The Family of SMF Metal Ion Transporters in Yeast Cells*

Metal ions are vital for all organisms, and metal ion transporters play a crucial role in maintaining their homeostasis. The yeast (Saccharomyces cerevisiae) Smf transporters and their homologs in other organisms have a central role in the accumulation of metal ions and their distribution in different tissues and cellular organelles. In this work we generated null mutations in each individual SMF gene in yeast as well as in all combinations of the genes. Each null mutation exhibited sensitivity to metal ion chelators at different concentrations. The combination of null mutants ΔSMF1 + ΔSMF2and the triple null mutant Δ3SMF failed to grow on medium buffered at pH 8 and 7.5, respectively. Addition of 5 μmcopper or 25 μm manganese alleviated the growth arrest at the high pH or in the presence of the chelating agent. The transport of manganese was analyzed in the triple null mutant and in this mutant expressing each Smf protein. Although overexpression of Smf1p and Smf2p resulted in uptake that was higher than wild type cells, the expression of Smf3p gave no significant uptake above that of the triple mutant Δ3SMF. Western analysis with antibody against Smf3p indicated that this transporter does not reach the plasma membrane and may function at the Golgi or post-Golgi complexes. The iron uptake resulting from expression of Smf1p and Smf2p was analyzed in a mutant in which its iron transporters FET3and FET4 were inactivated. Overexpression of Smf1p gave rise to a significant iron uptake that was sensitive to the sodium concentrations in the medium. We conclude that the Smf proteins play a major role in copper and manganese homeostasis and, under certain circumstances, Smf1p may function in iron transport into the cells.

Smf2p resulted in uptake that was higher than wild type cells, the expression of Smf3p gave no significant uptake above that of the triple mutant ⌬3SMF. Western analysis with antibody against Smf3p indicated that this transporter does not reach the plasma membrane and may function at the Golgi or post-Golgi complexes. The iron uptake resulting from expression of Smf1p and Smf2p was analyzed in a mutant in which its iron transporters FET3 and FET4 were inactivated. Overexpression of Smf1p gave rise to a significant iron uptake that was sensitive to the sodium concentrations in the medium. We conclude that the Smf proteins play a major role in copper and manganese homeostasis and, under certain circumstances, Smf1p may function in iron transport into the cells.
Transition metals are essential for many metabolic processes, and their homeostasis is crucial for life processes. Metal ion transporters play a major role in maintaining the correct concentrations of the various metal ions in the different cellular compartments. Recent studies of yeast (Saccharomyces cerevisiae) mutants revealed key elements in metal ion homeostasis, including novel transport systems. Several of the proteins discovered in yeast are highly conserved, and defects in some of the yeast mutants could be complemented by their human homologs (1)(2)(3)(4). Some yeast and human homologous proteins were found to be related to copper and iron transport (5-7), among which the most ubiquitous are the SMF family of genes encoding metal ion transporters (8 -10). SMF1 was originally cloned as a high copy number suppressor of a temperaturesensitive mif1-1 mutant (11). Later it was shown that the growth arrest at 37°C could be relieved by supplementing the media with Mn 2ϩ or overexpressing SMF1 that transports Mn 2ϩ from the medium and elevates its concentration in the cytoplasm (8,12,13). The temperature-sensitive mif1-1 mutant may have resulted from reduced stability of the processing peptidase under limited manganese concentrations in the medium (8). Further studies indicated the SMF1 is a general metal ion transporter and can transport not only Mn 2ϩ , Zn 2ϩ , and Cu 2ϩ (8), but also Fe 2ϩ , Cd 2ϩ , Ni 2ϩ , and Co 2ϩ (9,12,14). Yeast cells contain additional two genes of this family, SMF2 and SMF3, and indirect evidence indicates that they are also broad range metal ion transporters but exhibit different specificity from SMF1 (10), suggesting a specific function for each of them.
Expression of Smf1p in Xenopus oocytes demonstrated that this protein mediates H ϩ -dependent divalent metal ion transport (14). In addition, a large Na ϩ leak through Smf1p was observed, and sodium competed with the activity of metal ion uptake. Because the Smf family of proteins transports a wide range of divalent metal ions and specific and highly regulated transport systems exist in yeast for several of those metals (5-7), the function of the Smf family of proteins is not apparent. Our approach, since the discovery of these family members as metal ion transporters (8), was to identify conditions and yeast mutants that will help to elucidate specific functions of the various genes.
