Sphingolipid Metabolism in Cultured Fibroblasts: Microscopic And

Proc. Nati Acad. Sci. USA Vol. 80, pp. 2608-2612, May 1983 Cell Biology

Sphingolipid metabolism in cultured fibroblasts: Microscopic and biochemical studies employing a fluorescent ceramide analogue (Golgi apparatus/sphingomyelin/cerebrosides/liposomes/fluorescence) NAOMI G. LPSKY AND RICHARD E. PAGANO Department of Embryology, Carnegie Institution of Washington, 115 West University Parkway, Baltimore, Maryland 21210 Communicated by Harden M. McConnell, December 30, 1982

ABSTRACT A fluorescent analogue of ceramide, N-[7-(4-ni- HI trobenzo-2-oxa-1,3-diazole)]-e-aminocaproyl sphingosine (C6-NBD- ceramide), was used to investigate sphingolipid metabolism in HO-C-H Chinesehamster fibroblasts. C6-NBD-ceramide was incorporated | /~~~~0 into small unilamellar dioleoyl phosphatidylcholine vesicles and N N incubated with cells in monolayer culture at 20C, resulting in rapid 0 /9 and preferential transfer of the labeled ceramide from vesicles to HI ° cells. The cells were then washed and subsequently incubated at H-C-N- C-(CH2)5-N NO2 37°C for various intervals. The metabolism of C6-NBD-ceramide was monitored by lipid extraction and analysis, and the intracellular distribution of the labeled molecule was followed by fluo- H rescence microscopy. Initially, fluorescence was detected almost HO-C-C =C-(CH2)12- CH3 exclusively in mitochondria, with over 90% of the extractable lipid I fluorescence due to C6-NBD-ceramide. After 30 min at 370C, in- H H tense fluorescence. appeared in the Golgi apparatus. This organ- elle was identified by colocalization of NBD fluorescence with a FIG. 1. Structure of C6-NBD-ceramide. Golgi-apparatus-specific stain. At later times the plasma membrane became visibly labeled as well, at which point 90% of the orescent of the cell-associated fluorescence was recovered as NBD-labeled sphin- tracer allows direct microscopic observation gomyelin and NBD-labeled cerebroside. These metabolites were translocation and sequestering of the labeled precursor and its identified by enzymatic and biochemical analysis and by thin-layer metabolites as it occurs in living cells. Thus, artifacts and delays chromatography of the fluorescent lipid extracts. The finding that due to tissue fixation or subcellular fractionation are avoided. C6-NBD-ceramide is used by these cells in standard pathways of Here we report that C6-NBD-ceramide is metabolized by sphingolipid biosynthesis suggests that this fluorescent precursor cultured fibroblasts to fluorescently labeled sphingomyelin and will be a valuable tool for correlating the metabolism of sphin- cerebroside. During this metabolism, the intracellular distri- golipids with their intracellular distribution and translocation. In bution of fluorescence, involving the mitochondria, Golgi ap- addition, during its metabolism by Chinese hamster fibroblasts, paratus, and plasma membrane, also changes in a striking man- this compound acts as a vital stain for the Golgi apparatus. ner. The sphingolipids are important membrane lipids which have METHODS been implicated in a number of cellular properties, including Materials. Chemicals were purchased as indicated: Clostri- blood group specificity, receptor action, nerve conduction, dium perfringens phospholipase C, human placenta sphingo- membrane stability, oncogenic transformation, and aging (for myelinase, taurodeoxycholate, sphingosine-l-phosphocholine, reviews see refs. 1-4). Although all of the biosynthetic path- psychosine, and bovine brain sphingolipids (Sigma); Eagle's ways for the sphingolipids are not yet completely understood minimal essential medium (GIBCO); lectins (Vector Labora- (5-7), it is well known that defects in their metabolism or break- tories, Burlingame, CA); and rhodamine 3B (Eastman). down can have serious consequences, such as Tay-Sachs or Lipids and Lipid Vesicles. N-Rhodamine