Involvement of the VDE Homing Endonuclease and Rapamycin in Regulation of the Saccharomyces cerevisiae GSH11 Gene Encoding the High Affinity Glutathione Transporter*

The Saccharomyces cerevisiae gene HGT1/GSH11 encodes the high affinity glutathione transporter and is repressed by cysteine added to the culture medium. It has been found previously that a 5′-upstream cis-element, CCGCCACAC, is responsible for regulating GSH11 expression and that several proteins bind to this element (Miyake, T., Kanayama, M., Sammoto, H., and Ono, B. (2002) Mol. Genet. Genomics 266, 1004–1011). In this report we present evidence that the most prominent of these proteins is VDE, known previously as the homing endonuclease encoded by VMA1. We show also that GSH11 is not expressed in a VDE-deleted strain and that inability to express the GSH11 of this strain is overcome by introduction of the coding region of VDE or the entire VMA1 gene. It is also found that VDE does not cut DNA in the vicinity of the GSH11 cis-element. Rapamycin, an inhibitor of the target of rapamycin (TOR) signal-transduction system, is found to enhance expression of GSH11 in a VDE-dependent manner under conditions of sulfur starvation. These results indicate that GSH11 is regulated by a system sensitive to sulfur starvation (presumably via cysteine depletion) and a more general system involving the nutritional starvation signal mediated by the TOR system. Both systems need to be operational (inhibition of TOR and sulfur starvation) for full expression of GSH11.

Glutathione plays essential roles in the detoxification of various toxic agents (2)(3)(4)(5) including reactive oxygen species (6 -8). Depletion of glutathione causes an increase of reactive oxygen species and results in death of the cell (9 -11). Glutathione is the most abundant sulfur-containing compound in many organisms (12)(13)(14). It contains cysteine and produces cysteine by degradation. Thus, it acts as a reservoir of cysteine, thiol groups, and sulfur (15). Under conditions of sulfur starvation, glutathione is the last of various sulfur-containing compounds to decrease in the cell. 1 It is also known that, under conditions of sulfur starvation, Saccharomyces cerevisiae utilizes sulfate, produced by the consumption of glutathione, as a sulfur source (16). Therefore, it appears that maintenance of the intracellular level of glutathione is of great importance for viability in terms of sustaining sulfur metabolism and protecting the cell from oxidative stress.
Because many organisms have the ability to utilize exogenous glutathione, we have focused our attention on the glutathione uptake activity of S. cerevisiae and have shown previously that this organism has two kinetically distinguishable glutathione transport systems, i.e. an inducible high affinity transporter (GSH-P1) and a constitutive low affinity one (GSH-P2) (17). GSH-P1 mediates transport of pentapeptides as well as enkephalin, but it should be mentioned that the affinity of GSH-P1 to the oligopeptides is much lower than that to glutathione (1). GSH-P1 has characteristics common to the OPT family of transporters present in fungi but not in mammals and bacteria; this is the reason why it is also referred to as OPT1 (18,19). Recently, Arabidopsis has been shown to encode proteins belonging to this family (20). Glutathione and oligopeptides are convenient and useful sulfur sources if they are available in the environment. For this reason also, we are interested in the GSH-P1 transporter, particularly in the regulation of its production. We have already found that the gene coding for GSH-P1, i.e. HGT1/GSH11, has a novel cis-acting regulatory motif, CCGCCACAC, in its 5Ј upstream region (1).
In this study, we analyzed proteins that bound to the regulatory motif of GSH11 and found that the most prominent was VDE, the homing endonuclease derived from VMA1. We investigated the role of VDE in the regulation of GSH11 and also examined the effect of rapamycin, an inhibitor of the target of rapamycin (TOR) 2 signal transduction system, on the regulation of GSH11.

EXPERIMENTAL PROCEDURES
Strains and Media-The S. cerevisiae strains used in this study are listed in Table I. Strains SF1-1C and YPH500 are wild types in regard to the GSH-P1 activity and regulation of GSH-P1. They have the wild type allele in the VMA1 locus. Strain YOC2176 is a gene-manipulated derivative of strain YPH500; it has the VMA1-101 (vde-delta) allele in which the whole VDE coding region is deleted from VMA1 (21,22). Escherichia coli strain JM109 (TaKaRa) was used throughout to amplify plasmids. pMC1587, a plasmid containing the coding region of lacZ (23), was used for analysis of the promoter region of HGT1/GSH11. Plasmids pYO314-VMA1 and YCpTV-VDEc carry the wild type allele of VMA1 and the entire coding region of VDE, respectively (21,22).
