The CDS1 gene encoding CDP-diacylglycerol synthase in Saccharomyces cerevisiae is essential for cell growth.

An open reading frame (CDS1) residing on chromosome II of Saccharomyces cerevisiae encodes a hydrophobic protein with a predicted molecular mass of 51,789 Da, which exhibits 29 and 37% amino acid sequence identities with CDP-diacylglycerol synthases reported from Escherichia coli and Drosophila, respectively. Induction of expression of a GAL1 promoter-driven CDS1 gene on a multicopy plasmid in a cds1 null mutant background resulted in synthase activity 10 times that of wild-type cells and an elevation in the apparent initial rate of synthesis of phosphatidylinositol relative to phosphatidylserine. Without induction, activity was reduced to 10% of wild-type levels, which was sufficient to support growth but resulted in an inositol excretion phenotype, and had an opposite effect on the above phospholipid synthesis. Null cds1 mutants were incapable of spore germination or vegetative growth and could not be complemented under uninduced conditions with a GAL1 promoter-driven CDS1 gene on a low copy plasmid. Therefore, the essential CDS1 gene encodes the majority, if not all, of the synthase activity. The lack of consensus RNA splice sites derived from the genomic CDS1 sequence predicts that the multiple subcellular locations for synthase activities do not arise through RNA processing events.

CDP-diacylglycerol is an important branch point intermediate in glycerophosphate-based phospholipid biosynthesis in both prokaryotic and eukaryotic organisms (1)(2)(3)(4). In eukaryotic cells, phosphatidic acid is converted either to CDP-diacylglycerol by the CDP-diacylglycerol synthase or to diacylglycerol by a phosphatase. In mammalian cells, CDP-diacylglycerol is the precursor to phosphatidylinositol (and its polyphosphorylated derivatives), phosphatidylglycerol, and cardiolipin; in yeast, it is also the precursor in the endoplasmic reticulum to the de novo synthesis of phosphatidylserine, phosphatidylethanolamine, and phosphatidylcholine. Diacylglycerol is the precursor to triacylglycerol, phosphatidylethanolamine, and phosphatidylcholine in all eukaryotic cells. Therefore, the partitioning of phosphatidic acid between CDP-diacylglycerol and diacylglycerol must be an important regulatory point in eukaryotic phospholipid metabolism. In all eukaryotic cells, CDP-diacylglycerol is required in the mitochondria for phosphatidylglycerol and cardiolipin synthesis and in the endoplasmic reticulum and possibly other organelles for the synthesis of phosphatidylinositol.
A cDNA (derived from the CDS gene) encoding a photoreceptor cell-specific isoform of CDP-diacylglycerol synthase has been isolated from Drosophila (5). The gene product shares sequence identity with CDP-diacylglycerol synthase from Escherichia coli (6,7), suggesting that this enzyme has been highly conserved during evolution. This particular isoform is an important regulator of the reutilization of phosphatidic acid for the formation of phosphatidylinositol 4,5-bisphosphate, which is the substrate for a phospholipase C-mediated signal cascade linked to a G-protein-initiated signal. Overexpression of the CDS gene increases the amplitude of the light response of photoreceptor cells, and cds mutant cells undergo light-dependent retinal degeneration dependent on phospholipase C function. However, the mutant flies develop normally, indicating that the synthase isoform responsible for bulk phospholipid synthesis is unaffected by this mutation. These results are also consistent with there being multiple synthase activities in higher eukaryotic cells derived from either a single gene or multiple genes, which is in contrast to E. coli, which encodes a single synthase (6,7).
CDP-diacylglycerol synthase has been purified from the yeast Saccharomyces cerevisiae (8) and appears to be composed of two identical 56-kDa subunits (9), although some preparations also contain variable amounts of a 54-kDa species (8). The enzyme has been localized to the endoplasmic reticulum, the cytoplasmic side of the outer mitochondrial membrane, and the inner mitochondrial membrane (10). This enzymatic activity is also enriched in the plasma membrane over total membranes (11), consistent with the finding that the activity is also enriched in post-Golgi secretory vesicles (12). Its product, CDPdiacylglycerol, may play an important role as both a precursor to phosphoinositide biosynthesis in the plasma membrane and as a negative effector of phosphatidylinositol 4-kinase activity, thereby exerting an effect on cell proliferation via a lipid-dependent signal transduction cascade (13). The multiple locations of this enzyme in yeast mirrors the results seen in Drosophila.
In order to gain more in-depth understanding of the cellular distribution, function, and regulation of CDP-diacylglycerol synthase in eukaryotic cells, we report in this paper the isolation of the CDS1 gene from S. cerevisiae. The gene product was verified by overexpression of CDP-diacylglycerol synthase in yeast transformants. By gene interruption, we demonstrate that CDS1 is an essential gene for cell growth and encodes the majority, if not all, of the synthase activity in yeast.

