Heme and non-heme iron transporters in non-polarized and polarized cells
© Yanatori et al; licensee BioMed Central Ltd. 2010
Received: 26 March 2010
Accepted: 4 June 2010
Published: 4 June 2010
Skip to main content
© Yanatori et al; licensee BioMed Central Ltd. 2010
Received: 26 March 2010
Accepted: 4 June 2010
Published: 4 June 2010
Heme and non-heme iron from diet, and recycled iron from hemoglobin are important products of the synthesis of iron-containing molecules. In excess, iron is potentially toxic because it can produce reactive oxygen species through the Fenton reaction. Humans can absorb, transport, store, and recycle iron without an excretory system to remove excess iron. Two candidate heme transporters and two iron transporters have been reported thus far. Heme incorporated into cells is degraded by heme oxygenases (HOs), and the iron product is reutilized by the body. To specify the processes of heme uptake and degradation, and the reutilization of iron, we determined the subcellular localizations of these transporters and HOs.
In this study, we analyzed the subcellular localizations of 2 isoenzymes of HOs, 4 isoforms of divalent metal transporter 1 (DMT1), and 2 candidate heme transporters--heme carrier protein 1 (HCP1) and heme responsive gene-1 (HRG-1)--in non-polarized and polarized cells. In non-polarized cells, HCP1, HRG-1, and DMT1A-I are located in the plasma membrane. In polarized cells, they show distinct localizations: HCP1 and DMT1A-I are located in the apical membrane, whereas HRG-1 is located in the basolateral membrane and lysosome. 16Leu at DMT1A-I N-terminal cytosolic domain was found to be crucial for plasma membrane localization. HOs are located in smooth endoplasmic reticulum and colocalize with NADPH-cytochrome P450 reductase.
HCP1 and DMT1A-I are localized to the apical membrane, and HRG-1 to the basolateral membrane and lysosome. These findings suggest that HCP1 and DMT1A-I have functions in the uptake of dietary heme and non-heme iron. HRG-1 can transport endocytosed heme from the lysosome into the cytosol. These localization studies support a model in which cytosolic heme can be degraded by HOs, and the resulting iron is exported into tissue fluids via the iron transporter ferroportin 1, which is expressed in the basolateral membrane in enterocytes or in the plasma membrane in macrophages. The liberated iron is transported by transferrin and reutilized for hemoglobin synthesis in the erythroid system.
Iron has an essential function in mammalian metabolism because of the ease with which it can gain and lose electrons. It is involved in biological functions as a metal cofactor for many proteins and enzymes that are used in oxygen transport (hemoglobin and myoglobin), electron transfer (mitochondrial cytochrome), and DNA synthesis (ribonucleotide reductase). Thus, iron is indispensable for eukaryotes and prokaryotes; however, it is also potentially toxic because of the generation of the superoxide anion and hydroxyl radical. These oxygen metabolites readily react with biological molecules, including proteins, lipids, and DNA. Iron overload diseases owing to genetic misregulation of iron uptake are referred to as primary iron overload disease or hereditary hemochromatosis . On the other hand, an acquired anemia that is associated with iron deficiency is referred to as anemia of inflammation or anemia of chronic disease . Organisms have a system to maintain normal iron homeostasis; iron deficiency and overload are associated with cellular dysfunction. Therefore, all mammalian species tightly regulate the iron concentration in body fluids. Because humans lack a regulated pathway for iron excretion, regulation of iron absorption from the intestine and the recycling of iron from senescent red blood cells (RBCs) are crucial in maintaining iron balance. Normal iron loss in humans occurs through exfoliation of enterocytes and skin cells, and through menstruation and childbirth. The absorption of dietary iron, composed mostly of heme and non-heme iron, occurs predominantly in the duodenum and upper jejunum, and is highly regulated ; this involves transport of absorptive enterocytes across the apical membrane into the cytosol and across the basolateral membrane into body fluids. Divalent metal transporter 1 (DMT1) is the only known intestinal iron importer and is a member of the natural resistance-associated macrophage protein family [4–7]. DMT1 is highly conserved from prokaryotes to eukaryotes, expressed in the apical membrane of absorptive enterocytes in the small intestine, and is also present in the endosomes of all human cells . Proper endosomal recycling of DMT1 is important for efficient uptake of iron and depends on a retromer-mediated sorting mechanism . Iron imported by DMT1 enters into the cytosol of the absorptive cells where it can be stored in the cytosolic iron-storage molecule ferritin or exported into body fluids through the basolateral iron exporter ferroportin 1 (FPN1) [10–12]. FPN1, the only known cellular iron exporter, is found on all cell types, including the duodenal mucosa, macrophage, and placenta. The bactericidal peptide hepcidin functions as an iron regulatory hormone [13, 14]. The hepcidin gene encodes an 84-amino-acid pre-pro-peptide that is cleaved to form a bioactive 25-amino-acid peptide found in the plasma and urine. Hepcidin is synthesized in the liver, and its gene expression is increased by iron overload and inflammation, especially interleukin 6 and interleukin 1 , and decreased by hypoxia and anemia . Hepcidin induces irreversible internalization of FPN1 through lysosomal degradation, which results in a depletion of plasma iron and an accumulation of iron in duodenal enterocytes and macrophages .
