Importance of a Novel Oxidative Mechanism for Elimination of Intracellular Cholesterol in Humans
Abstract We have recently demonstrated that cultured human alveolar macrophages efficiently convert cholesterol into excretable 27-oxygenated products. We show here that increasing the intracellular concentration of cholesterol by a factor of 10 leads to about a twofold increase in the excretion of 27-oxygenated products from cultured macrophages. Inhibition of the sterol 27-hydroxylase caused a significant intracellular accumulation of cholesterol. A direct comparison was made between flux of cholesterol and 27-oxygenated products from macrophages preloaded with [4-14C]cholesterol. Under the specific conditions employed with fetal calf serum in the culture medium, the flux of 27-oxygenated products was about 10% of that of cholesterol. Since the sterol 27-hydroxylase, which converts cholesterol to 27-oxygenated products, is present in many cell types, we suggest that 27-oxygenation is a general mechanism for removal of intracellular cholesterol. To evaluate this hypothesis, we measured the net uptake by the human liver of circulating 27-oxygenated products, which was found to be about 20 mg/24 h. This uptake corresponds to ≈4% of the bile acid production, assuming quantitative conversion into bile acids. It is concluded that the 27-hydroxylase pathway is of significance for elimination of extrahepatic cholesterol.
- Received May 12, 1995.
- Accepted November 3, 1995.
Cholesterol homeostasis is achieved through a highly sophisticated regulation of the uptake, synthesis, and esterification of cholesterol in the body. Excess intracellular cholesterol may be stored as cholesteryl esters or may be transported out from the cell. The liver has the unique capacity to degrade cholesterol into bile acids that can ultimately be removed from the body. The complicated flux of lipoprotein-bound cholesterol to and from the liver, as well as the secretion of cholesterol and bile acids from the liver into the bile, are dependent upon the overall intake and synthesis of cholesterol.1
At the cellular level, the most important mechanism for removal of excess cholesterol is believed to be reverse cholesterol transport involving HDL. Part of the HDL cholesterol reaching the liver may be degraded into bile acids and thereby finally removed from the body.2
In some cells, however, an alternative mechanism seems to be of importance for the elimination of cholesterol. Recently, we demonstrated that human cultured macrophages have a very high capacity to convert cholesterol into the more polar metabolites 27-hydroxycholesterol and 3β-hydroxy-5-cholestenoic acid and excrete them into the medium.3 Sterol 27-hydroxylase is likely to be responsible for the formation of the two products, as shown by use of immunoblotting, a specific inhibitor, and an oxygen-18 technique. Since the two products of the reaction are present in plasma4 5 and since they are efficiently converted into bile acids in the liver,6 we suggested that conversion of cholesterol into 27-oxygenated products may constitute a defense mechanism for macrophages and possibly also other peripheral cells exposed to cholesterol. We7 8 and others9 have shown that sterol 27-hydroxylase is present in several tissues and cell types, including fibroblasts, endothelial cells, brain, kidney, and lung. The hypothesis is further supported by the demonstration of 7α-hydroxylation of oxysterols in human diploid fibroblasts.10
The aim of the present work was to evaluate the quantitative importance of this novel mechanism for cholesterol removal.
Isolation and Culture of Alveolar Macrophages
Human alveolar macrophages were isolated from bronchoalveolar lavage fluid as described.3 11 The macrophages were harvested from patients with pulmonary malignancies undergoing bronchoscopy for diagnostic purposes. The bronchoalveolar lavage was performed on the contralateral side of the tumor. From a clinical point of view, the patients represented a homogenous population. The cells were allowed to adhere to plastic culture flasks and cultured in minimum essential medium supplemented with 10% fetal calf serum, benzylpenicillin (400 U/mL), and streptomycin (0.2 mg/mL) as described in Reference 3. In some experiments, as shown in Table 1⇓, 20% fetal calf serum or 2% Ultroser was used (compare Reference 3). In these experiments, [4-14C]cholesterol (55 mCi/mmol, Radiochemical Centre, Amersham) was added to the medium, as described in Table 1⇓.
Synthesis of Cholestenoic and Norcholestenoic Acids
Cholestenoic acid was synthesized via standard Wittig-Horner reaction of 3β-tert-butyldimethylsilyloxycholen-24-al (synthesis described in Reference 12) with triethyl 2-phosphonopropionate and base. For synthesis of 27-norcholestenoic acid, triethyl 2-phosphonopropionate was replaced by triethyl phosphonoacetate.
