Surplus Zinc Is Handled by Zym1 Metallothionein and Zhf Endoplasmic Reticulum Transporter in Schizosaccharomyces pombe *

Homeostatic mechanisms prevent the accumulation of free zinc in the cytoplasm, raising questions regarding where surplus zinc is stored and how it is delivered to and from these stores. A genetic screen for zinc hypersensitivity inSchizosaccharomyces pombe identified a missense mutation truncating Zhf, an endoplasmic reticulum transporter. These cells were ∼5-fold more zinc-sensitive than other independent mutants. The targeted disruption of zhf prevented growth on low zinc medium and caused hypersensitivity to elevated zinc/cobalt but resistance to cadmium. The exposure to elevated zinc but not copper also promotes the accumulation of transcripts encoding a metallothionein designated Zym1. The Sty1 pathway is required for maximal zym1 expression but is not obligatory for zinc perception. The targeted disruption of zym1impaired cadmium tolerance but only slightly impaired zinc tolerance, whereas zym1 overexpression substantially rescued zinc hypersensitivity of zhf− cells. Four equivalents of zinc were displaced from Zym1 by up to 12 equivalents ofp-(hydroxymercuri)phenylsulphonate. Zym1 thiols react rapidly with 5,5′-dithiobis-(2-nitrobenzoic acid) compared with bacterial zinc metallothionein (6.8 and 0.2 × 10−4s−1, respectively). Zym1 is unlike known fungal metallothioneins that are induced by and sequester copper but not zinc. Less zinc but normal cadmium was accumulated by zym1Δ, consistent with zinc sequestration by Zym1 in vivo.

Proteins that require zinc for catalysis pervade metabolism, and zinc is a common element of protein structure (1). However, excess zinc is toxic, and cells restrict the availability of zinc. In mammalian cells, cytosolic-free zinc content has been estimated as picomolar (2), and the sensitivity of cellular zinc sensors suggests that there is also effectively no free zinc in Escherichia coli (3). Surplus zinc must be sequestered and stored until it is required for the incorporation into nascent proteins. It remains to be fully defined how zinc is delivered to intracellular stores and mobilized from them.
In higher eukaryotes but not Saccharomyces cerevisiae, surplus zinc can be sequestered in small cysteine-rich metal-binding proteins called metallothioneins (4). The exposure to zinc and a diversity of other factors promotes a rapid increase in both metallothionein mRNA and protein (5,6). Metallothioneins are capable of binding a range of metals, but in mammals they predominantly coordinate zinc with the exception of cells challenged with supraoptimal concentrations of metals that can displace zinc from metallothionein clusters (7). Mice that lack MT-I 1 and MT-II exhibit a small increase in sensitivity to zinc (8), but in cell lines that lack functional zinc transporters, metallothioneins confer some residual resistance (9). A number of roles have been proposed for MT-I and MT-II, but uncertainty remains around the evolutionary advantage that has selected for so many metallothionein genes in such a wide diversity of organisms (4). Importantly, in S. cerevisiae, the metallothioneins Cup1 and Crs5 coordinate copper are induced in response to elevated copper concentrations by the transcriptional regulator Ace1 and function predominantly in the detoxification of excess copper (10,11). There is no evidence for any role for either protein in zinc homeostasis.
