Organisms require mechanisms to survive under adverse conditions of extreme heat, osmolarity, nutrient limitation, and other stresses. Recent studies have begun to probe the cellular responses to the stress of zinc deficiency. Zinc is a critical cofactor for many proteins and plays important roles in myriad biological processes. Therefore, when zinc becomes limiting, cells must respond to maintain zinc homeostasis. In addition, cells may alter their metabolic processes to adapt to growth under conditions where certain zinc-dependent proteins are less active. We are examining the cellular responses to zinc deficiency in the yeast Saccharomyces cerevisiae.

In this yeast, the Zap1 transcription factor is a central player in the response to zinc deficiency [1]. For many of its target genes, Zap1 acts as an activator of transcription and increases gene expression when zinc levels are low. To perform this function, Zap1 binds to Zinc-Responsive Elements or "ZREs" in the promoters of its target genes [2]. The consensus sequence for a ZRE is ACCTTNAAGGT. While some Zap1 target genes contain multiple functional ZREs, many others have only a single binding site [2, 3].

The Zap1 protein is 880 amino acids long. A DNA binding domain consisting of five zinc fingers is found at its carboxy terminus [4, 5]. In addition, Zap1 contains two independent activation domains, designated AD1 and AD2, that mediate the increased transcription of target genes [6]. Zap1 is a direct sensor of cellular zinc levels. The protein resides in the nucleus under all conditions of zinc status [6]. When zinc levels rise, the metal binds to ligand residues in the AD1 and AD2 regions of the protein and this binding inhibits the ability of these domains to promote transcription [6–10]. Alteration of these regulatory zinc-binding ligands by mutation decreases the ability of Zap1 to respond to zinc and the mutant protein constitutively activates transcription [7, 10].

Previous studies have identified a large number of potential Zap1 target genes in the yeast genome [3, 11, 12]. Many of these genes act to maintain sufficient levels of cytosolic zinc available for cell growth. For example, the ZRT1, ZRT2, and FET4 genes encode zinc transporters responsible for zinc uptake across the plasma membrane [13–15]. These genes are induced by Zap1 in zinc-limited cells. Zap1 also induces expression of the ZRT3 gene in low zinc; ZRT3 encodes a vacuolar membrane protein responsible for transporting zinc stored in the vacuole to the cytoplasm for its utilization [16]. As a final example, Zap1 induces transcription of its own gene in a positive autoregulatory loop [1]. Thus, Zap1 levels rise in zinc-limited cells and this may lead to increased expression of other target genes.

In addition to its role in activating gene expression, Zap1 can also act as a transcriptional repressor. Previous studies have identified two different mechanisms of Zap1-mediated repression. The ZRT2 gene provided the first example. ZRT2 is unusual among Zap1 target genes in that it is induced by mild zinc limitation and repressed by more severe zinc deficiency [17]. This paradoxical pattern of regulation is due to the presence of three ZREs in the ZRT2 promoter. Two high affinity ZREs, ZRE1 and ZRE2, are located upstream of the TATA box and these elements mediate Zap1-dependent activation of gene expression. The third ZRE, ZRE3, has a low affinity of Zap1 binding and is located downstream of the TATA box. ZRE3 is essential for repression of ZRT2 expression. Under mild conditions of zinc deficiency, Zap1 binds to ZRE1 and ZRE2 and activates gene expression. Under severe zinc deficiency, Zap1 levels rise due to autoregulation and the protein then binds to ZRE3 and interferes with ZRT2 expression possibly by blocking transcription initiation.

The ADH1 and ADH3 genes provide examples of a second mechanism of Zap1-mediated repression. ADH1 and ADH3 encode zinc-dependent alcohol dehydrogenases. These genes are highly expressed in zinc-replete cells but are repressed in zinc-deficient cells [18]. Zap1 mediates ADH1 and ADH3 repression in low zinc by means of intergenic transcripts that are activated by Zap1 and transcribed through the ADH1 and ADH3 promoters. These intergenic transcripts, designated ZRR1 and ZRR2 respectively, do not encode protein products but rather their synthesis results in the transient displacement of transcription factors normally required for ADH1 and ADH3 expression. This results in the reduced expression of two of the most abundant zinc-binding proteins in the cell. Conversely, the ADH4 gene is induced by Zap1 and this gene encodes a potential iron-dependent alcohol dehydrogenase [3, 18, 19]. By switching from zinc-dependent to zinc-independent ADH isozymes, the cell may conserve zinc for other uses. Alternatively, Adh4 may use zinc as its cofactor [20] but this protein is predicted to bind only one zinc per monomer while Adh1 and Adh3 each bind two. Thus, zinc conservation could occur under this scenario as well.