Sodium hydride (25 mg) was suspended in 2,2-dimethoxypropane (2 mL). Triethyl 2-phosphonopropionate (100 μL, or the same amount of triethyl phosphonoacetate) was added and the mixture was stirred for 20 minutes. 3β-tert-Butyldimethylsilyloxycholen-24-al, 100 mg in 2 mL of tetrahydrofuran/2,2-dimethoxypropane, 1:1 (vol/vol) was added dropwise during 10 minutes and the reaction mixture stirred for 1 hour at room temperature. The reaction was terminated with addition of moist acidic ether, and the product was extracted with ether, washed with water, and the solvent removed under reduced pressure. The resulting ethyl-3β-tert-butyldimethylsilyloxy-Δ24-(nor-)cholestenoate was saturated in the Δ24 double bond by bubbling hydrogen gas through a solution of this compound dissolved in ethanol. This procedure was readily accomplished since the Δ24 double bond is more easily saturated than the Δ5 double bond. The reaction was monitored by gas chromatography. The resulting product, ethyl-3β-tert-butyldimethylsilyloxy-(nor-)cholestenoate was stirred in tetrabutylammonium fluoride–containing tetrahydrofuran overnight to remove the tert-butyldimethylsilyloxy group. The product, ethyl-(nor-)cholestenoate, was purified on an aluminum oxide column deactivated with 3% water and eluted with toluene/ethyl acetate, 8:2 (vol/vol). Finally, the ethyl ester group was removed via alkaline hydrolysis. The product was >95% pure, as determined by gas chromatography of the trimethylsilyl (TMS) ether/methyl ester. Mass spectra of cholestenoic acid/TMS ether/methyl ester were consistent with those of an authentic compound with prominent peaks at m/z 502, 412, 473, and 129, and mass spectra of norcholestenoic acid showed m/z peaks 14 mass units lower than corresponding ions of cholestenoic acid containing the steroid side chain. The presence of an intense ion in the mass spectrum of both compounds at m/z 129 demonstrated that the Δ5 double bond was intact.
Determination of Cholesterol in Cells
Free and esterified cholesterol were determined by using gas chromatography/mass spectrometry with [26,26,26,27,27,27-2H6]cholesterol as the internal standard, essentially as described.13
Determination of 27-Oxygenated Cholesterol in Plasma
7α-Hydroxycholesterol and 27-hydroxycholesterol in plasma were determined by isotope dilution/mass spectrometry using deuterium-labeled internal standards.14 3β,7α-Dihydroxy-5-cholestenoic acid and 7α-hydroxy-3-oxo-4-cholestenoic acid were determined as described.15
Western Blot Analysis
Western blot analysis using human sterol 27-hydroxylase antibodies was performed as described in Reference 3. Antibodies to human sterol 27-hydroxylase were a kind gift of Dr D. Russell, University of Texas Southwestern Medical Center, Dallas.
All experiments involving human volunteers and patient materials were reviewed and approved by the ethics committee at the Huddinge Hospital.
Results and Discussion
Cultured macrophages have a high capacity to convert cholesterol into 27-oxygenated products and secrete them into the medium. As shown earlier,3 human alveolar macrophages cultured in the presence of 10% fetal calf serum had a high capacity to secrete 27-oxygenated cholesterol into the medium (Table 2⇓). There was no further metabolism of 27-hydroxycholesterol into 7α-hydroxylated products, as found in diploid fibroblasts.10 The 27-oxygenated products recovered from the cell medium were always unesterified. The ratio between 27-hydroxycholesterol and 3β-hydroxy-5-cholestenoic acid varied between 0.1 and 0.5 in the experiments. The medium was found to contain 82±13% of the 27-hydroxycholesterol and 99±1% of the 3β-hydroxy-5-cholestenoic acid (mean±SEM, n=4). It is evident that 3β-hydroxy-5-cholestenoic acid is secreted from the macrophages more efficiently than the less polar 27-hydroxycholesterol.
Macrophages isolated from eight patients secreted about 4 fmol/cell of the 27-oxygenated products during 24 hours of incubation. The cholesterol content in these macrophages was about 10 fmol/cell. Thus, the oxidative mechanism has the potential to eliminate about 40% of the cell content of cholesterol in the cell in 24 hours.
We have previously shown that a major portion of the 27-oxygenated products secreted from cultured macrophages was originally derived from extracellular cholesterol.3 Accumulation of 27-oxygenated metabolites increased with increasing concentration of cholesterol in the medium up to ≈0.1 mmol/L. Further addition of free cholesterol in ethanol did not increase the flux of 27-oxygenated metabolites into the medium. As shown in Table 2⇑, addition of cholesterol-containing calf serum to the macrophages gave about the same flux of 27-oxygenated products into the medium as did the addition of free cholesterol in ethanol in the previous work.
