Atherosclerosis,
95
84 (1990) 95-100
Elsevier Scientific Publishers Ireland, Ltd. ATHERO 04523
Potential problems in the use of commercial preparations of radiolabeled cholesterol Florence H. Mahlberg, Annabelle Rodriguez-Oquendo, David W. Bernard, Jane M. Glick and George H. Rothblat Department
of Physiologyand Biochemistry, The Medical College of Pennsylvania, 3300 Henry Avenue, Philadelphia, PA 19129 (U.S.A.) (Received 29 November, 1989) (Revised, received 1 May, 1990) (Accepted 4 May, 1990)
Through a series of biological and analytical procedures, we demonstrate that a compound purchased from a commercial supplier as [7-3H]cholesterol was not cholesterol. In mouse peritoneal macrophages, this compound was metabolized differently than other radiolabeled cholesterol preparations and was accumulated in the steryl ester pool. In contrast, Fu5AH rat hepatoma cells did not discriminate this compound from cholesterol. Further analysis of the anomalous [7-3H]cholesterol by TLC after cholesterol oxidase treatment and by HPLC indicated that this radiochemical was less polar than cholesterol standard and other radiolabeled cholesterol preparations tested. Mass spectrometry analysis disclosed that the chemical has a similar fragmentation pattern and the same molecular weight (386) as cholesterol.
Key words: Cholesterol; Radiolabeled cholesterol; Commercial cholesterol preparations
Introduction
Both in vitro and in vivo studies on the metabolism of cholesterol are often dependent on the use of either 14C- or 3H-radiolabeled cholesterol. Although it is accepted practice to repurify these compounds to free them of autooxidation products prior to use, the investigator is
Correspondence to: Dr. F.H. Mahlberg, Department of Physiology and Biochemistry, The Medical College of Pennsylvania, 3300 Henry Avenue, Philadelphia, PA 19129 (U.S.A.)
0021-9150/90/$03.50
largely dependent on the manufacturers to ensure the quality of the radiolabeled compounds which they supply. In the course of cell culture studies designed to investigate aspects of cholesterol metabolism, we have observed anomalies in the intracellular metabolism of cholesterol which led us to suspect that a product which we were using, supplied as [7- 3H]cholesterol, was not cholesterol. Through the use of a series of biological, enzymatic and chromatographic procedures we demonstrated that this was indeed the case and that the metabolism of this compound was significantly different from that of cholesterol, depending on the cell type under investigation.
0 1990 Elsevier Scientific Publishers Ireland, Ltd.
Materials and methods
Cholesterol preparations The various preparations of radiolabeled cholesterol were purchased from commercial suppliers, designated Supplier 1 to Supplier 4. The cellular and biochemical analyses described in this study were initiated upon receipt of the labeled cholesterol from the suppliers. All radioactive compounds were stored in toluene at -2O’C. The radiolabeled cholesterol treated with cholesterol oxidase or analyzed by HPLC was used as supplied by the manufacturers. The cholesterol preparations used for cell culture studies were purified on silica gel glass plates developed in diethyl ether. Radiolabeled bands were eluted from the silica gel [l] and used within 1 week of purification. Cells
All media were buffered with 24 mM bicarbonate and contained gentamicin (50 pg/ml). Thioglycollate-elicited peritoneal macrophages were isolated from B6C3Fl mice. Four days prior to collection, sterile thioglycollate broth (10%) was injected intraperitoneally. Following cervical dislocation, peritoneal lavage was performed using 5 ml sterile phosphate-buffered saline (PBS) containing heparin (2 units/ml) per mouse. The cells were collected by low speed centrifugation and suspended in RPMI-1640 (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Sigma). Two hours after plating, the dishes were rinsed 5 times with serum-free medium, and the adherent cells were used for experiments. FuSAH rat hepatoma cells were grown as previously described [l] and were used 2 days after plating. Loading of both types of cells with labeled cholesterol was accomplished by incubating cell monolayers for 24 h in medium supplemented with 1% heat inactivated FBS, 1% bovine serum albumin (BSA) and either 50 pg protein/ml LDL for the hepatoma cells, or 50 I-18 protein/ml acetylated LDL (acLDL) for the macrophages. Ethanol solutions of the various preparations of [ 3H]cholesterol or [ “C]cholesterol (as indicated in the text) were added to the FBS prior to the dilution of the serum into the medium.’ Both [3H]cholesterol and [‘4C]cholesterol were added to
