OUP user menu

The Pmr1 protein, the major yeast Ca2+-ATPase in the Golgi, regulates intracellular levels of the cadmium ion

Cláudio Marcos Lauer Júnior, Diego Bonatto, Albanin Aparecida Mielniczki-Pereira, Ana Zilles Schuch, Johnny Ferraz Dias, Maria-Lúcia Yoneama, João Antonio Pêgas Henriques
DOI: http://dx.doi.org/10.1111/j.1574-6968.2008.01214.x 79-88 First published online: 1 August 2008

Abstract

Cadmium is a nonessential, highly toxic heavy metal that shows ionic properties similar to calcium. These ionic similarities imply that the cadmium ion, Cd2+, is a calcium ion, Ca2+, receptor-agonist, affecting the same biochemical pathways involved in Ca2+ homeostasis. In the yeast Saccharomyces cerevisiae, the PMC1 and PMR1 genes encode vacuolar and Golgi Ca2+-ATPases, respectively. The PMR1 protein product Pmr1p is involved in both Ca2+ and Mn2+ homeostasis. This study investigated the importance of Pmc1p and Pmr1p for Cd2+ cellular detoxification. Using the standard techniques of yeast molecular research and a multielemental procedure named particle-induced X-ray emission, Pmr1p was identified as a protein that directly participates in the detoxification of Cd2+, possibly through the secretory pathway. The results allow us to posit a model of Cd2+ detoxification where Pmr1p has a central role in cell survival in a Cd2+-rich environment.

Keywords
  • Saccharomyces cerevisiae
  • Pmr1p
  • secretory pathway
  • Ca2+/Mn2+-ATPases
  • cadmium
  • particle-induced X-ray emission

Introduction

Calcium is an essential nutrient that plays a key role in the transduction of external signals through the cytoplasm of eukaryotic cells. Fluctuations in cytosolic free calcium ion concentration ([Ca2+]c) directly elicit a cellular response by altering the functions of Ca2+-binding proteins and their targets (Anraku et al., 1991). A variety of stimuli can trigger the opening of Ca2+-specific channels in the plasma membrane or the endoplasmic reticulum (ER), causing massive Ca2+ influx and accumulation in the cytoplasm. After stimulation, basal [Ca2+]c levels are restored by Ca2+-ATPases and antiporters that transport Ca2+ from the cytoplasm through the plasma membrane to the extracellular environment as well as through several internal membranes into organelles capable of storing excess Ca2+. Cells use a wide array of ion transporters and sensory factors to regulate [Ca2+]c and produce appropriate responses to Ca2+ signals (Cunningham & Fink, 1994).

Currently, a relatively small number of Ca2+ transporters are known to maintain [Ca2+]c homeostasis in the yeast Saccharomyces cerevisiae, including: the vacuolar Ca2+-ATPase Pmc1p (Cunningham & Fink, 1994); the vacuolar Ca2+/H+ exchanger Vcx1p/Hum1p (Cunningham & Fink, 1996; Pozos et al., 1996); the ER Ca2+-ATPase Cod1p/Spf1p (Suzuki & Shimma, 1999; Cronin et al., 2000, 2002; Bonilla et al., 2002); and the Golgi Ca2+-ATPase Pmr1p (Rudolph et al., 1989; Antebi & Fink, 1992). These transporters respond to the calmodulin/calcineurin-signaling pathway and are controlled by the transcription factor Tcn1p/Crz1p, keeping the intracellular levels of Ca2+ between 50 and 200 nM, although the environmental concentrations of this ion may range from <1 to >100 mM (Eilam et al., 1985; Dunn et al., 1994; Batiza et al., 1996; Miseta et al., 1999; Yoshimoto et al., 2002).

Given that two of the four proteins responsible for yeast [Ca2+]c homeostasis are located in the vacuole, it is not surprising that this organelle serves as the major location for Ca2+ sequestration and normally contains more than 95% of the total cellular Ca2+ (Eilam et al., 1985; Dunn et al., 1994). Loss of the Golgi-associated Pmr1p increases sensitivity to high environmental Ca2+ levels when vacuolar Ca2+ transport is compromised, indicating that the Golgi complex also plays an important role in Ca2+ sequestration and expulsion from the cell (Tanida et al., 1995; Marchi et al., 1999; Miseta et al., 1999). Pmr1p was the first member of the secretory pathway Ca2+-ATPases (SPCAs) to be identified in eukaryotes (Sorin et al., 1997; Kathryn et al., 2004). In humans, mutations in the Pmr1-homologous protein SPCA1 cause the acantholytic skin condition Hailey–Hailey disease (HHD) (Hu et al., 2000; Sudbrak et al., 2000).

