In mammalian cells, exposure to a low-oxygen environment triggers a hypoxic response pathway centered on the regulated expression of the hypoxia-inducible transcription factor (HIF). HIF-1 is a heterodimer composed of α and β subunits (1). Under normoxia, HIF-1α is destabilized by a mechanism involving prolyl hydroxylation and targeted for proteasomal degradation (2). In hypoxia, prolyl hydroxylase activity is reduced and the nonhydroxylated form of HIF-1α is stabilized. HIF-1α then integrates to HIF-1β subunit to bind hypoxia-responsive elements (HREs) in target genes. Then HIF-1 activates transcription of its target genes that allow for adaptation to hypoxia (3). To date, there are more than 100 HIF-1 downstream genes identified with varying functions. Moreover, by using DNA microarrays, it has recently been reported that more than 2% of all human genes are regulated by HIF-1 in arterial endothelial cells, directly or indirectly (4).
Iron is needed for several essential functions including cellular growth and survival. Iron is also potentially dangerous as a catalyst of reactive oxygen species (ROS) production, so it is toxic when present in excess (5). Cells have evolved a mechanism to maintain iron homeostasis via iron transporter proteins. The divalent metal transporter 1 (DMT1), also known as natural resistance–associated macrophage protein 2 (Nramp2), is a protein recently shown to play a pivotal role in iron uptake from both transferrin (Tf) and non-Tf sources in different anatomic sites (6, 7). Ferrous iron is transported across the endosomal membrane via DMT1 (8). There are two splice variants of DMT1 (8). One form contains an iron responsive element (IRE) in the 3′-untranslated region of the mRNA capable of binding iron response proteins (IRPs) resulting in the stabilization of the mRNA. Accordingly, this form of DMT1 may be similar to the transferrin receptor (TfR) in that it potentially can be regulated by iron. The second mRNA form, lacking the IRE (−IRE), is presumably incapable of being regulated by iron, at least by an IRE/iron response protein interaction (9). DMT1 gene also has been reported to contain two different promoters with two alternative exon1 (1A and 1B) (10).
Hypoxia inducible changes in the expression of different isoforms of DMT1 are already described in PC12 cells. Lis et al. concludes that expression of the 1A containing species of DMT1 is increased in hypoxic treatment (11). The studies by Mastrogiannaki et al. (12) and Shah et al. (13) demonstrate that DMT1 exon1A is specifically regulated by HIF-2 but not by HIF-1. Previous studies conducted by our work revealed a high correlation between the expression of HIF-1α and DMT1 proteins in HepG2 cells treated with chemical (CoCl2) or physical hypoxia what led us to speculate that DMT1 might also be one of the target genes of HIF-1 (14). Lee at al. suggested that there are two motifs (CCAAAGTGCTGGG) that are similar to HIF-1 binding sites (HBS) in the 5′ regulatory region of human DMT1 exon1B (between −412 and −570), because he thought that they were similar to the HBS of EPO (15). However, these two motifs (CCAAAGTGCTGGG) did not contain the core HRE sequence (-A/GCGTG-). Lee did not show any other evidence to demonstrate the sequences were HBS.
In this study, we searched for a potential element of the DMT1 exon1B promoter responsible for transcriptional induction under hypoxia. We identified a functional HRE at position −327 to −323 of the DMT1 exon1B promoter, which is necessary for stimulation of DMT1 gene by hypoxia. We also demonstrated that total cellular iron levels increased by hypoxic treatment and there was a remarkable increase of ferrous iron uptake simultaneously. Meanwhile, our studies showed that ROS level changed by transfected with DMT1 exon1B expression plasmid. These results verified our presumption that hypoxia might affect cell iron homeostasis through regulating the expression of DMT1 exon1B. Excess of iron uptake inducing by DMT1 exon1B overexpression might lead to intensify cell death through generated profuse free radical.
MATERIALS AND METHODS
Cell Culture and Hypoxic Induction
Human HepG2 hepatoma cells were cultured in Dulbecco's Modified Eagle's Medium (GIBCO, Grand Island, NY) supplemented with 10% fetal bovine serum (GIBCO, Grand Island, NY) at 37°C in 5% CO2 atmospheric air incubator (Froma Series II, Thermo). For hypoxic treatment, cell culture plates were incubated in at 5% CO2 level with 1% O2 balanced with N2 incubator (Model 7101FC-1, NAPCO, USA).