provide a final concentration of 0.5 pCi/rnl, and the final concentration of ethanol was 0.1%. All media were incubated for 24 h at 37O C prior to use in the cell culture studies. Following a 24-h incubation period in the presence of the loading media, cell monolayers were washed 5 times with serum-free medium and replicate dishes were harvested for determination of the specific activities of the cellular free and esterified cholesterol. Analytical methods Lipids were extracted from media using the method of Bligh and Dyer [2] and from washed monolayers using isopropanol as previously described [3]. Free and total cholesterol mass determinations were conducted by gas liquid chromatography using coprostanol as an internal standard [3]. The distribution of radiolabel between free and esterified cholesterol was determined by thin-layer chromatography on ITLCSA plates (Gelman Sciences) developed in petroleum ether/diethyl ether/acetic acid (85 : 15 : 1, v/v). 3H and 14C were measured by liquid scintillation counting in a Beckman LS7500 using Scintiverse counting fluid. Protein concentrations were determined by the procedure of Markwell et al. [4] using BSA as a standard. Cholesterol oxidase assay [*4C]Cholesterol and [ 3H]cholesterol preparations, as indicated in the text, were incubated with cholesterol oxidase (EC1.1.3.6), and the reaction products were then compared following TLC analysis. For each assay, a mixture of 0.5 PCi radiolabeled cholesterol, 200 pg unlabeled cholesterol (Sigma) and 200 pg egg phosphatidylcholine (Sigma) was dried under nitrogen. This lipid mixture was then dissolved in 50 ~1 of ethanol, after which 0.8 ml PBS was added to the tube. To start the reaction, 20 IU cholesterol oxidase (Beckman) dissolved in 0.31 M sucrose buffered with 0.5 mM sodium phosphate, pH 7.5, were added to the solubilized lipids. The reaction mixture was incubated at 37” C for 1.5 h, and the lipids were then extracted by the method of Bligh and Dyer [2]. The extracted lipids were separated on Bakerflex silica gel TLC plates developed for 20 cm in hexane/diethyl ether/methanol (98 : 7.5 : 15, v/v). The radiolabeled compounds Were identified and
97 quantitated by scanning the TLC plates with a Radiomatic Thin Layer scanner. Unlabeled cholesterol and cholestenone (Sigma) used as standards were identified by iodine staining. HPLC analysis
Reversed-phase HPLC analysis of the radiolabeled sterols was performed using a 25 cm X 4.6 mm Isco ODS (C18) column with methanol as the mobile phase at a flow rate of 1.5 ml/min. Radioactivity in the column effluent was detected using a Radiomatic Flo-one A200 radioactive flow detector. Mass spectrometry
analysis
The major component of 3H compound from Supplier 1 was isolated by HPLC and analyzed on a Finnigan-MAT TSQ70B mass spectrometer in the electron impact mode (EI, 70 ev). Results and discussion
Table 1 shows the results of a study in which mouse peritoneal macrophages and Fu5AH rat hepatoma cells were cultured in media containing various preparations of radiolabeled cholesterol. The data are presented as the ratio of the specific activity of esterified cholesterol to that of the free cholesterol in the cells after a 24-h labeling period. If all the intracellular pools of free and esterified cholesterol reached equilibrium, the esterified to free cholesterol specific activity ratio should approximate 1. When macrophages were incubated TABLE 1 RATIOS OF ESTERIFIED TO FREE CHOLESTEROL SPECIFIC ACTIVITIES IN CELLS EXPOSED TO RADIOLABELED CHOLESTEROL Data are expressed as the ratios of the specific activities of the cellular esterified and free cholesterol, determined at the end of a 24-h labeling period. Values are the mean f SD (n = 3). Cell
Macrophage Fu5AH
Specific activity ratio Esterified cholesterol/free
cholesterol
[7-3H]Chclesterol (Supplier 1)
[4-i4C]Cholesterol (Supplier 4)
[1,2-3H]Cholesterol (Supplier 4)
2.7kO.6 0.8kO.l
0.9f0.2 0.6kO.l
1.2*0.2 0.6kO.l
in medium containing [7-3H]cholesterol from Supplier 1, the ratio of the specific activities of esterified to free cholesterol was 2.7. A number of hypothetical models invoking the selective utilization of cholesterol pools could be suggested to explain how the cellular cholesteryl ester could acquire a specific activity greater than the cellular free cholesterol that serves as its precursor. However, since it has been demonstrated that macrophages have an active cholesteryl ester cycle [5,6] this specific activity ratio after 24 h loading is somewhat surprising. Moreover, when similar experiments were performed using [4-‘4C]cholesterol or [1,2- 3H]cholesterol from Supplier 4, specific activity ratios in all cholesterol pools approximated 1 (Table 1). In addition, the divergence in specific activities of cholesterol pools observed in macrophages with the Supplier 1 [7-3H]cholesterol was not observed in the FuSAH hepatoma cells; as shown in Table 1, specific activity ratios of the cholesterol pools were similar in this cell line regardless of the source of radiolabeled cholesterol. To further probe the reason for the unexpected specific activity data obtained with macrophages labeled with [7-3H]cholesterol from Supplier 1, the uptake and cellular metabolism of that preparation and [1,2-3H]cholesterol from Supplier 4 were compared to that of [4-‘4C]cholesterol from Supplier 4. In these experiments hepatoma cells and macrophages were exposed