EXPERIMENTAL PROCEDURES
Materials-All chemicals were of reagent grade or better. Radiochemicals, Hybond N nylon membranes, and CTP were obtained from Amersham Corp. Liquiscint was purchased from National Diagnostics. Restriction endonucleases were from Promega Corp., New England * This work was supported in part by National Institutes of Health Grant GM20487. 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.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EMBL Data Bank with accession number(s) Z35898.
Biolabs, Stratagene, and Boehringer Mannheim. The Gene Amp PCR 1 reagent kit was from Perkin Elmer Cetus. The Genius 1 kit (DNA Labeling and Detection Kit, Nonradioactive), digoxigenin-labeled DNA molecular weight markers, positively-charged nylon membranes, and Lumi-Phos 530 were purchased from Boehringer Mannheim. Oligonucleotides were prepared commercially by Genosys Biotechnologies. Geneclean II kit, YEP broth, and synthetic media for yeast growth and selection were from BIO 101. Yeast nitrogen base without amino acids was from Difco. The BCA kit was from Pierce. L-␣-phosphatidic acid was from Sigma.
Strains, Plasmids, and Growth Conditions-A list of the strains and plasmids used in this work is given in Table I. Methods of yeast growth, sporulation, and tetrad analysis were as reported previously (14,15). YEPD medium consisted of 1% of Bacto-yeast extract, 2% Bacto-peptone, and 2% dextrose. In YEPR or YEPG medium, 2% of raffinose or galactose (glucose free), respectively, replaced dextrose as the carbon source. The induction medium YEPRG or repression medium YEPRD contained 2% glucose-free galactose or 2% glucose, respectively, in addition to raffinose. Complete synthetic media (CSM) was constituted as described previously (15) and contained the indicated sugar carbon source (D, G, or R). Yeast selection media contained the components of CSM except those noted for selection purposes (i.e. minus tryptophan (ϪTrp), minus leucine (ϪLeu), or minus uracil (ϪUra)). Yeast strains were grown at 30°C. E. coli strain DH5␣ was grown in LB medium (1% Bacto-tryptone, 0.5% Bacto-yeast extract, 1% NaCl, pH 7.4) at 37°C. Ampicillin (200 g/ml) was added to cultures of DH5␣ carrying plasmids. All the above media were supplemented with 2% agar (yeast) or 1.5% agar (E. coli) for preparation of agar plates.
DNA Manipulations-Methods for plasmid and genomic DNA preparation, restriction enzyme digestion, and DNA ligation have been described previously (16,17). Prior to transformation, E. coli cells were washed with cold, sterile water and concentrated in 10% glycerol as described in the Biotechnologies and Experimental Research, Inc. protocol for E. coli transformation. Concentrated cells were transformed by electroporation using a BTX electroporation system at 1.0 kV/mm and resistance of 129 ⍀. Yeast strains were transformed using a modified method based on the protocol for E. coli transformation above. Yeast cells were grown to stationary phase, washed with ice-cold water, and resuspended in sterile 10% glycerol. Cells were transformed by electroporation at 1.0 kV/mm and 186 ⍀.
Amplification of DNA by Polymerase Chain Reaction-For both analytical and preparative purposes, PCR was performed after optimizing conditions as described by Innis and Gelfand (18). Amplification of the CDS1 gene from a YES yeast cDNA library (19), from a YEp13 yeast genomic library (20), or from yeast chromosomal DNA employed the following primers: primer 1 (5Ј-CCATATCTCGAGAATGTCTGACAAC-CCTGAG-3Ј) and primer 2 (5Ј-CCGGTCTAGATCAAGAGTGATTGGT-CAATG-3Ј). They were designed according to the DNA sequence of open reading frame YBR029c (GenBank number Z35898); the underlined codons in primers 1 and 2 indicate the start and stop codons, respectively, for this open reading frame.
DNA Labeling and Detection-The Genius 1 kit was used according to the manufacturer's directions for preparation and detection of digoxigenin-labeled DNA probes. The kit utilized random priming of template DNA and incorporation of digoxigenin-dUTP into the probe. Template DNA was produced by PCR and isolated by agarose gel electrophoresis. The desired band was excised from the gel and extracted by using the Geneclean II kit. An antibody against digoxigenin coupled to alkaline phosphatase, which in the presence of Lumi-Phos 530 produces a chemiluminescent signal, was used to permit detection of hybridized probe by x-ray film.
