INTRODUCTION

Top Abstract Introduction Procedures Results Discussion References

Syringomycin E is a member of a family of cyclic lipodepsipeptides produced by strains of the plant bacterium Pseudomonas syringae pv. syringae (1). Traditionally regarded as a virulence factor in a variety of bacterial necrotic diseases of plants (2), syringomycin E and its analogs also possess antifungal properties, and it has been suggested that these metabolites are fungal antagonists that aid survival of the producing bacteria on plants (3, 4).

How these compounds produce their toxic effects is unknown, but past physiological studies have shown that syringomycin E targets primarily the plasma membrane (1, 5, 6). To further investigate the molecular mechanisms of action of this bioactive compound, resistant mutants of Saccharomyces cerevisiae were isolated to identify genes that encode proteins necessary for growth inhibition by syringomycin E (7). Several of the mutants were deficient in sterols, and one group was complemented by the gene SYR1 (identical to ERG3), which encodes sterol C-5,6 desaturase of the ergosterol biosynthetic pathway (8). These findings, when combined with results from binding (9) and lipid bilayer (10) studies, indicate that sterols influence the interaction of syringomycin E with the target plasma membrane.

Syringomycin E action in yeast was more recently shown to require a second, nonsterol biosynthetic gene, SYR2 (11). SYR2 is identical to SUR2, which was identified in a screen for mutants that suppress the impaired recovery of rvs161 strains from nutritional starvation (12). Syringomycin E-resistant syr2 mutants showed altered glycerophospholipid levels, and the SYR2 gene product was localized to the endoplasmic reticulum (11). Nevertheless, the precise function of Syr2p was unclear from these studies.

In addition to sterols and glycerophospholipids, sphingolipids are major lipid components of the plasma membrane (13). Ubiquitous in eukaryotic cells, sphingolipids all possess a sphingoid long chain base with mainly, in fungi and plants, a hydroxyl group at the C-4 position (phytosphingosine) or, in animals, a double bond at the C-4,5 position (sphingosine). Sphingolipids serve numerous roles, including mediating cell-cell interactions, anchoring membrane proteins, acting as enzyme co-factors (14), and serving as receptors for Escherichia coli verotoxin (15-17). In addition, sphingolipids are becoming recognized as significant players in the control of cell growth, differentiation, and response to stress through the second messenger action of sphingolipid metabolites sphingosine, sphingosine-1-phosphate, and ceramide (18-20). Despite their importance, numerous gaps remain in the knowledge of sphingolipid metabolism, including the nature of the enzymes directly responsible for phytoceramide or ceramide formation from the presumed immediate precursor dihydroceramide (Fig. 1) (21).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Sphingolipid biosynthesis in S. cerevisiae. Shown is the likely pathway for de novo sphingolipid biosynthesis in yeast, along with the genes, when known, involved at each step. The two possible pathways for long chain base 4-hydroxylation, proposed to be catalyzed by Syr2p, and long chain base acylation are shown. Hydroxylation of the very long fatty acid chain is thought to occur at some point after acylation. Very long fatty acid chains of 24 and 26 carbon length and with zero, one, or two hydroxyl groups have been reported in S. cerevisiae (39). The predominant monohydroxy-C-26 fatty acid is pictured. R can be phosphoinositol, phosphoinositol-mannose, or diphosphoinositol-mannose. AUR1 is necessary for inositol phosphorylceramide synthase activity (26), whereas SYR4/IPT1 catalyzes addition of the terminal phosphoinositol (Ref. 57 and S. D. Stock, J. A. Radding, D. A. Young, and J. Y. Takemoto, unpublished results). Additional genes involved in head group assembly have not yet been identified.

In this report we present evidence that S. cerevisiae SYR2 is required for 4-hydroxylation of sphingoid bases and that this activity is necessary for syringomycin E action. We show that strains mutant in SYR2 produce sphingolipids missing the hydroxyl group at the C-4 position of the long chain base moiety, that supplying such cells with C-4 hydroxylated long chain base suppresses the syringomycin E-resistant phenotype of syr2 strains, and that strains that overexpress Syr2p are enriched in 4-hydroxylase activity.