B Farber diseases (8, 9). All the sphingolipids have in common sulfonyl dioleoyl the ceramide backbone, which is either.glycosylated to yield phosphatidylethanolamine (N-Rh-PtdEtn) was synthesized as cerebrosides and gangliosides or coupled to phosphocholine to described (10). C6-NBD-ceramide, C6-NBD-galactosylceram- give sphingomyelin. The sites of synthesis of these molecules ide, and NBD-sphingomyelin were synthesized (9), using sphin- inside the cell remain to be determined, as do the molecular Abbreviations: NBD, 4-nitrobenzo-2-oxa-1,3-diazole; NBD-sphingo- mechanisms underlying translocation and insertion of the dif- myelin, N-[7-(4-nitrobenzo-2-oxa-1,3-diazole)]aminoacyl sphingosine-1- ferent sphingolipids into different subcellular membranes. phosphocholine; NBD-cerebroside, N-[7-(4-nitrobenzo-2-oxa-1,3-dia- In order to examine the correspondence between sphingo- zole)]aminoacyl-sphingosine monoglycoside; C6-NBD fatty acid, N-[7- lipid metabolism and the intracellular distribution of these lip- (4-nitrobenzo-2-oxa-1,3-diazole)]-6-aminocaproic acid; C6-NBD-cer- ids, we have initiated a study using N-[7-(4-nitrobenzo-2-oxa- amide, N-[7-(4-nitrobenzo-2-oxa-1,3-diazole)]-6-aminocaproyl sphingosine; C6-NBD-galactosylceramide, N-[7-(4-nitrobenzo-2-oxa-1,3-dia- 1,3-diazole)]-6-aminocaproyl sphingosine (C6-NBD-ceramide), zole)]-6-aminocaproyl sphingosine galactoside; N-Rh-PtdEtn, N-rho- a fluorescent derivative of ceramide (Fig. 1). The use of a flu- damine B sulfonyl dioleoyl phosphatidylethanolamine; Oe2-PtdCho, dioleoyl phosphatidylcholine; HMEM, 18 mM Hepes-buffered Eagle's The publication costs ofthis article were defrayed in part by page charge minimal essential medium, pH 7.4, with 0.62 mM phosphate; HMEMB, payment. This article must therefore be hereby marked "advertise- HMEM with choline, ethanolamine, serine, and myo-inositol at0.5 mM ment" in accordance with 18 U.S.C. §1734 solely to indicate this fact. each. 2608 Downloaded by guest on September 27, 2021 Cell Biology: Lipsky and Pagano Proc. Natl. Acad. Sci. USA 80 (1983) 2609

gosine, psychosine, or sphingosyl phosphocholine, respec- ments, filter combinations were chosen that eliminated cross- tively. N-[7-(4-nitrobenzo-2-oxa-1,3-diazole)]-6-aminocaproic acid over between rhodamine and NBD fluorescence channels. (C6-NBD fatty acid) and dioleoyl phosphatidylcholine (Ole2- Thin-Layer Chromatography. Lipid extracts were routinely PtdCho) were purchased from Avanti Biochemicals. Concen- applied to silica gel 60 thin-layer plates (Merck) or, where in- trations of lipid stock solutions were determined by phosphorus dicated, to silica gel G/1% sodium borate plates (Analtech, measurement (11), or, for C6-NBD-ceramide, by reference to Newark, DE). For two-dimensional chromatography, solvent 1 known concentrations of fluorescent standards (see below). was CHC13/CH3OH/H2O, 65:25:4 (vol/vol), and solvent 2 was Small unilamellar vesicles were formed by ethanol injection tetrahydrofuran/methylal/CH30H/H20, 10:6:4: 1 (vol/vol) (20). (12): 12 mol % C6-NBD-ceramide and 88 mol % Ole2-PtdCho One-dimensional plates were routinely developed with solvent (unless otherwise indicated) were mixed, dried under argon, 1. Anothersolventsystem was solvent3, CHC13/CH3OH/H2O/ and then, under reduced pressure, made to 2.8 mM in ethanol NH40H, 72:48:9:2 (vol/vol) (6). Fluorescent lipids were de- and injected while vortex mixing into 13 vol of 10 mM Hepes- tected on plates by UV illumination. The spots were scraped buffered calcium- and magnesium-free Puck's saline. This and the lipids were extracted as described for cell extracts. preparation was dialyzed at 20C overnight against the same Identification of Fluorescent Lipid Metabolites. All pro- buffered saline, then diluted to a final concentration of 22-44 cedures were first tested on appropriate bovine brain sphin- AtM with 18 mM Hepes-buffered Eagle's minimal