Research article

Cadmium regulates copper homoeostasis by inhibiting the activity of Mac1, a transcriptional activator of the copper regulon, in Saccharomyces cerevisiae

Dong-Hyuk Heo, In-Joon Baek, Hyun-Jun Kang, Ji-Hyun Kim, Miwha Chang, Mi-Young Jeong, Tae-Hyoung Kim, Il-Dong Choi, Cheol-Won Yun


Cadmium is a toxic metal and the mechanism of its toxicity has been studied in various model systems from bacteria to mammals. We employed Saccharomyces cerevisiae as a model system to study cadmium toxicity at the molecular level because it has been used to identify the molecular mechanisms of toxicity found in higher organisms. cDNA microarray and Northern blot analyses revealed that cadmium salts inhibited the expression of genes related to copper metabolism. Western blotting, Northern blotting and chromatin immunoprecipitation experiments indicated that CTR1 expression was inhibited at the transcriptional level through direct inhibition of the Mac1 transcriptional activator. The decreased expression of CTR1 results in cellular copper deficiency and inhibition of Fet3 activity, which eventually impairs iron uptake. In this way, cadmium exhibits a negative effect on both iron and copper homoeostasis.

  • cadmium
  • copper
  • Ctr1
  • iron
  • Mac1
  • Saccharomyces cerevisiae


Physiological processes in virtually all organisms require metal ions as cofactors, and the cellular depletion of particular metals causes defects in numerous biological functions. However, excessive accumulation of certain metal ions generates ROS (reactive oxygen species) and ultimately leads to cell death [1,2]. Living organisms rely on precise mechanisms to maintain intracellular metal homoeostasis. Certain metal ions, such as cadmium, arsenate, lead and mercury, are harmful to organisms. Thus cells have unique mechanisms for protection from these deleterious metal ions, such as exporting them from the cell [3]. Cadmium is one of the better-characterized toxic metals that affect living organisms, and cadmium toxicity has been studied using various model systems from bacteria to humans. The sources of cadmium are numerous, and the diseases that are caused by cadmium are diverse and include gastroenteritis [4], Itai-Itai disease, osteoporosis [5], renal dysfunction [6] and bone pain [7]. Cadmium directly damages the cell membrane and induces apoptosis [8], mitochondrial dysfunction [9], cancer [10,11] and ultimately cell death. Interestingly, cadmium cannot directly produce ROS, but it can induce ROS formation by inhibiting the scavenging activities of superoxide dismutase and catalase [11]. Furthermore, cadmium induces mutations both directly by damaging DNA and indirectly by inhibiting the activity of repair systems [12], although the mechanism of this inhibition is not completely understood.

The mechanism of cadmium toxicity has recently been studied at the molecular level, and a link to sulfur metabolism was identified. Interestingly, sulfur compounds, such as methionine, cysteine and glutathione, and their complexes, are involved in the detoxification of cadmium [13,14]. For example, cDNA microarray data showed that cadmium up-regulates the expression levels of genes that are involved in sulfur metabolism in Saccharomyces cerevisiae [14]. In addition, cadmium inhibits the SCFMet30 pathway, preventing the degradation of Met4 [13]. Met4 is the primary transcription factor for genes involved in sulfur metabolism. When cells are exposed to cadmium, Met4 is activated and it up-regulates the biosynthesis of molecules that are involved in sulfur metabolism, such as glutathione. These molecules then scavenge the cadmium. These results suggest that sulfur metabolism is a key detoxification pathway in S. cerevisiae.

Another aspect of cadmium toxicity is the ability of the metal ion to arrest the cell cycle, specifically through the Rad53 pathway [15]. Cadmium activates the Mec1/Rad53 pathway and ribonucleotide reductase, increases the copy number of mtDNA (mitochondrial DNA) and ultimately induces cell death.

Cadmium competes with physiologically important metal ions in the uptake pathway, and this competition has lethal consequences for organisms. This process is one of the principal mechanisms of cadmium toxicity. Furthermore, the P-type ATPase Pca1 pumps cadmium out of the cell and into the extracellular environment and thereby contributes to cadmium tolerance [16].

In S. cerevisiae, the intracellular copper concentration is strictly regulated because copper is an essential metal for enzyme activity [17], defence against oxidative stress [18] and the electron transport system [19,20]. Copper is taken up from the environment by two copper transporters, Ctr1 and Ctr3, which are localized to the plasma membrane [21,22], and regulated by the copper-sensing transcription factor Mac1 [21]. Another copper transporter, Ctr2, is localized to the vacuole membrane and has low-affinity copper-binding activity [23]. Interestingly, CTR3 is not transcribed in laboratory S. cerevisiae strains due to insertion of transposable elements in the promoter region [23]. Thus Ctr1 is the only high-affinity copper transporter in S. cerevisiae. The copper taken up by transporters is delivered to target organelles by intracellular delivery systems, such as those initiated by Atx1, Ccs and Cox17 [1720] for copper delivery to the Golgi, the Sod1 protein and the mitochondria respectively.