	EXPERIMENTAL PROCEDURES

Top Abstract Introduction Procedures Results Discussion References

Yeast Strains and Growth Conditions-- Yeast strains used in this study are listed in Table I. Yeast were grown at 28-30 °C with shaking. YPD, SC-ura, SG-ura (as SC-ura with glucose replaced by galactose), and SC-trp were as described by Kaiser et al. (22). The lcb1 mutants were grown in modified YPD (1% yeast extract, 1% peptone, 4% glucose, 50 mM sodium succinate, pH 5.0, 0.05% Tergitol (U. S. Biochemical Corp.), 25-50 µM phytosphingosine·HCl (PHS)¹ or DL-erythro-dihydrosphingosine (DHS) (both from Sigma) or synthetic medium as described by Pinto et al. (23) minus either uracil or tryptophan. Stock solutions (100 mM) of PHS and DHS in 95% ethanol were diluted into 0.5% Tergitol and then added to the media to yield the indicated final long chain base and Tergitol concentrations.

Long Chain Base Analysis-- Yeast were grown in YPD and harvested at 1.5 × 10⁸ cells/ml. Methanol-HCl hydrolysates of wild type (W303C and KZ1-1C) and syr2 cells (WSYR2 and 13N-F2), as well as C-18 DHS and C-18 PHS standards were derivatized with the UV-absorbing 4-biphenylcarbonyl chloride by the method of Jungalwala et al. (24) as adapted by Dickson et al. (25). Samples were resolved by HPLC on a 250 × 4.6-mm Alltech Econosil C18 column with a 10-mm guard column. Elution was isocratic with methanol/water (90:10, v/v) as the solvent at a flow rate of 1.0 ml/min. Effluent was monitored by absorbance at 280 nm. Electrospray ionization mass spectrometry was performed by the Utah State University Biotechnology Center.

Sphingolipid Analysis-- W303C and WSYR2 cells were inoculated to 5 × 10⁶ cells/ml in modified YPD containing 0.5 mCi of [³H]inositol (New England Nuclear Co., 20 mCi/µmol) or 0.5 mCi of [4,5-³H]DHS (2.56 mCi/µmol) derived by hydrolysis of [³H]N-acetyl-DHS prepared as described previously (26). After culture for 18 h at 30 °C the cells reached a density of 1.9-2.4 × 10⁸ cells/ml, and the reaction was stopped by adding trichloroacetic acid to a final concentration of 5%. The cells were processed to deacylate the ester lipids followed by extraction of the sphingolipids as described previously (27). To further purify the acidic sphingolipids, the sphingolipid extract was bound to and eluted from AG4 resin (Bio-Rad) as described previously (28). The AG4 eluate was dried, dissolved in 1 ml of chloroform/methanol/water (16:16:5, v/v/v), and 3-7-µl aliquots were subjected to thin layer chromatography on 20-cm silica gel plates (Whatman HP-K) with the solvent chloroform/methanol/4.2 N aqueous NH₄OH (9:7:2, v/v/v). Each lane contained a mixture of yeast PHS-containing sphingolipids: 2 nmol each of inositolphosphoryl ceramide (IPC) and mannosylinositolphosphoryl ceramide (MIPC) species containing mono- and di-OH fatty acids and mannosyl-di(inositolphosphoryl) ceramide (M(IP)₂C) with a mono-OH fatty acid. Radioactivity was measured with a BioScan apparatus. The standards were located by charring after spraying with 10% (w/v) CuSO₄·5H₂O in 8% H₃PO₄ followed by heating at 160 °C for 30 min (29). The mannosylated sphingolipids in the deacylated lipid extract were also detected after thin layer chromatography of larger aliquots (125 µl) on 20-cm Whatman K5 plates developed with the same solvent as above. The plate was first treated with orcinol reagent (30) to detect the carbohydrate containing sphingolipids, MIPC and M(IP)₂C, and then treated with the CuSO₄/phosphoric acid reagent as above.

Fatty Acid Analysis-- Fatty acids from an acidic sphingolipid fraction (prepared as described above without the addition of radioisotopes) or whole cells were liberated by saponification and converted to UV-absorbing phenacyl derivatives that were resolved and quantitated by reverse phase HPLC as described previously (25).