Screening of a Genomic DNA Library-E. coli colonies bearing a yeast genomic DNA library carried on the E. coli-yeast shuttle vector YEp13 (20) were transferred to positively charged nylon membranes and screened for hybridization to the labeled PCR probe generated by using genomic DNA as template. Transfer of colonies to membranes, hybridization, and development of blots were carried out using the manufacturer's instructions for use of positively charged nylon membrane and the Genius 1 kit. SSC dilutions were prepared from 20 ϫ SSC (3 M NaCl, 0.3 M sodium citrate, pH 7.0). Hybridization was performed overnight at 68°C in hybridization solution (5 ϫ SSC, 0.5% Genius 1 kit blocking reagent, 0.1% N-lauroylsarcosine, 0.02% SDS) containing the labeled PCR probe (10 ng/ml). Following hybridization, membranes were washed twice for 5 min in 2 ϫ SSC, 0.1% SDS at room temperature and twice for 15 min in 0.1% SSC, 0.1% SDS at 68°C. Colonies corresponding to the positive signals on the blot were picked from the original plates for further screening, including a second round of hybridization screening and Southern blot analysis of their DNA.
Southern Analysis of Genomic DNA-DNA samples were digested with restriction enzymes and separated by agarose gel electrophoresis. DNA was transferred to positively charged nylon membranes by capillary transfer using 20 ϫ SSC at room temperature and then crosslinked to the membrane by using a UV Stratalinker 1800. The labeled PCR product used for library screening was also used for hybridization to the Southern blots. Methods for hybridization and development of blots were the same as those in library screening. For high stringency hybridization, membranes were washed twice for 5 min in 2 ϫ SSC, 0.1% SDS at room temperature and twice for 15 min in 0.1% SSC, 0.1% SDS at 68°C. For low stringency hybridization, membranes were washed twice for 5 min in 2 ϫ SSC, 0.1% SDS at room temperature, and twice for 15 min in the same wash buffer at 68°C.
Preparation of Cell Fractions and Measurement of CDP-diacylglycerol Synthase Activity-All cell fractionation procedures were carried out at 4°C. For the measurement of CDP-diacylglycerol synthase activity in the total membrane fraction, S. cerevisiae cells were grown to the exponential phase of growth, and the cells were collected by centrifugation in tared containers. Cells were washed in 50 mM Trismaleate, pH 6.5, 0.25 M sucrose, 1 mM phenylmethylsulfonyl fluoride; centrifuged; and resuspended in the same solution. The cell suspension was mixed with an equal volume of pre-chilled silicon beads (diameter, 0.3 mm) and disrupted in a Mini-Beadbeater (Biospec Products) by six 15-s bursts with a 2-min pause between bursts. Silicon beads and unbroken cells were removed by centrifugation at 2,000 ϫ g for 1 min. The total membrane fraction was separated from the cytoplasmic fraction by centrifugation at 100,000 ϫ g for 1 h. The membrane pellet was suspended in 50 mM Tris-maleate, pH 6.5, containing 40 mM MgCl 2 . CDP-diacylglycerol synthase activity was measured by the incorporation of [ 3 H]CTP into chloroform-soluble material dependent on phosphatidic acid as described previously (9). For the measurement of CDP-  diacylglycerol synthase activity in the mitochondria and microsomes, yeast cells were grown to an optical density (A 600 ) of 1.0. Isolation of yeast organelles was carried out by the method of Zinser and Daum (21). Mitochondria were isolated from the cell-free homogenate (3,000 ϫ g supernatant) by centrifugation at 9,000 ϫ g for 10 min, while microsomes were collected by centrifugation of the postmitochondrial supernatant at 100,000 ϫ g for 1 h. Plasmid Shuffling and Overexpression of CDS1 by Galactose Induction-Plasmids (URA3) carrying the CDS1 gene under the regulation of the GAL1 promoter were introduced into the cds1 null mutant by "plasmid shuffling" as follows. Briefly, strain YSD90A/YEp30 was transformed with a P GAL1 -CDS1 plasmid. A transformant carrying both plasmids was grown in YEPG medium for 2 overnights before plating for single colonies on CSMG-URA plates to select cells still containing the P GAL1 -CDS1 plasmid; resulting colonies were screened for lack of growth on CSMD-LEU and CSMG-LEU plates to verify the absence of plasmid YEp30. Induction of the expression of CDS1 from the GAL1 promoter was carried out as follows. Cell transformants growing exponentially in either CSMR-URA or YEPR medium were diluted to an A 600 of 0.1 unit with the same medium. Galactose (2%) was added to the cell culture when A 600 reached 0.75 unit, and growth was continued before harvesting at various times; parallel growth after addition of glucose to 2% was carried out for those transformants that can grow in the presence of glucose. After preparation of membrane fractions, samples were diluted to 0.2-0.8 mg/ml of protein concentration before they were assayed for CDP-diacylglycerol synthase activity.
