Vol. 14,No. 3 May/June 1990
0145-6oO8/YO/ 1403-0490$2.00/0 ALCOHOLISM: CLINICAL A N D EXPERIMENTAL RESEARCH
RAPID COMMUNICATION
Oxysterols and Alcoholic Liver Disease Maria T. Ryzlak, Henry M. Fates, William L. Russell, Carl P.Schaffner
A new theory is presented implicating oxidative cholesterol metabolism and oxysterols as possible factors in the development of alcoholic liver disease. Our present studies have revealed the accumulation of cholesta-3,5-dien-7-one, 13.05 +. 2.75 pg/g (n = 8), and cholesta-4,6-dien-3-one,2.26 f 0.88 pglg (n = 8) in fatty alcoholic liver, as compared wiht controls, 0.21 f 0.12 pg/g (n = 7) and 0.3 f 0.33 pg/g (n = 7), respectively. Acetaldehyde at 1 to 6 micromolar concentration in the blood and tissues of alcoholics cannot account for the extent of tissue damage, nor can it adequately explain liver steatosis characterized by accumulation of cholesterol and fatty acids and their esters in the liver of alcoholics known for their poor dietary habits. Oxysterols may be the primary cause for the development of alcoholic liver diseases and damage to accessory tissues. Significantly lower levels of 7-ketocholesterolin fatty liver, 6.8 f 3.5 pglg (n = 8), as compared with control, 36.85 f 22.25 pg/g ( n = 7), may be responsible for the increased cholesterol content of the alcoholic liver due to the inhibitory propertiesof this sterol on HMGCoA reductase in cholesterol biosynthesis.
tivity of the citric acid cycle,*-1°thus affecting metabolism of acetate by this pathway. At the same time there is increased esterification of cholesterol.” LeEvre et aL6 showed in alcohol-fed rats that cholesterol esters accumulate in the liver, even on cholesterol-free diets, indicating that de novo cholesterol is made from ethanol-derived acetate. In these studies6when the diet was supplemented with cholesterol, ethanol feeding resulted in a breakdown of bile acid synthesis evidenced by reduced elimiiiation of bile acids and cholesterol esters from the liver. It has been recognized for sometime that increased lipid peroxidation is associated with free radical formation as the toxic manifestations of alcohol consumption. Thus, the lipid composition of the alcoholic fatty and cirrhotic livers has been extensively studied with particular emphasis on the saponifiable lipid fraction containing peroxides of straight chain fatty acids.” The nonsaponifiable lipid THANOL INGESTION is one of the major causes of fraction, which contains primarily cholesterol and neutral hepatic injury in man. The extent of the liver damage sterols, has attracted relatively minor attention. depends mainly on the amount and the duration of Cholesterol, one of the major constituents of biological ethanol intake. With prolonged ethanol consumption, membranes, has one unsaturated double bond (A 5 ) and fatty accumulations appear in the liver followed by inflam- theoretically is subject to attack by free radicals, superoxmation and liver cirrhosis, the major cause of death among ides, peroxides, molecular oxygen, and all other reactive alcoholic men and women. The mechanism and causes of oxygen species produced during lipid peroxidation. Aualcoholic liver disease are unknown. toxidation of cholesterol (in vitro) is a subject of multiple Ethanol is a rather unreactive compound and can cause investigations and is reviewed.I3 In most of the studies, tissue injury mainly via metabolism. Acetaldehyde, the highly unphysiological and frequently drastic conditions first product of ethanol oxidation, is reactive and toxic. were used which resulted in multiplicity of cholesterol However, in alcoholics, even after heavy drinking, levels oxidation products. When some of these cholesterol oxiof acetaldehyde in the blood remain low ( 1 to 6 p ~ ) , ’ . ’ dation products were evaluated for in vitro biological but there is a significant increase in the acetate p o ~ l .In~ , ~ activity, some proved to be highly cytotoxic, steatotic, man and rat maintained on a chronic ethanol diet, mutagenic and carcinogenic.14-’6 c h ~ l e s t e r o land ~ . ~fatty acid biosyntheses7are increased, as Cholesterol epoxide was shown to be carcinogenic when shown by incorporation of 14C-acetateinto cholesterol and administered subcutaneously.’7 Increased cholesterol fatty acids. Also, ethanol feeding results in depressed acepoxide levels were also observed in prostatic secretions and tissues of men with benign prostatic hyperplasia and From the Waksman Institute, Rutgers, The State University of New prostatic carcinomaI8 and in breast secretions of women Jersey, Piscataway, New Jersqv, and the National Heart and Lung with breast cancer and women in high risk groups’’ thus Institute, National Institutes of Health. Bethesda. Maryland. Received for publication August 16, 1989; revised manuscript received possibly linking cholesterol oxides to cancer. March 1. 1990; accepted March 2. 1990 In view of the fact that alcoholic fatty steatosis of the This work was supported by The Hyde and Watson Foundation. and liver precedes liver cirrhosis,” and that in our part of the the Charles and Johanna Busch Memorial Fund (MTR). Reprint requests: Carl P. Schaffner, Waksman Institute, Rutgers, The world, approximately half of the liver cancers occur in alcoholics with liver cirrhosis,21’22 possibilities exist that State University of New Jersey, Piscataway, NJ 08855-0759. Copyright 0 1990 by The Research Soeiety on Alcoholism. cholesterol oxidation products might play some role in
E
490
.4lcohol Clin Exp Res. Vol 14,No 3. 1990 pp 490-495
OXYSTEROLS AND ALCOHOLIC LIVER DISEASE
these disease processes. For these reasons our current investigations are concerned exclusively with the isolation and characterization of cholesterol peroxidation metabolites from the human alcoholic liver. MATERIALS AND METHODS Malerials Standard cholesta-3,5-dien-7-one was obtained from Steraloids, Inc., cholesta-4,6-dien-3-one and 7-ketocholesterol were from Sigma, and ['HI -cholesterol from New England Nuclear. Primary standard cholesterol purified through dibromide and recrystallized from ethanol was from Eastman. All chemicals used as standards were purified by chromatography prior to use and, with the exception of [3H]-cholesterol, were analyzed by gas chromatography-mass spectrometry (GC-MS) to confirm purity. All solvents used were HPLC grade.
49 I
at: 100 eV electron ionization potential, 1 mA filament emission current, 250'C source temperature, 1.2 kV electron multiplier voltage, scan rate of 1.0 seconds per decade with an interscan time of 0.8 seconds and a scan range of m/z 45 to 500. Data processing was conducted on a Finnigan MAT S5300 DATA SYSTEM.
Quantilution Quantitation of cholesta-3,5-dien-7-one, cholesta-4,6-dien-3-one,and 7-ketocholesterol was achieved from standard curves developed by HPLC on LiChrosorb RP-18 column (250 X 10 mm, 10 pm particle size) in methanokwater, (94:6, v/v) solvent system and a Waters HPLC system composed of the 600 Multisolvent Delivery System, 990 Photodiode Array Detector, 7 12 WISP autosampler, and the NEC APC-I11 computer. Solvent flow rate of 1 ml per min at room temperature was used. The oxysterols were identified from their retention times and Amax as compared with standards.
Liver Sumples Human liver samples, 2 to 10 hr after death, were obtained at autopsy or during surgery from local hospitals. To avoid autoxidation of cholesterol, specimens were immediately frozen and were stored under nitrogen at -20°C until used. Fatty and cirrhotic livers were obtained from individuals whose death was directly alcohol-related. The normal control livers were from individuals with no known history of alcoholism.
Lipid Extraction One to 3 g (alcoholic and control, respectively) liver were homogenized at 4°C by means of a Brinkman Polytron in (l:3, w/v), in chloroform (NaOH washed and dried): methanol (2: I , v/v). Lipids were extracted 3 times for 30 min under nitrogen. The extract was filtered through a scintered glass funnel and the pellet was re-extracted with the same solvent. Solvent from the combined extracts was evaporated and the residue dissolved in 1 to 2 ml chloroform.
