Hepatocyte Nuclear Factor 4 (cid:2) Controls Iron Metabolism and Regulates Transferrin Receptor 2 in Mouse Liver *

Iron is an essential element in biological systems, but excess iron promotes the formation of reactive oxygen species, resulting in cellular toxicity. Several iron-related genes are highly expressed in the liver, a tissue in which hepatocyte nuclear factor 4 (cid:2) (HNF4 (cid:2) ) plays a critical role in controlling gene expression. Therefore, the role of hepatic HNF4 (cid:2) in iron homeostasis was examined using liver-specific HNF4 (cid:2) -null mice ( Hnf4a (cid:3) H mice). Hnf4a (cid:3) H mice exhibit hypoferremia and a significant change in hepatic gene expression. Notably, the expression of transferrin receptor 2 ( Tfr2 ) mRNA was markedly decreased in Hnf4a (cid:3) H mice. Promoter analysis of the Tfr2 gene showed that the basal promoter was located at a GC-rich region upstream of the transcription start site, a region that can be transactivated in an HNF4 (cid:2) -independent manner. HNF4 (cid:2) -dependent expression of Tfr2 was mediated by a proximal promoter containing two HNF4 (cid:2) -binding sites located between the transcription start site and the translation start site. Both the GC-rich region of the basal promoter and the HNF4 (cid:2) -binding sites were required for maximal transactivation. Moreover, siRNA knockdown of HNF4 (cid:2) suppressed TFR2 expression in human HCC cells. These results suggest that Tfr2 is a novel target gene for HNF4 (cid:2) , and hepatic HNF4 (cid:2) plays a critical role in iron homeostasis.

Iron is an essential metallic element in living organisms and is required for the transport of oxygen by hemoglobin in red blood cells and for redox reactions by catalase and various cytochrome P450s. Because most of the iron in the human body is contained in hemoglobin and is recycled from aged red blood cells in the reticuloendothelial system, only 1-2 mg of iron are absorbed at the duodenum, and the amount exactly matches that of iron loss (1,2). Iron deficiency causes anemia, whereas iron overload has a cytotoxic effect due to the production of reactive oxygen species (ROS). 2 Thus, the amount of iron in the organism must be tightly regulated by various processes such as absorption, transport, cellular uptake, and storage (1)(2)(3).
Liver is the most important organ for iron metabolism, and many iron-related genes encoding transferrin (TF), transferrin receptor 2 (TFR2), and hepcidin (HAMP), for example, are highly expressed in hepatocytes. TF is a plasma iron carrier, and iron-loaded TF is mainly endocytosed into most tissues through transferrin receptor 1 (TFR1) (2)(3)(4). Congenital apotransferrinemia, deficiency of serum TF, causes iron deficiency anemia (5). Unlike TFR2, TFR1 is ubiquitously expressed in many tissues and is required for iron delivery to cells (4,6). Mutations of the TFR2 gene cause iron overload disease, hereditary hemochromatosis, and the affinity of TFR2 to holo-TF was 25-fold lower when compared with that of TFR1, indicating that TFR2 functions as an iron sensor unlike TFR1 (7)(8)(9). TFR2 is also an upstream regulator of HAMP, a peptide hormone secreted from liver (10). HAMP binds to iron exporter, ferroportin 1 (FPN1), and induces internalization and degradation of FPN1, resulting in decreased efflux of iron from the duodenum (11). In addition to the TFR2 gene, mutations of the HAMP and FPN1 genes also cause hereditary hemochromatosis (12). Thus, the HAMP-FPN1 axis is a central coordinator and an iron sensor in the body. Agonists and antagonists of these factors, including HAMP, are under development as therapeutic targets for iron overload disease and anemia of inflammation (1).
HNF4␣ is an orphan member of the nuclear receptor superfamily and positively regulates many genes involved in liverspecific functions (13,14). HNF4␣ was shown to transactive at the TF gene through an HNF4␣-binding site within the promoter region in the human hepatoma cell line (15). Moreover, an HNF4␣-binding site in the Hamp promoter was essential for bone morphogenetic protein-and hemojuvelin-induced transactivation (16), and hepatic expression of Hamp mRNA was increased in liver-specific Hnf4a-null (Hnf4a ⌬H ) mice (17). These results suggest that hepatic HNF4␣ could be involved in the control of iron metabolism through these target genes, but the in vivo relationship between HNF4␣ and iron metabolism in adult liver has not been explored. Because Hnf4a ⌬H mice exhibit many phenotypes that attenuate liver-specific functions such as lipid and ammonia metabolism and bile acid synthesis (18 -20), dysregulation of iron metabolism was also suspected in Hnf4a ⌬H mice.
