If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Both Sphingolipids and Cholesterol Participate in the Detergent Insolubility of Alkaline Phosphatase, a Glycosylphosphatidylinositol-anchored Protein, in Mammalian Membranes (∗)
To whom correspondence and reprint requests should be addressed. Present address: Dept. of Biochemistry and Cell Biology, National Institute of Health, 1-23-1, Toyama, Shinjuku-ku, Tokyo 162, Japan. Fax: 81-3-5285-1157.
Affiliations
Department of Biochemistry and Cell Biology, National Institute of Health, 1-23-1, Toyama, Shinjuku-ku, Tokyo 162, JapanDepartment of Embryology, Carnegie Institution, Baltimore, Maryland 21210
∗ This work was supported in part by a grant-in-aid from the Ministry of Education, Science and Culture of Japan and also by a long-term fellowship from International Human Frontier Science Program (to K. H.) and United States Public Health Service Grant R37 GM22942 (to R. E. P.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ¶ Present address: Mayo Clinic and Foundation, Guggenheim 6, Rochester, MN 55905-0001.
SPB-1, a Chinese hamster ovary cell variant defective in serine palmitoyltransferase activity for sphingolipid synthesis, provides a useful system for studying the effects of sphingolipids and/or cholesterol deprivation on cellular functions and membrane properties. To investigate whether there was an interaction among sphingolipids, cholesterol, and glycosylphosphatidylinositol (GPI)-anchored proteins in biological membranes, we introduced human placental alkaline phosphatase (PLAP) in SPB-1 and in wild type cells by stable transfection and examined the effects of sphingolipid and/or cholesterol deprivation on the solubility of PLAP in Triton X-100. Although the PLAP solubility of the membranes isolated from the control cells in Triton X-100 was only 10%, deprivation of sphingolipid and cholesterol further enhanced the solubility, which reached 50% when both sphingolipids and cholesterol were deprived. The enhanced solubility was suppressed to the control level by metabolic complementation with exogenous sphingosine and cholesterol. The sphingolipid and cholesterol content of the isolated membranes changed independently, eliminating the possibility that sphingolipid deprivation induced a reduction in cellular cholesterol and enhanced PLAP solubility and vice versa. It was also unlikely that the enhanced solubility was due to structural changes in PLAP molecules since, regardless of sphingolipid and cholesterol deprivations, almost all PLAP had the GPI-anchor moiety and there were no differences in the apparent molecular weight of the protein in supernatant and precipitate fractions of the detergent-treated membranes. In addition, the expression level of caveolin in the isolated membranes was not significantly affected by sphingolipids and/or cholesterol depletion. These results indicated that both sphingolipids and cholesterol were involved in the PLAP insolubility and suggested that these lipids coordinately played a role in formation of Triton X-100resistant complexes.
INTRODUCTION
Lipids are now recognized not only as constituents of the fluid matrix of biological membranes but also are thought to play a role in the formation of microdomains in membranes(
). Membrane lipids of mammalian cells consist mainly of three different classes of lipids (glycerolipids, sterols, and sphingolipids), which are categorized according to the structure of their hydrophobic backbones. Although the lipid composition of biological membranes depends on cell type and on the types of intracellular organelles, glycerolipids are the most abundant class of membrane lipids with ratios of phosphatidylcholine and phosphatidylethanolamine to total phospholipids of 30-60% and 10-30%, respectively, in various organelles(
). Cholesterol, which is the major sterol in most mammalian cells, is preferentially localized to the plasma membrane and amounts to about 10-20% of the total plasma membrane lipid(
). Like cholesterol, complex sphingolipids (sphingomyelin and glycosphingolipids) are preferentially localized to the plasma membrane, and, moreover, both cholesterol and complex sphingolipids are highly enriched in the exoplasmic leaflet of the plasma membrane bilayer(