The Mac1 protein [21] is a transcriptional activator that regulates the expression of the copper metabolism genes mentioned above. High concentrations of copper bind to Mac1 and inhibit its DNA-binding activity [24], thereby inhibiting the transcription of its target genes. Multiple cysteine-rich domains are found in the C-terminus of Mac1 [25], and these domains may confer copper-binding activity.

Interestingly, copper metabolism is regulated in a concerted manner with iron metabolism [26]. For example, copper can affect iron metabolism through the regulation of the ferroxidase Fet3 protein. The Fet3–Ftr1 complex comprises the reductive iron uptake system of S. cerevisiae, in which the ferrous form of iron is oxidized by the surface ferroxidase Fet3 and then taken up by the iron permease Ftr1. The enzymatic activity of Fet3–Ftr1 depends largely upon cellular copper utilization.

In the present study, we have identified a novel mechanism of cadmium toxicity in which cadmium inhibits Mac1 activity, which thereby reduces CTR1 transcription. In addition, cadmium interferes with iron utilization by inhibiting Fet3 activity, and this leads to changes in iron and copper homoeostasis in S. cerevisiae.


Strains and culture conditions

The strains used in the present study were WT (wild-type) BY4741 (MATa 9his3Δ1 leu2Δ0 met15Δ0 ura3Δ0), deletion mutants of BY4741 (Clontech) and mac1-1 [27]. These yeast strains were cultured in YPD medium (1% yeast extract, 2% peptone and 2% dextrose) or SD medium (0.67% yeast nitrogen base and 2% dextrose) supplemented with metabolites on the basis of the auxotrophic requirements of the stains. Media with defined iron concentrations were prepared as described previously [28]. For immunofluorescence and immunoblotting, the HA (haemagglutinin) epitope or GFP (green fluorescent protein) were fused to the coding sequences for Fet3, Ctr1 and Sod1 (Fet3–HA, Ctr1–GFP and Sod1–HA) by PCR on the basis of the gene modification method [29].

To construct the MAC1 WT strain and a mutant strain with a dominant constitutively active allele (MAC1-1up), plasmids containing WT MAC1 (pMAC1-WT) and MAC1-1up (pMAC1-1up) were obtained from Professor Dennis R. Winge (Department of Biochemistry, University of Utah Health Sciences Center, Salt Lake City, UT, U.S.A.). These plasmids contain the c-Myc epitope flanked by the AD (activation domain) and the DNA-binding domain. The Δmac1 cells were transformed with these plasmids, and the resulting strains were named pMAC1-WTmac1 and pMAC1-1upmac1.

To constitutively express CTR1, the pYPGE15 plasmid was obtained from Catherine Curie (Laboratoire de Biochimie et Physiologie Moléculaire des Plantes, Institut de Biologie Intégrative des Plantes, Montpellier, France). In this plasmid, the CTR1 open reading frame is flanked by the PGK1 promoter and the CYC1 terminator. The pYPGE15 empty vector and pYPGE15-CTR1 vector were transformed into the WT or Δctr1 strains, and the resulting transformants were designated the WT/EV, WT/CTR1, Δctr1/EV and Δctr1/CTR1 strains.

To measure the CTR1 promoter strength, the YEp24 plasmid containing the CTR1 promoter/LacZ fusion protein was obtained from Professor Dennis R. Winge. To subclone the CTR1 promoter/LacZ gene into pRS415, the Yep24-derived CTR1 promoter/LacZ was digested with SalI/NheI and introduced into the SalI/NheI-digested pRS415 plasmid (pLacZ). Furthermore, a LacZ gene without a CTR1 promoter was subcloned into pRS415 as a negative control. To construct the recombinant MAC1 DNA-binding domain/VP16 AD, the PGK1 promoter region was amplified by PCR and introduced into pRS316, and the resulting plasmid was named 316-PPGK1. The VP16 AD and the MAC1 DNA-binding domain truncations were amplified by PCR. The VP16 AD was introduced into the 316-PPGK1 plasmid and designated VP16 AD. The MAC1 full coding sequence and C-terminal truncations of MAC1 were subcloned into VP16 AD. These plasmids were called MAC1-WT, MAC1-101 and MAC1-125. These constructs (MAC1-WT, MAC1-125, MAC1-101 and VP16 AD) and the reporter constructs (pLacZ and LacZ) were co-transformed into the Δmac1 strain to exclude the endogenous Mac1p activity.

Plate assay and iron uptake assay

For screening, strains from the BY4741 deletion collection (Clontech) were grown on YPD containing 50 μM CdSO4. To confirm the screening results, WT cells and the Δaft1, Δfet3 and Δftr1 strains were grown on YPD with 0, 10, 50 or 100 μM CdSO4. To test growth recovery with exogenous iron, synthetic medium containing 100 μM BPS (bathophenanthroline disulfonate), various concentrations of ferrous ammonium sulfate (0, 10, 100 and 500 μM) as an iron source and CdSO4 (0 and 50 μM) were used.