essential me- golipid standards to ensure their efficacy. Acid hydrolysis was dium, pH 7.4, with 0.62 mM phosphate and choline, ethanol- performed as described (21). Lability to sphingomyelinase was amine, serine, and myo-inositol at 0.5 mM each (HMEMB) for assayed (22, 23) with modifications as follows: 0.25-3.0 nmol of incubations. fluorescent compound were desiccated in a screw-capped test Cell Culture, Vesicle-Cell Incubations, and Lipid Extrac- tube, then 200 A1 of taurodeoxycholate (62 ,ug/ml in 0.1 M po- tion. Monolayer cultures of Chinese hamster V79 fibroblasts tassium acetate, pH 5.0) was added and the solution was briefly (13) were grown to confluency in Eagle's minimal essential me- sonicated. Twenty units of enzyme or an equal volume of buff- dium supplemented with 12% horse serum in a water-saturated er was added, and the mixture was incubated at room tem- atmosphere of 5% CO2 in air. The plates were washed two times perature for 2 hr. Lipids were extracted with 1.2 ml of CHC13/ with Hepes-buffered Eagle's minimal essential medium, pH CH30H, 2:1 (vol/vol). Under these conditions, no contami- 7.4, with 0.62 mM phosphate (HMEM). Monolayers were in- nating phospholipase C activity on natural or fluorescent phos- cubated with vesicles at 20C for 90 min, then washed three times phatidylcholine was detected. Phospholipase C hydrolysis of with HMEM. For time course experiments, the plates were lipid substrates (24) and cerebroside ring oxidation, reduction, subsequently incubated at 370C in HMEMB. Cells were re- and removal (25) were performed. moved by trypsin treatment (14) for counting in a hemacytome- ter or were scraped for extraction, in medium, with a rubber RESULTS policeman, and lipids were extracted (15). Total NBD (excita- Cellular Uptake of C6-NBD-Ceramide from Vesicles. When tion at 470 nm; emission at 530 nm) and rhodamine (excitation cells were incubated at 20C with vesicles containing Ole2-PtdCho, at 555 nm and emission at 583 nm) fluorescence of lipid extracts C6-NBD-ceramide, and the nonexchangeable lipid N-Rh-PtdEtn was measured in an Aminco-Bowman spectrofluorimeter against (10, 14, 26), there was a rapid and preferential uptake of the appropriate fluorescent standards. fluorescent ceramide by the cells. The ratio of NBD to rho- Microscopy. Fibroblasts were grown on glass coverslips coated damine in the cell lipid extract was approximately 10 times that with rat tail collagen. For time course studies, cells were in- of the added vesicles for two different concentrations of C6-NBD- cubated at 20C for 90 min with C6-NBD-ceramide-containing ceramide (Table 1). Thus, NBD uptake by cells was due pri- vesicles, rinsed, and incubated in HMEMB at 37°C for various marily to uptake of C6-NBD-ceramide molecules rather than intervals. For mitochondrial colocalization studies, incubations mere association of intact vesicles with cells. If the latter had with C6-NBD-ceramide-containing vesicles at 20C for 90 min been the case, the ratio of NBD to rhodamine in the cells would were followed by 2-min incubations at 2°C with rhodamine 3B have been the same as that of the initial vesicle suspension (27). at 0.1 ,g/ml (16). The cells were washed and photographed, Metabolism of C6-NBD-Ceramide. When cells were incu- first with optics appropriate for viewing rhodamine fluores- bated with C6-NBD-ceramide-containing vesicles for 90 min at cence and then under conditions appropriate for NBD fluo- 20C, washed, and incubated at 37°C for 1 hr, three major flu- rescence (see below). For Golgi colocalization studies, cells were orescent products appeared in the lipid extracts. Fig. 2 shows incubated with C6-NBD-ceramide-containing vesicles for 90 min a two-dimensional chromatogram of such an extract, with the at 2°C, rinsed, incubated in HMEMB for 20 min at 37C, ex- fluorescent spots "C,' "Cb," and "Sm" labeled. The three amined, and photographed under