Zinc and cadmium can promote rapid nuclear translocation of metal response element-binding transcription factor-1.

Metal response element-binding transcription factor-1 (MTF-1) is a six-zinc finger protein that plays an essential role in activating metallothionein expression in response to the heavy metals zinc and cadmium. Low affinity interactions between zinc and specific zinc fingers in MTF-1 reversibly regulate its binding to the metal response elements in the mouse metallothionein-I promoter. This study examined the subcellular distribution and DNA binding activity of MTF-1 in cells treated with zinc or cadmium. Immunoblot analysis of cytosolic and nuclear extracts demonstrated that in untreated cells, about 83% of MTF-1 is found in the cytosolic extracts and is not activated to bind to DNA. In sharp contrast, within 30 min of zinc treatment (100 microM), MTF-1 is detected only in nuclear extracts and is activated to bind to DNA. The activation to bind to DNA and nuclear translocation of MTF-1 occurs in the absence of increased MTF-1 content in the cell. Furthermore, immunocytochemical localization and immunoblotting assays demonstrated that zinc induces the nuclear translocation of MTF-1-FLAG, expressed from the cytomegalovirus promoter in transiently transfected dko7 (MTF-1 double knockout) cells. Immunoblot analysis of cytosolic and nuclear extracts from cadmium-treated cells demonstrated that concentrations of cadmium (10 microM) that actively induce metallothionein gene expression cause only a small increase in the amount of nuclear MTF-1. In contrast, an overtly toxic concentration of cadmium (50 microM) rapidly induced the complete nuclear translocation and activation of DNA binding activity of MTF-1. These studies are consistent with the hypothesis that MTF-1 serves as a zinc sensor that responds to changes in cytosolic free zinc concentrations. In addition, these data suggest that cadmium activation of metallothionein gene expression may be accompanied by only small changes in nuclear MTF-1.

which play a role in zinc homeostasis, cadmium detoxication, and protection from reactive free radicals (1)(2)(3)(4)(5)(6). The rapid induction of MT-I and -II gene transcription by heavy metals (1) is mediated by metal response elements (MREs), present in multiple copies in the proximal promoters of MT genes (7). A protein that binds directly and specifically to MREs (8) and transactivates MT gene expression is referred to as MTF-1 (9).
MTF-1 is a six-zinc finger protein in the Cys 2 His 2 family of transcription factors. Human, mouse, and pufferfish MTF-1 have been cloned (10 -12), and this protein has been highly conserved, particularly in the zinc finger domain. The C terminus of mammalian MTF-1 contains three transactivation domains, which are acidic, proline-rich, and serine/threonine-rich, respectively (13). The DNA binding activity of native and recombinant MTF-1 is reversibly modulated by zinc interactions with the finger domain (14). The zinc fingers are heterogeneous in function and at least two exhibit low affinity binding of zinc (15,16).
Treatment of cells with zinc in vivo results in a rapid, dramatic increase in the DNA binding activity of MTF-1 measured in vitro (10,14) and the concomitant occupancy of MREs in the MT-I promoter in vivo (3). In contrast, the DNA binding activity of MTF-1 is apparently not activated by transition metals other than zinc (8,17,18), although cadmium is a particularly potent inducer of MT gene expression. Homozygous deletion of the mouse MTF-1 gene revealed that MTF-1 is essential for zinc and cadmium induction, as well as for basal expression of the mouse MT-I and -II genes in embryonic stem cells (19). MTF-1 is also essential for induction of these genes by oxidative stress (3) and hypoxia (20). Thus, several signal transduction pathways may impinge on the activities of MTF-1. Mice homozygous for targeted deletions of the MTF-1 gene die in utero due to failure of liver development, demonstrating that the MTF-1 gene is an essential gene (21), unlike the mouse MT-I and -II genes (22).
Transition metal regulation of gene expression has been documented in species from every kingdom of organisms. In many instances, the transition metal itself directly interacts with a preexisting metalloregulatory protein, and this interaction leads to a change in conformation of the protein and an alteration in the DNA or RNA binding activity of the protein (23). The available evidence suggests that MTF-1 is a metalloregulatory protein that serves as an intracellular zinc sensor to activate gene expression. This model predicts that MTF-1 would be located in the cytoplasm to facilitate direct interaction with free zinc. Since previous studies have not addressed the subcellular localization of MTF-1, it is not clear whether this cellular response to metal ions is initiated in the cytosol or nucleus. Furthermore, it is unclear how MTF-1 senses the toxic metal cadmium. Therefore, we examined the effects of zinc and cadmium treatment on the subcellular localization and DNA binding activity of mouse MTF-1.