Assay of 4-Hydroxylase Activity-- The previously constructed SYR2 overexpression plasmid, pYSYR2, placed a 5'-truncated SYR2 gene under the control of the galactose-inducible promoter GAL1 (11). For this work the truncated SYR2 insert was removed and replaced with an AccI-SphI fragment containing the entire SYR2 coding region. Expression of Syr2p from this construct, pYSYR2a, was confirmed by observation of galactose-inducible complementation of syringomycin E resistance of a syr2 strain.

Construction of lcb1 Disruptant Strains-- A disrupted LCB1 allele, lcb1-3, was constructed by replacing in pTZ18-LCB1 (32), a 1.5-kilobase pair SalI/BamHI fragment that contains the C-terminal 80% of the LCB1 gene, with a 1.4-kilobase pair TRP1-containing fragment to yield plasmid pLCB1-3. Plasmids were propagated in E. coli DH5. 3 µg of pLCB1-3 were digested with restriction endonuclease NdeI, the fragments were separated on a 0.8% agarose gel, and the 3.1-kilobase pair fragment containing the lcb1-3 allele was isolated using an Elu-Quik DNA Purification Kit as directed by the manufacturer (Schleicher & Schuell). The DNA fragment was transformed into diploid strain W303-1A by electroporation and plated onto SC-trp medium. Trp⁺ colonies were isolated and sporulated, and the resultant tetrads were dissected onto modified YPD plates containing 25 µM PHS. One Trp⁺ long chain base auxotrophic colony was selected and designated WLCB1. Replacement of the LCB1 allele with lcb1-3 was confirmed by Southern blotting using enhanced chemiluminescence detection (Amersham Pharmacia Biotech). The double disruptant, WLCB1SYR2, was produced by crossing WLCB1 with WSYR2a and selecting for Trp⁺ Ura⁺ long chain base auxotroph progeny. Standard genetic procedures were as described by Kaiser et al. (22).

Syringomycin E Treatment-- WLCB1SYR2 cells were grown overnight in modified YPD containing either 50 µM DHS or 50 µM PHS. WSYR2 and W303C were grown in modified YPD with no long chain base addition. Cells were washed once with sterile water and then transferred to modified YPD minus Tergitol and long chain base and with the indicated amounts of syringomycin E, prepared as described previously (33). Final cell densities were A₆₀₀ = 0.2 as measured in a Shimadzu UV-1201 spectrophotometer (1 A₆₀₀ unit equals approximately 3.7 × 10⁷ cells/ml). Incubation was continued at 28 °C. After 1 h Tergitol was restored to 0.05%, and DHS or PHS was restored to 50 µM as appropriate to prevent starvation for long chain base. Syringomycin E was found to be ineffective if added directly to medium containing working concentrations of Tergitol and long chain base. 16 h after syringomycin E addition, aliquots were removed and diluted 10-fold to measure growth by turbidity at A₆₀₀.

	RESULTS

Top Abstract Introduction Procedures Results Discussion References

Sequence Analysis Suggests That Syr2p Is a Diiron Binding Lipid Hydroxylase or Desaturase-- BLAST algorithm (34) comparisons of the deduced amino acid sequence of Syr2p with those in protein data bases showed significant similarities to endoplasmic reticulum proteins associated with lipid metabolism. In particular, close similarities were found with S. cerevisiae SYR1/ERG3 (C-5 sterol desaturase, score 51, p = 0.0098, 33% identities, 56% positives), Arabidopsis thaliana ERG3 (C-5 sterol desaturase, score 77, p = 0.014, 44% identities, 68% positives), and yeast ERG25 (C-4 sterol methyl oxidase, score 65, p = 4.5 × 10⁸, 52% identities, 65% positives). Similar findings were reported by Bard et al. (35) and Li and Kaplan (36). Despite the similarity of Syr2p to enzymes of sterol biosynthesis, an involvement for Syr2p in sterol metabolism could not be found. Comparisons of sterol profiles of syr2 mutants with similarly grown wild type strains revealed no differences (11),² indicating that Syr2p functions in some other metabolic pathway.

