(Received for publication, July 1, 1996, and in revised form, December 30, 1996)
From the Department of Biochemistry and the Lucille P. Markey Cancer Center, University of Kentucky Medical Center, Lexington, Kentucky 40536-0084
We have identified a Saccharomyces cerevisiae gene necessary for the step in sphingolipid synthesis in which inositol phosphate is added to ceramide to form inositol-P-ceramide, a reaction catalyzed by phosphatidylinositol:ceramide phosphoinositol transferase (IPC synthase). This step should be an effective target for antifungal drugs. A key element in our experiments was the development of a procedure for isolating mutants defective in steps in sphingolipid synthesis downstream from the first step including a mutant defective in IPC synthase. An IPC synthase defect is supported by data showing a failure of the mutant strain to incorporate radioactive inositol or N-acetylsphinganine into sphingolipids and, by using an improved assay, a demonstration that the mutant strain lacks enzyme activity. Furthermore, the mutant accumulates ceramide when fed exogenous phytosphingosine as expected for a strain lacking IPC synthase activity. Ceramide accumulation is accompanied by cell death, suggesting the presence of a ceramide-activated death response in yeast. A gene, AUR1 (YKL004w), that complements the IPC synthase defect and restores enzyme activity and sphingolipid synthesis was isolated. Mutations in AUR1 had been shown previously to give resistance to the antifungal drug aureobasidin A, leading us to predict that the drug should inhibit IPC synthase activity. Our data show that the drug is a potent inhibitor of IPC synthase with an IC50 of about 0.2 nM. Fungal pathogens are an increasing threat to human health. Now that IPC synthase has been shown to be the target for aureobasidin A, it should be possible to develop high throughput screens to identify new inhibitors of IPC synthase to combat fungal diseases.
Fungal pathogens present an increasing threat to human health, particularly for immunocompromised individuals. Current drugs are not efficacious, cause serious side effects, and are becoming less useful because of increased resistance to them. Our research has focused on enzymes catalyzing the synthesis of phosphoinositol-containing sphingolipids that are present in fungi but absent in humans and, therefore, likely to offer unique targets for antifungal drugs. One enzyme meeting this criterion is inositol phosphorylceramide synthase (phosphatidylinositol:ceramide phosphoinositol transferase, IPC synthase)1 which catalyzes the transfer of phosphoinositol from phosphatidylinositol to ceramide to give IPC (1).
Since sphingolipid synthesis is vital for growth and viability of the yeast Saccharomyces cerevisiae it is likely that blockage of synthesis by a drug would efficiently inhibit growth and induce cell death. Besides S. cerevisiae, other fungi including the human pathogens Histoplasma capsulatum (1) and Candida albicans (2) contain IPC. Other pathogenic fungi are likely to contain it as well since it has been found in all fungi studied to date (1).
The sphingolipid biosynthetic pathway of S. cerevisiae is
diagrammed in Fig. 1. Work presented in this paper
focuses on the modification of the 1-hydroxyl of phytoceramide by the
addition of myo-inositol phosphate to form IPC which is then
mannosylated to yield mannose inositol-P-ceramide (MIPC). The final
step in S. cerevisiae sphingolipid synthesis is the addition
of inositol-P to MIPC to yield the major sphingolipid
mannose-(inositol-P)2-ceramide (M(IP)2C
(3-5)). The later steps in sphingolipid synthesis in S. cerevisiae (Fig. 1) are tentative because none of the enzymes have
been purified, the reaction requirements are poorly defined, and the
stoichiometry has not been determined (reviewed in Ref. 1).
We previously searched for mutants defective in sphingolipid synthesis
by screening for cells requiring exogenous sphingoid long chain base
for growth (Lcb phenotype). Approximately 50 Lcb
mutants fell into only two complementation groups,
designated LCB1 and LCB2 (6, 7), both of which
are necessary for serine palmitoyltransferase activity and are thought
to encode subunits of the enzyme (8). No mutants were found for the
next step, 3-ketosphinganine reductase, possibly because of
accumulation of lethal concentrations of 3-ketosphinganine. Mutants
defective in the later steps in sphingolipid synthesis, ceramide
synthesis, and beyond, would not have been recovered because neither
exogenous ceramide nor sphingolipids containing phosphoinositol were
able to support growth of Lcb
mutants (7). In addition,
accumulation of ceramide should have inhibited growth since exogenous
N-acetylceramide does inhibit growth of S. cerevisiae cells (9, 10).
In the experimental approach described here for isolating strains defective in the later steps in sphingolipid biosynthesis, the potential problems of growth inhibition and killing by the build up of a sphingolipid intermediate were circumvented by using a sphingolipid compensatory (SLC) strain. Unlike normal yeast strains that require sphingolipids for growth and viability, SLC strains can grow without making sphingolipids (11). SLC strain 7R6 carries two mutations, a deletion of the lcb1 gene and a point mutation that creates the suppressor gene SLC1-1 (12). The suppressor gene enables cells, in the absence of exogenous long chain base, to make novel glycerophospholipids that are thought to compensate for one or more functions of sphingolipids essential for vegetative growth (13). The lcb1 mutation blocks the first committed step in sphingolipid synthesis, so no toxic sphingolipid intermediate should accumulate when cells are grown on medium lacking a long chain base. Thus if a gene encoding an enzyme later in the sphingolipid biosynthetic pathway is mutated, the strain should still be able to grow without making sphingolipids because of the SLC1-1 suppressor gene, and should not accumulate a toxic intermediate because of the lcb1 mutation.