Cell Culture-Mouse Hepa cells were maintained in Dulbecco's modified Eagle's medium-high glucose (DMEM) supplemented with 2% fetal bovine serum (FBS). The mouse dko7 cell line is a simian virus 40 large T-antigen-immortalized fibroblast derived from embryonic stem cells lacking MTF-1 (MTF-1 double knockout) and was a generous gift of Dr. Walter Schaffner, University of Zurich (Zurich, Switzerland) (13). These cells were maintained in DMEM supplemented with 10% FBS. For nuclear and cytosolic extract preparations, cells (2 ϫ 10 6 ) were plated in 15-cm Petri dishes and grown to 80% confluency. For transfection followed by extract preparation, cells (1.2 ϫ 10 5 ) were plated in six-well plates (9.4 cm 2 ) and grown to 50% confluency. For transfection and subsequent immunocytochemistry, cells (2.5 ϫ 10 4 ) were plated in four-chamber glass slides (1.8 cm 2 ) and grown to 50% confluency. All FBS was heat-inactivated prior to use; all media were supplemented with 50 units/ml penicillin, 50 g/ml streptomycin, and 2 mM L-glutamine.
Preparation of Cell Extracts-Whole cell extracts, nuclear extracts (NEs), and cytosolic extracts (CEs) were prepared essentially as described (14,24). Briefly, for preparation of nuclear extracts, treated cells were placed on ice, the medium was removed, and cells were washed once with cold PBS. Cells were scraped off the dish and collected by centrifugation at 1,500 ϫ g for 5 min. The cell pellet was resuspended in 5 ml of cell lysis buffer (10 mM HEPES (pH 7.9), 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM dithiothreitol, and 0.2 mM phenylmethylsulfonyl fluoride), and immediately centrifuged at 1,500 ϫ g for 5 min. Cells were resuspended in 2 times the original packed cell volume of cell lysis buffer, allowed to swell on ice for 10 min, and homogenized with 10 strokes of a Dounce homogenizer (B pestle). Nuclei were collected by centrifugation at 3,300 ϫ g for 15 min at 4°C, and supernatant was saved for cytosolic extracts. The nuclei were resuspended, using six strokes of a Teflon-glass homogenizer, in 3 volumes (about 750 l) of nuclear extraction buffer (20 mM HEPES, pH 7.9, 1.5 mM MgCl 2 , 400 mM KCl, 0.5 mM dithiothreitol, 0.2 mM phenylmethylsulfonyl fluoride, and 25% glycerol). The nuclear suspension was stirred on ice for 30 min and then centrifuged at 89,000 ϫ g for 30 min. The supernatant was collected and concentrated in a Microcon 10 concentrator by centrifugation at 14,000 ϫ g for 3 h at 4°C. For preparation of CE, the supernatant obtained after removal of nuclei was mixed thoroughly with 0.11 volume of 10ϫ cytoplasmic extraction buffer (1ϫ cytoplasmic extraction buffer: 30 mM HEPES (pH 7.9) at 4°C, 140 mM KCl, 3 mM MgCl 2 ) and then centrifuged at 89,000 ϫ g for 1 h. The supernatant was collected and concentrated in a Microcon 10 concentrator by centrifugation at 14,000 ϫ g for 1 h at 4°C. Protein concentration was determined using Bio-Rad Protein Assay reagent with bovine serum albumin as the standard.
In transfection experiments, NE-PER nuclear and cytoplasmic extraction reagents was used to prepare extracts. However, because of the presence of EDTA, the addition of exogenous zinc was required to activate MTF-1 DNA binding activity in these extracts.
Preparation of Total Protein SDS Extracts-SDS lysis of cells was performed on plates using 1ϫ SDS sample buffer without reducing agent or bromphenol blue. Protein concentration was determined using BCA protein assay reagent and bovine serum albumin as the standard.
