Zinc-Regulated Genes in Saccharomyces cerevisiae Revealed by Transposon Tagging

Corresponding author: Daniel S. Yuan, Johns Hopkins University School of Medicine, 725 N. Wolfe St., Baltimore, MD 21205-2185., dyuan{at}jhmi.edu (E-mail)

Communicating editor: A. G. HINNEBUSCH

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The biochemistry of human nutritional zinc deficiency remains poorly defined. To characterize in genetic terms how cells respond to zinc deprivation, zinc-regulated genes (ZRG's) were identified in yeast. Gene expression was probed using random lacZ reporter gene fusions, integrated by transposon tagging into a diploid genome as previously described. About half of the genome was examined. Cells exhibiting differences in lacZ expression on low or moderate (0.1 vs. 10 µM) zinc media were isolated and the gene fusions were sequenced. Ribonuclease protection assays demonstrated four- to eightfold increases for the RNAs of the ZAP1, ZRG17 (YNR039c), DPP1, ADH4, MCD4, and YEF3B genes in zinc-deficient cells. All but YEF3B were shown through reporter gene assays to be controlled by a master regulator of zinc homeostasis now known to be encoded by ZAP1. ZAP1 mutants lacked the flocculence and distended vacuoles characteristic of zinc-deficient cells, suggesting that flocculation and vacuolation serve homeostatic functions in zinc-deficient cells. ZRG17 mutants required extra zinc supplementation to repress these phenotypes, suggesting that ZRG17 functions in zinc uptake. These findings illustrate the utility of transposon tagging as an approach for studying regulated gene expression in yeast.

ZINC is one of the principal trace elements in biology, with structural or enzymatic roles in hundreds of proteins (VALLEE and FALCHUK 1993 ). Zinc finger proteins are especially numerous in eukaryotic genomes and play many roles in protein-DNA or protein-RNA interactions (BERG and SHI 1996 ). The enzymatic repertoire of zinc is extraordinary and includes many important phosphatases and metalloproteinases (LIPPARD and BERG 1994 ). Zinc also functions in a large number of oxidoreductases and transferases (VALLEE and FALCHUK 1993 ), and additional roles for zinc in cysteinyl transfer and methylation reactions are now known (MATTHEWS and GOULDING 1997 ).

Many studies, dating back to 1869, have confirmed the importance of zinc in nutrition (reviewed in VALLEE and FALCHUK 1993 ). In humans, zinc deficiency has gradually come to be recognized as a clinically significant and common form of malnutrition, particularly in chronically ill patients and in the Third World (AGGETT and COMERFORD 1995 ; BHUTTA et al. 1999 ). The clinical manifestations of zinc deficiency are diverse, with effects on immune function, epithelial integrity, appetite, cognitive function, and embryonic development (WALSH et al. 1994 ). Unfortunately it remains unclear how zinc deficiency relates to any of these changes at a biochemical level, even though a large variety of correlations have been made (e.g., SHANKAR and PRASAD 1998 ).

Studies of the biochemistry of zinc deficiency have been attempted for many years. "Throughout the period of discovery of zinc enzymes, there has been a diligent search for alterations of their activities in organs and tissues of zinc-deficient animals. The results have been almost uniformly disappointing" (VALLEE and FALCHUK 1993 , p. 81). Various biochemical indicators of zinc status continue to be studied in a research setting (e.g., GRIDER et al. 1990 ; LICASTRO et al. 1996 ; BECK et al. 1997 ). However, there still appears to be no way of predicting which zinc proteins are affected most by zinc deprivation. A promising new approach involves the identification of novel genes whose expression is regulated by zinc status. In particular, subtractive hybridization and differential display techniques have been used to identify several zinc-regulated genes in rats rendered zinc deficient or zinc replete through dietary manipulation (SHAY and COUSINS 1993 ; BLANCHARD and COUSINS 1996 ). It is not yet clear why or how these genes are regulated by zinc status, however.

This article describes a novel application of yeast genetics to the problem of identifying zinc-regulated genes. The original impetus for this study was the emerging realization that many features of iron and copper metabolism are conserved between yeast and mammals (reviewed in ASKWITH and KAPLAN 1998 ; CULOTTA et al. 1999 ), and indeed this conservation is now being extended to zinc (EIDE 1997 ). In this study, zinc-regulated genes in yeast were identified on a quasi-genomic scale using an existing transposon tagging technique (BURNS et al. 1994 ). A total of 17 genes were found to be differentially regulated at least 10-fold, as assessed by randomly generated translational fusions with the lacZ reporter gene. Five were subsequently shown to be regulated by the same transcription factor. One of these appears to encode a novel protein with a role in zinc uptake. These findings provide new starting points for characterizing the biochemical effects of zinc deficiency in eukaryotic cells.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Plasmids and yeast strains:
See Table 1 and Table 2, respectively.

Preparation of yeast growth media:
Glassware was scrubbed with Alconox detergent and acid washed before use; plasticware was used without further treatment. A defined growth medium lacking added zinc, iron, copper, phosphate, dextrose, and amino acids was prepared as a custom-made powder (Bio101, Vista, CA; following DANCIS et al. 1994 ). All but the dextrose, zinc, and selectable markers were added back with the aid of heating to 70° to restore the composition of standard defined medium. The resulting solution (prepared as a 5x stock solution) was cooled to 40°, sterilized by filtration, and stored in the dark at room temperature for up to 1 year. Stock solutions of glucose (dextrose), MES buffer (2-[N-morpholino]ethanesulfonic acid hemisodium salt; Sigma, St. Louis), and nutritional supplements were treated with Chelex-100 resin and sterilized by filtration. Solid medium was prepared with 10 g/liter agarose (Biochemika grade; Fluka, Buchs, Switzerland), autoclaved for 15 min, and cooled to 50°. This was supplemented with 2% w/v glucose, 50 mM MES, nutritional supplements, and the indicated amounts of zinc (as zinc sulfate) immediately before dispensing in petri dishes. Liquid medium was prepared without agarose and refiltered instead of autoclaved.

YPD growth medium was prepared as described (SHERMAN 1991 ) and synthetic defined medium was prepared from yeast nitrogen base, glucose, and supplements as recommended (Bio101), except that glucose was added as a separately autoclaved or filtered stock solution after the medium was cooled to 50°.

Yeast transformation with the transposon insertion library:
Cells of the diploid strain YPH274 (SIKORSKI and HIETER 1989 ) were transformed to leucine prototrophy with NotI-digested DNA that had been amplified once (Maxiprep; QIAGEN, Valencia, CA) from pool 21 of a library of transposon insertions (BURNS et al. 1994 ), graciously provided by the laboratory of Michael Snyder (Yale University). The digested DNA was used without heat inactivation or further purification. A high-efficiency transformation protocol was used routinely (GIETZ and WOODS 1994 ; http://www.umanitoba.ca/faculties/medicine/units/human_genetics/gietz/Trafo.html). Cells were washed twice with low-zinc medium before plating on low-zinc medium lacking added leucine. Using 3.4 µg digested DNA with 25 pooled aliquots of cells, 3500 colonies on each of 23 Leu-selective plates were obtained, indicating a transformation efficiency of 2 x 10⁴ colonies/µg digested DNA.