In mammalian cells, surplus intracellular zinc is also sequestered to compartments such as vacuoles and specialized vesicles, and this is also true of S. cerevisiae (12)(13)(14)(15). The mutants of S. cerevisiae that are defective in either vacuolar biogenesis or acidification consequently exhibit increased sensitivity to metals including zinc (16,17). Two important determinants of zinc tolerance in S. cerevisiae are the vacuolar metal transporters Zrc1 and Cot1 (18). The overexpression of either of these transporters increases zinc tolerance, whereas deletion reduces zinc tolerance (19 -21). Furthermore, a loss of these proteins severely impairs the accumulation of intracellular zinc (15). Zrc1 and Cot1 are members of the cation diffusion facilitator family that also includes the mammalian zinc transporters ZnT-1 to ZnT-5 (22,23). ZnT-1 is located at the plasma membrane and functions in zinc efflux (9). ZnT-2 and ZnT-3 are involved in the intracellular trafficking of zinc (12,13,24). ZnT-3 functions in zinc uptake into vesicles within neurones (13). ZnT-2 contributes to zinc efflux and the loading of vesicles in the intestine, kidney, and testis (12). Transfection with ZnT-2 confers a degree of zinc resistance, indicating that storage in intracellular compartments maintains a tolerable number of zinc atoms within the cytosol (12). ZnT-4 performs specialized functions in the brain, intestine, and mammary gland (25), whereas ZnT-5 transports zinc into secretory granules in pancreatic beta cells (23). During the writing of this paper, Clemens et al. (26) reported a cation diffusion facilitator transporter Zhf located in the endoplasmic reticulum and linked to * This work was supported in part by research grants from the Biological Sciences and Biotechnology Research Council. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
‡ To whom correspondence may be addressed. the accumulation of zinc within this compartment in Schizosaccharomyces pombe. We had coincidentally recovered the gene encoding the same protein from a screen designed to identify the major determinant(s) of zinc tolerance, confirming similar phenotypes and additionally observing impaired growth of zhf⌬ on low zinc medium (LZM).
Here we examine the action of a gene designated zym1, encoding a 50-residue cysteine-rich polypeptide. By analogy to fungal metallothioneins, it might be anticipated that zym1 transcripts would accumulate in response to copper, but we establish that zinc not copper enhances transcript abundance. Expression is substantially reduced in either wis1⌬ or pcr1⌬ cells, implicating the Sty1 MAPK pathway (27) in zym1 regulation, although this pathway is not obligatory for zinc perception because zinc-dependent expression is retained albeit at a lower level. The deletion of zym1 confers a "minimal" reduction in zinc tolerance in a wild-type background. Surprisingly, zym1⌬ does show some reduction in cadmium tolerance, even though phytochelatins are the principal determinants of resistance in S. pombe (28,29). We have exploited zhf Ϫ cells to demonstrate that zym1 overexpression does substantially rescue zinc hypersensitivity of this genotype. The deletion of zym1 also reduces zinc accumulation but does not alter the accumulation of cadmium. We report in vitro analyses of the zincbinding properties of Zym1 and establish that 10 -12 equivalents of p-(hydroxymercuri)phenylsulphonate (PMPS) displace four equivalents of zinc, confirming the coordination to most if not all 12 cysteine residues. As a zinc metallothionein from a fungal species, Zym1 provides the opportunity for functional and regulatory studies in a genetically tractable organism.
Plasmids-A plasmid (pRep1-Zym1) for the over-expression of zym1 in S. pombe was created by using PCR products corresponding to zym1, generated using primers 5Ј-zym1 (5Ј-GCGATCATATGGAACACAC-TACCCAATG-3Ј) and zym1-3Ј (5Ј-CGTCTGGATCCGTCAAGGCCA-GAACTACAGTC-3Ј). The resulting fragment was digested with NdeI and BamHI followed by ligation to the NdeI and BamHI sites of the pRep1 vector to create pRep1-Zym1. A plasmid for the overexpression of zym1 in E. coli was produced by ligating the zym1 PCR product into the NdeI and BamHI sites of pGEM-T (Promega). The zym1 DNA was then excised using NdeI and EcoRI and ligated to the prokaryotic expression vector pET29a to create pMHNR6.1.