NBD fluorescence condi- products were identified as C6-NBD-ceramide, NBD-cerebro- tions. A modified Golgi-apparatus-staining procedure (17) was then followed: Cells were incubated 90 at 20C with con- min Table 1. Transfer of fluorescent lipids from vesicles to cells canavalin A at 76 and rinsed with HMEM. re- ,g/ml The at 20C maining steps were performed at room temperature. Cells were fixed (18) for 30 min, then incubated 5 min with HMEM, 5 min C6-NBD-ceramide/N-Rh- with 0.1 M glycine, and 5 min with HMEM. Cells were made PtdEtn, mol/mol permeable by 3-min incubation with 0.1% Triton X-100 in Applied Lipid, pmol per 106 cells phosphate-buffered saline containing 1 mM calcium and mag- Exp. vesicles Cell extracts N-Rh-PtdEtn C6-NBD-ceramide then rinsed nesium, for 15 min with three changes of HMEM. A 0.9 5.9 ± 0.2 1.7 + 0.2 10.4 ± 2.1 Cells were then incubated for 20 with min rhodamine-conju- B 4.4 46.7 ± 11.2 1.3 0.2 59.2 ± 3.8 gated wheat germ agglutinin (25 ,g/ml of phosphate-buffered ± saline with calcium and magnesium), rinsed for 15 min in Monolayer cultures ("3 x 106 cells per dish) were incubated in tri- HMEM, fixed for 20 min, and rinsed for 15 minin HMEM. plicate in HMEMB at 2°C for 90 min with vesicles composed of N-Rh- again PtdEtn/C6-NBD-ceramide/Ole2-PtdCho at mol ratios of 1.6:1.5:96.9 The previously photographed field was reexamined and pho- (Exp. A) or 1.9:8.4:89.7 (Exp. B) for a final C6-NBD-ceramide concen- tographed with optics appropriate for rhodamine fluorescence. tration of 0.50 or 2.35 pM, respectively. Cells were then washed and A Zeiss Universal microscope equipped with epi-illumina- fluorescent lipids were extracted and measured. Values are the mean tion for fluorescence was used (19). For colocalization experi- + SD of three determinations. Downloaded by guest on September 27, 2021 2610 Cell Biology: Lipsky and Pagano Proc. Natl. Acad. Sci. USA 80 (1983)

C,)

LL -j

- < 50 Sm

x Cb

2 0 FIG. 2. Fluorescent metabolites of C6-NBD-ceramide. Monolayer cultures ofChinese hamster fibroblasts were incubated with C6-NBD- ceramide-containing vesicles for 90 min at 20C, then washed and incubated in HMEMB at 370C for 1 hr. Fluorescent lipids were extracted C and subjected to two-dimensional thin-layer chromatography and the chromatogram was photographed under UV light. The respective Rfs in solvent 1 and solvent 2 are C, 70 and 83; Cb, 50 and 83; and Sm, 5 0 1 2 and 4, corresponding to C6-NBD-ceramide, NBD-cerebroside, and NBD- HOURS AT 370C sphingomyelin (see text). FIG. 3. Levels of C6-NBD-ceramide metabolites. Monolayer cul- side, and NBD-sphingomyelin, respectively (Table 2). After acid tures ofChinese hamsterfibroblasts were incubated with C6-NBD-cer- hydrolysis, 90-100% of the fluorescence of each of the three amide-containing vesicles for 90 min at 20C, rinsed, and incubated in HMEMB at 370C for the indicated times. Fluorescent lipids were ex- was as a its spots recovered mixture of C6-NBD fatty acid and tracted at each point, and equal amounts of fluorescence were applied methyl ester, indicating that the fluorescence of all the com- to silica gel 60 chromatographic plates. Plates were developedwith Sol- pounds was due to C6-NBD fatty acid. The appearance of free vent 1, and fluorescence of individual spots was quantified. The per- fatty acid and its methyl ester are expected after acid hydrolysis cent of total fluorescence due to each metabolite during a typical time of sphingolipids. When all three fluorescent products were course is shown: A, C6-NBD-ceramide; e, NBD-cerebroside; *, NBD- subjected to periodate oxidation followed by sodium borohy- sphingomyelin. dride reduction and acid hydrolysis, a procedure that results in Further, NBD-cerebroside was not affected by galactose oxi- the removal of the sugar residue of cerebrosides (25), only Cb dase under conditions in which the synthetic C6-NBD-galac- was altered. This