Protection Against Cadmium Toxicity in Yeast by Alcohol Dehydrogenase Wei Yu, Ian G. Macreadie, * and Dennis R. Winge University
of Utah Medical Center,
School of Medicine,
Salt Lake City, Utah
-ABSTRACT A cDNA expression library from Schizosaccharomyces pombe was transformed into Saccharomyces cerevisiae to screen for genes capable of conferring cadmium resistance to S. cerevisiae cells. The cDNA library was cloned into the S. cerevisiae expression vector pDB20 which is designed to express cDNAs via the constitutively-expressed promoter of the gene for alcohol dehydrogenase I (ADHl). Terminator and polyadenylation signals are also provided by the ADHl gene. Cadmium resistant colonies were shown to arise by a recombination event leading to the exchange of the S. pombe DNA with the chromosomal ADH 1 gene and a consequent dramatic increase in the ADH 1 gene expression due to the high copy number of the plasmid. The overexpression of ADHl effectively buffered the cells for cadmium ions by formation of Cd-ADH.
The fission yeast Schizosaccharomyces pombe detoxifies cadmium by a mechanism similar to that found in plants. Cadmium is sequestered within a complex involving isopeptides called phytochelatin or cadystin whose structure is (~Glu-Cys),,Gly [1,2]. Synthesis of these peptides is considered to be enzymatic by virtue of the isopeptide linkage within the dipeptide repeat. A transpeptidase has been isolated in plants that is capable of synthesizing (yGlu-Cys),Gly peptides from glutathione [3]. The intriguing aspect of the synthesis is the apparent regulation by metal ions. Sequestration of cadmium ions by the (~Glu-Cys),Gly peptides is the predominant mechanism conferring cadmium resistance in S. pombe [1,4]. The Cd:(yGluCys),Gly peptide complexes in fungi also contain labile sulfur present as a CdS crystalline lattice coated with the isopeptides [5]. Mutant cells defective in the production of sulfide were shown to be hypersensitive to cadmium salts [6]. Other potential mechanisms of metal resistance include restricted plasma membrane transport, facilitated efflux, and compartmentalization within the vacuole [7]. The genet-
*On leave from CSIRO Division of Biomolecular Engineering, Victoria, Australia. Address reprint requests and correspondence to: Dr. Dennis R. Winge, University Center, School of Medicine, 50 N. Medical Drive, Salt Lake City, UT 84132.
of Utah Medical
155 Journal of Inorganic Biochemistry, 44, 155 161 (1991) 0 1991 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, NY, NY 10010 0162-0134/91/$3.50
156
W. Yu et al.
its of cadmium resistance in S. pombe are poorly described and no gene has been cloned from this organism that is functional in metal ion resistance. In order to clone genes conferring metal resistance in S. pombe, we performed heterologous complementation by transformations of Saccharomyces cerevisiae with an S. pombe cDNA expression library. Metal resistance is restricted in S. cerevisiae to copper salts as it is the main metal ion capable of inducing the CUP1 metallothionein gene [8,9]. Most strains of S. cerevisiae cells are very sensitive to cadmium-induced toxicity, although one cadmium-resistant strain has been isolated in which cadmium resistance is mediated by metallothionein expression (IO]. We predicted that the constitutive expression of certain cDNAs involved in cadmium resistance in S. pombe should impart metal tolerance to a cadmium-sensitive strain of S. crrevisiae. We presently report the characterization of cadmium resistant clones of S, cerp-
vi.siae.
MATERIALS
AND METHODS
Cell Culturing S. cerevisiae strain DYl.50 (a ura3-52 leu2-3,112 trpl-1 a&2-1 his3-11 canl-100) was the recipient strain for transformations with the S. pombe cDNA library provided by Dr. J. Fikes [ 111. The cells were transformed according to the procedure of Ito et al. [ 121. Colonies resistant to cadmium salts on solid medium were tested for metal resistance in liquid synthetic complete medium lacking uracil. The effective concentration of 50% growth inhibition (EC,,,) was defined as the concentration of cadmium salt that limited growth as determined by turbidity by 50 %. Polymerase
Chain
Reaction
Reaction conditions for PCR were those recommended by Perkin-Elmer/Cetus with the primers S’CTGCACAATATTTCAAG and SGATTGGAGACTTGACCA. Thermal cycling was 94°C for 1 mitt, 50°C for 1 min, and 72°C for 3 min for 30 cycles. Other Analyses Sequences in recombinant plasmids were determined directly from the plasmids utilizing the primers used for PCR and the U.S.B. Sequenase protocol. ADH activity was measured by the method of Racker 1131. Activity was quantified as the change in absorbance per min per ml with one unit defined as the change in optical density of I per min. Protein was quantified by the method of Lowry et al. 1141. Metal analysis was performed by atomic absorption spectroscopy. Size permeation chromatography was carried out on a Superdex 200 FPLC column from Pharmacia using an elution buffer of 20 mM Tris Cl, pH 7.4 containing 50 mM KCI. Fractions of E.5 ml were collected.