Heavy metal ions such as Al3+, Pb2+, Hg2+, and Cd2+ can interact metabolically with Ca2+-binding molecules. Cd2+ competes with Ca2+ in the nervous system and this interaction can impair cognitive development; this Cd2+–Ca2+ interaction can also produce osteodystrophies in the skeletal system (Goyer, 1997; Kazantzis, 2004). Furthermore, Cd2+ exposure can induce lung, kidney, prostate, and testicular cancers in murine models (Waalkes et al., 1992). Human epidemiological data suggest that Cd2+ causes tumors of the male reproductive system and induces respiratory tumors (Peters et al., 1986).

Despite the vast amount of physiological information about the effects of Cd2+ in biological systems, the cellular routes for uptake from the environment are not yet well understood. Cd2+ could be captured and internalized through Ca2+ channels of the cytoplasmic membranes (Beyersmann & Hechtenberg, 1997), because both Ca2+ and Cd2+ have similar ionic radii (Rainbow & Black, 2005). In this study, we examined how Cd2+ affects the phenotype of yeast strains proficient and deficient for the PMR1 and PMC1 genes. For this purpose, we used established yeast cytotoxicity assays to compare the Cd2+ sensitivity of these yeast strains. We also analyzed the intracellular concentration of Cd2+ using the multielemental technique called particle-induced X-ray emission (PIXE), which has been successfully used to quantify metals present in biological samples (Kern et al., 2005).

Materials and methods

Unless otherwise stated, all chemicals for yeast and bacterial cultures as well as for molecular biology procedures were purchased from Sigma-Aldrich Corp. (St Louis, MO), BD Biosciences Co. (San Jose, CA), and Invitrogen Corp. (Carlsbad, CA).

Strains, plasmids, growth conditions, and molecular biology procedures

The yeast S. cerevisiae strains used in this work are all isogenic derivatives of wild-type (WT) strain W303 (Table 1), and have been described previously (Cunningham & Fink, 1994). Yeast was routinely grown on synthetic complete (SynCo) medium (0.17 g L−1 of yeast nitrogen base without amino acids or ammonium sulfate, 65 g L−1 of ammonium sulfate, and 20 g L−1 of glucose with appropriate amino acids or/and nitrogen bases added at 20 mg L−1. For solid SynCo, the medium was supplemented with 20 g L−1 of bacto-agar). All yeast strains containing the plasmids YCpLac33 and pPMR1 were grown in SynCo minus uracil (SynCo-ura) medium.

View this table:
1

Saccharomyces cerevisiae and Escherichia coli strains used in this study

OrganismsStrainsGenotypesPhenotypesSources
S. cerevisiae
W303Mata ade2-1 his3-11 leu2-3,112 ura3-1 trp1-1 can1-100Wild-type (WT)Cunningham & Fink (1994)
W303TW303; YCpLac33WT; empty vectorThis study
K605Mata ade2-1 his3-11 leu2-3,112 ura3-1 trp1-1 can1-100 pmc1TRP1pmc1Δ mutantCunningham & Fink (1994)
K610Matαhis3-11,15 leu2-3,112 ade2-1 ura3-1 trp1-1 can1-100 pmr1TRP1pmr1Δ mutantCunningham & Fink (1994)
CML100K610; YCpLac33pmr1Δ mutant; empty vectorThis study
CML200K610; pPMR1Rescue pmr1Δ;PMR1This study
E. coliXL1-BluerecA1 endA1 gyrA96 thi-1 hsdR17 supE44 relA1 lac [F' proAB lacIqZ. M15 Tn10 (Tetr)]Invitrogen Corp.