). Although it is still unclear whether the similarity in distribution of these lipids has a physiological significance, several investigators have shown that cholesterol has a stronger affinity for sphingomyelin than for glycerophospholipids in model membrane systems (reviewed in (
)-anchored proteins, which are widely present in various types of cells from lower eukaryotes to mammalian cells, associate with the plasma membrane by integration of their phosphatidylinositol moiety in the exoplasmic leaflet of the membrane bilayer(
) recently demonstrated that, when Madin-Darby canine kidney cells expressing PLAP are treated with Triton X-100, most PLAP molecules are recovered as insoluble membranous materials where sphingomyelin and glycosphingolipids are also enriched. Their findings raised the possibility that the insolubility of GPI-anchored proteins might be conferred by the lipid environment(
). Interestingly, lipid composition analysis showed that cholesterol is also enriched, compared to glycerophospholipids, in the Triton X-100-insoluble fraction, although sphingolipids are enriched in the insoluble fraction to a much greater extent than cholesterol(
) who demonstrated that treatment of cells with saponin, a detergent which extracts cholesterol, enhances solubility of a GPI-anchored protein in Triton X-100; however, no analysis of sphingolipids was presented in their study. While an interaction between sphingolipids and GPI-anchored proteins was suggested by our previous findings that sphingolipid deficiency induces hypersensitivity of a GPI-anchored protein to PI-PLC(
), in the previous study we did not show any evidence for participation of sphingolipids in the Triton X-100 insolubility of GPI-anchored protein. Thus, there is no direct evidence that sphingolipids participate in the Triton X-100 insolubility of GPI-anchored proteins or that there is an additive or synergistic effect of sphingolipids and cholesterol on the insolubility of GPI-anchored proteins. To address these points, we introduced human placental alkaline phosphatase (PLAP), a typical GPI-anchored protein, into wild type CHO and mutant CHO cells defective in sphingolipid biosynthesis and examined the effects of deprivation of sphingolipids and/or cholesterol on the insolubility of PLAP in Triton X-100.
EXPERIMENTAL PROCEDURES
Materials
D-erythro-Sphingosine, cholesterol, cholesterol oleate, glucosylceramide, and GM3 were purchased from Matreya (Pleasant Gap, PA), and egg sphingomyelin was from Avanti Polar Lipids (Alabaster, AL). DL-Mevalonic acid lactone, p-nitrophenyl phosphate, and Triton X-114 were purchased from Sigma and Triton X-100 from Pierce. Rabbit anti-PLAP polyclonal antibody and rabbit anti-human caveolin polyclonal antibody were from Zymed Laboratories (San Francisco, CA) and Transduction Laboratories (Lexington, KY), respectively. Plasmids PLAP513 and PLAP489HA were provided by Dr. Deborah Brown (Dept. of Biochemistry and Cell Biology, State University of New York, Stony Brook), and compactin was from Dr. Robert Simoni (Dept. of Biological Sciences, Stanford University, Stanford). PI-PLC, which was purified from Escherichia coli transfectants overexpressing Bacillus thuringiensis PI-PLC, was the gift of Dr. Michael Edidin (Dept. of Biology, The Johns Hopkins University, Baltimore).
Cell Culture and Isolation of CHO Transfectants Expressing PLAP
The CHO mutant SPB-1, which is defective in sphingoid base biosynthesis(
), the wild type CHO-K1, and their variants expressing PLAP were maintained in Ham's F-12 medium supplemented with 2 mML-glutamine, 5% fetal bovine serum, penicillin G (100 units/ml), and streptomycin sulfate (100 μg/ml) at 33°C in a water-saturated atmosphere of 5% CO2 in air. Plasmid PLAP513 is a recombinant plasmid which expresses GPI-anchored PLAP(
)CHO cells were co-transfected with plasmid PLAP513 or PLAP489HA and pSV2neo by lipofection, and G418-resistant transfectants were seeded to form replica colonies on polyester disks as described previously(