Immunoblotting-Cell extracts (50 -100 g of protein) were separated by 10 or 12% SDS-polyacrylamide gel electrophoresis (25) under reduced conditions and transferred to nitrocellulose membranes. The membranes were blocked overnight at 4°C in 10% nonfat dry milk in PBS, 0.1% Tween 20 and probed with primary antibody diluted in 3% nonfat dry milk in PBS, 0.1% Tween 20 for 1 h at room temperature. Membranes were then incubated with an appropriate secondary antibody conjugated to horseradish peroxidase diluted in 3% nonfat dry milk in PBS, 0.1% Tween 20 for 30 min at room temperature, developed by chemiluminescence, and exposed to hyperfilm ECL. Relative band intensities were quantitated using Biomax 1D image analysis software (Kodak Scientific Imaging Systems). Equal protein loading and transfer was verified visually by staining membranes with Ponceau solution.
Expression Vector Construction-The CMV-MTF-1 expression vector was described previously (14). The MTF-1-FLAG construct was created by polymerase chain reaction amplification from this template using a sense primer that encompassed the translation start codon and an antisense primer against the carboxyl terminus that also incorporated the FLAG coding sequence. The amplified product was cloned into the CMV vector. Vectors were verified by DNA sequencing.
Transient Transfection-dko7 cells were transfected using Lipo-fectAMINE according to the manufacturer's instructions. Cells were grown to ϳ50% confluence. After washing the cells with serum-free DMEM, DNA and LipofectAMINE mixture prepared in serum-free DMEM was added. For immunocytochemistry, the mixture consisted of 2 l/well LipofectAMINE, 4 ng/well CMV-MTF-1-FLAG expression vector, and 300 ng/well SV-␤-gal, as an internal control for transfection efficiency, in 250 l of DMEM. In experiments in which preparation of nuclear and cytosolic extracts was performed, cells were treated with 4 l/well LipofectAMINE, 100 ng/well CMV-MTF-1 expression vector, and 1 g/well SV-␤-gal in 1.2 ml of DMEM. After 5 h, an equal volume of DMEM containing 2ϫ FBS was added, and the incubation was continued overnight. The following morning, the medium was removed and replaced with fresh DMEM containing 1ϫ FBS. Zinc treatment was initiated in the afternoon of day 2. After 1 h, cells were processed for either immunocytochemistry, as described below, or for nuclear and cytosolic protein isolation.
Immunocytochemistry-Twenty-four hours after transfection, dko7 cells were washed twice with serum-free medium and then incubated for 6 h in DMEM containing 1% (w/v) bovine serum albumin. Cells were then treated for 30 min with 100 M ZnSO 4 in this medium, washed with PBS, and fixed with 70% ethanol for 5 min. Slides were blocked for 1 h at room temperature with 10% goat serum in PBS-Triton X-100 and incubated overnight at 4°C with rabbit polyclonal FLAG antibody or Sp1 antibody diluted 1:500 or 1:100, respectively. Slides were then incubated with anti-rabbit IgG conjugated to peroxidase and stained with a DAB kit.
In Vitro Transcription/Translation of Mouse MTF-1-Synthesis of recombinant mouse MTF-1 was performed using the TnT coupled reticulocyte lysate transcription/translation system (TnT lysate), as described in detail previously (18).

RESULTS
Specificity of Mouse MTF-1 Antisera-Previous studies demonstrated that the rabbit polyclonal antisera against bacterially expressed glutathione S-transferase-MTF-1 was specific for MTF-1 in supershift EMSA (20). The specificity of this MTF-1 polyclonal antisera was examined by immunoblotting ( Fig. 1). Recombinant mouse MTF-1 synthesized in vitro in a TnT lysate system was used as a positive control (Fig. 1, lane  1), and an extract from dko7 (MTF-1 double knockout) cells was used as a negative control (Fig. 1, lane 2). Mouse MTF-1 migrates with an apparent molecular mass of ϳ100 kDa (Fig. 1,  lane 3), despite its predicted size of 72.5 kDa (10). This aberrant mobility may reflect the clustering of acidic, serine, and proline residues in the structure of MTF-1 (10, 13) and is not unique among transcription factors (26,27). The MTF-1 band was absent in extracts from dko7 cells (Fig. 1, lane 2) but was detected in whole cell extracts from mouse Hepa cells (Fig. 1,  lane 3) and from dko7 cells transiently transfected with an MTF-1 expression vector (see below). Two other bands with apparent molecular masses of ϳ200 and ϳ57 kDa were detected in cell extracts from both dko7 and Hepa cells (Fig. 1).

