The Isolation and Characterization of a Mutant Strain of Saccharomyces cereuisiae That Requires a Long Chain Base for Growth and for Synthesis of

A mutant of Saccharomyces cerevisiae has been ob- tained that shows an absolute growth requirement for long chain bases found in sphingolipids. In the absence of a long chain base, the cells are unable to synthesize the phosphoinositol-containing sphingolipids charac-teristic of yeast. These results suggest that one or more of the yeast sphingolipids plays a vital biological role. membrane in higher plants and fungi are rare but not unknown in portion of all these lipids an long chain base attached wide polar head exact class organism. of to

A mutant of Saccharomyces cerevisiae has been obtained that shows an absolute growth requirement for long chain bases found in sphingolipids. In the absence of a long chain base, the cells are unable to synthesize the phosphoinositol-containing sphingolipids characteristic of yeast. These results suggest that one or more of the yeast sphingolipids plays a vital biological role.
Sphingolipids are widespread membrane constituents in animals ( l ) , higher plants (2), and fungi (3); they are rare but not unknown in bacteria (4). The apolar portion of all these lipids is an N-fatty acyl long chain base (ceramide) to which is attached a wide variety of polar head groups, the exact nature of which depends on the class of the organism. Although their ubiquity implies a vital biological role, no specific molecular functions have been ascribed to any sphingolipid; in animals, interest actively centers on cell-cell interaction and recognition phenomena in growth and differentiation (5). Saccharomyces cereuisiae contain a unique set of sphingolipids related to those of other fungi and plants (6) but not reported to occur in animals. Schematically, these are IPC,' MIPC, and M(IP),C. We report herein the discovery of a mutant of S. cereuisiae that has an absolute requirement for a long chain base for sphingolipid synthesis and for growth; this suggests a vital role for one or more of the yeast sphingolipids.

Strains and Culture
Conditions-An inositol auxotroph of S. cereuisiae MC6A was obtained from Dr. Susan Henry (Albert Einstein College of Medicine) and is referred to as wild type. Cells were cultured at 30 "C with shaking, and turbidity was monitored by absorbance measurement at 650nm (1 cm) with a Zeiss PMQ-2 spectrophotometer. Basal medium consisted of Difco Vitamin-Free Yeast Nitrogen Base (16.7 g/l), glycylglycine (pH 3.2) (0.05 M), *This research was supported by National Institutes of Health Grant AI12299. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Mutant Production-Wild type cells were mutagenized and recovered according to Henry et al. (7). The cells were plated on basal medium plus 200 p~ DL-erythrodihydrosphingosine containing 1.5% agar. The colonies were scored for the desired mutation after replica plating onto plates containing basal medium with and without 200 p~ DL-erythrodihydrosphingosine.
Lipid Analysis-Lipids were extracted from cells killed with 5% trichloroacetic acid as previously described (8). The extract was treated to remove nonlipid impurities (9). The purified extract was analyzed for individual acyl ester phospholipids by two-dimensional paper chromatography (10); the 32P-lipid spots were visualized by autoradiography, cut out, and measured with a scintillation counter. The purified lipid extract was also subjected to mild alkaline methanolysis (9) which yields a measurement of the ratio of acyl ester Plipids to the alkali-stable phospholipids as well as yielding a preparation of the alkali-stable lipids suitable for liquid chromatography. Liquid chromatography of the alkali-stable phospholipids was carried out with a column (0.45 X 30 cm) packed with 5-pm silica gel (Lichrosorb Si 60, E. Merck) and maintained at 56 "C. The equipment (Waters Associates) consisted of a Model U6K injector and two Model 6000A pumps controlled by a Model 660 programmer. A flow rate of 2 ml/min was used throughout. The alkali-stable lipids were dissolved in solvent B, and 75-~1 aliquots were applied to the column which had been equilibrated with solvent A (CHCl,, 95% C2H,0H, concentrated NH3 (50:44:6)). The elution schedule was 2 min of solvent A, followed by a 20-min nonlinear gradient (program 4) formed by mixing solvents A and B (CHCl,, 95% C,H,OH, concentrated NH3, H20 (3052:6:12)) to a final composition of 35% solvent B. During this gradient the solvent composition at t minutes is given by: % solvent B = 35 (t/2O)'I3. After continuing 35% solvent B for 2 min, a linear gradient from 35 to 70% solvent B is pumped for 10 min followed by 70% B for 21 min. The column is recycled with a 5-min linear gradient from 70 to 100% solvent B, 5 min of 100% solvent B, a 5-min linear gradient to 100% solvent A, and finally 15 min of 100% solvent A. Both solvents contain 0.6 g/of NH,Cl. Samples of 1.0 ml were collected in shell vials, and the radioactivity was evaluated by Cerenkov counting in a scintillation counter.