). Thereafter, immunological screening for PLAP-expressing colonies were performed with the replica disks at room temperature. After rinsing with PBS, the disks were incubated in 5 ml of 4% formaldehyde/PBS for 15 min to fix the colonies. The disks were then rinsed three times with 5 ml of PBS, incubated in 5 ml of 100 mM NH4Cl/PBS for 15 min, rinsed with PBS, and incubated in 5 ml of 3% BSA/PBS-T (PBS containing 0.1% Tween 20) for 1 h. The disks were then incubated with 2.5 ml of 0.5% BSA/PBS-T containing anti-PLAP antibody for 1 h with gentle shaking. After washing four times in 5 ml of 0.5% BSA/PBS-T for 10 min, the disks were incubated in 2.5 ml of 0.5% BSA/PBS-T containing horseradish peroxidase-conjugated goat anti-rabbit IgG (Bio-Rad Laboratories) for 1 h. After washing the disks, they were incubated with 2 ml of chemiluminescence reagents (DuPont NEN) and exposed to x-ray film, where colonies expressing PLAP were visualized. These colonies were retrieved from the master dishes and purified as described(
Deprivation of Sphingolipids and Cholesterol of Cells
In a typical experiment, 8 × 105 cells (CHO-K1 variants) or 2 × 106 cells (SPB-1 variants) in 10 ml of F-12 containing 5% fetal bovine serum were seeded into a 150-mm diameter tissue culture dish on day 0 and cultured for 24 h at 33°C. On day 1, after washing the cell monolayers three times with 10 ml of PBS, the medium was changed to 25 ml of Nutridoma-BO, a sphingolipid-deficient medium, (
) supplemented with gentamicin (25 μg/ml). Then, the cells were grown in this lipid-deficient medium for 3 days at 39°C to about 70% confluency. To deprive cellular cholesterol, 25 μl of 2 mM compactin in ethanol and 0.25 ml of 10 mM mevalonate were added to the Nutridoma-BO medium on day 1, and 0.25 ml of 10 mM mevalonate was also added to the medium on days 2 and 3. To replenish cholesterol in compactin-treated cells, 2.5 ml of cholesterol suspension, which was made by dilution of 0.1 ml of a 12.5 mg/ml ethanolic solution of cholesterol with 5 ml of Nutridoma-BO just before use, was added to the medium on day 3, and the cells were cultured for 1 more day before harvest. For metabolic complementation of sphingolipids, sphingosine as a complex with defatted BSA (
) was added daily at a concentration of 2 μM to the medium.
Membrane Preparation
Membranes were prepared from the monolayers at 4°C or on ice. The monolayers were rinsed twice (or four times when exogenous cholesterol had been added to the culture medium) with 10 ml of PBS and harvested in PBS by scraping. After centrifugation (300 × g, 5 min), the cells from two subconfluent dishes were suspended in 3 ml of buffer L (10 mM sodium Hepes, pH 7.5, containing 0.25 M sucrose, 5 mM EDTA, 50 μM 4-amidinophenylmethanesulfonyl fluoride, and 10 μg/ml leupeptin), and lysed with a probe-type sonicator by 20 sonic pulses. After centrifugation of the lysate at 2500 × g for 15 min, membranes were separated from the supernatant by centrifugation at 105× g for 30 min. The membranes were suspended in buffer L with a 26-gauge needle and stored at −80°C until use.
Solubility of PLAP in Triton X-100
All manipulations were carried out at 4°C or on ice unless otherwise noted. Membranes (50 μg of protein) in 100 μl of buffer L were placed in a microcentrifuge tube (Beckman), mixed with 400 μl of buffer S (10 mM sodium Hepes, pH 7.5, containing 150 mM NaCl, 5 mM EDTA, 50 μM 4-amidinophenylmethanesulfonyl fluoride, and 10 μg/ml leupeptin) containing 1.25% Triton X-100, and incubated for 30 min. Half of the mixture was withdrawn and kept as a total fraction, and the residual half was centrifuged at 105× g for 30 min with a Beckman TLA100.3 rotor. Aliquots of these total and supernatant fractions were used for assays of alkaline phosphatase, determination of protein, and Western blotting. Solubility of PLAP was represented as the ratio of alkaline phosphatase activity in the supernatant fraction to that in the total fraction. Where indicated, the precipitates were suspended in 250 μl of buffer S with a 26-gauge needle and also subjected to Western blotting.
Alkaline phosphatase activity was determined as described previously (
) with the following modifications. Briefly, 100 μl of sample (containing up to 10 μg of protein in 1% Triton X-100/buffer S) was added to 800 μl of 0.1 M diethanolamine containing 2 mM MgCl2 and 5 mMp-nitrophenyl phosphate, and absorbance at 410 nm of the mixture was monitored at room temperature for 30 s with a Shimadzu UV-160 spectrophotometer using a time-scanning mode. One arbitrary unit of alkaline phosphatase activity was defined as the activity producing 0.1 A410 per min.