Escherichia coli strain XL1-Blue (Table 1) was used as a recipient for cloning procedures and was grown in Luria–Bertani (LB) medium (10 g L−1 of tryptone, 5 g L−1 of yeast extract, and 10 g L−1 of NaCl). For selection and propagation of plasmid-containing bacterial cells, the LB medium was supplemented with ampicillin solution at a final concentration of 150 μg mL−1.

Techniques for yeast genetics were performed as described by Rose et al. (1990). Standard molecular procedures for yeast and E. coli transformation, plasmid isolation, and DNA manipulation were performed as described by Sambrook et al. (1989) and Ausubel et al. (1996). The bifunctional yeast–E. coli centromeric vector YCpLac33 (Gietz & Sugino, 1988) was used for yeast transformation and PMR1 cloning.

Yeast PMR1 gene cloning and analysis

The PMR1 gene was amplified by the PCR using phenol–chloroform purified genomic DNA from W303 yeast (Table 1) as a template. For PCR amplification, the proofreading Pfx enzyme was used in order to avoid the introduction of mutations into the PMR1 sequence. The following primers were used for PMR1 amplification: sense primer (5′-G GGTACC CCC CCT CAT CAC AAC GAC ATC GTG TTT CAT CT-3); and antisense primer (5′-GGGCCC CAA GCT TGG TCT CTT TGT AAA GTT GAG AAT-3′). The underlined bases in both the sense and the antisense primers indicate the restriction sites for the KpnI and XmaI enzymes, respectively. Once amplified, the product was digested with KpnI and XmaI, gel purified, and the resulting fragment was cloned into the plasmid YCpLac33. The PMR1-YCpLac33 plasmid was named pPMR1, and the presence of insert was confirmed by both double digestion of pPMR1 with KpnI and XmaI and by PCR amplification, followed by direct agarose gel analysis of PMR1 fragment.

Yeast Cd2+ sensitivity assay by drop test

The Cd2+ sensitivity of yeast strains was determined by the drop test standard technique. Briefly, yeast cultures in an early stationary phase of growth (1 × 108 cells mL−1) were obtained after 2 days of growth in SynCo or SynCo-ura media at 30 °C. The yeast cultures were serially diluted (1 : 10 at each step) in a sterile saline solution (0.9% of NaCl) and 10 μL of each suspension was plated on SynCo or SynCo-ura media supplemented or not with 0.02 mM of CdCl2. Negative controls were performed on media without CdCl2. Plates were photographed after 2 days of growth at 30 °C.

Yeast survival curves

Yeast cells from an early stationary phase (1 × 108 cells mL−1) were reinoculated in SynCo-ura medium containing different concentrations of CdCl2 (from 20 to 160 μM) at a final density of 5 × 106 cells mL−1. The yeast cultures were incubated for 24 h in a shaker (180 r.p.m.) at 30 °C. The yeast cells were harvested by centrifugation, washed twice with sterile ultrapure water, and diluted to a final concentration of 2 × 103 cells mL−1. Aliquots of 100 μL were plated in triplicate on Synco-ura medium and incubated at 30 °C for 3 days for colony counting. The data for survival curve were obtained from three independent measurements.

PIXE

For PIXE analysis, cells were cultured on plates for 7 days at 30 °C, then harvested by scraping and suspended in sterile double-distilled water. The yeast cells were centrifuged, washed twice with sterile ultrapure water, resuspended, and diluted to a final density of about 5 × 106 cells mL−1 in 200 mL of liquid SynCo-ura medium. CdCl2 was added to the medium at a final concentration of 100 μM and the cultures were incubated at 30 °C for 24 h. After incubation, the yeast cultures were harvested by centrifugation, washed three times with sterile ultrapure water, freeze-dried, crushed, and pressed into pellets (Viau et al., 2006). Two independent samples were prepared, and PIXE analyses were carried out on each one individually. The average weight of the pellets was 0.085 g.

The PIXE analysis was carried out at the 3 MV Tandetron accelerator facility at the Physics Institute of the Universidade Federal do Rio Grande do Sul, Brazil. All measurements were performed using a 2 MeV proton beam with an average current of 5 mA. The acquisition time for each sample was 10–20 min. The beam spot at the target position was about 9 mm2. The samples containing the yeast cells, the blank, and the calibration targets were placed in a target holder, which accommodates up to 10 specimens. Each sample was positioned in the proton beam by means of an electric-mechanical system. The characteristic X-rays induced by the proton beam were simultaneously detected by an HPGe detector from EG&G (GLP series, EG&G Ortec, CA), with an energy resolution of 180 eV at 5.9 keV, and a lithium-doped silicon detector (SLP series, EG&G Ortec), with an energy resolution of about 155 eV at 5.9 keV. The GUPIX code (Campbell et al., 2000) was used for data analysis. The standardization procedure was carried out using a bovine liver standard NIST (SRM-1577b) (Viau et al., 2006).