As we show here, our mutant isolation procedure worked as expected and allowed the isolation of a mutant S. cerevisiae strain defective in IPC synthase activity. In addition, a diploid version of the mutant strain made it possible to isolate a gene, AUR1, which complemented the IPC synthase defect and restored IPC synthase activity. Recent studies of drug resistance in S. cerevisiae had shown that mutations in AUR1 (14, 15) confer resistance to the antifungal drug aureobasidin A (AbA), a cyclic depsipeptide produced by Aureobasidium pullulans R106. Because the function of the Aur1 protein was not known it was not clear why AbA was fungicidal. We show here that AbA is a potent inhibitor of IPC synthase and offer an explanation for its fungicidal activity.
S.
cerevisiae strains are: 7R6 (MATa ura3-52 leu2-3, 112 lcb1::URA3 SLC1-1 ade1, Ref. 12) which was derived from
wild type strain SJ21R (MATa ura3-52 leu2-3, 112 ade1);
YPH2 (MAT ura3-52 lys2-801amber
ade2-101ochre, Ref. 16); AG27-61 (MATa ura3-52
leu2-3, 112 lcb1::URA3 SLC1-1 lys2-801amber ade1
ipc1-1) was derived from 7R6 by the procedure described below. A
diploid version of strain AG27-61 (termed AGD27-61) was made by
transforming haploid cells with a plasmid (pHO-12, Ref. 17) carrying
the HO endonuclease gene, responsible for switching of the
mating type (18). Leu+ transformants were streaked onto
PYED plates and colonies containing MATa/MAT
diploid cells were identified by their large, ellipsoidal morphology.
Diploids were screened on defined medium lacking leucine for
Leu
cells, indicating loss of the plasmid carrying the
HO gene, and tested for transformation efficiency using
pRS315 (LEU2 CEN4, Ref. 16).
Construction of pLCB1-5 began by inserting the S. cerevisiae ADE2 gene as a PstI-SpeI DNA fragment into pRS315 cut with the same restriction endonucleases. The resulting plasmid was cut with NaeI and SacI and ligated to the LCB1 gene obtained from pTZ18-LCB1 (19) as an NruI-SacI DNA fragment. pIPC1 was isolated from the recombinant DNA library described below in which the 4307-base pair insert corresponds to S. cerevisiae Chromosome XI between coordinates 432,813 and 437,119 as described in the Saccharomyces cerevisiae genome data base at Stanford University (http://genome-www.stanford.edu).
A recombinant DNA library containing about 160,000 plasmids, 95% of which carried an insert, was constructed. Genomic DNA from derivatives of strain 4R3 resistant to pH 4.12 was isolated, pooled, and 10 µg was partially digested with Sau3AI. DNA fragments of 5-10 kilobases were isolated from an agarose gel and ligated with 1 µg of BamHI-digested, alkaline phosphatase-treated pRS315. Ligated DNA was purified using GeneClean (Bio-101, La Jolla, CA), electroporated into Escherichia coli XL1-Blue cells (Stratagene, La Jolla, CA), and plasmid DNA was prepared from ampicillin-resistant colonies selected on Petri plates.
Yeast were grown on modified PYED (buffered to pH 5.0, PYED-5.0) which contained 1% yeast extract (Difco), 2% Bacto-Peptone (Difco), 2 or 4% glucose, 50 mM sodium succinate (pH 5.0), inositol (50 mg/liter), and potassium phosphate monobasic (0.5 g/liter) (19), or on defined medium supplemented as described (19) and containing, when necessary, 25 µM phytosphingosine (PHS). PYED buffered to pH 4.1 (PYED-4.1) was made by mixing 290 ml of autoclaved agar (2% for plates only), 300 ml of autoclaved yeast extract (1%) plus peptone (2%), 100 ml of filter sterilized glycylglycine (0.5 M, pH 3.1), 200 ml of glucose (20%), 10 ml of inositol (0.5%), and 100 ml of potassium phosphate monobasic (0.5%). In a previous publication (20) we erroneously referred to this medium as having a pH of 3, but the actual pH is 4.1. For some experiments indicated under "Results," the agar in Petri plates was replaced with agarose (Fisher, BP160-500).