In this study, Hnf4a ⌬H mice were found to exhibit iron metabolism-related phenotypes, resulting in hypoferremia, but no iron deficiency anemia and hemochromatosis was observed. Gene expression profiling revealed that hepatic expression of Tfr2 mRNA was markedly decreased in Hnf4a ⌬H mice. Promoter analysis showed that the Tfr2 distal promoter region containing GC-rich sequences was required for basal transactivation of the Tfr2 gene. In addition, two HNF4␣-binding sites in the proximal promoter region were essential for the maximal transactivation in combination with the proximal GC-rich sequences, revealing that Tfr2 is a novel HNF4␣ target gene in liver.

Experimental Procedures
Animal-Liver-specific Hnf4a-null (Hnf4a ⌬H ) mice and liver-specific Cebpa-null (Cebpa ⌬H ) mice were described previously (18,21). All experiments were performed with 45-day-old male Hnf4a-floxed (Hnf4a f/f ) and Hnf4a ⌬H mice and 2-monthold male Cebpa-floxed (Cebpa f/f ) and Cebpa ⌬H mice. Mice were housed in a pathogen-free animal facility under standard 12-h light/12-h dark cycle with water and chow ad libitum. All experiments with mice were carried out under Association for Assessment and Accreditation of Laboratory Animal Care guidelines with approval of the Animal Care and Use Committee of the NCI, National Institutes of Health, and Gunma University Animal Care and Experimentation Committee.
Serum and Liver Iron Measurements-Mice were anesthetized with 2.5% avertin and decapitated, and trunk blood was collected for the measurement of red blood cells (RBC), hemoglobin, hematocrit, and the mean cell volume, or collected in a serum separator tube (BD Biosciences). The serum was separated by centrifugation at 7,000 ϫ g for 5 min and stored at Ϫ20°C prior to analysis. The serum iron concentration and the unsaturated iron-binding capacity were determined colorimetrically using iron and the total iron-binding capacity kit (Sigma). The total iron-binding capacity was calculated by adding the serum iron concentration and the unsaturated ironbinding capacity. Transferrin saturation (%) was calculated as (serum iron concentration/total iron-binding capacity) ϫ 100. Livers were perfused with phosphate-buffered saline (PBS) to remove serum iron, excised, and dried at 100°C for 16 h. Dried liver samples were dissolved in acidic solution (3 N HCl, 0.6 M trichloroacetic acid) and incubated at 65°C for 20 h. Extracted iron solution was reacted with chromogen solution (1.25 M acetic acid (pH 4.7), 0.008% bathophenanthroline, 0.08% glyceric acid) at room temperature for 30 min, and the liver iron was determined colorimetrically at 510 nm.
IronChip Analysis-For the cDNA microarray experiments, the "IronChip" (version 3.0) was used. Experimental details concerning the selection of the cDNA clones, the preparation of the microarray platform, the synthesis of fluorescent cDNA probes, prehybridization, and hybridization conditions of the microarrays as well as scanning and data analysis are described elsewhere (22)(23)(24). The symbols of the genes mentioned and RefSeq accession number in the text are summarized in Table 4.
Transfection of siRNA-Two kinds of siRNA for human HNF4␣ (10 nM, Sigma, Tokyo, Japan) were independently transfected into HepG2 cells with Lipofectamine RNAiMAX (Life Technologies, Inc.). After 48 h, cells were trypsinized and re-transfected with 10 nM of the same siRNA. After 48 h of re-transfection, cells were harvested, and total RNA was extracted using Tripure Isolation Reagent for RT quantitative PCR. Nucleotide sequences for the siRNA duplexes targeting human HNF4␣ are follows: GGCAGUGCGUGGUGGA-CAA for siHNF4␣-1 and AGAGAUCCAUGGUGUUCAA for siHNF4␣-2.
Western Blot-Liver samples from Hnf4a f/f and Hnf4a ⌬H mice were homogenized in lysis buffer (7 M urea, 2 M thiourea, 1% Triton X-100) and allowed to sit on ice for 30 min. The homogenate was centrifuged at 12,000 ϫ g for 30 min at 4°C, and the supernatants were used as whole cell lysates. The whole cell lysates and serum protein (40 g), determined by Quick Start TM Bradford Dye Reagent (Bio-Rad), were diluted with Laemmli sample buffer, incubated at 65°C for 15 min, and fractionated by 10% SDS-PAGE. The gels were stained with Coomassie Brilliant Blue R-250 or transferred onto a PVDF membrane (GE Healthcare, Tokyo, Japan). The membrane was incubated for 1 h with PBS containing 0.1% Tween 20 and 5% skim milk and then incubated for 1 h with anti-transferrin (Santa Cruz Biotechnology, Dallas, TX), anti-transferrin receptor 2 (Abcam, Tokyo, Japan), and anti-␥-tubulin (TUBG) (Sigma) antibodies. After washing, the membrane was incubated for 1 h with horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology), and the reaction product was visualized using SuperSignal West Pico Chemiluminescent Substrate (Pierce). Expression of TF and TFR2 proteins was quantified by densitometric analysis using ImageJ software and the expression in Hnf4a ⌬H was presented as expression differences relative to the Hnf4a ff/f mice.