Here we report on the properties of deletion mutants in each of the genes encoding SMF1, SMF2, and SMF3, as well as the combination of multiple deletions, including a triple mutant lacking the three genes. Our study revealed that the Smf proteins take part in the transport of Cu 2ϩ and Mn 2ϩ into yeast cells, and in their absence a growth arrest occurs due to a shortage in these metal ions.
The cells were grown in a YPD medium containing 1% yeast extract, 2% bactopeptone, and 2% dextrose. For metal ion limitation experiments, the cells were grown in a medium containing 0.25% yeast ex-tract, 0.5% bactopeptone 2% dextrose, 50 mM MES, 1 and the pH was usually adjusted to pH 6 by NaOH (16,17). Agar plates were prepared by the addition of 2% agar to the YPD buffer medium at the given pH. Yeast transformation was performed as described previously (18), and the transformed cells were grown on minimal plates containing a 0.67% yeast nitrogen base, 2% dextrose, 2% agar, and the appropriate nutritional requirements.
Gene Disruption-The gene knockout of the new strains was performed as follows: All or part of the target gene was replaced by the selectable marker URA3, leaving flanking DNA sequences of about 0.3 kb. When PCR was used for the construct, the DNA fragments were cloned into the TA plasmid or pGEM-T Easy (Promega). Transformed colonies that grew on the selective medium were selected, checked by PCR for homologous recombination, and analyzed for their phenotype. The genes containing approximately 0.3-kb flanking sequences were cloned by PCR into YEP24 or YPN2 or BFG plasmids (17). Sequential gene disruption was obtained by inactivation of the URA3 gene and selection on minimal plates containing 1 mg/ml 5-fluorouracil. The colonies that grew under this condition were analyzed for lack of growth on minimal plates without uracil. These yeast strains were used for subsequent gene disruption with a URA3-selectable marker. The various null mutants were analyzed for the disruption of each gene by PCR using one primer from the URA3 gene and one primer flanking the interrupted gene.
The interruption of SMF1 was performed as described previously (8). SMF2 was interrupted by the PCR-only (HANNAH) method (19). The 5Ј-and 3Ј-flanking regions of the gene, as well as the selectable marker, were amplified by PCR using oligonucleotides that are partially complementary to yield, in a second PCR, a DNA fragment composed of 5Ј-SMF2 URA3 3Ј-SMF2. Most of the reading frame of the gene was deleted leaving DNA fragments encoding 14 and 18 amino acids at the N and C termini of the transporter, respectively. SMF3 was interrupted by introducing the URA3 gene in the StyI site of SMF3. The same constructs were used for sequential interruption of more than one SMF gene.
Yeast Transformation-Yeast transformation was performed either by the method of Ito et al. (18) or by a bench-top method according to Elble (20). Yeast cells were grown overnight in 5 ml of YPD medium (pH 5.5) to stationary phase. The cells were centrifuged for 10 s in an Eppendorf microcentrifuge at 13,000 rpm. 10 l of salmon sperm (10 mg/ml) was added to the pellet as a DNA carrier. Then about 1 g of plasmid or DNA construct was added. Finally, the pellet was suspended in 0.5 ml of medium containing 10 mM Tris, pH 7.5, 1 mM EDTA, 40% polyethylene glycol 4000, and 0.1 M lithium acetate. The suspension was incubated overnight at room temperature and plated on the appropriate plates (17).
DNA Isolation from Yeast-Yeast cells were grown in 5 ml of selective or YPD medium to stationary phase. The cells were harvested by centrifugation for 2 min at 2500 rpm. The pellet was suspended in 100 l of STET solution containing 50 mM Tris (pH 8), 50 mM EDTA, 5% Triton X-100, and 8% sucrose. Glass beads (about 0.2 g) were added, and the suspension was vortexed for 20 min. Then, an additional 100 l of STET solution was added, and the mixture was boiled for 3 min, cooled for 1 min on ice, and centrifuged for 10 min at 18,000 ϫ g. 100 l was removed from the supernatant and 50 l of 7.5 M ammonium acetate was added. The mixture was incubated for 1 h at Ϫ20°C and centrifuged for 10 min at 18,000 ϫ g. 100 l of supernatant was removed to a fresh tube, 200 l of cold ethanol was added, and the mixture was centrifuged for 30 min at 18,000 ϫ g. The pellet, containing the DNA, was washed with 70% ethanol and dissolved in 20 l of 10 mM Tris and 1 mM EDTA (pH 8).