MTF-1 Is Localized in the Cytosol in Untreated Cells and in the Nucleus in
Zinc-treated Cells-The subcellular distribution of MTF-1 in untreated and zinc-treated cells was investigated by immunoblotting. Nuclear and cytosolic extracts were prepared from mouse Hepa cells, and it was noted that approximately 4-fold more protein was extracted in the cytosolic versus the nuclear extracts obtained from the same number of cells. On average, these extraction procedures recovered 88 pg of cytosolic protein and 21 pg of nuclear protein per cell. Immunoblotting of these extracts was performed using MTF-1 antiserum, and extracted proteins were normalized per cell for analysis. To account for differences in the amount of protein recovered in nuclear versus cytosolic extracts, 50 g of nuclear and 200 g of cytosolic protein were loaded ( Fig. 2A, right panel). Quantitation of relative intensities of the MTF-1 bands in these samples suggested that 17 Ϯ 9% of the immunoreactive MTF-1 was extracted in the nuclear fraction whereas 83 Ϯ 9% of the MTF-1 was extracted in the cytosolic fraction from control cells. In the remaining figures, equal quantities of protein per lane were applied to the gels. Some variability between experiments in the amount of MTF-1 extracted from nuclei versus cytosol was noted, as is demonstrated by the S.D. value shown above. This variability was accentuated in the nuclear extracts relative to cytoplasmic extracts and may reflect subtle differences in cell density, cell passage number, or culture conditions. Each experiment contained an internal control of untreated cells cultured in parallel under identical conditions.
In contrast to the results obtained using extracts from untreated Hepa cells, immunoblotting revealed that all MTF-1 im-munoreactivity was present in nuclear extracts from cells treated with 100 M ZnSO 4 for 1 h. The cytosolic extract was devoid of detectable MTF-1 ( Fig. 2A). Fig. 2A (left panel) is an immunoblot where equal amounts of nuclear and cytosolic proteins were applied to the gel. The amount of immunoreactive MTF-1 detected in nuclear extracts increased about 4-fold after zinc treat-  USF1. B, upper panel, nuclear extracts were analyzed for DNA binding activity using a labeled MTF-1-specific MRE-s oligonucleotide, as described under "Experimental Procedures." Lower panel, the same extracts were assayed using an Sp1-specific oligonucleotide. The arrows point to specific MTF-1 and Sp1 complexes with their respective oligonucleotides. ment of the cells. This is consistent with the data suggesting that about 83% of MTF-1 is initially found in the cytosolic fraction from untreated cells. Immunoblot analysis of proteins remaining in the nuclear pellet fraction after extraction revealed no MTF-1, Sp1, and USF1. Thus, MTF-1 is not preferentially extracted from nuclei of zinc-treated cells (data not shown). In contrast to MTF-1, two other transcription factors, known for their constitutive nuclear localization, Sp1 (Fig. 2B) and USF1 (Fig. 2C), were detected only in nuclear extracts, and the amount of immunoreactive protein was unaffected by zinc treatment. Thus, the cytosolic extracts were not significantly contaminated with nu-clear transcription factors. Furthermore, immunoblotting with antisera against the predominantly cytosolic heat shock protein 90 (Hsp90) (28,29) revealed that the majority of Hsp90 was detected in cytosolic extracts (Fig. 2D). Thus, nuclear extracts were not significantly contaminated with cytosolic proteins.