Colony assays for zinc-regulated lacZ activity:
Each transformation plate was replica plated to (1) a nylon membrane (Biotrans, 1.2 µm, 82-mm circles; ICN, Costa Mesa, CA) laid on a low-zinc plate, (2) a second membrane laid on low-zinc plates supplemented with 10 µmol/liter zinc sulfate, and (3) a master YPD plate. Velveteen squares used for replica-plating were scrubbed clean by hand, machine washed in hot water with chlorine bleach and a commercial laundry detergent, rinsed in hot water for three cycles, dried in a clothes dryer, and autoclaved in foil. To obtain the best replica fidelity, both the source and destination plates were allowed to dry out to 80% of their former thickness before use, and the velveteen was underlaid with two circles cut from gel blotting paper (GB004; Schleicher and Schuell, Keene, NH). For the genetic screen, nylon membranes were boiled in 1 mM EDTA before rinsing with water and autoclaving, although similar results were obtained later using untreated membranes.

After 24 hr of incubation at 30° to elicit the color differences associated with zinc status (see Fig 2B), membranes were transferred to a surgical clamp and dipped twice for 10 sec each into liquid nitrogen to permeabilize the cells (Matchmaker protocol; Clontech Laboratories, Palo Alto, CA). The frozen membranes were gently thawed over a small flame, laid on agar plates containing X-gal (BURNS et al. 1994 ) with 10 mM sodium azide, and incubated at 30° in a closed bag for up to 4 wk. The pink or orange pigmentation associated with the ade2 genetic marker in zinc-treated cells faded completely after a day of incubation and did not interfere with color development. The sodium azide was added to inhibit artifacts due to bacterial overgrowth and had no effect on yeast-associated lacZ activity.

View larger version (23K): [in this window] [in a new window] [Download PPT slide]   Figure 1. Pigmentation phenotypes in zinc-deficient cells. See MATERIALS AND METHODS. Wild-type cells (YPH252) were grown to stationary phase in 1-ml cultures containing the indicated nominal concentrations of zinc cultures. Cells were photographed after transfer to a microtiter plate.

View larger version (174K): [in this window] [in a new window] [Download PPT slide]   Figure 2. Illustration of the genetic screen for ZRGs. (A) Creation of a ZRG-lacZ fusion construct by insertion of a modified Tn3 transposon into yeast genomic DNA. The arrow schematizes the ZRG promoter (adapted from BURNS et al. 1994 ). (B) Zinc-dependent pigmentation. Shown is a representative pair of replicas grown on low-zinc (no added zinc) or moderate-zinc (10 µmol/liter added zinc) medium. (C) Zinc-dependent lacZ activity. Shown are colonies from the same plates in B after cell permeabilization in liquid nitrogen and incubation with the lacZ substrate X-gal at 30° for 2 wk. Arrowheads highlight colonies displaying differences in lacZ activity.

Isolation of cells differentially expressing lacZ activity:
After development of the X-gal color for 2 wk, the membranes were photographed using Ektapan (Kodak, 4162) film and the resulting pairs of negatives were superimposed with a slight offset and examined by eye against a clear incandescent lightbulb. By applying strips of removable tape, the negatives were readily scanned for colonies putatively exhibiting differential lacZ expression. Colonies on the master plate were located by comparison with prints from the negatives. To purify the clones and document zinc-regulated lacZ expression, clones of interest were dispersed into 1 ml of low-zinc, Leu-selective medium and 1 µl of this suspension spread as sectors on another low-zinc, Leu-selective plate. After colonies reached full size, the replica-plating procedure was repeated. About 70% of the clones exhibited perceptible differences in X-gal color in the two growth conditions and were kept for further study.

Identification of sequences upstream of genomic lacZ insertions by inverse PCR:
Yeast genomic DNA was prepared from each purified clone essentially as described but at 1/10 scale (PHILIPPSEN et al. 1991 ). Templates for inverse PCR were prepared by digesting aliquots of 200 ng genomic DNA (2 µl) with 0.4 units NlaIII or NspI enzyme in its recommended reaction buffer (New England Biolabs, Beverly, MA) in a total volume of 50 µl (giving a DNA concentration of 4 µg/ml) for 1 hr at 37°. After heat inactivation at 65°, the mixture was supplemented with ATP (to 1 mM), KCl (to 30 mM to promote intramolecular circularization; DUGAICZYK et al. 1975 ), and 0.5 units T4 DNA ligase (GIBCO BRL, Grand Island, NY) and incubated overnight at 14°. A 10-µl aliquot of this reaction mixture was diluted fivefold with 8 standard units of KlenTaq1 enzyme (Ab Peptides, St. Louis), components of the KlenTaq1 enzyme buffer without magnesium, deoxynucleotides, and 50 pmol each of the oligonucleotides lacZ-5'INV (GGCGATTAAGTTGGGTAACGCCAGGG, directed retrograde toward upstream promoter sequences) and lacZ-3'INV (CCGACTACACAAATCAGCGA, directed anterograde toward the first NlaIII restriction site in the lacZ coding sequence). After 40 cycles of PCR, reaction mixtures were cleaned up with PCR Select II columns (5 Prime 3 Prime, Boulder, CO) and analyzed by agarose gel electrophoresis. PCR products were sequenced using a cycle sequencing kit (GIBCO BRL) and a 96-well thermal cycler (MJ Research, Watertown, MA) in conjunction with the oligonucleotides lacZ-5'SEQ (CGTTGTAAAACGACGGGATCCCCCT; BURNS et al. 1994 ) or lacZ-5'INV (see above). The sequences obtained were matched to yeast sequences in GenBank or the Saccharomyces Genome Database using the BLAST program (ALTSCHUL et al. 1990 ).

Preparation of RNA probes for ribonuclease protection assays:
Sequences spanning the lacZ fusion sites for each zinc-regulated gene (ZRG) were amplified from genomic DNA samples corresponding to each ZRG clone using the lacZ oligonucleotide GGGAAAGCCGGCtaatacgactcactatagggATTAAGTTGGGTAACGCCAGGGT (T7 promoter in lowercase letters and lacZ sequences underlined) and a ZRG oligonucleotide (with CCCGAGCTC preceding sequences from each ZRG) designed to amplify a fragment of defined length (150 bases for ZRG1, 160 bases for ZRG2, etc.). Cycle numbers were optimized for each reaction to avoid saturation. All products were pure and of the expected size except for the ZRG12 fragment, which was gel purified to remove a smaller contaminant. Fragments for ZRG's 1, 2, 4, 6, 10, and 17 were digested with NgoMIV and SacI for directional cloning in pRS416, which was digested with the same enzymes to excise the endogenous T7 promoter. The resulting plasmids were validated by sequencing, linearized with SacI, and purified for in vitro transcription. Fragments for the other ZRG's were used directly without cloning. Fragments for TDH3 were synthesized with the oligonucleotides GGGAAAGCCGGCtaatacgactcactatagggATGGTAGAGTAACCGTATTCG (T7 promoter sequence in lowercase letters and TDH3 sequences underlined) and CCCGAGCTCCCTCTGACTTCTTGGGTGAC, designed to amplify 120 bases near the 3' end of the TDH3 coding sequence.

RNA probes labeled with [-³²P]CTP were synthesized at 1/4 scale with 10 µM total CTP using an in vitro transcription kit (Maxiscript; Ambion, Austin, TX). The TDH3 probe was synthesized with 500 µM total CTP. Probes of validated length were gel purified as recommended (RPA III; Ambion) except that the elution step was performed in 1/2 volume and with two freeze-thaw cycles to hasten elution.

