Negative Control of Heavy Metal Uptake by the Saccharomyces cerevisiae BSD2 Gene*

(Received for publication, January 30, 1997)

Xiu Fen Liu Dagger §, Frantisek Supek , Nathan Nelson par and Valeria Cizewski Culotta Dagger **

From the Dagger  Division of Toxicological Sciences, Department of Environmental Health Sciences, Johns Hopkins University School of Public Health, Baltimore, Maryland 21205, the  Department of Molecular and Cell Biology, Howard Hughes Medical Institute, University of California, Berkeley, California 94720-3202, and the par  Department of Biochemistry, Tel Aviv University, Tel Aviv 69978, Israel

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES


ABSTRACT

We have previously shown that mutations in the Saccharomyces cerevisiae BSD2 gene suppress oxidative damage in cells lacking superoxide dismutase and also lead to hyperaccumulation of copper ions. We demonstrate here that bsd2 mutant cells additionally accumulate high levels of cadmium and cobalt. By biochemical fractionation and immunofluorescence microscopy, BSD2 exhibited localization to the endoplasmic reticulum, suggesting that BSD2 acts at a distance to inhibit metal uptake from the growth medium. This BSD2 control of ion transport occurs independently of the CTR1 and FET4 metal transport systems. Genetic suppressor analysis revealed that hyperaccumulation of copper and cadmium in bsd2 mutants is mediated through SMF1, previously shown to encode a plasma membrane transporter for manganese. A nonsense mutation removing the carboxyl-terminal hydrophobic domain of SMF1 was found to mimic a smf1 gene deletion by eliminating the copper and cadmium toxicity of bsd2 mutants and also by precluding the bsd2 suppression of superoxide dismutase deficiency. However, inactivation of SMF1 did not eliminate the elevated cobalt levels in bsd2 mutants. Instead, this cobalt accumulation was found to be specifically mediated through the SMF1 homologue, SMF2. Hence, BSD2 prevents metal hyperaccumulation by exerting negative control over the SMF1 and SMF2 metal transport systems.


INTRODUCTION

Transition metals such as copper and manganese serve as essential cofactors for a variety of enzymatic reactions and play important structural and functional roles in cell metabolism. However, these same ions can be toxic when present at elevated levels. One mechanism of toxicity is believed to involve the metal-catalyzed generation of hydroxyl radicals in the Fenton reaction (1). Nonessential metals such as cadmium have no known biological function and are highly toxic at relatively low concentrations. To balance the stimulatory and inhibitory effects of essential ions and to counteract the toxicity of nonessential metals, all organisms possess homeostatic mechanisms that properly control the cellular accumulation, distribution, and detoxification of metals.

The bakers' yeast Saccharomyces cerevisiae provides an ideal system in which to study the factors controlling metal homeostasis. A number of heavy metal transport proteins have been identified in yeast that function in transporting and delivering the ions to cellular targets. For example, the plasma membrane protein CTR1 is required for high affinity copper uptake (2). Copper uptake is also controlled by the S. cerevisiae CTR3 gene, although this gene is inactivated in most laboratory strains of yeast through insertion of a transposable element (3). In addition to these pathways for the high affinity uptake of copper, a low affinity divalent metal transporter encoded by the FET4 gene has been identified in yeast (4). FET4 exhibits specificity for a subset of divalent metal ions including cobalt, cadmium, and iron (4). Another gene that directly participates in metal ion uptake is SMF1. SMF1 was originally identified as an extragenic suppressor of the mif1 mutation affecting mitochondrial protein processing (5). Subsequent studies indicated that SMF1 is a plasma membrane transporter for manganese. Overexpression of SMF1 was shown to increase manganese accumulation and facilitate a manganese-dependent step in mitochondrial protein processing (6). A close homologue to SMF1, known as SMF2, can also suppress the mitochondrial protein processing defect of mif1 mutants, suggesting that SMF1 and SMF2 may be functionally redundant (5).

An additional yeast gene that functions in heavy metal accumulation is BSD2. We originally isolated BSD2 (ypass OD efficiency) as a gene which, when inactivated by mutation, suppressed oxidative damage in yeast lacking the copper/zinc superoxide dismutase (SOD)1 (7). Yeast sod1Delta mutants exhibit a number of aerobic defects including oxygen-dependent auxotrophies for the amino acids methionine and lysine (8-11), and these markers of oxidative damage are completely suppressed by mutations in BSD2 (7). Additionally, bsd2 mutant strains exhibited increased sensitivity toward copper and cadmium toxicity and also accumulated elevated levels of copper (7). We therefore proposed that bsd2 mutations suppress oxidative damage by increasing the concentration of redox active metal ions that might neutralize reactive oxygen in lieu of a functional SOD (7).

To understand how BSD2 functions in metal metabolism, we have investigated the metal transporting properties of bsd2 mutants and have examined the intracellular localization of the BSD2 polypeptide. We provide evidence that BSD2 localizes to the endoplasmic reticulum (ER) and acts at a distance to control the uptake of divalent metal ions from the growth medium. The BSD2 protein exerts negative control over the distinct pathways of metal transport involving the SMF1 and SMF2 metal transport genes.