Nuclear Localization of MTF-1 Is Accompanied by Activation of DNA Binding Activity-EMSA was used to detect the MRE binding activity of MTF-1 (3,10,14) in extracts from untreated and zinc-treated Hepa cells. Nuclear and cytosolic extracts from untreated cells contained little MTF-1 that was active to bind to DNA (Fig. 3A). Previous studies demonstrated that MTF-1 in whole cell extracts from control cells can be activated in vitro by the addition of zinc (5-30 M) followed by incubation at 37°C (14). After zinc treatment, however, the DNA binding activity of MTF-1 increased 8 -12-fold in nuclear extracts, while cytosolic extracts exhibited no detectable MTF-1 DNA binding activity (Fig. 3A). The identity of the MTF-1⅐MRE-s complex was confirmed by supershift EMSA using the MTF-1 antisera (data not shown). The rapid and dramatic increase in MTF-1 DNA binding activity in nuclear extracts from cells treated with zinc correlates with the immunoblotting data and suggest that zinc induces the nuclear translocation and activation of MTF-1. Sp1 was detected only in nuclear extracts, and its DNA binding activity was unaffected by exogenous zinc (Fig. 3B).
Zinc-induced Nuclear Accumulation of MTF-1 Is Rapid-A time course for zinc-dependent nuclear accumulation of MTF-1 protein and of MRE binding activity of MTF-1 was determined using Hepa cells treated with 100 M ZnSO 4 (Fig. 4). The amount of immunoreactive MTF-1 in the nucleus increased about 2-fold by 5 min after the addition of zinc and 4-fold by 30 min (Fig. 4A). As a control for potential differences in protein loading, MTF-1 was compared with immunoreactive USF1 in these same extracts (Fig. 4A). Nuclear USF1 levels remained constant during zinc treatment. MRE binding activity of MTF-1 was monitored by EMSA (Fig. 4B). A 2.5-fold increase in MRE binding activity was detected by 5 min after zinc treatment and a 7-fold increase was detected by 30 min, consistent with previous observations (18). Sp1 DNA binding activity remained constant (Fig. 4B).
Zinc Treatment Does Not Alter the Amount of MTF-1 Protein-Immunoblotting of total cell SDS extracts was used to determine whether zinc causes a rapid change in the steady state levels of MTF-1 in Hepa cells. Hepa cells were treated with 100 M ZnSO 4 for 1 h, and the cells were lysed in situ in SDS sample buffer. Equal amounts of SDS-extracted proteins were then examined by immunoblotting. There was no detectable change in the amount of immunoreactive MTF-1 after this zinc treatment (Fig. 5). Likewise, there were no detectable changes in the relative amount of immunoreactive Sp1 or USF1. In addition, levels of immunoreactive MTF-1 in whole cell extracts from control and zinc-treated Hepa cells were unaffected by pretreatment of the cells with cycloheximide (10 nM to 100 M) for 1 h (data not shown). Taken together, these data indicate that zinc does not cause a detectable increase in the steady state levels of MTF-1 protein within 1 h in these cells, nor is MTF-1 protein rapidly degraded.
MTF-1 Expressed in Transiently Transfected dko7 Cells Also Translocates to the Nucleus after Zinc Treatment-To further address the effects of zinc on the subcellular distribution of MTF-1, dko7 cells, which have no endogenous MTF-1, were transiently transfected with a CMV-MTF-1 expression vector and then treated with 100 M ZnSO 4 for 1 h. Nuclear and cytosolic extracts were prepared using the NE-PER reagent and analyzed by immunoblotting (Fig. 6). In untreated cells, MTF-1 was detected in both the nuclear and cytosolic extracts. In contrast, after zinc treatment, about 3-fold more immunoreactive MTF-1 was detected in nuclear extracts, and cytosolic extracts were devoid of detectable MTF-1. These extraction buffers contained EDTA, which inactivates the DNA binding activity of MTF-1. This process is reversible by the readdition of zinc. The addition of zinc also serves to activate cytosolic MTF-1 previously not detected by EMSA. MTF-1 DNA binding was activated by incubating the extracts with 100 M ZnSO 4 for 15 min at 37°C prior to EMSA. Cytosolic extracts from untreated cells contained MTF-1 that was activated to bind to DNA in vitro, whereas cytosolic extracts from zinc-treated cells were devoid of MTF-1 binding activity. The amount of MTF-1 binding activity in nuclear extracts increased 3.5-fold in cells treated with zinc compared with untreated cells (Fig. 6B).
