Reduction of CDP-diacylglycerol synthase activity results in the excretion of inositol by Saccharomyces cerevisiae.

A yeast mutant, cdg1, was isolated on the basis of an inositol excretion phenotype. This mutant exhibited pleiotropic deficiencies in phospholipid biosynthesis, including reduced levels of CDP-diacylglycerol (DAG) synthase activity (Klig, L. S., Homann, M. J., Kohlwein, S. D., Kelley, M. J., Henry, S. A., and Carman, G. M. (1988) J. Bacteriol. 170, 1878-1886). In this study we present evidence that the molecular basis for the inositol excretion phenotype is a G305/A305 point mutation (Cys102 → Tyr substitution) within the CDS1 gene (encodes CDP-DAG synthase) of this mutant. Expression of CDP-DAG synthase activity from a plasmid-borne copy of the CDS1 gene in the cdg1 mutant was not down-regulated, and this expression also corrected the inositol excretion phenotype. Introduction of the above mutated gene (CDS1*) controlled by its endogenous promoter on a single copy plasmid into a cds1-null background reconstituted a transformant with the cdg1 phenotype, including reduced CDP-DAG synthase activity, elevated phosphatidylserine synthase activity, and inositol excretion into the growth medium. Expression of CDS1* in a single copy in the cdg1 mutant raised CDP-DAG synthase activity from 15 to 30% of derepressed wild-type yeast levels but still did not correct the inositol excretion phenotype. CDP-DAG synthase activity was not regulated in response to precursors of phospholipid biosynthesis in the cdg1 mutant either with or without a trans copy of the CDS1* gene. An open reading frame was identified 5′ to the CDS1 locus, YBR0314, which also resulted in inositol excretion when present in trans in multiple copies.

phenotype. Introduction of the above mutated gene (CDS1*) controlled by its endogenous promoter on a single copy plasmid into a cds1-null background reconstituted a transformant with the cdg1 phenotype, including reduced CDP-DAG synthase activity, elevated phosphatidylserine synthase activity, and inositol excretion into the growth medium. Expression of CDS1* in a single copy in the cdg1 mutant raised CDP-DAG synthase activity from 15 to 30% of derepressed wild-type yeast levels but still did not correct the inositol excretion phenotype. CDP-DAG synthase activity was not regulated in response to precursors of phospholipid biosynthesis in the cdg1 mutant either with or without a trans copy of the CDS1* gene. An open reading frame was identified 5 to the CDS1 locus, YBR0314, which also resulted in inositol excretion when present in trans in multiple copies.
CDP-diacylglycerol (DAG) 1 is an important branch point intermediate in phospholipid biosynthesis in both prokaryotic and eukaryotic organisms (1)(2)(3)(4). It is the precursor to the de novo synthesis of all the sphingolipids and glycerophosphatebased phospholipids in yeast (1,5,6). Besides being an early precursor to the formation of phosphoinositide derivatives in yeast, CDP-DAG may play a regulatory role in their formation and thereby may be important in the control of cell growth in this organism (7). In yeast, CDP-DAG synthase activity is found predominantly in the endoplasmic reticulum and mitochondria (8,9), although it has been reported in the plasma membrane and post-Golgi secretory vesicles (7,10). The yeast enzyme has been purified to near homogeneity (9), is composed of two subunits, and has a native molecular mass of 114 kDa (9,11). CDP-DAG synthase activity is regulated by water-soluble precursors of phospholipid biosynthesis. Inclusion of inositol in growth medium reduces synthase activity, and this repressive effect is enhanced by the addition of ethanolamine or choline (12). The CDP-DAG synthase activity also decreases as cells enter the stationary phase in inositol-containing medium (13).