View larger version (9K): [in this window] [in a new window]   Fig. 2.   Putative metal binding domains of Syr2p and similarity to sterol biosynthetic enzymes. The sequences were obtained as part of a BLAST analysis and are from S. cerevisiae (S.c.) or A. thaliana (A.t.). Histidine residues identified by Shanklin et al. (58) as being involved in metal binding are indicated in bold type. Histidine residues identified by Li and Kaplan (36) as a fourth histidine cluster motif specific to the ERG3/ERG25/SYR2 family of sequences are underlined.

Yeast Strains Deficient in SYR2 Lack the Sphingoid Long Chain Base Phytosphingosine-- Sphingolipids in yeast are normally composed of the sphingoid long chain base PHS (D-4-hydroxysphinganine), typically of 18 or 20 carbon chain length (C-18 and C-20, respectively), an amide-linked very long chain fatty acid, primarily mono-hydroxy-C-26 chains with lesser amounts of di- and nonhydroxyl forms, and a phosphoinositol-containing head group (39) (Fig. 1). To test the involvement of Syr2p in the sphingoid base 4-hydroxylation, the sphingoid base compositions of mutant strain Wsyr2 and the isogenic wild type strain W303C were determined (25). Reverse phase HPLC separation of biphenylcarbonyl-derivatized long chain bases derived from the wild type strain W303C revealed two peaks of UV-absorbing material, as expected, with retention times of 15 and 25 min (Fig. 3B). Coelution with a derivatized C-18 PHS standard and electrospray mass spectral analysis of the material collected from the two peaks verified their identities as C-18 PHS at 15 min and mainly C-20 PHS at 25 min. The derivatized long chain bases from the syr2 mutant strain also separated into two peaks, with one again eluting at 25 min but the other eluting at 42 min (Fig. 3C). Little material with a retention time of 15 min was apparent (<0.3% of total long chain base). Authentic C-18 DHS treated in the same manner as the lipid extracts eluted with a retention time of 25 min (Fig. 3A). It was not possible to distinguish between the C-20 PHS and C-18 DHS derivatives by the chromatography system used, but mass spectral analysis of the syr2 material eluting at 25 min revealed a molecular ion mass of 482.3 Da, consistent with its assignment as the N-biphenylcarbonyl derivative of C-18 DHS. The mass of the material eluting at 42 min was 510.4 Da, as expected for N-biphenylcarbonyl-C-20 DHS. Similar results were obtained with an independently isolated syr2 mutant strain in a different genetic background (13N-F2 (syr2) and KZ1-1C (SYR2); data not shown).

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Sphingoid long chain bases produced by wild type and syr2 strains. Methanol-HCl hydrolysates of wild type strain W303 (B) and Wsyr2 cells (C), as well as C-18 DHS and C-18 PHS standards (A), were derivatized with the UV-absorbing 4-biphenylcarbonyl chloride and resolved by reverse phase HPLC as described under "Experimental Procedures." Effluent was monitored by absorbance at 280 nm.

Fatty Acid Analysis of syr2 Strain-- The effect of the syr2 mutation on hydroxylation of the very long chain fatty acid component of sphingolipids was also examined as described under "Experimental Procedures." Hydroxylation of the very long chain fatty acids was observed in the absence of Syr2p activity. For example, the percentage of distribution of non-, mono-, and di-hydroxyl fatty acids for wild type cells was 8, 78, and 14%, respectively, and for syr2 cells, it was 53, 47, and 0%, respectively. Clearly syr2 cells hydroxylate the very long chain fatty acid in sphingolipids. Why the distribution of hydroxylated species differs from wild type is not known.