to media containing mixtures of the 3H and 14C compounds, and the “C/3H ratio of the total, free and esterified cholesterol were determined after a 24-h incubation period. The results from such an experiment are presented in Table 2 and are normalized to the isotope ratio of the starting incubation media which was set equal to 1.0. Any significant divergence from 1.0 indicates a selective utilization of either the [14C]- or [ 3H]cholesterol. Such divergent ratios were clearly evident in both the free and esterified cholesterol extracted from macrophages exposed to the mixture of [‘4C]cholesterol (Supplier 4) and [7-3H]cholesterol (Supplier 1). The isotope ratio was increased in the case of the free cholesterol and reduced in the esterified cholesterol. However, the isotope ratio calculated for the total cellular cholesterol isolated from this macrophage preparation was similar to that of the starting medium. If it is assumed that the metabo-
98 TABLE 2 COMPARATIVE UPTAKE AND METABOLISM OF [‘4C]CHOLESTEROL AND [3H]CHOLESTEROL BY MACROPHAGES AND FuSAH CELLS Data are expressed as the [‘4C]/[3H] ratios of the ceIlular total (TC), free (FC) and esterified (EC) cholesterol radioactivity, determined after a 24-h labeling period and normalized to that of the incubation medium, which was set equal to 1.0. Values arethemeanfSD(n=3). 14C/‘H TC
ratio FC
EC
[4-‘4C]cholesterol (Supplier 4) and [7-3H]cholesterol (Supplier 1) 1.1 kO.1 1.5 *0.1 Macrophage 0.8fO.l Fu5AH 1.0*0.0
0.6 f 0.1 1.1* 0.1
[4-14C] and [1,2-3Hjcholesrerol (Supplier 4) 1.0fO.l 0.9iO.l Macrophage 1.0fO.l 1.0fO.l Fu5AH
0.9fO.l 0.9rtO.l
lism of the [‘4C]cholesterol accurately reflected the metabolism of cholesterol, then the data presented in Table 2 indicate that there was no selective uptake of the Supplier 1 [7-3H]cholesterol by the macrophages, but that there was an accumulation of the [7-3H]sterol in the cellular steryl ester pool. This observation can reflect either a selective esterification of the compound from Supplier 1 or a failure to hydrolyze its ester form. In contrast to the data obtained with the macrophage cultures, the rat hepatoma cell line FuSAH exhibited no selective utilization of the [7-3H]cholesterol from Supplier 1. In addition, when both cell types were grown in media supplemented with a mixture of [4-‘4C]cholesterol and [1,2-3H]cholesterol (both from Supplier 4), the isotope ratios of the cellular total, free and esterified cholesterol were similar to that of the starting incubation media, demonstrating that these two labeled compounds were metabolized by the cells in an identical manner. Since the cell studies demonstrated that the [7- ‘HIcholesterol from Supplier 1 selectively accumulated in the steryl ester pool in macrophages, we subjected different radiolabeled cholesterol preparations to further study in order to establish if the radiolabeled sterol from Supplier 1 was indeed cholesterol. The radioactivity scans of thin-layer chromatograms of the [4-14C]cholesterol from Supplier 3 (Fig. 1, panel A) and [7-
3H]cholesterol from Supplier 1 (panel B) developed in a solvent system that separates a number of sterol oxidation products from cholesterol demonstrated that the two preparations had a similar mobility with a Rf value of 0.35. However, thinlayer chromatography of the reaction products formed upon incubation of the compounds with cholesterol oxidase showed that 90% of the oxidized [7-3H]cholesterol from Supplier 1 (Fig. 1, panel D) had a mobility (R, = 0.56) greater than that of known cholestenone, the reaction product of cholesterol oxidase on cholesterol (R, = 0.50). The oxidized [4-‘4C]cholesterol from Supplier 3 (Fig. 1, panel C) migrated like cholestenone with R, = 0.50. [1,2-3H]Cholesterol from Supplier 3 and [7-3H]cholesterol from Supplier 2, also analyzed by TLC, showed migration characteristics identical to [4-‘4C]cholesterol from Supplier 3 and known standards before oxidation, and to cholestenone after oxidation (data not shown). HPLC analysis also demonstrated that [73H]cholesterol from Supplier 1 did not chromatograph with the same retention time as the other labeled cholesterol preparations or with a known standard. As shown on the HPLC elution profiles (Fig. 2), more than 90% of the compound from Supplier 1 was eluted after 12.7 min (panel B) which is longer than the retention time of 11.5 min observed for [4-‘4C]cholesterol from Supplier 4 (panel A) and known cholesterol standard. For each of the other preparations of radiolabeled cholesterol analyzed by HPLC (Supplier 3 [4-14C], Supplier 3 and Supplier 2 [7-3H], Supplier 3 and 1001
501
I
h
Ali
I
I
A
cl
Fig. 1. TLC migration profiles (see Materials and Methods) of [4-‘4C]cholesterol from Supplier 3 before (A) and after (C) cholesterol oxidase treatment, and [7-‘HIcholesterol from Supplier 1 before (B) and after (D) cholesierol oxidase treatment.
99
Elutlon
ttme (men)
Fig. 2. HPLC elution profiles (see Materials and Methods) of [4-‘4C]cholesterol from Supplier 4 (A) and [7-‘HIcholesterol from Supplier 1 (B).