The PIXE spectrum for a yeast sample treated with CdCl2 is described in Fig. 1. The Kα line (c. 23.2 keV) of Cd was used in order to obtain the elemental concentration of this element in the W303T, CML100, and CML200 strains. The relationship among the elemental concentration C, the experimental X-ray yield Y (obtained from the Cd-peak area of the PIXE spectrum; Fig. 1), and the theoretical yield YT (which contains all physical parameters relevant to the process) follows the mathematical expression:

1

PIXE spectrum of a yeast sample. This spectrum clearly shows the presence of Cd X-ray peaks. Other elements such as Mn, Fe, Cu, Zn, and Mo were also detected. The most intense peaks of each element are Kα-lines and the less intense ones are Kβ-lines. This spectrum was only used to analyze Cd high-energy X-ray peaks. Therefore, the Ca X-ray peak is not present in this spectrum. Ep, energy of the proton.

C=Y/(Q×Ω×ɛ×t×YT)

where t is the X-ray transmission through any absorbers placed between the target and the detector's crystal, ɛ and Ω are the detector's efficiency and solid angle, respectively, and Q is the charge accumulated during the irradiation. For yeast cells, the stoichiometric ratio of intracellular metals was estimated considering the cell density retained in the target. The metal quantity was calculated from the PIXE results and expressed as ppm g−1 of cell dry weight. Other elements such as Mn, Fe, Cu, Zn, and Mo were also present in yeast samples (Fig. 1).

Results and discussion

Phenotypic analysis for Cd2+ sensitivity of S. cerevisiae strains proficient and deficient in vacuolar and Golgi Ca2+-ATPases

To analyze the sensitivity of yeast strains proficient and deficient in vacuolar and Golgi Ca2+-ATPases (Table 1), we used the drop test assay (Fig. 2a and b), which is currently used for an initial screening of specific yeast phenotypic characteristics, like Cd2+ toxicity. The results indicated that the yeast strain K610, which has the pmr1Δ mutation (see Table 1), had the highest sensitivity to CdCl2 at a final concentration of 0.02 mM (Fig. 2a). On the other hand, both strains W303 (WT; Table 1) and K605 (pmc1Δ; Table 1) showed the same tolerance to CdCl2 at 0.02 mM (Fig. 2a).

2

Sensitivity of Saccharomyces cerevisiae to chronic exposure to CdCl2 in solid medium. (a) Control (no CdCl2) and 0.02 mM CdCl2. The number of cells per milliliter used in the drop test varied from 108 to 103. Yeast strains: (1) W303, (2) K605, and (3) K610.(b) Control (no CdCl2) and 0.02 mM CdCl2. Yeast strains: (1) W303T, (2) CML200, and (3) CML100.

In order to confirm that the mutation pmr1Δ is responsible for Cd2+ sensitivity, we transformed the strain K610 with a centromeric plasmid carrying the complete yeast PMR1 gene (pPMR1) amplified by PCR (strain CML200). In addition, the W303 and K610 strains were also transformed with the empty vector YCpLac33 (W303T and CML100, respectively) to be used as a negative control. While the CML100 strain (harboring the empty vector YCpLac33) was highly sensitive to CdCl2 exposure, the CML200 strain (with a fully functional ‘rescue’ copy of PMR1) exhibited reduced sensitivity that was comparable with that observed in the WT strain W303T (Fig. 2b).