A yeast mutant (cdg1) was isolated on the basis of an inositol excretion phenotype (14). This mutant exhibited pleiotropic deficiencies in phospholipid biosynthesis, including a reduced rate of CDP-DAG synthesis, an elevated phosphatidic acid content, constitutive levels of inositol 1-phosphate synthase, and elevated phosphatidylserine (PS) synthase protein and activity. Biochemical and immunoblot analyses revealed that the defect in in vivo CDP-DAG synthase activity in the cdg1 mutant was due to a reduced level of the 56-kDa CDP-DAG synthase subunit rather than to an alteration in the enzymological properties of the enzyme. Genetic analysis indicated that a defect in a single gene was responsible for the mutant phenotype, and this gene was not allelic to any previously described genetic locus involved in phospholipid synthesis or regulation. Since no regulatory gene products had been identified in yeast that exhibited simultaneous negative and positive regulation of different parts of the phospholipid biosynthetic machinery, as appears to be affected in the cdg1 mutant, Klig et al. (14) argued that it is unlikely that the mutation in cdg1 is in a regulatory process. One possibility suggested was that this lesion might be in one of multiple copies of the structural gene(s) encoding CDP-DAG synthase activity, but their data were not sufficient to rule out the possibility that the mutation is in a regulatory gene the product of which controls expression of several phospholipid biosynthetic enzymes (3,5).
The CDS1 gene encoding CDP-DAG synthase has been isolated in the yeast Saccharomyces cerevisiae (15) based on its sequence homology with CDP-DAG synthases from Escherichia coli (16) and Drosophila (17). A UAS INO regulatory sequence is found in the 5Ј-untranslated region of the CDS1 gene. This unique sequence serves as a binding site for INO2 and INO4 gene products and thus mediates coordinated regulation of phospholipid biosynthetic gene expression (1, 18 -20), which may explain the decreased level of CDP-DAG synthase activity when yeast cells are supplied with inositol, choline, and ethanolamine in growth medium (12). The CDS1 gene encodes the majority, if not all, of the CDP-DAG synthase activity in yeast. A null cds1 mutant is incapable of spore germination or vegetative growth. Overexpression of CDP-DAG synthase activity results in an elevation in the apparent initial rate of synthesis of phosphatidylinositol (PI) relative to PS. A reduction of CDP-DAG synthase activity to 10% of the wild-type yeast level has an opposite effect on the above phospholipid biosynthesis. Reduced synthase levels also result in inositol excretion, a phe-notype similar to that of the cdg1 mutant. Complete repression of a single copy GAL1 promoter-driven CDS1 gene in a cds1null background results in arrest of cell growth. Therefore, all evidence is consistent with a single gene locus in yeast encoding CDP-DAG synthase activity, making the mutated locus in the cdg1 mutant either allelic with CDS1 or a regulatory gene. Given the dramatic effect this mutation has on regulation of phospholipid metabolism in yeast, it is important to determine the molecular basis for these perturbations.
In this report, we examine the relationship between CDS1 and the mutation in cdg1 cells. We present evidence that overexpression of CDP-DAG synthase activity complements the cdg1 mutant phenotype and that a G 305 /A 305 point mutation in the CDS1 locus (CDS1*) is responsible for the cdg1 phenotype. By introducing a single copy number plasmid bearing the mutant CDS1* gene in a cds1-null background, we reconstituted a cell with the properties of the cdg1 mutant. We also report that in addition to reduction of in vivo CDP-DAG activity, overproduction of an open reading frame upstream of CDS1 also results in inositol efflux into the growth medium.

EXPERIMENTAL PROCEDURES
Materials-All chemicals were reagent grade or better. Radiochemicals and CTP were obtained from Amersham Corp. Liquiscint TM was purchased from National Diagnostics. Restriction endonucleases were from Promega Corp., New England Biolabs, Stratagene, and Boehringer Mannheim. The Gene Amp polymerase chain reaction (PCR) reagent kit was from Perkin-Elmer. Oligonucleotides were prepared commercially by Genosys. A QIAEX II gel extraction kit and QIAprep spin plasmid miniprep kit were from Qiagen. Yeast growth medium broth and synthetic media for yeast growth and selection were from Bio 101, Inc. Yeast nitrogen base without amino acids and ammonium sulfate was from Difco. L-␣-Phosphatidic acid (derived from egg lecithin), L-serine, and myo-inositol were from Sigma. CDP-DAG was synthesized by reaction of phosphatidic acid and CMP-morpholidate and purified by phase separation and silica gel chromatography (21). The BCA protein assay kit was from Pierce.