RNA Analysis-RNA was extracted from 0.3 ϫ 10 9 cells as described previously (33) with some modification. Cell pellets were washed in H 2 O and resuspended in 200 l of RNA buffer (50 mM Tris-HCl, pH 8.0, 100 mM NaCl, 50 mM EDTA, pH 8.0, 0.25% (w/v) SDS) with 200 l of phenol:chloroform in a 2-ml screw-capped Eppendorf tube. The cells were ruptured with 0.75 ml of 0.5-mm glass beads (Biospec) in a Ribolyser (Hybaid) using two 10-s bursts at full power. An additional 0.75 ml of RNA buffer was added followed by spinning in a microfuge for 5 min. The aqueous layer was subjected to further phenol:chloroform extractions before the RNA was precipitated with 0.1 volumes of sodium acetate, pH 5.2, 0.6 volumes of isopropyl alcohol. RNA pellets were washed in 70% (v/v) ethanol and resuspended in H 2 O. A 10-15-g sample of total RNA was denatured with glyoxal, separated on either a 1.2 or 1.4% (w/v) agarose gel prepared in 15 mM sodium phosphate, pH 6.5, and transferred to a GeneScreen Plus hybridization membrane (PerkinElmer Life Sciences). DNA probes were produced by PCR amplification from genomic DNA using the appropriate primers. All of the probes were labeled with [␣-32 P]dCTP by use of a Prime-a-Gene labeling kit (Promega).
Analysis of Growth in Response to Metals-Exponentially growing cells (ϳ2 ϫ 10 4 ) were inoculated into 5 ml of medium supplemented with increasing concentrations of metal. A 595 was measured either after an overnight incubation at 30°C for rich medium (YE5S) or minimal medium (EMM) or after a 20-h incubation at 30°C for LZM.
Production of Recombinant Zym1-pMHNR6.1 was used to transform BL21 (DE3) E. coli cells and protein expression induced in cells grown to 0.7 (A 595 ) by the addition of 1 mM isopropyl-1-thio-␤-D-galactopyranoside. Zinc sulfate (0.5 mM) was added to the E. coli growth media, and the cells were allowed to grow for an additional 5 h before harvesting by centrifugation. The aliquots (2 ml) of cleared lysate were applied to a gel filtration column (Sephadex G-75, 2.5 ϫ 60 cm, Amersham Biosciences) equilibrated with 25 mM Tris, pH 9.7, 1 mM dithiothreitol. Fractions containing Zym1 were pooled, concentrated (Vivaspin 20, 2.5 kDa of molecular mass cut off, Sartorius), and exchanged into 25 mM Tris, pH 9.7, by gel filtration (PD10, Amersham Biosciences). Zym1 was applied to an anion exchange column (1-ml HiTrapQ, Amersham Biosciences) equilibrated in 25 mM Tris, pH 9.7), and eluted with 40 ml of 0 -500 mM linear NaCl gradient. The purity and identity of the recombinant protein were confirmed by polyacrylamide gel electrophoresis. SmtA was expressed and purified as described previously (34).

Identification of Zhf as a Major Determinant of Zinc
Tolerance-To identify proteins responsible for either sequestering or storing surplus zinc ions, we undertook a genetic screen to isolate mutants with reduced zinc tolerance. S. pombe cells were subjected to random mutagenesis via UV irradiation, and colonies were identified that exhibited impaired growth on rich agar (YE5S) supplemented with 2 mM ZnSO 4 . Approximately, 50,000 colonies were screened from which 12 strains with increased sensitivity to zinc were isolated. One strain subsequently designated zhf-537 was non-viable on rich agar (YE5S) containing 0.4 mM ZnSO 4 (Fig. 1A). An isogenic wild type formed colonies on rich agar (YE5S) containing 3 mM ZnSO 4 . Exponentially growing cells were inoculated into rich medium (YE5S) containing increasing concentrations of zinc and were incubated overnight. The growth of the wild type was not inhibited by up to 150 M ZnSO 4 , whereas zhf-537 showed some inhibition at 5 M ZnSO 4 and complete inhibition at 50 M (Fig.  1B). The mutant had impaired growth on medium enriched with cobalt but enhanced growth in the presence of cadmium compared with control cells (data not shown).