RESULTS Isolation
of Cadmium
Resistant
Transformants
The host strain chosen for the transformation, S. cerevisiae strain DY 150 is sensitive to cadmium chloride (Fig. 1A)? its growth is totally inhibited in the
ADHl
FIGURE 1. Growth of suspended in sterile water, cadmium and photographed DY150 cadmium-resistant DY150 retransformed with
AND CADMIUM
TOXICITY
157
S. cerevisiue strains on 0.5 mM cadmium plates. Strains were spotted onto solidified synthetic complete medium plus 0.5 mM after two days incubation at 30°C. Strains are DY150 (A), a transformant (B), a transformant after plasmid shedding (C), and the purified plasmid pDB20-Cd’ (D).
of 150 PM Cd(I1) on solid media. Following transformation with the cDNA library of S. pombe, a total of 21,000 Ura+ transformants from five separate transformations were obtained. These were replica plated onto solid media containing 0.15, 0.35, and 0.5 mM CdCl,. A higher than anticipated number of colonies (0.09%) were cadmium resistant. Eight of these seventeen colonies grew on plates containing 0.5 mM CdCl, (Fig. 1B). The EC,, of the eight most resistant transformants in liquid medium ranged between 110 and 300 PM, while the liquid EC,, of the parental strain was 20 PM. Only these eight transformants were further analyzed in this study. presence
Genetic
Analysis
of Cadmium
Resistant
Transformants
The resistance in each transformant was shown to be plasmid-borne by genetic analyses. Initially, Ura- auxotrophic cells were recovered that had shed the plasmid by growth of the transformants in rich media. Concomitant with the plasmid shedding was the loss of the cadmium resistance (Fig. 1C). Second, the plasmids from the transformants were purified by passage through E. co/i and subsequently retransformed into S. cerevisiue DY 150. Each purified plasmid transformed DY 150 to cadmium resistance phenotype (Fig. 1D) establishing that resistance is integrally associated with the plasmid. We have denoted these plasmids as pDB20-Cd’.
158
W. Yu et al.
Molecular
Analysis
of Plasmids
Determining
Cadmium
Resistance
In order to characterize the cDNA in pDB20-Cd’ plasmids we performed restriction site and PCR and DNA sequence analyses. The PCR primers were complementary to sequences in pDB20’s ADHl promoter and terminator. respectively, and together they were used to amplify the intervening sequences. Analysis of the PCR products by electrophoresis revealed that each recombinant plasmid contained an insen at the multiple cloning site of identical size at 1 kb while the parental vector, pDB20. yielded a PCR product of about 0.2 kb The conclusion drawn frc~l thck,c data 1s thar all inserts were of identical ttizc. Analyses with restriction enzyme digests were carried out to characterize the inserts. The iVot1 restriction sites that should have flanked the inserts were absent. In addition, Hind111 digests did not yield the expected completr liberation of the cDNA. Instead Hind111 fragments of approximately 0 15 kh and ii kh were obtained for each plasmid. identical sites in each plasmid were also observed in digests using SspI, EcoRV, and NanrKl restriction endonucleases. ‘This indicate\ Ihat ail thr pDB20-Cd’ plasmids arc identical. Finally, nucleotide sequence analysis was carried out utilizing the primers used previously for the PCR. This analysis showed that the onI> novel sequence present was an intact S. ceTvIsl(Ip ADNi gene 11-51. The S. ~~r?he I.DNX insert had. therefore, been exchanged for the ,S. cereuisiae RDH I gent h> recombination utilizing the homoiogous A DH1 sequences in the pDB3O plasmid. Flgurc _3 illustrates a scheme how this might come about. T’hl~ %%emu Would lead to
ADHI
TERMINATOR
PROMOTER
,;’
yDB20 library
FIGURE 2. Possible mechanism of insert replacement in pDB20 library. The structure of a typical pDB20 recombinant is shown. S. pombe cDNAs are inserted at the BstXI site of a polylinker in pDB20 whose structure is Hi&III-Not I-&t XI-Not I -: Hind111 A homologous double crossover with the chrornusomal S. cerevisiw ADff I locr~swould lcaci to replacement of S. pombe sequence< with an intact ADNI gene.