Labeling and Analysis of Phospholipids-The CDS1 haploid strain YPH102 and the cds1::TRP1/P GAL1 -CDS1 transformant YSD90A/ pSDG1 were grown in YEPR medium to the exponential phase of growth (A 600 about 1.0 -1.2). For steady state labeling, 5-ml aliquots of YEPRD and YEPRG medium were inoculated with the above strains to an A 600 of 0.05. Following the addition of 50 Ci of [ 32 P]orthophosphate, cells were grown for 16 h (six generations) to assure uniform labeling before harvesting by centrifugation at 1,500 ϫ g. The cell pellets were resuspended in 1 ml of 80% ethanol and incubated at 80°C for 15 min. After centrifugation at 1,500 ϫ g, the pellets were suspended in 0.67 ml of chloroform, methanol, 0.1 N HCl (1:2:0.8 (v/v)). Cells were lysed using glass beads, and the phospholipids were extracted as described previously (22). Isolated radiolabeled phospholipids were applied to boric acid-impregnated silica gel plates (8), which were developed in one dimension with chloroform/methanol/water/ammonium hydroxide (60: 37.5:3:1) as the solvent system. Labeled phospholipids were detected and quantified directly from the thin layer plate using a Betascope (Betagen Corp.). For pulse labeling of phospholipids, 250 Ci of [ 32 P]orthophosphate was added to each 5-ml cell culture (A 600 ϭ 1.0) in either YEPRD or YEPRG medium. Cells were grown for 30 min before harvesting. Phospholipid extraction and analysis were by the same procedure as described above.
Assessment of Inositol-Excretion Phenotype-The inositol excretion capacity of yeast strains was tested on CSMD-URA or CSMG-URA plates lacking inositol, choline, and ethanolamine (I Ϫ C Ϫ E Ϫ ) as reported previously (23) using growth of the inositol auxotrophic reporter strain MC13 (ino1). The yeast strain to be tested was patched onto the I Ϫ C Ϫ E Ϫ tester plate and permitted to grow for 24 h. The inositol auxotrophic reporter strain was then streaked away from the patch as described previously (24), and cross-feeding was scored after an additional 48-h incubation.

RESULTS AND DISCUSSION
Screening of the Genomic Library-Amino acid sequences of CDP-diacylglycerol synthases both from Drosophila (5) and from E. coli (6) were compared with DNA data bases by using the TBLAST search protocol (25) at the National Center for Biotechnology Information (Bethesda, MD). An unassigned open reading frame on chromosome II from S. cerevisiae corresponding to the open reading frame denoted as YBR0313 (26) or YBR029c (GenBank number Z35898) encodes a protein with high homology to the above two enzymes. PCR primers were designed according to the DNA sequence of the above open reading frame to generate a DNA fragment encoding the putative yeast CDP-diacylglycerol synthase. Only one PCR product with a length of about 1.4 kb was generated either from a YES yeast cDNA library (19) or from a YEp13 yeast genomic library (20). Digoxigenin-labeled PCR product was used for Southern blot analysis and for genomic library screening. Southern blot analysis of strain YPH102 genomic DNA digested either with EcoRI or HindIII restriction enzyme revealed only one hybridization-positive DNA fragment of either 4.9 or 3.7 kb, respectively, under both low-or high-stringency hybridization conditions (data not shown). This result is consistent with the restriction digestion map of the open reading frame YBR0313 (26) and its surrounding region on chromosome II. Screening of the YEp13 yeast genomic DNA library carried in E. coli strain DH5␣ was performed as described under "Experimental Procedures." Out of 70,000 colonies screened, two positive colonies containing plasmids YEp2 and YEp30 were identified as potential candidates carrying the putative yeast CDS1 gene. Restriction mapping of these two clones revealed that they were identical (data not shown).
Overexpression of CDP-diacylglycerol Synthase in Yeast-Total membrane fractions from exponential phase cultures of the strain YPH102 transformed with either plasmid YEp2 or YEp30 and grown on CSMD-LEU medium were examined for CDP-diacylglycerol synthase activity. Enzyme activity in strain YPH102 was increased about 10-fold when carrying either plasmid YEp2 or plasmid YEp30 as shown in Table II. Since both plasmids exhibited identical restriction patterns and brought about a similar overproduction of enzyme activity, they appeared to carry the same CDS1 structural gene. A 1.7-kb SspI fragment from the genomic clone was subcloned into the plasmid pYES2 downstream of P GAL1 to yield plasmid pSDG1, which was introduced into the wild-type yeast strain YPH102; this construct carries 108 bp of CDS1 gene upstream sequence after the P GAL1 . Induction of the CDS1 gene in CSMG-URA medium brought about an 10-fold overexpression of CDP-diacylglycerol synthase activity (Table II) when compared with either strain YPH102 alone or strain YPH102/ pSDG1 (not shown) grown on CSMD-URA medium (uninduced).
