Ctr1 (copper transporter 1) mediates high-affinity copper uptake. Ctr2 (copper transporter 2) shares sequence similarity with Ctr1, yet its function in mammalian cells is poorly understood. In African green monkey kidney COS-7 cells and rat tissues, Ctr2 migrated as a predominant band of ∼70 kDa and was most abundantly expressed in placenta and heart. A transiently expressed hCtr2–GFP (human Ctr2–green fluorescent protein) fusion protein and the endogenous Ctr2 in COS-7 cells were mainly localized to the outer membrane of cytoplasmic vesicles, but were also detected at the plasma membrane. Biotinylation of Ctr2 with the membrane-impermeant reagent sulfo-NHS-SS-biotin [sulfosuccinimidyl-2-(biotinamido)ethyl-1,3-dithiopropionate] confirmed localization at the cell surface. Cells expressing hCtr2–GFP hyperaccumulated copper when incubated in medium supplemented with 10 μM CuSO4, whereas cells depleted of endogenous Ctr2 by siRNAs (small interfering RNAs) accumulated lower levels of copper. hCtr2–GFP expression did not affect copper efflux, suggesting that hCtr2–GFP increased cellular copper concentrations by promoting uptake at the cell surface. Kinetic analyses showed that hCtr2–GFP stimulated saturable copper uptake with a Km of 11.0±2.5 μM and a K0.5 of 6.9±0.7 μM when data were fitted to a rectangular hyperbola or Hill equation respectively. Competition experiments revealed that silver completely inhibited hCtr2–GFP-dependent copper uptake, whereas zinc, iron and manganese had no effect on uptake. Furthermore, increased copper concentrations in hCtr2–GFP-expressing cells were inversely correlated with copper chaperone for Cu/Zn superoxide dismutase protein expression. Collectively, these results suggest that Ctr2 promotes copper uptake at the plasma membrane and plays a role in regulating copper levels in COS-7 cells.
- copper chaperone for Cu/Zn superoxide dismutase (CCS)
- copper transport
- copper transporter 2 (Ctr2)
- plasma membrane
- RNA interference (RNAi)
Copper is an essential micronutrient that plays an important role as a catalytic cofactor for metalloenzymes [1,2]. The redox chemistry of copper also makes it a toxic metal if allowed to accumulate to high levels. Because copper is an essential, but potentially toxic, metal, copper transporters and chaperones tightly regulate its uptake, distribution and elimination in order to maintain copper homoeostasis in cells [1,3,4].
From yeast to specialized cells of mammals, all have evolved specific mechanisms to transport copper efficiently across the plasma membrane into the cell from the extracellular environment. Considerable advancement has been made on the characterization of copper uptake systems in yeast. In the budding yeast Saccharomyces cerevisiae, copper uptake is mediated by the copper transporters Ctr1 or Ctr3 [5,6], whereas in the fission yeast Schizosaccharomyces pombe, copper is imported by a Ctr4–Ctr5 heteromeric complex .
In contrast with yeast, less is known about the systems involved in copper uptake in mammalian cells. However, the human orthologue of the S. cerevisiae Ctr1 has been identified and shown to complement phenotypes of yeast deficient in high-affinity copper uptake . Overexpression experiments in HEK-293 (human embryonic kidney) cells have demonstrated saturable high-affinity copper uptake by human Ctr1 (hCtr1) . Further analyses showed that copper import by hCtr1 is specific, energy-independent and stimulated by acidic pH and high extracellular K+ concentrations .
Despite the critical role of Ctr1 in copper import, copper levels are unaffected in some tissues of Ctr1+/− mice [10,11], and mouse embryonic cells deficient in Ctr1 exhibit ∼30% residual copper transport activity that is characterized by a Km of ∼10 μM . Furthermore, studies indicate that the Km for copper uptake varies in different cell types [13–16], and intestinal epithelial cells from intestinal epithelial cell-specific Ctr1-knockout mice hyperaccumulate copper . Together, these observations indicate the presence of Ctr1-independent copper uptake systems in mammalian cells.
Ctr2 is one candidate gene that may play a role in cellular copper acquisition. In S. cerevisiae, Ctr2 overexpression failed to complement the growth defect of yeast deficient in high-affinity copper uptake . However, yeast lacking Ctr2 were shown to be more resistant to copper toxicity, and yeast deficient in high-affinity copper uptake overexpressing Ctr2 were more resistant to copper starvation . On the basis of these experiments, Ctr2 was initially thought to function as a low-affinity copper transporter.
Subsequent studies have localized S. cerevisiae Ctr2 [19,20] and its S. pombe orthologue Ctr6  to the vacuolar membrane and have indicated that these proteins do not stimulate copper uptake at the plasma membrane. It was demonstrated that overexpression of Ctr2 in a δctr2 strain decreases vacuolar copper concentrations , and yeast lacking Ctr6 display a reduction in activity of the cuproenzyme SOD1 (Cu/Zn superoxide dismutase) . These data, together with experiments showing that Ctr2 functions as an upstream copper donor for copper chaperones , indicate that, in yeast, Ctr2/Ctr6 function to mobilize copper from the vacuole to the cytosol for incorporation into cuproproteins.