procedure transformed Cb into a fluorescent was altered Spot C was identified compound that had the chromatographic properties of C6-NBD- tosylceramide (not shown). as C6-NBD-ceramide by its migration with the standard in all ceramide; thus, Cb was identified as NBD-cerebroside. This solvent systems. Finally, spot Sm was identified as NBD-sphin- NBD-cerebroside was tentatively defined as glucocerebroside, gomyelin on the basis of both its migration with an NBD-sphin- on its behavior on borate-treated silica gel thin-layer the basis of a of solvents and its to plates (28). In solvent 3, bovine brain galactocerebroside and gomyelin standard in variety lability sphingomyelinase or phospholipase C. Either enzyme altered C6-NBD-galactosylceramide are reduced in mobility synthetic the starting material to a fluorescent compound having the on borate-treated plates. In contrast, glucocerebroside has in- chromatographic properties of C6-NBD-ceramide, whereas creased mobility. NBD-cerebroside manifested an increase in had no effect on C6-NBD-ceramide or NBD- similar to that of (data not shown). sphingomyelinase mobility glucocerebroside cerebroside (phospholipase C activity towards these was not Table 2. Identification of fluorescent metabolites of determined). C6-NBD-ceramide Levels of NBD-Metabolites. The relative amounts of fluorescent ceramide and its two metabolites in lipid extracts were Rfof fluorescent products Procedure C Cb Sm Table 3. Change in total extractable fluorescence over time Lipid extract 72 51 5 % total fluorescence After acid hydrolysis 54, 80 54, 80 54, 80 remaining after After sphingomyelinase incubation at 370C treatment 72 51 72 Temp., 0 0.5 1 2 After phospholipase C 0C Extract hr hr hr hr treatment - - 72 After sugar ring cleavage 72 72 5 2 Cell lipids 100 - 98 90 37 Cell lipids 100 58 54 32 Monolayers were incubated with C6-NBD-ceramide-containing ves- 37 Cell lipids plus 100 52 52 35 icles, total lipids were extracted, and fluorescent lipids were isolated. incubation medium Fluorescent lipids were then subjected to one of the procedures indicated (see text), and the reaction products were separated by chro- Monolayer cultures of Chinese hamster fibroblasts were incubated matography on silica gel-60 in at least two solvent systems. Typical Rfs with C6-NBD-ceramide-containing vesicles for 90 min at 20C, rinsed, offluorescent products in solvent 1 are shown. Under these conditions, and further incubated in HMEMB for the indicated times and tem- authentic C6-NBD-ceramide and C6-NBD fatty acid and its methyl es- peratures. Lipids were extracted and total fluorescence was measured. ter had Rfs of 72, 54, and 80, respectively. Values at 370C are the means of two independent determinations. Downloaded by guest on September 27, 2021 Cell Biology: Lipsky and Pagano Proc. Nati Acad. Sci. USA 80 (1983) 2611

FIG. 4. Intracellular distribution of fluorescence over time. Monolayer cultures of Chinese hamster fibroblasts were incubated with C6-NBD- ceramide-containingvesicles for90 min at 20C, rinsed, and incubated at 37TC in HMEMB for 0 min (A), 15 min (B), 30 min (C), or 1 hr (D). In D, the plane of focus is at the plasma membrane level. (Bar represents 5 ,um.) quantified after the incubation of monolayers with vesicles for membrane could be clearly observed by focusing in a plane be- 90 min at 20C, rinsing, and subsequent incubation at 37C in yond the nucleus (Fig. 4D). HMEMB for various times. Total fluorescence was determined To identify highly labeled structures, colocalization studies in the lipid extracts at each time point, then equal amounts of were performed with stains known to be specific for different fluorescent extracts from each point were subjected to thin-layer subcellular organelles. After incubation with C6-NBD-ceram- chromatography. The percent of the total recovered fluores- ide at 20C, cells were briefly labeled with rhodamine 