Both Pmc1p and Pmr1p belong to the yeast SPCA family, which are Mn2+- and Ca2+-transporting P-type ATPases located in the vacuole membrane and Golgi apparatus, respectively (Ton et al., 2002; Wuytack et al., 2003). These proteins are responsible for high-affinity transport of Mn2+ and Ca2+ into the Golgi, where these ions are sequestered from the cytoplasm (Andal et al., 2003). SPCA genes have been identified in diverse organisms, including S. cerevisiae, Caenorhabditis elegans, Drosophila melanogaster, and Homo sapiens. In S. cerevisiae, the Pmr1p has distinct biochemical characteristics, which clearly distinguish it from the well-studied mammalian sarcoendoplasmic reticulum (SERCA) and plasma membrane Ca2+-ATPases (PMCA) (Sorin et al., 1997; Wei et al., 1999). The yeast Pmr1p has received more attention in recent years because its mutated human homolog, hSPCA1, has been found to be responsible for HHD (Hu et al., 2000; Sudbrak et al., 2000).

To confirm the results of the drop test, we exposed the yeast strains W303T, CML100, and CML200 to increasing doses of Cd2+ for 24 h. The resulting survival curves clearly showed that all strains with a functional copy of the PMR1 gene were resistant to Cd2+ doses from 20 to 160 μM (Fig. 2). On the other hand, the strain harboring the pmr1Δ mutation exhibited sensitivity to the lowest dose of CdCl2 (Fig. 3). It should be noted that, despite the sensitivity of the pmr1Δ strain to CdCl2 relative to the WT strain, there is no dose–response effect to increased CdCl2 concentrations in the culture medium. This may be due to the activation of other mechanisms related to Ca2+ homeostasis when the intracellular levels of Cd2+ surpass a determined threshold.

3

Sensitivity of proficient and deficient PMR1 yeast strains to different concentrations of CdCl2. The names of the yeast strains used are included in the inset. The data represent an average of three independent experiments.

PIXE assay of intracellular Cd2+ concentration

The phenotypic results of Cd2+ toxicity in yeast strains proficient and deficient in Pmr1p (Figs 2 and 3) suggest that Cd2+ is not detoxified by the pmr1Δ mutant. To answer this question, we compared the levels of Cd2+ in the W303T, CML100, and CML200 strains using the PIXE method, which allowed us to correlate the intracellular quantity of Cd2+ with its toxicity. PIXE has been used conventionally to estimate metal content in organic and inorganic materials (Kozai et al., 2003; Przybylowicz et al., 2003).

The PIXE results indicate that the [Cd2+]c is approximately threefold higher in the pmr1Δ strain (4.37 ppm g−1 cell dry weight) than in both the W303T (1.62 ppm g−1 cell dry weight) and the pmr1Δ strain containing the plasmid pPMR1 (1.54 ppm g−1 cell dry weight; Fig. 4a). These data support our previous phenotypic results, indicating that the sensitivity of the pmr1Δ mutant is due to an elevated intracellular Cd2+ content compared with the W303T cells. One interesting aspect of this work is that the [Ca2+]c is lower in the CML100 strain (203 ppm g−1 cell dry weight) when compared with both W303T (317 ppm g−1 cell dry weight) and CML200 (274 ppm g−1 cell dry weight) strains (Fig. 4b) after exposure to Cd2+. Cd2+ being a Ca2+-agonist, it is possible that the elevated [Cd2+]c of the CML100 strain inhibits the uptake of extracellular Ca2+, resulting in diminished [Ca2+]c.

4

Cytoplasmic Cd2+ ([Cd2+]c; (a) and Ca2+ ([Ca2+]c; (b) content of yeast cells, proficient and deficient in Pmr1 protein, as analyzed by the PIXE technique. The data were obtained from two independent experiments and are presented as mean±SD.