DNA microarrays have been remarkably useful in assessing the transcriptional responses of an organism such as yeast to stress conditions [3, 11, 12]. In a previous study, we used microarrays to identify likely Zap1 target genes in the yeast genome, a group of genes that we referred to as the Zap1 "regulon" [3]. In that study, we identified a total of 46 genes in yeast that are potential targets of Zap1 activation. As described below, we have further addressed this issue using revised experimental approaches. This new analysis has led to the identification of many new potential Zap1 target genes. In addition, we have characterized the differential regulation of Zap1 target genes. We found that some genes respond to mild zinc deficiency and act as a first line of defense against this stress. These first-line defense genes participate in various mechanisms of zinc homeostasis. Other Zap1 target genes respond only to severe conditions of zinc limitation and serve as a second line of defense. Second-line defense genes act largely in the adaptation to conditions where zinc availability is insufficient to maintain optimal cell function.

Zinc deficiency causes drastic changes in yeast gene expression. We previously reported that ~15% (934) of all yeast genes increased or decreased in expression in zinc-limited vs. zinc-replete yeast grown in batch cultures [3]. Similarly, De Nicola et al. reported that 381 genes were affected by zinc status in chemostat cultures [12]. Because zinc plays so many functional roles, the majority of these effects are likely to be indirect responses to changes in cellular processes resulting from decreased activity of key zinc-dependent proteins. To define the direct responses to zinc deficiency, we are identifying genes under the control of the Zap1 transcription factor. In this way, we can learn how cells respond specifically to the stress of zinc-limiting conditions.

We previously used a combination of microarray analysis and a motif identification algorithm to identify 46 potential Zap1 targets [3]. In this current study, we made important modifications to our previous approach by using alternative growth conditions for the microarray experiments and a different motif analysis algorithm that was better able to detect potential Zap1 binding sites. Using these new tools, we have extended many aspects of our previous work. First, this study provides additional confirmation for many genes previously proposed to be Zap1 targets (Table 1). For example, genes such as MOH1, TKL2, etc. were suggested to be direct Zap1 targets based on their Zap1-dependent induction in zinc-limited cells [3]. Their increased expression in cells expressing the Zap1^TC allele provides strong additional support for this hypothesis. In all, 33 of our original 46 potential targets were confirmed in this way. Several of the remaining 13 genes (e.g. YBL048W, COS4, COS8, RAD27) showed increased expression in Zap1^TC-expressing cells although these effects did not meet our minimum cut-off value of 1.5-fold changes (Additional file 1).

In addition, we have added 31 potential new members to the Zap1 regulon. Sixteen of these genes were observed in our previous study to be zinc-responsive and Zap1-dependent [3] but were not considered to be direct Zap1 targets because we were unable to identify potential ZREs in their promoters using MEME (Table 2). With RSAT, we identified their potential ZREs and went on to show that many of these sequences were specifically bound by Zap1 in vitro. The response of these genes to the Zap1^TC allele also provides additional evidence for their direct regulation by Zap1. The other 15 new target genes responded to more severe zinc-limiting conditions than that used in our previous study (Table 3). These genes, plus four additional genes identified as Zap1 targets using other methods, i.e. EKI1, PIS1, ZRR1, and ZRR2 [18, 21, 22], brings the total of confirmed and potential Zap1 targets in the yeast genome to 81 genes.

With the identification of many new targets of Zap1 activation, we are building a comprehensive picture of how yeast cells respond to zinc-limiting conditions. A summary of the functional roles of many of these genes is provided in Table 4 and a figure showing the general relationship of these genes to their response over a range of zinc levels is provided in Fig. 6. We can separate these responses conceptually into first- and second-lines of defense against the stress of zinc deficiency. The first-line defense genes play key roles in zinc homeostasis while many second-line defense genes allow the cells to adapt to zinc-limiting conditions when they can no longer obtain sufficient zinc for optimal growth. One caveat to this analysis is that we are assuming that changes in transcript levels are accompanied by similar changes in protein abundance and this assumption is not necessarily true. Nonetheless, the following discussion provides a clear and testable framework for understanding the cellular responses to zinc deficiency.