In Vitro Measurement of 4-Hydroxylase Activity-- Sphingoid base 4-hydroxylase activity was measured in microsomal fractions, because Syr2p has previously been localized to the endoplasmic reticulum (11). Microsomes, prepared from SYR2 wild type, overexpression, and deletion strains, were supplied with substrate, either DHS or dihydroceramide solubilized in CHAPS, along with NADH and NADPH. Both NADH and NADPH have been reported to be cofactors for activity of other putative diiron proteins involved in oxygen-dependent reactions of hydrocarbon substrates. After 90 min at 25 °C, sphingoid long chain bases were released by methanol-HCl hydrolysis and extracted, and their 4-biphenylcarbonyl derivatives were separated and quantitated by reverse phase HPLC. Hydroxylated product was apparent when DHS or dihydroceramide were supplied to microsomes from a SYR2 overexpressing strain, W303C(pYSYR2a). Using the same amount of protein, 3-4-fold less hydroxylated product was produced if NADH and NADPH were omitted from the reaction or if the source of microsomes was a wild type strain, W303C(pYES2), containing only the chromosomal copy of SYR2 and a control plasmid (Table II). No 4-hydroxyl products were detected if microsomes were from the deletion strain Wsyr2.

View this table: [in this window] [in a new window]   Table II In vitro 4-hydroxylase activity Microsomes were prepared from the indicated strains, then incubated with either 50 nmol DHS or dihydroceramide with or without 0.1 mM each NADH and NADPH. Sphingoid long chain bases were treated with methanol-HCl, extracted, derivatized, and analyzed by reverse phase HPLC as described under "Experimental Procedures." Values are the averages of two to three trials with standard deviations for 90-min reactions containing 0.7 mg of microsomal protein. ND, none detected. , not measured.

Growth in the Presence of PHS Suppresses the syr2 Phenotype-- To confirm that the syringomycin E resistance phenotype of the syr2 mutant is due to a loss of hydroxylase activity rather than an additional unknown activity of Syr2p, we wished to test whether syringomycin E sensitivity is restored by supplying syr2 cells with the product of the hydroxylase. It is still uncertain, however, if the hydroxylase substrate is phytoceramide, PHS, or both (see "Discussion"). As yeast do not readily utilize exogenous ceramides but will incorporate exogenous sphingoid long chain bases into the sphingolipid biosynthetic pathway (40), syringomycin E sensitivity was tested following growth in medium containing hydroxylated (PHS) or nonhydroxylated (DHS) long chain bases rather than hydroxylated or nonhydroxylated ceramides. Also, to ensure that all sphingolipids were built on the exogenous long chain base, the LCB1 gene was deleted in strain Wsyr2a (see "Experimental Procedures"). LCB1 encodes a subunit of serine palmitoyltransferase, the first enzyme in sphingolipid biosynthesis (Fig. 1). An lcb1 mutation results in auxotrophy for long chain bases (32).

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Effect of PHS and DHS on the syringomycin E resistance syr2 phenotype. WLCB1SYR2, cultured in the presence of either PHS () or DHS (), W303C () and WSYR2 () cells were diluted into medium with the indicated amounts of syringomycin E. Syringomycin E resistance, as measured by cell growth (turbidity (O.D.600)), was evaluated 16 h after syringomycin E addition. The example shown is representative of three independent experiments.

Sphingolipid Analysis of syr2 Strain-- The fact that syr2 mutants are viable, whereas sphingolipids are essential, raised questions about the mutant lipid composition. The loss of long chain base hydroxylation could result in the simple substitution of nonhydroxylated long chain base for 4-hydroxy-long chain base in the sphingolipid pools or, alternatively, in an alteration of the overall levels or types of sphingolipids. The recovery of similar quantities of long chain base from mutant and wild type strains (Fig. 3) would argue against a large change in total sphingolipid content, but in order to more directly investigate the quantity and nature of sphingolipids produced by the syr2 mutant, wild type and syr2 cells were cultured overnight with either [³H]inositol or [³H]DHS to label the sphingolipids. The cells were processed to deacylate the ester lipids, such as phosphatidylinositol (PI), and sphingolipids were extracted. Based on the radioactivities of [³H]inositol-labeled extracts before and after PI deacylation, the distribution and amounts of PI and inositol sphingolipids were about the same in wild type (25.7 × 10⁶ cpm PI, 15.3 × 10⁶ cpm sphingolipid) and syr2 cells (30.6 × 10⁶ cpm PI, 14.4 × 10⁶ cpm sphingolipid), confirming that the mutant strain was able to make as much total sphingolipid as the wild type.