Some studies have shown that the yeast pmr1Δ strain is unable to properly maintain normal Ca2+ and Mn2+ levels in compartments of the secretory pathway. This Ca2+ transport defect leads to an increased rate of Ca2+ uptake and accumulation by a mechanism that resembles the capacitative Ca2+ entry response in nonexcitable mammalian cells (Csutora et al., 1999; Locke et al., 2000). Moreover, it was demonstrated that the pmr1 mutant strain, when grown on culture media containing lower or normal levels of Ca2+, accumulates high quantities of this ion in the vacuole as a consequence of the elevated activity of Pmc1p, which explains the low Ca2+-sensitive phenotype of pmr1 strains (Halachmi & Eilam, 1996). In this sense, both Pmr1p and Pmc1p are positively controlled by the calmodulin and calcineurin complex in the presence of an elevated [Ca2+]c (Cunningham & Fink, 1996). Interestingly, a yeast pmc1pmr1 double mutant is inviable even at low Ca2+ concentrations (Cunningham & Fink, 1996), a condition that can be reverted by: (1) the action of calcineurin-inactivating immunosuppressive drugs FK506/FKBP-12 and cyclosporin A/cyclophilin A, (2) by null mutations in the calcineurin A or B subunit, or (3) by point mutations in calmodulin (Cunningham & Fink, 1994) that destroy its high-affinity Ca2+-binding sites (Geiser et al., 1991). The inactivation of both calmodulin and calcineurin induces the vacuolar Vcx1 protein, a low-affinity H+/Ca2+ exchanger that is related to Ca2+ homeostasis and signal transduction (Cunningham & Fink, 1996).

Pmr1p is the major route for eliminating toxic Mn2+; mutant cells lacking Pmr1p are extremely sensitive to Mn2+ toxicity and accumulate very high cytosolic Mn2+ levels (Lapinskas et al., 1995; Ton et al., 2002; Kellermayer, 2005). It was postulated that the excess Mn2+ is pumped into the Golgi by Pmr1p, and then exits the cell in secretory vesicles that merge with the plasma membrane and release the Mn2+ back into the extracellular environment (Culotta et al., 2005). In addition to its role in Mn2+ detoxification, the Golgi apparatus acts in the transport of Ca2+ into the exocrine fluid, a physiological process used in the mammary glands for milk synthesis. The Ca2+ concentration in rabbit milk can approach 100 mM (Shennan & Peaker, 2000), which is approximately a million times higher than the intracellular free Ca2+ level and 50 times higher than total Ca2+ in blood. Human breast milk also contains high levels of Ca2+, around 12 mM (Greger & Windhorst, 1996). The Ca2+ secretion needed for milk synthesis occurs synchronously with the secretion of casein and lactose, indicating that the transport occurs via Golgi-derived secretory vesicles (Neville & Peaker, 1979). Ca2+ is also released in millimolar concentrations, mainly by exocytosis from the stimulated pancreas to produce pancreatic juice (Teufel et al., 1979; Marteau & Gerolami, 1994). In light of the similar ionic properties of Ca2+ and Cd2+ (Simkiss & Taylor, 1995; Williams & da Silva, 1996; Rainbow, 1997) and the results described herein in yeast cells with mutated Pmr1p, it is not surprising that Golgi complex proteins associated with Ca2+ homeostasis would be implicated in Cd2+ detoxification.

A model for Cd2+ detoxification by Pmr1p in S. cerevisiae

Based on our results, we propose the following models for the control of intracellular Cd2+ levels by Pmr1p. When a WT cell faces a Ca2+- and Cd2+-containing environment (Fig. 5a), the cytoplasmic transmembrane Ca2+ transporters (e.g. Cch1p and Mid1p) take up Ca2+ and Cd2+. It is expected that both divalent ions should compete for the Ca2+ transporters. Supporting this fact, a study performed by Lansdown et al. (2001) using a murine model for skin wound healing showed that wounds receiving Cd2+ show a local increase in Zn2+ content and a low Ca2+ level as a consequence of Cd2+ interaction with transmembrane and carrier proteins.

5

Models proposed for Cd2+ detoxification mediated by Pmr1p in Saccharomyces cerevisiae in culture medium containing Ca2+ and Cd2+. In (a), the growth of a WT yeast cell in the presence of both Cd2+ and Ca2+ induces the uptake of these ions by known transmembrane proteins, especially Ca2+ transporters. It is also expected that both ions compete for Ca2+ transporters. Once inside the cell, Cd2+ and Ca2+ activate the calmodulin/calcineurin complex that: (i) controls the stress-associated transcription factor Tcn1p/Crz1p and (ii) induces the activity of Pmr1p and Pmc1p. Once the activity of Pmr1p is increased, the Cd2+ ions could be transported to the Golgi and excreted extracellularly by vesicles of the exocytic pathway. In (b), a pmr1Δ strain that grows in the presence of low divalent ion concentration increased the uptake of these ions. Despite the extracellular competition between Ca2+ and Cd2+, there is an accumulation of free Cd2+ in the cytoplasm (thick arrow) and a lower influx of Ca2+ (dotted arrow), diminishing the activity of the calmodulin and calcineurin complex and inducing a shift in the redox balance. In (c), a pmr1Δ strain that grows in the presence of a high divalent ion concentration has an elevated cytoplasmic concentration of Cd2+ (thick arrows) that blocks the function of the calmodulin/calcineurin complex. The high [Cd2+]c promotes a shutdown of glutathione and thiol-compounds synthesis, altering the redox homeostasis and leading to: (i) a stress response by the activation of Msn2p/Msn4p and Yap1p transcription factors and (ii) inhibition of PKA, resulting in cell cycle arrest and increased resistance to Cd2+ and Ca2+. For additional details about the models, please refer to the main text of the manuscript.