EXPERIMENTAL PROCEDURES

Yeast Strains and Media

Strains of S. cerevisiae used in this study are listed in Table I. The isogeneic strains EG103, EG133, VCSUP2, XL103Delta b (8), and KS107 (12) have been described, as have strains 1783 (13) and AA255 (14). The bsd2Delta strains XL101, XL110, and XL115 were obtained by deleting BSD2 gene in AA255 using the bsd2Delta ::LEU2 construct (7) and by using the bsd2Delta ::HIS3 plasmid with strains KS107 and AA255, respectively. The FET4 gene was replaced with LEU2 as described (4) in strains VCSUP2, AA255, and XL115 to create XL107, XL103, and XL104, respectively. Using a ctr1Delta ::URA3 construct (2), the CTR1 gene was deleted in VCSUP2 and XL107 to obtain strains XL108 and XL109. XLSBS1 represents strain XL110 containing the smf1 allele with a nonsense mutation at nucleotide +1558. The SMF1 gene was deleted in strains XL110, AA255, XL101, and XL103Delta b2 using the smf1Delta ::URA3 construct (6) to create strains XL111, XL112, XL113, and XL122. The smf2Delta ::HIS3 plasmid was used to delete SMF2 in XL101, AA255, and XL103Delta b2, creating XL116, XL117, and XL123.

Table I. Yeast strains used in this study

Strains are categorized into two major isogenic groups. Strains are categorized into two major isogenic groups.
Strain Genotype Source

1788 MATa/MATalpha leu2-3 112 his4 trp1-1 ura3-52 canIr 13
EG103 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ 8
EG133 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::URA3 sod2Delta ::TRP1 8
VCSUP2 MAT cdeu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::URA3 sod2Delta ::TRP1 bsd2-1 8
XL103Delta b2 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ bsd2Delta ::LEU2 7
KS107 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::TRP1 12
XL107 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::URA3 sod2Delta ::TRP1 bsd2-1 fet4Delta ::LEU2 This study
XL108 MATalpha leu3-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::URA3 sod2Delta ::TRP1 bsd2-1 ctr1Delta ::URA3 This study
XL109 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta URA3 sod2Delta ::TRP1 bsd2-1 fet4Delta ::LEU2 ctr1Delta URA3 This study
XL110 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::TRP1 bsd2Delta :HIS3 This study
XL111 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::TRP1 bsd2Delta :HIS3 smf1Delta ::URA3 This study
XL122 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ bsd2Delta ::LEU2 smf1Delta ::URA3 This study
XL123 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ bsd2Delta ::LEU2 smf2Delta ::HIS3 This study
XLSBS1 MATalpha leu2-3 112 his3Delta 1 trp-289a ura3-52 GAL+ sod1Delta ::TRP1 bsd2Delta :HIS3 smf1-1 This study
AA255 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 14
XL101 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 bsd2Delta ::LEU2 This study
XL103 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 fet4Delta ::LEU2 This study
XL104 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 bsd2Delta ::HIS3 fet4Delta ::LEU2 This study
XL112 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 smf1Delta ::URA3 This study
XL113 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 bsd2Delta ::LEU2 smf1Delta ::URA3 This study
XL115 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura-352 bsd2Delta ::HIS3 This study
XL116 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 bsd2Delta ::LEU2 smf2Delta ::HIS3 This study
XL117 MATalpha ade2 his3Delta 200 leu2-3 112 lys2Delta 201 ura3-52 smf2Delta ::HIS3 This study

Stocks of strains were maintained on standard yeast extract/peptone/dextrose (YPD) medium (15). All yeast transformations were carried out by electroporation (16). Transformation of strains containing sod1Delta mutations required initial cultivation in anaerobic chambers (BBL GasPAK (8)). Atomic absorption analyses of metal ion accumulation were conducted on cells grown to a mid-logarithmic stage (A600 of 1.2-1.7) in a synthetic dextrose (SD) medium (15). Prior to harvesting, cells were treated with specific concentrations of CdCl2 (7.5 µM), CuSO4 (50 µM), CoSO4 (50 µM), MnSO4 (100 µM), or FeCl3 (100 µM). Following metal treatment, 1.5-ml aliquots of cells were harvested, washed, and resuspended in distilled water and subjected to atomic absorption spectrophotometry as described earlier (7, 17).

Plasmids

Construction of the Bsd2-HA fusion plasmid pXL36 involved polymerase chain reaction (PCR) amplification of BSD2 sequences -360 to the stop codon using a downstream primer that changed the termination sequence CATAAG to an NdeI site, CATATG. The product was first ligated to the pCR II vector (Invitrogen), then mobilized by NdeI and SalI digestion, and inserted at these same sites in pVC36 (18, 19). This 2-µm LEU2 vector contains the S. cerevisiae CDC23 gene (20) fused at a carboxyl-terminal NdeI site to two copies of HA, a 9-residue epitope from the influenza viral hemagglutinin protein (21). This resulted in the replacement of all CDC23 upstream and coding sequences with BSD2. To obtain the CEN Bsd2-HA construct, pXL36 was digested with SacI and SalI and inserted at these same sites into pRS315 (22). The bsd2Delta ::HIS3 deletion construct was constructed in a manner similar to that described for the bsd2Delta ::LEU2 plasmid pDB2E (7), except plasmid pRS303 was used rather than pRS305 (22). The resultant bsd2Delta ::HIS3 construct resulted in the deletion of BSD2 sequences +203 to +1091 with respect to the start codon. Gene deletion was confirmed through Southern blot analysis as described previously (7). To construct the smf2Delta ::HIS3 plasmid, SMF2 sequences -65 to +1676 were amplified by PCR and engineered to contain flanking EcoRI and SphI sites. The PCR product was digested with EcoRI and SphI and inserted into these same sites of pGEM3Zf(+) (Promega). The construct was then digested with BamHI and BglII at SMF2 sequences +216 and +442, and a BamHI fragment containing S. cerevisiae HIS3 was then inserted at these sites. To delete the chromosomal SMF2, the resultant smf2Delta ::HIS3 plasmid was linearized by digestion with EcoRI and SphI and used to transform yeast cells. Deletion of SMF2 sequences +216 to +442 was verified by PCR.