Computerized Search for Nuclear Factor-Binding Sites in the DMT1 Promoter
The DNA sequences from the human DMT1 exon1B promoter region were obtained from GeneBank (Accession #AF064475). Potential nuclear factor-binding sites were found using MatInspector software online.
The pIRESneo, pIRESneo-DMT1 isoform I (+IRE DMT1 exon1B), pIRESneo-DMT1 isoform I (−IRE DMT1 exon1B) were kind gifts from Dr. Mitsuaki Tabuchi (Kawasaki Medical School, Japan). Various lengths of DNA fragments were amplified from genomic DNA by polymerase chain reaction (PCR) using the primers given in Table 1. The PCR products were cloned into pGL3-Basic, pGL3-SV40 (Promega, Madiso, WI), and pcDNA3.1 (Invitrogen, Carlsbad, CA). The first construct (DMT1-prom) contained the promoter from −386 to +50 including three putative HREs. The second construct (DMT1-prom-HRE1) contains one putative HRE located from −386 to −216. The third construct (DMT1-prom-HRE2) contains two putative HREs located from −224 to −48. Site-directed mutagenesis of the putative HRE in DMT1 exon1B promoter (CAGTACCTAACGTGGCGCCA→CAGTACCTAAAATGGCG CCA). The construction containing the site-directed mutagenesis was referred to as DMT1-prom-HRE2mut. HIF-1α fragments were amplified with the Phusion High Fidelity PCR Kit (Finnzymes) and subcloned into the pcDNA3.1 expression vector. All the constructs were sequenced to confirm accuracy.
Table 1. Primers used for cloning of all constructs
Transient Transfection and Luciferase Assay
One day before transfection, HepG2 cells were plated into 24-well plates. The cells were grown to 90% confluence, and then plasmid constructs were cotransfected with an internal control vector pRL-TK (Promega, Madiso, WI) (100:1 ratio) to the cells by Lipofectamine 2000 (Invitrogen, Carlsbad, CA). After different treatments, the cells were subsequently harvested and then luciferase activity was quantitated (Synergy 2, BioTek, USA) using Dual-Luciferase Reporter System (Promega, Madiso, WI).
One day before transfection, HepG2 cells were plated into 6-well plates. The cells were grown to 50% confluence and then transfected with 100 nM of human DMT1-specific siRNA duplex (Dharmacon, Thermo fisher Scientific, USA), using X-tremeGENE siRNA transfection reagent (Roche, Mannheim, Germany).
Quantification of Iron Content by Atomic Absorption Spectroscopy
The cells were washed twice with phosphate-buffered saline and harvested. Aliquots of the lysate were heated in 100°C overnight and then used to quantify the total amount of iron by atomic absorption spectroscopy.
Calcein Loading of the Ferrous Uptake Assay
The cells were loaded with calcein-AM according to a method described previously (16). Briefly, the cells were washed twice with medium, and then incubated with 0.125 μM calcein-AM in serum-free medium for 10 min at 37°C. Excess calcein-AM on cell surface was removed by three washes with Hank's balanced salt solution (HBSS, pH 7.4). Before measurements, 100 μL of calcein-loaded cell suspension and 2 mL Hepes were added to a shade selection cuvette. The fluorescence was measured by Ultraviolet spectrophotometer (Shimadzu RF-5301PC, Japan) with λex = 485 nm and λem = 520 nm at 37°C. After initial baseline of fluorescence intensity was collected, ferrous ammonium sulfate (FAS; 40 μM, final concentration) was added to the cuvette. The quenching of calcein fluorescence was recorded in every 5 min for 30 min. The fluorescence descent degree reflects the ferrous uptake of cells. Data were normalized to the steady state (baseline) values of fluorescence.
ROS Level Detection
Cells were incubated in 5 μM hydroethidine (dihydroethidium; HE) or 25 μM 2,7-dichlorodihydrofluorescein (DCFH) for 30 min at 37°C. Then cells washed twice with phosphate-buffered saline. The fluorescence was measured with λex = 520 nm and λem = 610 nm for HE, λex = 498 nm and λem = 522 nm for DCFH (Synergy 2, BioTek).