CDS1 Gene and CDP-diacylglycerol Synthase-Restriction digestion mapping and DNA sequencing from both ends (total of approximately 800 bp) of the gene (1371 bp) confirmed that the CDS1 gene is identical with open reading frame YBR029c (GenBank number Z35898). Inspection of the DNA sequence did not reveal any sequence motifs (27), suggesting the existence of introns in or near the CDS1 gene, which minimizes the possibility of multiple synthases being derived from alternate splicing of a common RNA transcript. In the 3Ј-region, two potential transcriptional termination sequences (28, 29) (1416-TAG⅐⅐⅐T-rich region⅐⅐⅐TAG⅐⅐⅐TATGT⅐⅐⅐AT-rich region⅐⅐⅐TTT-1502 and 1465-TAGNNTATGTA-1475) and a polyadenylation site (30) (1548-AATAAA-1553) were found. A sequence (ATGT-GAAAA) homologous to the upstream activation sequence UA-S INO (1,31,32) was found beginning 161 bp 5Ј to the gene. No recognizable TATA promoter element was observed in this region. UAS INO has been found 5Ј to several genes (INO1, CHO1, PEM1, PEM2, PIS, and PSD1) related to phospholipid metabolism (1,33). This sequence appears to be related to the coordinate transcriptional depression of the expression of phospholipid biosynthetic enzymes via the products of the INO2 and INO4 genes (31) when cells are grown in the absence of inositol and either choline or ethanolamine. Its presence upstream of the CDS1 gene may explain the decreased level of CDP-diacylglycerol synthase when cells are grown in the presence of inositol, choline, and ethanolamine (34). Although multiple copies of this regulatory sequence exist in the promoter region of many of the genes reported above, only one conserved UA-S INO is located 5Ј to the CDS1 gene. This may explain the smaller response in synthase activity (about a 2-fold range) to these water-soluble precursors to phospholipids (34) when compared with the response of other enzymes of phospholipid metabolism (3-4-fold range). The predicted open reading frame encodes a protein of 457 amino acids with a molecular mass of 51,789 Da. A comparison of the deduced yeast CDP-diacylglycerol synthase amino acid sequence with the sequences of the CDP-diacylglycerol synthases from E. coli (6) and Drosophila (5) revealed a high degree of homology as shown in Fig. 1. The amino acid sequence from S. cerevisiae shares 37% identity and 60% similarity with the Drosophila enzyme, and 29% identity and 56% similarity with the E. coli enzyme. In regions that were highly conserved among the three enzymes, identities of approximately 90% were observed. Hydrophobicity analysis of the yeast enzyme by the method of Kyte and Doolittle (35) showed a highly hydrophobic protein containing several potential membrane spanning domains and a hydrophilic N terminus, which is absent in the E. coli CDP-diacylglycerol synthase. The profile of the yeast and Drosophila enzymes are remarkably similar. Analysis of the sequence by the PSORT program (36) showed no potential mitochondrial targeting sequence within the N-terminal hydrophilic region or any potential endoplasmic retention sequence (KKXX or HDEL) at the C terminus (37), although this enzyme activity has been localized primarily to the mitochondria and the endoplasmic reticulum (10). The lack in E. coli of the N-terminal and C-terminal extensions found in the eukaryotic enzymes may indicate that these sequences are required for organelle targeting. The PSORT program did predict possible plasma membrane and Golgi body localization for this sequence, which is consistent with finding the activity in secretory vesicles and the plasma membrane (11,12).