Supplier 4 [1,2-3H]cholesterol), more than 90% of the radioactivity had a retention time of 11.5 min, similar to known cholesterol (data not shown). Both TLC and HPLC data clearly indicated that the [7-3H]cholesterol obtained from Supplier 1 was not cholesterol, but rather a closely related compound that had TLC-chromatographic properties similar to that of cholesterol. Mass spectrometry analysis of the compound purchased from Supplier 1 indicated that this compound had the same molecular weight (386) and general fragmentation as cholesterol. The specific activity of the material was such that approximately 30% of the molecules were tritiated and the tritium label was apparent in the more intense ion clusters. Thus, the MS data coupled with the biological and chromatographic data indicate that the compound from Supplier I is probably an isomer of cholesterol. Epimerization of the hydroxyl group at position three and/or migration of the double bond from 5,6 to 4,5 or 7,8 are reasonable possibilities. However, reference spectra for these compounds do not differ significantly enough to allow unequivocal assignment by comparison with our rather low intensity spectrum obtained on the small amount (2 ng) of HPLC-isolated material. Studies by DiBussolo et al. [7] have shown that on a reverse phase Cl8 HPLC column, the retention times of 5,3acholestenol (epicholesterol) and of
4,3fi-cholestenol (allocholesterol) are shorter than that of 5,3/3-cholestenol (cholesterol), whereas retention time of 7,3&cholestenol (lathosterol) is longer than that of cholesterol. However, in our HPLC system, the retention time of known lathosterol is not similar to that of the unknown compound. Thus the identity of the compound remains unresolved. A re-examination of data on cellular cholesterol metabolism in mouse macrophages, collected in this laboratory over a three year period, indicated a pattern of divergent specific activities that could be correlated with three different lots of radiolabeled cholesterol from Supplier 1. Our results demonstrate that a screening of labeled cholesterol preparations, either by TLC after oxidation with cholesterol oxidase or by HPLC, is necessary to confirm the authenticity of the labeled cholesterol obtained from commercial suppliers. Published reports have appeared demonstrating instances in which [3H]cholesterol has been an unreliable tracer. Lacko [8] reported that some preparations of [ 3H]cholesterol were inefficiently esterified by lecithin : cholesterol acyltransferase and suggested that these preparations represented contaminated or deteriorated products. Davidson et al. [9], using a “C/3H ratio approach, demonstrated that a number of different preparations of [ 3H]cholester01 appeared to be unreliable tracers for cholesterol in human studies. In the latter investigation, the labeled cholesterol was shown to be radiochemically pure, and no explanation for the discrepancy in the isotope ratios could be provided. In the present study mass spectrometry data demonstrated that labeled sterol from Supplier 1 did not arise from auto-oxidation of cholesterol. Although the identity of the labeled compound from Supplier 1 has not been established, it should be noted that there was a cell specificity in its selective metabolism since it accumulated in the steryl ester pool in macrophages but not in hepatoma cells. Whether this differential accumulation was a reflection of enhanced esterification by the macrophage acyl-CoA:cholesterol acyltransferase or reduced hydrolysis of the labeled steryl ester by the cholesteryl ester hydrolase has not been established. However, the results suggest that biologically relevant sterols, other than cholesterol, may be preferentially deposited in some cell types.
100 Acknowledgments We are very grateful to Dr. Jerrold M. Liesch for performing mass spectrometry analysis. This work was supported by Program Project Grant HL-22633 and Training Grant HL-07443 from the National Heart, Lung and Blood Institute of the National Institutes of Health. F.H.M. was partly supported by a grant from the ComitC des Aides A la Recherche Fournier-Dijon-France and by the Heinz Foundation.
4
5
6
7
References Bamberger, M., Lund-Katz, S., Phillips, M.C. and Rothblat, G.H., Mechanism of the hepatic lipase induced accumulation of high density lipoprotein cholesterol by cells in culture, Biochemistry, 24 (1985) 3693. Bligh, E.G. and Dyer, W.J., A rapid method of total lipid extraction and purification, Can. J. B&hem., 37 (1959) 911. Johnson, W.J., Bamberger, M.J., Latta, RA., Rapp, P.E., Phillips, M.C. and Rothblat, G.H., The bidirectional flux of
8 9
cholesterol between cells and lipoproteins, J. Biol. Chem., 261 (1986) 5766. Markwell, M.A.K., Haas, SM., Bieber, L.L. and Tolbert, N.E., A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples, Anal. B&hem., 87 (1978) 206. Brown, MS., Ho, Y.K. and Goldstein, J.L., The cholesteryl ester cycle in macrophage foam cells. Continual hydrolysis and reesterification of cytoplasmic cholesteryl esters, J. Biol. Chem., 255 (1980) 9344. Glick, J.M., Adelman, S.A. and Rothblat, G.H., Cholesteryl ester cycle in cultured hepatoma cells, Atherosclerosis, 64 (1987) 223. DiBussolo, J.M. and Nes, W.R., Structural elucidation of sterols by reversed-phase liquid chromatography I. Assignment of retention coefficients to various groups, J. Chromatogr. Sci., 20 (1982) 193. Lacko, A.G., A test for the biochemical efficiency of isotopic cholesterol, Lipids, 15 (1980) 983. Davidson, N.O., Ahrens, Jr., E.H., Bradlow, H.L., McNamara, D.J., Parker, and Samuel, P., Unreliability of tritiated cholesterol studies with [1,2-3H]cholesterol and [24,253HJcholesterol in humans, Proc. Natl. Acad. Sci. USA, 77 (1980) 2255.