Standard yeast growth media were used (24). YPD medium contained 1% yeast extract, 2% peptone, and 2% glucose. SD medium is a synthetic minimal medium (25). Sulfur-free (SF) medium was prepared by substituting all sulfate salts in SD medium with the corresponding chloride salts (26). To provide for the nutritional requirements of the strains used in this study, adenine (20 g/ml), lysine (30 g/ml), histidine (20 g/ml), leucine (30 g/ml), tryptophan (20 g/ml), and uracil * The work was carried out as a part of the Bio-Venture project and the 21st century Center of Excellence (COE) program of Ritsumeikan University. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
§ To whom correspondence should be addressed. (20 g/ml) were added to the SD and SF media. To obtain a medium with a defined sulfur source, cysteine was added to SF medium at a concentration of 100 M. Rapamycin (Sigma) was used at a final concentration of 200 nM. A 1 mg/ml stock solution of rapamycin was prepared in 90% ethanol and 10% Tween 20. For solid medium, 2% agar was added. The growth temperature was 30°C, and liquid cultures were rotary shaken at 120 rpm. LB medium (1% tryptone, 0.5% yeast extract, and 0.5% NaCl, pH 7.2) was used for growth of E. coli, (27). Ampicillin was added to LB medium at a concentration of 50 g/ml to screen for Amp r clones. For solid medium, 1.5% agar was added. The growth temperature was 37°C, and liquid cultures were rotary shaken at 120 rpm.
DNA Manipulations-Extraction of S. cerevisiae genomic DNA and transformation of S. cerevisiae were carried out as described (28,29). Standard DNA manipulation procedures were adopted (27).
Polymerase Chain Reaction-Amplification of DNA fragment by PCR was carried out using TaKaRa ExTaq TM polymerase and a thermal cycler (Atto). The PCR program we adopted was as follows: step 1, denaturation at 93°C for 1 min; step 2, 30 cycles of denaturation at 93°C for 1 min, annealing at 55°C for 1 min, and extension at 72°C for 1 min; and step 3, extension at 72°C for 10 min. The PCR reaction mixture (50 l) was prepared as described in the manufacturer's instructions.
Identification of DNA-binding Proteins-Cells of strain SF1-1C were grown overnight in YPD medium and transferred to SF medium. After incubation for 16 h, they were harvested by centrifugation and washed once with homogenization buffer (20 mM Tris-HCl, pH 8.0, 1 mM EDTA, 1 mM DTT, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride, and 1 mM benzamidine-HCl). The cells were resuspended in 0.3 ml of the same buffer to which was added 0.4-ml glass beads. The suspension was vortexed in an ice bath for 10 cycles of 30 s of shaking and 60 s of cooling. Cell debris was removed by centrifugation, and the supernatant was used as cell extract.
Two complementary 41-mer DNAs, corresponding to the GSH11 sequence from Ϫ371 to Ϫ331 (5Ј-TTCCGGCCCGCCACACCTCCGAC-TACAAGACGCCACATCTA-3Ј) containing the cis-element (under-lined), were custom synthesized, annealed, and used as a double strand DNA probe, referred to as ce16. The probe was end-labeled with terminal transferase and digoxigenin (DIG)-11-ddUTP according to the manufacturer's instructions (Roche Diagnostics). Following this, the DIGlabeled DNA probe, ce16-DIG, was bound to anti-DIG magnetic particle as described by the manufacturer (Roche Diagnostics).
Binding reactions were carried out in a 200-l binding buffer (10 mM Hepes (pH 7.6), 1 mM EDTA, 10 mM (NH 4 ) 2 SO 4 , 1 mM DTT, 1% (w/v) Tween 20, and 30 mM KCl) containing 7.5 pmol of ce16-DIG/anti-DIG particles and ϳ0.72 mg of cell extract protein. The binding reactions were incubated at 25°C for 30 min. Complexes of magnetic particles and the DIG-DNA probe with bound proteins were separated with a magnetic separator and washed four times with binding buffer.