Subcellular Fractionation and Immunofluorescence Microscopy

To separate membrane and soluble components of yeast cells, we utilized fractionation by high speed centrifugation (23). Strain EG103 transformed with pRS425 or with pXL36 was grown to an A600 of 1.0 in SD medium. A culture of 50 ml was harvested and washed in distilled cold water, and cells extracts were prepared in the presence or absence of 1% Triton X-100 as described previously (19). The resultant extracts were then clarified by centrifugation and subjected to 100,000 × g centrifugation; the same cell equivalents of supernatant and pellet were then analyzed by SDS-polyacrylamide gel electrophoresis and Western blot, as described (19). The first antibody consisted of a monoclonal anti-HA antibody from mouse (BABCO, kind gift of David Levin) diluted 1:10,000, and the secondary antibody consisted of sheep anti-mouse IgG conjugated to horseradish peroxidase. Detection employed the enhanced chemiluminescence kit (Amersham Corp.), according to manufacturer's specifications.

Sucrose gradient fractionation was conducted using strain AA255 transformed with pXL36 and grown in SD medium to an A600 of 0.5-0.75. Preparation and fractionation of cell extracts over a 18-54% sucrose gradient was carried out as described (14). Protein concentration measurements were obtained, and GDPase and NADPH cytochrome c reductase activities were monitored according to published procedures (14, 18, 19).

Indirect immunofluorescence microscopy utilized strain 1788 transformed with pXL36. Yeast cells were grown to an A600 of 1.0 in SD medium; cells were fixed with formaldehyde and treated with zymolyase and analyzed by antibody staining according to standard procedures (15, 19). Incubation with the first antibody (mouse anti-HA diluted 1:1000) proceeded for 2 h. Detection was carried out with a FITC-labeling kit (Boehringer Mannheim) and utilized a goat anti-mouse secondary antibody (diluted 1:500) coupled to fluorescein isothiocyanate (FITC). Nucleic acids were stained by DAPI (Sigma), and FITC and DAPI staining were monitored by fluorescence microscopy, as described (15).

Genetic Studies

To isolate cadmium-resistant suppressors of bsd2, eight single colonies of the bsd2Delta sod1Delta strain XL110 were individually plated onto YPD medium containing 30 µM CdCl2, and after 3 days of incubation at 30 °C, a large number of cadmium-resistant colonies were isolated. These colonies were tested for co-resistance to copper by plating onto SD medium containing 175 µM CuSO4 and also for reversal of the bsd2 suppression of SOD deficiency by plating on SD medium lacking lysine. The cadmium-resistant clones that showed growth on the copper plates and also lack of aerobic growth on lysine-deficient medium were isolated. Genetic analysis was carried out on three independently isolated "sbs" mutants (uppressor of d2). The mutants were crossed to a sod1Delta bsd2Delta strain of opposite mating type, and the heterozygous diploids were found to be aerobic lysine prototrophs (due to bsd2 suppression of sod1Delta ), demonstrating that the sbs mutations were recessive. The diploids were induced to sporulate, and segregation of the sbs mutation was scored by the ability to suppress the lysine prototrophy of the bsd2Delta sod1Delta cells. The noted 2:2 segregation of aerobic lysine auxotrophy demonstrated that the sbs suppression of bsd2Delta in each sbs isolate reflects a single nuclear gene mutation.

To clone the wild type allele of SBS1, a genomic library present on the LEU2 CEN vector P366 (kind gift of F. Spencer) was used to transform by electroporation the original bsd2 suppressor strain XLSBS1, and approximately 100,000 transformants were obtained by growth on sorbitol SD without leucine. Following 4 days of growth in an anaerobic chamber, the transformants were replica plated onto SD medium without lysine and leucine and were allowed to grow for 2 days in air, after which 97 lysine prototrophic colonies were identified and isolated. Approximately half of these were also found to be sensitive to growth on SD medium containing 25 µM CdCl2. The plasmids from 12 of these transformants were isolated and subjected to restriction analysis.


RESULTS

The Effect of bsd2 Mutations on Heavy Metal Transport

We previously reported that mutations in BSD2 caused yeast cells to accumulate elevated levels of copper and also rendered these cells hypersensitive toward copper and cadmium toxicity (7). It was unclear whether the cadmium toxicity was related to the higher accumulation of copper or was the direct product of elevated cadmium uptake. In the present studies, the total cellular level of cadmium was measured by atomic absorption spectrophotometry in a bsd2Delta null mutant and an isogenic wild type strain. As shown in Fig. 1, cadmium levels were increased in the bsd2 mutant within 5 min of treatment with the metal and continued to rise over a 40-min period, compared with the isogenic wild type strain.


Fig. 1. Cadmium accumulation in bsd2 mutant and wild type yeast. The indicated strains of yeast were treated with 7.5 µM CdCl2 for the designated time intervals and were subjected to measurements of cadmium accumulation using atomic absorption spectrophotometry. Values obtained represent the averages of two independent samples where error bars indicate range. Strains used: BSD wt, AA255; bsd2Delta , XL101.
[View Larger Version of this Image (11K GIF file)]

To study whether other metal ions were affected by the bsd2 mutation, we used atomic absorption spectrophotometry to measure the cellular levels of copper, cadmium, cobalt, manganese, and iron in wild type and bsd2 null mutant strains. As is the case with copper and cadmium, the bsd2 mutation caused a notable elevation in cobalt accumulation (nearly 4-fold; Table II). Manganese ion accumulation was slightly increased in the bsd2 mutant (1.5-fold; Table II), and there was no reproducible elevation in iron accumulation up to 100 µM treatment with the metal (not shown). Hence, bsd2 mutations cause the cell to accumulate a subset of divalent heavy metals.