As a first step to identify the mutation, we searched the S. pombe genomic data base for homologues of genes with known function in the cell biology of zinc. Open reading frame SPAC23C11.14 encodes a polypeptide with greater than 40% identity to Cot1 and Zrc1 of S. cerevisiae (19 -21). Therefore, we recovered and sequenced SPAC23C11.14 from zhf-537 and identified a nonsense cytosine to thymine substitution converting codon Gln-298 to stop and truncating the carboxyl terminus of Zhf by 89 residues. To confirm that zinc sensitivity resulted from loss of Zhf, we created a strain in which zhf was replaced with the ura4-selectable marker. It was noted that the introduction of ura4 alone enhanced growth in rich medium without added zinc, although supplementation with ZnSO 4 in the range of 50 -150 M restored growth at rates more similar to strains deficient in ura4 (Fig. 1, B and C, compare control strains). Most importantly, the zhf⌬ strain was hypersensitive to zinc showing impaired growth on rich medium (YE5S) compared with the equivalent control strain and was unable to grow in medium supplemented with 20 M ZnSO 4 (Fig. 1C). A genetic cross between zhf-537 and zhf⌬ yielded only zinc-hypersensitive progeny, confirming that the mutation of zhf is the major determinant of zinc resistance and responsible for the aberrant phenotype of zhf-537.
Zhf Is Required for Growth on Low Zinc-Intracellular sequestration by cation diffusion facilitator transporters in S. cerevisiae provides a supply of zinc for efficient use, especially when cells are transferred to zinc-depleted conditions (15). Does Zhf also contribute toward the efficient use of zinc, and does zhf⌬ have impaired growth in zinc-limiting medium? The ura4 containing control strain had poor growth on LZM unless supplemented with at least 20 M ZnSO 4 (Fig. 1D). Significantly, zhf⌬ failed to grow on LZM, although slight growth was restored in 20 M ZnSO 4 . This was not to control the levels (Fig.  1D, inset). In the absence of Zhf, S. pombe is sensitive to both high and low zinc such that in minimal medium with limited buffering capacity, we were unable to establish conditions supporting growth equivalent to the respective control strain.
Zinc but Not Copper Induce zym1 Metallothionein Transcripts-S. pombe synthesizes cadmium-binding phytochelatins termed class III metallothioneins (36 -38), but there is also open reading frame SPAC22H10.13 within the S. pombe genome encoding a hypothetical protein with similarity to mammalian class I metallothioneins (Fig. 2). The predicted protein contains 12 cysteine residues, is shorter than mammalian metallothioneins (50 versus 61 amino acids), and the arrangement and spacing of the cysteine residues differ. In addition, the hypothetical S. pombe metallothionein also contains a single histidine residue. Is SPAC22H10.13, hereafter referred to as zym1, transcribed, and if so do the transcripts accumulate in response to elevated concentrations of copper in common with other fungal metallothioneins?
Zinc caused the accumulation of zym1 transcripts with up to 10 -20-fold increase 30 min after the addition of ZnSO 4 (Fig. 3,  A and C). There was a direct correlation between zym1 transcript abundance and zinc concentration. In contrast, equivalent exposures to CuSO 4 did not result in a detectable increase in the abundance of zym1 transcripts (Fig. 3B). To detect any subtle change in zym1 transcript abundance in response to copper, the blot was subjected to an extended exposure and this revealed low abundance and longer zym1-specific transcripts in control cells. The proportion of longer zym1 transcripts increased in cells exposed to copper, although the overall zym1 transcript abundance did not increase (Fig. 3B).
The Sty1 Pathway Positively Controls zym1 Expression but Is Not Obligatory for Induction by Zinc-The copper-binding transcriptional activator Ace1 modulates copper metallothionein expression in S. cerevisiae, but the induction of zym1 by zinc (Fig. 3A) implies a different mode of control (10,11). In S. pombe, the Sty1 MAPK pathway plays a central role in coordinating changes in gene expression in response to a wide range of environmental stresses (27). We reasoned that this pathway could control zym1 and therefore examined transcript abundance in a strain lacking Wis1, a MAPK kinase critical for the activation of Sty1 (38 -40). Both the basal and zinc-induced levels of zym1 transcripts were severely reduced in wis1⌬ (Fig.  3C). The Sty1 pathway clearly controls the expression of zym1, but this pathway is not obligatory for zinc perception because zinc induction was retained in Wis1 mutants albeit at reduced magnitude. The abundance of zym1 transcripts was also reduced in cells lacking Pcr1, a bZIP transcription factor that in conjunction with Atf1 functions downstream of Sty1 (41, 42) (Fig. 3C).