Immunocytochemistry was used to localize MTF-1-FLAG, expressed in transiently transfected dko7 cells. This approach was taken because the polyclonal antisera against MTF-1 produced too much background to allow unambiguous detection of endogenous MTF-1. Therefore, cells were transiently transfected with a CMV-MTF-1-FLAG vector, treated with 100 M ZnSO 4 for 1 h, and processed for immunocytochemistry with commercially available FLAG probe antisera. Untreated and zinc-treated, mock-transfected cells served as a control for specificity of the antisera (Fig. 7, A and B, respectively). Slight background immunostaining was detected in mock-transfected cells. However, strong cytoplasmic FLAG immunostaining was detected in 5-10% of the cells after transfection with the CMV-MTF-1-FLAG vector (Fig. 7C). This percentage of cells approximated the transfection efficiency based on ␤-galactosidase staining of cells in situ (data not shown). After treatment of transfected cells with zinc, this immunostaining was localized in nuclei (Fig. 7D). By comparison, intense nuclear staining of Sp1 was detected in untreated, as well as zinc-treated transfected cells (Fig. 7, E and F,  respectively). These results provide convergent evidence that MTF-1 translocates to the nucleus in response to zinc.

An Overtly Toxic Concentration of Cadmium Can Induce Nuclear Translocation and MRE Binding Activity of MTF-1 in
Hepa Cells-Several studies have reported that cadmium does not activate the DNA binding activity of MTF-1 either in vivo or in vitro (8,17,18). However, MTF-1 is essential for cadmium activation of MT gene expression (19). On a molar basis, cadmium is at least 10 times more toxic and more potent than zinc as an inducer of MT-I gene expression (30,31). To further investigate the effects of cadmium on MTF-1, Hepa cells were treated with 10 M CdCl 2 , which results in nearly maximal induction of MT gene expression, or treated with 20 or 50 M CdCl 2 . The latter concentration is very toxic to Hepa cells under these culture conditions and will kill them within 24 h. Cadmium caused a dose-and time-dependent increase in the amount of immunoreactive MTF-1 in the nucleus (Fig. 8). Based on comparisons of relative intensities of the MTF-1 bands, 10 M cadmium caused less than a 2-fold increase in immunoreactive MTF-1 in the nuclear extract (Fig. 8A). Increased nuclear MTF-1 was apparent within 15 min of the addition of the cadmium (Fig. 8B). We previously reported that treatment of cells with 10 M cadmium only modestly activates MTF-1 DNA binding activity (Ͻ2-fold) (14,18). In contrast, 50 M cadmium caused a 7.5-fold increase in immunoreactive nuclear MTF-1 within 1 h of treatment, and increased MTF-1 was evident by 15 min (Fig. 8B). The DNA binding activity of MTF-1 was examined using the nuclear extracts from cells treated with 50 M cadmium for 15 min and 1 h (Fig. 9). Under these conditions, 50 M cadmium caused about a 5-fold induction of DNA binding. Sp1 DNA binding activity was not altered by cadmium treatment (Fig. 9, lower panel). DISCUSSION Previous studies have documented that MTF-1 is essential for metal ion regulation of MT gene expression (19) and that this metalloregulatory protein is activated by zinc to bind to MREs in the MT promoter (3). Treatment of cells with zinc results in a rapid increase in the amount of DNA binding activity of MTF-1 detected in nuclear extracts. However, it has been shown that zinc-induced MT gene expression is not dependent on de novo protein synthesis (32), that MTF-1 mRNA is not induced by zinc (10,19,33), and that whole cell extracts from untreated cells contain latent MTF-1 that can be activated to bind to DNA by exogenous zinc (14). Taken together, these observations demonstrate that MTF-1 is a preexisting cellular protein that is activated to bind to DNA by metal ions. The data reported herein reveal that a significant portion of MTF-1 protein is present in the cytoplasm of unstressed cells and that exposure of the cells to metal ions results in the rapid translocation of MTF-1 protein to the nucleus and the activation of its DNA binding activity. These results are consistent with the concept that MTF-1 functions, in part, as a zinc sensor.