Chromatography The extracts were initially purified by column chromatography on silicic acid or Silica Gel 60 F254 (22 x 2 cm) columns equilibrated with chloroform by modification of the procedure.23 The lipids were eluted with chloroform (10 x 12 ml fractions) then with chloroform:acetone (9:l, v/v). The most polar components were eluted with methanol. Solvent was removed on a rotary evaporator. Individual fractions were dissolved in 1 ml of chloroform, and the lipid composition was examined by thin layer chromatography (TLC) on Silica Gel 60 F254 (20 X 20 cm) plates (Merck) applying a 5 p1 sample. Steroid standards, 5 pl (2 mg/ml). were applied as controls for crude lipid identification. The plates were developed in ch1oroform:acetone (9: I , v/v). Cholesta-3,5-dien-7-one, cholesta-4,6-dien-3-one, and 7-ketochelesterol were visible at 254 nm. Other steroids were visualized by spraying TLC plates with 15% orthophosphoric acid and heating for 15 min at 120°C. Standards and liver steroids were visible at 366 nrn as fluorescent spots. Individual fractions containing lipids under investigation were further purified by TLC. Standards were run on a separate ( 5 x 20 cm) plate. The desired components were scraped off the TLC plate and eluted with chloroform:methanol (2: I , v/v) on a scintered glass funnel.
RESULTS
Analysis of lipid extracts of alcoholic fatty and normal control human livers by thin layer chromatography (TLC) initially revealed (Fig. 1) significant differences in the number and concentration of steroid components. On an equivalent per gram wet liver weight basis, most steroids were either nondetectable in the control liver extracts or were present at significantly lower concentrations when compared with that of the alcoholic fatty liver. The cholesterol esters, triglycerides, and free fatty acids were easily separated from free cholesterol and its oxidation products by silicic acid or Silica Gel 60 F254 columns by chloroform elution. These eluates when analyzed by TLC revealed the rapid elution of cholesterol esters whereas the triglycerides were less readily detected after spraying with 15% orthophosphoric acid and heating. Further analysis of these cholesterol ester and triglyceridecontaining chloroform eluates by gas-liquid/mass spec+SOLVENT
+POINT
TRACK FIO.
Lipid Identification The identification of cholesterol oxidation products was obtained by a GC-MS, by on column injection procedure and a GC-Varian 3400 Gas Chromatograph coupled to a Finnigan MAT Mass Spectrometer Model 8230. Gas chromatography was conducted on a 10 x 0.32 nm id DB-I capillary column of 0.25 nm film thickness at 200 to 320"C, with 10°C per min temperature programming. The mass spectrometer was operated
FRONT
OF 4PPLICATIO:l
I 2 3 4 5 6 7 8 9 1 0 1 1
Fig. 1. Fluorescent thin layer chromatogram. Lipid extracts from human liver and standards on 20 X 20 cm. 250 pm precoated Silica Gel 60, F254 TLC plates (E. Merck. Darmstadt. Germany), developed with chloroform: acetone (9:1, v/v). Tracks contain the following: (1) cholesta-3(3.5n.6(3-triol; (2) cholesterol-5n.6aepoxide; (3) 7-hydroxy-cholesterol;(4) 7-ketocholesterol; ( 5 ) cholesterol; (6) cholesta-4.6-dien-3-one; (7) cholesterol acetate; (8) cholesterol palmitate; (9) cholesterol linoleate; (10) and (11) lipid extracts from control and alcoholic human fatty livers, respectively.The comparisons of liver extracts were made on an equivalent wet liver weight basis.
492
trometry (GC-MS) revealed in the normal control livers predominantly triglycerides whereas in the alcoholic fatty liver predominantly cholesterol esters. The chloroform-acetone eluate fractions of the silicic acid or Silica Gel 60 F254 columns contained additional cholesterol and its oxidation products. Analysis of these fractions by TLC revealed for the alcoholic fatty liver a number of cholesterol oxidation products either not seen in the fractions derived from normal livers or they were present in much lower concentrations. These cholesterol oxidation products from individual eluate fractions were further purified by preparative TLC and were identified by GC-MS. Several interesting cholesterol oxidation products were thus identified from the alcoholic fatty human liver including: 7-ketocholesterol (m/z 400, molecular ion, mass to charge ratio), cholesta-3,5-dien-7-one (m/z 382), and cholesta-4,6-dien-3-one (m/z 382). The mass spectra of isolated oxysterols are shown (Fig. 2). These were initially identified with the aid of a computer and were identical to those obtained with pure oxysterol standards. It is further interesting to note that we found these keto-diene steroids to be unstable to alkali and hence they would not survive lipid saponification procedures at reflux temperatures generally employed in the analysis of lipids. Quantitation of cholesta-4,6-dien-3-one, cholesta-3,5dien-7-one, and 7-ketocholesterol was achieved by use of high performance liquid chromatography (HPLC), and the results of these analyses with normal control and alcoholic fatty livers are given in Table 1 and Fig. 3. While the GC-MS procedure was indeed more sensitive than the HPLC procedure employed for the quantitation of the oxysterols in this study, the latter procedure proved to be adequate for the analysis of the lipid extracts derived from the normal and fatty livers under study. All oxysterol levels were determined on a per gram wet liver weight basis. The levels of cholesta-3,5-dien-7-one and cholesta4,6-dien-3-one were significantly elevated in the alcoholic fatty liver as compared with the controls whereas by sharp contrast, 7-ketocholesterol levels were greatly reduced in the alcoholic fatty liver. Neither cholesta-4,6-dien-3-one nor cholesta-3,Sdien7-one can be detected in the cirrhotic fibrous liver of alcoholics (Table 1) by the HPLC procedure employed in this study. However, minor amounts of cholesta-4,6-dien3-one and 7-ketocholesterol were identified in the cirrhotic liver including a surgical specimen by the GC-MS procedure. In view of multiple reports of spontaneous air oxidation of cholesterol,” we tried to establish the reliability of our results by undertaking evaluation of nonspecific oxysterol formation from our experimental conditions. Employing pure oxysterol primary standards, cholesterol purified through the dibromide and 3H-cholesterol,we conducted the experiments at 4°C and did not find any measurable amounts of oxidation products formed as a result of our
RYZLAK ET AL.
procedures. Therefore, we believe that all the cholesterol oxidation products we isolated were present in the liver and not produced by our procedures. Additional consideration was given to the possibility that the time interval between specimen collection and lipid analysis may have been responsible for the presence of the oxysterols we detected in the alcoholic fatty liver. There were no differences between alcoholic fatty livers taken at the time of surgery and those taken at autopsy 2 to 10 hr after death. This was also true for control normal livers where the time factor between death and specimen collection also varied. All specimens were frozen and maintained under nitrogen. DISCUSSION
The results presented here represent the first report on the isolation, identification, and quantitation of cholesta3,5-dien-7-one, cholesta-4,6-dien-3-one,and 7-ketocholesterol from the alcoholic human fatty liver. Cholesta-3,5dien-7-one, reported as one of the most cytotoxic and steatotic cholesterol oxidation products tested in L-cells,” was found in the fatty liver (Table 1) at concentrations exceeding reported cytotoxicity level^.'^.'^ A 10-fold increase in the concentration of this sterol in the fatty liver of an alcoholic as compared with a normal liver might possibly be responsible for the fatty steatosis and the metabolic failure of the liver associated with alcoholic liver diseases in man. The presence of these oxysterols is not the result of experimental manipulation. Analyses of normal or fatty liver specimens taken at the time of surgery and of those specimens collected at the time of autopsy 2 to 10 hr postmortem produced for each group similar results. For autopsy samples, since the time interval between death of the patient and liver freezing varied as much with the control normal and alcoholic fatty livers, the likelihood of autoxidation of tissue cholesterol as a factor in the formation of these oxysterols also seems to be a little remote. It is also well-known that at death and thereafter there is hypoxia and accumulation of carbon monoxide and carbon dioxide in tissues and body fluids, resulting in the inhibition of cytochrome P450 oxidases and other oxidation reactions. Therefore, we believe that all cholesterol oxidation products isolated in this study were present endogenously in the liver. Our results also confirm that in the alcoholic fatty liver the cholesterol esters represent a better part of the total lipid fraction.” There is also 4 to 6 times as much free cholesterol as compared with the normal liver. The increase in cholesterol concentration at the time of increased lipid peroxidative stress during ethanol metabolism might be the two most important factors for the increased oxysterol formation observed in the fatty liver of alcoholics. In view of the excessive cholesterol and cholesterol ester accumulation in the fatty liver, our results appear to
OXYSTEROLS AND ALCOHOLIC LIVER DISEASE
493
6
i? III
I I
I
11..
