Hereditary Hemochromatosis Protein, HFE, Interaction with Transferrin Receptor 2 Suggests a Molecular Mechanism for Mammalian Iron Sensing

Abstract

HFE and transferrin receptor 2 (TFR2) are membrane proteins integral to mammalian iron homeostasis and associated with human hereditary hemochromatosis. Here we demonstrate that HFE and TFR2 interact in cells, that this interaction is not abrogated by disease-associated mutations of HFE and TFR2, and that TFR2 competes with TFR1 for binding to HFE. We propose a new model for the mechanism of iron status sensing that results in the regulation of iron homeostasis. HFE and transferrin receptor 2 (TFR2) are membrane proteins integral to mammalian iron homeostasis and associated with human hereditary hemochromatosis. Here we demonstrate that HFE and TFR2 interact in cells, that this interaction is not abrogated by disease-associated mutations of HFE and TFR2, and that TFR2 competes with TFR1 for binding to HFE. We propose a new model for the mechanism of iron status sensing that results in the regulation of iron homeostasis. All mammalian cells have an absolute requirement for iron. Both cellular iron deficiency and iron-overload are pathological and iron concentration in cells and body fluids is tightly regulated. Chronic malaccumulation of iron in tissues results in systemic iron-overload diseases that are collectively called hemochromatosis. Hereditary hemochromatosis in humans is linked to mutations in several genes namely HFE, transferrin receptor 2(TFR2), hemojuvelin (HJV), and hepcidin (HAMP) (reviewed in Ref. 1Hentze M.W. Muckenthaler M.U. Andrews N.C. Cell. 2004; 117: 285-297Abstract Full Text Full Text PDF PubMed Scopus (1401) Google Scholar). Hepcidin, a hepatocyte-derived soluble factor, negatively regulates intestinal absorption of dietary iron and the release of recycled iron into circulation from macrophages (2Nemeth E. Tuttle M.S. Powelson J. Vaughn M.B. Donovan A. Ward D.M. Ganz T. Kaplan J. Science. 2006; 306: 2090-2093Crossref Scopus (3637) Google Scholar). Also, hepcidin expression responds to body iron status (3Nicolas G. Chauvet C. Viatte L. Danan J.L. Bigard X. Devaux I. Beaumont C. Kahn A. Vaulont S. J. Clin. Investig. 2002; 110: 1037-1044Crossref PubMed Scopus (1351) Google Scholar). Hepcidin, therefore, plays a pivotal role as a regulator of whole-body iron homeostasis. Since disruption of HFE, TFR2, or HJV causes decreased hepcidin production these gene products appear to be involved in the upstream regulation of hepcidin (reviewed in Ref. 4Le Gac G. Férec C. Eur. J. Hum. Genet. 2005; 13: 1172-1185Crossref PubMed Scopus (56) Google Scholar). HFE, an atypical major histocompatibility complex class I molecule, associates with transferrin receptor 1 (TFR1), 3The abbreviations used are: TFR, transferrin receptor; FBS, fetal bovine serum; CHO, Chinese hamster ovary; HEK, human embryonic kidney; TBS, Tris-buffered saline; Fe2-TF, diferric-transferrin. a type II transmembrane glycoprotein that is the primary effector of cellular iron uptake (5Bennett M.J. Lebrón J.A. Bjorkman P.J. Nature. 2000; 403: 46-53Crossref PubMed Scopus (296) Google Scholar). TFR2 is a homolog of TFR1 and, like TFR1, can bind and internalize diferric-transferrin (Fe2-TF) (6Kawabata H. Yang R. Hirama T. Vuong P.T. Kawano S. Gombart A.F. Koeffler H.P. J. Biol. Chem. 1999; 274: 20826-20832Abstract Full Text Full Text PDF PubMed Scopus (567) Google Scholar). However, while TFR1 is widely expressed, TFR2 is expressed predominantly in hepatocytes, hematopoietic cells, and duodenal crypt cells, overlapping with HFE expression (6Kawabata H. Yang R. Hirama T. Vuong P.T. Kawano S. Gombart A.F. Koeffler H.P. J. Biol. Chem. 1999; 274: 20826-20832Abstract Full Text Full Text PDF PubMed Scopus (567) Google Scholar, 7Griffiths W.J.H. Cox T.M. J. Histochem. Cytochem. 