ADHl
AND CADMIUM
TOXICITY
159
transplacement of the S. pombe DNA to the chromosomal ADNl locus. PCR analysis on total DNA of yeast transformants did not reveal extra products other than the 1 kb ADHl fragment. Therefore, it is conceivable that further recombination may have led to restoration of the ADHl gene in the nucleus. Localization
of Cadmium
in Transformants
Molecular genetic evidence suggested a direct correlation between cadmium resistance and the presence of > 50 copies of the ADHl gene. The latter could be expected to lead to increased ADH protein levels. Measurement of the ADH enzymatic activity in transformants (grown in the absence of cadmium) revealed an activity 214 times higher than the ADH level in nontransformed cells. The relationship of cadmium resistance and enhanced ADH activity was examined by analyzing lysates of transformed cells grown in cadmium by size permeation chromatography. Elution fractions were measured for ADH activity, protein, and zinc and cadmium concentrations (Fig. 3). ADH activity was found in only one region (apparent Mr 120 kDa) that coincidentally contains the peak zinc. The coelution of Zn2+ and ADH is expected as ADH is a zinc metalloprotein. Examination of the cadmium profile, however, showed the major cadmium peak was also co-incident with ADH. The molar Cd:Zn ratio in the ADH peak tube (fraction 48) was 3.5, suggesting that most of the ADH molecules contain bound Cd2+ ions. . Additional Cd-binding components were observed in the column-excluded volume and a volume near the internal volume. These undefined Cd-complexes were observed in Cd-treated S. cereuisiae cells that were not transfected, so they were not investigated further. Transformants were grown in increasing amounts of cadmium and assayed for intracellular concentrations of Cd(I1) and Zn(I1) ions as well as ADH enzymatic
1.4 r
IO
20
30
40 FRACTION
FIGURE
50
60
70
00
90
NUMBER
3. Chromatography of a cell lysate of cadmium-resistant cells. A clarified cell extract from a culture of cells pregrown to mid-logarithmic phase and incubated subsequently with 0.5 mM CdCl, for 24 hr. The extract was applied to Superdex 200 for FPLC at a flow rate of 1 ml/min. Elution fractions of 1.5 ml were analyzed for ADH activity as the absorbance change at 340 nm per ml per min (--), A,,, nm (--), zinc (- - - -), and cadmium (- * -) concentrations as pg per ml.
160
W. Yu et al.
activity (Table 1). Increasing amounts of cadmium in the culture medium resulted in increased intracellular cadmium levels but an inverse relationship to ADH activity. ADH activity is constitutive and should be directly related to total protein. However. as intracellular cadmium increased, even by 40 times, ,4DH activity dropped to almost one-third of its initial value. The zinc concentration in the lysate paralleled ADH levels. Thus, yeast Cd-ADH has a reduced specific activity relative to Zn--ADH as has been reported with the mammalian ADH enzyrne [ 161.