IS3
W'l
8 4 ? 8.1
--.
Fig. 2. The mass spectra of oxysterols isolated from the fatty liver of alcoholics. The identified compounds are: (1) 7-hydroxycholesterol; (2) 7-ketocholesterol; (3) cholesta-4.6dien-3~ne;(4) cholesta-3,5dien-7-one.
Table 1. Concentrations of Cholesta-3.5-Dien-7-One,Cholesta-4,6-Dien-3-One.and 7-Ketocholesterol in Control and Alcoholic Human Livers
Cholesta9,Bdien-7-one rg/g Control' Alcoholic Fatty+ Cirrhotics TLC-Rf HPLC-Retention Time: hcu
0.21 k 0.12 (n = 7)t 13.05 2.75 (n = 8)t NDll(n = 4)
*
0.71 33.50 min 277 nm
Cholesta-4.6dien-3-one PSIS 0.30 2.26
* 0.33 (n = 7)t * 0.88 (n = 8)t ND (n = 4)
0.67 315 6 min 285 nm
7-Ketocholesterol ag/g 36.85 6.80
22.25 (n = 7)t
* 3.50 (n = 8)t ND (n = 4)
0.13 14.00 min 238 nm
Lipids extracted with chloroform (NaOH washedpmethanol (2:1, v/v) from 1 to 3 g liver, purified by chromatography on silicic acid or Silica Gel 60 F254 column. Fractions containing sterols, identified by TLC. were rechromatographed on TLC plates (20 x 20 cm). Silica Gel 60 F254, 250 pm. in ch1oroform:acetone (9:1, v/v). Bands of Rf 0.60 to 0.71 containing cholesta-3.5dien-7-one and cholesta-4.6dien-3-one were scraped off and sterols eluted with methanokacetone (1:1, v/v). Eluates were used for quantitative analysis. Seven-ketocholesterol was in the band-Rf 0.133. Quantitation of cholesta-3,5dien-7-one. cholesta-4,5dien-3-one. and 7ketocholesterolwas achieved with standard curves developed by HPLC on LiChrosorb RP-18 column (250 x 10 mm, 10 pm particle size) in methano1:water. (94:6. v/v) solvent system and Waters HPLC system composed of the 600 Multisolvent Delivery System, 990 Photodiode Array Detector, 712 WISP autosampler and the NEC APC-Ill computer. Solvent flow rate of 1 ml per min at room temperature. The oxysterols were identified from retention times, and h,as compared with standards (see Fig. 3). * The control livers were from subjects 9 months to 65 years of age. One subject, 9 years old, who died of Reye's Syndrome, had unusually high 7-ketocholesterol content. t Results were mean (*) so. $ The fatty alcoholic livers were from subjects 29 to 48 years old. 5 Cirrhotic/fibrous livers were from men 29 to 39 years old. T ND. not detectable by the present procedure.
494
RYZLAK ET AL.
oxidation products, cholesta-3,5-dien-7-one and cholesta4,6-dien-3-one. These lipid peroxidation steroid metabolites possibly might be responsible for the permanent liver damage leading to liver cirrhosis and cancer as well as damage to other organs. It is of particular interest that a similar minor pathway leading to the accumulation of cholesta-4,6-dien-3-one was identified in Cerebrotendinous xanfhomatosis, a disease resulting from a defect in ... ... the bile acid synthesis pathway.28 ... ... .. . It appears that some animal studies possibly provide ,. ... 0.010 0.010 . . partial support to our hypothesis. When rats were fed ; i b 0.5 % levels of 7-ketocholesterol or 7-ketocholesterol ace0.m tate, they exhibited loss of appetite and reduced liver and I .. .. I h body weights,29whereas rabbits developed signs of hepatic N h .. .. 0.006 toxicity and cirrhosis,30and guinea pigs, liver necrosi~.~' . . .. :. Renal malignancy was also observed in rabbits.30Animals 0.m . . fed cholesta-3,5-dien-7-one had elevated serum cholesterol .... ..... 0.002 0.002 levels.32 . . .., ... In view of increasing evidence linking loss of regulation 0.000 of cholesterol biosynthesis just prior to, or inevitably with, 0 20 w) O' 0 20 co ? a 3 3 3 6 the development of cancer,33the accumulation of cholesterol in the fatty liver of alcoholics cannot be ignored. TIE (MlN) Reports show that there is enhanced cholesterol biosynFig. 3. Quantitation of cholesta-4,6-dien-3-one and cholesta-3.5dien-7-one by thesis in cancer cells such as spontaneous hepatomas in the HPLC procedure as described in Table 1. (a) Shows resolution and retention times of (1) cholesta4.6-dien-3-one. 31 5 6 min; and (2) cholesta-3.5-dien-7-one. mice, rat hepatomas, aflatoxin-induced hepatomas, mice 33.50 min. (b) Scan of two peaks identified as choIesta-4.6dien-3-one. 285 nm. lymphocytes in spontaneous leukemia, in lymphoid cells and cholesta-3.5dien-7-one,277 nm. present at very low concentrationsfollowing in various forms of human leukemia, and in tumors isolation by HPLC from human liver. (c) and (d) peaks of cholesta-3,5-dien-7-one at the arrow, (c) control and (d) fatty liver, respectively. Cholesta-3,5-dien-7-one. induced by carcinogen^.^^ Also animals fed aflatoxin do 24,400. All analyses were made on the basis of equivalent wet liver weights. not express the ability to regulate the hepatic cholesterol synthesis in response to dietary c h o l e ~ t e r o l . ~ ~ support the regulatory properties of 7-ketocholesterol in Lipid peroxidation and enhanced free radical formation cholesterol bio~ynthesis.~~ The current results (Table 1) are associated with various forms of cancer.3s Therefore, show that 7-ketocholesterol levels in the control normal increased tissue cytoxicity in vital organs and increased livers are significantly higher than in the alcoholic fatty incidence of cancer by alcohol consumption might possiliver. As identified in Fig. 2, the metabolite, 7-hydroxybly be considered as an effect of lipid peroxidation and in cholesterol (I), is a major intermediate in the bile acid particular increased oxysterol formation. Cholesta-3,5synthesis pathway and is a direct precursor to 7-ketochocholesta-4,6-dien-3-0ne,~~ like cholesdien-7-0ne~~ and lesterol (2). Both 7-ketocholesterol and 7-hydroxycholesterol have been implicated as regulatory sterols of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA reductase).2s,26As an inhibitor of HMG-CoA reductase, the rate-limiting enzyme of cholesterol formation, 7-ketocholesterol is a potent inhibitor of de novo cholesterol biosynthesis. According to additional reports'3.247-ketocholesterol is involved in the regulation of cholesterol biosynthesis whereas at relatively low concentrations it appears to enhance cholesterol esterification. l 3 The oxysterols, &;n# IV 111 'OH cholesta-3,5-dien-7-one, and cholesta-4,6-dien-3-one were shown to be much less effective as inhibitors of HMGCoA r e d ~ c t a s e . ~ ' . ~ ~ By comparing the oxidized cholesterol metabolites in O Wl the fatty liver of alcoholics with the normal metabolites BILE ACIDS formed in the pathway of bile acid synthesis,27we came Fig. 4. Suggested minor pathway of cholesterol metabolism in human liver: (I) to the conclusion that in alcoholic steatosis, part of the cholesterol; (11) 7-a-hydroxycholesterol; (Ill) 7-a-hydroxy-4cholesten-3-one; (IV) cholesterol might be metabolized via two minor pathways cholesta-4.6dien-3-one (V) 7-ketocholesterol (VI) cholesta-3,5dien-7-one; (VII) (Fig. 4),which lead to the potentially dangerous cholesterol 3~~,4~epoxycholestane-7-one.