Strains, Plasmids, and Growth Conditions-A list of the strains and plasmids used in this work is given in Table I. Plasmid structures of YEp30 and its derivatives are shown in Fig. 1. Except for pSDG1, in which expression of the CDS1 gene is under GAL1 promoter control, expression of the CDS1 gene in YEp30, pSD10, pSD20, and pSD40 is under the control of its endogenous promoter. Plasmids pSD10 and pSD20 contain the HindIII-BamHI fragment including the CDS1 open reading frame and its endogenous promoter region from plasmid YEp30. Plasmid pSD20* contains the same DNA fragment as pSD20 except for a G 305 /A 305 point mutation in the CDS1 open reading frame. Plasmid pSD30 was constructed by inserting NheI-PstI fragment from YEp30 into XbaI-PstI sites of the vector YCpGAL. This region of DNA includes the two open reading frames upstream of the CDS1 gene, YBR0314 and YBR0315 (27). Wild-type yeast and cdg1 mutant cell cultures were maintained in YEPD (1% bacto-yeast extract, 2% bactopeptone, and 2% dextrose). Yeast cell transformants were grown in complete synthetic media (CSM) with indicated nutrient dropouts for selection purposes (i.e. minus tryptophan, minus leucine (ϪLeu) or minus uracil (ϪUra)) supplemented with yeast nitrogen base containing 2 mg/ml (11 M) inositol based on the Difco manual (28), except where indicated. Yeast cell cultures were maintained at 30°C. E. coli strain DH5␣ was used for plasmid amplification and subcloning purposes. It was grown in LB medium (1% bacto-tryptone, 0.5% bacto-yeast extract, and 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.
Chromosomal DNA Preparation and PCR Analysis-Chromosomal DNA was prepared as described previously (29,30). PCR was performed after optimizing conditions as described by Innis and Gelfand (31). Amplification of the CDS1 gene from yeast chromosomal DNA and plasmids used the following primers: primer 1 (5Ј-TTGTCTAGACAC-CCAATCCACCGAG-3Ј) and primer 2 (5Ј-CCGGTCTAGATCAAGAGT-GATTGGTCAATG-3Ј). They were designed according to the DNA sequence of the CDS1 open reading frame and its promoter region (GenBank number Z35898); primer 1 begins 670 bp 5Ј to the CDS1 start codon, and the underlined codon in primer 2 indicates the stop codon for the CDS1 open reading frame.
DNA Sequencing-Plasmid DNA was purified by using the Wizard 373 kit (Promega), and reactions were performed by the Taq Dye-deoxy Terminator (Applied Biosystems) method and run on an Applied Biosystems Sequenator. All plasmids were derivatives of pBluescript II KS (Stratagene) and were sequenced by using the t7 and t3 primers and specific primers derived from the determined sequence. Sequence analysis was carried out with the Pileup program in the Genetics Computer Group (32).
Plasmid Shuffling in Yeast-Plasmids (LEU2) carrying the CDS1 or CDS1* gene under the regulation of its endogenous promoter were introduced into the cds1-null mutant YSD90A by plasmid shuffling (33). Briefly, strain YSD90A/pSDG1 (URA3 on plasmid but requiring leucine) was transformed with a LEU2-containing plasmid and plated on CSM-Leu plates with glucose as carbon source. Colonies formed on CSM-Leu plates were screened for loss of plasmid pSDG1 by lack of growth on CSM-URA plates with galactose as a carbon source.