RESULTS
Isolation of the Mutant-The possibility that an auxotrophic mutant could be obtained stemmed from our observation that wild type S. cereuisiae cells growing on the agar solidified culture medium described under "Materials and Methods" could take up labeled dihydrosphingosine and incorporate it with high efficiency into all the yeast sphingolipids. Selection of a long chain base auxotroph by the inositolless death procedure (7) was unsuccessful. Therefore, we screened over io5 mutagenized colonies for long chain base auxotrophs and found several with growth comparable to wild type in the presence of long chain base. One of these, Lcb-BOA, is the subject of this communication.  bases was examined (Fig. 2). Phytosphingosine, the natural product, was significantly more effective than either the erythro-or threo-isomers of DL-dihydrosphingosine. If only one of the DL-pair is effective, then the threo-and erythrodihydrosphingosines would be about as effective as phytosphingosine. D-Sphingosine supported only marginal growth.
Synthesis of Phospholipids by Lcb-ZOA-We wished to verify that long chain base auxotrophy was associated with an inability to make the complex phosphosphingolipids of S.
cereuisiae rather than a correction of some other unrelated defect. Phospholipid synthesis was studied by measuring the incorporation of 32Pi into the lipids of the mutant and parental strain cultured in the presence and absence of dihydrosphingosine. The data in Table I indicate that in the presence of DL-erythrodihydrosphingosine, incorporation of 32Pi into both total acyl ester phospholipid and alkali-stable phospholipids proceeds at the same rates in parental and mutant strains throughout the course of exponential growth. Acyl ester phospholipid synthesis with the mutant is scarcely diminished during the initial 5.0 h of long chain base starvation but declines sharply thereafter. In contrast, total alkali-stable phospholipid synthesis is sharply depressed between 0.5 and 2.5 h post-starvation and declines to less than 10% of the long chain base fed control between 5.5 and 7.5 h post-starvation (Table I) even though the turbidity is still 90% of the control after 7.5 h (Fig. 1). Curiously, the absence of long chain base appears to significantly stimulate phospholipid synthesis in the parental strain.
Incorporation of 32Pi into individual phospholipids was also measured during culture of the mutant in the presence and absence of DL-erythrodihydrosphingosine. The acyl ester phospholipids were estimated after two-dimensional chromatography on silica gel-impregnated paper. From 0.5 to 2.5 h, no significant effect of long chain base starvation could be detected in the synthesis of the six major acyl ester phospholipids (Table 11). After 5 h of starvation, significant depression in the rate of synthesis of all acyl ester phospholipids was observed ranging from a 73% decrease for phosphatidylcho-TABLE I A comparison of 32P, incorporation into phospholipid classes of wild type and mutant cells cultured in the presence and absence of dihydrosphingosine Cells were grown overnight in basal medium plus 100 p~ DLerythrodihydrosphingosine reaching an absorbance of 7.77 (MCGA) and 5.4 (Lcb-BOA). The cells were washed three times with basal medium, and each strain was suspended in two 50-ml cultures: basal medium and basal plus 50 pM DL-erythrodihydrosphingosine. At 0.5, 3, and 5.5 h, 10-ml aliquots were removed from each culture, added to a flask containing 0.6 mCi of 32Pi (carrier-free, New England Nuclear) and incubated for 2 h at 30 "C with shaking. The incorporation was terminated with trichloroacetic acid, and the lipid was extracted and processed as indicated under "Materials and Methods." From knowledge of the specific activity of the 32Pi in the culture medium and from the absorbance of the culture a t the end of the pulse period, the rate of incorporation was calculated. and absence (X-X) of DLerythrodihydrosphingosine and pulsed for 2 h (0.5-2.5 h post-starvation with 32P1 after transfer to fresh medium exactly as described in the legend to Table   I. the alkali-stable phospholipids were obtained and separated on a silica gel column as described under "Materials and Methods." EFFLUENT VOLUME (ML)