Binding proteins were eluted by successive incubation with a total of 30 l (3ϫ 10 l) of homogenization buffer containing KCl, the concentration of which was increased stepwise from 0.4 to 1 M. Aliquots of the eluted fractions were fractionated on SDS-PAGE (5-20% gradient gel, 64 V, ϳ3 h), and the gels were silver-stained for observation.
For identification of the protein bands, the binding reaction was scaled up to a total of 900 l containing 105 pmol of magnetic particles and the DIG-DNA probe and ϳ3.9 mg of protein extract and concentrated by elution in a final volume of 20 l. After SDS-PAGE, the proteins were transferred to polyvinylidene difluoride membranes by electroblotting (Nova-blot: Amersham Biosciences). The membranes were stained with Coomassie Brilliant Blue, and bands of interest were cut out for protein sequencing, which was carried out by APRO Life Science Institute, Inc., Naruto, Japan.
Reporter Gene Assay-A DNA fragment corresponding to the segment from Ϫ371 to 33 in the 5Ј-upstream region of GSH11 was PCRamplified using the genomic DNA of strain SF1-1C as template. The primers used are 5Ј-AATACCCGGGTTCCGGCCCGCCACACCTCCGA-CTA-3Ј (forward) and 5Ј-CACAGGATCCAACGAGTCGCTCTCCCTAT-AA-3Ј (reverse). The PCR product was digested with SmaI and BamHI and inserted into the SmaI/BamHI site proximal to the coding region of lacZ in pMC1587 to generate plasmid EUG11-9; note that in the resultant fusion protein the first eight amino acids of the ␤-galactosidase were replaced with the first 11 amino acids of GSH11.
The structure of plasmid EUG11-9 was confirmed by sequencing, and EUG11-9 was used to transform strains YPH500, YOC2176, YOC2176/ pYO314-VMA1, and YOC2176/YCpTV-VDEc. After incubation of the  resultant transformants in SF medium or SF medium supplemented with 0.1 mM cysteine for 16 h, the cells were harvested and lysed. Then, the ␤-galactosidase activity was measured by mean of the hydrolysis of o-nitrophenyl-␤-D-galactopyranoside (ONPG) to produce o-nitrophenol (ONP) and galactose. ␤-Galactosidase activity (in units) was calculated using the formula A 420 ϫ 1000/min/ml/OD 600 at 37°C.
Northern Blot Hybridization-Strains YPH500 and YOC2176 were grown overnight in YPD medium. The cells were harvested and suspended in SF medium or SF medium supplemented with 0.1 mM cysteine. After incubation for 16 h, total RNA was extracted by the method of Elder et al. (30). RNA (20 g per lane) was fractionated by agarose gel electrophoresis and subjected to Northern blot hybridization using a 741-bp fragment corresponding to a portion (710 -1450) of GSH11 as the probe (1).
VDE Cleavage-An 846-bp DNA fragment including the cis-element of GSH11 was PCR-amplified using genomic DNA of strain SF1-1C as template. Primers used are 5Ј-AATACCCGGGTAGAGCCATAGTGTG-GCAGGA-3Ј (forward) and 5Ј-CACAGGATCCAACGAGTCGCTCTC-CCTATAA-3Ј (reverse). About 0.22 g of the PCR product was incubated at 37 or 25°C for 16 h with 2 units of purified VDE (New England Biolabs) dissolved in 20 l of buffer (100 mM KCl, 10 mM Tris-Cl, 10 mM MgCl 2 , and 1 mM DTT, pH 8.6) supplemented with 100 g/ml bovine serum albumin. The reaction was stopped by the addition of 0.5% SDS, and the samples were subjected to electrophoresis in 1ϫ Tris acetate buffer with EDTA (TAE) on a 0.7% agarose gel. After electrophoresis, the gel was stained with SYBR Gold (Molecular Probes) and photographed.