Lipid Analysis
For sphingolipid determination, membranes were suspended in H2O and extracted by adding chloroform and isopropyl alcohol to the suspension at a final ratio of chloroform/isopropyl alcohol/H2O = 7:11:2 (v/v/v)(
). Lipid extracts were analyzed by thin layer chromatography on Silica Gel 60 F-254 plates (Merck) with chloroform/methanol/0.2% CaCl2 = 80:30:5 (v/v/v) as the developing solvent. Sphingomyelin and GM3 separated on the silica gel plates were stained with Coomassie Blue and quantified by densitometric analysis(
), using known amounts of sphingomyelin and GM3, respectively, as the standards. For cholesterol determination, lipids were extracted from the membranes by the method of Bligh and Dyer(
), using known amounts of cholesterol and cholesterol oleate, respectively, as the standards. The amount of membrane cholesterol ester was estimated by subtracting the amount of free cholesterol from the total cholesterol.
PI-PLC Treatment and Partitioning by Triton X-114 Phase Separation
Precondensation of Triton X-114 and phase separation of membrane proteins in Triton X-114 were performed as described previously (
) with the following modifications to prevent incomplete partitioning of PLAP. Membranes (10 μg of protein) were incubated in 50 μl of buffer L containing 0.4% Triton X-100 with or without PI-PLC (10 units) at 37°C for 1 h. After addition of 150 μl of buffer L containing 0.5 M NaCl and 20 μl of precondensed Triton X-114, the mixture was cooled on ice for 1 min, warmed at 37°C for 3 min, and centrifuged at 104× g for 1 min at room temperature without a sucrose cushion (these procedures are hereafter referred as to phase separation). The upper phases were mixed with 20 μl of precondensed Triton X-114 and, after phase separation once more, the second upper phases were recovered as aqueous phase fractions and the second lower phases were discarded. Similarly, the lower phases were mixed with 200 μl of buffer L containing 0.5 M NaCl and, after phase separation once more, the second upper phases were discarded and the second lower phases were recovered as detergent phase fractions. The recovered detergent fractions were diluted with 200 μl of buffer L containing 0.5 M NaCl and 1% Triton X-100 before being subjected to alkaline phosphatase assay.
Protein Determination and Western Blotting
Protein content was determined by the method of Schaffner and Weissmann (
) using BSA as the standard except that the final concentration of SDS in the samples was increased from 0.1% to 1%. Under the modified conditions, the presence of at least 1 mg of Triton X-100, n-octyl β-glucopyranoside, or CHAPS in the samples did not affect the protein determination. (
), and the separated proteins were transferred to a polyvinylidene difluoride membrane with a Mini Trans-Blot electrophoretic transfer system (Bio-Rad)(
). PLAP and caveolin were detected on the membrane by enhanced chemiluminescence using rabbit anti-PLAP antibody and anti-caveolin antibody, respectively, as the primary antibodies and horseradish peroxidase-conjugated goat anti-rabbit IgG as the secondary antibody.
RESULTS
Effects of Sphingolipids and Cholesterol Deprivation on Insolubility of PLAP in Triton X-100
Strain SPB-1 is a temperature-sensitive CHO mutant cell line with a defect in serine palmitoyltransferase activity which catalyzes the first step for sphingolipid biosynthesis, so that the SPB-1 cells cease de novo synthesis of sphingolipids at nonpermissive high temperatures (
). SPB-1 and the wild type CHO-K1 cells were transfected with a plasmid encoding human placental alkaline phosphatase, a typical GPI-anchored protein, and transfectants (named SPB-1/PLAP and CHO-K1/PLAP, respectively) expressing PLAP were obtained as model cell systems to examine effects of sphingolipid deficiency on insolubility of GPIanchored proteins in Triton X-100. Since alkaline phosphatase activity of membranes prepared from the non-transfectants was less than 2% of that from these transfectants (Table 1), we used, hereafter, alkaline phosphatase activity as a measure of PLAP molecules of the transfectant membranes.