1
%I
A
p 5wR
P
t1
ck?
OXYSTEROLS AND ALCOHOLIC LIVER DISEASE
terol, can be oxidized further to epoxides. This may possibly happen during lipid peroxidation. Judging the structures of steroids IV and VI (Fig. 4), it appears that they might also be subject to free radical polymerization and co-polymerization with other molecules. They may also have the potential to form Schiff bases with -NH2 groups on DNA and proteins. Other tissues such as the brain, lungs, spleen, heart, etc., also can synthesize cholesterol. The possibility exists that damage caused to these tissues by ethanol might be due to cholesterol oxidation product formation within these tissues. Thus the present hypothesis may possibly explain the causes of tissue damages by ethanol to accessory organs, which until now could not be explained by any previous hypothesis. The presence of the cytotoxic oxysterols, cholesta-3,5-dien-7-one and cholesta-4,6-dien-3one, in the alcoholic fatty human liver and their potential in the development of alcoholic liver disease and cancer certainly deserves further investigation. ACKNOWLEDGMENTS We are grateful to Dr. Robert T. Rosen and Dr. Thomas G. Hartman of the Food Science Center, Rutgers University, for their help in some mass spectrometric analysis. Our thanks to Marietta Walsh and Clara Weitz for their assistance in the preparation of this manuscript.
REFERENCES 1. Eriksson, CJP: Problems and pitfalls in acetaldehyde detenninations . Clin Exp Res 4:22-29, 1980 2. Peters TJ, Ward RJ, Rideout J, Lim CK: Blood acetaldehyde and ethanol levels in alcoholism, in Genetics and Alcoholism. New York, Alan R Liss Inc, 1987, p 215 3. Nuutinen H, Lindros K, Hekali P, Salaspuro M: Elevated blood acetate as indicator of fast ethanol elimination in chronic alcoholics. Alcohol 2:623-626, 1985 4. Salaspuro MP, K o m U-M, Nuutinen H, Roine R: Blood acetate and urinary dolichols: New markers of heavy drinking and alcoholism, in Genetics and Alcoholism. New York, Alan R. Liss, Inc., 1987, p 231 5. Lieber CS, Jones DP, Mendelson J, DeCarli LM: Fatty liver, hyperlipemia produced by prolonged alcohol consumption despite adequate dietary intake. Trans Assoc Am Phys 76:289-30 1, I963 6. LeEvre AF, DeCarli LM, Lieber CS: Effect of ethanol on cholesterol and bile acid metabolism. J Lipid Res 13:48-55, 1972 7. Lieber CS, Spritz N: Effects of prolonged ethanol intake in man: Role of dietary, adipose, and endogenously synthesized fatty acids in the pathogenesis of the alcoholic fatty liver. J Clin Invest 45:1400-141 I , 1966 8. Forsander OA, MaenpG PH, Salaspuro MP: Influence of ethanol on the lactate/pymvate and (3-hydroxybutyrate/acetoacetate ratios in rat liver experiments. Acta Chem Scand 19: 1770- I77 I , 1965 9. Lieber LCS, LeEvre A, Spitz N, Feinman L, DeCarli LM: Difference in hepatic metabolism of long- and medium-chain fatty acids: The role of fatty acid chain length in the production of the alcoholic fatty liver. J Clin Invest 46:14511460, 1967 10. Ontko, JA: Effects of ethanol on the metabolism of free fatty acids in isolated liver cells. J Lipid Res 14:78-86, 1973 1 I . Venkatesan S, Simpson KJ, Peters TJ: Hepatic cholesterol synthesis and esterification in rats after chronic ethanol feeding. Clin Sci 74:407-4 12, 1988 12. Videla LA, Valenzuela A: Alcohol ingestion, liver glutathione and
495
lipoperoxidation: Metabolic interrelations and pathological implications. Life Sci 3 I :2395-2407, 1982 13. Smith L L Cholesterol Autoxidation. New York and London, Plenum Press, 198 1 14. Sevanian A, Peterson AR: Cholesterol epoxide is a direct-acting mutagen. Proc Natl Acad Sci USA 8 1:4 198-4202, 1984 15. Higley NA, Taylor SL: The steatotic and cytotoxic effects of cholesterol oxides in cultured L cells. Food Chem Toxicol 22:983-992, 1984 16. Kelsey MI, Pienta RJ: Transformation of hamster embryo cells by cholesterol-a-epoxide and lithocholic acid. Cancer Lett 6: 143-149, 1979 17. Bischoff F Carinogenic effects of steroids. Adv Lipid Res 7: 165244, 1969 18. Schaffner CP: Prostatic cholesterol metabolism: regulation and alteration. Prog Clin Biol Res 75A: 279-324, 198 1 19. Petrakis NL: Epidemiologic studies of mutagenicity of breast fluids- relevance to breast cancer risk, in Banbury Report 8: Hormones and Breast Cancer. New York, Cold Spring Harbor Laboratory, 1981, p 243 20. Lieber CS (ed): Metabolic Aspects of Alcoholism. St. Leonards House, St. Leonargate, Lancaster, England, MTP Press Ltd, 1977 2 I. Tamburro CH, Lee HM: Primary hepatic cancer in alcoholics. Clin Gastroenterol 10:457-477, 1981 22. Seitz HK, Simanowski UA: Alcohol and carcinogenesis. Ann Rev Nutr 8:99-119, 1988 23. Hanahan DJ, Watts RM, Pappajohn D: Some chemical characteristics of the lipids of human and bovine erythrocytes and plasma. J Lipid Res 1:421-432, 1960 24. Gibbons G F The role of oxysterols in the regulation of cholesterol biosynthesis. Biochem SOCTrans I 1:649-65 1, 1983 25. Kandutsch AA, Chen HW: Inhibition of sterol synthesis in cultured mouse cells by cholesterol derivatives oxygenated in the side chain. J Biol Chem 249:6057-6061, 1974 26. Kandutsch AA, Chen HW: Inhibition of cholesterol synthesis by oxygenated sterols. Lipids 13:704-707. 1978 27. Bjkorkhem I, Skrede S, Buchmann MS, East C, Grundy S: Accumulation of 7a-hydroxy-4-cholesten-3-one and cholesta-4.6-dien-3one in patients with Cerehrotendinous xanfhomatosis:Effect of treatment with chenodeoxycholic acid. Hepatology 7:266-27 1, 1987 28. Skrede S, Bjorkhem I, Buchmann MS, Hopen G, Fause 0 A novel pathway for biosynthesis of cholestanol with 7a-hydroxylated C2,steroids as intermediates, and its importance for the accumulation of cholestanol in Cerebrotendinousxanthomafosis. J Clin Invest 75:448455, 1985 29. Erickson SK, Cooper AD, Matsui SM, Could RG: 7-Ketocholesterol: its effects on hepatic cholesterogenesis and its hepatic metabolism in vivo and in vitro. J Biol Chem 252:s 186-5 19 I , 1977 30. Altschul R, Spender EY: Biological effect of 7-ketocholesterol. Rev Canad Biol I 1:250-258, 1952 31. Bing RJ, Sarma JSM, Chan SI: Inhibition of cholesterol uptake by the arterial wall in the intact animal. Artery 5: 14-28, 1979 32. Kantiengar NL, Lowe JS, Morton RA: The effects of administering cholesterol and cholesta-3:5-dien-7-one to cockerels. Biochem J 60:34-39, 1955 33. Gibbons GF, Mitropoulos KA, Myant NB: Biochemistry ofCholesterol. New York, Elsevier Biomedical Press, 1982, p 277 34. Siperstein MD, Gyde AM, Moms H P Loss of feedback control of hydroxymethylglutaryl coenzyme A reductase in hepatomas. Proc Natl Acad Sci USA 68:3 15-3 17, I97 I 35. Masotti L, Casali E, Galeotti T: Lipid peroxidation in tumour cells. Free Radical Biol Med 4:377-3867, 1988 36. Ahmad MS, Mushfiq M, Khan NZ: Reaction of cholesta-3,Sdien-7-one with peracids. Indian J Chem 34B:936-938, 1976 37. Mendelsohn D, Mendelsohn L, Staple E: The in vitro catabolism of cholesterol: a comparison of the formation of cholest-4-en-7~~-01-3one and 5/3-cholestan-7-a-ol-3-onefrom cholesterol in rat liver. Biochem J 5:1286-1289, 1966
0145-6oO8/YO/ 1403-0490$2.00/0 ALCOHOLISM: CLINICAL A N D EXPERIMENTAL RESEARCH
RAPID COMMUNICATION