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Thin layer chromatography of sphingolipids obtained by in vivo labeling of wild type and syr2 cells with [3H]DHS. Wild type W303C and Wsyr2 cells were grown in the presence of [4,5-3H]DHS and then processed to deacylate the ester lipids, followed by extraction and purification of the sphingolipids as described under "Experimental Procedures." Samples were subjected to thin layer chromatography on 20-cm silica gel plates (Whatman HP-K) with the solvent chloroform/methanol/4.2 N aqueous NH4OH (9:7:2, v/v). Each lane contained a mixture of yeast PHS-containing sphingolipids: 2 nmol each of IPC and MIPC species containing mono- and di-OH fatty acids and M(IP)2C with a mono-OH fatty acid. The standards were located by charring after spraying with 10% (w/v) CuSO4·5H2O in 8% H3PO4 followed by heating at 160 °C for 30 min (29). The radioactivity was evaluated with a BioScan apparatus (solid line, wild type; dotted line, mutant). The upper curve was displaced upward by 1500 counts for display purposes. G, H, I, J, and K refer to the locations of an unlabeled mixture of yeast PHS containing sphingolipids added to each lane and detected by charring. G and H, IPCs with mono- and di-OH fatty acids, respectively; I and J, MIPCs with mono- and di-OH fatty acids, respectively; K, M(IP)2C with a mono-OH fatty acid. A, B, C, D, E, and F are the locations of peaks in the mutant strain, all with DHS. A and B, IPCs with zero and mono-OH fatty acid, respectively; C and D, MIPCs with zero and mono-OH fatty acid, respectively; E and F, M(IP)2Cs with zero and mono-OH fatty acid, respectively.

	DISCUSSION

Top Abstract Introduction Procedures Results Discussion References

The present study of yeast gene SYR2 reveals that sphingolipids play a key role in the action of the antifungal syringomycin E. It also uncovers a previously uncharacterized activity of the sphingolipid biosynthetic pathway. The lack of PHS-based sphingolipids and accumulation of DHS-based sphingolipids in syringomycin E-resistant syr2 mutants, the greater 4-hydroxylase activity in a SYR2 overexpression strain, the lack of activity in a syr2 mutant strain, and the ability to bypass the syr2 defect with exogenously added PHS provide support for a biosynthetic role of Syr2p in 4-hydroxylation of sphingoid bases. In addition, the sequence similarity of Syr2p to membrane hydroxylases and desaturases and its localization in the endoplasmic reticulum (11), the site of early sphingolipid biosynthetic steps (41), suggest that Syr2p is the enzyme that catalyzes this specific hydroxylation step. An essential catalytic role in fatty acid hydroxylation is not likely because very long chain fatty acid hydroxylation was still detected in the syr2 mutant despite a complete loss of SYR2 transcripts (11). While this report was in preparation, Haak et al. (42) reported that an independent sur2/syr2 mutant produces sphingolipids lacking long chain base 4-hydroxylation, whereas a second gene product, Scs7p, is required for sphingolipid very long chain fatty acid hydroxylation.

Identification of Syr2p as the sphingolipid 4-hydroxylase will pave the way for isolation and characterization of the enzyme. This will, in turn, permit clarification of several issues regarding sphingolipid biosynthesis. One important question concerns the identity of the lipid substrate of Syr2p. The analogous reaction in animal cells, desaturation at the C-4 position, is thought to occur at the level of dihydroceramide (43). In yeast, however, it is not known if 4-hydroxylation occurs before or after long chain base acylation, i.e. if Syr2p converts DHS to PHS, dihydroceramide to phytoceramide, or both (Fig. 1). Inhibition of ceramide synthesis leads to accumulation of both DHS and PHS (44, 45), suggesting a direct conversion from DHS to PHS is possible, at least under conditions of limited ceramide formation. We have shown that incubation of either DHS or dihydroceramide with microsomes from a SYR2 overexpressing strain leads to hydroxylation. With crude microsome preparations, however, enzymes capable of catalyzing acylation or deacylation of the added substrate before hydroxylation may be present and obscure the true nature of the substrate.