For deprivation of membrane sphingolipids, SPB-1/PLAP cells were cultured in a lipid-deficient medium at 39°C for 3 days. Membranes prepared from these cells were incubated in 1% Triton X-100 at 4°C, and the recovery of alkaline phosphatase activity in the supernatant after high speed centrifugation was compared between CHO-K1/PLAP and SPB-1/PLAP cell membranes. Solubility of PLAP of the wild type cell membranes in Triton X-100 was only about 10% (Table 2). Interestingly, SPB-1 cell membranes showed three times higher solubility of PLAP than the wild type cell membranes (Table 2).
For deprivation of membrane cholesterol by inhibiting cholesterol synthesis, cells were cultured in lipid-deficient medium at 39°C in the presence of compactin, a potent inhibitor of 3-hydroxy-3-methylglutaryl-CoA reductase(
). During the 3-day compactin treatment, 0.1 mM mevalonate was added daily to the culture medium to complement non-sterol isoprenoid products essential for cell growth(
), so that the cells could sustain viability (estimated by trypan blue extrusion). Compactin treatment of CHO-K1/PLAP cells increased the solubility of PLAP about 2-fold (Table 2). Also, compactin treatment of SPB-1/PLAP cells increased the solubility to 50% (Table 2).
To verify that the GPI-anchor was an essential part of PLAP for the insolubility in Triton X-100, we used PLAP-HA, a chimeric protein which has a single membrane-spanning domain of influenza hemagglutinin at the carboxyl-terminal region of PLAP in place of GPI-anchor2 (Table 1). Nearly 100% of the activity of PLAP-HA of membranes prepared from CHO-K1 and SPB-1 transfectants was solubilized in Triton X-100 (Table 2), confirming that the GPI-anchor moiety of PLAP was an essential part of PLAP insolubility.
Restoration of PLAP Insolubility by Metabolic Complementation
The finding that PLAP of SPB-1 cell membranes showed much higher solubility in Triton X-100 than PLAP of the wild type cell membranes implied that the enhanced solubility was due to sphingolipid deficiency in the cells. However, since the SPB-1 cell line was isolated from mutagenized CHO-K1 cells(
), mutations unrelated to sphingolipid metabolism might cause the enhanced solubility in SPB-1/PLAP cells. This possibility could be eliminated if exogenous sphingosine, which metabolically bypasses the serine palmitoyltransferase defect in SPB-1 cells(
), restored the PLAP insolubility to wild type levels. Indeed, when cells were cultured in the presence of sphingosine, the PLAP solubility of the SPB-1/PLAP cell membranes was suppressed to the wild type level (Fig. 1) whereas the solubility of the wild type cells was not affected by exogenous sphingosine (data not shown). Further, the enhanced solubility of PLAP from membranes prepared from compactin-treated SPB-1/PLAP cells was partially suppressed when either exogenous sphingosine or cholesterol was supplied to the cells and almost completely suppressed to the wild type level when both exogenous sphingosine and cholesterol were supplied to the cells. Compactin treatment or exogenous lipid supplementation of the cells did not affect expression levels of PLAP activity (data not shown), which was further confirmed by Western blotting (see below). These results demonstrated the additive effects of sphingolipid and cholesterol on the insolubility of PLAP in Triton X-100. On the other hand, solubility of total membrane proteins, which was estimated by recovery of proteins in the supernatant fractions, showed similar levels (50-60%) among sphingolipid- and/or cholesteroldeprived membranes and the control membranes (Fig. 1). Thus, these results indicated that sphingolipid and cholesterol deprivations did not induce nonspecific enhancement of membrane protein solubilization.
Figure 1Restoration of PLAP insolubility in SPB-1/PLAP cell membranes by metabolic complementation. SPB-1/PLAP cells were cultured in Nutridoma-BO medium at 39°C for 3 days in the presence (+) or the absence(-) of compactin, sphingosine, and cholesterol, and membranes were prepared from these cells as described under “Experimental Procedures.” Solubility of PLAP (open bars) and membrane proteins (hatched bars) in Triton X-100 were determined and are represented as the means ± S.D. (n = 3).