Once inside the cell, both Ca2+ and Cd2+ activate the calmodulin/calcineurin complex (Suzuki et al., 1985), which promotes the expression of genes related to stress tolerance and Ca2+ homeostasis (including those encoding for Pmr1p and Pmc1p; Fig. 5a) by means of the transcription factor Tcn1p/Crz1p (Kellermayer et al., 2003). By contrast, the level of Vcx1p is reduced by environmental Ca2+, and its activity is further repressed by calcineurin activation through a posttranslational mechanism (Fig. 5a; Yoshimoto et al., 2002). The transcription factor Tcn1p/Crz1p also controls the expression of the GPX2 gene, which codifies for a phospholipid hydroperoxide glutathione peroxidase that protects the cells against oxidative stress (Fig. 5a; Tsuzi et al., 2004), demonstrating that the control of Ca2+ homeostasis is linked to antioxidative protection. As observed, Gpx2p exerts a protective effect against Cd2+ in mammalian cells (Croute et al., 2005). An other biochemical process that is induced by Cd2+ is DNA mismatch repair, which repairs oxidative base damages and allows the cells to tolerate Cd2+ (Fig. 5a; McMurray & Tainer, 2003).

In its turn, yeast mutants lacking Pmr1p exhibit high rates of Ca2+ influx and elevation of [Ca2+]c due to stimulation of Cch1p and Mid1p (Locke et al., 2000) when the cells are grown on culture media containing a normal divalent ion concentration (Fig. 5b). This process resembles the capacitative Ca2+ entry mechanisms in animal cells where depletion of secretory Ca2+ pools promotes Ca2+ influx through the plasma membrane channels and refilling of the depleted organelles (Putney & McKay, 1999). In yeast, excessive activity of the vacuolar Ca2+ transporters Pmc1p and Vcx1p can compete with Pmr1p for Ca2+ and activates the capacitative Ca2+ entry-like mechanism (Locke et al., 2000). Therefore, the activity of vacuolar Ca2+ transporters must be balanced with Pmr1p activity to avoid depletion of secretory organelles and inefficient use of energy for Ca2+ sequestration in the vacuole, which is not observed in the pmr1Δ strain. This model is in agreement with our data, where the pmr1Δ strain is sensitive to low doses of Cd2+ (Fig. 3) and has a high [Cd2+]c (Fig. 4a), indicating an increase of divalent ion uptake from the extracellular environment. Moreover, the extracellular competition between Ca2+ and Cd2+ for Ca2+ receptors located at the cellular membrane is also a factor that influences the tolerance of the cell against Cd2+ (Fig. 5b). As expected, we observed a lower [Ca2+]c in the presence of Cd2+ (Fig. 4b) when the pmr1Δ cells were grown in the standard concentration of Ca2+ normally found in SynCo medium. This model also considers that the sensitivity of cells against Cd2+ is induced by a mild oxidative stress generated by glutathione depletion. Cd2+ can promptly react with two molecules of glutathione inside the cell, generating bis(glutathionato)cadmium [(GS)2Cd] that accumulates, and promotes a redox shift towards oxidation. Thus, without Pmr1p to capture the excess of intracellular Cd2+ by means of secretory vesicles, all biochemical systems become unstable (McMurray & Tainer, 2003). In particular, the DNA mismatch repair system (MMR; Fig. 5b) is strongly inhibited by Cd2+ (McMurray & Tainer, 2003), leading to the accumulation of oxidized nucleotide residues in both the mitochondrial and the nuclear genomes (Fig. 5b). Moreover, it was demonstrated that oxidative stress is able to diminish the activity of calcineurin (Ullrich et al., 2003). Ultimately, increased production of reactive oxygen species due to thiol depletion and inhibition of thiol-dependent redox systems (Dormer et al., 2000), associated with functional alterations in DNA repair systems and a diminished activity of calmodulin/calcineurin complex (Fig. 5b), results in an impaired genic response as a consequence of the low activity of Tcn1p/Crz1p, followed by the action of DNA nucleases and caspase-related proteases, culminating in cell cycle arrest and cellular death by apoptosis (Madeo et al., 1999, 2002).