Overexpression of zym1 Suppresses Zinc Hypersensitivity of zhf-537-Zinc-dependent zym1 expression (Fig. 3A) suggests a role in the binding and storage of zinc in vivo. We investigated whether or not the overexpression of zym1 suppressed zinc hypersensitivity associated with the loss of Zhf. The zhf-537 strain was transformed with either pRep1-Zym1, which expresses zym1 from the nmt1 promoter (43) unlike the control strain (Fig. 4A). Suppression appeared incomplete, because wild-type cells grew better on 0.4 mM ZnSO 4 than zhf-537 containing pRep1-Zym1. Exponentially growing liquid cultures of zhf-537 containing pRep1-Zym1 showed enhanced cell division in the presence of zinc compared with equivalent cells containing pRep1 alone (Fig. 4B).
Deletion of zym1 Reduces Zinc Accumulation but Only Slightly Impairs Zinc Tolerance-A role in zinc homeostasis was further tested by creating a strain in which zym1 was inactivated by insertion. The growth of the resulting zym1⌬ showed only a small but reproducible impairment in rich medium (YE5S) supplemented with 10 -100 M zinc (Fig. 5A). The analyses of cellular metal content by atomic absorption spectroscopy revealed less atoms of zinc per zym1⌬ cell compared with wild-type cells (Fig. 5B).
Zym1 Affords Some Protection against Cadmium-Phytochelatins are the major determinants of cadmium resistance in S. pombe (28,29). We tested the growth of zym1⌬ in cadmium and observed a decrease in tolerance (Fig. 6A). However, zym1⌬ showed no difference in cadmium content as determined by atomic absorption spectroscopy (data not shown), and in this respect, the action of Zym1 is different in the cellular handling of cadmium and zinc. A loss of Zym1 has the greatest effect on resistance to cadmium (compare Fig. 5 with 6A) but only modulates the accumulation of zinc (Fig. 5B).

FIG. 3. Expression of zym1 is induced by zinc but not copper.
A, exponentially growing wild-type NT5 cells were treated with ZnSO 4 at the indicated concentration for 30 min. Cells were harvested, and total RNA was prepared and subjected to northern blotting using zym1 and cdc2 probes. B, zym1 mRNA is not induced by copper. The arrow indicates the presence of a longer transcript. Details are as for A with the exception that CuSO 4 was used in place of ZnSO 4 . C, exponentially growing wild-type NT5 (w. t.), wis1⌬ (GD1682), and pcr1⌬ (JX125) cells were treated with ZnSO 4 (2 mM) for either 15 or 30 min as indicated. Cells were harvested, and total RNA was prepared and subjected to northern blotting using zym1 and cdc2 probes. To detect background variability, two cell extracts (Ϫ) were prepared in each experiment. 2) in minimal medium (EMM) before adding ZnSO 4 to the appropriate concentration. A 595 was measured following incubation at 30°C for 6 h. Data are the means of triplicate determinations with S.E., and equivalent trends have been observed on two further occasions.
Cadmium Promotes Accumulation of Longer zym1 Transcripts-Evidence that Zym1 confers some protection against cadmium indicates that the gene product is present in cadmium-challenged cells. Basal zym1 expression in culture may be sufficient to mediate this level of resistance. Alternatively zym1 may be induced by cadmium, and this theory was tested. Cadmium increased the proportion of longer zym1 transcripts as observed previously following the exposure to copper (Fig.  6B). The longer transcript is unlikely to result from cadmiummediated inhibition of splicing, the zym1 gene containing no predicted introns. It is formally possible that cadmium selects an alternative zym1 transcription start or termination site. There is a precedent for stress in the form of DNA-damaging agents, influencing the selection among three optional termination sites in the rhp51 gene of S. pombe (44). Transcript abundance continues to increase up to 2 mM zinc ( Fig 3A). Of course, cadmium is toxic at lower concentrations than zinc (Figs. 5A and 6A, compare x axes on showing inhibition by cadmium under these exposure conditions at approximately two orders of magnitude lower concentrations than zinc), precluding comparative analysis of viable cells at 2 mM. However, it is notable that transcript abundance (the sum of both transcripts) does not substantially increase at cadmium concentrations greater than 0.1 mM, indicating that optimal induction has been achieved (Fig. 6B). The fold increase in zym1 transcript abundance in response to zinc (Fig. 3A) is greater than that observed in response to cadmium (Fig. 6). It is formally possible that a response to cadmium could indirectly result from the displacement of intracellular zinc.