Oxysterols and Alcoholic Liver Disease Maria T. Ryzlak, Henry M. Fates, William L. Russell, Carl P.Schaffner
A new theory is presented implicating oxidative cholesterol metabolism and oxysterols as possible factors in the development of alcoholic liver disease. Our present studies have revealed the accumulation of cholesta-3,5-dien-7-one, 13.05 +. 2.75 pg/g (n = 8), and cholesta-4,6-dien-3-one,2.26 f 0.88 pglg (n = 8) in fatty alcoholic liver, as compared wiht controls, 0.21 f 0.12 pg/g (n = 7) and 0.3 f 0.33 pg/g (n = 7), respectively. Acetaldehyde at 1 to 6 micromolar concentration in the blood and tissues of alcoholics cannot account for the extent of tissue damage, nor can it adequately explain liver steatosis characterized by accumulation of cholesterol and fatty acids and their esters in the liver of alcoholics known for their poor dietary habits. Oxysterols may be the primary cause for the development of alcoholic liver diseases and damage to accessory tissues. Significantly lower levels of 7-ketocholesterolin fatty liver, 6.8 f 3.5 pglg (n = 8), as compared with control, 36.85 f 22.25 pg/g ( n = 7), may be responsible for the increased cholesterol content of the alcoholic liver due to the inhibitory propertiesof this sterol on HMGCoA reductase in cholesterol biosynthesis.
tivity of the citric acid cycle,*-1°thus affecting metabolism of acetate by this pathway. At the same time there is increased esterification of cholesterol.” LeEvre et aL6 showed in alcohol-fed rats that cholesterol esters accumulate in the liver, even on cholesterol-free diets, indicating that de novo cholesterol is made from ethanol-derived acetate. In these studies6when the diet was supplemented with cholesterol, ethanol feeding resulted in a breakdown of bile acid synthesis evidenced by reduced elimiiiation of bile acids and cholesterol esters from the liver. It has been recognized for sometime that increased lipid peroxidation is associated with free radical formation as the toxic manifestations of alcohol consumption. Thus, the lipid composition of the alcoholic fatty and cirrhotic livers has been extensively studied with particular emphasis on the saponifiable lipid fraction containing peroxides of straight chain fatty acids.” The nonsaponifiable lipid THANOL INGESTION is one of the major causes of fraction, which contains primarily cholesterol and neutral hepatic injury in man. The extent of the liver damage sterols, has attracted relatively minor attention. depends mainly on the amount and the duration of Cholesterol, one of the major constituents of biological ethanol intake. With prolonged ethanol consumption, membranes, has one unsaturated double bond (A 5 ) and fatty accumulations appear in the liver followed by inflam- theoretically is subject to attack by free radicals, superoxmation and liver cirrhosis, the major cause of death among ides, peroxides, molecular oxygen, and all other reactive alcoholic men and women. The mechanism and causes of oxygen species produced during lipid peroxidation. Aualcoholic liver disease are unknown. toxidation of cholesterol (in vitro) is a subject of multiple Ethanol is a rather unreactive compound and can cause investigations and is reviewed.I3 In most of the studies, tissue injury mainly via metabolism. Acetaldehyde, the highly unphysiological and frequently drastic conditions first product of ethanol oxidation, is reactive and toxic. were used which resulted in multiplicity of cholesterol However, in alcoholics, even after heavy drinking, levels oxidation products. When some of these cholesterol oxiof acetaldehyde in the blood remain low ( 1 to 6 p ~ ) , ’ . ’ dation products were evaluated for in vitro biological but there is a significant increase in the acetate p o ~ l .In~ , ~ activity, some proved to be highly cytotoxic, steatotic, man and rat maintained on a chronic ethanol diet, mutagenic and carcinogenic.14-’6 c h ~ l e s t e r o land ~ . ~fatty acid biosyntheses7are increased, as Cholesterol epoxide was shown to be carcinogenic when shown by incorporation of 14C-acetateinto cholesterol and administered subcutaneously.’7 Increased cholesterol fatty acids. Also, ethanol feeding results in depressed acepoxide levels were also observed in prostatic secretions and tissues of men with benign prostatic hyperplasia and From the Waksman Institute, Rutgers, The State University of New prostatic carcinomaI8 and in breast secretions of women Jersey, Piscataway, New Jersqv, and the National Heart and Lung with breast cancer and women in high risk groups’’ thus Institute, National Institutes of Health. Bethesda. Maryland. Received for publication August 16, 1989; revised manuscript received possibly linking cholesterol oxides to cancer. March 1. 1990; accepted March 2. 1990 In view of the fact that alcoholic fatty steatosis of the This work was supported by The Hyde and Watson Foundation. and liver precedes liver cirrhosis,” and that in our part of the the Charles and Johanna Busch Memorial Fund (MTR). Reprint requests: Carl P. Schaffner, Waksman Institute, Rutgers, The world, approximately half of the liver cancers occur in alcoholics with liver cirrhosis,21’22 possibilities exist that State University of New Jersey, Piscataway, NJ 08855-0759. Copyright 0 1990 by The Research Soeiety on Alcoholism. cholesterol oxidation products might play some role in
E
490
.4lcohol Clin Exp Res. Vol 14,No 3. 1990 pp 490-495
OXYSTEROLS AND ALCOHOLIC LIVER DISEASE
these disease processes. For these reasons our current investigations are concerned exclusively with the isolation and characterization of cholesterol peroxidation metabolites from the human alcoholic liver. MATERIALS AND METHODS Malerials Standard cholesta-3,5-dien-7-one was obtained from Steraloids, Inc., cholesta-4,6-dien-3-one and 7-ketocholesterol were from Sigma, and ['HI -cholesterol from New England Nuclear. Primary standard cholesterol purified through dibromide and recrystallized from ethanol was from Eastman. All chemicals used as standards were purified by chromatography prior to use and, with the exception of [3H]-cholesterol, were analyzed by gas chromatography-mass spectrometry (GC-MS) to confirm purity. All solvents used were HPLC grade.
49 I
at: 100 eV electron ionization potential, 1 mA filament emission current, 250'C source temperature, 1.2 kV electron multiplier voltage, scan rate of 1.0 seconds per decade with an interscan time of 0.8 seconds and a scan range of m/z 45 to 500. Data processing was conducted on a Finnigan MAT S5300 DATA SYSTEM.
Quantilution Quantitation of cholesta-3,5-dien-7-one, cholesta-4,6-dien-3-one,and 7-ketocholesterol was achieved from standard curves developed by HPLC on LiChrosorb RP-18 column (250 X 10 mm, 10 pm particle size) in methanokwater, (94:6, v/v) solvent system and a Waters HPLC system composed of the 600 Multisolvent Delivery System, 990 Photodiode Array Detector, 7 12 WISP autosampler, and the NEC APC-I11 computer. Solvent flow rate of 1 ml per min at room temperature was used. The oxysterols were identified from their retention times and Amax as compared with standards.