Table II. Metal accumulation in bsd2 mutant and wild type yeast

Metal accumulation in nmol/109 cells following 30 min of metal treatment is as specified under "Experimental Procedures." Values reflect the averages of 2 independent trials with 4 samples. ±, maximal range of values obtained. Strains: wild type, AA255; bsd2 mutant, XL101. Metal accumulation in nmol/109 cells following 30 min of metal treatment is as specified under "Experimental Procedures." Values reflect the averages of 2 independent trials with 4 samples. ±, maximal range of values obtained. Strains: wild type, AA255; bsd2 mutant, XL101.
Metal Wild type bsd2 mutant

Cd2+ 0.49 (±0.12) 1.22 (±0.38)
Cu2+ 0.26 (±0.14) 1.28 (±0.25)
Co2+ 1.92 (±0.54) 7.45 (±0.45)
Mn2+ 4.23 (±0.62) 6.14 (±0.7)

Localization of the BSD2 Polypeptide

To help understand how BSD2 controls heavy metal accumulation, we studied the intracellular localization of the BSD2 polypeptide. Hints as to a possible cellular localization of BSD2 were obtained from examination of the primary amino acid sequence. As previously noted (7), the protein contains in the carboxyl terminus three hydrophobic regions that are characteristic of trans-membrane spanning domains. Additionally, probable signals for retention/recycling in the ER were noted. A typical signal for ER retention is a carboxyl-terminal dilysine motif (KKXX) preceded by a hydrophobic trans-membrane segment (24). BSD2 contains two such motifs in tandem (KKLEKKYL) adjacent to the carboxyl terminus of the polypeptide (7).

The localization of BSD2 was examined experimentally by epitope tagging the protein and using biochemical fractionation and fluorescence microscopy methods of immunodetection. Two copies of the hemagglutinin (HA) epitope (21) were fused in frame to BSD2 at the stop codon. This Bsd2-HA fusion present on either a low copy CEN or high copy 2-µm plasmid did not affect BSD2 function, as judged by the ability to fully complement the metal sensitivity conferred by a bsd2 mutation (data not shown). Total extracts from cells expressing this fusion protein were analyzed by Western blot. We noted that BSD2 may be a low abundant protein since we could not detect Bsd2-HA by Western analysis when the protein was expressed on the CEN vector, even though protein levels were sufficient for full complementation of the bsd2 mutation. Bsd2-HA was readily detected when produced from the 2-µm vector (Fig. 2A).


Fig. 2. Western blot analysis of BSD2. Extracts from strain EG103 expressing Bsd2-HA were analyzed by SDS-polyacrylamide gel electrophoresis and Western blot using as first antibody mouse anti-HA serum and as secondary antibody horseradish peroxidase-coupled IgG (anti-rabbit or mouse). A, whole extracts from cells transformed with either pXL36 (Bsd2-HA: +) or empty vector pRS425 (Bsd2-HA: -). Sizes of molecular mass standards run in parallel are indicated. Bsd2-HA migrates anomalously on SDS gels with an apparent molecular mass (52 kDa) somewhat larger than the anticipated size (41 kDa); this mobility is not altered by glycosylase treatment or by expressing Bsd-HA in sec53 (33) and dpm1 (34) glycosylation mutants. B, extracts from the Bsd2-HA (B) or pRS425 (Co) transformed cells were treated with 1% Triton X-100 where indicated (+Tx) and were subjected to centrifugation at 100,000 × g. The resultant pellets and soluble fractions (Spnt) were analyzed by Western blot. Arrow marks position of Bsd2-HA.
[View Larger Version of this Image (36K GIF file)]

To investigate whether BSD2 behaves as a membrane protein, we prepared membrane fractions from cells expressing the Bsd2-HA fusion protein and analyzed these by Western blot. As shown in Fig. 2B, Bsd2-HA fractionated to the pellet upon a 100,000 × g centrifugation. In contrast, this protein fractionated to the supernatant when extracts were treated with Triton X-100 which is known to solubilize membrane proteins (23). These data are consistent with our hydropathy analysis of the amino acid sequence, indicating that BSD2 is a membrane protein.

To study BSD2 localization biochemically, cells expressing BSD2-HA were fractionated on a sucrose gradient (14). Fractionation of the Golgi apparatus in the gradient was monitored by a GDPase assay, and NADPH cytochrome c reductase served as a marker for the ER. As previously reported (18, 19), a soluble form of both marker enzymes cofractionated with bulk protein at the gradient top, whereas most of the GDPase and NADPH cytochrome c reductase fractionated in the pattern characteristic of yeast Golgi and ER (Fig. 3). GDPase activity exhibited a peak in a middle segment of the gradient, and cytochrome c reductase activity appeared in later fractions (Fig. 3). The position of Bsd2-HA analyzed by Western blot completely superimposed the biochemical marker for ER (Fig. 3, top).


Fig. 3. Sucrose gradient fractionation of Bsd2-HA expressing cells. Extracts were prepared from strain AA255 transformed with pXL36 and were subjected to sucrose gradient fractionation. Fractions collected from the top (1) to the bottom (16) were analyzed for the presence of Bsd2-HA by Western blot (top). Fractions were also analyzed for total protein concentration in mg/ml (bottom; closed squares), GDPase activity in nanomoles of Pi liberated per 20 min (bottom; X), and for NADPH cytochrome c reductase activity in nanomoles of cytochrome c reduced per 5 min (bottom; open squares).
[View Larger Version of this Image (29K GIF file)]

The intracellular localization of BSD2 was investigated further by indirect immunofluorescence microscopy using a secondary antibody coupled to fluorescein isothiocyanate (FITC). By FITC labeling, Bsd2-HA exhibited a punctate staining pattern in the cytosol and a striking rim surrounding the nucleus (as defined by co-staining with DAPI, Fig. 4A). This fusion protein produced from the 2-µm vector did not exhibit the diffuse staining pattern expected for a protein mis-localized to the cytosol when overproduced. The distinct perinuclear staining observed with Bsd2-HA is considered a hallmark for the ER in yeast (25). These immunofluorescence studies together with our biochemical cofractionation studies strongly indicated that BSD2 is an ER membrane protein.