The iron uptake assays were performed as described previously [30]. Briefly, the cells (WT, mutants and plasmid-harbouring strains) were grown to mid-exponential phase in the described conditions. The cells were washed two times with citrate buffer (50 mM sodium citrate and 5% glucose, pH 6.5). Prior to the assay, 1 μM 55FeCl2 was reduced with ascorbic acid. The cells were incubated with the reduced radioactive iron for 1 h at 30 °C and then at 4 °C to measure the passively imported iron. The internalized iron was quantified with a liquid scintillation counter (Beckman Coulter) in duplicate. This experiment was independently performed at least three times.

Ferric reductase assay

WT and Δfre1 cells were grown to mid-exponential phase in YPD medium and then were incubated for 2 h in the absence or presence of 50 μM CdSO4. The assay was performed in duplicate as described previously [30]. The amount of Fe2+ produced was estimated with a standard curve constructed from solutions of known Fe2+ ion concentrations. The ferric reductase assay was performed independently at least three times.

Immunofluorescence microscopy and immunoblot analysis

BY4741 WT and BY4741/Fet3–HA cells were grown in 10 ml of YPD to mid-log phase and incubated for 120 min in the absence or presence of 50 μM CdSO4. Formaldehyde (1 ml, 37%) was added to the medium, and the culture was incubated for 10 min in an orbital shaker at room temperature (25 °C). The cells were harvested by centrifugation and resuspended gently in 10 ml of Buffer A (40 mM potassium phosphate, pH 6.5, and 500 μM MgCl2) containing 1 ml of 37% formaldehyde to fix the cells. The cells were incubated for a further 1 h in an orbital shaker and then harvested. The cells were washed twice with Buffer A without formaldehyde. To make spheroplasts, the fixed cells were resuspended in 500 μl of Buffer B (1.2 M sorbitol and 40 mM potassium phosphate, pH 6.5) containing 10 μl of 100T zymolyase (2 mg/ml) and incubated for 15 min at 30 °C. Finally, the spheroplasts were suspended in 0.8–1.2 ml of Buffer B for washing.

The spheroplasts were attached to a glass slide covered with poly-L-lysine and incubated with blocking buffer (1% BSA and 0.5% Tween 20 in PBS, pH 7.4) for 15 min to permeabilize the spheroplasts and to block non-specific interactions with the antibodies. To detect the localization of Fet3p, an anti-HA monoclonal IgG antibody (1:500; Santa Cruz Biotechnology) and an Alexa Fluor® 568-conjugated rabbit anti-mouse IgG antibody (1:1000; Invitrogen) were used. The immunostained cells were examined using an HBO 100 microscope illuminating system (Zeiss).

For immunoblotting, the CTR1–GFP strain was grown in YPD with or without 50 μM CdSO4. Total proteins were extracted by the alkaline lysis method and resolved by SDS/PAGE (9 % gels). To detect the Ctr1 protein and Pgk1 as a loading control, anti-GFP (1:2000; Santa Cruz Biotechnology), anti-Pgk1 (1:5000; Santa Cruz Biotechnology), HRP (horseradish peroxidase)-conjugated anti-mouse IgG antibody (1:5000; Amersham) followed by enhanced chemiluminescence (Amersham) were used.

Ferroxidase assay

The ferroxidase assay was performed as described previously [31]. The WT and WT Fet3-HA, Δccc2 Fet3-HA, mac1-1 Fet3-HA and Δfet3 cells were grown in YPD medium with or without 50 μM CdSO4. To isolate the membrane-rich fraction, the cells were washed and homogenized in ice-cold buffer {150 mM NaCl, 25 mM Tris/HCl, pH 7.4, 1 mM AEBSF [4-(2-aminoethyl)benzenesulfonyl fluoride], 10 μM pepstatin A, 30 μM leupeptin and 1 mM PMSF} by vortexing with glass beads. After unbroken cells were removed, the lysates were centrifuged (16000 g, 4 °C) for 30 min. The pellets were washed with buffer and resuspended in buffer containing 1% Triton X-100. The insoluble particles were removed by centrifugation, and 30 μg of protein was separated by SDS/PAGE under non-reducing (oxidase assay) or reducing and denaturating (Western blot analysis) conditions. The oxidase assay was performed with 3 mM p-phenylenediamine dihydrochloride as a substrate. To detect the Fet3–HA, the anti-HA (Santa Cruz Biotechnology) and anti-mouse antibodies were used at dilutions of 1:2000 and 1:5000 respectively. The immunocomplexes were detected by enhanced chemiluminescence.