The cell viability was assessed using a 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. Briefly, a total of 100 μL MTT (final concentration 0.5 mg/mL in 0.01 M pH 7.2 phosphate-buffered saline) was added to each well and another 4 h of incubation at 37°C. Then, the supernatant was removed and 100 μL of dimethyl sulfoxide was added into each well. Optical density was measured at the 570 nm wavelength by the use of the ELX-800 microplate assay reader (Bio-tek). The results were expressed as a percentage of absorbance measured in control cells.
The statistical analyses were performed using SPSS 10.0. Data are presented as mean ± SD. The difference between means was determined by One-Way analysis of variance followed by a Student-Newman-Keuls test for multiple comparisons. A probability value of *P < 0.05 was taken to be statistically significant.
Identification of Sequences Required for Hypoxia-inducible Transcription From DMT1 Promoter
We focused our attention on detailed DMT1 exon1B promoter analysis, being of particular interest the search for putative HIF-binding sequences. We used the online data base MatInspector (www.genomatix.de). An extensive analysis of 5′-flanking region of DMT1 exon1B promoter sequence revealed the presence of several putative consensus binding sites for various transcription factors. Among all the putative response sequences found in the DMT1 exon1B promoter region, HRE was one of special interests on factors of the transcriptional regulation of the gene (Fig. 1). A HRE that 100% homologous to the consensus HIF binding site (-ACGTG-) was present. The previously reported analysis of published HRE sequences are 5′-R(A/G)CGTG-3′ (3). We also found that DMT1 exon1B promoter contains the other two putative HREs. The positions of HRE sequences located in the DMT1 promoter were shown (Tables 2 and 3). Spanning the region of DMT1 promoter revealed interesting new potential hypoxia binding sites at positions −327 to −323, −175 to −171, and −143 to −139 that seemed to be good candidates to contribute to the DMT1 stimulation by hypoxia.
Response of Nested Deletions in the 5′-Flanking Region of DMT1 Gene Promoter to Hypoxic Treatment
To delimit the promoter region mediating activation by hypoxia for 6 h, different fragments of the DMT1 exon1B promoter were generated and cloned into luciferase reporter vector (Fig. 2A). HepG2 cells were transiently transfected with these reporter constructs and pRL-TK vector to normalize transfection efficiencies. As shown in Fig. 2B, luciferase activities from cells transfected with construct pGL3-DMT1-prom treated with hypoxia were comparable and significantly higher than those transfected with pGL3-Basic. Thus, 3.11 ± 0.53 fold inductions were observed for pGL3-DMT1-prom construct after hypoxia treatment. Similar result in Fig. 2C, 2.97 ± 0.51-fold inductions were obtained for pGL3-DMT1-prom-HRE1 construct, after hypoxic treatment. We revealed that luciferase activities of pGL3-DMT1-prom-HRE2 and pGL3-SV40 were not significantly different.
To further determine whether the above putative HREs are essential for DMT1 exon1B responsed to hypoxia, we inactivated this HIF binding sequence by site-directed mutagenesis. Mutation of HRE (CAGTACCTAACGTGGCGCCA→CAGTACCTAAAATGGCGCCA) reduced the induction of reporter activity after stimulation with hypoxia from 3.15 ± 0.37 fold to 1.07 ± 0.06 fold (P < 0.01 vs. wt) (Fig. 2C).
To test whether exogenous HIF-1α overexpression could cause the same stimulatory effects on DMT1 exon1B promoter as those observed with hypoxia, HIF subunits expression vectors were assayed. DMT1 exon1B promoter constructs were cotransfected with pcDNA3.1 or pcDNA-HIF-1α and pRL-TK vector to normalize transfection efficiencies. The pcDNA-HIF-1α construct allows exogenous HIF-1α to express under normoxic conditions. Cotransfection of pGL3-DMT1-prom-HRE1 with the pcDNA-HIF-1α showed a 2.49 ± 0.65 fold in luciferase activity in contrast to the basic condition. On the other hand, no significant additive stimulation was observed when HRE mutated (Fig. 2D), indicating that the levels of exogenous HIF-1α are sufficient for full stimulation when cells are expressing constitutively active HIF-1α in normoxic condition.
Effect on Total Cellular Iron Levels by DMT1 Exon1B
To determine whether DMT1 exon1B could alter total cellular iron levels, HepG2 cells were transfected with two isoform of DMT1 exon1B expression plasmids. Results in Fig. 3 demonstrate that +IRE DMT1 exon1B increased total cellular iron levels significantly by approximately two-fold compared with control. Meanwhile, −IRE DMT1 exon1B might have no effect on total iron uptake. Specific silencing of DMT1 in HepG2 cells was documented by Western blot (unpublished data).