TABLE I1
The synthesis of phospholipids in mutant cells cultured in the presence and absence of DL-erythrodihydrosphingosine Lcb-BOA cells were cultured in the presence and absence of DLerythrodihydrosphingosine and pulsed with 32P, as described in the legend to Table I. The lipid extract freed of nonlipid impurities was chromatographed on silica gel-impregnated paper as described under "Materials and Methods" to measure the radioactivity in each acyl ester phospholipid. The alkali-stable phospholipids were obtained and separated by liquid chromatography as described under "Materials and Methods." The rate of incorporation was calculated from the specific activity of 32P, in the culture medium and the absorbance of the culture at the end of the pulse period. incorporation of 32Pi into MIPC species is less than 5% of the long chain base fed control ( Fig. 3 and Table 11), and strong inhibition of incorporation is observed for the IPCs (80%) and the M(IP)*Cs (67%). During the 5.5-7.5-h post-starvation interval, incorporation into both IPCs and MIPCs is less than 2% of fed control and incorporation into M(IP)&s about 10% of the fed control.
Thus, with Lcb-BOA, an early, severe, and differential inhibition of phosphosphingolipid synthesis is observed as a consequence of starvation for long chain base. The inhibition of sphingolipid synthesis precedes the restriction of growth.

DISCUSSION
Based on in vivo and in uitro experiments (9, 11, la), the pathway of sphingolipid synthesis in yeast was postulated to be a ceramide -+ IPC -+ MIPC -+ M(IP),C, with phosphatidylinositol and GDP mannose as the phosphoinositol and mannose donors, respectively. With this pathway in mind, one can expect that starvation for long chain base would first affect the incorporation of 32Pi into IPC and MIPC as ceramide is depleted and that incorporation into M(IP),C would be affected last because [32P]phosphatidylinositol could react with pre-existing unlabeled MIPC to give labeled M(IP)&. This conforms with the data observed in Fig. 3 and Table 11.
Since the sphingolipid products formed by culture of the mutant with dihydrosphingosine all appear to contain phytosphingosine (Fig. 3), the mutant must actively convert dihydrosphingosine to phytosphingosine (as does the parental strain).' It is logical to speculate that the mutant is blocked in the formation of long chain base with the synthetic defect either in the formation of ketodihydrosphingosine or in its reduction to dihydrosphingosine. The observation that the unnatural threo-isomer of dihydrosphingosine supports growth (Fig. 2) might be interpreted to mean that the cells can somehow epimerize the threo-and the erythro-isomers; if this is the case, then this may rule out reduction of ketodihydrosphingosine as the affected step. Direct enzyme assays should resolve these questions. The ineffectiveness of D-sphingosine, the animal long chain base, in supporting growth (Fig. 2) possibly suggests a tight specificity in its transport or in its conversion to complex sphingolipids; alternatively, yeast sphingolipids constructed with sphingosine may function poorly.
To our knowledge, this is the first mutant of this kind studied, and this study strongly suggests that sphingolipids are vital substances. We hope that a study of the molecular pathology of long chain base starvation in such mutants may shed light on the specific biological function(s) of the sphingolipids.