ChIP (chromatin immunoprecipitation)

WT BY4741 cells lacking an epitope tag and c-Myc-tagged MAC1-WT cells were grown to mid-exponential phase in 200 ml of SD medium lacking uracil. The cells were treated with CuSO4, CdSO4 and BCS (bathocuproine sulfonate) prior to fixation. The ChIP assay was performed as described previously [32], with some modifications. Each of our ChIP samples contained 2 mg of total cell lysate in 1 ml of lysis buffer, and 10 μl was used as the ‘input’. DNA from the input and immunoprecipitated samples were purified and resuspended in 50 μl of Tris/EDTA buffer, and PCR was performed in a total volume of 20 μl using 8 μl of purified DNA as the template. For the PCR cycles, an initial denaturation of 2 min at 95 °C was followed by 27 cycles of denaturation for 30 s at 95 °C, annealing for 30 s at 55 °C, and polymerization for 30 s at 72 °C and a final extension for 2 min at 72 °C. These conditions facilitated the amplification of the template without reaching the plateau phase. The PCR products were resolved by 1.5% agarose gel electrophoresis and visualized using ethidium bromide.

Microarrays and Northern blotting

Yeast cells were grown to mid-exponential phase in YPD medium and incubated for an additional 120 min in the presence or absence of 50 μM CdSO4. Total RNA was extracted using TRIzol® (Invitrogen) following the manufacturer's instructions. Total RNA was analysed using cDNA microarrays at Digital Genomics. Northern blotting was performed with probes against FET3, FET4, ARN1, CCC2, CTR1 and ACT1, and the primer sequences are shown in Table 1.

View this table:
Table 1 Names and sequences of the primers used in the present study

β-Galactosidase reporter assay for CTR1 expression

To measure the transcription activity of the CTR1 promoter, each transformant (MAC1-WT pLacZ, MAC1–125 pLacZ, MAC1–101 pLacZ, VP16AD pLacZ and MAC1-WT LacZ) was grown to early exponential phase and treated with the indicated metals (50 μM CuSO4, 20 μM CdSO4, 50 μM ZnSO4 and 50 μM CoCl2) and BCS (50 μM) for 2 h. The activity of β-galactosidase was quantified with ONPG (o-nitrophenyl-β-D-galactopyranoside) as a substrate, as described previously [33].

SOD (superoxide dismutase) activity assay

The WT, Δsod1 and Δsod2 strains were grown to early exponential phase and treated with the indicated metals (50 μM CdSO4 and 200 μM CuSO4) for 2 h. Each soluble extract (50 μg) was separated by 10% native gel electrophoresis at 4 °C and SDS/PAGE (15% gel) to detect Sod1–HA. The in-gel SOD assay was performed as previously described [34].


Cadmium up- and down-regulates genes involved in iron and copper metabolism respectively in S. cerevisiae

Heavy metals have lethal effects on living organisms, and we sought to determine the mechanism of cadmium toxicity using S. cerevisiae as a model system. First, we determined the genes involved in cadmium toxicity and performed a DNA microarray analysis using cells cultured in the presence of 50 μM cadmium for 120 min. We found that genes that are involved in detoxification were up-regulated by cadmium, which has been reported previously [14]. However, in this experiment, we found a novel effect of cadmium on iron and copper metabolism. As shown in Figure 1(A), cadmium up-regulated the expression of the iron regulon genes, including ARN1, FTR1, FET3 and CCC2 [35]. This result indicates that iron metabolism is affected by cadmium. Northern blot analysis was performed to determine whether cadmium perturbed the expression of the iron regulon genes FET3 and ARN1 [36]. The expression levels of FET3, FIT3, and ARN1 were enhanced in cadmium-treated cells (Figure 1B). In particular, the expression of CTH2 [37], which mediates the mRNA degradation of the genes involved in iron metabolism and is an indicator of iron depletion, was increased by cadmium. This strongly indicates that cadmium perturbs iron metabolism. Furthermore, as shown in Figure 1(A), cadmium down-regulates genes involved in copper ion uptake, including CTR1, FRE1 and FRE7 [38], but did not have a stimulatory effect on the expression of CUP1-1 compared with the expression level of ACE1 (Figure 1B). The CUP1-1 gene encodes the copper detoxifying protein Cup1. The expression of ACE1 was up-regulated, as shown by Northern blot and cDNA microarray analyses, and its mechanism of up-regulation should be further investigated. The cadmium-induced expression of the iron regulon genes may be a response to a cadmium-induced impairment in iron uptake. To test this possibility, we quantified the iron uptake activity in cadmium-treated cells (Figure 2A). Iron uptake was lower in cadmium-treated cells than in control cells, although FET3 expression was enhanced in the treated cultures (Figure 1B). The effect of cadmium may result from an effect on the Fet3/Ftr1 uptake system. To determine whether cadmium altered the Fet3 import system, the cellular localization of Fet3 was monitored in cadmium-treated cells. Fet3–HA was localized on the plasma membrane in cells treated with cadmium (Figure 2B), as in control cells. This result implies that there are other functions of cadmium in iron metabolism.