The CDP-diacylglycerol synthase isolated from the total membranes of yeast (minus the nuclear fraction) was reported to have a molecular weight of 56,000 (9). The discrepancy with the predicted molecular weight could be explained by the inaccuracy of SDS gel electrophoresis methods with membraneassociated proteins. However, any additional posttranslational processing normally associated with organelle targeting of proteins would only make this discrepancy greater. Since the CDS1 gene encodes most if not all of the synthase activity in the cell (see below) and there are no sequence motifs consistent with alternate splicing of the primary transcript, how is this activity directed to multiple sites in the cell? Are there additional synthase isoforms that have not been yet identified? There is precedence in yeast for the use of alternative AUG start sites on a common RNA transcipt for the synthesis of tRNA modification enzymes, which are localized to the mitochondria, cytoplasm, and nucleus (38 -40). The protein products from a common transcript have different N termini, which appears to account for their different locations in the cell. In contrast to the CDP-diacylglycerol synthase, the N termini of the complete open reading frames of these modifying enzymes contain predicted mitochondrial targeting sequences. Phosphatidylserine synthase of yeast is found associated with both the mitochondria (most likely the outer membrane) and the endoplasmic reticulum (41). Both forms have the same molecular weight of 30,804 and are derived from the CHO1 gene; whether this dual localization occurs in vivo and how it may happen has not been resolved. Removal of the first AUG of the open reading frame of the CHO1 gene results in localization of the majority of the protein (molecular weight of 22,400) to the cytoplasm in an inactive form but with sufficient membraneassociated active protein of reduced size to complement a cho1 mutant (42). This N terminus region cannot localize a soluble marker protein to either membranes or specific organelles, yet it appears to be important in localizing the CHO1 gene product. Translation beginning at the second methionine of the CDPdiacylglycerol synthase would only reduce the size of the protein by 600 mass units, which would be indistinguishable from the complete open reading frame by SDS gel electrophoresis methods but may inactivate an unrecognized mitochondrial targeting sequence. The third AUG lies 72 codons from the start of the open reading frame, which would encode a protein much too small to be in agreement with the size of major protein thus far isolated; however, a smaller minor isoform of the synthase may have been missed in the earlier work. Such a smaller product might still be active and membrane associated (with the plasma membrane for example) because it would still retain the regions homologous to the E. coli enzyme.
Disruption of the CDS1 Gene-The CDS1 gene was disrupted in vitro and introduced into the genome by homologous recombination as described below. A 3-kb EcoRI-BamHI fragment from plasmid YEp30 containing the CDS1 gene was subcloned into pUC19. The NruI-MscI region (853 bp in length) internal to the CDS1 gene was replaced by a 2 kb PstI-BspHI fragment carrying the TRP1 gene from plasmid pRS304 (43). The disrupted CDS1 gene (cds1::TRP1) was excised by EcoRI-BamHI digestion and used to transform the trp1 homozygous diploid YPH501 (43). Tryptophan prototrophy (growth on CSMD-TRP plates) was used to select for replacement of one of the CDS1 genes by homologous recombination with the cds1::TRP1 fragment. PCR reactions using genomic DNA of the interrupted diploid as template confirmed that the disrupted CDS1 gene had integrated at the CDS1 locus of one of the two chromosomes (Fig. 2).
The CDS1/cds1::TRP1 heterozygous diploid strain YSD3 was sporulated and subjected to tetrad analysis. Each of the 10 tetrads dissected gave rise to only two viable spores, all of which were tryptophan auxotrophs; no spores with a TRP1 phenotype survived. Inspection of the nonviable spores by mi-croscopy showed that none of them had undergone germination, single cell division, or budding, indicating that the residual amount of CDP-diacylglycerol synthase in the spore was not sufficient for germination. The 2:2 ratio of viable to nonviable spores, together with the segregation of tryptophan auxotrophy and the CDS1 gene with the viable spores, indicates that the CDS1 gene is essential for cell growth.
In order to rescue the nonviable spores, plasmid YEp30 (CDS1), which also carries a LEU2 marker, was transformed into the heterozygous diploid YSD3 (CDS1/cds1::TRP1) prior to sporulation. Each of the four tetrads dissected gave rise to four viable spores. These spores were tested for growth in the absence of leucine and tryptophan. Among the four spores within each tetrad, two were tryptophan auxotrophs (YSD90B, YSD90D) and the other two were prototrophic for tryptophan (YSD90A, YSD90C). All spores were leucine prototrophs, indicating that plasmid YEp30 segregated efficiently during meiosis and that a plasmid-borne copy of the CDS1 gene had rescued the nonviable spores. These spores were grown in YEPD medium for one or two overnights, and the cell cultures were sampled. After 24 h of growth, 90% of the tryptophan auxotrophs (CDS1 wild type) were leucine auxotrophs, indicating that the wild-type cells lost the YEp30 plasmid rapidly. The tryptophan prototrophs (cds1::TRP1) remained prototrophic for leucine even after 48 h of growth. Supplementation of the liquid growth medium and the selection plates with choline and ethanolamine did not result in the loss of the covering plasmid YEp30 from the null mutants. Therefore, lack of dependence on CDP-diacylglycerol for phosphatidylethanolamine and phosphatidylcholine biosynthesis by utilization of the diacylglycerol-dependent pathway does not suppress the need for the CDS1 gene.