hCtr2 was identified in a database search on the basis of its sequence similarity with hCtr1 . hCtr2 is a relatively small protein of 143 amino acids and, like Ctr1, contains three putative transmembrane domains . Although studies in yeast have indicated a role for Ctr2/Ctr6 as vacuolar copper transporters, there is presently little information on the function of Ctr2 in mammalian cells. In the present study, using biochemical techniques in combination with overexpression and RNAi (RNA interference) experiments, we investigated the role of Ctr2 in copper transport in African green monkey kidney COS-7 cells.
hCtr2 was expressed with the GFP (green fluorescent protein) fused to the C-terminus (hCtr2–GFP). The hCtr2 cDNA (GenBank® accession number U83461) was PCR-amplified from the pCMV-HA expression vector (BD Biosciences) containing the hCtr2 cDNA using the Advantage™ cDNA PCR kit and polymerase mixture (BD Biosciences). Primers (forward primer, 5′-AGCTTCGAATTCTGATGGCGATGCATTTCATCTTCTCAGATACAGCGGTG-3′; reverse primer, 5′-ACTGCAGAATTCGAGCTGTGCTGAGAAGTGGGTAAGCTAGGTAGTA-3′) were designed to produce the full-length hCtr2 cDNA with EcoRI restriction sites at the 5′ and 3′ ends. Following restriction digestion, the hCtr2 cDNA was subcloned into the pAcGFP-N1 expression vector (BD Biosciences) at the EcoRI site. The hCtr2 cDNA sequence was verified by DNA sequencing. GFP was expressed from the empty pAcGFP-N1 expression vector.
Generation of AbCtr263–78
The affinity-purified polyclonal antibody against hCtr2 was produced by Bethyl Laboratories. A peptide corresponding to amino acids 63–78 (S63QQTIAETDGDSAGSD78) of hCtr2 was generated with an additional N-terminal cysteine residue for conjugation to keyhole-limpet haemocyanin. The peptide was purified to > 90% by HPLC and verified by MS. The peptide preparation was injected into rabbits. Hyperimmune serum was immunoabsorbed to agarose-conjugated peptide to retrieve antibody activity directed against the peptide.
Western blotting and antibodies
All tissues, except placenta, were extracted from Zn-30 male rats fed on a copper-adequate diet, as described previously . Small intestine was obtained from ∼1 cm of intestinal segment starting 9 cm caudal to the pyloric sphincter. For preparation of intestinal mucosa, mucosal cells were gently scraped with a glass cover slide from a 10 cm intestinal segment starting just below the pyloric sphincter. Placenta was obtained from a pregnant female rat (at 21 days of gestation) fed on a copper-adequate diet. Rat tissues were homogenized with a Dounce homogenizer in 0.2% Triton X-100 containing a protease inhibitor cocktail (Roche Diagnostics). Cells were harvested by washing four times with cold PBS and then gently scraping the cells from the dish in PBS. Cells were spun down at 1000 g for 3 min at 4 °C and lysed with 0.2% Triton X-100. Supernatants of tissues and cell extracts were recovered following centrifugation at 13000 g for 10 min at 4 °C, and protein concentration was determined using the bicinchoninic acid method . Total protein extracts (20 μg) were mixed with SDS sample buffer [50 mM Tris/HCl (pH 6.8), 10% (v/v) glycerol, 2% (w/v) SDS, 144 mM 2-mercaptoethanol AND 0.008% Bromophenol Blue] and heated at 95 °C for 3 min prior to loading. Proteins were separated using 8–16% Tris/glycine gradient gels (Invitrogen) and electroblotted on to nitrocellulose membranes (Invitrogen). Membranes were blocked in TBS/Tween (20 mM Tris/HCl, 500 mM NaCl and 0.1% Tween 20, pH 7.5) supplemented with 5% (w/v) non-fat dried milk powder (Bio-Rad) at room temperature (20 °C) for 1 h. Membranes were probed with antibodies against Ctr2 (AbCtr263–78), CCS (copper chaperone for SOD1) (FL-274; Santa Cruz Biotechnology), actin [(I-19)-R; Santa Cruz Biotechnology], SOD1 (FL-154; Santa Cruz Biotechnology), Na+/K+-ATPase α (H-300; Santa Cruz Biotechnology) and GFP (JL-8; BD Biosciences) overnight at 4 °C at final concentrations of 2.5, 1, 0.5, 0.5, 0.5 and 0.5 μg/ml respectively in TBS/Tween supplemented with 0.5% non-fat dried milk powder. Membranes were washed with TBS/Tween and incubated with either a goat anti-rabbit or goat anti-mouse horseradish-peroxidase-conjugated secondary antibody (Bio-Rad) for 1 h at room temperature at a 1:5000 dilution. Blots were washed, and proteins were detected using ECL® (enhanced chemiluminescence) with ECL® Western blotting detection reagents (Amersham Biosciences) and exposure to film (Amersham Biosciences). Some membranes were stripped of antibodies by incubating 30 min at 55 °C in stripping buffer [62.5 mM Tris/HCl (pH 6.8), 2% (w/v) SDS and 100 mM 2-mercaptoethanol] and probed with a different antibody as described above. Some films were scanned, and band intensities were determined using Scion Image software (Scion Corporation).
COS-7 fibroblast cells were cultured in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% (v/v) fetal bovine serum, sodium pyruvate, penicillin and streptomycin (basal medium) at 37 °C in 5% CO2. BCS (bathocuproinedisulfonic acid) and metals were added directly to the medium from aqueous stock solutions.