Isolation of Strain AG27Strain 7R6 was grown overnight in PYED plus 25 µM PHS (referred to here as the medium) and then mutagenized with ethylmethanesulfonate to give 20% killing (7). Mutagenized cells were diluted to an absorbance at 600 nm (A600) of 0.4 with medium, incubated with shaking at 30 °C for 7 h, during which time the A600 increased to 2.5, centrifuged, resuspended in 2 ml of medium, and sonicated using a microtip (Heat Systems-Ultrasonic) for 2 min to disrupt clumped cells. One ml of cells was layered on 4 ml of 30% sodium diatriazoate (7) and centrifuged at 10 °C in a Sorvall RT6000B centrifuge for 4 min at 2000 rpm. Most cells were at the interface but a faint pellet of dense cells was present at the bottom of the tube. The liquid was aspirated and the cell pellet was carefully resuspended in 0.5 ml of medium to avoid mixing with cells stuck to the side of the tube. Resuspended pellets from two tubes were mixed and re-centrifuged on 30% sodium diatriazoate. The cell pellet was resuspended in medium and about 500 cells were spread on PYED plates containing 25 µM PHS. Two days later only about 15 colonies per plate were visible, suggesting only 3% of the dense cells were viable.
Inositol Labeling of Sphingolipids in VivoSphingolipid synthesis was measured by growing cells to saturation at 30 °C in PYED, diluting to an A600 of 0.1 with fresh PYED medium containing 10 µg/ml inositol and 10 µCi/ml myo-[2-3H]inositol (20 Ci/mmol, DuPont NEN), and grown overnight to saturation at 30 °C. Cells were washed with 25 ml of 0.1 M sodium phosphate (pH 5.5) by centrifugation and resuspended in the same buffer at 50 A600 units/ml. A reaction containing 0.4 ml of 0.5% Tergitol, 0.4 ml of 0.5 M sodium succinate (pH 5.5), 0.1 ml of 40% glucose, and 10 µl of 2.5 mM PHS in 95% ethanol was initiated by addition of 0.2 ml of the radiolabeled cells. At time 0, a 0.2-ml sample and at 1.5 h a 0.5-ml sample of radiolabeled cells were treated with trichloroacetic acid at a final concentration of 5% for 20 min at 0 °C. Cells were centrifuged, washed 3 times with 1 ml of 5% trichloroacetic acid, and finally with water. Cell pellets were extracted with 1 ml of 95% ethanol:water:diethyl ether:pyridine:concentrated NH4OH (15:15:5:1:0.018, v/v) for 60 min at 60 °C (21). After centrifuging while warm, a 0.5-ml sample of the supernatant fluid was evaporated under a stream of N2, dissolved in 1 ml of monomethylamine reagent, and heated for 30 min at 52 °C to deacylate the acylester phospholipids (22). The sample was dried and dissolved in 0.5 ml of chloroform:methanol:water (16:16:5). Qualitative analysis of 25-µl samples was carried out by silica gel thin layer chromatography on 20-cm Whatman LK5 plates developed with chloroform, methanol, 4.2 N NH4OH (9:7:2). Radioactivity was measured by using a BioScan apparatus.
[3H]N-Acetylsphinganine Labeling of Sphingolipids in VivoCellular conversion of [3H]N-acetylsphinganine to IPC was measured as follows. PYED medium was added to tubes containing solid [3H]N-acetylsphinganine (860 cpm/pmol) and the mixture was dissolved by treatment for 5 min in an ultrasonic water bath. Log phase cells, grown in PYED medium, were centrifuged, suspended in fresh medium and added to the radiolabeled medium to give a final cell density of 1.5 A600 units/ml. The final concentration of N-acetylsphinganine was 5.0 µM. At the indicated times, 1-ml samples were removed, quenched with trichloroacetic acid, and processed as in the inositol-labeling procedure.
Qualitative analysis of 25-µl samples was carried out on 20-cm Whatman HP-K plates developed with chloroform, methanol, 4.2 N NH4OH (9:7:2). Each lane contained 2 nmol of IPC-3 (IPC with phytosphingosine and a monohydroxylated fatty acid, Ref. 1) internal standard. Radioactivity was measured by using a BioScan apparatus and the plates were sprayed with a 10% solution of CuSO4·5H20 in 8% H3PO4 and charred at 160 °C to locate the IPC-3 standard.
Quantification of radioactive products was achieved by ascending chromatography of 5-µl samples on Whatman SG-81 paper washed with EDTA (23) and developed with chloroform, methanol, 4.2 N NH4OH (9:7:2). The lanes were cut into 1-cm zones and counted in a liquid scintillation spectrometer.
Phospholipase C Treatment of Putative IPC Synthesized from [3H]N-AcetylsphinganineThe deacylated lipid extract from the 5-h incubation of [3H]N-acetylsphinganine with strain 7R6 was dried, sonicated with 50 µl of buffer, and treated with 1 µl of phosphatidylinositol-specific phospholipase C (Bacillus thuringiensis, 390 units/ml in 50% glycerol) for 19 h precisely as described previously (13). Samples (5 µl) of each reaction mixture were chromatographed on 20-cm Whatman HP-K plates (200 µ) and radioactivity was located by using a BioScan apparatus.
In Vitro Assay of IPC SynthaseWe elected to develop a less
time consuming assay for measuring IPC synthase activity than the
published assay which uses chromatography to separate radiolabeled
substrate from product, IPC (5, 24). The improved procedure separates
the substrate and product by differential solvent extraction. A
substrate mixture containing 50,000 cpm of
[3H]N-acetylsphinganine and 20 nmol of
N-acetylsphinganine, was added to a plastic tube, and dried
using a stream of N2. The dry mixture was suspended in 50 µl of 0.2 M potassium phosphate (pH 7.0), 20 µl of 20 mM CHAPS (Sigma), 40 µl of 5 mM aqueous
phosphatidylinositol, membranes, and water to 200 µl of total volume.