Antibody Preparation and Western Analysis-Polyclonal antibody against Smf3p was obtained by injecting rabbits with a chimeric protein containing the maltose-binding protein and the hydrophilic sequence of amino acids 382-469 of the Smf3p. The DNA fragment encoding these amino acids was amplified by PCR with introduced EcoRI and HindIII restriction sites. The amplified DNA fragment was cloned in-frame to the maltose binding protein in the plasmid PMAL-C (New England BioLabs). Following sequence verification, 500 ml of bacterial culture was grown to OD 0.5 at 600 nm, induced with isopropyl-1-thio-␤-Dgalactopyranoside for 3 h, and harvested by centrifugation at 4000 ϫ g. The cells were disrupted by French press, and the protein was purified by using a column containing maltose agarose. The fractions containing the chimeric protein were dissociated by SDS, loaded on preparative gel, and electrophoresed. The gel was briefly stained by Coomassie Blue, the identified protein band was cut out, and the fusion protein was electroeluted. About 0.25 mg of fusion protein was injected into rabbits as described previously (21). Antibody to Pma1p was raised in rabbits using the purified protein that was electroeluted from polyacrylamide gels as described previously (21)(22)(23). Antibody against Sed5p was a generous gift from Dr. Randy Schekman (University of California, Berkeley, CA). The antibody detection system (ECL) was from Amersham Pharmacia Biotech. Western blots were performed according to the protocol of the ECL antibody detection system from the manufacturer. Samples were denatured by SDS sample buffer and electrophoresed on 12% polyacrylamide Mini gels (Bio-Rad) as described previously (24). Following electrotransfer at 0.5 A for 15 min, the nitrocellulose filters were blocked for 1 h in a solution containing 100 mM NaCl, 100 mM sodium phosphate (pH 7.5), 0.1% Tween 20, and 5% nonfat dried milk. Antibodies were incubated for 30 min at room temperature at a dilution of 1:1000 in a similar solution containing 2% dried milk. Following five washes in the same solution, peroxidase-conjugated second antibody or protein A was added to the filters. After incubation for 30 min and five washes with the same solution, the nitrocellulose filters were subjected to the ECL amplification procedure. The filters were exposed to Kodak X-Omat AR film for 5-60 s.
Membrane Preparations-Yeast cells were grown in 500 ml of YPD medium (pH 5.5) to OD 1 at 600 nm. The suspension was centrifuged at 3000 ϫ g for 5 min, and the pellet was washed with 200 ml of water and then washed again with 1 M sorbitol. The cell wall was digested by 2.5 units of zymolyase in 10 ml of solution containing 10 mM HEPES, pH 7.5, and 1 M sorbitol. After 30-min incubation at 30°C, the suspension was centrifuged in 15-ml Corex tubes at 3000 ϫ g for 5 min. 1 ml of glass beads was added to the pellet as well as 1 ml of solution containing 30 mM MOPS (pH 7), 1:100 protease inhibitor mixture (Sigma), 1 mM phenylmethylsulfonyl fluoride, 1 mM EDTA, and 1 mM EGTA. The suspension was vortexed five times for 30 s with incubation on ice for 30 s between each vortexing step. The solution was removed from the glass beads and placed in a new Corex tube. An additional 2 ml of the above solution was added to the tube with the glass beads, and the tube was vortexed briefly. The suspension was added to the previous one and centrifuged at 1000 ϫ g for 5 min to give a pellet containing the cell debris and nuclei. The supernatant was centrifuged at 12,000 ϫ g for 10 min, and the pellet was suspended in 0.3-0.5 ml of a solution containing 10 mM Hepes (pH 7.5) and 0.5 M sorbitol and stored as the mitochondrial fraction. The supernatant was centrifuged at 115,000 ϫ g for 30 min, and the pellet was suspended in 0.3-0.5 ml of a solution containing 10 mM Tris-Cl (pH 7.5), 1 mM EDTA, 2 mM dithiothreitol, and 25% glycerol and stored as the membrane fraction at Ϫ80°C. Sucrose gradients were also used to estimate the relative density of various membrane fractions. The gradients were made as described in Lupashin et al. (25), except that gradients of 20 -60% sucrose were used and the centrifugation was done for 14 h.