COS-7 cells transfected to express hCtr2–GFP or untransfected cells were incubated with PBS or PBS supplemented with 2 mg/ml sulfo-NHS-SS-biotin [sulfosuccinimidyl-2-(biotinamido)ethyl-1,3-dithiopropionate] (Pierce) for 20 min at room temperature. Cells were washed four times with 50 mM glycine (in PBS) and then incubated for 5 min at room temperature with 50 mM glycine to ensure complete quenching of the biotinylation reaction. Cells were scraped from the dish and lysed in 20 mM Tris/HCl (pH 7.6) 150 mM NaCl and 1% (v/v) Triton X-100. Protein extracts were incubated with streptavidin–agarose beads (Sigma) for 4 h at 4 °C with gentle rocking. Streptavidin–agarose beads were washed four times with PBS and then resuspended in SDS sample buffer and heated for 5 min at 95 °C to recover biotinylated proteins.
Cells were grown to 70–80% confluence in 60-mm-diameter dishes. Cells were transfected with 4 μg of hCtr2-GFP or GFP expression vectors using Lipofectamine™ 2000 reagent (Invitrogen). Transfections were performed in basal medium without antibiotics for 18 h essentially as described in the manufacturer's protocol. Following transfection, transfection medium was replaced with specific medium depending on the assay.
siRNA (small interfering RNA) duplexes targeted to an N-terminal [sense, r(CAGUGGUGCUUCUGUUUGA)dTdT; antisense, r(UCAAACAGAAGCACCACUG)dTdA] or middle region [sense, r(GGUGGUAUUUGUGUCACUU)dTdT; antisense, r(AAGUGACACAAAUACCACC)dTdG] of the COS-7 Ctr2 mRNA coding sequence were designed by and purchased from Qiagen. Cells at ∼40% confluence were transfected with 2 μg of siRNA duplex using Lipofectamine™ 2000 reagent. The negative control siRNA duplex was obtained from Qiagen (catalogue no. 1027310).
QPCR (quantitative PCR)
Total RNA was isolated from cells using the RNeasy® kit (Qiagen). cDNA from each sample was generated using an oligo(dT) primer (Ambion). Sequences for the COS-7 Ctr2 and β-actin genes were obtained by PCR amplifying these genes from total RNA isolated from COS-7 cells using the OneStep RT (reverse transcription)-PCR kit (Qiagen) using primers specific for hCtr2 (GenBank® accession number U83461) (forward primer, 5′-ATGGCGATGCATTTCATCTTCTCA-3′; reverse primer, 5′-AGCTGTGCTGAGAAGTGGGTAAGC-3′) and Cercopithecus aethiops β-actin (GenBank® accession number AB004047) (forward primer, 5′-ATGGATGATGATATCGCCGCGCTC-3′; reverse primer, 5′-CTAGAACCATTTGCGGTGGACGAT-3′) genes respectively. PCR products were sequenced and aligned against the hCtr2 and Cercopithecus aethiops β-actin genes to verify amplification of the correct genes. Primers used for QPCR were designed using PrimerQuest (Integrated DNA Technologies). QPCR primer sets for COS-7 Ctr2 (forward primer, 5′-TCCTGGCTGTACTGTATGAAGGCA-3′; reverse primer, 5′-GGCCAAAGTGACACAAATACCACC-3′) and β-actin (forward primer, 5′-TGGGCATGGGTCAGAAGGATTCAT-3′; reverse primer, 5′-TGTGGTGCCAGATCTTCTCCATGT-3′) produced single products of expected sizes of 183 and 132 bp respectively. QPCR was performed using a Mx4000 Multiplex Quantitative PCR System (Stratagene). Reactions were performed in duplicate using the Brilliant SYBR Green QPCR core reagent kit (Stratagene). Post-amplification dissociation curves confirmed amplification of a single homogeneous product for each reaction. Relative mRNA levels were determined using the standard curve method. Ctr2 mRNA expression was normalized to β-actin expression.
hCtr2–GFP and GFP localization was assessed by visualizing GFP fluorescence in living cells. Cells were grown and transfected on coverslips. Following transfection, cells were incubated in basal medium for 24 h before visualization. For indirect immunofluorescence, cells were grown on coverslips in basal medium and fixed with 3% (w/v) paraformaldehyde in PBS for 30 min in the dark at 4 °C. Cells were then incubated for 10 min at room temperature with 0.1 M glycine in PBS (pH 8.0) and permeabilized by incubating 30 min at room temperature with 0.5% Triton X-100. Cells were blocked for 1 h with 5% (v/v) normal goat serum (Vector Laboratories) and incubated overnight at 4 °C with AbCtr263–78 at a final concentration of 20 μg/ml. Following washing with PBS, cells were incubated for 1 h at room temperature with an FITC-conjugated anti-rabbit secondary antibody (Sigma) at a 1:100 dilution in PBS in the dark. Coverslips were washed, mounted and visualized with an Axiophot fluorescence microscope (Carl Zeiss). Photographs were taken with an AxioCam camera (Carl Zeiss) and images were analysed using AxioVision software (Carl Zeiss). Results presented represent a minimum of three independent experiments and observation of at least 500 cells.