DISCUSSION A relatively frequent (O.OY%) event during transformations with the pDB20 library leads to cadmium resistance by an intracellular increase in the copy number of the ADH 1 gene. The mechanism of the copy number increase is presumably by homologous recombination with the chromosomal ADHl genr and pDB20 as shown in Figure 1. Another mechanism that could be utilized is ,gap repair of fragmented DNA as described by &-r-Weaver et al. 1I?]. For example, simultaneous double strand breaks in the ADHI promoter and terminator of pDB2.0 could bc repaired in S. cerevkiae using the chromosomal ilL)H 1 gene as a template. This would obviously lead to the introduction of the intact r3DH 1 gene into pDB20. Alcohol dehydrogenase 1. synthesized at high levels by \,irtue of its ,opy number in this instance. can bind cadmium and provide a considerable degree of protection to that metal. The 2 p-based plasmids replicate to greater than 50 copiex per cell [ I8 1. Alcohol dehydrogenase ib known in animal cells to bind Cd(H) Ions in animal administered cadmium salts. presumably by Zn displacement 1191. The sequestration of Cd(lI) by alcohol dehydrogenase and glutathione appear to !~e the initial Cd butfering components prior TVthe induction of metallothionein [.X.2 11. The present results illustrate that a substantial level of resistance to Cd toxicity can be achicveci by an increase in the ADN gene copy number. It is likei! that over-expression of genes encoding other Zn-metalloproteins may also afford a similar protrction against heavy metal toxicity. A cell with only a single copy of the A DNl gene may not have the opportunity to exploit this effect in cadmium resistance unless it K to amplify that locus. Facile Limpiification of a chromosomal gene with at least two tandem repeats, such as the S. cerevisiae metallothionein CUPI locus. can occur under selective pressure through meiotic gene conversIon 12’2/. The use of heterologou5 complementation for cloning genes involved in metal resistance is an obvious and powerful approach. The success of the strategy assumes that the resistance process involves a single gene product and necessitates constitu-
TABLE 1. ADH Activity and Cd and Zn Concentrations in Lysates of Cells Grown in Various Cadmium Concentrations [Cd’+ 1 in Culture CcM
Protein in:! in11
Lysate [Cd’+ 1 Lysate (Zn’” j -._..-. _.-____~______... pg,/mg Protein
ADW Activit) units, rng Protein
ADHl
AND CADMIUM
TOXICITY
161
tive or induced expression of the gene library. The ADHl promoter may be used in expression plasmids if the termination sequences are provided from a non-ADH 1 gene to minimize chances of homologous recombination. We thank Dr. John Fikes and Dr. David Stihman for supplying the pDB20 library and strain DYISO, respectively. This work was supported by Public Health Service grant No. ES 03817 from the National Institutes of Health to DR W. IGM was the recipient of a Fulbright award.
REFERENCES 1. Y. Hayashi,
C. W. Nakagawa,
and A. Murasugi,
Environ.
Health Perspect. 65, 13
(1986). 2. W. Rauser, Ann. Rev. Biochem. 59, 61 (1990). 3. E. Grill, E.-L. Winnaker, and M. H. Zenk, Proc. Natl. Acad. Sci. USA 86, 6838 (1989). 4. R. N. Reese, R. K. Mehra, E. B. Tarbet, and D. R. Winge, J. Biol. Chem. 263, 4186 (1988).
5. C. T. Dameron, 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
R. N. Reese, R. K. Mehra, A. R. Kortan, P. J. Carroll, M. L. Steigerwald, L. E. Brus, and D. R. Winge, Nature 338, 596 (1989). N. Mutoh and Y. Hayashi, Biochem. Biophys. Res. Commun. 151, 32 (1988). R. K. Mehra and D. R. Winge, J. Cell. Biochem. 45, 30 (1991). T. R. Butt and D. J. Ecker, Mcrobiol. Rev. 51, 351 (1987). D. H. Hamer, Ann. Rev. Biochem. 55, 913 (1986). M. Inouhe, M. Hiyama, H. Tohoyama, M. Joho, and T. Murayama, Biochim. Biophys.
Acta 993, 51 (1989). J. D. Fikes, D. M. Becker, F. Winston, and L. Guarente, Nature 346, 291 (1990). H. Ito, Y. Fukuda, K. Murata, and A. Kimura, J. Bacterial. 153, 163 (1983). E. Racker, J. Biol. Chem. 184, 313 (1950). 0. H. Lowry, N. J. Rosenbrough, and R. J. Randall, J. Biol. Chem. 193, 265 (1951). J. L. Bennetzen and B. D. Hall, J. Biol. Chem. 257, 3018 (1982). M. F Dunn, H. Dietrich,
Biochemistry
A. K. H. MacGibbon,
S. C. Koerber,
and M. Zeppezauer,
21, 353 (1982).
17. T. L. Orr-Weaver,
J. W. Szostak, and R. J. Rothstein,
Proc. Natl. Acad. Sci. USA
78, 6354 (1981). 18. F. C. Volkert, D. W. Wilson, and J. R. Broach, Microbial. Rev. 53, 299 (1989). 19. K. T. Suzuki, S. Kawahara, H. Sunaga, and N. Shimojo, Toxicol. Applied Pharmacol.
105,413 (1990). 20. H. Sunaga, Y. Yamane, Y. Aoki, and K. T. Suzuki, Eisei Kagaku 35, 21 (1989). 21. R. K. Singhal, M. E. Anderson, and A. Meister, FASEB J. 1, 220 (1987). 22. S. Fogel, J. W. Welch, and E. J. Louis, Cold Spring Harbor Symp. Quant. Biol. XLIX, 55 (1984).