2003; 51: 613-623Crossref PubMed Scopus (52) Google Scholar). This and other differences in transcriptional regulation, Fe2-TF binding affinities and gene deletion phenotypes suggest that TFR1 and TFR2 have distinct roles in iron homeostasis. While TFR1 is a key mediator of iron uptake, TFR2 is postulated to play a regulatory role in whole-body iron homeostasis. Since hepcidin is produced predominantly by hepatocytes, it is likely that these cells express molecular determinants of iron sensing. Serum transferrin deficiency in hpx mice is associated with low hepcidin levels despite parenchymal iron loading, a phenotype corrected by transferrin administration (8Weinstein D.A. Roy C.N. Fleming M.D. Loda M.F. Wolfsdorf J.I. Andrews N.C. Blood. 2002; 100: 3776-3781Crossref PubMed Scopus (544) Google Scholar). Also, elevated serum Fe2-TF stabilizes TFR2 protein in liver (9Robb A. Wessling-Resnick M. Blood. 2004; 104: 4294-4299Crossref PubMed Scopus (169) Google Scholar), an effect recapitulated in vitro (9Robb A. Wessling-Resnick M. Blood. 2004; 104: 4294-4299Crossref PubMed Scopus (169) Google Scholar, 10Johnson M.B. Enns C.A. Blood. 2004; 104: 4287-4291Crossref PubMed Scopus (193) Google Scholar). Circulating Fe2-TF is therefore likely to be an iron signal sensed by hepatocyte membrane proteins that regulate hepcidin production and TFR2 may be part of the regulatory system sensing transferrin saturation. Previous investigators hypothesized that HFE and TFR2 might belong together in such a regulatory pathway. An earlier study, however, demonstrated that soluble, purified ectodomains of HFE and TFR2 do not interact in vitro (11West A.P. Bennett M.J. Sellers V.M. Andrews N.C. Enns C.A. Bjorkman P.J. J. Biol. Chem. 2000; 275: 38135-38138Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar), precluding the possibility that their roles in a convergent iron homeostasis pathway involve formation of an HFE-TFR2 complex. Here we investigate whether HFE and TFR2 interact when expressed in cells. Based on our findings, we propose that HFE plays an important role in iron sensing by conveying the whole-body iron status, reflected by transferrin saturation, from the HFE-TFR1 complex to TFR2, resulting in potential downstream signaling events. Cell Culture and Transfection—All cell culture media were supplemented with 100 units/ml penicillin, 100 μg/ml streptomycin, and 10 mm glutamine (Invitrogen) unless otherwise noted. AML12, a differentiated, non-transformed mouse hepatocyte cell line (12Wu J.C. Merlino G. Fausto N. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 674-678Crossref PubMed Scopus (250) Google Scholar) was maintained in 1:1 Dulbecco's modified Eagle's medium/Ham's nutrient F-12 mixture (F-12) supplemented with 10% fetal bovine serum (FBS, ATCC), 5 μg/ml human insulin (Sigma), 5 μg/ml human transferrin (Roche Applied Science), 5 ng/ml sodium selenite (Sigma), 40 ng/ml dexamethasone (Sigma), and 5 mm sodium pyruvate. CHO-TRVb-0 (TFR1-deficient Chinese hamster ovary cells) and TRVb-1 (human TFR1 stably transfected TRVb-0) cells (13McGraw T.E. Greenfield L. Maxfield F.R. J. Cell Biol. 1987; 105: 207-214Crossref PubMed Scopus (163) Google Scholar) were cultured in Ham's F-12 supplemented with 5% FBS and 0.2% glucose. Human embryonic kidney 293T (HEK293T) cells