For copper-uptake assays, following transfection, cells were depleted of copper by incubating for 24 h in basal medium. For the time-dependence assay, basal medium was replaced with basal medium supplemented with 10 μM CuSO4, and cells were harvested after 0, 20, 40, 80, 160 and 320 min. For the concentration-dependence assay, basal medium was replaced with basal medium supplemented with 0, 2.5, 5, 10, 20 or 50 μM CuSO4, and cells were harvested after 300 min. For metal-competition experiments, basal medium was replaced with basal medium supplemented with 10 μM CuSO4 or basal medium supplemented with 10 μM CuSO4 and 100 μM concentrations of ZnSO4·7H2O, FeCl3, AgNO3 or MnCl2·4H2O. Cells were harvested after 300 min. Copper concentrations in the cell lysates were determined by graphite furnace atomic absorption spectrophotometry (PerkinElmer 5100 PC) and normalized to total protein.
Following transfection, cells were loaded with copper by incubating in basal medium supplemented with 10 μM CuSO4 for 24 h. Copper-loaded cells were washed three times with PBS and incubated in basal medium containing 100 μM BCS. Cells were harvested after 0, 20 and 160 min.
Semi-logarithmic plots of the copper-efflux data gave straight lines for cells loaded with 3100±150 pmol of copper/mg of protein (hCtr2–GFP-expressing cells) or 950±50 pmol of copper/mg of protein (GFP-expressing cells), suggesting that copper excretion could be modelled as a first-order decay process. The rate constant for copper efflux (kout = 1.6 × 10−3 min−1) was determined as the average of two values obtained from cells expressing hCtr2–GFP or GFP. Copper uptake was modelled using the copper-efflux constant (kout) and a constant rate of copper uptake (vin): (1)
In eqn (1), Cu(t) is the cellular copper content at any time and Cu(t0) is the cellular copper content at zero time. A constant rate of copper uptake was assumed because of the negligible decrease in external copper concentration over the experimental time course (a maximum of 1.5% of the external copper was accumulated by the cells after 320 min). Data were analysed by one of two methods. For the time-dependence assay (see Figure 5), the rate of copper uptake was estimated by non-linear regression fitting (SigmaPlot 9.0; Systat Software) of the cellular copper against time progress curves to eqn (1) using kout = 1.6 × 10−3 min−1. In order to calculate vin values for hCtr2–GFP- or GFP-expressing cells using cellular copper concentrations measured at 300 min (stopped time assays, see Figure 7A), eqn (1) was inverted to give vin as a function of the Cu(t) value (using kout = 1.6 × 10−3 min−1). Copper-uptake rates for GFP-expressing cells were subtracted from those measured in hCtr2–GFP-expressing cells to obtain net rates of copper uptake by hCtr2–GFP (see Figure 7B). These data were then fitted to either a rectangular hyperbola (Michaelis–Menten equation) or Hill equation in order to obtain estimates of the copper affinity of hCtr2–GFP (by non-linear regression, SigmaPlot 9.0).
Statistical analyses were performed using Statistica 7 software (StatSoft). Data were analysed using Student's t test or one-way ANOVA followed by Fisher's least significant difference test. Results are means±S.E.M. Differences were considered significant at P<0.05.
Expression of Ctr2 in rat tissues and COS-7 cells
The tissue expression pattern of Ctr2 in rats fed on a nutritionally complete diet containing adequate levels of copper was examined using a Ctr2-specific antibody. A band of ∼70 kDa in size was detected in all tissues examined in response to AbCtr263–78 (Figure 1A). A reproducible band of weaker intensity of ∼50 kDa was observed in heart. Ctr2 was highly expressed in placenta and heart and was expressed at low levels in brain. A 70 kDa band was also detected in COS-7 cells with AbCtr263–78 (Figure 1B). To verify that this band represents Ctr2, siRNA duplexes were used to knock down expression of Ctr2. siRNA duplexes are specific and potent inducers of gene silencing. Cells were transfected with siRNA duplexes targeted to an N-terminal (siRNA-1) or middle (siRNA-2) region of the COS-7 Ctr2 mRNA coding sequence or a negative control siRNA duplex. Cells transfected with siRNA-1 or siRNA-2 showed a marked reduction in the intensity of the 70 kDa band compared with control cells (Figure 1B), confirming that the 70 kDa band represents Ctr2. Consistent with a reduction in Ctr2 protein, QPCR analysis revealed a ∼60% reduction in Ctr2 mRNA expression in siRNA-1 and siRNA-2 cells compared with control cells (Figure 1C).
Transient expression and cellular localization of Ctr2
To begin to examine the role of Ctr2 in copper transport, hCtr2 was expressed in COS-7 cells as a GFP-fusion protein by transient transfection of an expression plasmid containing the full-length hCtr2 cDNA under control of the cytomegalovirus promoter. Western blot analysis of total protein extracts from cells transfected with the hCtr2–GFP expression vector using AbCtr263–78 revealed bands of ∼40 and 80 kDa in size (Figure 2). Under these conditions, bands were not observed in untransfected cells or cells transfected with the empty plasmid expressing GFP alone, indicating that the endogenous Ctr2 in COS-7 cells is expressed at significantly lower levels compared with transiently expressed hCtr2–GFP.