Preparation of Cell Extracts and Enzyme Assays-Preparation of the membrane fraction was modified from our previous report (15). All cell fractionation procedures were carried out at 4°C. Yeast cell transformants were grown to the exponential phase of growth in the appropriate CSM nutrient dropout medium to select for the plasmid and either 2% glucose or galactose as indicated. Cells were harvested by centrifugation and washed in 20 mM Tris-maleate, pH 7.0, and 1 mM phenylmethylsulfonyl fluoride, centrifuged, and resuspended in the same buffer. The cell suspension was mixed with an equal volume of prechilled silicon beads (diameter, 0.3 mm) and disrupted in a Mini-Beadbeater TM (Biospec Products) by three 30-s bursts at 2,800 rpm with a 5-min pause between bursts. Silicon beads and unbroken cells were removed by centrifugation at 750 ϫ g for 10 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 20 mM Tris-maleate, pH 7.0. CDP-DAG synthase activity was measured by the incorporation of radiolabel into CDP-DAG in 50 mM Tris-maleate buffer, pH 6.5, 20 mM MgCl 2 , 1% Triton X-100, 1.0 mM [5-3 H]CTP (9,000 cpm/nmol), 0.5 mM phosphatidic acid, and 50 g membrane protein in a total volume of 0.1 ml (11). PS synthase activity was measured by the incorporation of radiolabel into PS in 50 mM Tris-HCl buffer, pH 8.0, 0.6 mM MnCl 2 , 0.2 mM CDP-DAG, 0.25% Triton X-100, 0.5 mM L-[3-3 H]serine (10,000 cpm/nmol), and 75 g of membrane protein in a total volume of 0.1 ml (34). PI synthase activity was measured by the incorporation of 0.5 mM myo-[2-3 H]inositol (10,000 cpm/nmol) into PI in 50 mM Tris-HCl buffer, pH 8.0, 2 mM MnCl 2 , 0.2 mM CDP-DAG, 0.15% Triton X-100, and 75 g of membrane protein in a total volume of 0.1 ml (21). All assays were carried out at 30°C for either 5 min (CDP-DAG synthase) or 20 min (PS and PI synthases). Assessment of Inositol Excretion Phenotype-The inositol excretion capacity of yeast strains was tested on CSM-Leu or CSM-Ura plates lacking inositol, choline, and ethanolamine as reported previously (35) using growth of the inositol auxotrophic reporter strain MC13 (ino1) (24). The yeast strain to be tested was patched onto the tester plate lacking inositol, choline, and ethanolamine and permitted to grow for 24 h. The inositol auxotrophic reporter strain was then dilution streaked away from the patch as described previously (14,36), and cross-feeding was scored after an additional 48-h incubation.

Complementation of the cdg1 Mutant by CDS1-Multi-copy
number plasmids bearing the CDS1 gene were transformed into wild-type yeast YPH102, the cds1-null mutant YSD90A, and the cdg1 mutant (Table II). Introduction of genomic clone YEp30 (multicopy plasmid; Fig. 1), which carries two open reading frames 5Ј and two open reading frames 3Ј to the CDS1 gene, resulted in overproduction of CDP-DAG synthase activity in the above mutants to the same level as in the wild-type strain. Surprisingly, this plasmid did not correct the inositol excretion phenotype of the mutants. In addition, growth of YEp30 transformants of the wild-type strain on galactose resulted in low but distinct inositol excretion. Induction of P GAL1 -CDS1 gene expression on pSDG1 in the cds1 null mutant and in the cdg1 mutant by growth on galactose not only brought about overproduction of CDP-DAG synthase activity to levels in excess of that in wild-type cells (strain YHP102 alone) but also corrected their inositol excretion phenotype. The pSDG1 transformant of the cds1-null mutant, when grown in a noninduction medium (in the presence of glucose) had only 10% CDP-DAG synthase activity relative to wild-type yeast. Under repressed conditions for CDS1 expression, it also excreted inositol into growth medium, as previously reported (15). A correlation between inositol excretion and CDP-DAG synthase activity raised the possibility that CDS1 and the mutated gene in the cdg1 mutant are allelic, but suppression of inositol excretion by expression of CDS1 from the GAL1 promoter but apparently not from its own promoter remained unresolved.