Humans are able to utilize 2 types of iron, heme and non-heme. Heme is an important nutritional source of iron and is believed to be more readily absorbed than non-heme iron. Heme is a ubiquitous molecule with an active iron center carrying a high affinity to oxygen, which allows for reversible binding and transport of oxygen in hemoglobin. Heme groups serve as the catalytic site; they tightly bind to a variety of proteins involved in aerobic metabolism, including respiratory chain cytochromes and numerous cytochrome P450 isoenzymes. Heme is mostly absorbed in the proximal half of the duodenum, the absorptive capacity of which is decreased in the distal position of the small intestine . In macrophage, senescent RBCs are phagocytosed and digested into heme in the lysosome. Heme degradation is catalyzed by heme oxygenases (HOs), the activities of which are particularly high in the spleen, testes, brain, and liver . At present, cDNAs encoding 2 isoenzymes, HO-1  and HO-2 , have been cloned. Although HO-1 and HO-2 catalyze the same reaction and have similar cofactor requirements (NADPH-cytochrome P450 reductase and O2) , they substantially differ in regulation and expression patterns. HO-1 and HO-2 proteins differ in molecular weight. HO-1 is an inducible isoenzyme, while HO-2 is constitutive. HO-1 has been identified as the major 32-kDa heat shock protein hsp32  and is highly sensitive to various stimuli, including oxidative stress, heavy metals, UV radiation, and inflammation. Several reports investigated HO localization to various subcellular compartments, including endoplasmic reticulum (ER) , nucleus , mitochondria , or caveola . HO-1 has also been reported to change its location under hemin treatment [23, 26].
Because the catalytic sites of HOs are supposed to be in the cytosol, heme needs appropriate transporters for its import into the cytosol through the plasma or endosomal membrane. Two heme transporters have thus far been reported--heme carrier protein 1 (HCP1)  and heme responsive gene-1 (HRG-1) . HCP1 is highly conserved and is a member of a large family of proton-coupled transporters known as the major facilitator superfamily. Within this family, HCP1 well resembles a bacterial protein that transports the antibiotic tetracycline . Notably, there are structural similarities between the planar heme ring and tetracycline-metal structures that must be transported across the apical membrane of absorptive enterocytes. Moreover, HCP1 has recently been identified as a transporter that mediates the translocation of folate across the plasma membrane and is suggested to be the possible molecular entity of the carrier-mediated intestinal folate transport system . HRG-1 and HRG-4 have been reported to be essential in heme homeostasis and heme sensing in Caenorhabditis elegans, and HRG-1 knockdown leads to profound defects in erythropoiesis in zebrafish . HCP1 was found to be expressed in the duodenum and small intestine , and HRG-1 in the brain, heart, kidney, and small intestine . The expression of HCP1 and HRG-1 was also investigated in some cultured cell lines and detected in macrophage and epithelial cell lines. It is not clear yet which transporter predominantly functions as the heme transporter in enterocytes or macrophages, both of which are main iron-regulatory cells in the human body.
To understand heme catabolism in humans, it is important to analyze the relationship among heme transporters, HOs, and iron transporters. In the current study, we investigate the expression and subcellular localization of HOs, DMT1, HRG-1, and HCP1 in non-polarized and polarized epithelial cells. Our results suggest that HCP1 functions on the apical membrane of enterocytes, HRG-1 transports heme from the inside of the lysosome into the macrophages, and HOs on smooth ER catalyze the degradation of heme in the cytosol.