were grown in Dulbecco's modified Eagle's medium containing 10% FBS. Total amounts of DNA transfected in each experiment were kept equal in all samples by adding appropriate vector DNA. AML12, TRVb-0, and TRVb-1 cells were transfected using Lipofectamine 2000 (Invitrogen) in antibiotic-free growth medium using methods described by the manufacturer. HEK293T cells were transfected with DNA:calcium phosphate co-precipitates using a HEPES-buffered calcium phosphate method (14Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (eds) Current Protocols in Molecular Biology, pp. 9.1.1–9.1.11 and 12.1.1–12.1.6, Wiley-Interscience, New YorkGoogle Scholar). Transfection medium was replaced with fresh culture medium 12 h post-transfection and cells were harvested and lysed 48 h post-transfection. Expression Plasmids and Mutagenesis—1 μg of mouse liver RNA was used in reverse transcription reactions using Superscript II (Invitrogen) following protocols described by the manufacturer. Specific oligonucleotide primers were used to amplify mouse HFE, TFR1, and TFR2 cDNAs. Myc epitope-tagged TFR1 and TFR2 cDNAs were constructed by replacing the respective stop codons with an XhoI site and cloning into pcDNA3.1 myc-His plasmid (Invitrogen). HFE cDNA was first cloned into pcDNA3.1 (Invitrogen). This parent HFE cDNA was mutagenized using the QuikChange mutagenesis kit (Stratagene) to insert sequence coding for a FLAG epitope downstream of the HFE signal peptide. Mutagenesis of the mouse cDNAs to produce HFE and TFR2 mutants (HFEH67D, TFR2M167K, TFR2Y245X, and TFR2K685P) corresponding to human hemochromatosis associated HFE and TFR2 mutations (HFEH63D, TFR2M172K, TFR2Y250X, and TFR2Q690P, respectively) were also performed using the QuikChange mutagenesis protocol. The TFR2Y245X mutant was generated by deletion of the sequences coding for amino acids 246–798 from the parent TFR2-myc construct. The resulting construct thus encoded amino acids 1–245 of mouse TFR2 followed directly by an inframe myc epitope tag. The conserved RGD motif in mouse TFR2 (spanning amino acids 673–675) was mutated to GGD. All plasmids were subjected to sequence analysis to confirm integrity. The T91-Display construct (encoding FSTL3-myc in pDisplay, Invitrogen) was a kind gift from Dr. Y. Sidis (Massachusetts General Hospital, Harvard Medical School) and is described in Ref. 15Sidis Y. Mukherjee A. Keutmann H. Delbaere A. Sadatsuki M. Schneyer A. Endocrinology. 2006; 147: 3586-3597Crossref PubMed Scopus (171) Google Scholar. Antibodies—Primary antibodies, anti-TFR1 monoclonal antibody (Zymed Laboratories Inc.), anti-TFR2 rabbit polyclonal antibody (Alpha Diagnostics), anti-FLAG M2 monoclonal antibody and agarose-conjuguated M2 antibody (Sigma), and anti-myc monoclonal antibody (Upstate Biotechnology) were purchased. Protein G-conjugated agarose beads and protein A-conjugated Sepharose beads were purchased from Roche Applied Science and Sigma, respectively. Horseradish peroxidase conjugated anti-mouse IgG and anti-rabbit IgG secondary antibodies were purchased from Amersham Biosciences. Mouse TrueBlot and rabbit TrueBlot secondary antibodies, which do not recognize denatured IgG, were purchased from eBiosciences. Preparation of Cell Lysates—Cells were washed once in ice-cold Dulbecco's phosphate-buffered saline and harvested. Cell pellets were lysed in ice-cold Triton X-100 lysis buffer (1% Triton X-100 (v/v), 50 mm Tris-HCl, pH 8.0, 10 mm KCl, 0.15 m NaCl, 20 mm NaF, 10 mm Na2P2O7, 1 mm Na3VO4) supplemented with a mixture of protease inhibitors (Complete Mini, Roche