3B, a mi- cence due to each sphingolipid at each time point is shown in tochondrial marker (16). Fig. 5 demonstrates the remarkable Fig. 3. Within 1 hr after the shift to 37°C, C6-NBD-ceramide coincidence of staining provided by these labels and indicates declined from about 90% of the total fluorescence to about 10%. that most of the NBD fluorescence was located in the mito- Both NBD-cerebroside and NBD-sphingomyelin increased to chondria. The independent behavior of these two stains and the about 45% of the total within this time period. The relative lev- lack of crossover between the separate fluorescence channels els at 2 hr remained unchanged at 4 hr (data not shown). were confirmed with appropriate controls. In addition, the ap- The amount of total extractable NBD fluorescence found in pearance of mitochondria stained with either fluorophore was cultures over this time course declined from 100% after 90 min similar to that seen with Janus green (29), a classical mito- at 20C to approximately 50% after further incubation for 1 hr chondrial vital stain (not shown). at 370C (Table 3). Furthermore, the lost fluorescence could not Fig. 6 demonstrates that the bright punctate fluorescence be recovered by extraction of the medium. This decline was not seen at 370C (corresponding to Fig. 4 B and C) was located in seen when cultures were maintained at 20C. Varying the ex- the Golgi apparatus. This fluorescence colocalizes with that ob- traction procedure, or examining the aqueous phases, did not tained in the same preparation after labeling with rhodamine- result in increased fluorescence recovery (not shown). These conjugated wheat germ agglutinin, a lectin specific for the Golgi results suggest that one or more of the labeled metabolites was apparatus (17). The same rhodamine pattern was obtained in rendered nonfluorescent at 37C. the presence or absence of NBD, indicating that this lectin was Intracellular Distribution of Fluorescence. Fig. 4 demon- not merely binding to NBD-sphingolipid (data not shown). strates the appearance of labeled fibroblasts during the time course described above. Immediately after incubation with C6- DISCUSSION NBD-ceramide at 2°C (Fig. 4A), NBD fluorescence was largely This study has shown that a fluorescent derivative of the sphin- localized in cytoplasmic structures. After 10-30 min at 37C golipid precursor ceramide was taken up by cultured fibroblasts (Fig. 4B), bright areas of punctate fluorescence appeared, ad- from phospholipid vesicles. The ceramide analogue was con- jacent to, or over, the nucleus. At later times, the plasma mem- verted to fluorescent derivatives of cerebroside and sphingo- brane also became highly labeled. Although the total cell-as- myelin, products expected from the known pathways (3) of sociated fluorescence decreased after 30-60 min (see above), sphingolipid metabolism. The fluorescent label allowed direct the labeling pattern was maintained (Fig. 4C). The plasma observation of the intracellular distribution of the analogue and

EAt _ FIG. 5. Localization of NBD fluorescence to mitochondria. Monolayer cultures were incu- F bated with C6-NBD-ceramide-containing vesicles for 90 min at 200, washed, incubated with rhodamine 3B at 0.1 ug/ml for 3 min at 2°C, washed, and examined in the fluorescence microscope. (A) NBD fluorescence, (B) rhodamine 3B fluorescence in the same cells. (Bar represents 5 ,m.) Downloaded by guest on September 27, 2021 2612 Cell Biology: Lipsky and Pagano Proc. Natl. Acad. Sci. USA 80 (1983)

FIG. 6. ocalization ofNBD fluorescence to Golgi apparatus. (A) Monolayer cultures were incubated with C6-NBD-ceramide-containing vesicles for 90 min at 20C, rinsed, incubated in HMEMB at 37TC for 20 min, and examined for NBD fluorescence. (B) The same cells, examined for rhodamine fluorescence, after staining with rhodamine-conjugated wheat germ agglutinin. (Bar represents 5 pn.)