Gel Mobility Shift Analysis-Nuclear extracts from HepG2 cells were prepared using NE-PER nuclear and cytoplasmic extraction reagents (Thermo Fisher Scientific, Yokohama, Japan), and gel shift analysis was carried out using LightShift Chemiluminescent EMSA kit (Thermo Fisher Scientific). The following double-stranded probes were used (mutations are indicated as bold and underlined); the HNF4␣-binding site at ϩ182/ϩ201 in the mouse Tfr2 promoter (wild type, gggtgaggtggaccctgaactttcccattcacagctgca and tgcagctgtgaatgggaaagttcagggtccacctcaccc; mutant, gggtgaggttgacgctgtgatttcccattcacagctgca and tgcagctgtgaatgggaaatcacagcgtcaacctcaccc); the HNF-4␣-binding site at ϩ830/ϩ842 in the mouse Tfr2 promoter (wild type, gaaccaagttggccaaagtctgctctgaggt and acctcagagcagactttggccaacttggttc; mutant, gaaccaagttggcccgcgtctgctctgaggt and acctcagagcagacgcgggccaacttggttc); and the HNF4␣-binding site at Ϫ203/Ϫ192 in the mouse ornithine transcarbamylase (Otc) promoter as a positive control (gttaggcttaaagttcaagtg and cacttgaactttaagcctaac) (19). Nuclear extracts (3 g) and the 5Ј-biotin-labeled probes of the HNF4␣-binding sites for the Tfr2 promoter (wild type) were added, and the reaction mixture was incubated on ice for 10 min. For competition experiments, a 50-fold excess of unlabeled probe was added to the reaction mixture, and the mixture was incubated on ice for 10 min prior to the addition of the 5Ј-biotin-labeled probe. For supershift analysis, 1 g of anti-HNF4␣ or anti-peroxisome proliferator-activated receptor ␣ antibodies (Santa Cruz Biotechnology) was added to the reaction mixture, and the mixture was incubated on ice for 10 min after the addition of the 5Ј-biotin-labeled probe. DNA-protein complexes were fractionated by 7% PAGE containing 5% glycerol and blotted onto a Biodyne B nylon membrane (Pall, Tokyo, Japan). After washing, DNA-protein complexes were visualized using a detection module in the kit on an ImageQuant LAS4000.
Chromatin Immunoprecipitation-HepG2 cells cultured in a 10-cm dish were fixed in 0.5% formaldehyde for 10 min and then quenched with 125 mM glycine for 5 min at room temperature. After washing with ice-cold PBS, the cells were resuspended in 3 ml of Lysis buffer 1 (50 mM HEPES-KOH (pH 7.5), 140 mM NaCl, 1 mM EDTA, 10% glycerol, 0.5% Nonidet P-40, 0.25% Triton X-100, and protease inhibitor (Roche Applied Science)) on ice for 10 min and then centrifuged at 1,400 ϫ g for 5 min. The cell pellet was resuspended in 3 ml of Lysis buffer 2 (10 mM Tris-HCl (pH 8.0), 200 mM NaCl, 1 mM EDTA, and 5 mM EGTA) for 10 min at room temperature and then centrifuged at 1,400 ϫ g for 5 min. The cell pellet was resuspended in 1 ml of Lysis buffer 3 (10 mM Tris-HCl (pH 8.0), 300 mM NaCl, 1 mM EDTA, 0.5 mM EGTA, and 0.1% sodium deoxycholate). Liver samples stored at Ϫ80°C were ground to pieces with a pestle and mortar under liquid nitrogen and fixed in PBS containing 20 mM sodium butyrate, 1% formaldehyde, and protease inhibitor mixture for 10 min at room temperature. After centrifugation, the pellet was resuspended in Lysis buffer (50 mM Tris-HCl (pH 8.0), 10 mM EDTA, 1% SDS, and 20 mM sodium butyrate). The cell lysate from HepG2 cells and liver samples was disrupted by an ultrasonicator (UR-20P, TOMY, Tokyo, Japan) for 5 min on ice and then 1% Triton X-100 was added, followed by centrifuging at 20,000 ϫ g for 10 min. A small volume of the supernatant was stored at 4°C as the input samples. The remaining supernatant was pre-cleared by adding of 15 l of a 50% slurry of protein G-Sepharose 4 Fast Flow (GE Healthcare) with sonicated salmon sperm DNA and rotated for 4 h at 4°C, followed by centrifuging at 1,900 ϫ g for 5 min. The supernatant was divided into two pieces and incubated with anti-HNF4␣ antibody (4 g, Santa Cruz Biotechnology) or control normal goat IgG for 16 h at 4°C. The reaction mixture was centrifuged at 1,900 ϫ g for 5 min at 4°C, and the pellet collected