Applied Science). After mixing on ice for 15 min, cell lysates were centrifuged at 8000 × g for 5 min in a cooled table-top microcentrifuge to sediment nuclei and debris, and supernatants were collected and used in subsequent experiments. Crude membrane preparations were isolated from cell pellets using a hypotonic lysis method as described before (14Ausubel, F. M., Brent, R., Kingston, R. E., Moore, D. D., Seidman, J. G., Smith, J. A. & Struhl, K. (eds) Current Protocols in Molecular Biology, pp. 9.1.1–9.1.11 and 12.1.1–12.1.6, Wiley-Interscience, New YorkGoogle Scholar). Prior to experiments crude membrane preparations were resuspended in 1% Triton X-100 lysis buffer. Immunoprecipitation, Gel Electrophoresis, and Western Blot Analysis—Total protein amounts in lysates were measured using the Bio-Rad protein assay and equivalent amounts of protein were used for immunoprecipitation. Immunoprecipitations were performed by first incubating cell lysates with the indicated primary antibodies overnight at 4 °C. Immune complexes were then precipitated by incubating reactions for an additional 2 h at 4°C with either Sepharose-conjugated protein A or agarose-conjugated protein G as appropriate. For FLAG immunoprecipitations, agarose-conjugated anti-FLAG M2 antibody was used and additional protein G incubations omitted. Immunoprecipitates were washed four times with 1% Triton X-100 lysis buffer and were resuspended in SDS-PAGE sample buffer for Western blotting. Lysates and immunoprecipitates were separated by SDS-PAGE using 10% polyacrylamide gels, and the proteins were transferred to pure nitrocellulose filters (Amersham Biosciences). The filters were stained with Ponceau S to confirm transfer and blocked with either 3% nonfat dry milk or 3% bovine serum albumin in Tris-buffered saline (TBS) containing 0.05% Tween 20 for 1 h at room temperature. Primary antibody incubations were performed overnight at 4 °C. Filters were washed three times with TBS containing 0.05% Tween 20 (TBST) and then incubated with appropriate secondary antibodies. The filters were again washed four times, and immune complexes formed on the blot were visualized by ECL (Amersham Biosciences). Immunoreactive protein bands were analyzed using the public domain NIH Image program (developed at the United States National Institutes of Health and available on the Internet at rsb.info.nih.gov/nih-image/). First, we transfected HEK293T cells with N-terminal FLAG epitope-tagged HFE (FLAG-HFE) cDNA together with C-terminal myc epitope-tagged TFR2 (TFR2-myc) cDNA. In both cases epitope tags were in the extracellular domains. Lysates from transfected cells were used in immunoprecipitation analyses with anti-FLAG or anti-myc antibodies (Fig. 1A, B, and E). TFR2-myc was co-immunoprecipitated by anti-FLAG antibody when co-expressed with FLAG-HFE. Singly transfected TFR2-myc was not immunoprecipitated by the anti-FLAG antibody. In reciprocal anti-myc immunoprecipitations FLAG-HFE was co-immunoprecipitated when co-transfected with TFR2-myc. Hence, we demonstrate that HFE and TFR2 interact when expressed in cells. In similar experiments, when coexpressed with FLAG-HFE, TFR1-myc co-immunoprecipitated with FLAG-HFE while an unrelated cell surface expressed C-terminal myc-tagged protein, T91-Display, did not (Fig. 1, A and E). Hence, HFE interacts specifically with TFR1 and TFR2. To extend our findings in a relevant cell-type, we assessed HFE-TFR2 interaction in crude membrane fractions of transfected AML12 mouse hepatocytes and confirmed TFR2-myc interaction with FLAG-HFE in