Typically 75-300 µg of membrane protein (25) were added per reaction
as the reaction was linear within this range. The reaction mixture was
sonicated for 1 min in an ultrasonic water bath before addition of
membranes and then incubated for 15 or 30 min at 30 °C with gentle
shaking. The reaction was stopped by addition of 2.8 ml of 96.43%
methanol (final methanol concentration of 90%). After standing for 10 min the mixture was centrifuged at room temperature. Two ml of the supernatant fluid were mixed with 4 ml of tert-butyl methyl
ether (Sigma-Aldrich, high performance liquid chromatography grade), then with 2 ml of water, followed by vortexing, and centrifugation. The
lower phase containing the product of the reaction was extracted twice
more with 4-ml portions of mock upper phase, prepared by mixing 2 parts
of reaction mixture (lacking N-acetylsphinganine, phosphatidylinositol, and membranes) with 4 parts of
tert-butyl methyl ether plus 2 parts of water. The volume of
the lower phase was measured, and 1 ml was added to 4 ml of
scintillation mixture and counted in a liquid scintillation
spectrometer. The mixture contained 4 g/liter of
2-(4-t-butylphenyl)-5-(4
-biphenylyl)-1,3,4-oxadiazole, 3 g/liter of 2-(4
-biphenylyl)-6-phenylbenzoxazole, 33.3% Triton X-100,
and 66.7% toluene. IPC synthase activity was expressed as picomoles of
IPC made per min/mg of protein.
Cells were grown to mid-log phase in PYED without long chain base, centrifuged, and suspended at an A600 of 0.2 in fresh medium containing 25 µM PHS. Cells were incubated at 30 °C with shaking. At 0, 2, and 4 h 100 A600 units were treated with a final concentration of 5% (v/v) trichloroacetic acid, washed several times by centrifugation with water, and the pellet was extracted with 5 ml of chloroform:methanol (1:1) for 30 min at 50 °C. The extract was dried and resuspended in 0.5 ml of chloroform and applied to a 1-ml column of Adsorbosil 100-200 mesh (Applied Sciences Inc.) in a Pasteur pipette packed in methanol and equilibrated with chloroform. After washing with 3 ml of chloroform, ceramides were eluted with 3 ml of chloroform:methanol (9:1) and the eluates dried. Samples as well as 5-20 nmol of ceramide 3 standard were dissolved in absolute ethanol and dried to remove traces of water. Perbenzoylation was achieved by treating each sample with 0.5 ml of benzoyl chloride:pyridine (1:9, v/v) (anhydrous, Aldrich) for 2 h at 70 °C followed by addition of 1 ml of methanol and heating for 30 min at 70 °C. After solvent removal, 1 ml of saturated sodium carbonate in methanol plus about 50 mg of solid sodium carbonate were added followed by extraction three times with 1 ml of hexane. The combined hexane extracts were evaporated to dryness and dissolved in 1.0 ml of hexane which was applied to a 1.0-ml column (in Pasteur pipette) of AG4-X4: acetate form, packed in water, equilibrated with absolute ethanol, then hexane. After elution with 3 ml of hexane, the eluates were dried and dissolved in 200 µl of hexane.
The benzoylated ceramide (25 µl) was separated and quantified by chromatography on a 0.45 × 30-cm, 5-µm Lichrosorb Si 60 column, eluted with hexane:dioxane (93:7) at a flow rate of 1.0 ml/min and monitored at 228 nm. Data were integrated by using Millennium software (Waters and Co.).
Preparation of [4,53H]N-AcetylsphinganineN-Acetylsphingosine (Matreya, Inc.) was reduced in ethanol with tritium gas (American Radiochemicals, Inc.) in the presence of Adams catalyst. The sample was dried, dissolved in 0.2 ml of chloroform, and applied to a 0.5 × 50-cm column of Adsorbosil (100/200 mesh) prewashed with chloroform:methanol (1:1) and equilibrated with chloroform. The elution schedule was: 14 ml of chloroform followed by chloroform:methanol (97:3), collecting 7-ml fractions. Fractions 11-16 were pooled, dried, dissolved in methanol, applied to silica gel thin layer plates (Whatman LK5), and developed with chloroform:methanol (95:5). The appropriate zone, located by autoradiography, was scraped and transferred to a syringe fitted with a Teflon filter (Acro LC3S, 0.45 µm, Gelman Sciences, Inc.). The [3H]N-acetylsphinganine was eluted with 8 ml of methanol. The thin layer chromatography step was repeated. The final product, with the same mobility as authentic N-acetylsphinganine prepared according to a published procedure (26), had a specific activity of 1.5 × 107 dpm/nmol.
Miscellaneous ProceduresYeast were transformed using LiOAc-treated cells (27). Protein concentration was determined by using the Bradford reagent with bovine serum albumin as the standard (Bio-Rad).