The SMF Family of Yeast Metal Ion Transporters-SMF1
encodes a hydrophobic protein of 63,258 Da with potentially eight to ten transmembrane domains. A search in the yeast genome data base with the Smf1p sequence revealed two homologous genes that were named SMF2 and SMF3. These genes encode proteins of 59,758 Da and 51,778 Da, respectively, and contain a similar number of transmembrane domains. Fig.  1 shows the multiple alignment of the predicted amino acid sequences of the three members of the yeast SMF gene family. The three proteins exhibit about 50% identity to each other, and the main difference between them was found at their N terminus. At this end, Smf3p is shorter then Smf1p by 70 amino acids and by 51 amino acids from Smf2p. These extra pieces are highly populated by charged amino acids. Up to 12 transmembrane segments have been proposed to constitute DCT1, the mammalian homolog of Smf1p (26). We assume that the number of transmembrane segments will be similar in all the family members and 10 or 12 transmembrane segments are likely to exist in these transporters. Multiple alignment of amino acid sequences of family members from bacteria, yeast, plants, insects, and mammals suggests that several conserved charged amino acids are present inside the membrane. Among them (using Smf1p amino acid sequence) are Asp-92 in TM1; Glu-160, Asp-167, and Glu-170 in TM3; and His-278 in TM6. The three negatively charged amino acids in TM3 face the same side of an ␣ helix, suggesting a role in translocation of positively charged ions across the membrane. Obviously, the above structural and functional assumptions have to be examined by multiple experimental approaches.
Properties of Yeast Null Mutants in the Various SMF Genes-The main feature of the SMF1 null mutant was its sensitivity to EGTA concentrations, toward which the parental wild-type strain was insensitive (8). The growth inhibition of ⌬SMF1 in the presence of EGTA could be alleviated by low concentrations of manganese or copper but not by any other metal ion. Deletion of the genes encoding SMF1 and SMF2 resulted in a double null mutant that was not able to grow at pH 8 but exhibited normal growth in YPD medium buffered at lower pH (27). We generated null mutations in each individual SMF gene as well as combinations of null mutations in these genes. The triple null mutant in the three SMF genes (⌬3SMF) was highly sensitive to EGTA and was not able to grow on YPD medium buffered at pH 7.5. Table I summarizes the sensitivity and resistance of the various null mutants to the chelator EGTA, pH, various metal ions, high osmolarity, and oxidative stress (obtained by the addition of H 2 O 2 to the medium). All the null SMF mutants exhibited some sensitivity to EGTA. Although ⌬SMF1 and ⌬SMF2 were quite sensitive to EGTA, ⌬SMF3 showed only marginal sensitivity to the chelating agent. The mutant strains ⌬SMF1, ⌬SMF3, and ⌬SMF1ϩ3 exhibited resistance to relatively high Co 2ϩ and Mn 2ϩ concentrations in the medium. ⌬SMF2 showed resistance to osmotic stress induced by 0.9 M NaCl or 1.6 M glycerol or 1.5 or 1.7 M sorbitol, as well as relative resistance to Mn 2ϩ , but only at pH 7.5. The resistance for certain metal ions can be explained in terms of reduced transport activity of these ions in the absence of one or more of the Smf metal ion transporters. It is more difficult to account for the sensitivity to metal ions in the various null mutants. Thus ⌬SMF2 mutant was sensitive to Zn 2ϩ and Ni 2ϩ ; ⌬SMF1ϩ2 was sensitive to Ni 2ϩ , Co 2ϩ , and Mn 2ϩ ; and ⌬SMF1ϩ3 was sensitive to Zn 2ϩ . Apparently, disturbances in metal ion homeostasis may elicit pleiotropic effects through alteration and different distribution of the other metal ion transporters and/or signal transduction mediators (7, 9, 13, 28 -31).