Direct GFP fluorescence was used to determine hCtr2–GFP localization in living COS-7 cells (Figure 3A). GFP was distributed throughout the cell (Figure 3A, panel a), as expected for a small protein lacking subcellular targeting motifs. Visualization of hCtr2–GFP-expressing cells under low magnification revealed intense staining at a small number of large cytoplasmic vesicles in some cells (Figure 3A, panel b, arrows). DIC (differential interference contrast) imaging and hCtr2–GFP fluorescence of a cell visualized at higher magnification demonstrated localization of hCtr2–GFP at the outer membrane of these large vesicles (Figure 3A, panels d and e). hCtr2–GFP was also localized to many smaller cytoplasmic vesicles (Figure 3A, panel c, arrow). A small proportion of hCtr2–GFP was localized to the plasma membrane, as was evident by staining at the cell periphery (Figure 3A, panel c, arrowhead).
Indirect immunofluorescence experiments using AbCtr263–78 revealed an identical staining pattern for endogenous Ctr2 in COS-7 cells (Figure 3B). Intense fluorescence in response to AbCtr263–78 was detected at the outer membrane of a few large cytoplasmic vesicles in some cells (Figure 3B, panel a, arrow). Staining at numerous smaller cytoplasmic vesicles was also observed (Figure 3B, panel b). Similarly to transiently expressed hCtr2–GFP, a small proportion of endogenous Ctr2 was localized to the plasma membrane (Figure 3B, panel b, arrowhead). All staining could be abolished by pre-incubating AbCtr263–78 with a 100-fold molar excess of the immunogen peptide (results not shown).
To investigate the cell-surface expression of Ctr2 in more detail, cells transfected to express hCtr2–GFP or untransfected cells were treated with the cell-impermeant biotinylation reagent sulfo-NHS-SS-biotin. Sulfo-NHS-SS-biotin reacts with the ϵ-amine of lysine residues and α-amine groups present on the N-termini of peptides to produce a stable product that can be recovered by precipitation with streptavidin–agarose beads. Both hCtr2–GFP (Figure 4A) and the endogenous Ctr2 in COS-7 cells (Figure 4B) were precipitated from protein extracts derived from cells treated with sulfo-NHS-SS-biotin, but not from extracts derived from control cells treated with PBS. The transmembrane protein Na+/K+-ATPase, which localizes to the plasma membrane , was also precipitated from extracts derived from cells treated with sulfo-NHS-SS-biotin, confirming biotinylation of cell-surface proteins. SOD1, an abundant cytosolic protein that does not localize to the cell surface, was not precipitated from extracts from cells treated with sulfo-NHS-SS-biotin, indicating that intracellular proteins were not biotinylated with sulfo-NHS-SS-biotin treatment. Notably, comparable proportions of hCtr2–GFP and endogenous Ctr2 were precipitated from the sulfo-NHS-SS-biotin-treated cells relative to the amount of Na+/K+-ATPase α (positive control) precipitated, suggesting that the proportions of hCtr2–GFP and endogenous Ctr2 present at the cell surface were similar. Together, these results demonstrate that both transiently expressed hCtr2–GFP and endogenous Ctr2 localize to the cell surface in COS-7 cells.
Copper uptake by Ctr2 in COS-7 cells
To examine whether Ctr2 plays a role in regulating cellular copper levels, copper accumulation in untransfected cells or cells transfected to express hCtr2–GFP or GFP was determined at various times (Figure 5). Transfected and untransfected cells were depleted of copper by incubating in basal medium (containing ∼0.2 μM copper) for 24 h. Basal medium was then replaced with basal medium supplemented with 10 μM CuSO4, and copper concentrations in the cells were determined after 0, 20, 40, 80, 160 and 320 min. Untransfected cells or cells transfected to express GFP showed little copper accumulation over 320 min. In contrast, copper concentrations in hCtr2–GFP-expressing cells increased substantially over 320 min and were > 12-fold higher compared with starting concentrations after 320 min. After 160 and 320 min, copper concentrations in hCtr2–GFP-expressing cells were > 3-fold higher compared with untransfected or GFP-expressing cells.
The marked accumulation of copper in cells expressing hCtr2–GFP suggested that Ctr2 plays a significant role in regulating copper concentrations in these cells. To determine whether the endogenous protein also functions in regulating copper concentrations, endogenous Ctr2 was depleted in these cells by RNAi. COS-7 cells were transfected with siRNA-1, siRNA-2 or the negative control siRNA duplex. After 30 h following transfection, transfection medium was replaced with basal medium supplemented with 10 μM CuSO4, and cells were incubated for 18 h before determination of cellular copper concentrations. Copper concentrations in cells transfected with siRNA-1 (718±56 pmol of copper/mg of protein) or siRNA-2 (702±38 pmol of copper/mg of protein) were ∼20% less (P<0.05; n = 6/group) compared with control cells (893±29 pmol of copper/mg of protein). QPCR analysis of cells treated in parallel revealed that Ctr2 mRNA levels were 33.1±2.4% (siRNA-1) and 25.8±4.0% (siRNA-2) of those of control cells. Ctr1 mRNA expression was similar in cells transfected with siRNA-1, siRNA-2 or the negative control siRNA duplex (results not shown). These data demonstrate a role for the endogenous Ctr2 in regulating copper levels in COS-7 cells.