Liver Sumples Human liver samples, 2 to 10 hr after death, were obtained at autopsy or during surgery from local hospitals. To avoid autoxidation of cholesterol, specimens were immediately frozen and were stored under nitrogen at -20°C until used. Fatty and cirrhotic livers were obtained from individuals whose death was directly alcohol-related. The normal control livers were from individuals with no known history of alcoholism.
Lipid Extraction One to 3 g (alcoholic and control, respectively) liver were homogenized at 4°C by means of a Brinkman Polytron in (l:3, w/v), in chloroform (NaOH washed and dried): methanol (2: I , v/v). Lipids were extracted 3 times for 30 min under nitrogen. The extract was filtered through a scintered glass funnel and the pellet was re-extracted with the same solvent. Solvent from the combined extracts was evaporated and the residue dissolved in 1 to 2 ml chloroform.
Chromatography The extracts were initially purified by column chromatography on silicic acid or Silica Gel 60 F254 (22 x 2 cm) columns equilibrated with chloroform by modification of the procedure.23 The lipids were eluted with chloroform (10 x 12 ml fractions) then with chloroform:acetone (9:l, v/v). The most polar components were eluted with methanol. Solvent was removed on a rotary evaporator. Individual fractions were dissolved in 1 ml of chloroform, and the lipid composition was examined by thin layer chromatography (TLC) on Silica Gel 60 F254 (20 X 20 cm) plates (Merck) applying a 5 p1 sample. Steroid standards, 5 pl (2 mg/ml). were applied as controls for crude lipid identification. The plates were developed in ch1oroform:acetone (9: I , v/v). Cholesta-3,5-dien-7-one, cholesta-4,6-dien-3-one, and 7-ketochelesterol were visible at 254 nm. Other steroids were visualized by spraying TLC plates with 15% orthophosphoric acid and heating for 15 min at 120°C. Standards and liver steroids were visible at 366 nrn as fluorescent spots. Individual fractions containing lipids under investigation were further purified by TLC. Standards were run on a separate ( 5 x 20 cm) plate. The desired components were scraped off the TLC plate and eluted with chloroform:methanol (2: I , v/v) on a scintered glass funnel.
RESULTS
Analysis of lipid extracts of alcoholic fatty and normal control human livers by thin layer chromatography (TLC) initially revealed (Fig. 1) significant differences in the number and concentration of steroid components. On an equivalent per gram wet liver weight basis, most steroids were either nondetectable in the control liver extracts or were present at significantly lower concentrations when compared with that of the alcoholic fatty liver. The cholesterol esters, triglycerides, and free fatty acids were easily separated from free cholesterol and its oxidation products by silicic acid or Silica Gel 60 F254 columns by chloroform elution. These eluates when analyzed by TLC revealed the rapid elution of cholesterol esters whereas the triglycerides were less readily detected after spraying with 15% orthophosphoric acid and heating. Further analysis of these cholesterol ester and triglyceridecontaining chloroform eluates by gas-liquid/mass spec+SOLVENT
+POINT
TRACK FIO.
Lipid Identification The identification of cholesterol oxidation products was obtained by a GC-MS, by on column injection procedure and a GC-Varian 3400 Gas Chromatograph coupled to a Finnigan MAT Mass Spectrometer Model 8230. Gas chromatography was conducted on a 10 x 0.32 nm id DB-I capillary column of 0.25 nm film thickness at 200 to 320"C, with 10°C per min temperature programming. The mass spectrometer was operated
FRONT
OF 4PPLICATIO:l
I 2 3 4 5 6 7 8 9 1 0 1 1
Fig. 1. Fluorescent thin layer chromatogram. Lipid extracts from human liver and standards on 20 X 20 cm. 250 pm precoated Silica Gel 60, F254 TLC plates (E. Merck. Darmstadt. Germany), developed with chloroform: acetone (9:1, v/v). Tracks contain the following: (1) cholesta-3(3.5n.6(3-triol; (2) cholesterol-5n.6aepoxide; (3) 7-hydroxy-cholesterol;(4) 7-ketocholesterol; ( 5 ) cholesterol; (6) cholesta-4.6-dien-3-one; (7) cholesterol acetate; (8) cholesterol palmitate; (9) cholesterol linoleate; (10) and (11) lipid extracts from control and alcoholic human fatty livers, respectively.The comparisons of liver extracts were made on an equivalent wet liver weight basis.
492
trometry (GC-MS) revealed in the normal control livers predominantly triglycerides whereas in the alcoholic fatty liver predominantly cholesterol esters. The chloroform-acetone eluate fractions of the silicic acid or Silica Gel 60 F254 columns contained additional cholesterol and its oxidation products. Analysis of these fractions by TLC revealed for the alcoholic fatty liver a number of cholesterol oxidation products either not seen in the fractions derived from normal livers or they were present in much lower concentrations. These cholesterol oxidation products from individual eluate fractions were further purified by preparative TLC and were identified by GC-MS. Several interesting cholesterol oxidation products were thus identified from the alcoholic fatty human liver including: 7-ketocholesterol (m/z 400, molecular ion, mass to charge ratio), cholesta-3,5-dien-7-one (m/z 382), and cholesta-4,6-dien-3-one (m/z 382). The mass spectra of isolated oxysterols are shown (Fig. 2). These were initially identified with the aid of a computer and were identical to those obtained with pure oxysterol standards. It is further interesting to note that we found these keto-diene steroids to be unstable to alkali and hence they would not survive lipid saponification procedures at reflux temperatures generally employed in the analysis of lipids. Quantitation of cholesta-4,6-dien-3-one, cholesta-3,5dien-7-one, and 7-ketocholesterol was achieved by use of high performance liquid chromatography (HPLC), and the results of these analyses with normal control and alcoholic fatty livers are given in Table 1 and Fig. 3. While the GC-MS procedure was indeed more sensitive than the HPLC procedure employed for the quantitation of the oxysterols in this study, the latter procedure proved to be adequate for the analysis of the lipid extracts derived from the normal and fatty livers under study. All oxysterol levels were determined on a per gram wet liver weight basis. The levels of cholesta-3,5-dien-7-one and cholesta4,6-dien-3-one were significantly elevated in the alcoholic fatty liver as compared with the controls whereas by sharp contrast, 7-ketocholesterol levels were greatly reduced in the alcoholic fatty liver. Neither cholesta-4,6-dien-3-one nor cholesta-3,Sdien7-one can be detected in the cirrhotic fibrous liver of alcoholics (Table 1) by the HPLC procedure employed in this study. However, minor amounts of cholesta-4,6-dien3-one and 7-ketocholesterol were identified in the cirrhotic liver including a surgical specimen by the GC-MS procedure. In view of multiple reports of spontaneous air oxidation of cholesterol,” we tried to establish the reliability of our results by undertaking evaluation of nonspecific oxysterol formation from our experimental conditions. Employing pure oxysterol primary standards, cholesterol purified through the dibromide and 3H-cholesterol,we conducted the experiments at 4°C and did not find any measurable amounts of oxidation products formed as a result of our
RYZLAK ET AL.