95
84 (1990) 95-100
Elsevier Scientific Publishers Ireland, Ltd. ATHERO 04523
Potential problems in the use of commercial preparations of radiolabeled cholesterol Florence H. Mahlberg, Annabelle Rodriguez-Oquendo, David W. Bernard, Jane M. Glick and George H. Rothblat Department
of Physiologyand Biochemistry, The Medical College of Pennsylvania, 3300 Henry Avenue, Philadelphia, PA 19129 (U.S.A.) (Received 29 November, 1989) (Revised, received 1 May, 1990) (Accepted 4 May, 1990)
Through a series of biological and analytical procedures, we demonstrate that a compound purchased from a commercial supplier as [7-3H]cholesterol was not cholesterol. In mouse peritoneal macrophages, this compound was metabolized differently than other radiolabeled cholesterol preparations and was accumulated in the steryl ester pool. In contrast, Fu5AH rat hepatoma cells did not discriminate this compound from cholesterol. Further analysis of the anomalous [7-3H]cholesterol by TLC after cholesterol oxidase treatment and by HPLC indicated that this radiochemical was less polar than cholesterol standard and other radiolabeled cholesterol preparations tested. Mass spectrometry analysis disclosed that the chemical has a similar fragmentation pattern and the same molecular weight (386) as cholesterol.
Key words: Cholesterol; Radiolabeled cholesterol; Commercial cholesterol preparations
Introduction
Both in vitro and in vivo studies on the metabolism of cholesterol are often dependent on the use of either 14C- or 3H-radiolabeled cholesterol. Although it is accepted practice to repurify these compounds to free them of autooxidation products prior to use, the investigator is
Correspondence to: Dr. F.H. Mahlberg, Department of Physiology and Biochemistry, The Medical College of Pennsylvania, 3300 Henry Avenue, Philadelphia, PA 19129 (U.S.A.)
0021-9150/90/$03.50
largely dependent on the manufacturers to ensure the quality of the radiolabeled compounds which they supply. In the course of cell culture studies designed to investigate aspects of cholesterol metabolism, we have observed anomalies in the intracellular metabolism of cholesterol which led us to suspect that a product which we were using, supplied as [7- 3H]cholesterol, was not cholesterol. Through the use of a series of biological, enzymatic and chromatographic procedures we demonstrated that this was indeed the case and that the metabolism of this compound was significantly different from that of cholesterol, depending on the cell type under investigation.
0 1990 Elsevier Scientific Publishers Ireland, Ltd.
Materials and methods
Cholesterol preparations The various preparations of radiolabeled cholesterol were purchased from commercial suppliers, designated Supplier 1 to Supplier 4. The cellular and biochemical analyses described in this study were initiated upon receipt of the labeled cholesterol from the suppliers. All radioactive compounds were stored in toluene at -2O’C. The radiolabeled cholesterol treated with cholesterol oxidase or analyzed by HPLC was used as supplied by the manufacturers. The cholesterol preparations used for cell culture studies were purified on silica gel glass plates developed in diethyl ether. Radiolabeled bands were eluted from the silica gel [l] and used within 1 week of purification. Cells
All media were buffered with 24 mM bicarbonate and contained gentamicin (50 pg/ml). Thioglycollate-elicited peritoneal macrophages were isolated from B6C3Fl mice. Four days prior to collection, sterile thioglycollate broth (10%) was injected intraperitoneally. Following cervical dislocation, peritoneal lavage was performed using 5 ml sterile phosphate-buffered saline (PBS) containing heparin (2 units/ml) per mouse. The cells were collected by low speed centrifugation and suspended in RPMI-1640 (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Sigma). Two hours after plating, the dishes were rinsed 5 times with serum-free medium, and the adherent cells were used for experiments. FuSAH rat hepatoma cells were grown as previously described [l] and were used 2 days after plating. Loading of both types of cells with labeled cholesterol was accomplished by incubating cell monolayers for 24 h in medium supplemented with 1% heat inactivated FBS, 1% bovine serum albumin (BSA) and either 50 pg protein/ml LDL for the hepatoma cells, or 50 I-18 protein/ml acetylated LDL (acLDL) for the macrophages. Ethanol solutions of the various preparations of [ 3H]cholesterol or [ “C]cholesterol (as indicated in the text) were added to the FBS prior to the dilution of the serum into the medium.’ Both [3H]cholesterol and [‘4C]cholesterol were added to