To enrich for strains defective in sphingolipid synthesis we relied on the observation that a cell's density increases when sphingolipid synthesis is blocked, presumably because the ratio of proteins to lipids increases (7). When SLC strains are fed PHS they make sphingolipids and have a normal density. If, however, a mutation has blocked sphingolipid synthesis the cell's density should increase because sphingolipids will not be made in the presence of PHS. This strategy allowed enrichment for SLC cells having a higher density following growth in the presence of PHS.
Putative mutants were screened to differentiate those specifically defective in sphingolipid synthesis from those defective in other lipid biosynthetic pathways which might also affect cell density. The screen was based upon the observation that strain 7R6 cannot grow at low pH when it lacks sphingolipids (no PHS present in the medium) but can grow when allowed to make sphingolipids (PHS present in the medium, Ref. 20). Thus, about 100 mutants unable to grow on PYED plates at pH 4.1 either in the presence or absence of 25 µM PHS were identified.
Mutant strains were examined further for a defect specific to sphingolipid synthesis relative to glycerophospholipid synthesis by looking for decreased incorporation of [3H]inositol into inositol-containing sphingolipids relative to phosphatidylinositol (28). By this assay, two strains, AG27 and AG84, appeared to be specifically defective in sphingolipid synthesis.
Both mutants tended to die rapidly even when stored in 15% glycerol at
70 °C. To preserve the mutants, each was crossed to strain YPH2
and the resulting diploids were stored frozen. Eventually the original
haploid AG27 strain could not be revived from a frozen stock, so
haploid offspring having the same phenotypes as AG27 were obtained by
sporulating the diploid and analyzing random spores. The phenotypes of
AG27 are: Ura+, indicating the presence of the
lcb1-
::URA3 allele; Lcb+,
indicating the presence of the SLC1-1 suppressor gene;
growth inhibition at 37 °C on PYED plates containing 25 µM PHS, indicating a block in sphingolipid synthesis
(20). One such offspring, AG27-61, was used for the remainder of this
research except for the data shown in Fig. 2, which were obtained using
the original AG27 strain. Identical data to that shown in Fig. 2 were
obtained using strain AG27-61 (data not shown).
Strain AG27 Is Defective in IPC Synthase Activity
The specific defect in sphingolipid synthesis in strain AG27 was determined in three steps. First, an in vivo radiolabeling procedure was used to determine if AG27 cells could make inositol-containing sphingolipids. Cells were cultured overnight, without PHS to prevent sphingolipid synthesis, but with [3H]inositol to label PI, a substrate for synthesis of IPC and M(IP)2C. Cells were separated from unincorporated radioisotope and incubated with or without PHS in a buffered solution containing glucose. The parental strain 7R6, but not the presumptive mutant strain AG27, should exhibit a phytosphingosine-dependent transfer of the radiolabel from phosphatidylinositol to ceramide to yield [3H]IPC and to [3H]MIPC to yield [3H]M(IP)2C. Strain 7R6 showed the three expected sphingolipids (peaks 3, 4, and 5; M(IP)2C, MIPC, IPC, respectively) only when the cells were incubated with PHS (Fig. 2). In contrast, identical amounts of AG27 cells did not exhibit these peaks indicating a defect in transferring the phosphoinositol of phosphatidylinositol to sphingolipids. Both strains showed deacylation products near the origin (Fig. 2, peaks 1 and 2) derived from deacylation of phosphatidylinositol, indicating that synthesis of phosphatidylinositol was normal. These results indicate that the mutation in strain AG27 either blocks uptake of long chain base, or inactivates ceramide synthase or IPC synthase.
Second, sphingolipid synthesis in mutant strain AG27, parental strain
7R6, and wild type strain SJ21R was compared using a membrane-permeable
ceramide, [3H]N-Acetylsphinganine, to
radiolabel sphingolipids, specifically IPC. After 5 h incubation,
strains SJ21R and 7R6, but not strain AG27-61, convert
[3H]N-acetylsphinganine in vivo to
putative inositolphosphoryl-N-acetylsphinganine, a
non-deacylatable product that chromatographs on a thin layer plate at
about 10 cm, slightly slower than the internal standard, yeast IPC-3
(Fig. 3). One would not expect these IPCs to migrate identically since the radioactive product would have 24 fewer methylenes and 2 fewer hydroxyls than the yeast IPC-3. In addition to
its stability under deacylation conditions, the putative
inositolphosphoryl-N-acetylsphinganine is susceptible to
"phosphatidylinositol-specific" phospholipase C; treatment with
this enzyme results in its disappearance and appearance of
[3H]N-acetylsphinganine as shown by thin layer
chromatography (data not shown).
The failure of AG27 cells to incorporate
[3H]N-acetylsphinganine into IPC is even more
dramatic when the time course of incorporation is examined
quantitatively (Fig. 4). Strain 7R6 converts
[3H]N-acetylsphinganine to the IPC product at
a much faster rate than does the wild type strain probably because
strain 7R6 has no endogenous ceramide to compete with the radiolabeled
non-physiological ceramide, whereas, the wild type strain contains
endogenous ceramide. These results indicate that AG27-61 cells lack IPC
synthase activity.