Detection of Proteins
Binding to the cis-Element of GSH11-We noted previously by using the gel mobility shift assay that S. cerevisiae has protein(s) that binds to the ciselement of GSH11 (1). We decided to extend this line of investigation. A DNA fragment (ce16) that contained the 5Ј-upstream regulatory element (CCGCCACAC) of GSH11 was DIGlabeled and adsorbed to magnetic beads (see "Experimental Procedures"). The magnetic beads were mixed with the cell extract, collected, thoroughly washed, and then suspended consecutively in homogenization buffer containing increasing concentrations of KCl (from 0.4 to 1 M). Each supernatant was subjected to SDS-PAGE. As shown in Fig. 1, a large proportion of the bound proteins eluted in the 0.4 M KCl fraction, and some eight bands could be distinguished in this fraction. Moreover, the 0.4 M KCl fraction caused a reduction in the electrophoretic mobility of the ce16 probe (data not shown). We conclude that this fraction contains protein(s) that bind to the ce16 fragment. The most prominent protein in this fraction had a size of ϳ50 kDa (Fig. 1). We determined its N-terminal amino acid sequence and obtained the sequence "*FAKGTNVL" (* indicates refractory to analysis). Using a homology search we found that VDE, the homing endonuclease encoded by VMA1, has the identical N-terminal amino acid sequence (Fig. 2).
Involvement of VDE in Regulation of GSH11-To investigate the involvement of VDE in the regulation of GSH11, we conducted reporter gene assays using a strain in which the VDE coding region of VMA1 was deleted. For this experiment we constructed a plasmid in which the 5Ј-upstream region (from Ϫ371 to 33) of GSH11 was inserted in front of the coding region of the E. coli lacZ gene of the expression vector, pMC1587. Strains YPH500 (VMA1) and YOC2176 (a vde-delta derivative of YPH500) were transformed with the resultant plasmid, designated EUG11-9. Representative transformants were incubated for 16 h in SF medium or SF medium supplemented with 0.1 mM cysteine. The transformants derived from the VMA1 strain had higher ␤-galactosidase activity in sulfur starvation conditions and markedly lower activity in the presence of cysteine in the growth medium. In contrast, those derived from the vde-delta strain had very low ␤-galactosidase activity in both sulfur starvation and cysteine-supplemented conditions (Fig. 3).
We assayed the mRNA contents of the above mentioned cells by Northern blot hybridization using GSH11 as the probe (see "Experimental Procedures"). The VMA1 strain revealed a substantial level of expression of GSH11 in SF medium, whereas the vde-delta strain revealed no detectable level of expression (data not shown). Moreover, we found that the vde-delta strain recovered the ability of regulation of GSH11 if transformed with either plasmid pYO314-VMA1 (carrying the entire region of VMA1) or plasmid YCpTV-VDEc (carrying the entire VDE coding region). When the representative transformants were subjected to the reporter gene assay, they, like the VMA1 strain (YPH500), yielded substantial levels of ␤-galactosidase activity in sulfur starvation conditions (Fig. 3). All these results clearly indicate that VDE is involved in the expression of GSH11. VDE Does Not Cut the GSH11 cis-Element-VDE is known as a homing endonuclease, but its target sequence remains obscure; that is, although single base substitutions reduced the susceptibility of target DNAs, none gained complete resistance to VDE (31,32). The precise boundary of the target sequence is also not clear. Nevertheless, the following sequence is thought to be the most likely target of VDE (2 indicates cleavage site) Because the cis-element (CCGCCACAC) of GSH11 (1) is somewhat similar to this sequence (underlined), we tested whether VDE binds to and cuts at the GSH11 cis-element. An 846-bp DNA fragment containing the GSH11 cis-element was PCR-amplified (see "Experimental Procedures"), and the product was mixed with VDE and incubated for 16 h at 37 or 25°C. We detected no sign of cleavage of the fragment (data not shown), leading to the conclusion that VDE does not cut DNA at this position. This is a strong indication that VDE acts as a transcriptional regulatory factor for GSH11.
Effect of Rapamycin on the Expression of GSH11-It is well established that S. cerevisiae cells respond to both the quality and quantity of nutrients and that the TOR proteins, via their kinase activities, are responsible for maintaining a balance between protein synthesis and degradation in response to changes in nutritional conditions (34). For example, the TOR system up-regulates genes involved in ribosome biosynthesis, and inhibition of the expression of these genes induces autophagy. It is also known that the TOR system modulates the expression of genes responsible for the biosynthesis of various amino acids and amino acid permeases (34). Because GSH11 is induced by deprivation of sulfur, we suspected that GSH11 would be under the control of the TOR system. To test this, we examined the effect of rapamycin, an inhibitor of the TOR system (34), on the expression of GSH11. We transformed strains YPH500 (VMA1), YOC2176 (vde-delta), YOC2176/ pYO314-VMA1, and YOC2176/YCpTV-VDEc with plasmid EUG11-9. The obtained transformants were incubated in SF medium in the presence or absence of 200 nM rapamycin and with or without 0.1 mM cysteine. It can be seen (Fig. 4) that in the VMA1 strain rapamycin caused an increase of ␤-galactosidase activity only in the absence of cysteine. Contrastingly, in the vde-delta strain rapamycin did not affect ␤-galactosidase regardless of the presence or absence of cysteine. This result strongly indicates that GSH11 is under the control of the TOR regulatory system and that this regulatory system is mediated by VDE. However, it appears that the expression of GSH11, unlike that of many other metabolic genes, is also coupled to another regulatory system specific for sulfur starvation and, especially, cysteine depletion (see "Discussion").