its metabolites within living cells and thus permitted a unique 5. Stoffel, W. & Melzner, I. (1980) Hoppe-Seyler's. Z. PhysioL Chem. insight into ceramide metabolism as it occurred. 361, 755-771. 6. Bernert, J. T. & Ullman, M. D. (1981) Biochim. Biophys. Acta 666, Our findings that C6-NBD-ceramide was metabolized by cells 99-109. along expected pathways and that it and its products were trans- 7. Marggraf, W.-D., Anderer, F. A. & Kanfer, J. N. (1981) Biochim. located within cells strongly suggest that the NBD label did not Biophys. Acta 664, 61-73. drastically interfere with the cell's ability to metabolize the cer- 8. Brady, R. 0. (1978) Annu. Rev. Biochem. 47, 687-713. amide moiety. This is consistent with recent studies (19, 30), 9. Chen, W W, Moser, A. B. & Moser, H. W. (1981) Arch. Biochem. which used an NBD-labeled derivative of phosphatidic acid to Biophys. 208, 444-455. a 10. Struck, D. K., Hoekstra, D. & Pagano, R. E. (1981) Biochemistry show that it, too, is metabolized to yield characteristic set of 20, 4093-4099. fluorescent products, with a pattern of intracellular labeling 11. Ames, B. N. & Dubin, D. T. (1960)J. Biol Chem. 235, 769-775. very different from that reported here. The results reported 12. Kremer, J. M. H., v. d. Esker, M. W. J., Pathmamanoharan, C. here appear to be specific to the ceramide base, rather than the & Wiersema, P. A. (1977) Biochemistry 16, 3932-3935. NBD moiety, and are consistent with C6-NBD-ceramide acting 13. Ford, D. K. & Yerganian, G. (1958)J. NatI Cancer Inst. 21, 393- as a reliable functional analogue for ceramide. 425. 14. Struck, D. K. & Pagano, R. E. (1980) J. Biol Chem. 255, 5404- A significant finding of this study is that, as C6-NBD-cer- 5410. amide was metabolized during incubations at 37TC, the intra- 15. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol 37, cellular distribution of fluorescence changed in a gradual and 911-917. progressive fashion, first involving mitochondria, then the Golgi 16. Johnson, L. V., Walsh, M. L., Bockus, B. J. & Chen, L. B. (1981) apparatus, and later the plasma membrane as the dominantly J. Cell Biol 88, 526-535. labeled structures in the cell. Although this redistribution of 17. Virtanen, I., Ekblom, P. & Laurila, P. (1980)J. Cell Biol. 85, 429- to with the of 434. fluorescence seemed correlate directly synthesis 18. Fambrough, D. M. & Devreotes, P. N. (1978) J. Cell BioL 76, 237- NBD-sphingomyelin and NBD-cerebroside, we do not as yet 244. know the precise subcellular location of each of these NBD-la- 19. Pagano, R. E., Longmuir, K. J., Martin, 0. C. & Struck, D. K. beled compounds. However, because many glycosyltransfer- (1981)1. Cell Biol 91, 872-877. ases (31-34) are located in the Golgi apparatus, it seems rea- 20. Robbins, P. W & Macpherson, I. A. (1971) Proc. R. Soc. London sonable to expect that the NBD-cerebroside was synthesized Ser. B 177, 49-58. Be- 21. Poulos, A., Hann, C., Phillipou, G. & Pollard, A. C. (1979) AnaL there prior to translocation to the cell surface or elsewhere. Biochem. 