Received March 4, 1991; accepted April 29, 1991
of Utah Medical Center,
School of Medicine,
Salt Lake City, Utah
-ABSTRACT A cDNA expression library from Schizosaccharomyces pombe was transformed into Saccharomyces cerevisiae to screen for genes capable of conferring cadmium resistance to S. cerevisiae cells. The cDNA library was cloned into the S. cerevisiae expression vector pDB20 which is designed to express cDNAs via the constitutively-expressed promoter of the gene for alcohol dehydrogenase I (ADHl). Terminator and polyadenylation signals are also provided by the ADHl gene. Cadmium resistant colonies were shown to arise by a recombination event leading to the exchange of the S. pombe DNA with the chromosomal ADH 1 gene and a consequent dramatic increase in the ADH 1 gene expression due to the high copy number of the plasmid. The overexpression of ADHl effectively buffered the cells for cadmium ions by formation of Cd-ADH.
The fission yeast Schizosaccharomyces pombe detoxifies cadmium by a mechanism similar to that found in plants. Cadmium is sequestered within a complex involving isopeptides called phytochelatin or cadystin whose structure is (~Glu-Cys),,Gly [1,2]. Synthesis of these peptides is considered to be enzymatic by virtue of the isopeptide linkage within the dipeptide repeat. A transpeptidase has been isolated in plants that is capable of synthesizing (yGlu-Cys),Gly peptides from glutathione [3]. The intriguing aspect of the synthesis is the apparent regulation by metal ions. Sequestration of cadmium ions by the (~Glu-Cys),Gly peptides is the predominant mechanism conferring cadmium resistance in S. pombe [1,4]. The Cd:(yGluCys),Gly peptide complexes in fungi also contain labile sulfur present as a CdS crystalline lattice coated with the isopeptides [5]. Mutant cells defective in the production of sulfide were shown to be hypersensitive to cadmium salts [6]. Other potential mechanisms of metal resistance include restricted plasma membrane transport, facilitated efflux, and compartmentalization within the vacuole [7]. The genet-
*On leave from CSIRO Division of Biomolecular Engineering, Victoria, Australia. Address reprint requests and correspondence to: Dr. Dennis R. Winge, University Center, School of Medicine, 50 N. Medical Drive, Salt Lake City, UT 84132.
of Utah Medical
155 Journal of Inorganic Biochemistry, 44, 155 161 (1991) 0 1991 Elsevier Science Publishing Co., Inc., 655 Avenue of the Americas, NY, NY 10010 0162-0134/91/$3.50
156
W. Yu et al.
its of cadmium resistance in S. pombe are poorly described and no gene has been cloned from this organism that is functional in metal ion resistance. In order to clone genes conferring metal resistance in S. pombe, we performed heterologous complementation by transformations of Saccharomyces cerevisiae with an S. pombe cDNA expression library. Metal resistance is restricted in S. cerevisiae to copper salts as it is the main metal ion capable of inducing the CUP1 metallothionein gene [8,9]. Most strains of S. cerevisiae cells are very sensitive to cadmium-induced toxicity, although one cadmium-resistant strain has been isolated in which cadmium resistance is mediated by metallothionein expression (IO]. We predicted that the constitutive expression of certain cDNAs involved in cadmium resistance in S. pombe should impart metal tolerance to a cadmium-sensitive strain of S. crrevisiae. We presently report the characterization of cadmium resistant clones of S, cerp-
vi.siae.
MATERIALS
AND METHODS
Cell Culturing S. cerevisiae strain DYl.50 (a ura3-52 leu2-3,112 trpl-1 a&2-1 his3-11 canl-100) was the recipient strain for transformations with the S. pombe cDNA library provided by Dr. J. Fikes [ 111. The cells were transformed according to the procedure of Ito et al. [ 121. Colonies resistant to cadmium salts on solid medium were tested for metal resistance in liquid synthetic complete medium lacking uracil. The effective concentration of 50% growth inhibition (EC,,,) was defined as the concentration of cadmium salt that limited growth as determined by turbidity by 50 %. Polymerase
Chain
Reaction
Reaction conditions for PCR were those recommended by Perkin-Elmer/Cetus with the primers S’CTGCACAATATTTCAAG and SGATTGGAGACTTGACCA. Thermal cycling was 94°C for 1 mitt, 50°C for 1 min, and 72°C for 3 min for 30 cycles. Other Analyses Sequences in recombinant plasmids were determined directly from the plasmids utilizing the primers used for PCR and the U.S.B. Sequenase protocol. ADH activity was measured by the method of Racker 1131. Activity was quantified as the change in absorbance per min per ml with one unit defined as the change in optical density of I per min. Protein was quantified by the method of Lowry et al. 1141. Metal analysis was performed by atomic absorption spectroscopy. Size permeation chromatography was carried out on a Superdex 200 FPLC column from Pharmacia using an elution buffer of 20 mM Tris Cl, pH 7.4 containing 50 mM KCI. Fractions of E.5 ml were collected.