Figure 1 Cadmium up- and down-regulates genes involved in iron and copper metabolism respectively in S. cerevisiae

(A) A cDNA microarray analysis was performed to identify cadmium-regulated genes. Cells were cultured to mid-exponential phase and then incubated for 2 h in the presence of 50 μM cadmium. Total RNA was extracted from the cells and used in the cDNA microarray experiments. (B) Northern blotting was performed to identify the expression patterns of genes that are involved in the reductive iron transport system. WT cells were cultured to mid-exponential phase in SD broth, and the cells were cultured for 2 h in the presence of increasing concentrations of cadmium. Total RNA was extracted from the cells, and Northern blotting was performed for the indicated genes. ACT1 (actin) was used as a loading control.

Figure 2 Cadmium does not induce mislocalization of Fet3 in S. cerevisiae

(A) Iron uptake assays were performed with cells cultured in SD broth containing the indicated concentrations of cadmium. The cells were cultured to mid-exponential phase in cadmium-free medium and then were incubated for 2 h in the presence of increasing concentrations of cadmium. The cells were washed with cadmium-free buffer, and the iron uptake assay was then performed using reduced 55FeCl2. (B) The cellular localization of Fet3–HA was investigated. FET3 was tagged with HA in the genomic DNA of S. cerevisiae, and the tagged cells were cultured in SD medium to mid-exponential phase. They were then treated with 50 μM cadmium for 2 h. The cellular localization of Fet3–HA was investigated using fluorescence microscopy. The nuclei were stained with DAPI (4′,6-diamidino-2-phenylindole). DIC, differential interference contrast.

Cadmium down-regulates genes involved in copper metabolism

The microarray data involving copper metabolism were confirmed by Northern and/or Western blot analyses (Figure 3A). To confirm the expression of FRE1 in particular, the ferric reductase assay was performed with the WT and Δfre1 strains to investigate the effect of cadmium on ferric reductase activity. Cadmium decreased the ferric reductase activity, and the level of ferric reductase activity was decreased to the level seen in Δfre1 cells (Figure 3B). Taken together, these results imply that there is a relationship between cadmium and copper metabolism.

Figure 3 Cadmium down-regulates genes involved in copper metabolism

(A) Northern and Western blotting was performed against CTR1 and Ctr1–GFP respectively. The cells were cultured to mid-exponential phase and incubated for 2 h in the presence of 50 μM cadmium. Total RNA and protein were extracted. An internal region of the CTR1 open reading frame was used as a probe for Northern blotting. Pgk1 and ACT1 (actin) were used as loading controls in the Northern and Western blots respectively. (B) A ferric reductase assay was performed with the WT and Δfre1 strains. The cells were grown to mid-exponential phase and then were treated with or without cadmium. A ferric reductase assay was performed.

Copper reverses the cadmium-mediated defect of Fet3 activity

Because Fet3 requires copper, we evaluated the ability of copper supplementation to reverse the effects of cadmium on Fet3 activity. Interestingly, we found that copper rescues the Fet3-related cadmium effects using iron uptake studies (Figure 4A). As shown in Figure 4(A), 200 μM copper suppressed the defect in iron uptake induced by cadmium and the lower panel shows the expression pattern of the Fet3 protein in each culture. Interestingly, the expression of Fet3 was increased by cadmium treatment, and this result was supported by the Northern blot results (Figure 1A). It has been reported that Fet3 requires copper to be active, although copper does not affect Fet3 localization [39]. To further support the effect of cadmium on Fet3 activity, we tested the ferroxidase activity of Fet3. Ferroxidase activity was not detected in the Δccc2 and mac1-1 strains (Figure 4B). Interestingly, the Δccc2 and mac1-1 strains were defective in copper metabolism, which implies that cadmium may cause a copper deficiency and result in the loss of ferroxidase activity.

Figure 4 Exogenous copper complements the cadmium-induced inhibition of Fet3 activity

Iron uptake was assessed in cells treated with cadmium, copper or both metals (A). The cells were cultured as stated in the text, and then they were subjected to the iron uptake assay. The expression pattern of the Fet3 protein from each culture was confirmed by Western blot analysis (lower panel). (B) The ferroxidase activity of Fet3 was assayed and showed that cadmium inhibited Fet3 activity. Cadmium was used to treat WT yeast, the Δccc2 deletion mutant and the mac1-1 strain, and the membrane fractions were partially purified. The ferroxidase assay, as described in the Experimental section, and Western blot analysis were performed. Pma1 and Dpm1 were used as marker proteins for the plasma membrane and endoplasmic reticulum respectively.

Cadmium inhibits the transcription of CTR1

Cadmium treatment of yeast appeared to result in a copper deficiency. Yeast respond to copper deficiency by activating the Mac1 transcriptional activator to induce the expression of the high-affinity copper uptake system, which involves Ctr1 and Fre1 [38]. We investigated whether the Mac1 transcription factor was perturbed in the cadmium-treated cells. Cells containing WT Mac1 or the constitutively active MAC1-1up allele [40] were treated with cadmium sulfate. The expression levels of CTR1 and FRE1 in these cells were assessed using Northern blot analysis (Figure 5A) and quantified using the ImageJ program ( Expression of both CTR1 and FRE1 were inhibited in cadmium-treated cultures, and FRE1 expression was even slightly decreased. These inhibitory effects were more obvious in the MAC1–1up strain than in WT cells. Furthermore, we assessed the extent at which mRNA degradation may affect the CTR1 RNA levels using an mRNA decay assay. Cadmium did not alter the degradation of the CTR1 mRNA (Figure 5B). These results imply that cadmium is involved in the transcriptional regulation of CTR1 by Mac1.