procedures. Therefore, we believe that all the cholesterol oxidation products we isolated were present in the liver and not produced by our procedures. Additional consideration was given to the possibility that the time interval between specimen collection and lipid analysis may have been responsible for the presence of the oxysterols we detected in the alcoholic fatty liver. There were no differences between alcoholic fatty livers taken at the time of surgery and those taken at autopsy 2 to 10 hr after death. This was also true for control normal livers where the time factor between death and specimen collection also varied. All specimens were frozen and maintained under nitrogen. DISCUSSION
The results presented here represent the first report on the isolation, identification, and quantitation of cholesta3,5-dien-7-one, cholesta-4,6-dien-3-one,and 7-ketocholesterol from the alcoholic human fatty liver. Cholesta-3,5dien-7-one, reported as one of the most cytotoxic and steatotic cholesterol oxidation products tested in L-cells,” was found in the fatty liver (Table 1) at concentrations exceeding reported cytotoxicity level^.'^.'^ A 10-fold increase in the concentration of this sterol in the fatty liver of an alcoholic as compared with a normal liver might possibly be responsible for the fatty steatosis and the metabolic failure of the liver associated with alcoholic liver diseases in man. The presence of these oxysterols is not the result of experimental manipulation. Analyses of normal or fatty liver specimens taken at the time of surgery and of those specimens collected at the time of autopsy 2 to 10 hr postmortem produced for each group similar results. For autopsy samples, since the time interval between death of the patient and liver freezing varied as much with the control normal and alcoholic fatty livers, the likelihood of autoxidation of tissue cholesterol as a factor in the formation of these oxysterols also seems to be a little remote. It is also well-known that at death and thereafter there is hypoxia and accumulation of carbon monoxide and carbon dioxide in tissues and body fluids, resulting in the inhibition of cytochrome P450 oxidases and other oxidation reactions. Therefore, we believe that all cholesterol oxidation products isolated in this study were present endogenously in the liver. Our results also confirm that in the alcoholic fatty liver the cholesterol esters represent a better part of the total lipid fraction.” There is also 4 to 6 times as much free cholesterol as compared with the normal liver. The increase in cholesterol concentration at the time of increased lipid peroxidative stress during ethanol metabolism might be the two most important factors for the increased oxysterol formation observed in the fatty liver of alcoholics. In view of the excessive cholesterol and cholesterol ester accumulation in the fatty liver, our results appear to
OXYSTEROLS AND ALCOHOLIC LIVER DISEASE
493
6
i? III
I I
I
11..
IS3
W'l
8 4 ? 8.1
--.
Fig. 2. The mass spectra of oxysterols isolated from the fatty liver of alcoholics. The identified compounds are: (1) 7-hydroxycholesterol; (2) 7-ketocholesterol; (3) cholesta-4.6dien-3~ne;(4) cholesta-3,5dien-7-one.
Table 1. Concentrations of Cholesta-3.5-Dien-7-One,Cholesta-4,6-Dien-3-One.and 7-Ketocholesterol in Control and Alcoholic Human Livers
Cholesta9,Bdien-7-one rg/g Control' Alcoholic Fatty+ Cirrhotics TLC-Rf HPLC-Retention Time: hcu
0.21 k 0.12 (n = 7)t 13.05 2.75 (n = 8)t NDll(n = 4)
*
0.71 33.50 min 277 nm
Cholesta-4.6dien-3-one PSIS 0.30 2.26
* 0.33 (n = 7)t * 0.88 (n = 8)t ND (n = 4)
0.67 315 6 min 285 nm
7-Ketocholesterol ag/g 36.85 6.80
22.25 (n = 7)t
* 3.50 (n = 8)t ND (n = 4)
0.13 14.00 min 238 nm
Lipids extracted with chloroform (NaOH washedpmethanol (2:1, v/v) from 1 to 3 g liver, purified by chromatography on silicic acid or Silica Gel 60 F254 column. Fractions containing sterols, identified by TLC. were rechromatographed on TLC plates (20 x 20 cm). Silica Gel 60 F254, 250 pm. in ch1oroform:acetone (9:1, v/v). Bands of Rf 0.60 to 0.71 containing cholesta-3.5dien-7-one and cholesta-4.6dien-3-one were scraped off and sterols eluted with methanokacetone (1:1, v/v). Eluates were used for quantitative analysis. Seven-ketocholesterol was in the band-Rf 0.133. Quantitation of cholesta-3,5dien-7-one. cholesta-4,5dien-3-one. and 7ketocholesterolwas achieved with standard curves developed by HPLC on LiChrosorb RP-18 column (250 x 10 mm, 10 pm particle size) in methano1:water. (94:6. v/v) solvent system and Waters HPLC system composed of the 600 Multisolvent Delivery System, 990 Photodiode Array Detector, 712 WISP autosampler and the NEC APC-Ill computer. Solvent flow rate of 1 ml per min at room temperature. The oxysterols were identified from retention times, and h,as compared with standards (see Fig. 3). * The control livers were from subjects 9 months to 65 years of age. One subject, 9 years old, who died of Reye's Syndrome, had unusually high 7-ketocholesterol content. t Results were mean (*) so. $ The fatty alcoholic livers were from subjects 29 to 48 years old. 5 Cirrhotic/fibrous livers were from men 29 to 39 years old. T ND. not detectable by the present procedure.
494
RYZLAK ET AL.
oxidation products, cholesta-3,5-dien-7-one and cholesta4,6-dien-3-one. These lipid peroxidation steroid metabolites possibly might be responsible for the permanent liver damage leading to liver cirrhosis and cancer as well as damage to other organs. It is of particular interest that a similar minor pathway leading to the accumulation of cholesta-4,6-dien-3-one was identified in Cerebrotendinous xanfhomatosis, a disease resulting from a defect in ... ... the bile acid synthesis pathway.28 ... ... .. . It appears that some animal studies possibly provide ,. ... 0.010 0.010 . . partial support to our hypothesis. When rats were fed ; i b 0.5 % levels of 7-ketocholesterol or 7-ketocholesterol ace0.m tate, they exhibited loss of appetite and reduced liver and I .. .. I h body weights,29whereas rabbits developed signs of hepatic N h .. .. 0.006 toxicity and cirrhosis,30and guinea pigs, liver necrosi~.~' . . .. :. Renal malignancy was also observed in rabbits.30Animals 0.m . . fed cholesta-3,5-dien-7-one had elevated serum cholesterol .... ..... 0.002 0.002 levels.32 . . .., ... In view of increasing evidence linking loss of regulation 0.000 of cholesterol biosynthesis just prior to, or inevitably with, 0 20 w) O' 0 20 co ? a 3 3 3 6 the development of cancer,33the accumulation of cholesterol in the fatty liver of alcoholics cannot be ignored. TIE (MlN) Reports show that there is enhanced cholesterol biosynFig. 3. Quantitation of cholesta-4,6-dien-3-one and cholesta-3.5dien-7-one by thesis in cancer cells such as spontaneous hepatomas in the HPLC procedure as described in Table 1. (a) Shows resolution and retention times of (1) cholesta4.6-dien-3-one. 31 5 6 min; and (2) cholesta-3.5-dien-7-one. mice, rat hepatomas, aflatoxin-induced hepatomas, mice 33.50 min. (b) Scan of two peaks identified as choIesta-4.6dien-3-one. 285 nm. lymphocytes in spontaneous leukemia, in lymphoid cells and cholesta-3.5dien-7-one,277 nm. present at very low concentrationsfollowing in various forms of human leukemia, and in tumors isolation by HPLC from human liver. (c) and (d) peaks of cholesta-3,5-dien-7-one at the arrow, (c) control and (d) fatty liver, respectively. Cholesta-3,5-dien-7-one. induced by carcinogen^.^^ Also animals fed aflatoxin do 24,400. All analyses were made on the basis of equivalent wet liver weights. not express the ability to regulate the hepatic cholesterol synthesis in response to dietary c h o l e ~ t e r o l . ~ ~ support the regulatory properties of 7-ketocholesterol in Lipid peroxidation and enhanced free radical formation cholesterol bio~ynthesis.~~ The current results (Table 1) are associated with various forms of cancer.3s Therefore, show that 7-ketocholesterol levels in the control normal increased tissue cytoxicity in vital organs and increased livers are significantly higher than in the alcoholic fatty incidence of cancer by alcohol consumption might possiliver. As identified in Fig. 2, the metabolite, 7-hydroxybly be considered as an effect of lipid peroxidation and in cholesterol (I), is a major intermediate in the bile acid particular increased oxysterol formation. Cholesta-3,5synthesis pathway and is a direct precursor to 7-ketochocholesta-4,6-dien-3-0ne,~~ like cholesdien-7-0ne~~ and lesterol (2). Both 7-ketocholesterol and 7-hydroxycholesterol have been implicated as regulatory sterols of 3-hydroxy-3-methylglutaryl coenzyme A reductase (HMG-CoA reductase).2s,26As an inhibitor of HMG-CoA reductase, the rate-limiting enzyme of cholesterol formation, 7-ketocholesterol is a potent inhibitor of de novo cholesterol biosynthesis. According to additional reports'3.247-ketocholesterol is involved in the regulation of cholesterol biosynthesis whereas at relatively low concentrations it appears to enhance cholesterol esterification. l 3 The oxysterols, &;n# IV 111 'OH cholesta-3,5-dien-7-one, and cholesta-4,6-dien-3-one were shown to be much less effective as inhibitors of HMGCoA r e d ~ c t a s e . ~ ' . ~ ~ By comparing the oxidized cholesterol metabolites in O Wl the fatty liver of alcoholics with the normal metabolites BILE ACIDS formed in the pathway of bile acid synthesis,27we came Fig. 4. Suggested minor pathway of cholesterol metabolism in human liver: (I) to the conclusion that in alcoholic steatosis, part of the cholesterol; (11) 7-a-hydroxycholesterol; (Ill) 7-a-hydroxy-4cholesten-3-one; (IV) cholesterol might be metabolized via two minor pathways cholesta-4.6dien-3-one (V) 7-ketocholesterol (VI) cholesta-3,5dien-7-one; (VII) (Fig. 4),which lead to the potentially dangerous cholesterol 3~~,4~epoxycholestane-7-one.