provide a final concentration of 0.5 pCi/rnl, and the final concentration of ethanol was 0.1%. All media were incubated for 24 h at 37O C prior to use in the cell culture studies. Following a 24-h incubation period in the presence of the loading media, cell monolayers were washed 5 times with serum-free medium and replicate dishes were harvested for determination of the specific activities of the cellular free and esterified cholesterol. Analytical methods Lipids were extracted from media using the method of Bligh and Dyer [2] and from washed monolayers using isopropanol as previously described [3]. Free and total cholesterol mass determinations were conducted by gas liquid chromatography using coprostanol as an internal standard [3]. The distribution of radiolabel between free and esterified cholesterol was determined by thin-layer chromatography on ITLCSA plates (Gelman Sciences) developed in petroleum ether/diethyl ether/acetic acid (85 : 15 : 1, v/v). 3H and 14C were measured by liquid scintillation counting in a Beckman LS7500 using Scintiverse counting fluid. Protein concentrations were determined by the procedure of Markwell et al. [4] using BSA as a standard. Cholesterol oxidase assay [*4C]Cholesterol and [ 3H]cholesterol preparations, as indicated in the text, were incubated with cholesterol oxidase (EC1.1.3.6), and the reaction products were then compared following TLC analysis. For each assay, a mixture of 0.5 PCi radiolabeled cholesterol, 200 pg unlabeled cholesterol (Sigma) and 200 pg egg phosphatidylcholine (Sigma) was dried under nitrogen. This lipid mixture was then dissolved in 50 ~1 of ethanol, after which 0.8 ml PBS was added to the tube. To start the reaction, 20 IU cholesterol oxidase (Beckman) dissolved in 0.31 M sucrose buffered with 0.5 mM sodium phosphate, pH 7.5, were added to the solubilized lipids. The reaction mixture was incubated at 37” C for 1.5 h, and the lipids were then extracted by the method of Bligh and Dyer [2]. The extracted lipids were separated on Bakerflex silica gel TLC plates developed for 20 cm in hexane/diethyl ether/methanol (98 : 7.5 : 15, v/v). The radiolabeled compounds Were identified and
97 quantitated by scanning the TLC plates with a Radiomatic Thin Layer scanner. Unlabeled cholesterol and cholestenone (Sigma) used as standards were identified by iodine staining. HPLC analysis
Reversed-phase HPLC analysis of the radiolabeled sterols was performed using a 25 cm X 4.6 mm Isco ODS (C18) column with methanol as the mobile phase at a flow rate of 1.5 ml/min. Radioactivity in the column effluent was detected using a Radiomatic Flo-one A200 radioactive flow detector. Mass spectrometry
analysis
The major component of 3H compound from Supplier 1 was isolated by HPLC and analyzed on a Finnigan-MAT TSQ70B mass spectrometer in the electron impact mode (EI, 70 ev). Results and discussion
Table 1 shows the results of a study in which mouse peritoneal macrophages and Fu5AH rat hepatoma cells were cultured in media containing various preparations of radiolabeled cholesterol. The data are presented as the ratio of the specific activity of esterified cholesterol to that of the free cholesterol in the cells after a 24-h labeling period. If all the intracellular pools of free and esterified cholesterol reached equilibrium, the esterified to free cholesterol specific activity ratio should approximate 1. When macrophages were incubated TABLE 1 RATIOS OF ESTERIFIED TO FREE CHOLESTEROL SPECIFIC ACTIVITIES IN CELLS EXPOSED TO RADIOLABELED CHOLESTEROL Data are expressed as the ratios of the specific activities of the cellular esterified and free cholesterol, determined at the end of a 24-h labeling period. Values are the mean f SD (n = 3). Cell
Macrophage Fu5AH
Specific activity ratio Esterified cholesterol/free
cholesterol
[7-3H]Chclesterol (Supplier 1)
[4-i4C]Cholesterol (Supplier 4)
[1,2-3H]Cholesterol (Supplier 4)
2.7kO.6 0.8kO.l
0.9f0.2 0.6kO.l
1.2*0.2 0.6kO.l
in medium containing [7-3H]cholesterol from Supplier 1, the ratio of the specific activities of esterified to free cholesterol was 2.7. A number of hypothetical models invoking the selective utilization of cholesterol pools could be suggested to explain how the cellular cholesteryl ester could acquire a specific activity greater than the cellular free cholesterol that serves as its precursor. However, since it has been demonstrated that macrophages have an active cholesteryl ester cycle [5,6] this specific activity ratio after 24 h loading is somewhat surprising. Moreover, when similar experiments were performed using [4-‘4C]cholesterol or [1,2- 3H]cholesterol from Supplier 4, specific activity ratios in all cholesterol pools approximated 1 (Table 1). In addition, the divergence in specific activities of cholesterol pools observed in macrophages with the Supplier 1 [7-3H]cholesterol was not observed in the FuSAH hepatoma cells; as shown in Table 1, specific activity ratios of the cholesterol pools were similar in this cell line regardless of the source of radiolabeled cholesterol. To further probe the reason for the unexpected specific activity data obtained with macrophages labeled with [7-3H]cholesterol from Supplier 1, the uptake and cellular metabolism of that preparation and [1,2-3H]cholesterol from Supplier 4 were compared to that of [4-‘4C]cholesterol from Supplier 4. In these experiments hepatoma cells and macrophages were exposed to