Ribonuclease protection assays (RPAs):
Total yeast RNA was prepared from matched low- and high-zinc cultures (100 ml) of the parental diploid strain, YPH274. The protocol used was chosen to allow the concurrent isolation of small RNAs (WISE 1991 ). RPAs were carried out with a kit (RPA III; Ambion). For each ZRG, 10 µg of total RNA from low-zinc or high-zinc cells and 10,000 cpm each of the purified ZRG and TDH3 probes were precipitated and solubilized in hybridization buffer. A negative control containing the same probes but no added RNA was prepared for each ZRG. Hybridization was carried out at 42° for 12–16 hr. Unhybridized RNAs were digested with RNase A/T1 as directed except that all samples were kept at 15° and 10 µg Torula yeast RNA was added to the RNase solution just before mixing with the negative controls. After 60 min, all samples were processed for electrophoresis and separated on denaturing polyacrylamide gels (Novex, Encinitas, CA). A 10-bp DNA ladder was used as standard (GIBCO BRL end labeled using T4 polynucleotide kinase); RNA size was determined to be the DNA size plus 8%. After drying the gels onto filter paper, signals were quantitated by phosphorimaging (Storm; Molecular Dynamics, Sunnyvale, CA), using the negative controls to determine the level of background signal in adjacent bands.

Cloning and disruption of the ZAP1 gene:
The ZAP1 open reading frame and 5' and 3' flanking sequences (283 and 309 bases, respectively) were cloned as a PCR product into the bacterial cloning vector pDirect (Clontech Laboratories, Palo Alto, CA), designed for ligation-independent cloning (ASLANIDIS and DE JONG 1990 ), as directed. Genomic DNA from strain YPH252 was used as the PCR template. This yielded pDY195. A ZAP1 disruption construct was prepared from pDY195 by inserting the URA3 gene flanked by Klenow-blunted HindIII sites (prepared from plasmid B728, from T. Donahue and M. Cigan, National Institutes of Health) in reverse orientation into the MscI site of ZAP1, 16 codons into the open reading frame. This yielded pDY233. The fragment released by digestion with ClaI and NotI was used to transform yeast to uracil prototrophy in synthetic defined medium. Transformants were validated by colony PCR using ZAP1 and URA3 oligonucleotides flanking the 5' end of the transforming fragment. A ZAP1 deletion construct was created and validated as for pDY233 except that ZAP1 sequences lying between MscI and NsiI (containing all but the first 16 codons of the open reading frame) were removed by digestion with these enzymes and treatment with T4 DNA polymerase, and the URA3 fragment was prepared by digestion with EcoRI and blunting with Klenow fragment. The resulting plasmid, pDY276, was used as for pDY233. Correct integration at the 3' end was also verified in the same manner as at the 5' end.

Cloning of synthetic promoter-lacZ fusion constructs:
A centromeric lacZ reporter vector, pDY269, was prepared from the high-copy number lacZ reporter vector, YEp368R (MYERS et al. 1986 ), by subcloning the lacZ gene and flanking sequences into the TRP1-marked plasmid pRS414 (SIKORSKI and HIETER 1989 ) and inserting the ligation-independent cloning site of pDirect (see above) into the SmaI site upstream of the lacZ gene. The construction and use of this vector will be described in detail elsewhere. DNA fragments of interest were synthesized by PCR from YPH252 (SIKORSKI and HIETER 1989 ) genomic DNA. Oligonucleotides were designed to create an in-frame fusion of yeast sequences containing a putative promoter and an initiation codon with the promoterless lacZ gene lacking a translation initiation codon. Ligation-independent cloning (ASLANIDIS and DE JONG 1990 ) was accomplished by direct transformation of competent yeast with a mixture of suitably digested vector and insert fragments. Colonies growing on tryptophan-free medium were screened directly for zinc-regulated lacZ expression using the replica-plating techniques described for the genetic screen (above), maintaining selection for the TRP1 plasmid. Representative colonies were cloned for further analysis.

Plasmid rescue:
Yeast cells containing plasmids of interest were grown to saturation in 20-ml cultures of synthetic defined medium with the appropriate selectable markers. Pelleted cells were digested for at least 1 hr at 37° with 100 µg Zymolyase 100-T spheroplasting enzyme (ICN), in 1 ml of a buffer containing 1.2 M sorbitol, 40 mM sodium phosphate, pH 7.0, 0.5 mM magnesium chloride, and 0.2% v/v 2-mercaptoethanol. After centrifugation at 1000 x g for 2 min, spheroplasts were subjected to a plasmid miniprep protocol (QIAprep; QIAGEN). The eluate was used to transform Escherichia coli.

Quantitative assay of zinc-regulated lacZ expression:
Cells from relatively fresh plates were washed twice in low-zinc medium and used to inoculate paired 10-ml cultures to calculated optical densities of 0.02 or 0.002 OD₆₀₀/ml. Zinc was then added (100 µmol/liter zinc sulfate) to the second culture. Cells were typically in late exponential growth phase at the end of a 24-hr growth period at 30° on a rotary shaker. Longer culture times were used as needed. After chilling to 0°, cells were collected in microcentrifuge tubes and stored in a buffer containing 5% glucose and 50 mM sodium citrate, pH 6.5. Measurements of lacZ activity were as described (GUARENTE 1983 ) except that 10 mM sodium azide was included. In control experiments this amount of azide had no effect on lacZ activity. OD₆₀₀ values were corrected for nonlinearities but not for zinc-dependent effects (1 OD₆₀₀ of diploid cells corresponded to 33.5 x 10⁶ or 38.8 x 10⁶ cells (± 1 x 10⁶) for cells grown in low- or high-zinc medium, respectively). Results are expressed as Miller units (1000 x OD₄₂₀/OD₆₀₀/min; MILLER 1972 ). Where presented, means and standard deviations represent measurements from six to eight aliquots of the same culture; otherwise means of duplicate aliquots are given.

Determination of cellular phenotypes affected by zinc:
Cells were prepared as for quantitative lacZ assays except as noted.

Pigmentation: Cells were grown from 0.02 OD₆₀₀/ml in 1-ml cultures for 40–48 hr and transferred to a microtiter plate for photography.

Flocculence: Cells were grown as for cell pigmentation assays and vigorously swirled without rotating the plate, then allowed to settle for 2 min before photography.

Vacuolization: Cells were grown as indicated and suspended at room temperature in glucose-citrate buffer. Representative fields of cells were photographed with a digital camera under Nomarski optics at x1600.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Validation of a low-zinc growth medium:
Methods for depleting cells of zinc are fundamental to studies of zinc deficiency. A low-zinc growth medium for this purpose was prepared by omitting the 1.3 µM zinc sulfate that is present in standard defined medium (SHERMAN 1991 ) and taking precautions against zinc contamination (see MATERIALS AND METHODS). Cells with the ade2 genetic marker lost their usual coloration (WEISMAN et al. 1987 ) when grown in this medium (Fig 1), providing a useful indicator of zinc status. Subtle darkening was evident with cells grown in medium containing 100 nmol/liter added zinc, suggesting that concentration of bioavailable zinc in the low-zinc medium was on the order of 0.1 µM. Cells were still able to grow in medium containing concentrations of zinc up to 1000 µM (Fig 1), although growth was initially delayed at 1000 µM and above. The ability of these cells to tolerate ambient zinc concentrations spanning the range 0.1–1000 µM demonstrated that robust mechanisms exist for zinc homeostasis in these cells.