For unknown reasons, macrophages isolated from one of the patients (Experiment 9 in Table 2⇑) had a cholesterol content about five times higher (53 fmol/cell) than the other cells, and 70% of this cholesterol was esterified. However, the secretion of 27-oxygenated products from these specific macrophages was only slightly higher than above, 4.9 fmol/cell per 24 hours. To further study the effect of a high intracellular concentration of cholesterol on the flux of 27-oxygenated metabolites, macrophages isolated from another patient were exposed to acetylated LDL16 (Experiment 10 in Table 2⇑). The total cholesterol content increased to 103 fmol/cell, 57% of which was esterified. The excretion of 27-oxygenated products from these macrophages was higher than above, 7.6 fmol/cell per 24 hours (about twice as high as in control macrophages). Thus, a higher cholesterol content in the macrophages did not lead to a marked upregulation of the sterol 27-hydroxylase within the time period studied.
In one experiment, macrophages with a high content of intracellular cholesterol were homogenized and analyzed by Western blotting, using antibodies specific for the human sterol 27-hydroxylase. The intensity of the band was similar to that obtained in a parallel analysis of control macrophages with a normal content of cholesterol. Thus, the concentration of sterol 27-hydroxylase protein also does not seem to increase as a consequence of a higher cholesterol content in the macrophages.
Inhibition of the sterol 27-hydroxylase in cultured macrophages leads to accumulation of intracellular cholesterol. If 27-hydroxylation is important for cholesterol removal, inhibition of this mechanism might increase the size of the intracellular pool of cholesterol. Cyclosporin is an efficient inhibitor of 27-hydroxylase.17 As shown in the Figure⇓, 20 μmol/L cyclosporin in the culture medium reduced the secretion of 27-oxygenated metabolites from macrophages during 24 hours by more than 90%, with a concomitant increase in the intracellular total cholesterol content by about 40%. This increase of intracellular cholesterol was of the same magnitude as the secretion of 27-oxygenated cholesterol metabolites from control macrophages during the same time. In both macrophages exposed to cyclosporin and control macrophages, the amount of esterified cholesterol was <8% (mean levels, 4.5% and 5.7%, respectively). Theoretically, the effect of cyclosporin may be the result of an effect on the flux of unmetabolized cholesterol from the cells to the medium. In a separate experiment, macrophages were preloaded with [4-14C]cholesterol, as described in the experiments below (Table 1⇑). The flux of [4-14C]cholesterol from such preloaded macrophages into the culture medium was not significantly affected by 20 μmol/L cyclosporin in the medium. The difference in flux between the macrophages exposed to cyclosporin and the control macrophages was thus only about 4% (mean of two experiments).
Comparison between flux of cholesterol and 27-oxygenated metabolites from cultured macrophages. Under the conditions employed, there is a flux of cholesterol both from the medium into the cultured macrophages3 and from the cultured macrophages into the medium. To compare the flux of cholesterol from the cultured macrophages with the corresponding flux of 27-oxygenated products, macrophages were preloaded with [4-14C]-labeled cholesterol (Table 1⇑). When such macrophages were cultured in medium containing fetal calf serum, there was a considerable flux of [4-14C]cholesterol from the cells into the medium. As shown in Table 1⇑, about 50% of the labeled intracellular cholesterol was recovered in the medium after 24 hours of culture. In the two experiments shown in Table 1⇑, the amount of 27-oxygenated products in the medium was about 10% of that of cholesterol (12% in Experiment I and 8% in Experiment II).
It should be emphasized that a 10-fold higher flux of cholesterol than of 27-oxygenated products from the macrophages does not mean that the former flux is 10-fold more important than the latter under in vivo conditions. The 27-oxygenated products are rapidly transported to the liver and rapidly eliminated as bile acids. In contrast, cholesterol may recirculate or be taken up by other cells before it is ultimately eliminated as bile acids.
There is a significant net flux of 27-oxygenated metabolites of cholesterol from extrahepatic cells to the liver. The present mechanism for removal of intracellular cholesterol is not restricted to macrophages. It was recently shown that cultured arterial endothelial cells have a high capacity to secrete 27-oxygenated products into the medium,18 and the mechanism is also active in human umbilical vein endothelial cells.3 In view of the broad tissue distribution of the sterol 27-hydroxylase, it seems likely that other types of cells may also utilize this mechanism for cholesterol elimination.