To investigate in more detail the role of hCtr2 in copper transport, we examined whether expression of hCtr2–GFP influences copper efflux in COS-7 cells. Following transfection, hCtr2–GFP- and GFP-expressing cells were loaded with copper by incubating the cells for 24 h in basal medium supplemented with 10 μM CuSO4. The medium was then replaced with basal medium containing 100 μM BCS, a cell-impermeant copper chelator, and copper concentrations in the cells were determined after 0, 20 and 160 min. BCS was added to the medium to eliminate copper uptake over the course of the assay. As expected, at zero time, copper concentrations in hCtr2–GFP-expressing cells were higher compared with GFP-expressing cells (3100 ±150 pmol of copper/mg of protein compared with 950±50 pmol of copper/mg of protein). Despite differences in the starting level of copper, the percentage of copper excreted from the cells at 20 and 160 min was similar for hCtr2–GFP- and GFP-expressing cells, indicating that expression of hCtr2–GFP does not affect copper efflux in these cells (Figure 6).
Plots of cellular copper content against time were linear when plotted semi-logarithmically (results not shown), showing that copper efflux could be modelled as a simple first-order decay process: (2) Modelling of copper accumulation over time using eqn (1), which was developed to take both copper uptake and efflux into account, shows a close correspondence between measured and fitted data (adjusted r2=0.94) (Figure 5). Taken together, these results demonstrate that hCtr2–GFP promotes copper accumulation in COS-7 cells by stimulating copper uptake at the cell surface rather than through a copper-retention mechanism.
Copper uptake by hCtr2–GFP is concentration-dependent and saturable
To investigate the concentration-dependence of copper uptake by hCtr2–GFP, transfected cells were depleted of copper by incubating in basal medium for 24 h. Copper accumulation was then determined in cells cultured in basal medium supplemented with 0, 2.5, 5, 10, 20 or 50 μM CuSO4 (Figure 7A). Copper concentrations were not significantly different between hCtr2–GFP- and GFP-expressing cells cultured in basal medium supplemented with 0 or 2.5 μM copper, indicating that hCtr2–GFP does not import copper efficiently at low copper concentrations. Cells cultured in medium supplemented with 5, 10, 20 or 50 μM copper showed higher copper concentrations in hCtr2–GFP-expressing cells compared with GFP-expressing cells. Copper concentrations in GFP-expressing cells were higher when cultured in medium supplemented with 2.5 compared with 0 μM copper and similar when cultured in medium supplemented with 2.5, 5, 10 or 20 μM copper, suggesting the presence of an endogenous high-affinity copper-uptake system in COS-7 cells that is saturated at low copper concentrations (i.e. < 2.5 μM copper). Copper concentrations in hCtr2–GFP-expressing cells increased in a dose-dependent manner up to 10 μM copper. The dose-dependent increase indicates that the copper added to the medium remained in a form that was available for uptake and not sequestered by components present in the medium. Of note, the increase in copper concentrations in hCtr2–GFP-expressing cells at 50 μM copper compared with 10 or 20 μM copper probably reflects copper uptake by an endogenous low-affinity mechanism, given that a similar increase was observed for GFP-expressing cells.
At any time, the cellular copper content is a balance between copper uptake and efflux. Because copper efflux experiments indicated that the amount of copper excreted from cells was higher at higher internal copper concentrations (Figure 6), rates of copper uptake by hCtr2–GFP-expressing cells were determined using eqn (1), which takes into account copper efflux (see the Experimental section). Rates of copper uptake by hCtr2–GFP-expressing cells were fitted to either a rectangular hyperbola or Hill equation to estimate the affinity of hCtr2–GFP for copper (Figure 7B). The rectangular hyperbola gave Km and Vmax estimates of 11.0±2.5 μM and 10.2±0.9 pmol of copper/mg of protein per min respectively. The Hill equation gave estimates of K0.5, Vmax and h (Hill coefficient) of 6.9±0.7 μM, 8.1±0.4 pmol of copper/mg of protein per min and 1.9±0.3 respectively.
Specificity of copper uptake by hCtr2-GFP
To investigate the specificity of copper uptake by hCtr2–GFP, transfected cells were depleted of copper by incubating for 24 h in basal medium. Copper-uptake experiments were then performed in the presence of a 10-fold molar excess of zinc, iron, silver or manganese (Table 1). Copper concentrations in control hCtr2–GFP-expressing cells were ∼2.5-fold higher (P<0.05) compared with control GFP-expressing cells cultured in basal medium supplemented with 10 μM CuSO4. In GFP-expressing cells, addition of these metals had no effect on copper concentrations compared with control cells. Similarly, zinc, iron or manganese had no effect on copper uptake by hCtr2–GFP-expressing cells, as copper concentrations were similar to control cells. In contrast, silver completely inhibited copper uptake by hCtr2–GFP, as shown by similar copper concentrations in hCtr2–GFP- or GFP-expressing cells.