Identification of genes differentially regulated by zinc status:
In 1994 a procedure for identifying differentially expressed genes on a genomic scale in yeast was described (BURNS et al. 1994 ; GOFFEAU 1994 ). This procedure involves the random insertion of the lacZ gene into the yeast genome for use as reporter gene fusion constructs. Specifically, yeast genomic DNA is subjected to transposition of a lacZ::LEU2::bla cassette, flanked at either end by inverted repeats of the 38-base mini-Tn3 transposon (m-Tn3), in E. coli (Fig 2A). The resulting library of random lacZ insertions is then digested with NotI to release the yeast genomic DNA from its vector and is used to transform a leu2/leu2 diploid yeast strain to leucine prototrophy. It is expected that the yeast genomic sequences flanking the lacZ::LEU2::bla cassette will mediate integration of the cassette into one copy of the diploid yeast host genome. In a minority of cases, the lacZ gene (lacking an initiation codon) will be inserted in frame within a translated open reading frame, resulting in lacZ activity driven by the upstream promoter. Transformants exhibiting regulated lacZ activity can then be studied to identify the gene fused to lacZ.

The efficiency of this procedure was improved through a number of technical modifications (see MATERIALS AND METHODS). One modification was to assess differential lacZ expression by replica-plating colonies onto pairs of plates, rather than by streaking colonies out individually. Thus, only a few hours were needed to process the 80,000 transformants in this study. A second modification was to sequence the lacZ insertion sites in individual clones by amplifying them directly by PCR from circularized genomic DNA fragments, rather than by cloning them through plasmid rescue. Two transformations and two plasmid preparations were saved in this way for each of the 100+ transformants that were later analyzed.

To apply this procedure to the identification of zinc-regulated genes, transformants were first grown on low-zinc (0.1 µM) medium and then replica plated in succession to low-zinc and moderate-zinc (10 µM) media. Growth of cells on the low-zinc solid medium resulted in loss of pigmentation (Fig 2B), consistent with the loss of pigmentation observed in cells grown in low-zinc liquid medium (Fig 1). lacZ reporter gene activity was detected in 8000 colonies, or 10% of transformants, similar to the fractions reported elsewhere (BURNS et al. 1994 ; ERDMAN et al. 1998 ). Of these 8000 colonies, 105, or 1%, exhibited visibly different levels of lacZ expression when subcloned and reassayed. The fact that 99% of the lacZ fusions examined in this genetic screen were not obviously affected by a 100-fold range in ambient zinc concentrations (Fig 2C) was reassuring, in view of a previous report describing markedly decreased protein content in zinc-deficient cells (OBATA et al. 1996 ). Because many zinc-dependent enzymes are involved in RNA biosynthesis, including RNA polymerase II (e.g., THURIAUX and SENTENAC 1992 ), it was conceivable that extreme zinc deprivation would result in global decreases in lacZ expression. That this was not observed suggested that zinc deprivation has specific effects on gene expression, at least in its early stages.

The 105 clones expressing zinc-regulated lacZ activity were analyzed by sequencing the lacZ insertion sites in these cells and measuring levels of lacZ expression after growth in low-zinc (0.1 µM) or high-zinc (100 µM) liquid media. Unambiguous identification of the lacZ insertion sites was achieved by BLAST searches in GenBank in almost every case. The only exceptions were one clone that contained two independent lacZ insertion sites and two clones in which the NlaIII site used for fragmenting genomic DNA was situated within a few bases of the lacZ cassette. In all identified clones the lacZ gene was situated in the same reading frame as upstream portions of the disrupted gene, as defined by the presence of an upstream ATG initiation codon without an interposed termination codon.

Assessment of the genetic screen:
The number of independent lacZ-positive colony transformants examined in this genetic screen, 8000, compared favorably with the 6000 genes in the Saccharomyces cerevisiae genome. As a first approximation Poisson statistics would imply that at least half of the 6000 genes were examined. Some examples of saturation of the genetic screen were also found: ZRG17 was isolated 6 times as 5 different insertions (Table 3); ENA1/PMR2, found as a cluster of 5 highly similar genes in the genome of strain S288C (an ancestor of the parental strain used in this study; WIELAND et al. 1995 ), was isolated 9 times as 5 different insertions (regulated weakly; not shown); and the TyA gene in the Ty1 retrotransposon, represented by 33 copies in the genome (KIM et al. 1998 ), was isolated 31 times in 17 different insertions (also regulated weakly, not shown). Together, these observations suggest that roughly half of the genes in the genome were examined.

Analysis of ZRG expression by ribonuclease protection assays:
Clones exhibiting a 10-fold or greater range of lacZ expression were arbitrarily chosen for further study. The 17 ZRG's represented by these clones are listed in Table 3.

To assess the contribution of transcriptional regulation (or other mechanisms affecting RNA abundance) to the regulated expression of ZRG-lacZ fusions, ribonuclease protection assays (RPAs) were undertaken for each of the ZRG's. The total RNA used in these experiments was derived from cells grown in the same low- or high-zinc media as for the quantitative lacZ assays. RNA was isolated using a procedure developed to ensure the concurrent isolation of small molecular weight species (WISE 1991 ). The internal standard chosen was TDH3 (encoding glyceraldehyde-3-phosphate dehydrogenase), a gene that was fortuitously isolated in the genetic screen as a strongly expressed control that was minimally sensitive to zinc status (1270 and 770 Miller units in low- and high-zinc media, respectively; TDH3 was disrupted by the modified transposon after codon 6).

Of the 17 ZRG's, ZRG's 1, 5, 7, 10, 16, and 17 were most clearly regulated by zinc at the level of RNA expression (Fig 3). These were induced 3.8-, 7.6-, 5.4-, 5.7-, 8.7-, and 6.6-fold, respectively, in zinc-deficient cells relative to TDH3. The ZRG7 data provide the first evidence that ZRG7 (YEF3B) is in fact expressed (cf. MAURICE et al. 1998 ; SARTHY et al. 1998 ). ZRG6 and ZRG14 RNAs were induced in zinc excess as expected in one experiment, but for unknown reasons this was not reproducible in a second experiment (not shown).

View larger version (82K): [in this window] [in a new window] [Download PPT slide]   Figure 3. Ribonuclease protection assays of zinc-regulated ZRG expression. The indicated radiolabeled ZRG probes (top bands) and a radiolabeled TDH3 probe (lower bands) were used in ribonuclease protection assays to probe RNA from zinc-deficient cells (left lane), RNA from zinc-replete cells (center lane), or no hybridizing RNA (right lane). Numerals denote the expected location of the hybridization signal for the ZRG indicated.

Of the remaining ZRG's, RNAs for ZRG's 8, 11, 12, 13, and 15 were induced 1.9-, 3.6-, 1.9-, 1.9-, and 2.0-fold, respectively, in zinc-deficient cells relative to TDH3 (Fig 3). Of the four ZRG's encoding unusual open reading frames, ZRG2 RNA (68 codons) was barely detectable. (In Fig 3 the ZRG2 hybridization was carried out without TDH3 probe, and imaging thresholds were decreased 10-fold.) ZRG3 RNA (9 codons) could not be confidently assessed due to unexplained and variable excesses in probe length, suggesting secondary structure in the probe; an A-rich sequence (43/51 bases) lies within the expected probe. ZRG4 RNA (12 codons) was undetectable. Interestingly, however, ZRG9 RNA (25 codons, antiparallel to the STU2 open reading frame) was clearly detected (Fig 3), with a 2.3-fold induction in zinc-deficient cells.