One of the key questions to be answered is how do seemingly unrelated metal ions elicit similar effect in mutants lacking one or more of the SMF genes? Inactivation of the Drosophila homolog of the SMF1 gene resulted in a loss of taste behavior (32). Addition of manganese or iron to the food of these mutants corrected the defect in their taste behavior (33). Similarly, addition of micromolar concentrations of Cu 2ϩ or Mn 2ϩ relieved the growth arrest of the yeast mutant in which the

TABLE I Phenotypes resulted from inactivation of each SMF gene and all the combination of SMF null mutations
The assay for resistance and sensitivity for metal ions was performed on a solid YPD medium buffered at pH 6 or the indicated pH. The medium was supplemented with the indicated metal ions in their chloride form at the following concentrations: Co 2ϩ , 0.25, 0.5, 0.75 mM; Mn 2ϩ , 1.5, 2, 2.5 mM; Zn 2ϩ , 5, 6, 7 mM; Ni 2ϩ , 1.5, 2, 2.5 mM. Sensitivity to EGTA was assayed as described in Fig. 2. Resistance to NaCl (0.9 M) and sorbitol (1.5 or 1.7 M) was analyzed at pH 6. Sensitivity or resistance to H 2 O 2 was measured in liquid medium buffered at pH 6 at concentrations of 0.5 and 2 mM. Cells were inoculated to give a cell density of 0.2 OD at 600 nm, and the growth rate was measured every 2 h by following the increase in optical absorption at the same wavelength. SMF1 gene was interrupted (8). To gain further information on the phenotype of the various SMF null mutants, we tested their growth under different EGTA concentrations. Fig. 2 shows a sensitivity order toward EGTA of ⌬SMF1 Ͼ ⌬SMF2 Ͼ ⌬SMF3 for the respective null mutants. The parental wild-type strain grew quite well at EGTA concentrations of up to 3 mM. Reduced growth was already detected at 0.5 mM EGTA for the mutants ⌬SMF1 and ⌬SMF2, and a complete growth arrest of these mutants was obtained at 2 mM EGTA. The mutant ⌬SMF3 grew well on plates containing 1 mM EGTA and a complete growth arrest of this mutant was obtained only at 3 mM EGTA. The triple mutant ⌬SMF1ϩ2ϩ3 was highly sensitive to the chelating agent and also exhibited pH sensitivity. It grew normally at pH 5.5, grew poorly at pH 6.5, and exhibited a complete growth arrest at pH 7.5 (Fig. 4). 2 The growth arrest of the various SMF null mutants was alleviated by the inclusion of micromolar concentrations of Cu 2ϩ or Mn 2ϩ . Fig. 3 shows that at 3 mM EGTA, where all the SMF mutants fail to grow, only 5 M Cu 2ϩ were sufficient to promote growth. Similarly, 5 M Mn 2ϩ promoted growth of all the SMF mutants except for the triple mutant ⌬SMF1ϩ2ϩ3, which required about 10 M Mn 2ϩ for normal growth in the presence of 3 mM EGTA. The ability of this mutant to grow in the presence of EGTA was specific for Cu 2ϩ and Mn 2ϩ , and none of the other cations such as Zn 2ϩ , Co 2ϩ , Ni 2ϩ , Fe 3ϩ , Mg 2ϩ , and Ca 2ϩ could induce growth under these conditions. We were not able to analyze the effect of Fe 2ϩ , because addition of ascorbic acid by itself promoted growth in the presence of EGTA. The effect of ascorbate may result from the reduction of the EGTA-bound Cu 2ϩ into Cu ϩ , which became available to an alternative copper uptake system. The transport properties of the different Smf proteins were analyzed by expressing the various SMF genes in the triple null mutant ⌬SMF1ϩ2ϩ3 and assaying the transport of Mn 2ϩ and Fe 2ϩ by radioactive tracing. As shown in Fig. 5, expression of Smf1p or Smf2p in the triple null mutant restored the Mn 2ϩ uptake activity above the wild-type level. Smf3 exhibited no transport activity and the short Smf1p (Smf1p-s), in which 68 amino acids were removed from its N terminus, exhibited 2-to 3-fold higher transport activity than either Smf1p or Smf2p. The lack of transport activity in the cells expressing Smf3p may be due to its different cellular distribution. Fig. 6 shows that the expressed Smf3p failed to reach the plasma membrane. Smf3p is present in the sucrose gradient fractions enriched with post-Golgi vesicles where Sed5p is present. Sed5p is a syntaxin (t-SNARE) homolog required to Golgi transport and is located in the Golgi membrane, which receives transport vesicles (34). Moreover, switching to nonpermissive temperature in the sec6 mutant resulted in similar shifts in the gradient position of both proteins. The mutant cells were grown overnight at 25°C, and then the temperature was raised to 37°C for 2 h. The cells were harvested and disrupted, and their membranes The experimental procedure was as described in Fig. 2, except that the medium was adjusted to pH 8, EGTA was omitted, and the indicated concentrations of CuCl 2 , MnCl 2 , or ZnCl 2 were added. were collected by differential centrifugation and separated on sucrose gradients of 40 -60%. Western analysis revealed a similar shift in the position in the gradient of Sed5p and Smf3p.