Figure 5 Cadmium inhibits the binding of Mac1 to the CTR1 promoter and CTR1 transcription

(A) To confirm the effect of cadmium on copper metabolism, Northern blot analysis was performed using a CTR1 probe. Δmac1 cells, which were transformed with plasmids containing the MAC1 or MAC1-1up gene, were grown in medium containing copper (20 μM), BCS (50 μM) and cadmium (50 μM) and then were used for Northern blot analysis. ACT1 (actin) was used as a loading control. +, treatment with copper, cadmium or BCS; −, no treatment. The inhibitory effects of cadmium on CTR1 and FRE1 expression were quantified using the ImageJ program (lower panels). (B) mRNA decay experiments were performed to identify the effect of cadmium on CTR1 expression. The cells were grown in medium containing 50 μM cadmium, and thiolutin, an effective transcription inhibitor, was added to stop RNA synthesis. The cells were harvested at the indicated times. Total RNA was then extracted, and Northern blot analysis was performed for CTR1 and ACT. rRNA was used as a loading control. (C) A ChIP assay was performed to assess whether the metals affected the binding of Mac1 to the CTR1 promoter. The MAC1-Myc plasmid was transformed into the Δmac1 deletion and MAC1 WT cells, and the promoter region of CTR1 was detected using a probe specific for CTR1. CUP1-1 was used as a negative control because its expression is not regulated by Mac1. The left-hand panel is input, and the right-hand panel is output. The results of the ChIP assay are quantified in (D). Cadmium-treated cells showed the same levels as untagged cells.

The Mac1 transcriptional activator contains both a DNA-binding domain and a transcriptional AD. Copper binding to the transactivation domain subsequently impairs DNA binding through an intramolecular interaction between the two domains [41]. To test whether the Mac1 DNA-binding activity was impaired in cadmium-treated cultures, a ChIP assay was performed. Using this experiment, it can be determined whether or not the Mac1 protein binds to a target promoter. Under conditions of copper depletion induced by 100 μM BCS, Mac1 bound the CTR1 promoter, as previously shown (Figure 5C, lanes 2–4) [24]. However, Mac1 binding to the CTR1 promoter was not detected by PCR in cells treated with either cadmium or copper (Figure 5C, lanes 5 and 6). As shown in the left-hand panel of Figure 5(C), the Ctr1 input was the same, and the loading control CUP1 was not affected by Mac1. The results of the ChIP assay were quantified and are shown in Figure 5(D). These results suggest that cadmium inhibits the expression of Ctr1 by impairing the DNA-binding activity of Mac1.

Cadmium prevents Fet3 function by inhibiting copper metabolism

The effects of cadmium were tested in other yeast strains, including the Δatx1, Δccc2 and Δctr1 mutants. Atx1 and Ccc2 are involved in intracellular copper transfer to Fet3 [40]. Although the Δatx1 or Δccc2 mutants showed slightly lower iron uptake than WT, cadmium still affected these cells (Figure 6A). However, there were relatively minimal effects of cadmium in the Δctr1 strain, and iron uptake was extremely low. To confirm the effect of CTR1 on iron uptake, a constitutively activated CTR1 allele was transformed into WT cells and the Δctr1 strain. An iron uptake assay was performed to examine the effect of cadmium on these cells. Although both strains constitutively expressed CTR1 that was regulated by the PGK1 promoter, neither exhibited changes in the presence of cadmium, and cadmium did not reduce iron uptake (Figure 6B). Furthermore, the cadmium-induced growth defect was abrogated by the overexpression of CTR1 (Figure 6C), and these results support the involvement of Ctr1 in cadmium toxicity.

Figure 6 Exogenous CTR1 suppresses cadmium toxicity

To identify the effect of CTR1 on cadmium toxicity, an iron uptake assay was performed. (A) Iron uptake activity was measured in the Δatx1, Δccc2 and Δctr1 deletion strains. The deletion strains were grown in medium containing 50 μM cadmium, and iron uptake was then measured. (B) CTR1 with a PGK1 promoter was introduced into the Δctr1 deletion strain, and iron uptake was measured. The growth conditions were the same as in (A). (C) The growth defect seen during cadmium treatment was recovered by CTR1 overexpression. The cells were grown on SD medium for 2 days and then transferred into medium containing 50 μM cadmium alone or both 50 μM cadmium and 200 μM copper. The plates were incubated for 3 days at 30 °C.