these cells (Fig. 1C). To eliminate the possibility that HFE-TFR2 interaction involves TFR1 (by TFR1-TFR2 heterodimerization) we examined HFE-TFR2 interaction in CHO-TRVb-0 cells, which lack TFR1 expression. Again, TFR2-myc was immunoprecipitated by anti-FLAG antibody in lysates from cells expressing both TFR2-myc and FLAG-HFE (Fig. 1D). To test whether addition or chelation of non-transferrin-bound iron influences the interaction we repeated the co-immunoprecipitation analyses after treating transfected HEK293T cells with or without iron ammonium citrate or desferrioxamine. We found that neither treatment affected binding of FLAG-HFE to TFR1-myc or TFR2-myc (Fig. 1E). These findings contrast the lack of interaction previously reported between purified soluble ectodomains of HFE and TFR2 (11West A.P. Bennett M.J. Sellers V.M. Andrews N.C. Enns C.A. Bjorkman P.J. J. Biol. Chem. 2000; 275: 38135-38138Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar). It is possible that truncated HFE and TFR2 have altered binding characteristics compared with the native proteins. Also, expression in cell membranes might promote structural conformations or provide other molecular components that are required for, or stabilize, HFE-TFR2 interaction, while the ectodomain structures alone are sufficient for HFE-TFR1 interaction. Additionally, although residues in TFR1 essential for HFE-TFR1 interaction are not conserved in TFR2 (11West A.P. Bennett M.J. Sellers V.M. Andrews N.C. Enns C.A. Bjorkman P.J. J. Biol. Chem. 2000; 275: 38135-38138Abstract Full Text Full Text PDF PubMed Scopus (216) Google Scholar), it appears that they are not necessary for HFE-TFR2 interaction. Indeed, while mutation of a conserved RGD motif in TFR1 abrogates its interaction with both HFE and transferrin (16West A.P. Giannetti A.M. Herr A.B. Bennett M.J. Nangiana J.S. Pierce J.R. Weiner L.P. Snow P.M. Bjorkman P.J. J. Mol. Biol. 2001; 313: 385-397Crossref PubMed Scopus (114) Google Scholar), we demonstrate (Fig. 1A) that this mutation in TFR2 does not affect binding of TFR2 to HFE. To address whether hemochromatosis-associated mutations (HFEH63D, TFR2M172K, TFR2Y250X, and TFR2Q690P) affect HFE-TFR2 interaction we introduced the corresponding mutations (HFEH67D, TFR2M167K, TFR2Y245X, and TFR2K685P) into the mouse proteins. We show that wild type FLAG-HFE interacts with all three TFR2 mutants tested (Fig. 1A). Most notably the TFR2Y245X mutant that lacks 553 amino acids of the extracellular domain retains ability to interact with HFE suggesting that known disease-associated mutations downstream of Y250X (17Roetto A. Daraio F. Alberti F. Porporato P. Cali A. De Gobbi M. Camaschella C. Blood Cells Mol. Dis. 2002; 29: 465-470Crossref PubMed Scopus (47) Google Scholar) also do not influence HFE-TFR2 binding. Similarly, the HFEH67D mutation does not abrogate HFE-TFR2 binding (Fig. 1B). Since iron homeostasis is compromised by these mutations in humans, our observations imply that in disease states, HFE and TFR2 interact but subsequent signal transduction leading to hepcidin production is impaired. This in turn suggests that signal transduction initiated by the HFE-TFR2 interaction may require other molecular components that together with HFE and TFR2 form an iron sensor and signal transduction effector. Fe2-TF competes with HFE for binding to TFR1 (16West A.P. Giannetti A.M. Herr A.B. Bennett M.J. Nangiana J.S. Pierce J.R. Weiner L.P. Snow P.M. Bjorkman P.J. J. Mol. Biol. 2001; 313: 385-397Crossref PubMed Scopus (114) Google Scholar) displacing HFE from the HFE-TFR1 complex. Also, increased serum transferrin saturation (9Robb A. Wessling-Resnick M. Blood. 