To study the discrepancy in the results between transformants with either plasmid pSDG1 or YEp30, plasmid pSD10 was constructed, which excluded the two open reading frames on both sides of the CDS1 gene in the genomic clone (Fig. 1). The cds1 (YSD90A) and cdg1 mutants bearing pSD10 (CDS1 expressed from its own promoter) had similar levels of overexpression of the CDP-DAG synthase activities as the YEp30 transformants of both the mutants and the wild-type cells. However, the pSD10 transformants did not excrete inositol. An autonomously replicating sequence-centromere sequence plas-mid (single copy plasmid), pSD20, which carries CDS1 expressed from its normal promoter, was introduced into YPH102, the cdg1 mutant, and the cds1 null strain. This single copy plasmid contributed a similar amount of additional synthase activity (0.67, 0.53, and 0.50 nmol/min/mg, respectively) to each of these three strains. Therefore, there appears to be no repression of expression in a cdg1 mutant of a single copy of the CDS1 gene expressed from its own promoter and carried in trans, and simple restoration of synthase activity to wild-type levels is sufficient to prevent inositol excretion.
The YBR0314 Gene and Inositol Excretion-To understand the inositol excretion phenotype caused by YEp30, the DNA fragment in the genomic clone was further digested to yield a 3.8-kb NheI-PstI fragment and a 5.3-kb PstI-SalI fragment, which were subcloned into vector YEpGAL to yield plasmids pSD30 and pSD40, respectively (Fig. 1). Growth of the cds1null mutant containing plasmid pSD40 on glucose resulted in overproduction of CDP-DAG synthase activity but not excretion of inositol (Table II). Plasmid pSD30 contains the two upstream open reading frames YBR0314 and YBR0315 but lacks the CDS1 gene. Gene YBR0315 encodes the ribosomal protein L2B (37). The gene product of YBR0314 does not show significant homology with the known sequences in GenBank by BLAST search (38). Introduction of plasmid pSD30 into strain YSD90A/pSDG1 reduced CDP-DAG synthase activity by 42%, but synthase activity was still 6-fold greater than wild-type yeast levels (Table II), yet the cells excreted inositol. The reduction of CDS1 overexpression was possibly due to a reduced FIG. 1. Plasmid structures. Genomic clone YEp30 containing five open reading frames in vector YEp13 were digested with different restriction enzymes to construct the plasmids used in this study, as described under "Experimental Procedures." CDS1 expression in pSDG1 is under the control of a GAL1 promoter. Plasmids pSD10 and pSD20 contain the CDS1 open reading frame and its endogenous promoter, illustrated by P CDS1 in the multicopy vector YEpGAL (2) and single copy vector YCpGAL (autonomously replicating sequence 1 centromere sequence 4 (ARS1 CEN4)), respectively. Plasmid pSD40 contains part of the P CDS1 region. Plasmid pSD30 contains the two open reading frames upstream of CDS1. Only YBR0314 is present in pSD50. The vector in plasmids pSD30, pSD40, and pSD50 is YEpGAL. copy number for plasmid pSDG1 resulting from the introduction of the 2 plasmid pSD30, since yeast cells regulate the total number of 2 plasmids (39). The 3.8-kb DNA fragment was further digested to yield a 2-kb PvuII fragment containing the open reading frame YBR0314 and 83 bp of its 5Ј-region, which was subcloned into YEpGAL (Fig. 1). This plasmid, pSD50, when transformed into YSD90A/pSDG1, caused inositol excretion in the presence of galactose but qualitatively less than that from the pSD30 transformant. It is possible that the 83-bp endogenous 5Ј-region in this construct was not long enough for efficient YBR0314 expression. Thus, the inositol excretion phenotype of YEp30 transformants appears to be due to multiple copies of YBR0314 carried in trans.