media containing mixtures of the 3H and 14C compounds, and the “C/3H ratio of the total, free and esterified cholesterol were determined after a 24-h incubation period. The results from such an experiment are presented in Table 2 and are normalized to the isotope ratio of the starting incubation media which was set equal to 1.0. Any significant divergence from 1.0 indicates a selective utilization of either the [14C]- or [ 3H]cholesterol. Such divergent ratios were clearly evident in both the free and esterified cholesterol extracted from macrophages exposed to the mixture of [‘4C]cholesterol (Supplier 4) and [7-3H]cholesterol (Supplier 1). The isotope ratio was increased in the case of the free cholesterol and reduced in the esterified cholesterol. However, the isotope ratio calculated for the total cellular cholesterol isolated from this macrophage preparation was similar to that of the starting medium. If it is assumed that the metabo-
98 TABLE 2 COMPARATIVE UPTAKE AND METABOLISM OF [‘4C]CHOLESTEROL AND [3H]CHOLESTEROL BY MACROPHAGES AND FuSAH CELLS Data are expressed as the [‘4C]/[3H] ratios of the ceIlular total (TC), free (FC) and esterified (EC) cholesterol radioactivity, determined after a 24-h labeling period and normalized to that of the incubation medium, which was set equal to 1.0. Values arethemeanfSD(n=3). 14C/‘H TC
ratio FC
EC
[4-‘4C]cholesterol (Supplier 4) and [7-3H]cholesterol (Supplier 1) 1.1 kO.1 1.5 *0.1 Macrophage 0.8fO.l Fu5AH 1.0*0.0
0.6 f 0.1 1.1* 0.1
[4-14C] and [1,2-3Hjcholesrerol (Supplier 4) 1.0fO.l 0.9iO.l Macrophage 1.0fO.l 1.0fO.l Fu5AH
0.9fO.l 0.9rtO.l
lism of the [‘4C]cholesterol accurately reflected the metabolism of cholesterol, then the data presented in Table 2 indicate that there was no selective uptake of the Supplier 1 [7-3H]cholesterol by the macrophages, but that there was an accumulation of the [7-3H]sterol in the cellular steryl ester pool. This observation can reflect either a selective esterification of the compound from Supplier 1 or a failure to hydrolyze its ester form. In contrast to the data obtained with the macrophage cultures, the rat hepatoma cell line FuSAH exhibited no selective utilization of the [7-3H]cholesterol from Supplier 1. In addition, when both cell types were grown in media supplemented with a mixture of [4-‘4C]cholesterol and [1,2-3H]cholesterol (both from Supplier 4), the isotope ratios of the cellular total, free and esterified cholesterol were similar to that of the starting incubation media, demonstrating that these two labeled compounds were metabolized by the cells in an identical manner. Since the cell studies demonstrated that the [7- ‘HIcholesterol from Supplier 1 selectively accumulated in the steryl ester pool in macrophages, we subjected different radiolabeled cholesterol preparations to further study in order to establish if the radiolabeled sterol from Supplier 1 was indeed cholesterol. The radioactivity scans of thin-layer chromatograms of the [4-14C]cholesterol from Supplier 3 (Fig. 1, panel A) and [7-
3H]cholesterol from Supplier 1 (panel B) developed in a solvent system that separates a number of sterol oxidation products from cholesterol demonstrated that the two preparations had a similar mobility with a Rf value of 0.35. However, thinlayer chromatography of the reaction products formed upon incubation of the compounds with cholesterol oxidase showed that 90% of the oxidized [7-3H]cholesterol from Supplier 1 (Fig. 1, panel D) had a mobility (R, = 0.56) greater than that of known cholestenone, the reaction product of cholesterol oxidase on cholesterol (R, = 0.50). The oxidized [4-‘4C]cholesterol from Supplier 3 (Fig. 1, panel C) migrated like cholestenone with R, = 0.50. [1,2-3H]Cholesterol from Supplier 3 and [7-3H]cholesterol from Supplier 2, also analyzed by TLC, showed migration characteristics identical to [4-‘4C]cholesterol from Supplier 3 and known standards before oxidation, and to cholestenone after oxidation (data not shown). HPLC analysis also demonstrated that [73H]cholesterol from Supplier 1 did not chromatograph with the same retention time as the other labeled cholesterol preparations or with a known standard. As shown on the HPLC elution profiles (Fig. 2), more than 90% of the compound from Supplier 1 was eluted after 12.7 min (panel B) which is longer than the retention time of 11.5 min observed for [4-‘4C]cholesterol from Supplier 4 (panel A) and known cholesterol standard. For each of the other preparations of radiolabeled cholesterol analyzed by HPLC (Supplier 3 [4-14C], Supplier 3 and Supplier 2 [7-3H], Supplier 3 and 1001
501
I
h
Ali
I
I
A
cl
Fig. 1. TLC migration profiles (see Materials and Methods) of [4-‘4C]cholesterol from Supplier 3 before (A) and after (C) cholesterol oxidase treatment, and [7-‘HIcholesterol from Supplier 1 before (B) and after (D) cholesierol oxidase treatment.
99
Elutlon
ttme (men)
Fig. 2. HPLC elution profiles (see Materials and Methods) of [4-‘4C]cholesterol from Supplier 4 (A) and [7-‘HIcholesterol from Supplier 1 (B).