97, 323-327. cause NBD-sphinogmyelin was present in lipid extracts in sig- 22. Yedgar, S. & Gatt, S. (1980) Biochem. J. 185, 749-754. nificant amounts prior to intense labeling of the cell surface, 23. Pentchev, P. G., Brady, R. O., Gal, A. E. & Hibbert, S. (1977) it can be argued that it is first synthesized intracellularly (4, 6, Biochim. Biophys. Acta 488, 312-321. 34, 35), then transported to the cell surface. Plasma membrane 24. Bergelson, L. D., ed. (1980) Lipid Biochemical Preparations (El- labeling became most conspicuous after about 30 min at 37TC, sevier/North-Holland Biomedical, Amsterdam), p. 228. 25. Carter, H. E., Rothfus, J. A. & Gigg, R. (1961) J. Lipid Res. 2, consistent with the length of time reported for an isotopically 228-234. labeled neuronal ganglioside (32), to appear at the plasma mem- 26. Pagano, R. E., Martin, 0. C., Schroit, A. J. & Struck, D. K. (1981) brane. However, is is also possible that fluorescent labeling of Biochemistry 20, 4920-4927. the plasma membrane occurred at much earlier times but was 27. Pagano, R. E., Schroit, A. J. & Struck, D. K. (1981) in Liposomes: obscured by the very bright fluorescence of the cytoplasm. From Physical Structure to Therapeutic Applications, ed. Knight, Consistent with this possibility are recent reports by Marggraf C. G. (Elsevier/North-Holland Biomedical, Amsterdam), pp. 323- 348. et al. (7), and Voelker and Kennedy (36), which suggest that the 28. Kishimoto, Y. (1978) in Research Methods in Neurochemistry, eds. plasma membrane is the primary site of synthesis of sphin- Marks, N. & Rodnight, R. (Plenum, New York), Vol. 4, pp. 411- gomyelin. Studies to detect the appearance with time of NBD- 436. sphingomyelin at the cell surface may resolve this issue. 29. Lazarow, A. & Cooperstein, S. J. (1953) Exp. Cell Res. 5, 56-97. Finally, in addition to providing a means for investigating 30. Pagano, R. E., Longmuir, K. J. & Martin, 0. C. (1983) J. Biol sphingolipid metabolism and translocation in cells, fluorescent Chem. 258, 2034-2040. 31. Keenan, T. W., Morre, D. J. & Basu, S. (1974)J. Biol Chem. 249, ceramide may be useful as a vital stain for the Golgi apparatus. 310-315. This work was supported by a grant from the Whitehall Foundation 32. Miller-Podraza, H. & Fishman, P. H. (1982) Biochemistry 21, 3265- and by U.S. Public Health Service Grant GM22942. 3270. 33. Tartakoff, A. M. (1980) Int. Rev. Exp. Pathol. 22, 227-251. 1. Morell, P. & Braun, P. (1972) J. Lipid Res. 13, 293-310. 34. Fleischer, B., Zambrana, F. & Fleischer, S. (1974) J. Supramol. 2. Hakomori, S. (1981) Annu. Rev. Biochem. 50, 733-764. Struct. 2, 737-750. 3. Kishimoto, Y. & Kawamura, N. (1979) MoL Cell Biochem. 23, 17- 35. Ullman, M. D. & Radin, N. S. (1974) J. Biol Chem. 249, 1506- 25. 1512. 4. Barenholz, Y. & Thompson, T. E. (1980) Biochim. Biophys. Acta 36. Voelker, D. R. & Kennedy, E. P. (1982) Biochemistry 21, 2753- 604, 129-158. 2759. Downloaded by guest on September 27, 2021