Independent Changes in Membrane Sphingolipids and Cholesterol
To know whether membrane sphingolipids and cholesterol were deprived independently, we determined the contents of sphingomyelin, GM3 (the sole ganglioside in CHO cells(
)), and cholesterol in the isolated membranes. As shown in Table 3, the sphingomyelin and GM3 contents of SPB-1/PLAP cell membranes were about 30% of the wild type levels whereas the cholesterol level of sphingolipid-deprived membranes was almost the same as that of wild type control membranes. Conversely, compactin treatment of the cells reduced the cholesterol content by 50-65% without appreciable effects on sphingolipid levels. Supplying exogenous sphingosine to SPB-1/PLAP cells reversed the sphingolipid contents to more than 70% of the wild type levels without affecting the cholesterol content. Even when SPB-1/PLAP cells were cultured in the presence of compactin, supplementation of the cells with both cholesterol and sphingosine restored both the cholesterol and sphingolipid content of the isolated membranes. Cholesterol ester levels were less than 10% of free cholesterol levels in all cases, and the amount of glucosylceramide and lactosylceramide, metabolic intermediates for GM3, were too low (<3 nmol/mg of protein) to be accurately determined (data not shown). These results confirmed that the sphingolipid and cholesterol content of the isolated membranes were reduced under the deprivation conditions and that they were restored by metabolic complementation. Another important point was that the mass levels of membrane sphingolipids and cholesterol changed independently (Table 3), eliminating the possibility that the sphingolipid deprivation might cause reduction of the cholesterol content and thereby enhance the solubility of PLAP or vice versa.
GPI-anchoring of PLAP in Sphingolipid- and Cholesterol-deprived Membranes
Conversion of the GPI-anchor to a membrane-spanning polypeptidyl domain abolished the insolubility of PLAP in Triton X-100, as shown by the difference in the solubility between PLAP and PLAP-HA (Table 2). Thus, it raised the possibility that a certain population of PLAP molecules in sphingolipid- and cholesterol-deprived membranes might be integrated in membranes via a membrane-spanning domain but not the GPI-anchor, increasing the apparent solubility of PLAP activity in Triton X-100. To test this possibility, the sensitivity of PLAP to PI-PLC, which cleaves GPI-anchors(
), was assessed from shifts in hydrophobicity of PLAP. Membranes were incubated with or without PI-PLC under membrane-permeabilizing conditions, and the partitioning of PLAP into Triton X-114 and aqueous phases was subsequently determined. When membranes were not treated with PI-PLC, more than 95% of the activity of CHO-K1/PLAP, SPB-1/PLAP, and compactin-treated SPB-1/PLAP cell membranes partitioned into the Triton X-114 phase, and PI-PLC treatment of these membranes resulted in an almost complete shift of PLAP from the detergent phase to the aqueous phase (Fig. 2). The possibility that the hydrophilic catalytic domain of PLAP was separated by proteases contaminating the PI-PLC sample was ruled out since most of PLAP-HA partitioned into the detergent phase regardless of PI-PLC treatment (Fig. 2). These results indicated that almost all of PLAP molecules were attached to membranes via the GPI-moiety in both the sphingolipid/cholesterol-deprived membranes and in control membranes.
Figure 2PI-PLC sensitivity of PLAP. Membranes prepared from the indicated cells (when indicated, the cells were cultured with compactin) were incubated with (+) or without(-) PI-PLC at 37°C for 1 h and subsequently subjected to partitioning by Triton X-114 phase separation as described under “Experimental Procedures.” Distribution of alkaline phosphatase activity to Triton X-114 and aqueous phases after partitioning are shown as the percentages of the total activity (Triton X-114 plus aqueous fractions). Bars shown are the means ± S.D. (n = 3).