DISCUSSION
We have found that at least eighteight proteins bind to the 5Ј-upstream region of GSH11. The most prominent turned out to be VDE, a protein produced by cleavage of the VMA1 gene product Vma1p. Vma1p consists of three segments, A, B, and C, and is processed to produce two peptides, B and AC. Peptide B corresponds to VDE, and AC to the vacuolar H ϩ -ATPase (35). VDE acts as a homing endonuclease that cuts the VDE-free VMA1 product at a specific site, and this triggers DNA repair using the VDE-containing VMA1 gene on the homologous chromosome as template. As a consequence, a heterozygous diploid whose genotype is VMA1 (ABC)/VMA1-101 (AC) is converted to a VMA1 (ABC)/VMA1 (ABC) homozygous diploid. VDE is thus thought to be a kind of self-propagating, or selfish, gene (21,32,36). Although the behavior of VDE as a homing endonuclease is well established, its biological significance remains obscure. Our data, however, indicate that VDE binds to the regulatory element of GSH11, the gene encoding the high affinity glutathione transporter, without cutting DNA at this position. We have further demonstrated that VDE is essential for expression of GSH11. This is the first pointer to the otherwise obscure role of VDE. It should be emphasized that VDE homing takes place in the process of sporulation that is initiated by nitrogen starvation (33,37), which implies that VDE may be translocated from the cytoplasm to the nucleus to perform its homing function in sporulation. This behavior of VDE may be the key to its function as a transcriptional activator of GSH11, because for VDE to act as a transcriptional factor, its localization in the nucleus is a prerequisite. We speculate further that the homing activity of VDE and, thus, an increased level of expression of GSH11 during sporulation is vital for the survival of the cell by means of a supply of a useful sulfur source, glutathione.
In relation to nutrient responses, it is well established that various starvation-induced metabolic genes are negatively regulated by the TOR signal transduction system and that rapamycin inhibits this system and induces expression of starvation-induced metabolic genes (34). We found in this study that rapamycin enhanced transcription of GSH11 only under conditions of sulfur starvation, and this enhancing effect of rapamycin was mediated by VDE. This finding indicates that GSH11 is under the control of the TOR signal transduction system, although in a different way from the regulation of other metabolic genes, because in the case of GSH11 the rapamycin treatment potentiates starvation only in the absence of a sulfur source. It has recently been demonstrated that translocation of VDE is stimulated and stabilized by nitrogen starvation and inactivation of the TOR regulatory system (22). Taking this and our findings into account, we propose a model of how VDE and rapamycin regulate the expression of GSH11 (Fig. 5). In this model the following actions take place. 1) VDE is transported to the nucleus when starved for sulfur or nitrogen or carbon. 2) It binds to the regulatory element of GSH11 but does not cut the DNA; and 3) DNA-bound VDE enhances binding of sulfur-or cysteine-specific regulatory factors that stimulate transcription of GSH11. In the light of this model we are now seeking to identify proteins, other than VDE, that bind to the cis-acting regulatory motif of GSH11. It is also necessary to test whether starvation of sulfur, like starvation of nitrogen or carbon, stimulates translocation of VDE into the nucleus.
It is worth mentioning that several genes (or open reading frames) have 5Ј-upstream sequences identical or similar to the cis-acting motif of GSH11. Of these, CYS3 (coding for cystathionine ␥-lyase), MET13 (coding for a putative 5,10-methylenetetrahydrofolate reductase), SPE2 (coding for S-adenosylmethionine decarboxylase), and MET2 (coding for homoserine O-acetyltransferase) are unquestionably involved in sulfur metabolism (1). Whether VDE is involved in the regulation of these genes is another interesting question, and examination of this possibility is underway in our group.