Functional Group	Zap1 Target Genes
First-line defenses: Zinc homeostasis
Zinc uptake	ZRT1, ZRT2, FET4, ZPS1
Mobilize zinc stores	ZRT3
Increase transcription response	ZAP1
Zinc sparing	ZRR1 (ADH1), ZRR2 (ADH3), ADH4
Zinc shock resistance	ZRC1
Other zinc homeostasis genes	IZH1, IZH2
Second-line defenses: Adaptive responses
Secretory pathway function	ZRG17, MNT2, MCD4, YJR061W
Cell wall function	PST1, BAG7, HSP12, SED1, SCW4, YOL155C, YIL169C, YJL171C
Stress resistance	GRE2, RAD27, TSA1, CTT1, UTH1, HSP26
Phospholipid metabolism	DPP1, EKI1, PIS1, YJL132W
Sulfur metabolism	LAP3, SAM3, YPR003C
Protein degradation	PRC1, PEP4, PRB1, ATG19, UBX6
Carbohydrate metabolism	NRG2, TKL2, ENO1, ENO2, TDH1
Purine/pyrimidine metabolism	URA10, ADE17, HNT1
Mitochondrial function	FMP43, PET20, PET10
Other	VEL1, MOH1, GPG1, COS6, ICY2, COS1, PHM7, PIG2, TIS11, ALD3, PUT4 , COS2, ZIP1, COS4, COS8, TPO5, COS3, YNL254C, YOR387C, YLL020C, YOL131W, YBL029W, YHR214W-A, YMR181C, YLL053C , YBL048C, YMR086W, YNL234W

Genes shown in bold are new targets in this study

Among the first-line defense genes, ZRT1, ZRT2, and FET4 are induced by mild zinc deficiency to increase the ability of the cell to accumulate zinc from its environment. ZPS1 is also induced under these mild conditions and we have previously proposed that its product, a secreted protein related to metalloproteases, may also be involved in zinc acquisition by degrading extracellular proteins and releasing any bound metals [3]. Similarly, ZRT3 is up-regulated by mild zinc deficiency and also responds rapidly to zinc withdrawal. Thus, mobilization of zinc stores from the vacuole is also a first-line response. Induction of the YOR387C gene in response to mild zinc deficiency suggests that this gene is also involved in the first-line defense against zinc limitation. The function of this protein is not yet known but its pattern of regulation by zinc suggests that it may play a role in zinc uptake or vacuolar zinc export. TIS11 ( also known as CTH2) is also induced in LZM + 300 μM ZnCl₂. Tis11 binds to specific mRNAs and signals their degradation in response to iron limitation [25]; its role in zinc deficiency is not yet determined.

Other first-line defenses to zinc deficiency include an increase in ZAP1 expression mediated by Zap1 autoregulation [1]. The resulting increase in Zap1 protein level may maximize target gene expression and is also likely to be responsible for repression of ZRT2 under these conditions [17]. In support of this latter hypothesis, we note that ZRT2 repression correlates well with the observed increase in ZAP1 mRNA levels. In addition, the coordinated switch from ADH1/ADH3 expression to ADH4 expression likely represents a mechanism to conserve zinc for other uses [18]. Finally, we observed an increase in ZRC1 expression in moderate-to-severely zinc-deficient cells. We showed previously that this induction is required for cells to survive the stress of zinc shock, i.e. when zinc-deficient cells are re-supplied with zinc [26]. The high activity of zinc uptake transporters in these cells results in rapid zinc overload and the increased activity of Zrc1 is needed to sequester the excess zinc in the vacuole.

Induction of many Zap1 target genes occurred only in response to the most severe zinc-limiting condition we tested (LZM + 3 μM ZnCl₂) and only slowly in response to zinc withdrawal. We consider these genes to be the second line of defense against zinc deficiency. Most of these second-line defense genes appear to be involved in the adaptation to zinc deficiency rather than playing a role in zinc homeostasis (see below). This transition from homeostatic to adaptive mechanisms correlates well with what we know about cell growth under these conditions. We previously showed that cells grow at their maximum growth rate in LZM supplemented with ≥ 10 μM ZnCl₂ but their growth rate decreases in media containing less zinc [14]. Thus, under conditions that induce the second-line defense genes, zinc homeostatic mechanisms are no longer capable of supplying sufficient zinc for optimal function of cellular processes. Our results indicate that it is under these conditions of suboptimal zinc that adaptive responses occur.