This may suggest that Smf3p functions in the Golgi or post-Golgi compartment and is likely to be involved in maintaining Cu 2ϩ and/or Mn 2ϩ homeostasis in these organelles.
Previously we reported that Na ϩ inhibits metal ion transport by Smf1p and prevents growth of the wild type or ⌬SMF2 strains in the present of EGTA (14). Addition of sodium to the manganese uptake system as described in Fig. 5 (even at pH 6.5) had little or no effect (not shown). Experiments with Xenopus oocytes revealed that the sodium inhibition of expressed Smf1p is highly dependent on the membrane potential and is drastically reduced at negative membrane potentials. 3 Because yeast cells maintain high values of negative membrane potential, we turned to iron uptake for analyzing the effect of sodium on the metal ion uptake. Iron uptake into yeast cells is primarily mediated by a system involving Fet3p and Fet4p (30). However, it was demonstrated that alternative transport systems for iron may exist (35). To decrease the iron uptake by its specific transport systems, we utilized the double mutant (⌬2FET) in which FET3 and FET4 were inactivated. Fig. 7 depicts a direct effect of Na ϩ on the Fe 2ϩ uptake activity of Smf1p. The expressed Smf2p and Smf3p in the double mutant ⌬2FET failed to show iron uptake above background. This observation suggests that Smf1p plays a role in an alternative iron uptake system.
The effect of different metal ions on manganese uptake by the Smf proteins expressed in the triple null mutant (⌬3SMF) was analyzed (Fig. 8) Finally, the cells were sedimented in an Eppendorf microcentrifuge, suspended in 50 l of water, and transferred to scintillation vials. 50 l of 1% SDS and 4 ml of scintillation fluid were added to the vials, and the radioactivity was counted in a Beckman scintillation counter. The uptake at 0°C was deduced from each experiment.
FIG. 6. Identification of Smf3p in the Golgi or post-Golgi membranes. The antibody against Smf3p was raised as described under "Materials and Methods." Wild type cells and the temperature-sensitive sec6 mutant were grown at 25°C and 37°C in YPD media with 2% and 0.2% glucose, respectively (44). The cells were harvested and disrupted, and their membranes were collected by differential centrifugation as described under "Materials and Methods." They were then separated on sucrose gradients of 40 -60%. Fourteen fractions were collected from the bottom of the tubes, and 5 l of the fraction was electrophoresed on SDS gels and subjected to Western analysis with the indicated antibodies.

FIG. 7. Sodium inhibits FeCl 2 transport by Smf1p expressed in
FET double disruptant mutant. The FET double mutant (⌬2FET) was transformed by shuttle vector containing each of the SMF genes as described in Fig. 5. The uptake experiment was similar to that of Fig. 5, except that 1 mM FeCl 2 and 55 Fe 2ϩ replaced the manganese and 2 mM sodium ascorbate was added to the uptake medium to maintain the iron in its reduced form. The pH of the uptake medium was 6.5. The uptake medium contained either 100 mM choline chloride (gray columns) or 100 mM NaCl (white columns). The experiment was repeated twice, and the bars with standard deviation represent the mean of all the experiments. strated Mn 2ϩ , Fe 2ϩ , and Co 2ϩ transport with K m values of 2, 5, and 11 M, respectively. 3 The order of potency of inhibition by different divalent metal ions of 54 Mn 2ϩ uptake was Cu 2ϩ Ͼ Mn 2ϩ Ͼ Fe 2ϩ Ͼ Zn 2ϩ . As shown in Fig. 8, Zn 2ϩ inhibited not only the manganese transport by Smf1p and Smf2p but also the transport by an additional unknown manganese transport system. The IC 50 of 54 Mn 2ϩ uptake by ⌬3SMF expressing either Smf1p or Smf2p was about 2 M Cu 2ϩ (not shown). Although copper and cobalt were likely to be transported by Smf1p and Smf2p, zinc was not transported and was likely to inhibit by blocking the transport pathway. DISCUSSION Redox reactions are fundamental life processes, and transition metals are essential for the function of most proteins involved in electron transport. The different metal ions may be grouped into redox active ions, such as Fe 2ϩ , Cu 2ϩ , Co 2ϩ , and to a less extent Mn 2ϩ , and non-redox-active ions such as Ca 2ϩ and Zn 2ϩ . The redox-active ions normally function in enzymes that participate in redox reactions and the conversion of active oxygen-containing components. These processes require defined amounts of specific metal ions at the right position and in a timely fashion. Living cells compete for metal ions, but excess amounts are toxic and can cause damage to the very function that they serve, as well as to proteins and nucleic acids that are present in their proximity. Metal ion transporters provide an efficient tool for competition in the limited resources and the control over their accumulation storage and secretion. Genetic screens have identified several yeast genes that encode metal ion transporters or are involved in metal ion homeostasis transport (5-7, 10). One of those transporters, Smf1p, was shown to transport Mn 2ϩ across the plasma membrane of yeast cells (8). SMF1 has originally been cloned as a high copy number suppressor of a temperature-sensitive mif1-1 mutant (11). MIF1 (MAS1) and MAS2 (MIF2) encode the mitochondrial processing enhancing protein and the matrix processing peptidase, respectively (36 -38). Later it was shown that three homologous genes, SMF1, SMF2, and SMF3, are present in the yeast genome (9,10,27). The presence of the three genes raises some interesting questions about their expression and function.