was washed for 5 min at 4°C with 1 ml of RIPA-1 buffer (50 mM Tris-HCl (pH 8.0), 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, and 0.1% sodium deoxycholate), RIPA-2 buffer (50 mM Tris-HCl (pH 8.0), 300 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% SDS, and 0.1% sodium deoxycholate), LiCl wash solution (10 mM Tris-HCl (pH 8.0), 0.25 M LiCl, 1 mM EDTA, 0.5% Nonidet P-40, and 0.5% sodium deoxycholate), and TE (10 mM Tris-HCl (pH 8.0) and 1 mM EDTA), respectively. Then the ChIP direct elution buffer (10 mM Tris-HCl (pH 8.0), 300 mM NaCl, 5 mM EDTA, and 0.5% SDS) was added to the pellet and incubated for 16 h at 65°C for decrosslinking. After treatment with RNase A for 30 min at 37°C and proteinase K for 2 h at 55°C, the DNA was purified and used for PCR and real time PCR. Enrichment of the HNF4␣ binding was normalized to the input samples using the ⌬⌬Ct method and expressed as fold-enrichment as compared with the control normal IgG antibody. The following primers were used for amplification of the HNF4␣-binding sites in the human TFR2 promoter (gggaactaggaggccaaagt and tctcccctgccaatctctc), the human MIR-194/192 gene without HNF4␣-binding site as a negative control (ccttgtgagggcacaccttt and aaagccaggcagtcagtgct), the HNF4␣-binding sites in the mouse Tfr2 promoter (caggaagaccggctaacg and tcttgggtctggttgctagg), and the mouse Hmgcs2 gene without HNF4␣-binding site as a negative control (gatccctgggactcacaca and gaatgcacatttatggaggtca).
Statistical Analysis-All values are expressed as the mean Ϯ S.D. All data were analyzed by the unpaired Student's t test for significant differences between the mean values of each group.

Results
Hnf4a ⌬H Mice Exhibit Hypoferremia-Hematological parameters are good indicators for alternation of iron homeostasis. Here, hepatic HNF4␣ was found to maintain serum iron levels ( Table 1). Hnf4a ⌬H mice exhibited a 60% reduction in serum iron compared with control Hnf4a f/f mice. Likewise, total iron binding capacity, an index of the total amount of serum iron that transferrin can bind, was also reduced in Hnf4a ⌬H mice, indicating a lower availability of transferrin protein in Hnf4a ⌬H mice. However, because no significant difference in transferrin saturation was observed in Hnf4a ⌬H mice, iron-binding transferrin is normal in Hnf4a ⌬H mice. Moreover, red blood cell counts (RBC), hemoglobin levels, the hematocrit, and the mean cell volume were unchanged in Hnf4a ⌬H mice compared with Hnf4a ⌬H mice (Table 2). Furthermore, no significant difference in liver iron content was detected in Hnf4a ⌬H mice ( Expression of Genes Involved in Iron Metabolism is Altered in Hnf4a ⌬H Mice-The expression profile of mRNAs encoded by iron metabolism-related genes was analyzed in Hnf4a ⌬H and Hnf4a f/f mouse livers using IronChip microarray analysis (Table 4). Hepatic expression of many mRNAs significantly changed in Hnf4a ⌬H mice. Among the down-regulated mRNAs were transferrin (Tf), transferrin receptor 2 (Tfr2), hemojuvelin (Hjv), ferritin light chain (Ftl), ferritin heavy chain (Fth), hemopexin (Hpx), catalase (Cat), Cu-Zn superoxide dismutase (Sod1), and glutathione peroxidase 1 (Gpx1). Up-regulated mRNAs included transferrin receptor 1 (Tfr1), hepcidin (Hamp), and ceruloplasmin (Cp) mRNAs. Microarray data were validated by quantitative RT-PCR (Fig. 1A). Because atransferrinemia/hypotransferrinemia is caused by recessive mutations in the Tf gene resulting in iron-deficient hypochromic anemia in humans and mice (25,26), reduced expression of TF may cause hypoferremia. Indeed, as expected from reduced Tf mRNA expression, serum TF protein in Hnf4a ⌬H mice was decreased by 66% compared with Hnf4a f/f mice (Fig. 1B). These data show that misregulation of a previously validated HNF4␣ target gene is preserved in Hnf4a ⌬H mice (15). Thus, HNF4␣dependent regulation of Tf could partially contribute to the hypoferremia observed in this mouse model.