If the mechanism is of general importance, one would expect the presence of 27-oxygenated cholesterol metabolites in the circulation and a net flux of these compounds to the liver, where they are known to be converted into bile acids.6 In addition to 27-hydroxycholesterol and 3β-hydroxy-5-cholestenoic acid, there are significant amounts of 3β,7α-dihydroxy-5-cholestenoic acid and 7α-hydroxy-3-oxo-4-cholestenoic acid in the peripheral circulation.5 The latter two metabolites may also be formed in nonhepatic cells.10
Table 3⇓ summarizes the results of measurements of 27-hydroxycholesterol, 3β-hydroxy-5-cholestenoic acid, 3β,7α-dihydroxy-5-cholestenoic acid, and 7α-hydroxy-3-oxo-4-cholestenoic acid in the hepatic vein and a peripheral artery in six healthy volunteers (compare Reference 19). The lowest concentrations of the four products were found in the hepatic vein, indicating an uptake of the compounds in the splanchnic area. The total uptake of the four 27-oxygenated compounds was 26±4 mg/24 h.
For reasons of comparison we also measured the concentration of 7α-hydroxycholesterol in the two vessels. This steroid is mainly produced in the liver and is also taken up by that organ and converted into bile acids.20 The concentration of 7α-hydroxycholesterol was practically identical in the two vessels (about 40 μg/L), and the difference in concentration was 2±2 μg/L in the six patients (mean±SEM).
Theoretically, elimination of the 27-oxygenated metabolites in the splanchnic region may occur in the liver and/or the intestine. If the intestine is involved, there should be lower levels in the portal vein than in a peripheral artery or vein. In a separate experiment, we collected blood from the portal vein and a peripheral vein from five patients subjected to cholecystectomy (compare Reference 21). The concentration of the 27-oxygenated metabolites was 274±26 μg/L in the portal vein and 280±28 μg/L in the peripheral vein. The difference was 6±3 μg/L (mean±SEM). With an assumed flux of plasma through the portal vein of about 400 mL/min, this corresponds to an uptake of <4 mg of 27-oxygenated products per 24 hours in the intestine.
The present results demonstrate that there is a significant net flux of 27-oxygenated products from extrahepatic sources to the liver. In the liver, these 27-oxygenated compounds are known to be efficiently converted into bile acids (for a review see Reference 6). If ≈20 mg of the 27-oxygenated products taken up by the liver is converted into bile acids, it would correspond to 4% of the total bile acid formation. This percentage could conceivably be higher in patients with liver disease and a downregulated cholesterol 7α-hydroxylase.
In the above calculations, it is assumed that the levels of the different 27-oxygenated products are relatively stable and that there is little or no diurnal variation. In previous work by one of us,22 it was shown that there is little diurnal variation in the case of 7α-hydroxy-3-oxo-4-cholestenoic acid and 3β,7α-dihydroxy-5-cholestenoic acid. In a study of three subjects in which the two acids were analyzed repeatedly over a 24-hour period, the coefficients of variation varied between 6% and 27%. In the present study, blood was collected from two healthy subjects at 4-hour intervals during 24 hours. The coefficients of variation for the level of 27-hydroxycholesterol in the circulation of the two subjects were 8% and 16%, respectively. The corresponding figures for 3β-hydroxy-5-cholestenoic acid were 14% and 13%, respectively. It is evident that the small variations in the levels of the different 27-oxygenated products do not significantly affect the above calculations of the total flux of these compounds during a 24-hour period.
Importance of the present mechanism. We have previously shown that patients with the rare inborn disease cerebrotendinous xanthomatosis (CTX) lack sterol 27-hydroxylase activity, and mutations in the sterol 27-hydroxylase gene have been defined (reviewed in Reference 23). Despite normal or low levels of circulating cholesterol, patients with CTX are predisposed to develop premature atherosclerosis.23 This correlation suggests a protective role of sterol 27-hydroxylase in the development of atherosclerosis. It is of interest that we3 and others24 have shown that 27-hydroxycholesterol is present in relatively high amounts in human atheromas. This accumulation may reflect a failure of macrophages to remove excess cholesterol at an early stage of development of an atheroma. In view of this, it would be of interest to investigate whether patients with atherosclerosis have higher or lower levels than normal of 27-oxygenated products in the circulation. According to preliminary experiments on patients with well-defined atherosclerosis in our laboratory, however, atherosclerosis in itself does not seem to affect the circulating levels of 27-oxygenated products.
To summarize, it is evident that the present pathway for elimination of cholesterol in extrahepatic cells is of quantitative importance. Further work is needed, however, to evaluate the exact role of this mechanism in relation to reverse cholesterol transport.