2004; 104: 4294-4299Crossref PubMed Scopus (169) Google Scholar) and treatment of cells with Fe2-TF (9Robb A. Wessling-Resnick M. Blood. 2004; 104: 4294-4299Crossref PubMed Scopus (169) Google Scholar, 10Johnson M.B. Enns C.A. Blood. 2004; 104: 4287-4291Crossref PubMed Scopus (193) Google Scholar) enhances TFR2 protein stability. It is possible that when serum Fe2-TF is elevated, increased cellular TFR2 competes with TFR1 for binding to HFE in hepatocytes. To test whether HFE-TFR1 interaction is competed by TFR2, CHO-TRVb-1 cells, stably overexpressing TFR1 but not expressing TFR2, were transfected with either FLAG-HFE alone or with added increasing amounts of TFR2-myc cDNA. Lysates were assayed for the co-immunoprecipitation of FLAG-HFE and TFR2-myc or endogenous TFR1. Increasing TFR2 expression the of TFR1 co-immunoprecipitated with HFE, as after to cellular TFR1 (Fig. A and The expression of T91-Display did not HFE-TFR1 Additionally, as treatment with Fe2-TF, of HFE-TFR1 interaction by co-immunoprecipitation of TFR1 with HFE. these results suggest that TFR2 competes with TFR1 for binding to HFE of Fe2-TF of HFE-TFR1 interaction was however, in the of both Fe2-TF and TFR2 expression. serum Fe2-TF TFR2 expression (9Robb A. Wessling-Resnick M. Blood. 2004; 104: 4294-4299Crossref PubMed Scopus (169) Google Scholar, 10Johnson M.B. Enns C.A. Blood. 2004; 104: 4287-4291Crossref PubMed Scopus (193) Google Scholar) with of TFR1 N. G. R. R. P. P. P. J. Full Text PDF PubMed Scopus Google Scholar) in hepatocytes. These the formation of an HFE-TFR2 complex an HFE-TFR1 complex. an iron status sensing role for HFE and the HFE-TFR2 interaction. HFE, TFR2, and HJV proteins appear to be upstream of hepcidin production (reviewed in Ref. 4Le Gac G. Férec C. Eur. J. Hum. Genet. 2005; 13: 1172-1185Crossref PubMed Scopus (56) Google Scholar) and Fe2-TF is likely the iron status signal (8Weinstein D.A. Roy C.N. Fleming M.D. Loda M.F. Wolfsdorf J.I. Andrews N.C. Blood. 2002; 100: 3776-3781Crossref PubMed Scopus (544) Google Scholar, D.M. Blood. 2006; PubMed Scopus Google Scholar). Since serum Fe2-TF Fleming M.D. Andrews N.C. Blood. 1999; PubMed Google Scholar) and TFR2 expression (9Robb A. Wessling-Resnick M. Blood. 2004; 104: 4294-4299Crossref PubMed Scopus (169) Google Scholar, 10Johnson M.B. Enns C.A. Blood. 2004; 104: 4287-4291Crossref PubMed Scopus (193) Google Scholar) are increased and and are in mice that show iron homeostasis Fleming M.D. Andrews N.C. Blood. 1999; PubMed Google Scholar), it appears that HFE is integral to the signal transduction mechanism that increased Fe2-TF and a cellular in of hepcidin expression. We therefore propose a new model for iron homeostasis (Fig. transferrin saturation, HFE with TFR1. serum Fe2-TF HFE is competed by Fe2-TF from TFR1, and thus HFE a molecular sensor for serum transferrin saturation. HFE the elevated Fe2-TF status by interaction with TFR2, leading to the formation of an iron sensor and signal transduction effector complex. signaling then to the production of hepcidin that to the regulation of whole-body iron homeostasis. In or human hemochromatosis and mouse this sensor and signal transduction effector complex not be and hepcidin production not be by Fe2-TF, of whole-body iron homeostasis and iron in parenchymal cells. In HFE and TFR2 interact in cells hepatocytes, and this interaction may form the of a systemic iron sensor in hepatocytes. We Dr. Y. Sidis for the cDNA and are to A. Mukherjee and for and