In this study, we showed the precise subcellular localizations of HOs. Previous reports on HO localizations indicated that HOs are located in the ER , nucleus , mitochondria , or caveola . To understand heme catabolism and iron recycling in cells, it is important to determine the localizations of HOs and other molecules related to heme and iron metabolism. Heme is a prosthetic group that consists of a protoporphyrin ring and an iron atom. Because the cell membrane is not freely permeable to heme, it is necessary to allocate certain heme transporters at appropriate locations and orientations in the cells so HOs can adequately access their substrates. We constructed recombinant HOs with 2 different types of tagging molecules on their N or C termini to examine their exact localizations in cells. Before comparing their localizations with appropriate marker molecules, it is necessary to make sure that the addition of tagging molecules has no effect on HO localizations. Our results show that HO-1 and HO-2 clearly colocalized with each other without any influence from the tagging molecules.
Both HO-1 and HO-2 did not colocalize with PDI (Figure. 1-A, d, e, f), which is mainly located in the rough ER, but partly colocalized with syntaxin 17 and calnexin (Figure. 1A, g-l), both of which are located in smooth and rough ER [33, 34]. NADPH-cytochrome P450 reductase, which supplies an electron to HOs and is reported to be located in smooth ER , showed a clear colocalization with HOs (Figure. 1B, a, b, c). Subcellular fractionation study confirmed the specific location of HOs in smooth ER, as obtained by immunofluorescence analyses (Figure. 1B). HO localizations are discussed in previous reports, and this study shows that HO-1 does not change its location even under hemin treatment.
Humans incorporate two-thirds of the total absorbed iron as heme in enterocytes and recycles iron from senescent RBCs in macrophages. Two candidate molecules are thus far reported as heme transporters [27, 28]: HCP1, which is believed to transport heme  or folate  through the plasma membrane, and HRG-1, which was identified in C. elegans by genome-wide microarrays as a heme-regulated gene. HRG-1 is also proven to be an essential molecule for erythropoiesis and development in zebrafish, and has a heme-uptake activity in human cultured cells and Xenopus laevis oocytes . To investigate on which organelles these 2 molecules function, we analyzed the localizations of HRG-1 and HCP1 in non-polarized epithelial cells. Both HRG-1 and HCP1 are cotransfected in MDCK cells, and we observed these 2 molecules to be localized to the plasma membrane and lysosome. HRG-1 is located almost equally in the plasma membrane and lysosome, whereas HCP1 is located mostly in the plasma membrane and slightly in the lysosome. A previous report on HRG-1 localization showed that HRG-1 is distributed in an intracellular compartment punctuated throughout the cytoplasm, with about 10% of total HRG-1 on the cell periphery , and another reported that HRG-1 is localized to the endosome and plasma membrane . In addition, the location of HRG-1 was changed under serum-starvation conditions . In our study, HRG-1 is localized to both the lysosome and plasma membrane, and we did not observe HRG-1 translocation under serum-starvation conditions. These differences possibly arise from the differences in the cell lines or in the expression constructs used. Both HRG-1 and DMT1 are localized to lysosome, and a more detailed analysis using endogenous HRG-1 and DMT1 will be needed in future works. DMT1, which functions as a non-heme iron transporter, has 4 isoforms . Alternative splicing of the DMT1 gene produces 2 distinct classes of DMT1 transcripts, which differ in the C-terminal amino acids and subsequent 3'-untranslated regions. One form, IRE (I) (Figure. 4A), has an iron responsive element (IRE) by which intracellular iron concentration can affect its translation; the other form, non-IRE (II), does not have IRE on its mRNA. Alternative use of DMT1 gene promoters generates 2 variant DMT1 transcripts that differ in nucleotide sequences encoding the 5'-untranslated region and their subsequent N-terminal amino acids. A polypeptide transcribed from the 5'-upstream promoter and exon 1A is indicated as 1A, and a polypeptide transcribed from another promoter and exon 1B is indicated as 1B in Figure. 4A. These 4 DMT1 isoforms showed distinct subcellular localizations. DMT1A-I is mainly located in the plasma membrane, DMT1B-I in late endosome and lysosome, and both DMT1A-II and DMT1B-II in the recycling endosome. The protein expression levels of these 4 isoforms differ among tissue types; DMT1A-I is expressed in the duodenum and kidney and absorbs non-heme iron into the cytosol, whereas DMT1B-I is expressed in the macrophage and transports non-heme iron from the lysosome into the cytosol. DMT1A-II is expressed in the duodenum; its expression level is considerably lower compared with other isoforms. DMT1B-II is expressed in peripheral tissues and transports iron released from transferrin in the recycling endosome . Notably, we investigated the crucial signal for DMT1A-I localization to the plasma membrane. DMT1A-I L16A mutant localizes to the late endosome and lysosome. We reported that the non-IRE (II) C-terminal cytosolic region conducts the proper endosomal recycling of DMT1A and DMT1B . As a next step, analysis of the detailed function of N-terminal cytosolic region in the sorting mechanism of DMT1A is needed.