Identification of the ZRG10 gene product as a regulator of zinc-regulated genes:
During the course of these studies, ZHAO and EIDE 1997 described the ZAP1 gene, encoding a transcriptional activator of the ZRT1 and ZRT2 genes and also of ZAP1 itself. ZRT1 and ZRT2 had been shown previously to encode high- and low-affinity zinc uptake transporters whose expression is highly regulated by zinc (ZHAO and EIDE 1996A , ZHAO and EIDE 1996B ). In zinc-deficient cells expression of ZRT1 and ZRT2 is elevated, consistent with a role for these transport proteins in maintaining cellular zinc homeostasis in the face of fluctuations in nutritional zinc bioavailability. The ZAP1 protein functions in vitro as a DNA-binding protein that binds specifically to a degenerate 11-base motif [zinc regulatory element (ZRE)], ACCYYNAAGGT, in the promoters of the ZRT1, ZRT2, and ZAP1 genes (ZHAO et al. 1998 ).

ZRG10 was found to be identical to ZAP1. This information enabled a search for ZREs in promoter sequences of the ZRG's. ZRE-like sequences (ACCTTNAAGGT, with one allowed mismatch in the underlined bases) were identified within the upstream 800 bases of ZRG's 1, 5, 10, 16, and 17. Two approaches were used to determine the role of ZAP1 in regulating ZRG's. In the first, haploid ZRG-lacZ strains were prepared as for ZRG17 (see above) and the effects of ZAP1 disruption on regulated lacZ expression were determined. In the second, the ZRG-lacZ fusion construct was prepared synthetically. Promoter fragments adjacent to an initiation codon were fused to the lacZ gene in a centromeric (low-copy-number) plasmid and introduced into ZAP1 null cells or their wild-type counterparts. This second approach was used when haploid mutants could not be used due to issues of conceptual clarity (ZRG10/ZAP1) or cell viability (ZRG16/MCD4).

Expression of these ZRG-lacZ fusion constructs was strongly zinc regulated in cells with an intact ZAP1 gene (Table 4). Cells lacking a functional ZAP1 gene exhibited greatly decreased lacZ expression. In the case of ZRG1, restoration of ZAP1 function via a centromeric plasmid containing ZAP1 sequences completely restored the zinc-regulated expression of a ZRG1-lacZ fusion construct (not shown). Additionally, MnSO₄, FeNH₄(SO₄)₂, and CuSO₄ (100 µM) failed to repress ZRG1-lacZ expression in zinc-deficient cells to the same extent as ZnSO₄ (100 ± 1%, 91 ± 1%, and 87 ± 3%, respectively; cf. 0.1 ± 0.1%). These findings indicated that expression of ZRG's 1, 5, 10, 16, and 17 requires ZAP1, as expected from the presence of ZREs in the promoters of these genes. Conversely, lacZ expression in various other haploid ZRG-lacZ strains (ZRG's 2, 3, 8, 14, and 15) or synthetic ZRG-lacZ constructs (ZRG's 6 and 11) was not visibly affected by disruption of ZAP1, and ZRE-like motifs were correspondingly absent from the upstream 2000 bases of these ZRG's (not shown). (ZRG7 remains to be tested; the clone from the genetic screen failed to sporulate.) Thus, the presence of ZRE-like consensus sequences in promoter sequences of the ZRG's was sufficient to predict the ZAP1-dependent expression of those genes.

The discrepancy between the 500-fold regulation of ZRG1 in the genetic screen and the 4-fold regulation of ZRG1 in RPAs was investigated using a panel of synthetic ZRG1-lacZ fusions (Table 5). A centromeric plasmid construct containing 800 bases of ZRG1 promoter sequence and a start codon for the lacZ gene was modestly zinc regulated (12-fold) and completely ZAP1 dependent. This degree of regulation was comparable to the 4-fold regulation observed in the RPAs, consistent with the fusion functioning as a transcriptional fusion. Expression was poor with just 353 bases of ZRG1 promoter sequence fused to lacZ, consistent with exclusion of the single ZRE (at base –452) from this sequence. Stringent (>100-fold) zinc-regulated expression was reconstituted, however, by including coding sequences from ZRG1 that were in the lacZ fusion from the genetic screen. This observation suggested that ZRG1 coding sequences play a role in the regulation of the ZRG1-lacZ fusions isolated from the genetic screen.

Control of phenotypes of zinc deficiency by ZAP1:
Two other phenotypes of zinc-deficient cells were observed in addition to zinc-dependent growth and loss of ade2-dependent pigmentation and zinc-dependent growth (Fig 1). First, cells grown in shaking cultures to stationary phase in low-zinc medium tended to flocculate. The flocs dispersed immediately when cells were resuspended in a glucose-citrate buffer. Minimal flocculence was observed in cells grown with added zinc (Fig 4A). Second, cells growing in low-zinc medium in exponential phase exhibited striking distension of the vacuole (Fig 4B). Vacuole size varied substantially from cell to cell but nonetheless appeared much larger than the dilated vacuoles that are commonly seen in cells in stationary phase or cells stored in water (ROBERTS et al. 1991 ). These drastic changes were also repressed by zinc supplementation (Fig 4B).

View larger version (218K): [in this window] [in a new window] [Download PPT slide]   Figure 4. ZAP1-dependent phenotypes of zinc-deficient cells. See MATERIALS AND METHODS. (A) Flocculence. Wild-type (YPH252) and congenic zap1 null (DYY1198) cells were shaken for 2 days in low-zinc medium supplemented with the indicated amounts of zinc (µmol/liter) and photographed. (B) Vacuolization. Same as A except that cells were grown for 2 days in defined medium supplemented with the indicated amounts of zinc (µmol/liter) and examined with Nomarski optics. Bar, 5 µm.

As noted above, ZAP1 mutants have a defect in expression of the high- and low-affinity zinc uptake transporters encoded by ZRT1 and ZRT2. They should be more zinc deficient than wild-type cells in conditions with limited zinc bioavailability, and indeed this has been demonstrated directly (ZHAO and EIDE 1997 ). It was therefore surprising to observe that ZAP1 mutants grown in low-zinc medium exhibited neither flocculence (Fig 4A) nor vacuolization (Fig 4B). This finding suggested that flocculation and vacuolar distension are not merely passive metabolic consequences of zinc deficiency, but rather are part of an active cellular response to zinc deprivation. The loss of flocculence and vacuolization was not observed in haploid cells with disruptions of known ZAP1-regulated genes (other than ZAP1 mutants themselves), i.e., ZRG1-lacZ, ZRG1 null, ZRG5-lacZ, or ZRG17-lacZ cells (not shown). [ZRG16-lacZ haploid strains were not viable, as reported for null mutants of this gene (MONDESERT et al. 1997 ; GAYNOR et al. 1999 ).] Therefore, the flocculence and vacuolization characteristic of zinc-deficient cells appears to be mediated by other ZAP1-regulated genes that have not yet been identified.

Evidence for a role for ZRG17 in zinc uptake:
To examine the function of the ZRG17 gene, haploid disruption mutants were conveniently prepared from the ZRG17 clones from the genetic screen by sporulating these heterozygous diploid cells and dissecting the products of meiosis. In liquid culture the ZRG17 mutants were equally as flocculent as wild-type cells in low-zinc medium, but 10-fold higher concentrations of zinc were needed to repress flocculence (Fig 5A). Cells from mutants grown to stationary phase on YPD plates contained distended vacuoles, and vacuolar distension was ameliorated if the cells were grown with zinc supplementation (Fig 5B). These phenotypes resembled those of zinc-deficient cells and suggested that the ZRG17 protein participates in zinc uptake, a function that is consistent with the identification of ZRG17 as a ZAP1-dependent gene.