Zym1 Binds Four Zinc Atoms in Vitro via up to 12 Cysteines-Several lines of evidence suggest that Zym1 can sequester zinc in vivo. To test the putative zinc-binding properties of Zym1, recombinant Zym1 was generated in E. coli (supplemented with 0.5 mM zinc) and purified from cell homogenates by gel filtration and anion exchange chromatography. Purity was tested by polyacrylamide gel electrophoresis, identity confirmed by amino-terminal sequence analysis, and association with zinc was confirmed by atomic absorption spectroscopy. To investigate stoichiometry and the nature of liganding, Zym1 was incubated with increasing concentrations of PMPS in the presence of the metallochromic indicator 4-(2)-(pyridylazo)resorcinol (35). Between 10 and 12 molar equivalents of PMPS displaced approximately four equivalents of zinc from Zym1, indicating that up to 12 cysteine A 595 was measured following overnight incubation at 30°C. No difference in cadmium content was detected using atomic absorption spectrophotometry (data not shown). Data are the means Ϯ S.E. of triplicate determinations, and equivalent trends have been observed on two further occasions. B, exponentially growing wild-type SW62, cells were treated with CdSO 4 for 30 min. Cells were harvested, and total RNA was prepared and subjected to northern blotting using zym1 and cdc2 probes. The arrow indicates the presence of a longer transcript. residues are involved in binding four atoms of zinc (Fig. 7A). All of the 12 cysteine residues appear to contribute ligands. Bridging ligands are probable, because there are insufficient cysteines to provide four independent tetravalent sites.
The Thiols in Zym1 are 40-fold More Reactive with 5-Thio-2-nitrobenzoate Than Those of SmtA-Is Zym1-bound zinc likely to act as a labile zinc store? Thiol reactivities of metallothioneins correlate with the propensity of the cluster to engage in zinc transfer reactions (45). Recombinant Zym1 was incubated with an excess of DTNB, and the formation of 5-thio-2-nitrobenzoate was monitored by the increase in absorbance at 405 nm. Values were calibrated by reacting known amounts of free cysteine, and bacterial metallothionein SmtA was also used for comparison (35). The thiols react faster in Zym1 than SmtA (Fig. 7B). An analysis of the data as first order reactions gave values of 6.8 ϫ 10 Ϫ4 s Ϫ1 for Zym1 and 0.2 ϫ 10 Ϫ4 s Ϫ1 for SmtA. Zym1 is the more probable to exchange zinc in vivo, although both proteins are less reactive than mammalian metallothionein at 28 ϫ 10 Ϫ4 s Ϫ1 (45).

DISCUSSION
Zym1 is a metallothionein revealed by its cysteine-sulfur metal coordination characteristics (Fig. 7), size, and sequence ( Fig. 2) and further supported by an accumulation of zym1 transcripts in response to metal (Fig. 3). Several lines of evidence (Figs. 3-5 and 7) implicate Zym1 in the cellular handling of zinc: zinc but not copper promotes a large accumulation of zym1 transcripts (Fig. 3, A and B); the overexpression of zym1 "substantively" restores zinc tolerance to zhf mutants (Fig. 4); and the deletion of zym1 "slightly" reduces zinc tolerance and causes a decrease in the number of zinc atoms cell Ϫ1 (Fig. 5). Surprisingly, zym1-deficient cells also show slight sensitivity to cadmium (Fig. 6B). A longer zym1 transcript predominates in response to cadmium. The Sty1 pathway up-regulates zym1 but is not essential for zinc signaling to zym1, because zinc-dependent expression is retained in wis1⌬ (Fig. 3C). The endoplasmic reticulum zinc transporter Zhf is confirmed to be the major determinant of zinc resistance recovered in a phenotypic screen of S. pombe mutants, and we established that zhf is required for rapid growth under zinc deficiency (Fig. 1B). Both Zhf and Zym1 contribute toward the intracellular handling of zinc.