Sphingolipid and Cholesterol Deprivations Did Not Affect the Apparent Molecular Weight of PLAP
Newly synthesized PLAP molecules are initially extractable in Triton X-100 with high efficiency, but subsequently acquire a Triton X-100-insoluble property soon after leaving from the endoplasmic reticulum or the pre-Golgi compartment(
). To examine whether an immature form(s) of PLAP accumulated in sphingolipid/cholesterol-deprived membranes and was selectively solubilized in Triton X-100, we performed Western blot analysis on the supernatant and the precipitate fractions of Triton X-100-treated membranes. As shown in Fig. 3, PLAP of Mr = 66,000 corresponding to the mature form (
) was virtually the only product detected in both fractions regardless of sphingolipid and cholesterol deprivation. Membranes from nontransfected CHO-K1 and SPB-1 cells showed no detectable PLAP signal by Western blotting (data not shown). These results suggested that sphingolipid/cholesterol deprivations did not cause any structural changes of PLAP molecules, ruling out the possibility that the enhanced solubility of PLAP might be due to the accumulation of immature forms of PLAP in sphingolipid/cholesterol-deprived membranes. It should also be noted that the solubility of PLAP molecules estimated by densitometric analysis of the Western blotting data was consistent with that estimated by alkaline phosphatase assays (Fig. 1 and Fig. 3). In addition, the total amount of PLAP in membranes was not affected by compactin treatment or exogenous lipid supplementation of the cells (Fig. 3).
Figure 3Western blotting of PLAP. Cells were cultured under conditions as described in the legend to Table 3 and membranes were prepared from these cells. Membranes incubated with 1% Triton X-100 were subjected to high speed centrifugation, and the supernatant (S) and the precipitate (P) fractions were analyzed by Western blotting using anti-PLAP polyclonal antibody as described under “Experimental Procedures.” Lanes 1 and 2, CHO-K1/PLAP cell membranes; lanes 3-10, SPB-1/PLAP cell membranes. Molecular mass standards used are rabbit muscle phosphorylase B (97 kDa), bovine serum albumin (66 kDa), and hen egg white ovalbumin (45 kDa).
Effects of Sphingolipid/Cholesterol Depletion on Expression of Caveolin and on the Cell-surface Distribution of PLAP
Previous observations that GPI-anchored proteins expressed in Fischer rat thyroid cells, which naturally lack caveolin, are efficiently extracted by Triton X-100 (
) raise the possibility that caveolin is necessary for GPI-anchored proteins to acquire Triton X-100 resistance. Western blot analysis of caveolin in membranes isolated from CHO cells showed that expression of caveolin was not significantly affected by sphingolipids and/or cholesterol depletion (Fig. 4). Thus, it was unlikely that the enhanced solubility of PLAP by sphingolipids and/or cholesterol depletion was due to the secondary effect of reduced expression of caveolin, even if caveolin was an essential component for Triton X-100 insolubility of GPI-anchored proteins.
Figure 4Western blotting of caveolin. Cells were cultured under conditions as described in the legend to Table 3. Membranes (5 μg of protein) prepared from these cells were analyzed by Western blotting using anti-caveolin antibody as described under “Experimental Procedures.” Lanes 1 and 2, CHO-K1/PLAP cell membranes; lanes 3-6, SPB-1/PLAP cell membranes. Molecular mass standards used are carbonic anhydrase (31 kDa) and soybean trypsin inhibitor (21 kDa).
We further examined whether there was a difference in the distribution of PLAP at the cell surface by indirect immunogold labeling electron microscopy. As shown in Fig. 5, A-C, PLAP molecules were almost randomly distributed on the cell surface, regardless of sphingolipid/cholesterol depletion, and no obvious difference in the distribution was observed between CHO-K1/PLAP, SPB-1/PLAP, and compactin-treated SPB-1/PLAP cells. Nontransfected SPB-1 cells showed no signal (Fig. 5D), confirming the specificity of the indirect immunogold labeling of PLAP.
Figure 5Distribution of PLAP at the cell surface. Monolayers of CHO cells were grown in Nutridoma-BO with or without compactin at 39°C for 3 days. The monolayers were incubated with rabbit anti-PLAP polyclonal antibody in Nutridoma-BO medium at 4°C for 1 h, rinsed with PBS, and fixed with PBS containing 0.5% glutaraldehyde and 3% formaldehyde. After incubation with 0.1 M NH4Cl in PBS, the monolayers were incubated with gold-conjugated protein A in Nutridoma-BO for 1 h. The samples were fixed again with 1% formaldehyde and 3% glutaraldehyde, stained with OsO4 and uranyl acetate, dehydrated in ethanol, embedded in Epon, sectioned, and viewed. Bar represents 0.5 μm. A, CHO-K1/PLAP cells; B, SPB-1/PLAP cells; C, compactin-treated SPB-1/PLAP cells; D, SPB-1 cells (PLAP-negative control).