Finally, IPC synthase activity was assayed in microsomal membranes
using the improved assay procedure described under "Experimental Procedures." This assay measures incorporation of
[3H]N-acetylsphinganine into total
sphingolipids and gives a sphingolipid profile similar to the in
vivo radiolabeling procedure as measured by thin layer
chromatography of the radiolabeled products (data not shown). Membranes
from wild type strain SJ21R and its SLC derivative 7R6 gave enzyme
activity (Fig. 5) of about 4 nmol of product/mg
protein/h which is in the range expected from enzyme activity
measurements made in vivo (5). In contrast, membranes from
strain AG27-61 contained little if any enzyme activity, the slight
increase at the 30-min time point being within background values found
over the course of many assays. Based upon the data shown in Figs. 3, 4, 5
we conclude that strain AG27 is unable to make sphingolipids because it
lacks or has very reduced IPC synthase activity.
Exogenous Phytosphingosine Kills AG27 Cells
Early in the
characterization of strain AG27 we noted sensitivity to exogenous PHS,
but it was not clear whether cells were growth-inhibited or killed. To
decide which of these possibilities was correct, cells were incubated
at a low cell density with PHS and colony forming ability, a measure of
viability, was determined. After 5 h of treatment with 10 µM PHS the viability of AG27-61 cells dropped from 100%
to about 0.3% (Fig. 6). At 24 h viability increased slightly to about 4% but there was no increase in culture density, indicating little or no growth. In the absence of PHS, AG27-61
cells grew after an initial lag and were 100% viable at all time
points. Growth of the 7R6 control cells was not inhibited by PHS and,
in fact, the cells grew faster and to a higher density in the presence
of PHS, probably because they were able to make sphingolipids.
Viability was not reduced by PHS at any time point. Thus 7R6 cells are
not killed by exogenous PHS but AG27-61 cells are.
The lcb1 gene is deleted in AG27-61 cells and they are thus unable to make endogenous long chain bases. If AG27-61 cells are killed by long chain base then they should not tolerate the wild type LCB1 gene. This hypothesis was verified by comparing the ability of AG27-61 and 7R6 to be transformed with a centromeric vector carrying the LCB1 gene (pLCB1-5, URA3). Only strain 7R6 gave transformants (Ura+) indicating that strain AG27-61 would not tolerate the LCB1 gene. The parental vector lacking the LCB1 gene (pRS316, URA3) was able to transform both strains to Ura+. Finally, strain AG27-61 was treated with a mixture of the two plasmids. Plasmid DNA from Ura+ transformants was rescued in E. coli and analyzed. All 12 Ura+ transformants contained pRS316 DNA, none contained pLCB1-5. We conclude that the LCB1 gene is toxic to AG27-61 cells.
AG27 Cells Accumulate CeramideA possible explanation for the
killing of AG27-61 cells by exogenous long chain base is that it is
metabolized to produce a high level of ceramide, a known growth
inhibitor of S. cerevisiae cells (9, 10). The level of the
predominant free ceramide (ceramide 3, comprised of PHS and OH26:O
fatty acid, Ref. 1) in cells was measured following incubation with
PHS. After 2 h of PHS treatment AG27-61 cells contained 8 times
more ceramide 3 than the parental 7R6 or wild type SJ21R cells (Fig.
7). The ceramide level continued to increase in 7R6 and
SJ21R cells but was still much lower than in AG27-61 cells at 4 h
(Fig. 7). The peak level of ceramide accumulating in AG27-61 cells is
in the range of total inositol-containing sphingolipids found in wild type cells, about 750 pmol/absorbance unit of cells (29), and may be
limited only by the available C26 fatty acid.
AUR1 complements the IPC synthase defect in AG27-61 cells
AG27-61 cells transform very poorly with plasmid DNA which prevented isolation of a gene that could complement the IPC synthase defect. Since diploid yeast strains are generally more vigorous than haploids, we examined a diploid version of AG27-61 and found that it transformed at a high enough frequency to allow screening of a recombinant DNA library. One diploid strain, designated AGD27-61, was transformed with a recombinant DNA library carried in pRS315. A pool of about 3000 Leu+ transformants was treated with 10 µM PHS to enrich for the desired transformant since AG27-61 cells die rapidly when treated with 10 µM PHS (Fig. 6). Following PHS treatment, transformants were spread at various densities on Petri dishes containing defined medium lacking leucine, agarose in place of agar, and 10 or 20 µM PHS. Agarose was used in place of agar to enhance the effective concentration of PHS which appears to bind to components in regular bacterial agar. PHS-resistant colonies were obtained from cells transformed with the genomic library but not from an equal number of cells transformed with the vector.
Plasmid DNA from four PHS-resistant colonies was rescued in E. coli. The four plasmids, termed pIPC1, gave the same EcoRI restriction fragment pattern, indicating that they carry the same genomic DNA region. Each of the four pIPC1 DNA samples gave transformants when retransformed into AGD27-61 and selected on defined medium lacking leucine and containing 1 or 12 µM PHS whereas AGD27-61 transformed with the vector pRS315 gave no PHS-resistant transformants at the higher PHS concentration. These data show that the PHS-resistance phenotype is carried on pIPC1.