One important functional category of second-line defense involves secretory pathway function (Table 4). The ZRG17 zinc transporter gene is up-regulated presumably to maintain the zinc status of the endoplasmic reticulum [27, 28]. MCD4 and MNT2 encode enzymes required for protein modifications (GPI anchor synthesis and O-linked glycosylation), that occur in the ER and the Golgi, respectively [29, 30]. YJR061W encodes an ER-localized protein [31] related to Mnn4, which is involved in both N- and O-linked glycosylation [32]. Increased expression of these proteins may help maintain the efficiency of these processes under the adverse conditions of severe zinc deficiency. Consistent with this hypothesis, Mcd4 is thought to be a zinc-dependent enzyme [33].

Another group of second-line defense genes are involved in cell wall function (Table 4). These include PST1, HSP12, SED1, and SCW4. While their specific functions are unclear, these proteins all reside in the cell wall suggesting that cell wall remodeling occurs under zinc-limiting conditions. These alterations may be important to maintain the structural integrity of the cell wall. Consistent with this hypothesis, disruption of zinc transport into the secretory pathway caused increased sensitivity to an inhibitor of cell wall function, calcofluor white [34]. This observation suggests that defects in cell wall synthesis occur in zinc-limited cells.

A third functional category of second-line defense genes are involved in stress resistance (Table 4). Our results implicate TSA1, CTT1, UTH1, and HSP26 as Zap1 targets along with previously identified targets RAD27 and GRE2. RAD27 encodes a nuclease that is required for the stability of minisatellite repeats in genomic DNA [35, 36]. The recent observation that zinc deficiency in yeast destabilizes these DNA repeats [37] is consistent with a role of Rad27 in promoting genome stability in zinc-limited cells. GRE2 encodes a methylglyoxal reductase [38]. Methylglyoxal is a byproduct of glycolysis that can react with proteins to disrupt their function [39]. We note with interest that two major targets of damage by methylglyoxal, the enolases encoded by ENO1 and ENO2 [39], are also up-regulated by Zap1 (Table 3). TSA1, CTT1, and UTH1 are involved in resistance of the cell to oxidative stress. We recently showed that yeast cells grown under severe zinc-limiting conditions experience increased oxidative stress [23]. In addition, we demonstrated that induction of TSA1, encoding the major cytosolic peroxiredoxin, in low zinc is required to resist that increased stress. CTT1, encoding the cytosolic form of catalase [40], may be up-regulated for similar reasons. Uth1 is a mitochondrial protein of unknown function that is required for superoxide resistance [41] so its induction by Zap1 may also be related to oxidative stress resistance. Lastly, Hsp26 is a member of the small heat shock protein family that has chaperone activity and protects proteins from misfolding [42]. The regulation of these genes by Zap1 may help temper the various stresses experienced by zinc-limited cells. Other functional categories of Zap1 target genes induced by severe zinc deficiency include phospholipid synthesis, protein degradation, and carbohydrate, sulfur, and purine/pyrimidine metabolism (Table 4). On a final note, more than half of the genes now identified as potential Zap1 targets (47 of 81) have mammalian orthologs. Regulating the expression of these genes may also be important for cell growth under low zinc conditions in mammals.

In this report, we have identified 31 new potential targets of Zap1 and have characterized the response of these and previously identified target genes to changes in zinc status and to time following zinc withdrawal. This analysis of Zap1 transcriptional regulation is providing unexpected and exciting new insights into the mechanisms of metal nutrient homeostasis in yeast and is telling us much about how these cells adapt to growth under the stress of zinc deficiency. Future studies will address the detailed roles of these various adaptive responses to cell growth under zinc-limiting conditions.

Multiple Expectation Maximization for Motif Elicitation (MEME) program: http://meme.sdsc.edu/meme/

Regulatory Sequence Analysis Tools (RSAT): http://rsat.ulb.ac.be/rsat/

PatMatch: http://seq.yeastgenome.org/cgi-bin/PATMATCH/nph-patmatch

WebLogo: http://weblogo.berkeley.edu/