RESULTS Isolation
of Cadmium
Resistant
Transformants
The host strain chosen for the transformation, S. cerevisiae strain DY 150 is sensitive to cadmium chloride (Fig. 1A)? its growth is totally inhibited in the
ADHl
FIGURE 1. Growth of suspended in sterile water, cadmium and photographed DY150 cadmium-resistant DY150 retransformed with
AND CADMIUM
TOXICITY
157
S. cerevisiue strains on 0.5 mM cadmium plates. Strains were spotted onto solidified synthetic complete medium plus 0.5 mM after two days incubation at 30°C. Strains are DY150 (A), a transformant (B), a transformant after plasmid shedding (C), and the purified plasmid pDB20-Cd’ (D).
of 150 PM Cd(I1) on solid media. Following transformation with the cDNA library of S. pombe, a total of 21,000 Ura+ transformants from five separate transformations were obtained. These were replica plated onto solid media containing 0.15, 0.35, and 0.5 mM CdCl,. A higher than anticipated number of colonies (0.09%) were cadmium resistant. Eight of these seventeen colonies grew on plates containing 0.5 mM CdCl, (Fig. 1B). The EC,, of the eight most resistant transformants in liquid medium ranged between 110 and 300 PM, while the liquid EC,, of the parental strain was 20 PM. Only these eight transformants were further analyzed in this study. presence
Genetic
Analysis
of Cadmium
Resistant
Transformants
The resistance in each transformant was shown to be plasmid-borne by genetic analyses. Initially, Ura- auxotrophic cells were recovered that had shed the plasmid by growth of the transformants in rich media. Concomitant with the plasmid shedding was the loss of the cadmium resistance (Fig. 1C). Second, the plasmids from the transformants were purified by passage through E. co/i and subsequently retransformed into S. cerevisiue DY 150. Each purified plasmid transformed DY 150 to cadmium resistance phenotype (Fig. 1D) establishing that resistance is integrally associated with the plasmid. We have denoted these plasmids as pDB20-Cd’.
158
W. Yu et al.
Molecular
Analysis
of Plasmids
Determining
Cadmium
Resistance
In order to characterize the cDNA in pDB20-Cd’ plasmids we performed restriction site and PCR and DNA sequence analyses. The PCR primers were complementary to sequences in pDB20’s ADHl promoter and terminator. respectively, and together they were used to amplify the intervening sequences. Analysis of the PCR products by electrophoresis revealed that each recombinant plasmid contained an insen at the multiple cloning site of identical size at 1 kb while the parental vector, pDB20. yielded a PCR product of about 0.2 kb The conclusion drawn frc~l thck,c data 1s thar all inserts were of identical ttizc. Analyses with restriction enzyme digests were carried out to characterize the inserts. The iVot1 restriction sites that should have flanked the inserts were absent. In addition, Hind111 digests did not yield the expected completr liberation of the cDNA. Instead Hind111 fragments of approximately 0 15 kh and ii kh were obtained for each plasmid. identical sites in each plasmid were also observed in digests using SspI, EcoRV, and NanrKl restriction endonucleases. ‘This indicate\ Ihat ail thr pDB20-Cd’ plasmids arc identical. Finally, nucleotide sequence analysis was carried out utilizing the primers used previously for the PCR. This analysis showed that the onI> novel sequence present was an intact S. ceTvIsl(Ip ADNi gene 11-51. The S. ~~r?he I.DNX insert had. therefore, been exchanged for the ,S. cereuisiae RDH I gent h> recombination utilizing the homoiogous A DH1 sequences in the pDB3O plasmid. Flgurc _3 illustrates a scheme how this might come about. T’hl~ %%emu Would lead to
ADHI
TERMINATOR
PROMOTER
,;’
yDB20 library
FIGURE 2. Possible mechanism of insert replacement in pDB20 library. The structure of a typical pDB20 recombinant is shown. S. pombe cDNAs are inserted at the BstXI site of a polylinker in pDB20 whose structure is Hi&III-Not I-&t XI-Not I -: Hind111 A homologous double crossover with the chrornusomal S. cerevisiw ADff I locr~swould lcaci to replacement of S. pombe sequence< with an intact ADNI gene.