hCtr2–GFP expression down-regulates CCS protein in COS-7 cells
We next sought to determine whether expression of hCtr2–GFP increases a cellular copper pool that is available to cuproenzymes. CCS delivers copper to the antioxidant enzyme SOD1 . We and others have shown previously that copper deficiency increases CCS protein levels in rodents [26,27]. We demonstrated further that copper promotes the degradation of CCS by the 26S proteasome . Recently, a study by Caruano-Yzermans et al.  has suggested that degradation of CCS is stimulated by direct binding of copper to CCS. Therefore we rationalized that, if hCtr2–GFP contributes copper to a pool that is available for incorporation into SOD1, expression of hCtr2–GFP should down-regulate CCS protein levels. Consistent with our previous findings in H4IIE and HepG2 cells , CCS protein was higher in COS-7 cells cultured in copper-deficient compared with copper-rich medium (results not shown), indicating that copper regulates CCS expression in these cells. Following transfection, cells were cultured in basal medium supplemented with 10 μM CuSO4 for 24 h before measuring CCS expression. CCS protein level was lower in hCtr2–GFP-expressing cells compared with GFP-expressing cells (Figure 8). Analysis of copper content in extracts from these cells showed that CCS expression was inversely correlated with cellular copper concentrations (Figure 8). Quantification of CCS immunoreactive protein revealed that CCS expression in hCtr2–GFP-expressing cells was 43.7±5.1% compared with GFP-expressing cells. Taken together, these data suggest that Ctr2 participates in a process that contributes copper to CCS and, consequently, a copper pool that is available for incorporation into SOD1 in COS-7 cells.
Significant progress has been made in the understanding of the role of Ctr1 in copper transport in mammalian cells. However, currently, little information exists on the function of Ctr2. Ctr1 exists as a homotrimer . Recently, van den Berghe et al.  have shown that transiently expressed epitope-tagged versions of hCtr2 also form homomultimers. In the present study, we have shown that endogenous Ctr2 in rat tissues and COS-7 cells migrated as a predominant 70 kDa band, which is consistent with a multimeric form of Ctr2. Furthermore, transiently expressed hCtr2–GFP was detected as 40 and 80 kDa bands, which are close to the predicted molecular masses of an hCtr2–GFP monomer and homodimer respectively. At present, the reason for the difference in migration pattern observed for endogenous and overexpressed hCtr2-GFP is unclear. Notably, however, endogenous and overexpressed Ctr2 in yeast have also been shown to migrate differently by SDS/PAGE , suggesting that the expression level of Ctr2 may influence its migration pattern.
Co-localization studies have shown that transiently expressed hCtr2 localizes to late endosomes and lysosomes in mammalian cells . Localization of hCtr2–GFP to the outer membrane of large and distinctly smaller cytoplasmic vesicles is consistent with a late endosomal and lysosomal distribution for Ctr2 in COS-7 cells. Importantly, in the present study using a Ctr2-specific antibody, we established a similar vesicular distribution for the endogenous Ctr2 protein. Lysosomes are cellular organelles that function to degrade ingested products and organelles such as mitochondria. Therefore one function of Ctr2 in COS-7 cells may be to export copper from degraded products in lysosomes for reuse in the cytosol. Because lysosomes are analogous to yeast vacuoles, such a function for Ctr2 in mammalian cells is consistent with experiments in yeast indicating that Ctr2/Ctr6 mobilize copper from the vacuole to the cytosol for incorporation into cuproproteins [19–21].
Although Ctr2 showed a predominately intracellular vesicular distribution, results of the present study suggest that one function of Ctr2 in COS-7 cells is to promote copper uptake at the cell surface. Several lines of evidence support a role for Ctr2 as a cell-surface copper importer in COS-7 cells. First, both transiently expressed hCtr2–GFP and endogenous Ctr2 showed steady-state localization at the plasma membrane by fluorescence microscopy. Localization to the cell surface was confirmed by biotinylation of hCtr2–GFP and endogenous Ctr2 with the cell-impermeant reagent sulfo-NHS-SS-biotin. Secondly, copper-depleted cells overexpressing hCtr2–GFP hyperaccumulated copper when incubated in medium supplemented with CuSO4, whereas cells depleted of endogenous Ctr2 by siRNAs accumulated lower levels of copper. Thirdly, the percentage of copper excreted from cells over time was similar for cells expressing hCtr2–GFP or control cells. These data indicated that expression of hCtr2–GFP did not affect copper efflux from cells, suggesting that copper accumulation in hCtr2–GFP-expressing cells resulted from increased copper uptake rather than a copper-retention mechanism. Lastly, given that silver inhibits copper import at the plasma membrane by hCtr1 [9,12], inhibition of hCtr2–GFP-dependent copper accumulation in cells in the presence of silver is further evidence suggesting a role for Ctr2 in copper uptake at the cell surface.
In rat tissues, Ctr2 protein was ubiquitously expressed and showed the highest expression in placenta and heart. These results are in agreement with previously reported Ctr2 mRNA expression levels in human tissues showing high expression in placenta and heart . The higher expression level of Ctr2 in placenta and heart compared with that in other tissues examined suggests that these tissues may have a greater demand for Ctr2 copper-transport activity. The exact function of Ctr2 accounting for its high expression in placenta and heart awaits characterization of Ctr2's role in these tissues.