In addition, decreased expression of hepatic Tfr2 mRNA to 30% was observed in Hnf4a ⌬H mice, although expression of Tfr1 mRNA encoding a receptor with affinity to holo-TF 25-fold higher than TFR2 was increased to 4-fold (Fig. 1A) (8). TFR2 protein in Hnf4a ⌬H mice was also decreased by 19% compared with Hnf4a f/f mice (Fig. 1C). Because TFR2 and HNF4␣ are highly expressed in the liver and mutations in the TFR2 gene cause type III hemochromatosis in humans (6, 7), further studies were focused on the Tfr2 gene. Previous reports suggest that the promoter activity of the Tfr2 gene was induced by C/EBP␣ expression (27). However, expression of hepatic Tfr2 was unchanged in liver-specific Cebpa-null (Cebpa ⌬H ) mice (Fig.  1D), indicating that C/EBP␣ is unlikely to be the main regulator of Tfr2 expression in the liver.
Promoter Analysis of the Distal Region of the Tfr2 Promoter-Promoter analysis of the mouse Tfr2 gene was performed to determine whether HNF4␣ directly regulates Tfr2 expression.   Analysis of the distal promoter was carried out in HepG2 cells that highly express endogenous HNF4␣ and TFR2 (6) revealing significant promoter activities with the region between Ϫ116 and Ϫ46 from the transcription start site (Fig. 2B). Activity of the Otc promoter that has two HNF4␣-binding sites was completely suppressed by HNF4␣ knockdown using siRNA; however, no repression of the promoter activities was detected in all Tfr2 promoters by siRNA against HNF4␣ (Fig. 2C), indicating that the distal promoter of the Tfr2 gene was activated in an HNF4␣-independent manner. Searching the JASPER database for transcription factor-binding sites in the region between Ϫ116 and Ϫ46 revealed three GC boxes (GC-1, 2, and -3) that were predicted between Ϫ91 and Ϫ53 (Fig. 2D). Furthermore, the stepwise transactivation of the promoter activity was observed in the presence of these GC boxes (Fig. 2B). To determine which regions of these GC boxes are sufficient for transactivation, mutations were introduced into each GC box. Mutations of the GC-2 region significantly suppressed the promoter activity, and mutations of all GC regions further suppressed the activity (Fig. 2E), indicating that the GC-2 region is a central cis-element for transactivation, and the GC-1 and GC-3 regions are required for the maximal transactivation of the distal promoter.
Promoter Analysis of the Proximal Region of the Tfr2 Promoter-As described above, regions that respond to HNF4␣ were not identified in the distal promoter. No transactivation was detected in this promoter region (ϩ505/ϩ1144 and Ϫ46/ ϩ1144) as compared with the promoterless and the distal promoter (Ϫ116/ϩ67, see Fig. 3A). Therefore, the proximal promoter was analyzed for the presence of an HNF4␣-dependent enhancer. At first, the full-length proximal promoter (Ϫ46/ ϩ1144) had enhancer activity in the presence of the core Tk promoter (Fig. 3B). In addition, the shorter promoter regions (Ϫ46/ϩ505 and Ϫ46/ϩ1144) had weak enhancer activities, showing that these regions might cooperatively contribute to enhancer transactivation. Next, the Ϫ46/ϩ504 and ϩ505/ ϩ1144 proximal promoters were transactivated by HNF4␣ expression vector, and the full-length promoter was further transactivated by HNF4␣ in the presence of the Tk mini promoter (Fig. 3C). These results indicate that the proximal promoter contains at least two HNF4␣-dependent elements located at Ϫ46/ϩ504 and ϩ505/ϩ1144. In addition, the proximal promoters with the distal promoter (Ϫ116/ϩ504, Ϫ116/ ϩ896, and Ϫ116/ϩ1144) was also transactivated by HNF4␣ as compared with the distal promoter alone (Ϫ116/ϩ67), with the result that HNF4␣ has the potential to transactivate the native Tfr2 promoter via two regions located at ϩ67/ϩ504 and ϩ505/ ϩ896 (Fig. 3D). By searching with the JASPER database, three potential HNF4␣-binding sites were found at ϩ182/ϩ194, ϩ189/ϩ201, and ϩ830/ϩ842, and the binding sites located at ϩ182/ϩ194 and ϩ189/ϩ201 overlapped (Fig. 3E). Of these, the binding site at ϩ830/ϩ842 was highly conserved among species (Fig. 4A). To determine whether these predicted HNF4␣-binding sites could transactivate the Tfr2 gene in an HNF4␣-dependent manner, mutations were introduced into the binding sites. Because the binding site at ϩ182/ϩ194 and ϩ189/ϩ201 overlapped, mutations were introduced into the middle region at ϩ182/ϩ201 so that HNF4␣ could not bind to both sites. Consequently, disruption of each binding site decreased HNF4␣dependent transactivation, and disruption of both binding sites further repressed the transactivation, indicating that these HNF4␣-binding sites were transactivated in an HNF4␣-dependent manner (Fig. 4B).