This work was supported by grants from the Swedish Medical Research Council, Marianne and Marcus Wallenberg Foundation, Hjärt-Lungfonden, Stiftelsen Ragnhild och Einar Lunströms Minne, Stiftelsen Sigurd och Elsa Goljes Minne, and the European Community (No. PL 931790).
Björkhem I, Andersson O, Diczfalusy U, Sevastik B, Xiu R, Duan C, Lund E. Atherosclerosis and sterol 27-hydroxylase: evidence for a role of this enzyme in elimination of cholesterol from human macrophages. Proc Natl Acad Sci U S A. 1994;91:8592-8596.
Javitt NB, Kok E, Cohen B, Burstein S. Cerebrotendinous xanthomatosis: reduced serum 26-hydroxycholesterol. J Lipid Res. 1982;23:627-630.
Axelson M, Mörk B, Sjövall J. Occurrence of 3β-hydroxy-5-cholestenoic acid, 3β,7α-dihydroxy-5-cholestenoic acid, and 7α-hydroxy-3-oxo-4-cholestenoic acid as normal constituents in human blood. J Lipid Res. 1988;29:629-641.
Skrede S, Björkhem I, Kvittingen EA, Buchmann MS, Lie SO, East C, Grundy S. Demonstration of 26-hydroxylation of C27-steroids in human skin fibroblasts, and a deficiency of this activity in cerebrotendinous xanthomatosis. J Clin Invest. 1986;78:729-735.
Andersson S, Davis DL, Dahlbäck H, Jörnvall H, Russell DW. Cloning, structure, and expression of the mitochondrial cytochrome P-450 sterol 26-hydroxylase, a bile acid biosynthetic enzyme. J Biol Chem. 1989;264:8222-8229.
Zhang J, Larsson O, Sjövall J. 7α-Hydroxylation of 27-hydroxycholesterol and 25-hydroxycholesterol in human fibroblast cultures. 13th International Bile Acid Meeting, Falk Symposium No. 80. San Diego, Calif; 1994:86.
Lund E, Breuer O, Björkhem I. Evidence that 24- and 27-hydroxylation are not involved in the cholesterol-induced down regulation of hydroxymethylglutaryl-CoA reductase in mouse liver. J Biol Chem. 1992;267:25092-25097.
Schaffer R, Sniegoski LT, Welch MJ, White E, Cohen VA, Hertz HS, Mandel J, Paule RC, Svensson L, Björkhem I, Blomstrand R. Comparison of two isotope dilution/mass spectrometric methods for determination of total serum cholesterol. Clin Chem. 1982;28:5-8.
Basu SK, Goldstein JL, Anderson RGW, Brown MS. Degradation of cationized low density lipoprotein and regulation of cholesterol metabolism in homozygous familial hypercholesterolemia. Proc Natl Acad Sci U S A. 1976;73:3178-3182.
Dahlbäck-Sjöberg H, Björkhem I, Princen HMG. Selective inhibition of mitochondrial 27-hydroxylation of bile acid intermediates and 25-hydroxylation of vitamin D3 by cyclosporin A. Biochem J. 1993;293:203-206.
Reiss AB, Martin KO, Javitt NB, Martin DW, Grossi EA, Galloway AC. Sterol 27-hydroxylase: high levels of activity in vascular endothelium. J Lipid Res. 1994;35:1026-1030.
Ahlborg G, Juhlin-Dannfelt A. Effect of beta-receptor blockade on splanchnic and muscle metabolism during prolonged exercise in men. J Appl Physiol. 1994;76:1037-1042.
Björkhem I. Mechanism of bile acid biosynthesis in mammalian liver. In: Danielsson H, Sjövall J, eds. Comprehensive Biochemistry. Amsterdam, Netherlands: Elsevier Publishing Co; 1985:231-278.
Björkhem I, Angelin B, Einarsson K, Ewerth S. Fasting levels of monoketonic bile acids in human peripheral and portal circulation. J Lipid Res. 1982;23:1020-1025.
Axelson M, Mörk B, Sjövall J. Differential distribution of unconjugated C27-bile acids in bile and blood. In: Paumgartner G, Stiehl A, Gerok W, eds. Trends in Bile Acid Research. London, UK: Kluwer Academic Publishers; 1988:79-88.
Björkhem I, Muri-Boberg K. Inborn errors in bile acid biosynthesis and storage of sterols other than cholesterol. In: Scriver CR, Beaudet AL, Sly WS, Valee D, eds. The Metabolic Basis of Inherited Diseases. New York, NY: McGraw-Hill; 1994:2073-2100.