In S. cerevisiae, Zrc1 and Cot1 reduce the pool of toxic cytosolic zinc and hence confer the resistance to high zinc concentrations (15). Zhf from S. pombe has an extensive sequence similarity to Cot1 and Zrc1 and has been localized to the endoplasmic reticulum (26) rather than the vacuole. The detailed analyses of the intracellular distribution of zinc clearly demonstrate that Zhf contributes toward the compartmentalization of zinc within the endoplasmic reticulum and thereby detoxifies surplus (26). Fig. 1D confirms that Zhf also acts in the efficient cellular supply of zinc, because the growth of zhf⌬ is severely impaired on LZM. Zym1 is fundamentally different from the metallothioneins of other fungal species that accumulate in response to and bind copper not zinc (10,11). However, a loss of zym1 resulted in only a small decrease in tolerance to elevated zinc (Fig. 5A), consistent with Zhf rather than Zym1 in representing the primary mechanism for detoxifying excess. This finding is similar to the situation in mammals in which cell lines or mutant mice that do not produce MT-I and MT-II exhibit only slight decreases in zinc tolerance with zinc transporters acting as the principal determinants of resistance (46). It is intriguing that zhf⌬ and zym1⌬ strains show inverse phenotypes with respect to excess cadmium, the zhf⌬ strain being cadmium resistance and the zym1⌬ strain being cadmium-sensitive. That fission yeast lacking zym1 shows any reduction in cadmium tolerance was unexpected because S. pombe rapidly produces cadmium-binding phytochelatins, (␥-GluCys) n Gly, from glutathione in response to cadmium challenge (36,37). Cadmium accumulation in contrast to zinc accumulation was unaffected by the loss of Zym1 (data not shown), indicating that this metallothionein does not influence cadmium sequestration by phytochelatins. The binding of cadmium by Zym1 might provide some supplemental buffering of cadmium in transit across the sensitive cytosol prior to vacuolar deposition as cadmium-phytochelatin complexes. Alternatively, Zym1 might scavenge cytosolic zinc displaced from metalloproteins by cadmium.
How is zym1 regulated? In addition to direct metalloregulation, MT-I and MT-II are also expressed in response to agents that generate reactive oxygen species (6), and metallothioneins protect against oxidative damage (47,48). Preliminary data (not shown) indicate that zym1 transcripts accumulate in response to H 2 O 2 . Cadmium (Fig. 6B) and copper stress (Fig. 3B) cause some production of longer transcripts (Fig. 3C). Critically, the Sty1 pathway substantially enhances the overall level of expression of zym1 in both the absence and presence of zinc. "Preemptive" up-regulation of zym1 by stresses that promote zinc liberation from metalloproteins may involve general stress signal pathways such as Sty1, elevating zym1 expression under culture conditions. However, the Sty1 kinase Wis1 is not obligatory for the perception of zinc (Fig. 3C). Some independent zinc-sensing factor also acts on zym1. The S. pombe genome does not encode proteins with a high homology or similar number of zinc fingers to the S. cerevisiae low zinc sensor Zap1 (49), and it is unclear how S. cerevisiae genes are activated in response to zinc surplus. In mammals, the zinc finger factor MTF-I regulates MT-I and MT-II in response to elevated zinc (6). MTF-I homologues have not been found in lower eukaryotes, but there are several uncharacterized open reading frames in S. pombe that have some similarity to the zinc finger regions of MTF-I and are therefore candidates for the control of zym1.
Compartmentalization by Zhf is required for optimal growth in zinc limitation and excess. Zinc released from metalloproteins as a result of normal turnover or by oxidants and metals of greater affinity is likely to form gratuitous partnerships with other cytosolic ligands and may not be accessible for transport by Zhf. One attractive hypothesis is that Zym1 scavenges zinc associated with adventitious sites to inhibit deleterious interactions, aiding cellular retention and recycling.