View larger version (228K): [in this window] [in a new window] [Download PPT slide]   Figure 5. Zinc-inhibitable phenotypes of ZRG17 mutants. (A) Flocculence. Cells of strain YPH252, a zrg17-242 haploid strain (DYY1206), and a zrg17-531 haploid strain (DYY617) were grown and photographed as for Fig 4A. (B) Vacuolization. Cells of strain YPH252 and the haploid zrg17 mutant strain zrg17-242 (DYY1179) were grown to stationary phase on the same YPD or zinc-supplemented YPD plates. Samples were taken from colonies with comparable and visually indistinguishable regions of growth and photographed under Nomarski optics. Bar, 5 µm.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Transposon tagging as an approach for identifying differentially regulated genes:
This study describes an application of transposon tagging to the identification of differentially regulated genes in yeast. Transposon tagging was first adapted from bacterial systems for application in yeast by several groups in the mid-1980s (RUBY and SZOSTAK 1985 ; SEIFERT et al. 1986 ; MCCAFFREY et al. 1987 ), and further useful refinements were introduced in 1994 that allowed studies of differential gene expression to be performed on a quasi-genomic scale (BURNS et al. 1994 ; DANG et al. 1994 ). However, there are still only a few studies of this type in the literature (for an excellent example, see ERDMAN et al. 1998 ).

Several newer methodologies for identifying differentially regulated genes have been applied recently on a large scale in yeast. These include microarray hybridization technology (DERISI et al. 1997 ), serial analysis of gene expression (SAGE; VELCULESCU et al. 1995 , VELCULESCU et al. 1997 ), and two-dimensional gel electrophoresis of proteins combined with mass spectrometry (GODON et al. 1998 ). When compared with these methodologies, transposon tagging has some advantages that are illustrated in this study: (1) Transposon tagging uses only instrumentation found in any molecular biology laboratory; (2) levels of gene expression can be quantitated precisely with simple assays that have a dynamic range spanning almost four orders of magnitude; (3) reporter gene constructs are embodied in the clones isolated from the genetic screen as robust and renewable reagents; and (4) potential disruption alleles are conveniently prepared from the clones in the genetic screen by sporulation and dissection, facilitating the functional characterization of novel genes.

Mechanistic classification of the ZRG's:
The identification of multiple ZRG's raises immediate questions about the mechanisms underlying zinc sensing and how such signals are coupled to the mechanisms of gene transcription and translation. In the case of the ZAP1-dependent genes, a partial answer is already at hand (ZHAO and EIDE 1997 ; ZHAO et al. 1998 ): The ZAP1 protein may have dual roles as zinc sensor and DNA binding protein. While much remains to be learned about how zinc binds to this protein and how zinc binding is structurally coupled to interactions with zinc-response elements, it is clear that such properties will ultimately account for many of the observed phenotypes of zinc deprivation in these cells.

A challenge that remains to be addressed is to reconcile the 10- to 500-fold regulation observed with the ZRG-lacZ fusion constructs with the much more modest 2- to 8-fold regulation of the endogenous ZRG RNAs measured in carefully performed ribonuclease protection assays. In addition to this exaggerated degree of regulation, lacZ expression levels correlated rather poorly with RNA expression levels. For example, ZRG10 and ZRG13 RNA expression levels were much stronger than suggested by the corresponding lacZ expression levels, while the opposite was true for ZRG2 and ZRG4. While lacZ fusion constructs have been used for many years in reporter gene assays of gene expression (ROSE and BOTSTEIN 1983 ), there are several reasons why RNA abundance may correlate poorly with lacZ-encoded ß-galactosidase activity:

1. ß-Galactosidase is active only as a tetramer (NICHTL et al. 1998 ). At very low monomer concentrations it is possible that this enzyme activity may be disproportionately weak, thus amplifying the -fold regulation observed at the RNA level. However, this nonlinearity does not explain how the RNA is regulated to begin with.

2. Biochemical idiosyncrasies associated with the lacZ fusions (e.g., transmembrane domains) could contribute to differences in lacZ expression levels. For example, lacZ activities varied 10-fold among the five different species of ZRG17 clones (not shown).

3. The ZRG-lacZ fusions obtained in the genetic screen may be susceptible to post-transcriptional modes of regulation. Indeed, coding sequences in ZRG1 appear to contribute substantially to the zinc-regulated expression of ZRG1 (Table 5). It is pertinent to note that ubiquitination of the zinc uptake transporter encoded by ZRT1 can be instigated by exposure of cells to zinc (GITAN et al. 1998 ).

Functional roles of ZRG's in the ZAP1 regulon:
Perhaps the most important questions still to be answered involve the functions of the ZAP1-dependent genes—whether they play some unsuspected role in zinc homeostasis, or whether they function to maintain some biochemical process that is sensitive to zinc status.

ZRG17 mutants have mutant phenotypes suggestive of cellular zinc deficiency (Fig 5), suggesting that the ZRG17 protein functions in zinc uptake. This idea is supported by the presence of seven potential transmembrane domains in the predicted protein and by a cluster of histidine residues admixed with acidic residues following the third potential transmembrane domain (not shown). The latter feature could represent a zinc-interacting domain, in view of the histidine clusters that are present in other zinc-transporting proteins (e.g., ZHAO and EIDE 1996B ). It is an open question why ZRG17 mutants exhibited phenotypes of zinc deficiency in the presence of unmutated alleles of ZRT1 and ZRT2, although it is noteworthy that the low-zinc growth medium used in the ZRG17 studies was different from the pH 4.2 EDTA-containing medium used to characterize ZRT1 and ZRT2 (ZHAO and EIDE 1996B ). Further characterization of the function of the ZRG17 protein in the context of the physiologic functions of ZRT1 and ZRT2 is underway.

DPP1 (ZRG1) was the most highly regulated of all the ZRG-lacZ fusions in the genetic screen. Expression of DPP1-lacZ fusions was completely dependent on ZAP1 function, suggesting a role in zinc homeostasis. The structure of the DPP1 protein suggested a role in zinc transport, with six transmembrane domains and a highly conserved set of three histidine residues. Surprisingly, however, overexpression or deletion of DPP1 (ZRG1) revealed no discernable phenotypes, zinc-related or otherwise, compared with controls (not shown). It is provocative to note that the DPP1 protein has a known enzymatic activity, diacylglycerol pyrophosphate phosphatase; in fact, DPP1 was first cloned after purification of this enzyme (TOKE et al. 1998 ). There is a single report of cell aggregation, impaired mitotic cytokinesis, and abnormal cell shape in a DPP1 mutant (KATAGIRI and SHINOZAKI 1998 ). These phenotypes recall the flocculence and vacuolization in zinc-deficient cells. It is unclear why they were not observed in any of the DPP1 mutants in this study or elsewhere (TOKE et al. 1998 ).

ADH4 (ZRG5) was also stringently regulated without an observable phenotype. The ADH4 protein has been characterized as a zinc-dependent alcohol dehydrogenase that was thought to be minimally expressed if at all (DREWKE and CIRIACY 1988 ). The induction of ADH4 expression in low-zinc conditions suggests that this protein functions in these conditions as a backup to the strongly expressed alcohol dehydrogenase encoded by ADH1, but this has not been tested.

Finally, MCD4 (ZRG16) encodes a protein required for the synthesis of glycosylphosphatidylinositol (GPI) anchors that mediate the cell-surface expression of various proteins (GAYNOR et al. 1999 ). The fact that MCD4 is an essential gene poses the challenge of explaining why zinc-supplemented cells and ZAP1 null cells are viable, given the strong repression of this gene in those cells as gauged by lacZ fusion constructs. Perhaps residual levels of expression are sufficient to maintain cell viability. Interestingly, the FLO1 protein, a cell-surface protein required for flocculence (see below), is believed to be GPI anchored before it is conjugated to carbohydrates in the cell wall (VAN DER VAART et al. 1996).