Regulated Expression of the CDS1 Gene-To study the cellular response to different CDP-diacylglycerol synthase levels, plasmid pSDG1 (P GAL1 -CDS1, multicopy) was introduced into the null mutant YSD90A (cds1::TRP1) by "plasmid shuffling" as described under "Experimental Procedures." This transformant also showed an 10-fold increase of the CDP-diacylglycerol synthase activity above the wild-type yeast background, which was dependent on growth in CSMG-URA induction medium (Table II). The increase in CDP-diacylglycerol synthase specific activity relative to a wild-type control was the same (9-fold) in the mitochondrial-and the endoplasmic reticulum-enriched fractions dependent on galactose induction of the only CDS1 gene in a haploid cds1 null background. One-third of the total enzyme activity was in the mitochondria, and the remainder was in the microsomal fraction as was also observed in wildtype cells lacking any plasmids. Although the reported distribution of activity between these two organelles varies (8,10,44), both fractions were proportionately enriched in synthase activity when the CDS1 gene was overexpressed, indicating that the CDS1 gene product is associated with both subcellular fractions. When strain YSD90A/pSDG1 was grown in either CSMD-URA or CSMR-URA medium (noninducing conditions), only 10% of the wild-type CDP-diacylglycerol synthase activity was detected (Table II), which was sufficient to support robust cell growth on agar plates and in liquid medium. Introduction of the low copy number plasmid pSDG2 (P GAL1 -CDS1), which carries a LEU2 marker into strain YSD90A, also supported growth and overproduction of synthase activity (10-fold) in CSMG-LEU medium; this plasmid contains no DNA derived from the 5Ј upstream region of the CDS1 gene. However, unlike the high copy number plasmid pSDG1, the latter plasmid could not complement the null allele when grown on CSMD-LEU agar plates. Plating cultures for single colonies resulted in the appearance after 4 days incubation of very small colonies, which when restreaked to CSMD-LEU plates did not form single colonies. Therefore, the original colonies appear to have resulted from utilization of residual synthase activity produced under induction conditions followed by cell arrest on glucosecontaining media once insufficient synthase activity was present to sustain growth. Similarly, liquid cultures of this transformant growing in CSMG-LEU arrested several generations after switching to CSMD-LEU media. These results are consistent with the earlier conclusion that the CDS1 gene is essential and encodes the majority of the synthase activity. The ability of plasmid pSDG1 and not plasmid pSDG2 to complement the null allele under repressed conditions is most likely due to leak through transcription, which would result in higher levels of transcript from the multicopy plasmid (about 10 -20 copies/cell) than the low copy number plasmid (about 1-2 copies/cell) (45).
Inositol Excretion Phenotype-The cdg1 mutant of yeast (46) exhibits about a 75% reduction in the derepressed level of CDP-diacylglycerol synthase activity. The synthase activity in this mutant also no longer responds to regulation by inositol and choline, and the mutant excretes inositol into the growth medium. The product of the CDG1 gene has not been established. In order to determine whether inositol excretion could be caused by simply reducing the steady state level of synthase activity, strain YSD90A/pSDG1 was streaked as patches to I Ϫ C Ϫ E Ϫ plates containing glucose (uninduced) or galactose (induced) and grown for 24 h to test for inositol excretion. The inositol auxotrophic strain MC13 (ino1) showed detectable and robust growth 24 -48 h after being streaked next to the patches of strain YSD90A/pSDG1 grown under uninduced conditions but not next to the patches grown under induced conditions; strain YPH102 did not exhibit an inositol excretion phenotype under either growth condition. Therefore, an inositol-excretion phenotype is exhibited by yeast with depressed levels of CDPdiacylglycerol synthase similar to that observed in the cdg1 mutant.
CDS1 Expression and Phospholipid Metabolism-Strain YSD90A/pSDG1 was pulse-labeled with [ 32 P]orthophosphate in both YEPG and YEPD media to examine the initial rate of phospholipid synthesis as a function of the capacity to make CDP-diacylglycerol (Fig. 3A) versus the wild-type strain YPH102 grown under similar conditions; the results for the latter strain were independent of the carbon source. Changes in the relative percent incorporation of label into the various phospholipid classes during a pulse labeling experiment should be related to changes in the initial rate of synthesis of each phospholipid. The most significant difference brought about by overproduction of the synthase (induced) is a marked increase in the rate of synthesis of phosphatidylinositol and a decrease in the rate of synthesis of phosphatidylserine and its downstream metabolic products. A 90% reduction in the level of the synthase over wild-type levels (uninduced) resulted in a significant increase in phosphatidylserine labeling with a reduction in phosphatidylinositol labeling. The increase in phosphatidic acid from 0.2% for wild-type and induced cells to 0.8% for uninduced synthase was significant and reproducible consistent with this synthase being involved in the major phospholipid biosynthetic pathways of the cell. To analyze phospholipid composition, both strain YSD90A/pSDG1 and strain YPH102 were labeled to steady state with [ 32 P]orthophosphate in the above media as described under "Experimental Procedures" (Fig. 3B). Strain YPH102 grown in YEPG medium (data not shown) gave the same results as cells grown in YEPD medium. Except possibly for phosphatidylethanolamine, the differences in labeling patterns among the strains reflected the pulse labeling results. Relative to wild-type cells, overproduction of the syn-thase increased the proportion of phosphatidylinositol while underexpression of the synthase reduced the proportion of phosphatidylinositol. Although the levels were low, phosphatidic acid appears to be elevated under uninduced conditions and cardiolipin levels were highest under induced conditions. These results are consistent with variations in either the steady state level of CDP-diacylglycerol or its rate of synthesis affecting the relative rate of synthesis of phospholipid at this branch point in metabolism. Accumulation of phosphatidic acid and increases in cardiolipin are also consistent with low and high levels, respectively, of synthase activity.