Cadmium is a potent inducer of MT gene expression, and MTF-1 is also an essential component of that signaling mechanism. However, MTF-1 is activated to bind to DNA by reversible interactions of zinc with specific zinc fingers in the DNAbinding domain and not by interactions with cadmium or other transition metals (8 -10, 14, 16 -19, 34). We previously reported that 6 M CdCl 2 has little effect on the amount of MTF-1 DNA binding activity in the nuclei of cultured cells (18). Cadmium (5-15 M) rapidly induces MT-I gene expression in these cells 2 and in other cell types (33). Herein, it was further shown that cadmium (10 M) exerts only a small effect (15% increase) on the amount of MTF-1 protein in the nucleus. However, higher concentrations of cadmium caused the complete activation of DNA binding and translocation of MTF-1 to the nucleus. This suggests that cadmium may cause the redistribution of zinc in the culture, which, in turn, may activate MTF-1 to bind to DNA and move to the nucleus. Other transition metals have been suggested to act in this manner (33). However, this mechanism may only account in part for the effects of cadmium on activation of the MT promoter by MTF-1.
The disassociation between the concentration of cadmium required for activation of MT-I gene expression and that required to cause most MTF-1 to be activated to bind to DNA and translocated to the nucleus suggests several possibilities including the following: 1) only small changes in MTF-1 levels in the nucleus are sufficient for maximal activation of gene expression; 2) cadmium causes a modification of MTF-1; 3) cadmium and zinc cause the formation of distinct MTF-1 promoter complexes; or 4) cadmium-responsive transcription factors other than MTF-1 may also activate MREs. The concentrationresponse curves for zinc activation of MT gene expression and increased MTF-1 DNA binding activity do not support the first possibility. However, the possibility of effects of cadmium on the transactivation capacity of MTF-1 cannot be excluded. The transactivation domains present in the carboxyl-terminal half of the protein are also important for transduction of the metal signal. Thus, the full biological functions of MTF-1 are dependent on a complex interplay of different functional domains (13). Little is known about those interactions or the interactions of MTF-1 with other proteins; thus, the second and third possibilities remain to be addressed. With regard to the fourth possibility, we recently found that MRE activity can be increased in the absence of detectable MTF-1 in IMR cells (35). Cadmium-responsive factor(s) that can interact with MREs in vitro have been reported previously (8, 36 -38), but the functional significance, if any, of those factors has not been determined. Therefore, other MRE-binding proteins may play a role in regulating MT gene expression in response to cadmium, at least in certain cell types.
The nuclear localization of many transcription factors is a key controlling point in regulating gene expression and accompanies differentiation or changes in the metabolic state of eukaryotic cells (39). Although the majority of transcription factors are localized to the nucleus (40,41), others predominantly reside in the cytoplasm and are translocated to the nucleus in response to stimulus (42)(43)(44). One mechanism associated with transcription factor transport to the nucleus is the nuclear localization signal (NLS) (45). MTF-1 has a putative NLS (KRKEVVKR) that immediately precedes the zinc finger domain (10). For a large protein to translocate to the nucleus, the NLS has to be exposed on the protein surface (45). Interactions between zinc and MTF-1 probably cause conformational changes leading to uncovering of the NLS. Another mechanism that exposes the NLS is phosphorylation of adjacent sites (39). MTF-1 has several potential phosphorylation sites in the vicinity of the NLS such as Thr 131 for protein kinase C, Tyr 139 for tyrosine kinase, and Thr 142 for casein kinase. Protein kinase C has been suggested to play a role in metal induction of MT gene expression (46). Finally, it is possible that the zinc fingers themselves are involved in metal-induced nuclear translocation of MTF-1. Several mutations of zinc fingers in MTF-1 were found to cause the cytoplasmic localization of MTF-1 expressed in transiently transfected cells (13). Three Cys 2 His 2 -type zinc fingers within the DNA-binding domain of nerve growth factorinduced transcription factor 1-A (also known as Erg1 or Krox24) are necessary for nuclear localization (47), and the vitamin D receptor's NLS signal is located between the two zinc fingers (41). The entire DNA-binding domain of the glucocorticoid receptor, as a functional unit, may be required for nuclear transfer and optimal retention in the nucleus (48).
In conclusion, the present study indicates that the majority of MTF-1 protein is located in the cytoplasm in cells cultured in medium replete with zinc. However, increases in zinc in the culture medium promote the rapid transport into the nucleus and the activation of DNA binding activity of MTF-1. These results are consistent with the concept that MTF-1 serves as a sensor of cytoplasmic metal ions and suggest that zinc and cadmium may utilize MTF-1 differently in the activation of gene expression.