We tried to determine the Mac1 binding affinity for the CTR1 promoter under cadmium stress using C-terminally truncated forms of Mac1 and a LacZ-fused Ctr1 promoter. As shown in Figure 7(A), the C-terminal truncated forms of Mac1 were constructed as previously described [33], and PCTR1-LacZ was provided by Professor Dennis R. Winge. The C-terminal truncations of Mac1 and PCTR1-LacZ were co-transformed into the Δmac1 strain, and β-galactosidase activity was then measured after treating cells with copper or cadmium for 2 h. Previously, it was reported that Mac1 has DNA-binding and copper-binding domains and that copper binds to the C-terminus of Mac1 and regulates its activity [33]. To identify whether the inhibitory mechanism of cadmium on Mac1 is similar to that of copper, we investigated CTR1 expression during cadmium treatment using plasmids that have C-terminal truncations of Mac1 (1–101, 1–125 and full-length Mac1) and PCTR1-LacZ. Full-length Mac1 inhibited CTR1 expression when cadmium or copper was added (Figure 7B). The C-terminal truncation of Mac1 (amino acids 1–125) failed to inhibit CTR1 expression in the presence of copper, as reported previously [33]. This can be explained by truncation of the cysteine-rich AD [25]. However, cadmium does inhibit the C-terminal truncation (amino acids 1–125) of Mac1, and these cells partially fail to activate the CTR1 promoter. These results indicate that cadmium may regulate Mac1 independently of the copper-dependent regulation mechanism. Recently, it was reported that Sod1 may affect Mac1 activity, but the detailed mechanism by which it does so has not been determined [34]. To identify the effect of Sod1 on Mac1 during cadmium treatment, we measured Sod1 activity during cadmium treatment. As shown in Figure 8(A), the activity and expression of Sod1 were not changed upon cadmium treatment, indicating that cadmium toxicity results from the direct binding of cadmium to Mac1.

Figure 7 Cadmium binds to the C-terminus of Mac1 and regulates CTR1 expression

The expression of CTR1 by Mac1 or by Mac1 with C-terminal truncations was measured with β-galactosidase using PCTR1-LacZ. (A) The C-terminal truncations of Mac1 were constructed as described in the Materials and methods section, and PCTR1-LacZ was provided by Professor Dennis R. Winge. (B) The Δmac1 strain, which was co-transformed with the C-terminal truncations of Mac1 (amino acids 1–101, 1–125, 1–159 and 1–194 and full-length Mac1) and PCTR1-LacZ, was treated with cadmium, and then β-galactosidase activity was measured. Mac1-WT plus LacZ and the VP16 AD plus PCTR1-LacZ served as the negative controls.

Figure 8 Sod1 is not involved in inhibition of Mac1 activity by cadmium

Sod1 activity was measured using an in-gel assay, and protein concentration was determined by Western blotting. (A) The indicated concentrations of cadmium or copper were added to the cells, and Sod1 activity was measured with an in-gel assay. Manganese SOD and copper SOD were both detected. (B) Sod1–HA was detected by Western blot with Pgk1 as a loading control.


Cadmium is known to be a toxic heavy metal, and the mechanism of its toxicity has been studied in many different model systems. Cadmium causes severe and potentially fatal diseases in humans. The toxic mechanism of cadmium has only been studied recently at the molecular level, and even low concentrations of the metal affect the functions of many proteins. Furthermore, the effects of cadmium on the metabolism of other metals have been reported previously, but the mechanisms through which cadmium acts remain unknown.

In S. cerevisiae, we have found novel effects of cadmium on iron metabolism. As shown in Figure 1(A), the expression of FET3 was up-regulated by cadmium, whereas iron uptake was reduced to the basal level (Figure 2A). It is still unclear why the iron uptake activity was decreased and the expression level of FET3 was increased. One possible explanation is that cadmium inhibits copper delivery to Fet3. As shown in Figure 4, the defects of iron uptake and growth by cadmium were recovered by addition of exogenous copper (Figure 4A). Furthermore, the ferroxidase activity of Fet3 was inhibited by cadmium treatment. These results strongly indicate that cadmium affects copper metabolism. In S. cerevisiae, cellular copper is taken up by two copper transporters, Ctr1 and Ctr3, which are localized to the plasma membrane [21,22]. However, intracellular copper is transported to target molecules by the copper chaperone Atx1 [17]. Ccc2 transfers copper to apo-Fet3, and Ccc2 is located in the post-Golgi or late-Golgi compartment [40].

We attempted to determine the step at which cadmium inhibits copper metabolism. The first possibility was that cadmium reduces the expression of the copper transporter. We therefore performed Northern blotting analysis to confirm the transcription level of CTR1 (Figure 5). Interestingly, cadmium decreased the transcription level of CTR1, and this result indicates that cadmium may function at the stage of transcription (Figure 5A). Furthermore, our chase experiment indicated that cadmium did not alter the degradation kinetics of the CTR1 mRNA. These results strongly suggest that cadmium inhibits the transcription of CTR1 (Figure 5C). These results were confirmed by introducing a constitutively overexpressed CTR1 gene into the WT or Δctr1 deletion strains (Figure 6). Both of these strains constitutively expressed CTR1, but cadmium did not inhibit iron uptake in either. The next question was whether Mac1, a transcriptional activator of the copper regulon, inhibited CTR1 expression. We found that Mac1 mediated the inhibitory effect of cadmium (Figure 5A); the metal bound to Mac1 and inhibited its ability to bind to the CTR1 promoter (Figure 5C). These results strongly suggest that cadmium inhibits CTR1 transcription and that Mac1 mediates this effect. Ultimately, cadmium affects both copper and iron homoeostasis via Mac1.