Prospects for identifying other genes in the ZAP1 regulon:
Because zinc is a required cofactor for hundreds of proteins throughout metabolism, it is not surprising that zinc deprivation elicits a variety of cellular phenotypes. However, it was unexpected that two of the most prominent phenotypes of zinc-deficient cells, flocculence and vacuolar dilatation, were missing or attenuated in ZAP1 mutants. As noted earlier, zinc uptake is impaired in these mutants, so if these phenotypes were consequences of zinc depletion in some zinc-dependent protein, exaggerated phenotypic expression should have been observed in ZAP1 mutants. That the phenotypes were missing or attenuated indicates instead that these phenotypes are directly controlled by ZAP1.

Flocculation has been studied for many years due to its importance in the brewing industry (reviewed in STRATFORD 1992 ). As fermentation mixtures age, yeast cells form large aggregates or flocs that sediment, thereby effecting a separation of the yeast from the brewery product. Two types of flocculation have been described in S. cerevisiae, FLO1 dependent and the NewFlo type, distinguished by their differing sensitivities to inhibition by glucose and other culture conditions. The flocculation observed in zinc-deficient cells in this study was evidently not inhibited by glucose, as glucose was present in the culture medium, and preliminary studies indicate that calcium can rescue the loss of flocculation observed when zinc-deficient cells are resuspended in a citrate buffer (not shown). These characteristics are therefore consistent with the FLO1-dependent type of flocculence that is observed in most laboratory strains (STRATFORD 1992 ). The FLO1 gene encodes a recently characterized mannose-specific and divalent cation-dependent lectin that appears to promote flocculence through its ability to bind to mannosylated constituents of the cell wall (KOBAYASHI et al. 1998 ). Expression of the FLO1 gene is known to be controlled by the general transcriptional repressors encoded by TUP1 and SSN6 (TEUNISSEN et al. 1995 ), but little is known about other modes of regulation. A consensus ZRE is present at base –401 in the FLO1 promoter, suggesting that FLO1 is another ZAP1-dependent gene. A synthetic FLO1 (–800 ... +3) promoter fusion with the lacZ gene was in fact found to be expressed at very low levels (5% of wild type) in a ZAP1 null mutant (not shown), possibly accounting for why ZAP1 mutants do not flocculate. However, unexpectedly, FLO1-lacZ expression was induced rather than repressed in zinc-replete cells, even in ZAP1 null cells (not shown). Understanding how ZAP1 controls flocculence therefore requires a more thorough investigation that is beyond the scope of this article.

Progress in understanding the phenotype of vacuolar dilatation has been rapid recently with the discovery of a biochemically characterized yeast mutant exhibiting hugely dilated vacuoles. This mutant was originally discovered in a genetic screen for cells defective in mitotic cytokinesis (cf. the discussion of DPP1, above), but this latter phenotype was later shown to be secondary to the presence of the hugely dilated vacuole (YAMAMOTO et al. 1995 ). The gene affected in this mutant, FAB1, encodes a phosphatidylinositol-3-phosphate 5-kinase, suggesting that the novel phospholipid produced by this enzyme participates in post-Golgi, prevacuolar membrane trafficking (GARY et al. 1998 ). The similarity between the distended vacuoles observed in zinc-deficient cells and those in FAB1 mutants may indicate that maintenance of cytosolic phosphatidylinositol-3,5-bisphosphate levels requires at least one ZAP1-dependent gene.

An important task that lies ahead is to develop a comprehensive list of the ZAP1-dependent genes. The promoters in the seven ZAP1-dependent genes identified so far all contain a sequence motif that closely resembles the consensus ZREs previously derived. Several other ZRG's lacking a ZRE in their promoters were found to be expressed independently of ZAP1, suggesting that ZREs have predictive value in identifying ZAP1-dependent genes. The number of ZAP1-dependent genes is not known. Five ZAP1-dependent genes were identified here in a genetic screen that examined approximately half of the genes in the genome, suggesting that there are perhaps 10 ZAP1-dependent genes. However, a search for the sequence ACCTTNAAGGT in the S. cerevisiae genome (http://genome-www2.stanford.edu/cgi-bin/SGD/PATMATCH/nph-patmatch) revealed 9 genes containing this sequence within 500 bases of their putative start codons, including ZRT1 and MCD4. The actual number of candidate ZAP1-regulated genes is considerably larger, since the ZRE recognized by the ZAP1 protein is clearly somewhat degenerate (e.g., for DPP1), and also since the promoters for some genes may be much longer than 500 bases (e.g., RUPP et al. 1999 ). DNA microarrays (DERISI et al. 1997 ) may provide a rapid experimental approach for testing which of these are in fact ZAP1-dependent genes.

	ACKNOWLEDGMENTS

I am grateful to R. Klausner for allowing me to initiate this project in his laboratory. I also thank A. Levine, J. Berg, and G. Dover for their support. The transposon-tagged genomic library was generously provided by M. Snyder. I thank A. Dancis, R. Binder, and S. Erdman for introducing me to the genetic screen, T. Dunn for helpful discussions, C. Yuan for sharing equipment, and J. Berg for comments on the manuscript. This work was supported in part by National Institutes of Health grants to the Johns Hopkins University Department of Pediatrics (Child Health Research Center) and to J. Berg, and by a Richard S. Ross Clinician Scientist Award from the Johns Hopkins University School of Medicine.

Manuscript received July 19, 1999; Accepted for publication May 11, 2000.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

AGGETT, P. J. and J. G. COMERFORD, 1995  Zinc and human health. Nutr. Rev. 53:S16-S22[Medline].

ALTSCHUL, S. F., W. GISH, W. MILLER, E. W. MYERS, and D. J. LIPMAN, 1990  Basic local alignment search tool. J. Mol. Biol. 215:403-410[Medline].

ASKWITH, C. and J. KAPLAN, 1998  Iron and copper transport in yeast and its relevance to human disease. Trends Biochem. Sci. 23:135-138[Medline].

ASLANIDIS, C. and P. J. DE JONG, 1990  Ligation-independent cloning of PCR products (LIC-PCR). Nucleic Acids Res. 18:6069-6074[Abstract/Free Full Text].

BAI, C., P. SEN, K. HOFMANN, L. MA, and M. GOEBL et al., 1996  SKP1 connects cell cycle regulators to the ubiquitin proteolysis machinery through a novel motif, the F-box. Cell 86:263-274[Medline].

BECK, F. W., J. KAPLAN, N. FINE, W. HANDSCHU, and A. S. PRASAD, 1997  Decreased expression of CD73 (ecto-5'-nucleotidase) in the CD8+ subset is associated with zinc deficiency in human patients. J. Lab. Clin. Med. 130:147-156[Medline].

BERG, J. M. and Y. SHI, 1996  The galvanization of biology: a growing appreciation for the roles of zinc. Science 271:1081-1085[Abstract].

BHUTTA, Z. A., R. E. BLACK, K. H. BROWN, J. M. GARDNER, and S. GORE et al., 1999  Prevention of diarrhea and pneumonia by zinc supplementation in children in developing countries: pooled analysis of randomized controlled trials. J. Pediatr. 135:689-697[Medline].

BLANCHARD, R. K. and R. J. COUSINS, 1996  Differential display of intestinal mRNAs regulated by dietary zinc. Proc. Natl. Acad. Sci. USA 93:6863-6868[Abstract/Free Full Text].