G 305 /A 305 Point Mutation in the cdg1 Mutant--Both a CDP-DAG synthase functional assay and correction of the inositol excretion phenotype indicated that the cdg1 phenotype was caused by a decreased level of CDP-DAG synthase activity inside the cell. However, the functional assay data clearly indicated that expression of CDS1 in the cdg1 mutant was not repressed (Table II), thereby excluding the possibility that a mutant regulatory gene caused reduced CDP-DAG synthase activity in the cdg1 mutant. Since CDS1 encodes the majority, if not all, of CDP-DAG synthase activity (15), and a defect in a single gene is responsible for the mutant phenotype (14), a reasonable explanation for the cdg1 phenotype would be a mutation either within the CDS1 gene or in its upstream regulatory region. Three independent PCR reactions were carried out to synthesize a 2-kb DNA fragment including the CDS1 gene and its 5Ј-promoter region by using genomic DNA from the cdg1 mutant as a template, as described under "Experimental Procedures." PCR products were then subcloned into the pBluescript vector, and the constructs were amplified in E. coli. Plasmids from two individual colonies of each of the PCR products were sequenced and analyzed. A G/A point mutation located 305 bp downstream of the initiation codon was found in all six plasmids sequenced (Fig. 2). As controls, PCR reactions using genomic DNA of YPH102 wild-type cells as a template were also carried out, and their products were sequenced. None of the products showed a G 305 /A 305 mutation. Thus, a G 305 /A 305 mutation exists in the CDS1 gene of the cdg1 mutant. This point mutation causes a Cys 102 3 Tyr substitution at the interphase between a hydrophobic and hydrophilic region (40) of the CDP-DAG synthase.
Activities and Regulation of Phospholipid Biosynthetic Enzymes-A 626-bp PstI-Nru I fragment in pSD20 was replaced with a DNA fragment from the same region in the above PCR product derived from cdg1 cells to create plasmid pSD20* containing the mutant CDS1* gene. Except for a G 305 /A 305 mutation, the rest of the DNA in pSD20* was identical with that of pSD20 as checked by DNA sequencing. Both pSD20 and pSD20* were introduced into the cds1-null mutant YSD90A by plasmid shuffling as described under "Experimental Procedures" and into the cdg1 mutant. As controls, YPH102 and cdg1 were transformed with the YCpGAL vector. The above transformants were grown in complete synthetic medium with or without inositol and choline where indicated. Membrane extracts were prepared from the above cells, and the activities of CDP-DAG synthase, PS synthase, and PI synthase were measured.
As expected, the specific activity of CDP-DAG synthase in the cdg1 mutant was 15% of the level found in YPH102 when inositol and choline were not supplied in growth medium, and inositol and choline repressed the activity in wild-type cells but had no effect on the already low synthase activity in the mutant (Fig. 3). A single copy of CDS1 in trans to the null chromosomal cds1 gene (YSD90A/pSD20) resulted in the same pattern of CDP-DAG synthase activity regulation as in YPH102, whereas placing the CDS1* gene in trans (YSD90A/pSD20*) mimicked the cdg1 mutant. When CDS1 was in trans to the chromosomal mutation in the cdg1 gene (cdg1/pSD20), the CDP-DAG synthase activities were higher than that for YPH102 under both derepressed and repressed growth conditions, consistent with a higher level of activity expressed from the combination of the wild-type CDS1 gene and the CDS1 gene in the cdg1 mutant,  but there was still significant repression by inositol and choline. However, when CDS1* was trans to the mutation in the cdg1 strain (cdg1/pSD20*), the level of synthase activity was double that of the cdg1 mutant strain with no plasmid copy of CDS*, but there was no repression by inositol and choline, consistent with duplication of a mutated CDS1 gene. The total activity in the strain was still only 30% of wild-type levels, which appears to be too low to correct the phenotype brought about by reduced CDP-DAG synthase activity. PS synthase activity was also regulated by inositol and choline both in wild-type yeast and in the cdg1 mutant (Fig. 4). Overall, PS synthase activities (under both derepressed and repressed conditions) were higher than wild-type levels in all strains lacking a wild-type copy of the CDS1, including the cdg1 strain with CDS1*. Strains expressing at least one copy of the wild-type CDS1 gene resembled strain YPH102 in the levels of PS synthase activity. PI synthase-specific activity was unaffected by additions of inositol and choline to the growth medium in all of these strains, including YPH102 (data not shown), which is consistent with the mutation in the cdg1 strain having no effect on PI synthase activity (14).