Effect on ROS Levels by DMT1 Exon1B
We have demonstrated that total iron level was increased by DMT1 exon1B. As we know, iron has its properties of auto-oxidation and free radical generation of active oxygen species capable of attacking other biomolecules. Now many fluorescent probes have been used for detecting for ROS (17). Here, we used two fluorescent probes, hydroethidine (dihydroethidium; HE) and 2,7-DCFH, to measure the ROS level. HE was chosen to detect the superoxide anion (O). DCFH was used to detect the hydrogen peroxide (H2O2) and hydroxyl radical (HO•). There were no changes with HE fluorescent activities both in +IRE DMT1 exon1B and –IRE DMT1 exon1B (Fig. 4A). Results in Fig. 4B showed that +IRE DMT1 increased DCFH fluorescent activities. Meanwhile, −IRE DMT1 exon1B might have no such effect. Then, MTT assay was performed to determine the growth inhibition rate. As shown in Fig. 4C, +IRE DMT1 exon1B could inhibit the proliferation of the HepG2 cells, and −IRE DMT1 exon1B could not reduce the growth of the cells.
Effect on Cellular Iron Level by DMT1 Under Hypoxia
To define the role of DMT1 in regulating iron uptake of cells under hypoxia, we used a siRNA approach to suppress DMT1 expression using human DMT1-specific siRNA duplex (Dharmacon). To control for nonspecific effects of the siRNA transfection, siRNAcontrol nontargeting siRNA duplex were used. There was a significant difference between siRNA DMT1 group and control group both under normoxia and hypoxia. The results showed that DMT1 RNA interference might reverse the increase of total iron content (Fig. 5A). To confirm that the calcein fluorescence method provides a valid measure of the ferrous uptake, a baseline signal was obtained from normal cells and those with no ferrous added cells. This indicated that the fluorescence was steady in the 30-min recording (Fig. 5B). We assessed the effects of downregulating DMT1 protein expression on ferrous uptake by HepG2 cells. The fluorescence of siDMT1 group decreased less than control group at the beginning of 15 min. The 2.71 ± 0.23-fold fluorescence degression was obtained for siDMT1 group at 30 min. The data indicated that the less DMT1 expression by RNA interference might lead to less ferrous iron uptake.
HIF-1 is a master regulator of oxygen homeostasis that plays critical roles in a multitude of developmental and physiologic processes. Several dozens of HIF-1 target genes have been identified to date that participate in responses to hypoxia (18).
The expression of a variety of proteins involved in iron homeostasis, such as erythropoietin (19), ferritin, cerruloplasmin, transferrin, and transferrin receptor, have been reported to be induced during hypoxia (20). These changes are compensatory to the low oxygen environment and presumably restore metabolism toward normal or functionally acceptable homeostatic conditions required for cell survival. DMT1, the principal transport protein for iron and other transition metals, behaves in an analogous fashion to these other essential components involved in iron homeostasis. DMT1 expression is modified by hypoxia in a compensatory manner presumably to help preserve normal iron balance in vivo (11).
Previous studies have provided an evidence of a high correlation between the expression of HIF-1α and DMT1 proteins in responding to CoCl2 or hypoxia in HepG2 cell (14). Sequence analysis revealed the presence of several putative HREs in human DMT1 exon1B promoter, which could explain the described effect of hypoxia on the induction of DMT1 exon1B gene. Luciferase activities from pGL3-DMT1-prom construct showed that it contained functional HREs. Compared pGL3-DMT1-prom-HRE1 with pGL3-DMT1-prom-HRE2, fold induction results show that promoter sequences site in −175 to −171 and −143 to −139 were negligible. These two putative HREs (5′-GCGTC-3′) of the DMT1 promoter are not essential in the hypoxic response. The putative HRE located at −327 to −323 may be relevant for physiological hypoxic response in the DMT1 exon1B gene. To confirm unequivocally the importance of HIF-1α, DMT1 exon1B gene expression was induced with the use of the transactivating factors HIF-1α. Significant difference in luciferase activities was observed when cotransfected with pcDNA-HIF-1α in normoxic condition. Thus, results from HepG2 cells indicate that the levels of exogenous HIF-1α are sufficient for full stimulation when cells are expressing constitutively active HIF-1α in normoxic condition. The mutation analyses indicate that this HRE sequence is essential for the DMT1 hypoxic response.
