 Received April 14, Accepted September 17, Reprint requests Address requests for reprints to: Anja-Katrin Bosserhoff, PhD, Institute of Pathology, University of Regensburg, Franz-Josef-Strauss- Allee 11, D Regensburg, Germany. anja.bosserhoff@ klinik.uni-regensburg.de; fax: (49) Acknowledgments The authors thank Elizabeth Robertson for kindly providing the BMP6-deficient mice; Eva Wacker, Simone Kaufmann, Martina Waeber, and Rudolf Jung for technical assistance; Daniel Wagner for expert animal care; and Richard Bauer for the design of Figure 6. C.H. and A.-K.B. contributed equally to this paper. Conflicts of interest The authors disclose no conflicts. Funding Supported by grants of the German Research Association and the Medical Faculty of the University of Regensburg (ReForM) (to C.H. and A.-K.B.).