Inositol Excretion Phenotype-Functional assays showed that CDP-DAG synthase and PS synthase activities of YSD90A/pSD20 resembled wild-type yeast, whereas those of YSD90A/pSD20* resembled the cdg1 mutant. The inositol excretion phenotype also follows the level of CDP-DAG synthase activity of the various strains. Strains expressing the wild-type CDS1 gene, resulting in normal or elevated levels of CDP-DAG synthase activity (Fig. 5, B, D, and F), do not cross-feed the inositol auxotrophic tester strain. Those strains carrying only cdg1 or CDS1* or both cdg1 and CDS1* (Fig. 5, A, C, and E) cross-feed the tester strain. The common determinant in all of the strains excreting inositol is reduced CDP-DAG synthase activity. DISCUSSION Pleiotropic deficiencies of the cdg1 mutant raised interesting questions as to the role in phospholipid biosynthesis and regulation of the mutated gene in this strain. Several lines of evidence suggested that the mutated locus might not be allelic with CDS1, and that the mutation was either in a regulatory element (3,5), or in a second copy of the CDS gene (14). Identification of the molecular basis for the cdg1 phenotype should extend the understanding of the factors regulating phospholipid metabolism in yeast. Cloning of the CDS1 gene made it possible to determine whether a mutated locus in the cdg1 mutant is allelic with CDS1, is another structure gene, or is a novel regulator of phospholipid metabolism. The following evidence supports the first possibility. Plasmids bearing a copy of the CDS1 gene could be fully expressed in the cdg1 mutant to the same extent as in wild-type cells. The increased CDP-DAG synthase activity in the cdg1 mutant background also corrected the inositol excretion phenotype of the mutant. The CDS1 locus in the cdg1 mutant was found to contain a G 305 / A 305 single base substitution. Introduction of this mutation into an otherwise wild-type CDS1 gene expressed from its own promoter from a single copy plasmid in a cds1-null background resulted in CDP-DAG synthase, PS synthase, and PI synthase FIG. 5. Mutant CDS1, but not wild-type CDS1, causes inositol excretion into the growth medium. The inositol excretion phenotype was tested on inositol-free medium as indicated below and described under "Experimental Procedures." An inositol-requiring strain, MC13, was dilution streaked (from left to right and from top to bottom) away from the patch of the strain to be tested. A, YSD90A/pSDG1 on CSM-URA, 2% glucose; B, YSD90A/pSDG1 on CSM-URA, 2% galactose; C, YSD90A/pSD20*; D, YSD90A/pSD20; E, cdg1/pSD20*; F, cdg1/ pSD20. Medium in C-F was CSM-Leu, 2% glucose. activities with properties similar to those in the cdg1 mutant. As in the cdg1 mutant, expression of CDP-DAG synthase activity from the mutant CDS1* gene on the plasmid was no longer regulated by precursors to phospholipid biosynthesis. This transformant also excreted inositol into growth medium, a pleiotropic deficiency typical of the cdg1 mutant. The common determinant of the cdg1 phenotype is a 70% or greater reduction in the level of CDP-DAG synthase activity due to either a mutation in the CDS1 gene or reduced gene expression. The defect in CDP-DAG synthase activity in the cdg1 mutant was previously correlated with a reduced level of synthase-specific antigen rather than with an alteration in the enzymological properties of the enzyme (14). This reduction in the gene product could be due to either reduced transcription and translation of the mutant gene or reduced stability of the mutant protein product.