Fig. 4. Localization of Bsd2-HA by immunofluorescence microscopy. The diploid strain 1788 transformed with the Bsd-HA plasmid pXL36 (A) or control 1788 cells (B) were fixed with formaldehyde and probed with either anti-HA and a FITC-conjugated anti-rabbit antibody (FITC), or with DAPI for visualization of nucleic acids. Cells were analyzed by fluorescence microscopy at 1000 × magnification.
[View Larger Version of this Image (42K GIF file)]

BSD2 and the Plasma Membrane Transporters for Metal Ions

Our localization studies on BSD2 raised the question of how an ER protein could regulate the accumulation of heavy metals. One possibility is that BSD2 functions in the microsomal transport of metal ions. To test this, cadmium accumulation studies were conducted on isolated microsomes from wild type and bsd2 mutant yeast. However, we failed to observe any partitioning of cadmium between soluble and microsomal pools that was dependent upon BSD2 (not shown), indicating that this protein does not control metal trafficking in the ER. Alternatively, BSD2 may act indirectly to modulate the plasma membrane uptake of metal ions.

We investigated whether known metal transport systems in yeast are affected by the bsd2 mutation. S. cerevisiae FET4 encodes a low affinity divalent metal transporter for the uptake of cadmium, cobalt, and iron (4). To test whether the FET4 gene product was involved in bsd2-mediated metal uptake, the FET4 gene was deleted in a bsd2Delta mutant and an isogenic wild type strain, and the resultant mutants were tested for cadmium toxicity. As shown in Fig. 5A, the fet4Delta mutation did not reverse the cadmium sensitivity of the bsd2Delta strain. By atomic absorption spectrophotometry, the bsd2Delta fet4Delta mutant still accumulated elevated levels of cadmium (not shown). Furthermore, the fet4Delta mutation did not interfere with the ability of bsd2 to suppress sod1Delta -related defects. As seen in Fig. 5B, bsd2 mutations suppress the aerobic lysine auxotrophy of sod1Delta cells, and this suppression of oxygen toxicity was still seen in a sod1Delta bsd2Delta fet4Delta triple mutant. We also tested whether the copper accumulation in bsd2 mutants involves the high affinity copper transporter CTR1 (2, 26). We observed that the bsd2 mutation still conferred copper sensitivity and caused an increase in copper accumulation in strains lacking CTR1 (not shown). We additionally noted that loss of CTR1 did not preclude the suppression of SOD deficiency by bsd2 mutations, as a sod1Delta bsd2Delta ctr1Delta triple mutant still exhibited aerobic lysine prototrophy (Fig. 5B). Through a similar series of genetic studies, we were able to demonstrate that the bsd2 influx of metals is not dependent upon endocytosis, since the bsd2 mutation still conferred metal sensitivity in an end3 mutant (27) defective for endocytosis (not shown). Together, these results demonstrated that BSD2 does not work through the CTR1 and FET4 transport systems or through endocytosis to control metal ion uptake.


Fig. 5. Dependence of bsd2 mutations on metal ion transporters. The indicated yeast strains were plated onto minimal medium and aerobic growth monitored following 3 days of incubation at 30 °C. A, medium was supplemented where indicated with 30 µM CdCl2. Strains utilized are as follows: Delta b (bsd2Delta ), XL115; wild type, AA255; Delta b Delta f (bsd2Delta fet4Delta ), XL104; Delta f (fet4Delta ), XL103. B-C, medium was supplemented with lysine where indicated. Strains utilized are as follows: B, wild type, EG103; Delta s (sod1Delta ), EG133; Delta s Delta b Delta f (sod1Delta bsd2-1 fet4Delta ), XL107; Delta s Delta b (sod1Delta bsd2-1), VCSUP2; Delta s Delta b Delta f Delta c (sod1Delta bsd2-1 fet4Delta ctr1Delta ), XL109; Delta s Delta b Delta c (sod1Delta bsd2-1 ctr1Delta ), XL108; C, Delta s (sod1Delta ), KS107; Delta s Delta b (sod1Delta bsd2Delta ), XL110; Delta s Delta b Delta smf1 (sod1Delta bsd2Delta smf1Delta ), XL111.
[View Larger Version of this Image (58K GIF file)]

Isolation of Suppressor Mutations of bsd2

If BSD2 controls a particular metal transport system, then inactivation of this putative transporter should suppress the cadmium and copper sensitivity associated with bsd2 and may also suppress the ability of bsd2 mutations to overcome SOD deficiency. We therefore sought to isolate suppressors of the bsd2 mutation that would serve to identify candidates for the metal transport gene.

A sod1Delta bsd2Delta strain was plated onto cadmium containing medium, and cadmium-resistant suppressors were isolated. These mutants that exhibited suppression of bsd2 cadmium sensitivity were also tested for copper sensitivity and for the ability to suppress the aerobic lysine auxotrophy of sod1Delta bsd2Delta cells. One complementation group designated sbs1 (uppressor of d2) reversed all three of these phenotypes associated with loss of BSD2 and was chosen for further study. As seen in Fig. 6, the sbs1 mutation also suppressed the elevated cadmium accumulation associated with loss of BSD2. Genetic analysis demonstrated that sbs1 represents a recessive mutation in a single nuclear gene (see "Experimental Procedures").