To investigate whether the minimal promoter region in the distal promoter would be essential for transactivation by HNF4␣, mutations were introduced into both the GC boxes and HNF4␣-binding sites (Fig. 4C). In the presence of the HNF4␣-binding sites in the proximal promoter, mutations of  Direct Binding of HNF4␣ to the Proximal Region of the Tfr2 Promoter-To determine whether HNF4␣ can directly bind to both HNF4␣-binding sites in the Tfr2 promoter, gel mobility shift analysis was performed (Fig. 5A). Nuclear extracts from HepG2 cells bound to the distal (ϩ182/ϩ201) and proximal (ϩ830/ϩ842) HNF4␣-binding sites (Fig. 5A, lane 2, lower  arrows). This complex was diminished by the addition of excess amounts of unlabeled Tfr2 competitor and the Otc competitor that contains a bona fide HNF4␣-binding site (Fig. 5A, lanes 3  and 4) (19) but not the competitor that has mutations in the HNF4␣-binding site of the Tfr2 promoter (Fig. 5A, lane 5). Moreover, the complex was supershifted by anti-HNF4␣ antibody (Fig. 5A, lane 6, upper arrow) but not the unrelated antiperoxisome proliferator-activated receptor ␣ antibody (lane 7). These results indicate that HNF4␣ indeed binds to the distal and proximal regions of the Tfr2 promoter. Next, chromatin immunoprecipitation (ChIP) was used to determine whether HNF4␣ directly binds to the Tfr2 promoter in vivo. HNF4␣ bound to the predicted HNF4␣-binding sites in HepG2 cells as compared with IgG control (Fig. 5B). In addition, ChIP analysis using the livers of Hnf4a f/f and Hnf4a ⌬H mice indicated that hepatic HNF4␣ in Hnf4a f/f mice significantly bound to the promoter region, and the binding was lower in Hnf4a f/f mouse livers (Fig. 5C), suggesting that HNF4␣ directly and physiologically binds to the promoter region of the Tfr2 gene in human and mouse liver.
To determine whether regulation of hepatic Tfr2 expression by HNF4␣ observed in mice could be also maintained in a human hepatocyte-derived cell line, HNF4␣ knockdown by siRNAs was performed revealing that HNF4␣ was suppressed in HepG2 cells (Fig. 5D). Expression of TFR2 mRNA was decreased by HNF4␣ knockdown, indicating that expression of the TFR2 mRNA is positively correlated with HNF4␣ expression in HepG2 cells (Fig. 5D).

Discussion
The present data show that serum iron levels are decreased in Hnf4a ⌬H mice without significant changes in RBC, HBG, and hematocrit and no evidence for iron deficiency anemia. Changes in the expression of multiple genes in Hnf4a ⌬H mice make it difficult to ascribe a phenotype to loss of expression of a single gene. However, expression of hepatic Tf mRNA and serum TF protein was decreased in Hnf4a ⌬H mice, due to loss of direct transactivation of TF by HNF4␣ (15). Hypotransferrinemic mice carrying a mutation in the Tf gene die from iron deficiency anemia before weaning, but these mice can be rescued by injection of serum or purified TF (26). Thus, the reduced expression of TF would only partially cause hypoferremia in Hnf4a ⌬H mice. Serum TF protein still remains about 70% of normal as revealed by Western blot data, suggesting that the symptoms in Hnf4a ⌬H mice would not be severe enough to develop iron deficiency anemia. Because serum TF protein has FIGURE 1. Hepatic expression of iron metabolism-related genes in Hnf4a ⌬H mice. A, real time PCR for Hnf4a, Tf, Tfr2, Hjv, Ftl, Fth, Hpx, Cat, Sod1, and Gpx1 (upper panel), and Tfr1, Hamp, Hfe, Fpn1, Heph, and Cp (lower panel) mRNAs from total liver RNA of Hnf4a ⌬H and Hnf4a f/f mice (n ϭ 5 for each genotype). The normalized expression in Hnf4a ⌬H mice was presented relative to that in Hnf4a f/f mice. B, Western blot of TF protein from serum of Hnf4a ⌬H and Hnf4a f/f mice (n ϭ 3 for each genotype) (upper panel). Coomassie Brilliant Blue stain of serum total protein of Hnf4a ⌬H and Hnf4a f/f mice as loading controls (lower panel) is shown. The