The fact that a mutation within the CDS1 gene resulting in reduced CDP-DAG synthase activity causes pleiotropic changes in phospholipid biosynthesis brings about interesting issues as to the role of CDP-diacylglycerol and its synthase in the regulation of phospholipid biosynthesis in yeast. From the combined results on the cdg1 phenotype, a decrease in CDP-DAG synthase activity is sufficient to dramatically affect the regulation of the phospholipid biosynthetic system, including elevated levels of PS synthase activity, constitutive expression of inositol 1-phosphate synthase, lack of response to inositol plus choline in the growth medium, and excretion of inositol (14). In vitro studies have shown that PS synthase activity is inversely related to the PI/PS ratio in the membrane. An increase of the PI/PS ratio from 2:1 to 8:1 reduced PS synthase activity to 50% of its original level (41). When CDS1 gene expression is regulated through a GAL1 promoter, there is a proportional relationship between the PI/PS ratio and the level of CDP-DAG synthase activity (15). A 40% reduction in PI content was observed as a result of a 75% decrease of CDP-DAG synthase activity in the cdg1 mutant, whereas PS levels were barely changed compared with the wild-type yeast (14). Thus, a decrease of the PI/PS ratio as a secondary effect of a reduction of CDP-DAG synthase activity may partly contribute to the elevated PS synthase activity in cells bearing a mutant CDS1* gene either in the chromosome (cdg1 mutant) or on the covering plasmid (pSD20* in the cds1-null mutant). However, PS synthase may also be regulated by posttranslational modification, since phosphorylation of the 23-kDa PS synthase subunit affects its activity (42,43). Alternatively, expression of the gene encoding PS synthase may be derepressed at low cellular CDP-DAG synthase levels, as is INO1 (inositol 1-phosphate synthase) expression in some mutants of phospholipid metabolism. Constitutive expression of INO1 causes inositol excretion, the opi phenotype (44), which has been found in a series of mutants involved in synthesis of the aminophospholipids and inositol uptake (36,(45)(46)(47)(48). Low cellular CDP-DAG synthase in cdg1 and CDS1* mutants may affect the biosynthesis of the aminophospholipids due to the limited amount of CDP-DAG available for de novo synthesis and consequently may result in derepression of INO1 gene expression. Preliminary results indicate that INO1 expression is enhanced at low CDP-DAG synthase levels in the cds1-null mutant, which can be reversed by CDS1 overexpression from a covering plasmid. 2 Therefore, the molecular basis for the lack of down-regulation of CDP-DAG synthase and PS synthase activities by the presence of inositol and choline in the growth medium of the cdg1 and CDS1* mutants appears to be a result of interruption of aminophospholipid biosynthesis and/or constitutive INO1 expression (1,46). The initial observation that genomic clone YEp30 resulted in overexpression of CDP-DAG synthase in all strains but did not correct the inositol excretion phenotype in the cdg1 mutant and caused excretion in wild-type cells grown on galactose lead to the identification of an open reading frame 5Ј to the CDS1 locus YBR0314, which results in inositol excretion when present in multiple copies. No information is available as to the possible functions of this gene locus. The inositol excretion phenotype has been associated with mutations in inositol and phospholipid metabolism but never with multiple copies of normal yeast genes (36, 44 -48). Further examination of this gene and its product may provide additional information on the complex interaction between inositol and phospholipid metabolism in yeast.