The next question involved the mechanism by which cadmium inhibits Mac1 activity. Recently, it was reported that Sod1 regulates Mac1 transcriptional activity [34]. One possibility was that cadmium affected Sod1 activity, which then inhibited Mac1 activity. To evaluate this hypothesis, we determined the involvement of Sod1 in Mac1 activity during cadmium treatment. As shown in Figure 8(A), enzyme activity or changes of Sod1 expression were not found when cadmium was added, which indicates that cadmium does not affect Sod1 activity. Next we determined whether cadmium bound to Mac1 directly. Mac1 has multiple metal-binding domains that can interact with copper or zinc [41]. Cadmium may also bind to one of these domains. If cadmium binds to these domains [44], which modulate the DNA-binding activity of Mac1, it would inhibit the binding of Mac1 to the CTR1 promoter. To support this hypothesis, we used MAC1 truncations and PCTR1-LacZ plasmids. As shown in Figure 7(B), cadmium affected Mac1 activity in the N-terminal region of Mac1 [33]. Mac1 contains a zinc-finger-like motif in its N-terminal region, and residues Cys-23 and His-25 are required for binding to the CTR1 promoter in vitro [42]. This zinc-finger-like motif is found in Ace1, which is a copper-dependent transcriptional activator. However, the copper-binding domain of Ace1 may regulate the DNA-binding activity directly, rather than via a zinc-finger-like motif [43]. On the basis of these results, we propose that cadmium inhibits Mac1 activity via a zinc-finger-like domain or another region that is separate from the DNA-binding domain involved in CTR1 expression and the iron uptake pathway.

hCTR1 (human CTR1) [44] is the homologue of yeast CTR1, and the function of hCTR1 is similar to that of yeast Ctr1. hCTR1 also confers cisplatin resistance, similarly to that reported for yeast Ctr1 [45,46], and the mechanism of hCTR1 gene expression induction by cisplatin and copper treatments has been studied. hCTR1 functions as a copper and cisplatin influx transporter, and the regulation of its expression is important for the activity of platinum-based anticancer reagents. Interestingly, a balance between the levels of the proteins hCTR1 and ATP7a or ATP7b is a key factor for cisplatin accumulation in cells. However, a transcriptional activator that is equivalent to the Mac1 protein in yeast has not been identified in humans, and the exact mechanism of hCTR1 regulation has not been reported. It is known that hCTR1 is not regulated at the transcriptional level, but rather by protein degradation in response to copper- and platinum-based anticancer reagents [46]. However, Sp1, which is a transcriptional activator, may function as a transcriptional activator of hCTR1 expression, and it responds to the cellular copper concentration. Investigating the relationship between cadmium and hCTR1 expression will provide numerous clues to copper metabolism and resistance to platinum-based anticancer reagents in humans. Our findings may help to explain the relationship between copper metabolism and the metabolism of other heavy metals in humans.

Furthermore, copper is involved in many human genetic diseases. There are two families of P1B-type ATPases that confer copper secretion or uptake, the ATP7a and ATP7b families, which are described above. Genetic mutations in these genes result in fatal genetic diseases, such as Menkes disease or Wilson's disease [47,48]. Patients suffering from these diseases either cannot acquire copper or accumulate high levels of copper respectively. Interestingly, the skin fibroblasts of patients with Menkes disease show greater resistance to cadmium [49], which results from a defect in the delivery of cadmium by ATP7a.


Dong-Hyuk Heo carried out most of the experiments. In-Joon Baek performed the experiments shown in Figure 1. Hyun-Jun Kang, Ji-Hyun Kim and Miwha Chang carried out the Northern and Western blotting experiments. Mi-Young Jeong, Tae-Hyoung Kim and Il-Dong Choi analysed results and carried out experiments. Cheol-Won Yun designed all the experiments and wrote the paper.


This work was supported by the Korea Research Foundation Grant funded by the Korean Government (MOEHRD, Basic Research Promotion Fund) [grant number KRF-2007-313-C00456].


We thank Professor Dennis R. Winge for experimental advice and a critical review of this manuscript prior to submission.

Abbreviations: AD, activation domain; BCS, bathocuproine sulfonate; ChIP, chromatin immunoprecipitation; GFP, green fluorescent protein; HA, haemagglutinin; hCTR1, human CTR1; HRP, horseradish peroxidise; ROS, reactive oxygen species; SOD, superoxide dismutase; WT, wild-type


View Abstract