Fig. 6. Cadmium accumulation in a suppressor strain of bsd2. The indicated strains grown in minimal medium were treated with 2.5 µM CdCl2 for the designated times and were analyzed for cadmium accumulation by atomic absorption spectrophotometry. Results shown represent the averages of two independent samples, and error bars represent range of values obtained. Strains utilized are as follows: BSD2 wt, KS107; bsd2Delta , XL110; bsd2Delta sbs1, XLSBS1.
[View Larger Version of this Image (12K GIF file)]

The wild type SBS1 gene was isolated by functional complementation. A genomic library was screened for genes that reversed the sbs1 suppression of bsd2. Library transformants that restored the aerobic lysine prototrophy and copper and cadmium sensitivity of a sod1Delta bsd2Delta sbs1 triple mutant were obtained. The corresponding plasmids from 12 such transformants were isolated, and restriction analysis indicated that they harbored overlapping fragments of the same region of genomic DNA. The common region was found to span the complete open reading frame of the S. cerevisiae SMF1 gene, previously shown to encode a plasma membrane transporter involved in manganese uptake (6). To confirm that our sbs1 suppressor contained a mutation in SMF1, we isolated and sequenced SMF1 from the original sbs1 suppressor strain. A single C to T transition mutation was noted at SMF1 nucleotide +1558 in the sbs1 mutant but not in the corresponding parental strain. This transition mutation changes a glutamine at amino acid 521 to a stop codon and is predicted to delete the carboxyl-terminal trans-membrane spanning domain of the SMF1 transporter (6).

To test whether inactivation of SMF1 was responsible for suppression of bsd2, we constructed bsd2Delta smf1Delta mutants. As shown in Fig. 5C, deletion of SMF1 completely abolished the ability of bsd2 to suppress SOD deficiency, as a sod1Delta bsd2Delta smf1Delta triple mutant did not grow aerobically in the absence of lysine. Furthermore, deletion of SMF1 reversed the cadmium and copper sensitivity associated with a bsd2 mutation (Fig. 7). Inactivation of SMF1 also dampened the elevated copper and cadmium accumulation associated with loss of BSD2 (not shown). Overall, the smf1Delta deletion and the original sbs1 suppressor mutation were identical in their abilities to suppress bsd2. Therefore, removal of the SMF1 carboxyl-terminal hydrophobic domain in the sbs1 suppressor strain inactivated SMF1.


Fig. 7. Suppression of bsd2 metal toxicity. The indicated strains were seeded at an A600 of 0.05 in minimal medium supplemented with the designated concentrations of CdCl2 (A) or CuCl2 (B), and total growth following a 24-h incubation period was measured turbidimetrically at an O.D. at 600 nm. Strains utilized are as follows: bsd2Delta , XL103Delta b; bsd2Delta smf1Delta , XL122; bsd2Delta smf2Delta , XL123. The symbols used are: black-square, bsd2Delta ; , bsd2Delta smf1Delta ; , bsd2Delta smf2Delta .
[View Larger Version of this Image (25K GIF file)]

In comparison to results obtained with SMF1, deletion of the SMF1 homologue, SMF2, had no significant effect on the copper or cadmium toxicity associated with loss of BSD2 (Fig. 7). Furthermore, a smf2Delta mutation did not interfere with the ability of bsd2 to suppress SOD deficiency, as a sod1Delta bsd2Delta smf2Delta triple mutant is still an aerobic lysine prototroph (not shown). These studies demonstrate that contrary to previous notions (5), SMF1 and SMF2 are not functionally redundant.

In addition to causing elevated copper and cadmium accumulation, bsd2 mutations result in the hyperaccumulation of cobalt ions (Table II and Fig. 8). We were surprised to find that this effect on cobalt was not suppressed by inactivation of SMF1. As seen in Fig. 8, a bsd2Delta smf1Delta double mutant accumulated the same high levels of cobalt observed with a bsd2Delta single mutant. In comparison, deletion of the SMF1 homologue, SMF2, completely reversed the elevated cobalt accumulation associated with the bsd2 mutation (Fig. 8), further confirming the notion that SMF1 and SMF2 operate through distinct metal transport pathways. Collectively, these studies demonstrate that a mutation in BSD2 results in the activation of at least two independent metal transport systems, the uptake of cadmium and copper by SMF1 and the uptake of cobalt ions by SMF2.


Fig. 8. Cobalt accumulation in bsd2, smf1, and smf2 mutant strains. The indicated strains of yeast were treated with 50 µM CoCl2 for 30 min, and total cobalt accumulation was measured by atomic absorption spectrophotometry. Strains utilized are as follows: wild type, AA255; bsd2Delta , XL101; bsd2Delta smf1Delta , XL113; bsd2Delta smf2Delta , XL116; smf1Delta , XL112; smf2Delta , XL117.
[View Larger Version of this Image (12K GIF file)]


DISCUSSION

Evidence presented herein and in the previous paper (7) demonstrate that S. cerevisiae BSD2 is a modulator of heavy metal uptake. First, bsd2 mutations were associated with a hypersensitivity toward copper and cadmium toxicity. Second, inactivation of BSD2 caused cells to substantially accumulate increased levels of copper, cadmium, and cobalt. However, BSD2 does not directly participate in plasma membrane transport of metals. The polypeptide contains no obvious metal binding sequences, and our immunofluorescence microscopy and biochemical analyses failed to show a plasma membrane localization for the protein. Instead, BSD2 exhibited localization to the ER. BSD2 contains a dilysine signaling motif at the carboxyl terminus of the polypeptide, as would be expected for a ER protein in yeast (24).

How might an ER localized protein participate in metal ion accumulation? Our studies indicate that BSD2 does not function in microsomal metal ion transport but rather acts at a distance to control the uptake of metals from the growth medium. Through genetic suppressor analysis, we demonstrated that one source of elevated metal accumulation in bsd2 mutants is SMF1, a plasma membrane transporter previously shown to participate in the uptake of manganese ions from the medium (6).