The sphingolipid synthesis defect in AG27-61 cells, haploid or diploid, prevents them from growing on PYED plates having a pH of 4.1 (PYED-4.1) when PHS is present whereas the parental strain 7R6 is able to grow because it makes sphingolipids (20). If pIPC1 restores sphingolipid synthesis then AGD27-61 cells transformed with the plasmid should behave like 7R6 cells and grow on PYED-4.1 plates containing 25 µM PHS. This expectation was fulfilled by all 30 AGD27-61 transformants tested while cells transformed with the vector did not grow. These data indicate that pIPC1 carries a gene that complements the IPC synthase defect in AGD27-61 cells and restores enzyme activity.
To identify the genomic region carried in pIPC1, the nucleotide sequence at the ends of the genomic insert was determined and used to search the Saccharomyces genome data base at Stanford University. This search identified a region of 4307 bases located on Chromosome XI between coordinates 432,813 and 437,119. There is only one complete open-reading frame (YKL004w, AUR1) in this interval.
Because AGD27-61 diploids do not sporulate it was not possible to show
by genetic analysis that AUR1 is linked to the IPC synthase
defect. Instead, complementation (measured as resistance to PHS and
growth at low pH) using portions of AUR1 localized a
mutation to the 3 of the gene. For example, a purified 946-base pair
NsiI-PvuII DNA fragment, representing the 3
end
of AUR1, restored PHS resistance and growth at low pH when
transformed into AGD27-61 cells, but a purified DNA fragment
corresponding to the 5
end of the gene did not complement. The
aur1 gene was retrieved from strain AGD27-61 by gap repair
(30) of pIPC1. The DNA sequence of two independent clones was
determined between the unique NsiI site in the coding region
and the stop codon. Both clones lacked the C found in codon 325 of the
wild type allele. The deletion allele is predicted to encode a variant
protein of only 367 amino acids instead of 401 (see
"Discussion").
Membranes from AG27-61 cells transformed with pIPC1 contained about 50% more IPC synthase activity than did membranes from the 7R6 positive control cells (Table I) while membranes from AG27-61 cells contained barely detectable enzyme activity (Table I). Restoration of IPC synthase activity by pIPC1 was also examined by measuring incorporation of [3H]N-acetylsphinganine by cells into sphingolipids, specifically IPC. The concentration of radiolabeled IPC increased in A27-61 cells transformed with pIPC1 in an almost linear fashion over the 3-h course of the experiment and, thus, these cells behaved like the positive control cells containing IPC synthase activity (7R6 transformed with pRS315) whereas AG27-61 cells transformed with pRS315 showed no synthesis of radiolabeled product (Fig. 8). Based upon the data presented in Table I and Fig. 8 we conclude that pIPC1 carries a gene capable of restoring IPC synthase activity in AG27-61 cells.
|
IPC Synthase Activity Is Inhibited by Aureobasidin A
Mutations in the AUR1 gene have been shown to make
S. cerevisiae cells resistant to AbA (14, 15). An important
prediction of these results and our results is that AbA should inhibit
IPC synthase activity. The data shown in Fig. 9
demonstrate that AbA strongly inhibits IPC synthase activity, with 50%
inhibition (IC50) occurring at about 0.2 nM
AbA.
Using the experimental rationale outlined under Introduction we have isolated a mutant strain defective in IPC synthase activity, the first time such a mutant has been isolated in any organism. A defect in IPC synthase activity was verified by demonstrating in vivo that mutant cells are unable to transfer [3H]inositol from [3H]PI to ceramide to make [3H]IPC (Fig. 2) and are unable to incorporate [3H]N-acetylsphinganine into IPC (Figs. 3 and 5) whereas parental 7R6 and wild type SJ21R cells can incorporate both radiolabels into IPC. Direct assay of IPC synthase activity showed that membranes from AG27 cells contain a very reduced level of enzyme activity (Fig. 5 and Table I). The observation that AG27 cells cannot convert [3H] inositol-PI to either MIPC or M(IP)2C (Fig. 2) is the strongest evidence to date that IPC is the obligate precursor of both MIPC and M(IP)2C as shown in Fig. 1.