expression of TF protein was quantified by densitometric analysis using ImageJ software. The normalized TF expression in Hnf4a ⌬H was presented as relative to that in Hnf4a f/f mice. C, Western blot of TFR2 (upper panel) and TUBG (lower panel) protein from livers of Hnf4a ⌬H and Hnf4a f/f mice (n ϭ 3 for each genotype). The TFR2 expression normalized by TUBG in Hnf4a ⌬H was presented as relative to Hnf4a f/f mice. D, real time PCR for Cebpa and Tfr2 mRNAs from total liver RNA of Cebpa ⌬H and Cebpa f/f mice (n ϭ 5 for each genotype). The normalized expression in Cebpa ⌬H mice was presented relative to that in Cebpa f/f mice. Data are mean Ϯ S.D. *, p Ͻ 0.05; **, p Ͻ 0.01; ***, p Ͻ 0.001. a relatively long half-life (8 -10 days), the majority of serum TF protein may remain in Hnf4a ⌬H mice despite that TF protein is mainly expressed in the liver, and hepatic expression of Tf mRNA in Hnf4a ⌬H mice was decreased to 20% compared with control mice. Furthermore, expression of Tfr1 mRNA was markedly increased in Hnf4a ⌬H mice. Tfr1 mRNA is post-transcriptionally regulated via iron-response elements in the 3Ј-untranslated regions by iron-regulatory proteins during low iron levels, resulting in Tfr1 mRNA stabilization and increased TFR1 protein (3,28). However, both liver iron content and expression of IRP-1 and -2 remained unchanged in Hnf4a ⌬H mice, suggesting that TFR1 protein expression would also be increased in Hnf4a ⌬H mice as well as Tfr1 mRNA without posttranscriptional regulation. Thus, increased iron uptake into the liver by increased TFR1 may cause the hypoferremia found in Hnf4a ⌬H mice.
The other cause of hypoferremia in Hnf4a ⌬H mice might be due to up-regulation of Hamp expression. Excess HAMP degrades or internalizes duodenal FPN1 and inhibits iron export into the bloodstream, leading to induction of hypoferremia, and thus the HAMP-FPN1 axis is the most important regulator of iron homeostasis (1,12). Expression of HAMP is positively regulated at the transcriptional level through hemojuvelin, TFR2, and the pro-inflammatory IL6/STAT3 pathway (1,2,10,29). However, despite the decreased expression of Hjv and Tfr2 mRNA, Hamp mRNA is elevated 3-fold in Hnf4a ⌬H mice. In contrast, Hnf4a ⌬H mice showed hepatosteatosis and increased serum alanine transaminase, an indicator of liver dysfunction and inflammation (18). Moreover, ROS are known to induce inflammation through the production of proinflammatory cytokines (30), and hepatic expression of CAT, SOD1, and GPX1 that degrade ROS was decreased in Hnf4a ⌬H mice (31). These data indicate that the accumulation of ROS could lead to liver inflammation, and the resultant up-regulation of Hamp in Hnf4a ⌬H mice could be caused by enhanced inflammatory signaling.
Even though reduced expression of hepatic Tfr2 that causes hereditary hemochromatosis when mutated in humans (7), Hnf4a ⌬H mice did not show evidence of this disorder. Because hepatic Hamp expression in both the complete and liver-specific Tfr2-null mice was decreased at a similar age as the Hnf4a ⌬H mice used in this study (32,33) and because HAMP is the downstream regulator of TFR2, up-regulation of Hamp might be the main reason why Hnf4a ⌬H mice do not show hemochromatosis despite the down-regulation of Tfr2 expression. It is noteworthy that liver-specific Tfr2-null mice also exhibited elevated transferrin saturation (33), suggesting that absorption of iron from the duodenal FPN1 would be increased by reduced expression of Hamp. In contrast, Hnf4a ⌬H mice exhibited normal transferrin saturation. Thus, the phenotype of Hnf4a ⌬H mice does not overlap with those of liver-specific Tfr2-null mice, despite the fact that HNF4␣ is an upstream regulator of TFR2. A potential reason for this discrepancy is that lack of HNF4␣ may have a profound effect on hepatic iron metabolism compared with that of TFR2.