BURNS, N., B. GRIMWADE, P. B. ROSS-MACDONALD, E. Y. CHOI, and K. FINBERG et al., 1994  Large-scale analysis of gene expression, protein localization, and gene disruption in Saccharomyces cerevisiae. Genes Dev. 8:1087-1105[Abstract/Free Full Text].

CULOTTA, V. C., S. J. LIN, P. SCHMIDT, L. W. KLOMP, and R. L. CASARENO et al., 1999  Intracellular pathways of copper trafficking in yeast and humans. Adv. Exp. Med. Biol. 448:247-254[Medline].

DANCIS, A., D. S. YUAN, D. HAILE, C. ASKWITH, and D. EIDE et al., 1994  Molecular characterization of a copper transport protein in S. cerevisiae: an unexpected role for copper in iron transport. Cell 76:393-402[Medline].

DANG, V.-D., M. VALENS, M. BOLOTIN-FUKUHARA, and B. DAIGNAN-FORNIER, 1994  A genetic screen to isolate genes regulated by the yeast CCAAT-box binding protein Hap2p. Yeast 10:1273-1283[Medline].

DERISI, J. L., V. R. IYER, and P. O. BROWN, 1997  Exploring the metabolic and genetic control of gene expression on a genomic scale. Science 278:680-686[Abstract/Free Full Text].

DREWKE, C. and M. CIRIACY, 1988  Overexpression, purification and properties of alcohol dehydrogenase IV from Saccharomyces cerevisiae. Biochim. Biophys. Acta 950:54-60[Medline].

DUGAICZYK, A., H. W. BOYER, and H. M. GOODMAN, 1975  Ligation of EcoRI endonuclease-generated DNA fragments into linear and circular structures. J. Mol. Biol. 96:171-184[Medline].

EIDE, D., 1997  Molecular biology of iron and zinc uptake in eukaryotes. Curr. Opin. Cell Biol. 9:573-577[Medline].

ERDMAN, S., L. LIN, M. MALCZYNSKI, and M. SNYDER, 1998  Pheromone-regulated genes required for yeast mating differentiation. J. Cell Biol. 140:461-483[Abstract/Free Full Text].

GARY, J. D., A. E. WURMSER, C. J. BONANGELINO, L. S. WEISMAN, and S. D. EMR, 1998  Fab1p is essential for PtdIns(3)P 5-kinase activity and the maintenance of vacuolar size and membrane homeostasis. J. Cell Biol. 143:65-79[Abstract/Free Full Text].

GAYNOR, E. C., G. MONDESERT, S. J. GRIMME, S. I. SEED, and P. ORLEAN et al., 1999  MCD4 encodes a conserved endoplasmic reticulum membrane protein essential for glycosylphosphatidylinositol anchor synthesis in yeast. Mol. Biol. Cell 10:627-648[Abstract/Free Full Text].

GIETZ, R. D., and R. A. WOODS 1994 High-efficiency transformation with lithium acetate, pp. 121–134 in Molecular Genetics of Yeast: A Practical Approach, edited by J. R. JOHNSTON. IRL Press, New York.

GITAN, R. S., H. LUO, J. RODGERS, M. BRODERIUS, and D. EIDE, 1998  Zinc-induced inactivation of the yeast ZRT1 zinc transporter occurs through endocytosis and vacuolar degradation. J. Biol. Chem. 273:28617-28624[Abstract/Free Full Text].

GODON, C., G. LAGNIEL, J. LEE, J. M. BUHLER, and S. KIEFFER et al., 1998  The H₂O₂ stimulon in Saccharomyces cerevisiae. J. Biol. Chem. 273:22480-22489[Abstract/Free Full Text].

GOFFEAU, A., 1994  Yeast genes in search of functions. Nature 369:101-102[Medline].

GRIDER, A., L. B. BAILEY, and R. J. COUSINS, 1990  Erythrocyte metallothionein as an index of zinc status in humans. Proc. Natl. Acad. Sci. USA 87:1259-1262[Abstract/Free Full Text].

GUARENTE, L., 1983  Yeast promoter and lacZ fusions designed to study expression of cloned genes in yeast. Methods Enzymol. 101:181-191[Medline].

HODGES, P. E., A. H. MCKEE, B. P. DAVIS, W. E. PAYNE, and J. I. GARRELS, 1999  The Yeast Proteome Database (YPD): a model for the organization and presentation of genome-wide functional data. Nucleic Acids Res. 27:69-73[Abstract/Free Full Text].

KATAGIRI, T. and K. SHINOZAKI, 1998  Disruption of a gene encoding phosphatidic acid phosphatase causes abnormal phenotypes in cell growth and abnormal cytokinesis in Saccharomyces cerevisiae. Biochem. Biophys. Res. Commun. 248:87-92[Medline].

KIM, J. M., S. VANGURI, J. D. BOEKE, A. GABRIEL, and D. F. VOYTAS, 1998  Transposable elements and genome organization: a comprehensive survey of retrotransposons revealed by the complete Saccharomyces cerevisiae genome sequence. Genome Res. 8:464-478[Abstract/Free Full Text].

KOBAYASHI, O., N. HAYASHI, R. KUROKI, and H. SONE, 1998  Region of FLO1 proteins responsible for sugar recognition. J. Bacteriol. 180:6503-6510[Abstract/Free Full Text].

LICASTRO, F., L. F. DAVIS, E. MOCCHEGIANI, and N. FABRIS, 1996  Impaired peripheral zinc metabolism in patients with senile dementia of probable Alzheimer's type as shown by low plasma concentrations of thymulin. Biol. Trace Elem. Res. 51:55-62[Medline].

LIPPARD, S. J., and J. M. BERG 1994 Principles of Bioinorganic Chemistry. University Science Books, Mill Valley, CA.

LUSSIER, M., A. M. WHITE, J. SHERATON, T. DI PAOLO, and J. TREADWELL et al., 1997  Large scale identification of genes involved in cell surface biosynthesis and architecture in Saccharomyces cerevisiae. Genetics 147:435-450[Abstract].

MATTHEWS, R. G. and C. W. GOULDING, 1997  Enzyme-catalyzed methyl transfers to thiols: the role of zinc. Curr. Opin. Chem. Biol. 1:332-339[Medline].

MAURICE, T. C., C. E. MAZZUCCO, C. S. RAMANATHAN, B. M. RYAN, and G. A. WARR et al., 1998  A highly conserved intraspecies homolog of the Saccharomyces cerevisiae elongation factor-3 encoded by the HEF3 gene. Yeast 14:1105-1113[Medline].

MCCAFFREY, G., F. J. CLAY, K. KELSAY, and G. F. SPRAGUE, 1987  Identification and regulation of the gene required for cell fusion during mating of the yeast Saccharomyces cerevisiae. Mol. Cell. Biol. 7:2680-2690[Abstract/Free Full Text].

MILLER, J. H. 1972 Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

MONDESERT, G., D. J. CLARKE, and S. I. REED, 1997  Identification of genes controlling growth polarity in the budding yeast Saccharomyces cerevisiae: a possible role of N-glycosylation and involvement of the exocyst complex. Genetics 147:421-434[Abstract].

MOSCH, H. U. and G. R. FINK, 1997  Dissection of filamentous growth by transposon mutagenesis in Saccharomyces cerevisiae. Genetics 145:671-684[Abstract].

MYERS, A. M., A. TZAGOLOFF, D. M. KINNEY, and C. J. LUSTY, 1986  Yeast shuttle and integrative vec