By complementation of the IPC synthase defect in a diploid version of strain AG27-61 we were able to isolate the AUR1 gene. Our data demonstrated that a 4307-base pair region of S. cerevisiae Chromosome XI, containing only one complete open-reading frame (YKL004w), restored IPC synthase activity to membranes prepared from AG27-61 cells transformed with pIPC1 to a slightly higher level than was seen in parental 7R6 cells (Table I). Restoration of IPC synthase activity was also demonstrated in vivo by showing that AG27-61 cells transformed with pIPC1 incorporate [3H]N-acetylsphinganine into IPC whereas cells transformed with the pRS315 control plasmid fail to do so (Fig. 8). The parent of AG27-61, strain 7R6, which contains IPC synthase activity, converts [3H]N-acetylsphinganine to IPC (Fig. 8). In addition, AG27-61 cells transformed with pIPC1 grow on agar plates containing PYED-4.1 when fed 25 µM PHS but fail to do so when transformed with the vector pRS315. This behavior indicates that pIPC1 has restored sphingolipid synthesis to AG27-61 because we had shown previously that growth at pH 4.1 only occurs if sphingolipids are made (20). Finally, AbA is known to kill S. cerevisiae cells and resistance to the drug is conferred by mutations in the AUR1 gene (14, 15). Based upon these results we predicted that AbA should inhibit IPC synthase activity. One nanomolar AbA completely inhibited IPC synthase activity (Fig. 9) indicating that it is a very potent IPC synthase inhibitor.
IPC synthase is known to be a membrane-bound enzyme (5) and the protein
would be expected to contain one or more membrane-spanning domains. The
predicted Aur1 protein was analyzed for membrane-spanning domains by
using three algorithms (31-33), each of which identified 3 or more
similar transmembrane domains (Fig. 10). This analysis strongly indicates that AUR1 encodes a membrane-bound
protein.
A search of current nucleic acid data bases using the BLAST algorithm (34) identified one putative homolog of Aur1p, a Schizosaccharomyces pombe protein (Z69086). The two proteins show about 40% amino acid identity which is spread throughout the sequences (Fig. 10). The predicted membrane-spanning domains occurring in similar regions of the proteins are shown in Fig. 10. Conservation of these domains suggests a similar conformation for the two proteins.
Using the Propsearch algorithm (35), the yeast enzymes cholinephosphotransferase (CPT1, EC 2.7.8.4) and ethanolaminephosphotransferase (EPT1, EC 2.7.8.1), which show less than 15% amino acid identity with the Aur1 protein, were found to have other properties similar to Aur1p. These similarities and the fact that the three enzymes transfer a polar phosphoalcohol group to a lipid alcohol acceptor forming a phosphodiester, suggest that Aur1p is a new member of a family of phospho-X transferases.
Our strategy for isolating mutants defective in the later steps in sphingolipid synthesis, starting with an SLC strain that can grow without making sphingolipids and cannot accumulate toxic sphingolipid intermediates, appears to have been justified based upon the findings that AG27-61 cells accumulate ceramide (Fig. 7) and are killed (Fig. 6) when fed exogenous PHS. A prediction of these results is that AG27-61 cells, which are unable to make sphingoid long chain base intermediates due to the lcb1 deletion, should not survive if transformed with a functional LCB1 gene because it should enable cells to make a toxic level of ceramide. We verified this prediction by showing that AG27-61 could not be transformed with a centromeric vector carrying LCB1 (pLCB1-5) but could be transformed with the same vector lacking LCB1 (pRS316).
Comparison of our enrichment and screening procedure with the PHS-induced cell killing experiment presented in Fig. 6 presents an apparent conundrum: how did AG27 cells survive the enrichment and screening procedure since the agar plates contained 25 µM PHS? We have observed that a higher concentration of PHS is needed to kill cells in solid than in liquid medium, so we think some mutants like AG27 survived because the concentration of PHS was too low to kill all of them. Had we used a lower concentration of PHS, more mutants like AG27 probably would have been recovered.
Ceramide, produced intracellularly from sphingomyelin, or exogenously added C2-ceramide, induces programmed cell death (apoptosis) in a variety of animal cells by an incompletely characterized signal transduction pathway that includes a ceramide-activated protein kinase (36-38). It is not known if S. cerevisiae has a ceramide-induced protein kinase nor if it has a process analogous to apoptosis, but based upon the ability of exogenous PHS to produce a large increase in the intracellular concentration of ceramide and kill AG27 cells or the intolerance of AG27-61 cells for a functional LCB1 gene, it seems possible that S. cerevisiae has a ceramide-activated death response. Ceramide accumulation in AG27 cells may prevent synthesis of suppressor lipids which are thought to compete with ceramide synthesis for C26 fatty acid (13). Thus blockage of both sphingolipid and suppressor lipid synthesis, in addition to the ceramide-activated death pathway(s), could contribute to cell death.
Aureobasidin A inhibits growth of a wide range of fungi including the human pathogens C. albicans, Cryptococcus neoformans, H. capsulatum, and Blastomyces dermatitidis, and is fungicidal against Candida (39) and S. cerevisiae (15). Previous data from our laboratory explain why inhibition of IPC synthase is fungicidal. When sphingolipid synthesis is blocked, either by withdrawal of a long chain base from an auxotrophic strain (7) or by drug inhibition of the first enzyme in the pathway (serine palmitoyltransferase) (28), S. cerevisiae cells die rapidly. These observations imply that any inhibitor of IPC synthase should be a broad spectrum fungicide. Now that AUR1 has been shown to be necessary for IPC synthase activity, it should be possible to develop high throughput screens to identify new IPC synthase inhibitors which are efficacious antifungal drugs.
This paper is dedicated to the memory of Bharath Srinivasan (deceased 1990), who isolated the original AG27 strain and found its growth to be inhibited by PHS.