Confirmation that Syr2p is in fact an iron-containing oxidative enzyme as predicted from the sequence (Fig. 2) will also be afforded by biochemical analysis. The stimulation of 4-hydroxylase activity by reduced pyridine nucleotide, as shown here, is consistent with Syr2p being a member of this enzyme family. Molecular oxygen has been shown to be the main source of oxygen added to dihydroceramide (46), but little else is known of the mechanism of this reaction. Knowledge about Syr2p will also reveal mechanisms about mammalian sphingolipid biosynthesis. Phytoceramide is produced by certain mammalian tissues (47) as well as by fungi and plants, and these tissues are predicted to contain a Syr2p homolog. The primary mammalian sphingolipids based on ceramide contain the long chain base sphingosine, which has a C-4,5 double bond rather than the 4-hydroxyl group. Recent reports concerning the enzyme in rats that catalyzes this reaction, dihydroceramide desaturase, suggest it also has properties similar to diiron-containing lipid desaturases and hydroxylases (43, 48).

How sphingolipids, and more specifically sphingolipid 4-hydroxylation, allow susceptibility to syringomycin E can only be speculated. One possibility is that 4-hydroxylated sphingolipids directly bind this antifungal compound at the cell surface. The 4-hydroxyl group is expected to influence the degree of sphingolipid exposure on the membrane surface, but it will also affect lipid and protein nearest neighbor interactions in the plane of the membrane. Another possibility is that 4-hydroxylated sphingolipids indirectly influence syringomycin E-cell interaction by modulating sterol or glycerophospholipid compositions or both. Syringomycin E action is influenced by sterols (8, 10), phospholipid bilayers facilitate ion channel formation by syringomycin E molecules (49), and syr2 mutants have lowered cellular glycerophospholipid levels (11). Despite evidence for cross-regulation of the biosynthetic pathways of these various lipid classes in yeast (50), it is difficult to predict precisely how an alteration in the hydroxylation state of sphingolipids could influence cellular sterol and phospholipid composition. Furthermore, 4-hydroxylation could influence the insertion and assembly of lipids as well as proteins into the plasma membrane. Finally, the requirement for sphingolipid 4-hydroxylation may reflect the involvement of phytoceramide-mediated growth inhibition processes in syringomycin E action. Phytoceramide and ceramide, but not dihydroceramide, are reported to mediate cell death in animal cells and growth inhibition in yeast (18, 51, 52), although the phenomenon is not always observed (53). Exposure of yeast cells to syringomycin E may cause increased cellular levels of phytoceramide (perhaps by activating sphingolipid turnover), which in turn may activate specific protein kinases and phosphatases (18, 51, 52), leading to growth arrest. Without the ability to hydroxylate dihydroceramide to phytoceramide and the consequent substitution of dihydroceramide into mature sphingolipids, syr2 mutants would be incapable of undergoing this process.

The observation that SYR2 encodes a nonessential function raises questions about the cellular roles of 4-hydroxylated sphingolipids in yeast growth and survival. Normal SYR2 strains produce sphingolipids that are based almost exclusively on phytoceramide (this study and Ref. 54), but syr2 mutants grow well with dihydroceramide-based sphingolipids. Sphingolipids are indicated to be required for maintaining proton permeability barriers across the membrane or for proton extrusion (55) and for maturation of glycosylphosphatidylinositol-anchored proteins (56). Preliminary observations, however, show that syr2 mutants display wild type growth phenotypes under conditions where proper functioning of these processes may be essential, namely at acidic pHs (4.1), high temperatures (39 °C), and high salt concentrations (0.75 M NaCl).³ Also, Calcofluor staining of chitin, a probe of cell wall structure, was unperturbed in the syr2 mutant, although growth of the syr2 mutant was slightly retarded by Calcofluor.⁴ The two phenotypes previously associated with syr2/sur2 mutations, resistance to syringomycin E, and suppression of rvs161 mutations can now be said to be associated with a loss of 4-hydroxylation of the long chain base moiety of sphingolipids. The only apparent commonality of these two phenotypes is growth restoration under conditions that inhibit growth of wild type cells. The mechanism(s) responsible for these effects await elucidation, as does a clear definition of the role of 4-hydroxy-sphingolipids in yeast biology and syringomycin E action.