This study further shows that HNF4␣ binds and activates transcription of Tfr2. Previous studies showed that the mouse Tfr2 proximal and distal promoter was transactivated in the FIGURE 4. Requirement of the minimal promoter of the Tfr2 gene for transactivation by HNF4␣. A, sequence alignment of the proximal promoter of the Tfr2 gene in mouse, human, rat, rabbit, gorilla, chimpanzee, cow, dolphin, horse, panda, and elephant. Predicted HNF4␣-binding site is boxed. Completely conserved nucleotides among the species are shown by asterisks. B, mutations were introduced into the HNF4␣-binding sites in the proximal promoter of the Tfr2 gene. Empty vector or HNF4␣ expression vector was co-transfected with the mutated proximal promoter and the minimal distal promoter, and the enhancer activity was measured. C, mutations were introduced into the HNF4␣-binding sites and GC-rich sequences in the Tfr2 promoter. Empty vector or HNF4␣ expression vector was co-transfected with the mutated promoter, and the enhancer activity was measured. The normalized activity Ϯ S.D. (n ϭ 3) is presented as promoter activity based on empty vector-transfected control.
presence of GATA-1 and C/EBP␣ (27). Of these factors, GATA-1 would not be an essential factor to transactivate Tfr2 in the liver because GATA-1 is a master regulator for erythroid differentiation and transactivates erythroid-specific genes (34,35). Moreover, C/EBP␣ is enriched in the liver and positively regulates many liver-enriched genes (36). Although the expression of hepatic Cebpa was decreased by 45% in Hnf4a ⌬H mice (37), no significant change of hepatic Tfr2 mRNA expression was detected in Cebpa ⌬H mice. Additionally, the transactivation potential of C/EBP␣ alone for the Tfr2 promoter without GATA-1 is much lower as compared with that with GATA-1 (27), and the contribution of C/EBP␣ to Tfr2 regulation would be minor in the liver. Although details on transcriptional regulation of the Tfr2 gene in the liver are not clear, HNF4␣ could directly regulate the Tfr2 expression through at least two HNF4␣-binding sites. Unlike the highly conserved binding site at ϩ830/ϩ842, the binding site at ϩ182/ϩ201 is not conserved among the species, and disruption of both binding sites did not completely inhibit the HNF4␣-dependent transactivation. However, HNF4␣ knockdown greatly suppressed TFR2 expression in human hepatoma cells, indicating that there might be additional HNF4␣-binding sites within the proximal promoter. This study revealed that GC-rich regions in the distal promoter were also important for maximal transactivation of the Tfr2 gene promoter. It was reported that cooperation of the HNF4␣ and GC-rich regions is important for transactivation of the human APOC3 gene and mouse Slc27a5 gene (38,39), indicating that the GC-rich regions and the HNF4␣-binding sites in the Tfr2 promoter can be the basal cis-elements and the liverspecific enhancer elements, respectively.
Although HNF4␣ was found to directly regulate Tfr2 expression, the mechanism for repressed expression of hepatic Hamp expression in Hnf4a ⌬H mice could not be determined in this study. Because HNF4␣ is a master regulator for maintenance of liver-specific functions and severe liver dysfunction and inflammation were actually observed in Hnf4a ⌬H mice, the possibility cannot be excluded that complicated secondary effects would transactivate Hamp expression and lower the effect of reduced expression of Tfr2 in Hnf4a ⌬H mice. Despite the fact that Tfr2 is a direct target for HNF4␣, remarkable changes of gene expression override the potential effects and phenotypes caused by decreased expression of Tfr2 in Hnf4a ⌬H mice. Consequently, HNF4␣ is expected to transactivate Hamp through modulation of Tfr2 expression in the normal liver.
In conclusion, hepatic HNF4␣ significantly contributes to iron homeostasis and directly regulates Tfr2 expression. Thus, augmentation of hepatic HNF4␣ signaling could be of value for the treatment of hypoferremia and anemia.   6 and 7). B, chromatin immunoprecipitation using HepG2 cells were performed with anti-HNF4␣ antibody and normal goat IgG. The regions with the proximal HNF4␣-binding sites (ϩHNF4␣) and the regions without HNF4␣-binding site (ϪHNF4␣) in the human MIR-194/192 gene were amplified. The data from real time PCR were normalized relative to the input and expressed as fold-enrichment. ***, p Ͻ 0.001. C, chromatin immunoprecipitation using the livers of Hnf4a ⌬H and Hnf4a f/f mice with anti-HNF4␣ antibody and normal goat IgG. The regions between the proximal and the distal HNF4␣-binding sites in the mouse Tfr2 promoter and the regions without HNF4␣-binding site in the mouse Hmgcs2 gene were amplified by real time PCR. The data were normalized relative to the input and expressed as fold-enrichment over data from the IgG control. ***, p Ͻ 0.001. D, 10 nM siRNA for HNF4␣ (siHNF4␣) or negative control (siControl) was transfected into HepG2 cells. Real time PCR for HNF4a (white bar) and TFR2 mRNA (black bar) from total RNA of siRNA-transfected cells (n ϭ 3) is shown. Data are mean Ϯ S.D. **, p Ͻ 0.01; ***, p Ͻ 0.001 compared with siControl-treated cells.