ADHl
AND CADMIUM
TOXICITY
159
transplacement of the S. pombe DNA to the chromosomal ADNl locus. PCR analysis on total DNA of yeast transformants did not reveal extra products other than the 1 kb ADHl fragment. Therefore, it is conceivable that further recombination may have led to restoration of the ADHl gene in the nucleus. Localization
of Cadmium
in Transformants
Molecular genetic evidence suggested a direct correlation between cadmium resistance and the presence of > 50 copies of the ADHl gene. The latter could be expected to lead to increased ADH protein levels. Measurement of the ADH enzymatic activity in transformants (grown in the absence of cadmium) revealed an activity 214 times higher than the ADH level in nontransformed cells. The relationship of cadmium resistance and enhanced ADH activity was examined by analyzing lysates of transformed cells grown in cadmium by size permeation chromatography. Elution fractions were measured for ADH activity, protein, and zinc and cadmium concentrations (Fig. 3). ADH activity was found in only one region (apparent Mr 120 kDa) that coincidentally contains the peak zinc. The coelution of Zn2+ and ADH is expected as ADH is a zinc metalloprotein. Examination of the cadmium profile, however, showed the major cadmium peak was also co-incident with ADH. The molar Cd:Zn ratio in the ADH peak tube (fraction 48) was 3.5, suggesting that most of the ADH molecules contain bound Cd2+ ions. . Additional Cd-binding components were observed in the column-excluded volume and a volume near the internal volume. These undefined Cd-complexes were observed in Cd-treated S. cereuisiae cells that were not transfected, so they were not investigated further. Transformants were grown in increasing amounts of cadmium and assayed for intracellular concentrations of Cd(I1) and Zn(I1) ions as well as ADH enzymatic
1.4 r
IO
20
30
40 FRACTION
FIGURE
50
60
70
00
90
NUMBER
3. Chromatography of a cell lysate of cadmium-resistant cells. A clarified cell extract from a culture of cells pregrown to mid-logarithmic phase and incubated subsequently with 0.5 mM CdCl, for 24 hr. The extract was applied to Superdex 200 for FPLC at a flow rate of 1 ml/min. Elution fractions of 1.5 ml were analyzed for ADH activity as the absorbance change at 340 nm per ml per min (--), A,,, nm (--), zinc (- - - -), and cadmium (- * -) concentrations as pg per ml.
160
W. Yu et al.
activity (Table 1). Increasing amounts of cadmium in the culture medium resulted in increased intracellular cadmium levels but an inverse relationship to ADH activity. ADH activity is constitutive and should be directly related to total protein. However. as intracellular cadmium increased, even by 40 times, ,4DH activity dropped to almost one-third of its initial value. The zinc concentration in the lysate paralleled ADH levels. Thus, yeast Cd-ADH has a reduced specific activity relative to Zn--ADH as has been reported with the mammalian ADH enzyrne [ 161.