We compared the localization of HRG-1 with that of HCP1 in epithelial cells: HRG-1 colocalized with HCP1 in the plasma membrane in non-polarized epithelial cells. On the other hand, HRG-1 and HCP1 show different localizations in polarized epithelial cells; HCP1 is located in the apical membrane and HRG-1 is in the basolateral membrane and lysosome. HRG-1 can be detected very slightly in apical membrane in this system, and we will examine in the next stage whether endogenous HRG-1 is localized in apical membrane and can function to uptake heme from diet. HCP1 is able to transport folate more efficiently than heme [27, 29, 41]. HCP1 has a higher affinity for folate (Km = 1.67 μM) than heme (Km = 125 μM), and thus folate may be the more physiologically relevant target of HCP1. However, our localization study indicates that HRG-1 cannot function as a heme transporter to absorb heme from diet because of its location in epithelial cells, and that HCP1 may play a role in dietary heme uptake, because heme concentration in meat is roughly estimated to be 100 - 250 μM .
In this study, we analyze the localizations of HOs, HRG-1, HCP1, and DMT1 in non-polarized and polarized cells, and add new knowledge concerning heme transport and iron recycling system. Future work will be needed to further define the functions of HRG-1 and HCP1 in enterocytes and macrophages, the significance of HRG-1 localization to the basolateral membrane in enterocytes, the capability of HRG-1 to transport heme from body fluids or lysosome in macrophages, and the capability of HCP1 to transport heme from diet.
Mouse anti-human transferrin receptor (TfR) monoclonal antibody (mAb) (N-2) was prepared as described previously . Mouse anti-human EEA1 mAb, mouse anti-human GW130 mAb, mouse anti-human calnexin mAb, and mouse anti-human α-tubulin mAb were purchased from BD Transduction Laboratories (San Jose, CA). Mouse anti-HA mAb was purchased from Covance (Berkeley, CA), and mouse anti-human protein disulfide isomerase (PDI) mAb was purchased from Daiichi Fine Chemical (Toyama, Japan). Mouse anti-human LAMP2 mAb (H4B4, developed by Drs. J.E.K. Hildreth and J.T. August) was obtained from the Developmental Studies Hybridoma Bank (Baltimore, MD). Alexa 594-labeled anti-rabbit IgG and anti-mouse IgG, Alexa 488-labeled anti-rabbit IgG and anti-mouse IgG, and MitoTracker Deep Red 633 were purchased from Invitrogen Corp. (Carlsbad, CA).
The amino-acid-coding regions of human HOs, NADPH-cytochrome P450 reductase, syntaxin 17, HCP1, and HRG-1 were amplified using HEp-2 or Caco-2 cell cDNAs as templates. The fragments containing the full-length ORF were ligated into pEGFP-C1, pEGFP-N1, and pIRES-HA vectors (Clontech, Palo Alto, CA). pRSET-B-mCherry vector was kindly provided by Dr. Roger Y. Tsien (University of California, San Diego, CA). The mutant forms of DMT1 were made by PCR mutagenesis using KOD plus DNA polymerase (Toyobo, Osaka, Japan). Nucleotide sequences of PCR-oriented constructs were confirmed by the dideoxynucleotide chain-termination method using an ABI 3100 automated DNA sequencer.
Human HEp-2 epithelial cells and Madin-Darby canine kidney (MDCK) cell line were maintained in high-glucose Dulbecco's minimal essential medium containing 10% fetal calf serum, 50 μg/ml penicillin, and 50 μg/ml streptomycin. FuGENE 6 transfection reagent (Roche Molecular Biochemicals, Mannheim, Germany) was used for the transfection of HEp-2 cells and MDCK cells, which was done according to the manufacturer's instructions. After transfection, cells were cultured for 48 h on glass coverslips. For the polarity studies and cell surface labeling assay, clonal MDCK cells were cultured as confluent monolayers on polycarbonate filter chamber (Transwell, Corning, NY) for 6 days.