Although SMF1 was originally documented as a manganese transporter, evidence suggests that the protein also functions in the transport of other metals. First, smf1 mutants are reportedly sensitive to the metal chelator EGTA, and this sensitivity can be relieved by treating yeast cells with either copper or manganese ions (6). Moreover, we demonstrate here that SMF1 is necessary for the hyperaccumulation of copper and cadmium in bsd2 mutants. bsd2 mutants also exhibit a slight elevation in manganese ion accumulation, consistent with metal transport by SMF1. Therefore, SMF1 appears to operate on a range of divalent heavy metals. It is quite conceivable that under normal conditions, SMF1 contributes little to heavy metal accumulation, and the principal route of uptake for metals is through the major transport systems such as CTR1. In accordance with this model, a mutation in SMF1 has little impact on manganese uptake, but hyperactivation of this gene by expression on a multicopy plasmid causes cells to dramatically increase the rate of manganese accumulation (6). We propose that bsd2 mutations may have an analogous effect on SMF1, causing hyperactivation of this transporter such that SMF1 now becomes a major contributor to cellular metal accumulation.

Inactivation of SMF1 also abolishes the ability of bsd2 mutations to suppress oxidative damage in sod1Delta strains. Furthermore, overexpression of SMF1 in a BSD2 wild type strain mimics a bsd2 mutation and suppresses the oxygen toxicity associated with loss of SOD1.2 These findings genetically establish a connection between heavy metal accumulation by SMF1 and resistance to oxidative stress and corroborate our other reports on metal-mediated suppression of SOD deficiency (18, 28, 29). At this point, it is still unclear as to which metal suppresses oxygen toxicity in bsd2 mutants, as these strains accumulate elevated levels of two redox active metals, copper and manganese. Our preliminary studies indicate that both ions contribute to oxygen resistance, presumably through the neutralization of toxic oxygen radicals.

The SMF1 gene product is not the only transporter affected by BSD2. Mutations in bsd2 also cause an increase in cobalt accumulation, and this effect on cobalt is specifically dependent upon the SMF1 homologue, SMF2. The role of SMF2 in metal transport has not been previously investigated, but this protein was assumed to function analogously to SMF1 (5). Our studies on cobalt accumulation clearly demonstrate that SMF1 and SMF2 are not redundant metal transport systems. This is further supported by the observation that the bsd2 suppression of SOD deficiency is specifically dependent on SMF1 but not SMF2. As we have suggested for SMF1, we propose that the contribution of SMF2 to metal uptake is normally very minor. In fact, we show here that a deletion in SMF2 has no effect on cobalt uptake in strains containing a wild type BSD2 gene. Yet in strains lacking bsd2, SMF2-dependent transport is somehow up-regulated, and the contribution of SMF2 to cobalt accumulation becomes very great.

The remaining question regards the mechanism by which an apparent ER protein can impact on cell surface transporters for metals. Since metal transport by SMF1 and SMF2 is enhanced by a mutation in BSD2, we conclude that BSD2 normally acts to prevent the uncontrolled uptake of metals by these two transporters. As one possibility, BSD2 might serve as a sensor for intracellular metal pools. Through a signaling pathway, BSD2 may lead to SMF1 and SMF2 down-regulation, thereby preventing the hyperaccumulation of potentially toxic metals. Alternatively, the effect of BSD2 may be more direct. This protein may participate in the quality control mechanism functioning at the ER. This control system promotes the folding and assembly of newly synthesized proteins and prevents the export of improperly assembled proteins (30). As SMF1 and SMF2 pass through the ER, BSD2 could exert control over the stability of the proteins or may impact on their localization. Consistent with this hypothesis, Chang3 discovered that bsd2 mutations result in the mislocalization of a third ion transporter, a mutant variant of the cell surface proton ATPase PMA1 (Pma1-7) (31). This modified PMA1 is normally targeted to the vacuole (32) but accumulates in the plasma membrane as an active proton transporter when BSD2 is mutated.3 A protein sorting aberration of this type could explain the elevation in SMF1- and SMF2-mediated transport observed in bsd2 strains. As has been found with PMA1-7, bsd2 mutations may cause an increase in the level of SMF1 and SMF2 that associate with the cell surface. Although the precise mechanism by which BSD2 regulates SMF1, SMF2, and PMA1-7 is still unclear, these studies have demonstrated that ion transport can fall under many tiers of control. It is quite conceivable that other metal transport proteins are modulated by chaperone or effector molecules acting at a distance.


FOOTNOTES

*   This work was funded in part by the Johns Hopkins NIEHS center, by National Institutes of Health Grants GM 50016 and ES 05794 (awarded to V. C. C.), and by a grant from the Israel Science Foundation (to N. N.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§   Supported by NIEHS Training Grant ES 07141.
**   To whom correspondence should be addressed: The Johns Hopkins University, 615 N. Wolfe St., Rm. 7032, Baltimore, MD 21205. Tel.: 410-955-3029; Fax: 410-955-0116; E-mail: vculotta{at}phnet.sph.jhu.edu.
1   The abbreviations used are: SOD, superoxide dismutase; ER, endoplasmic reticulum; PCR, polymerase chain reaction; FITC, fluorescein isothiocyanate; HA, hemagglutinin; DAPI, 4,6-diamidino-2-phenylindole.
2   S. J. Lin and V. C. Culotta, unpublished data.
3   A. Chang, personal communication.

ACKNOWLEDGEMENTS

We are indebted to Dr. D. Eide for the fet4Delta deletion construct and for helpful discussions, to Dr. A. Dancis for the ctr1Delta deletion plasmid, to Dr. A. Chang for communicating unpublished findings, and to Dr. K. Slekar for critical reading of this work.