In this work, we demonstrated that inactivation of each of the three SMF genes induced various degrees of EGTA sensitivity in the respective mutants. The growth arrest caused by EGTA in all the three mutants could be suppressed by the addition of 5 M Cu 2ϩ or 25 M Mn 2ϩ (Fig. 3). Addition of other divalent cations such as Zn 2ϩ , Co 2ϩ , Ni 2ϩ , Fe 3ϩ , Mg 2ϩ , and Ca 2ϩ failed to induce growth under these conditions. The effect of Fe 2ϩ could not be examined under the same conditions, because inclusion of ascorbic acid by itself suppressed the growth arrest caused by EGTA (not shown). However, the failure of high iron concentrations to induce growth and the lack of sensitivity to bathophenantroline sulfonate suggest that the growth arrest was not due to iron limitation. It is well established that yeast growth requires a constant supply of copper (39,40). However, very small amounts of copper are sufficient for sustaining growth, and it is argued that less than one free copper ion is present in each yeast cell (41). If so, how can disruption of each SMF gene cause copper shortage in the presence of EGTA? The answer is not apparent, especially in light of the presence of high affinity copper transport systems in the form of Ctr1p and Ctr3p (40). Inactivation of these proteins causes a severe disruption in iron homeostasis through the failure of Fet3p to assemble into the plasma membrane (30). Although the effect of Ctr1p and Ctr3p on iron transport was studied in detail, the function of these two proteins as copper transporters was not directly demonstrated. Inactivation of the three SMF genes resulted in a mutant strain (⌬3SMF) that exhibited high sensitivity to several chelating agents such as EGTA, EDTA, bathophenantroline sulfonate, and bathocuproinedisulfonic acid. In addition, the mutant failed to grow in YPD medium buffered at pH 7.5. Inclusion of only 5 M Cu 2ϩ or 0.1 M ascorbate in the presence of the chelating agents or at pH 7.5 promoted the growth of the mutant cells. The effect of ascorbate can be explained by the reduction of Cu 2ϩ to Cu ϩ , thus making it more available to the cells. Therefore, our experiments suggest that the Smf proteins function in Cu 2ϩ transport and that their inactivation causes copper shortage under oxidizing conditions.
Manganese also promoted growth of the various SMF null mutants in the presence of chelating agents or high pH. We demonstrated that both Smf1p and Smf2p function in Mn 2ϩ uptake into yeast cells (8,12). Most probably, Smf3p does not function in metal ion transport across the plasma membrane and may be involved in metal ion transport in the membranes of internal organelles such as the Golgi or post-Golgi vesicles and secondary endosomes (Fig. 6). These locations were shown to be critical for the assembly of iron-transporting proteins, especially through the incorporation of copper into iron oxidoreductases (30,39). The Golgi and post-Golgi vesicles are also important for the correct glycosylation of membrane and secreted proteins. Recently it was demonstrated that the glycosylation complex is vital for yeast cells and that the activity of this complex is manganese-dependent (42,43). Therefore, manganese deficiency in these compartments may result in growth arrest. Alternatively, the growth promotion of SMF null mutants in the presence of chelating agents by the addition of micromolar concentrations of manganese may result from the release of bound copper at the level of the Golgi and post-Golgi vesicles by the added manganese.