Cells grown on glass coverslips and Transwell were fixed with 4% paraformaldehyde (PFA) in PBS for 15 min at room temperature, and permeabilized with 0.2% Triton X-100 in PBS for 20 min. The coverslips and Transwell were washed and blocked in 0.1% fish skin gelatin in PBS. Cells were incubated with primary antibodies for 60 min at room temperature. Coverslips and Transwell were washed with 0.1% fish skin gelatin in PBS. Secondary antibodies coupled to Alexa 488 or Alexa 594 were incubated on cells for 60 min at room temperature. MitoTracker Deep Red was used to stain the mitochondria. HEp-2 cells grown on coverslips were incubated in fresh medium with 250 nM MitoTracker Deep Red at 37°C for 45 min. Coverslips and Transwell were washed and mounted on slides with VECTASHIELD (Vector Laboratories, Burlingame, CA). The XY and XZ images were obtained by using a Leica TCS SP2 AOBS confocal laser scanning microscope system.
Cells grown on Transwell were washed with ice-cold PBS containing 0.1 mM CaCl2 and 1 mM MgCl2 [PBS(+)]. For selective labeling of the apical or the basolateral surface, sulfo-NHS-LC-biotin was added either to the apical or the basolateral compartment of the filter chamber. The compartment not receiving sulfo-NHS-LC-biotin was filled with an equivalent volume of PBS(+). Three filter chambers were used per experimental condition. Cells were washed with Tris-buffered saline with mild agitation. Then, cells were extracted in RIPA buffer [150 mM NaCl, 50 mM Tris (pH 8.0), 5 mM EDTA, 1% Nonidet P-40, 0.5% deoxycholate, and 0.1% SDS] and the extracts were clarified by centrifugation at 14,000 × g for 15 min. The resulting supernatants were added to anti-green fluorescent protein (GFP) or anti-mCherry antibody conjugated protein A beads. After 2-h incubation, the beads were washed and the proteins were eluted with Laemmli buffer. Eluants were analyzed by immunoblotting. Biotinylated proteins were detected by ImmunoPure Avidin, Horseradish Peroxidase, Conjugated (Pierce Biotechnology, Rockford, IL).
HEp-2 cells stably expressing HA-tagged HO-2 were grown to confluency in a 100-cm2 dish. Cells were washed once with PBS and scraped in homogenizing buffer (0.25 M sucrose and 10 mM Tris-HCl, pH 7.4). Cells were homogenized in glass-Teflon Potter homogenizer rotating at 1,000 rpm, 25 strokes. The homogenate was centrifuged at 1,000 × g for 10 min to obtain post nuclear supernatant (PNS). PNS was centrifuged at 10,000 × g for 20 min to obtain the post-mitochondrial fraction (PMF). Then, to obtain smooth and rough microsomal fractions, PMF was loaded onto 3.5 ml of 1.3 M sucrose. Samples were centrifuged at 100,000 × g for 2 h at 4°C. Aggregated rough microsomes sedimented through the 1.3 M sucrose layer and pelleted at the bottom of the tubes, whereas smooth membranes collected on top of the 1.3 M sucrose layer .
HEp-2 cells were cultured with 100 μM hemin for 36 h. After incubation, cells were washed in cold PBS. Proteins were collected in tube and solubilized in 2% SDS-Laemmli buffer; protein contents were determined by the method of Lowry. The samples were resolved by SDS-PAGE using 12% acrylamide gels and blotted onto polyvinylidene difluoride membranes. HO-1 was detected by rabbit anti-HO-1 polyclonal antibody (Stressgen, Victoria, BC, Canada).
The recombinant constructs used in this study are indicated in Additional file 1 Figure. S1A.
divalent metal transporter 1
green fluorescent protein
human influenza hemagglutinin
heme carrier protein 1
heme responsive gene-1
phosphate buffered saline
protein disulfide isomerase
This work was supported in part by Grant-in-Aid for scientific research from the Ministry of Education, Science and Culture of Japan, and Research Project Grants from Kawasaki Medical School.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.