Supplier 4 [1,2-3H]cholesterol), more than 90% of the radioactivity had a retention time of 11.5 min, similar to known cholesterol (data not shown). Both TLC and HPLC data clearly indicated that the [7-3H]cholesterol obtained from Supplier 1 was not cholesterol, but rather a closely related compound that had TLC-chromatographic properties similar to that of cholesterol. Mass spectrometry analysis of the compound purchased from Supplier 1 indicated that this compound had the same molecular weight (386) and general fragmentation as cholesterol. The specific activity of the material was such that approximately 30% of the molecules were tritiated and the tritium label was apparent in the more intense ion clusters. Thus, the MS data coupled with the biological and chromatographic data indicate that the compound from Supplier I is probably an isomer of cholesterol. Epimerization of the hydroxyl group at position three and/or migration of the double bond from 5,6 to 4,5 or 7,8 are reasonable possibilities. However, reference spectra for these compounds do not differ significantly enough to allow unequivocal assignment by comparison with our rather low intensity spectrum obtained on the small amount (2 ng) of HPLC-isolated material. Studies by DiBussolo et al. [7] have shown that on a reverse phase Cl8 HPLC column, the retention times of 5,3acholestenol (epicholesterol) and of
4,3fi-cholestenol (allocholesterol) are shorter than that of 5,3/3-cholestenol (cholesterol), whereas retention time of 7,3&cholestenol (lathosterol) is longer than that of cholesterol. However, in our HPLC system, the retention time of known lathosterol is not similar to that of the unknown compound. Thus the identity of the compound remains unresolved. A re-examination of data on cellular cholesterol metabolism in mouse macrophages, collected in this laboratory over a three year period, indicated a pattern of divergent specific activities that could be correlated with three different lots of radiolabeled cholesterol from Supplier 1. Our results demonstrate that a screening of labeled cholesterol preparations, either by TLC after oxidation with cholesterol oxidase or by HPLC, is necessary to confirm the authenticity of the labeled cholesterol obtained from commercial suppliers. Published reports have appeared demonstrating instances in which [3H]cholesterol has been an unreliable tracer. Lacko [8] reported that some preparations of [ 3H]cholesterol were inefficiently esterified by lecithin : cholesterol acyltransferase and suggested that these preparations represented contaminated or deteriorated products. Davidson et al. [9], using a “C/3H ratio approach, demonstrated that a number of different preparations of [ 3H]cholester01 appeared to be unreliable tracers for cholesterol in human studies. In the latter investigation, the labeled cholesterol was shown to be radiochemically pure, and no explanation for the discrepancy in the isotope ratios could be provided. In the present study mass spectrometry data demonstrated that labeled sterol from Supplier 1 did not arise from auto-oxidation of cholesterol. Although the identity of the labeled compound from Supplier 1 has not been established, it should be noted that there was a cell specificity in its selective metabolism since it accumulated in the steryl ester pool in macrophages but not in hepatoma cells. Whether this differential accumulation was a reflection of enhanced esterification by the macrophage acyl-CoA:cholesterol acyltransferase or reduced hydrolysis of the labeled steryl ester by the cholesteryl ester hydrolase has not been established. However, the results suggest that biologically relevant sterols, other than cholesterol, may be preferentially deposited in some cell types.
100 Acknowledgments We are very grateful to Dr. Jerrold M. Liesch for performing mass spectrometry analysis. This work was supported by Program Project Grant HL-22633 and Training Grant HL-07443 from the National Heart, Lung and Blood Institute of the National Institutes of Health. F.H.M. was partly supported by a grant from the ComitC des Aides A la Recherche Fournier-Dijon-France and by the Heinz Foundation.
4
5
6
7
References Bamberger, M., Lund-Katz, S., Phillips, M.C. and Rothblat, G.H., Mechanism of the hepatic lipase induced accumulation of high density lipoprotein cholesterol by cells in culture, Biochemistry, 24 (1985) 3693. Bligh, E.G. and Dyer, W.J., A rapid method of total lipid extraction and purification, Can. J. B&hem., 37 (1959) 911. Johnson, W.J., Bamberger, M.J., Latta, RA., Rapp, P.E., Phillips, M.C. and Rothblat, G.H., The bidirectional flux of
8 9
cholesterol between cells and lipoproteins, J. Biol. Chem., 261 (1986) 5766. Markwell, M.A.K., Haas, SM., Bieber, L.L. and Tolbert, N.E., A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples, Anal. B&hem., 87 (1978) 206. Brown, MS., Ho, Y.K. and Goldstein, J.L., The cholesteryl ester cycle in macrophage foam cells. Continual hydrolysis and reesterification of cytoplasmic cholesteryl esters, J. Biol. Chem., 255 (1980) 9344. Glick, J.M., Adelman, S.A. and Rothblat, G.H., Cholesteryl ester cycle in cultured hepatoma cells, Atherosclerosis, 64 (1987) 223. DiBussolo, J.M. and Nes, W.R., Structural elucidation of sterols by reversed-phase liquid chromatography I. Assignment of retention coefficients to various groups, J. Chromatogr. Sci., 20 (1982) 193. Lacko, A.G., A test for the biochemical efficiency of isotopic cholesterol, Lipids, 15 (1980) 983. Davidson, N.O., Ahrens, Jr., E.H., Bradlow, H.L., McNamara, D.J., Parker, and Samuel, P., Unreliability of tritiated cholesterol studies with [1,2-3H]cholesterol and [24,253HJcholesterol in humans, Proc. Natl. Acad. Sci. USA, 77 (1980) 2255.