We also considered an alternative model where a pmr1Δ strain can grow and tolerate the presence of a high extracellular concentration of Cd2+ and Ca2+ (Fig. 5c). We observed that increasing levels of Cd2+ do not lower the survival of the pmr1Δ strain as expected (Fig. 2), indicating that other Ca2+ and Cd2+-tolerance mechanisms are activated in this condition. It was observed that Pmc1p is strongly upregulated in a pmr1Δ strain, this condition being necessary for cellular proliferation in the presence of excessive extracellular Ca2+ (Matheos et al., 1997). In addition, the excessive Ca2+ also activates the vacuolar Vcx1p transporter, which accounts for Ca2+ tolerance in the Pmr1p-defective strain (Cunningham & Fink, 1996). Considering this information, it should be possible that a high extracellular concentration of Cd2+ and Ca2+ decreases the sensitivity of a pmr1Δ strain towards the opposite ion as a result of competition between both divalent ions at the cytoplasmic membrane (Fig. 5c). This phenomenon was already described for a yeast pmr1-deficient strain grown in the presence of Ca2+ and Mg2+ (Szigeti et al., 2005). On the other hand, an increase in the uptake of Cd2+ in the absence of Pmr1p induces a strong oxidative stress response that inactivates the calmodulin/calcineurin complex and stimulates the activity of Vcx1p (Fig. 5c). This same oxidative stress generated by Cd2+ and/or the osmotic stress caused by a high concentration of Ca2+ can induce other stress-related transcriptional activators, like the proteins Msn2/Msn4 and Yap1.

Msn2p and Msn4p activate transcription through stress-response elements and upregulate the expression of 180 genes in response to environmental stress (Martinez-Pastor et al., 1996). As with Tcn1p/Crz1p, Msn2p/Msn4p are regulated primarily through their subcellular localization. Under optimal conditions, both proteins are phosphorylated by protein kinase A (PKA) and are cytosolic. Upon inactivation of PKA, Msn2p/Msn4p rapidly accumulate in the nucleus, where they activate stress-induced gene transcription (Fig. 5c; Kafadar & Cyert, 2004). Another transcription factor that is activated by Cd2+-generating oxidative stress is Yap1p (Fig. 5c; Wemmie et al., 1994). The genes regulated by Yap1p include those that belong to the antioxidant protection network (Wemmie et al., 1994). Among the genes controlled by Yap1p is YCF1, which codes for the Ycf1 protein, an ATP-binding cassette (ABC) protein that can transport glutathione-conjugates to the vacuole (Li et al., 1997). It was described that mutations in this protein render yeast cells hypersensitive to Cd2+ at high doses compared with the concentrations used for the pmr1Δ strain (Perego & Howell, 1997). In this sense, the activation of Msn2/Msn4p and Yap1p under conditions of elevated Ca2+ and Cd2+ concentration should increase the resistance of the pmr1Δ strain against both divalent ions, a condition that was observed in our work (Fig. 2).

Considering the models obtained in this work about the functions of Pmr1p on Cd2+ tolerance, we are now performing additional biochemical experiments in order to refine the data collected in this work.

Acknowledgements

We thank Dr G.R. Fink (Whitehead Institute for Biomedical Research, Massachusetts Institute of Technology, Cambridge) for kindly providing yeast strains. This work was supported by Fundação de Amparo à Pesquisa do Estado do Rio Grande do Sul (FAPERGS), Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), and GENOTOX/ROYAL (Laboratório de Genotoxicidade/Instituto Royal – UFRGS).

Footnotes

  • Editor: Derek Jamieson

References

View Abstract