DISCUSSION A relatively frequent (O.OY%) event during transformations with the pDB20 library leads to cadmium resistance by an intracellular increase in the copy number of the ADH 1 gene. The mechanism of the copy number increase is presumably by homologous recombination with the chromosomal ADHl genr and pDB20 as shown in Figure 1. Another mechanism that could be utilized is ,gap repair of fragmented DNA as described by &-r-Weaver et al. 1I?]. For example, simultaneous double strand breaks in the ADHI promoter and terminator of pDB2.0 could bc repaired in S. cerevkiae using the chromosomal ilL)H 1 gene as a template. This would obviously lead to the introduction of the intact r3DH 1 gene into pDB20. Alcohol dehydrogenase 1. synthesized at high levels by \,irtue of its ,opy number in this instance. can bind cadmium and provide a considerable degree of protection to that metal. The 2 p-based plasmids replicate to greater than 50 copiex per cell [ I8 1. Alcohol dehydrogenase ib known in animal cells to bind Cd(H) Ions in animal administered cadmium salts. presumably by Zn displacement 1191. The sequestration of Cd(lI) by alcohol dehydrogenase and glutathione appear to !~e the initial Cd butfering components prior TVthe induction of metallothionein [.X.2 11. The present results illustrate that a substantial level of resistance to Cd toxicity can be achicveci by an increase in the ADN gene copy number. It is likei! that over-expression of genes encoding other Zn-metalloproteins may also afford a similar protrction against heavy metal toxicity. A cell with only a single copy of the A DNl gene may not have the opportunity to exploit this effect in cadmium resistance unless it K to amplify that locus. Facile Limpiification of a chromosomal gene with at least two tandem repeats, such as the S. cerevisiae metallothionein CUPI locus. can occur under selective pressure through meiotic gene conversIon 12’2/. The use of heterologou5 complementation for cloning genes involved in metal resistance is an obvious and powerful approach. The success of the strategy assumes that the resistance process involves a single gene product and necessitates constitu-
TABLE 1. ADH Activity and Cd and Zn Concentrations in Lysates of Cells Grown in Various Cadmium Concentrations [Cd’+ 1 in Culture CcM
Protein in:! in11
Lysate [Cd’+ 1 Lysate (Zn’” j -._..-. _.-____~______... pg,/mg Protein
ADW Activit) units, rng Protein
ADHl
AND CADMIUM
TOXICITY
161
tive or induced expression of the gene library. The ADHl promoter may be used in expression plasmids if the termination sequences are provided from a non-ADH 1 gene to minimize chances of homologous recombination. We thank Dr. John Fikes and Dr. David Stihman for supplying the pDB20 library and strain DYISO, respectively. This work was supported by Public Health Service grant No. ES 03817 from the National Institutes of Health to DR W. IGM was the recipient of a Fulbright award.
REFERENCES 1. Y. Hayashi,
C. W. Nakagawa,
and A. Murasugi,
Environ.
Health Perspect. 65, 13
(1986). 2. W. Rauser, Ann. Rev. Biochem. 59, 61 (1990). 3. E. Grill, E.-L. Winnaker, and M. H. Zenk, Proc. Natl. Acad. Sci. USA 86, 6838 (1989). 4. R. N. Reese, R. K. Mehra, E. B. Tarbet, and D. R. Winge, J. Biol. Chem. 263, 4186 (1988).
5. C. T. Dameron, 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
R. N. Reese, R. K. Mehra, A. R. Kortan, P. J. Carroll, M. L. Steigerwald, L. E. Brus, and D. R. Winge, Nature 338, 596 (1989). N. Mutoh and Y. Hayashi, Biochem. Biophys. Res. Commun. 151, 32 (1988). R. K. Mehra and D. R. Winge, J. Cell. Biochem. 45, 30 (1991). T. R. Butt and D. J. Ecker, Mcrobiol. Rev. 51, 351 (1987). D. H. Hamer, Ann. Rev. Biochem. 55, 913 (1986). M. Inouhe, M. Hiyama, H. Tohoyama, M. Joho, and T. Murayama, Biochim. Biophys.
Acta 993, 51 (1989). J. D. Fikes, D. M. Becker, F. Winston, and L. Guarente, Nature 346, 291 (1990). H. Ito, Y. Fukuda, K. Murata, and A. Kimura, J. Bacterial. 153, 163 (1983). E. Racker, J. Biol. Chem. 184, 313 (1950). 0. H. Lowry, N. J. Rosenbrough, and R. J. Randall, J. Biol. Chem. 193, 265 (1951). J. L. Bennetzen and B. D. Hall, J. Biol. Chem. 257, 3018 (1982). M. F Dunn, H. Dietrich,
Biochemistry
A. K. H. MacGibbon,
S. C. Koerber,
and M. Zeppezauer,
21, 353 (1982).
17. T. L. Orr-Weaver,
J. W. Szostak, and R. J. Rothstein,
Proc. Natl. Acad. Sci. USA
78, 6354 (1981). 18. F. C. Volkert, D. W. Wilson, and J. R. Broach, Microbial. Rev. 53, 299 (1989). 19. K. T. Suzuki, S. Kawahara, H. Sunaga, and N. Shimojo, Toxicol. Applied Pharmacol.
105,413 (1990). 20. H. Sunaga, Y. Yamane, Y. Aoki, and K. T. Suzuki, Eisei Kagaku 35, 21 (1989). 21. R. K. Singhal, M. E. Anderson, and A. Meister, FASEB J. 1, 220 (1987). 22. S. Fogel, J. W. Welch, and E. J. Louis, Cold Spring Harbor Symp. Quant. Biol. XLIX, 55 (1984).
Received March 4, 1991; accepted April 29, 1991