In this paper, we developed CHO cell systems to lower cellular sphingolipids and/or cholesterol, which should be useful for investigating the coordinate roles of these lipids in mammalian membranes. We demonstrated that deprivation of sphingolipids and cholesterol in CHO cell membranes enhanced the solubility of PLAP in Triton X-100 (Table 2), and this enhanced solubility was suppressed by metabolic complementation with exogenous sphingosine and cholesterol (Fig. 1). Moreover, determination of the mass levels of sphingolipids and cholesterol in the isolated membranes revealed independent changes in sphingolipid and cholesterol levels (Table 3), eliminating the possibility that deprivation of sphingolipids might induce a reduction of cholesterol and thereby enhance the solubility of PLAP and vice versa. It was unlikely that the enhanced solubility was due to structural changes in PLAP molecules since, regardless of sphingolipid and cholesterol depletion, almost all the PLAP molecules had the GPI-anchor moiety (Fig. 2) and there were no differences in the apparent molecular weight of the protein in the supernatant and precipitate fractions of the detergent-treated membranes (Fig. 3). Furthermore, our findings that sphingolipid and/or cholesterol depletion did not affect the solubility of total membrane proteins (Fig. 1) and that conversion of the GPI-anchor to a membrane-spanning domain abolished the insolubility of PLAP (Table 2) indicated that sphingolipids and cholesterol conferred the insolubility to a limited set of membrane proteins including GPI-anchored proteins, and therefore ruled out the possibility that the insolubility of PLAP resulted from nonspecific inclusion of membrane proteins into Triton X-100-insoluble sphingolipid/cholesterol aggregates. From these results, we conclude that both sphingolipids and cholesterol are involved in the insolubility of PLAP in Triton X-100 and suggest that these lipids may coordinately play a role in the formation of putative Triton X-100-resistant membrane microdomains where GPI-anchored proteins acquire their insolubility.
Both cholesterol and complex sphingolipids are enriched in the exoplasmic leaflet of the plasma membrane of intact cells(
)). The insolubility of GPI-anchored proteins is observed in glycosphingolipid-poor cells such as CHO cells (this study) as well as glycosphingolipid-abundant cells like Madin-Darby canine kidney cells(
). These previous findings combined with the present study revealing participation of both membrane sphingolipids and cholesterol in the insolubility of PLAP lead us to suggest that an interaction between cholesterol and sphingolipids, especially sphingomyelin, may play a role in the initial formation for Triton X-100-resistant complexes, into which GPI-anchored proteins are subsequently integrated. However, it is still unknown whether additional membrane components are also involved in the formation of Triton X-100-resistant microdomains. Caveolin/VIP21 is an integral membrane protein, which associates with caveolae (noncoated invaginations at the plasma membrane) (
), the enhanced solubility of PLAP by sphingolipids and/or cholesterol depletion was not an indirect effect of reduced expression of caveolin since expression of caveolin was not significantly affected by sphingolipids and/or cholesterol depletion (Fig. 4). Recent findings that myristoylated/palmitoylated non-receptor kinases and multimeric GTP-binding proteins are also enriched in nonionic detergent-resistant complexes (
) raise the intriguing possibility that putative cholesterol/sphingolipid-enriched membrane microdomains play an important role in signal transduction. Model cell systems such as the ones presented here will hopefully be useful in investigating this possibility in the future.
ACKNOWLEDGEMENTS
We thank Deborah Brown, Robert Simoni, and Michael Edidin for gifts of plasmids, compactin, and PI-PLC, respectively. We also thank Ona Martin for critical reading of this manuscript and Mike Sepanski for technical assistance with electron microscopy studies.
81489-2&id=tbl1.jpg){kind=link}
81489-2&id=tbl2.jpg){kind=link}
81489-2&id=gr1.jpg){kind=link}
81489-2&id=tbl3.jpg){kind=link}
81489-2&id=gr2.jpg){kind=link}
81489-2&id=gr3.jpg){kind=link}
81489-2&id=gr4.jpg){kind=link}
81489-2&id=gr5.jpg){kind=link}