Previous studies found that HIF-2α played a crucial role in maintaining iron balance in the organism by directly regulating the transcription of the gene encoding DMT1 exon1A, the principal intestinal iron transporter. Specific deletion of HIF-2α led to a decrease in serum and liver iron levels and a marked decrease in liver hepcidin expression, indicating the involvement of an induced systemic response to counteract the iron deficiency (12, 13). Altogether, these results demonstrate that both DMT1 exon1A and exon1B are hypoxia-inducible genes that are stimulated through HIF factor interaction.
Our pervious data showed that when cells exposed to hypoxia, +IRE DMT1 protein expression increased and reached peak at 6 h. –IRE DMT1 protein expression reached the highest level at 6–12 h hypoxia (14). Both DMT1 exon1A and exon1B were regulated by hypoxia. The function of human DMT1 different exons is the equivalent efficiency as metal ion transporters (21). The content of total iron and ferrous influx under hypoxia has been little reported. We showed in this article that total and ferrous iron levels were significantly higher under hypoxia or +IRE DMT1 exon1B express in excess. However, continued delivery of iron to cells can overwhelm the capacity of ferritin to store and sequester the metal, inducing oxidative injury to cells. Indeed, iron can act as a catalyst in the Fenton reaction to potentiate oxygen and nitrogen toxicity by the generation of a wide range of free radical species, including hydroxyl radicals, or the peroxynitrite anion, produced by the reaction between NO and the superoxide anion (22). Hydroxyl radicals are the most reactive free radical species known and have the ability to react with a wide range of cellular constituents, including aminoacid residues and purine and pyrimidine bases of DNA, as well as attacking membrane lipids to initiate a free radical chain reaction known as lipid peroxidation (22). The data in our experiments also show that cells might be damaged by the ROS production when iron was present in excess. Several studies have reported that DMT1 functions as a transporter for a variety of metals including manganese, cobalt, copper, cadmium, nickel in addition to iron (23).
The fluorescence of siDMT1 group decreased less than control group at the beginning of 20 min under normoxia. The 1.39 ± 0.19-fold fluorescence degression was obtained for siDMT1 group at 30 min under normoxia (data not shown). Compared with our pervious results, ferrous influx had been delayed when cells treated in normoxia. Besides, the fluorescent degression fold was obviously higher under hypoxia than normoxia. These experiments indicated that there is a significant increase of iron content and ferrous influx in hypoxia. These results suggested that hypoxic stimulation had an important effect on cellular iron transport maybe principal through regulating DMT1 expression. Thus, we supposed that hypoxia might induce the rapid localization of DMT1 transported to the plasma membrane (14). Subsequently, iron uptake occurred through a pathway involving DMT1. The increasing of iron content and ferrous influx displayed a sufficient consistent with DMT1 protein expression under hypoxia. Such a rapid response to hypoxia might allow cells to sequester sufficient iron to maintain enzyme function and cellular survival during a potentially extended period of low oxygen concentration. Our favored explanation for the regulation of DMT1 during hypoxia was increased iron uptake.
Several studies have reported that DMT1 functions as a transporter for a variety of metals including iron, manganese, cobalt, copper, cadmium, and nickel (24, 25). At present, it is little known that whether the upregulation of DMT1 by hypoxia has effect on cell uptake of these ions. If the answer is positive, DMT1 may be more important and necessary for adaptation of cellular metabolism under hypoxia.
In conclusion, we have performed a detailed analysis of the DMT1 exon1B promoter demonstrating that oxygen-regulated function depends on HIF-1-binding sites. Hypoxia increased total and ferrous iron levels through DMT1. ROS changed by +IRE DMT1 exon1B overexpression might cause the death of cells. Although iron is involved in multiple physiological processes, the biological significance of enhanced DMT1 expression for iron during hypoxia remains to be determined.
This study was financially supported by the National Natural Science Foundation of China (Grant No. 30770806 & 30971197), Postgraduate Project of Jiangsu Province (CX08S-027Z), and Natural Science Fund of Nantong University (No. 09Z052). The plasmids of pIRESneo, pIRESneo-DMT1 isoform I, and pIRESneo-DMT1 isoform II were kind gifts from Dr. Mitsuaki Tabuchi (Kawasaki Medical School, Japan).