The fact that only 10% of the wild type level of the synthase FIG. 3. Dependence of phospholipid metabolism on the level of CDP-diacylglycerol synthase activity. A, the relative initial rate of biosynthesis of individual phospholipids was estimated by the percent distribution of [ 32 P]orthophosphate into the indicated phospholipids in 30 min. Individual determinations varied by less than 4% of the respective mean of each duplicate determination. B, steady state phospholipid composition was estimated by the percent distribution of [ 32 P]orthophosphate incorporated into the indicated phospholipids after long term labeling as described under "Experimental Procedures." Individual determinations varied by less than 7% of the respective mean of each duplicate determination. Wild type strain YPH102 was grown in YEPD (f) and the cds1::TRP1 mutant strain YSD90A transformed with the overexpression plasmid pSDG1 was grown in either YEPG (u induced) or YEPD (Ⅺ, uninduced). Phospholipids were extracted and separated as described under "Experimental Procedures." can support normal growth and near normal lipid composition is consistent with similar results for other phospholipid biosynthetic enzymes seen in E. coli (47), which indicates that the catalytic capacities of these biosynthetic enzymes are in large excess and that their activities, as has been demonstrated in yeast (1), are highly regulated in response to growth conditions. However, changes over a 100-fold range in the level of synthase activity, which are presumably somewhat reflected in the steady state level of CDP-diacylglycerol, should have significant effects on the partitioning of product at this branch point in metabolism. The excretion of inositol under repressed expression of the P GAL1 -CDS1 gene in the null background is consistent with the level of CDP-diacylglycerol being limiting for phosphatidylinositol synthesis. The phosphatidylinositol synthase would appear to be more sensitive to changes in the in vivo concentration of CDP-diacylglycerol than the phosphatidylserine synthase, even though the affinity for CDP-diacylglycerol by these two enzymes measured in vitro at saturation for their second substrates is the same (48 -50). With two substrate enzymes, the apparent K m of one substrate is inversely related to the concentration of the second substrate when the latter is below its saturation concentration. At physiological concentrations of serine, the phosphatidylserine synthase should be saturated for serine and operating at its minimum apparent K m for CDP-diacylglycerol (50). However, the physiological concentration of inositol is 9-fold below its K m in the case of the phosphatidylinositol synthase (50). Therefore, the apparent K m of the latter enzyme for the CDP-diacylglycerol should be much higher than the former enzyme under physiological conditions, which is reflected in the effects brought about by overproduction and repression of CDP-diacylglycerol synthase activity under the control of P GAL1 .
Conclusions-Clearly the CDS1 gene encodes an essential CDP-diacylglycerol synthase activity associated with both the yeast endoplasmic reticulum and mitochondrial fractions. Results with complementation of the cds1 null mutant with plasmids pSDG1 and pSDG2 demonstrate that the CDS1 gene encodes more than 90% of the synthase activity in the cell and does not encode an activity that is targeted solely to either the mitochondria or the endoplasmic reticulum. If a second CDS gene exists, as is the case with the expression of phosphatidylserine decarboxylase (PSD1 and PSD2) activity (33,51,52), it would account for significantly less than 10% of the total activity. Unlike the PSD genes (52,53), which can complement each other, a possible second CDS gene does not support growth in the absence of the CDS1 gene.
There is a clear difference in the germination phenotype of null mutants of the CDS1 and PIS (encoding phosphatidylinositol synthase) genes. Although both genes are essential for vegetative growth, null spores derived from heterozygous null/ wild-type diploids of the PIS gene undergo sporulation and at least one cell division before arresting with buds (54), while spores containing the null cds1 gene do not germinate. Therefore, supplying CDP-diacylglycerol for functions other than bulk phosphatidylinositol biosynthesis may be crucial to cell viability. One such function might be in supplying substrate for the plasma membrane-associated signal transduction pathway responsible for phosphoinositide formation, which has been linked to regulation of cell growth in yeast (13,55).