Knowledge of the copper-efflux kinetics concomitant with copper accumulation in cells allowed analyses of the copper-transport kinetics of hCtr2–GFP. The rate of copper uptake by hCtr2–GFP was concentration-dependent and saturable. Fitting copper-uptake data to a rectangular hyperbola or Hill equation gave estimates of the copper-binding affinity of 11.0±2.5 and 6.9 ±0.7 μM respectively. These estimates are significantly higher than the Km of 1.71±0.39 μM reported previously for hCtr1 , suggesting that hCtr2 transports copper with lower affinity than does hCtr1. In the absence of a detailed kinetic investigation, it is difficult to determine the exact mechanism of transport (rectangular hyperbola or co-operative). However, it should be noted that the sum-of-squares error term was much lower when a Hill equation (positive co-operativity) model was used to fit the data, suggesting that a simple relationship between copper binding and activity may not exist for this transporter.
The lack of inhibition of copper uptake by hCtr2–GFP in the presence of a 10-fold molar excess of zinc, iron or manganese indicated that hCtr2 is a metal-specific transporter. The complete inhibition of copper uptake by silver suggests that, in addition to copper, silver is also a substrate for hCtr2 transport. Since silver is isoelectric to Cu(I), these data suggest that reduced univalent copper is transported by hCtr2. These data are consistent with a recent study indicating that yeast Ctr2 functions in concert with a metalloreductase to mobilize copper from the vacuole to the cytosol . Although hCtr2 appears to have a lower affinity for copper compared with hCtr1, together these data suggest that Ctr2 transports Cu(I) similarly to Ctr1 , and Ctr2 and Ctr1 may transport copper by a similar mechanism. This notion is also supported by mutational studies demonstrating that conserved amino acid residues between Ctr1 and Ctr2 proteins are important for copper transport by Ctr1 and Ctr2 [20,31,33]. However, from our experiments, we cannot exclude the possibility that hCtr2 also transports bivalent copper.
It is well established that cells possess multiple mechanisms for copper uptake. Different cell types may have particular requirements for copper and thus use distinct copper-uptake mechanisms. Interestingly, copper levels are unaffected in kidney, liver and intestine of Ctr1+/− mice, but are decreased in spleen and brain [10,11]. Given the results of the present study suggesting that Ctr2 functions in copper import at the plasma membrane in COS-7 kidney cells, it is tempting to speculate that Ctr2 may compensate for the loss of Ctr1 copper-uptake activity in kidney of Ctr1+/− mice.
The lower affinity of Ctr2 for copper compared with that of Ctr1 suggested by the Km and K0.5 values determined in the present study and recent experiments by van den Berghe et al.  would suggest a predominant role for Ctr2 in copper transport when copper is present in relatively high concentrations, whether in intracellular vesicles or extracellular fluid. Given that normal serum copper concentrations in humans range from ∼10 to 25 μM and can decrease significantly in deficient states, Ctr2 may play a more significant role in copper import at the cell surface under normal or excess copper conditions. It has been demonstrated that copper stimulates endocytosis and degradation of a myc-tagged hCtr1 fusion protein in HEK-293 cells . Removal of Ctr1 from the cell surface and degradation of the protein when cells are exposed to elevated copper concentrations may be regulatory mechanisms to decrease copper import by Ctr1 to avoid copper toxicity. It would be interesting to determine whether Ctr2 replaces Ctr1 at the plasma membrane as a major copper importer under conditions of excess copper. If Ctr2 is present at low levels at the cell surface (as shown for COS-7 cells), it may be more advantageous for certain cell types to acquire copper through Ctr2 rather than Ctr1 when copper is in excess.
Once imported into cells, copper chaperones distribute copper to specific cuproenzymes and subcellular compartments . Since CCS protein levels were inversely correlated with cellular copper concentrations, reduced CCS expression in hCtr2–GFP-expressing cells may have resulted from higher cellular copper concentrations due to copper import by hCtr2–GFP at the plasma membrane. However, because hCtr2–GFP was also localized at the outer membrane of cytoplasmic vesicles, it is possible that mobilization of copper from these vesicles into the cytosol by hCtr2–GFP also contributed to increase the efficiency of copper transfer to CCS and consequently decrease CCS protein. Further research is required to determine whether Ctr2 delivers copper directly or indirectly to CCS.
In the present paper, we have shown evidence implicating Ctr2 as a specific copper-uptake protein at the cell surface that regulates copper levels in COS-7 cells. Since Ctr2/Ctr6 in yeast do not appear to function at the cell surface to import copper, copper uptake at the plasma membrane may be a unique property of Ctr2 in specialized cells of higher eukaryotes or mammals. It will next be important to explore whether Ctr2 has a similar role in other cell types and the interactions between Ctr2 and other copper-trafficking proteins, namely Ctr1. Nonetheless, our results provide novel insight into the mechanisms by which cells acquire copper.
We are grateful to Rudi Mueller for assistance with the fluorescence microscopy. This is publication number 613 of the Bureau of Nutritional Sciences.
Abbreviations: BCS, bathocuproinedisulfonic acid; (h)Ctr, (human) copper transporter; DIC, differential interference contrast; GFP, green fluorescent protein; HEK-293, human embryonic kidney; QPCR, quantitative PCR; RNAi, RNA interference; siRNA, small interfering RNA; SOD1, Cu/Zn superoxide dismutase; CCS, copper chaperone for SOD1; sulfo-NHS-SS-biotin, sulfosuccinimidyl-2-(biotinamido)ethyl-1,3-dithiopropionate
- © The Authors Journal compilation © 2008 Biochemical Society