REFERENCES

  1. Halliwell, B. (1991) Am. J. Med. 91, 14S-22S [Medline] [Order article via Infotrieve]
  2. Dancis, A., Yuan, S., Haile, D., Askwith, C., Eide, D., Moehle, C., Kaplan, J., and Klausner, R. (1994) Cell 76, 393-402 [Medline] [Order article via Infotrieve]
  3. Knight, S., Labbe, S., Kwon, L. F., Kosman, D. J., and Thiele, D. J. (1996) Genes Dev. 10, 1917-1929 [Abstract]
  4. Dix, D. R., Bridgham, J. T., Broderius, M. A., Byersdorfer, C. A., and Eide, D. J. (1994) J. Biol. Chem. 269, 26092-26099 [Abstract/Free Full Text]
  5. West, A. H., Clark, D. J., Martin, J., Neupert, W., Hartl, F.-U., and Horwich, A. L. (1992) J. Biol. Chem. 267, 24625-24633 [Abstract/Free Full Text]
  6. Supek, F., Supekova, L., Nelson, H., and Nelson, N. (1996) Proc. Natl. Acad. Sci. U. S. A. 93, 5105-5110 [Abstract/Free Full Text]
  7. Liu, X. F., and Culotta, V. C. (1994) Mol. Cell. Biol. 14, 7037-7045 [Abstract]
  8. Liu, X. F., Elashvili, I., Gralla, E. B., Valentine, J. S., Lapinskas, P., and Culotta, V. C. (1992) J. Biol. Chem. 267, 18298-18302 [Abstract/Free Full Text]
  9. Gralla, E. B., and Kosman, D. J. (1992) Adv. Genet. 30, 251-319 [Medline] [Order article via Infotrieve]
  10. Bilinski, T., Krawiec, Z., Liczmanski, L., and Litwinska, J. (1985) Biochem. Biophys. Res. Commun. 130, 533-539 [Medline] [Order article via Infotrieve]
  11. Chang, E., and Kosman, D. (1990) J. Bacteriol. 172, 1840-1845 [Medline] [Order article via Infotrieve]
  12. Culotta, V. C., Joh, H.-D., Lin, S.-J., Slekar, K. H., and Strain, J. (1995) J. Biol. Chem. 270, 29991-29997 [Abstract/Free Full Text]
  13. Lee, K. S., Patton, J. L., Fido, M., Hines, L. K., Kohlwein, S. D., Paltauf, F., Henry, S. A., and Levin, D. E. (1994) J. Biol. Chem. 269, 19725-19730 [Abstract/Free Full Text]
  14. Antebi, A., and Fink, G. R. (1992) Mol. Biol. Cell 3, 633-654 [Abstract]
  15. Rose, M. D., Winston, F., and Hieter, P. (1990) Methods in Yeast Genetics: A Laboratory Course Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY
  16. Becker, D., and Guarente, L. (1991) Methods Enzymol. 194, 182-187 [Medline] [Order article via Infotrieve]
  17. Strain, J., and Culotta, V. C. (1996) Mol. Gen. Genet. 251, 139-145 [CrossRef][Medline] [Order article via Infotrieve]
  18. Lin, S. J., and Culotta, V. C. (1996) Mol. Cell. Biol. 16, 6303-6312 [Abstract]
  19. Lapinskas, P. J., Lin, S. J., and Culotta, V. C. (1996) Mol. Microbiol. 21, 519-528 [CrossRef][Medline] [Order article via Infotrieve]
  20. Lamb, J. R., Michaud, W. A., Sikorski, R. S., and Hieter, P. A. (1994) EMBO J. 13, 4321-4328 [Abstract]
  21. Wilson, I. A., Niman, N. H., Houghten, R. A., Cherenson, A. R., Connolly, M. L., and Lerner, R. A. (1984) Cell 37, 767-778 [Medline] [Order article via Infotrieve]
  22. Sikorski, R. S., and Hieter, P. (1989) Genetics 122, 19-27 [Abstract/Free Full Text]
  23. Feldheim, D., Rothblatt, J., and Schekman, R. (1992) Mol. Cell. Biol. 12, 3288-3296 [Abstract]
  24. Nothwehr, S. F., and Stevens, T. H. (1994) J. Biol. Chem. 269, 10185-10188 [Free Full Text]
  25. Nishikawa, S., and Nakano, A. (1991) Biochim. Biophys. Acta 1093, 135-143 [CrossRef][Medline] [Order article via Infotrieve]
  26. Stearman, R., Yuan, D., Yamaguchi-Iwan, Y., Klausner, R. D., and Dancis, A. (1996) Science 271, 1552-1557 [Abstract]
  27. Raths, S., Rohrer, J., Crausaz, F., and Riezman, H. (1993) J. Cell Biol. 120, 55-65 [Abstract]
  28. Lapinskas, P. J., Cunningham, K. W., Liu, X. F., Fink, G. R., and Culotta, V. C. (1995) Mol. Cell. Biol. 15, 1382-1388 [Abstract]
  29. Lin, S., and Culotta, V. C. (1995) Proc. Natl. Acad. Sci. U. S. A. 92, 3784-3788 [Abstract/Free Full Text]
  30. Klausner, R. D., and Sitia, R. (1990) Cell 62, 611-624 [Medline] [Order article via Infotrieve]
  31. Serrano, R., Kielland, B. M., and Fink, G. R. (1986) Nature 319, 689-693 [Medline] [Order article via Infotrieve]
  32. Chang, A., and Fink, G. R. (1995) J. Cell Biol. 128, 39-49 [Abstract]
  33. Kepes, F., and Schekman, R. (1988) J. Biol. Chem. 263, 9155-9161 [Abstract/Free Full Text]
  34. Orlean, P. (1990) Mol. Cell. Biol. 10, 5796-5805 [Medline] [Order article via Infotrieve]

©1997 by The American Society for Biochemistry and Molecular Biology, Inc.