1
%I
A
p 5wR
P
t1
ck?
OXYSTEROLS AND ALCOHOLIC LIVER DISEASE
terol, can be oxidized further to epoxides. This may possibly happen during lipid peroxidation. Judging the structures of steroids IV and VI (Fig. 4), it appears that they might also be subject to free radical polymerization and co-polymerization with other molecules. They may also have the potential to form Schiff bases with -NH2 groups on DNA and proteins. Other tissues such as the brain, lungs, spleen, heart, etc., also can synthesize cholesterol. The possibility exists that damage caused to these tissues by ethanol might be due to cholesterol oxidation product formation within these tissues. Thus the present hypothesis may possibly explain the causes of tissue damages by ethanol to accessory organs, which until now could not be explained by any previous hypothesis. The presence of the cytotoxic oxysterols, cholesta-3,5-dien-7-one and cholesta-4,6-dien-3one, in the alcoholic fatty human liver and their potential in the development of alcoholic liver disease and cancer certainly deserves further investigation. ACKNOWLEDGMENTS We are grateful to Dr. Robert T. Rosen and Dr. Thomas G. Hartman of the Food Science Center, Rutgers University, for their help in some mass spectrometric analysis. Our thanks to Marietta Walsh and Clara Weitz for their assistance in the preparation of this manuscript.
REFERENCES 1. Eriksson, CJP: Problems and pitfalls in acetaldehyde detenninations . Clin Exp Res 4:22-29, 1980 2. Peters TJ, Ward RJ, Rideout J, Lim CK: Blood acetaldehyde and ethanol levels in alcoholism, in Genetics and Alcoholism. New York, Alan R Liss Inc, 1987, p 215 3. Nuutinen H, Lindros K, Hekali P, Salaspuro M: Elevated blood acetate as indicator of fast ethanol elimination in chronic alcoholics. Alcohol 2:623-626, 1985 4. Salaspuro MP, K o m U-M, Nuutinen H, Roine R: Blood acetate and urinary dolichols: New markers of heavy drinking and alcoholism, in Genetics and Alcoholism. New York, Alan R. Liss, Inc., 1987, p 231 5. Lieber CS, Jones DP, Mendelson J, DeCarli LM: Fatty liver, hyperlipemia produced by prolonged alcohol consumption despite adequate dietary intake. Trans Assoc Am Phys 76:289-30 1, I963 6. LeEvre AF, DeCarli LM, Lieber CS: Effect of ethanol on cholesterol and bile acid metabolism. J Lipid Res 13:48-55, 1972 7. Lieber CS, Spritz N: Effects of prolonged ethanol intake in man: Role of dietary, adipose, and endogenously synthesized fatty acids in the pathogenesis of the alcoholic fatty liver. J Clin Invest 45:1400-141 I , 1966 8. Forsander OA, MaenpG PH, Salaspuro MP: Influence of ethanol on the lactate/pymvate and (3-hydroxybutyrate/acetoacetate ratios in rat liver experiments. Acta Chem Scand 19: 1770- I77 I , 1965 9. Lieber LCS, LeEvre A, Spitz N, Feinman L, DeCarli LM: Difference in hepatic metabolism of long- and medium-chain fatty acids: The role of fatty acid chain length in the production of the alcoholic fatty liver. J Clin Invest 46:14511460, 1967 10. Ontko, JA: Effects of ethanol on the metabolism of free fatty acids in isolated liver cells. J Lipid Res 14:78-86, 1973 1 I . Venkatesan S, Simpson KJ, Peters TJ: Hepatic cholesterol synthesis and esterification in rats after chronic ethanol feeding. Clin Sci 74:407-4 12, 1988 12. Videla LA, Valenzuela A: Alcohol ingestion, liver glutathione and
495
lipoperoxidation: Metabolic interrelations and pathological implications. Life Sci 3 I :2395-2407, 1982 13. Smith L L Cholesterol Autoxidation. New York and London, Plenum Press, 198 1 14. Sevanian A, Peterson AR: Cholesterol epoxide is a direct-acting mutagen. Proc Natl Acad Sci USA 8 1:4 198-4202, 1984 15. Higley NA, Taylor SL: The steatotic and cytotoxic effects of cholesterol oxides in cultured L cells. Food Chem Toxicol 22:983-992, 1984 16. Kelsey MI, Pienta RJ: Transformation of hamster embryo cells by cholesterol-a-epoxide and lithocholic acid. Cancer Lett 6: 143-149, 1979 17. Bischoff F Carinogenic effects of steroids. Adv Lipid Res 7: 165244, 1969 18. Schaffner CP: Prostatic cholesterol metabolism: regulation and alteration. Prog Clin Biol Res 75A: 279-324, 198 1 19. Petrakis NL: Epidemiologic studies of mutagenicity of breast fluids- relevance to breast cancer risk, in Banbury Report 8: Hormones and Breast Cancer. New York, Cold Spring Harbor Laboratory, 1981, p 243 20. Lieber CS (ed): Metabolic Aspects of Alcoholism. St. Leonards House, St. Leonargate, Lancaster, England, MTP Press Ltd, 1977 2 I. Tamburro CH, Lee HM: Primary hepatic cancer in alcoholics. Clin Gastroenterol 10:457-477, 1981 22. Seitz HK, Simanowski UA: Alcohol and carcinogenesis. Ann Rev Nutr 8:99-119, 1988 23. Hanahan DJ, Watts RM, Pappajohn D: Some chemical characteristics of the lipids of human and bovine erythrocytes and plasma. J Lipid Res 1:421-432, 1960 24. Gibbons G F The role of oxysterols in the regulation of cholesterol biosynthesis. Biochem SOCTrans I 1:649-65 1, 1983 25. Kandutsch AA, Chen HW: Inhibition of sterol synthesis in cultured mouse cells by cholesterol derivatives oxygenated in the side chain. J Biol Chem 249:6057-6061, 1974 26. Kandutsch AA, Chen HW: Inhibition of cholesterol synthesis by oxygenated sterols. Lipids 13:704-707. 1978 27. Bjkorkhem I, Skrede S, Buchmann MS, East C, Grundy S: Accumulation of 7a-hydroxy-4-cholesten-3-one and cholesta-4.6-dien-3one in patients with Cerehrotendinous xanfhomatosis:Effect of treatment with chenodeoxycholic acid. Hepatology 7:266-27 1, 1987 28. Skrede S, Bjorkhem I, Buchmann MS, Hopen G, Fause 0 A novel pathway for biosynthesis of cholestanol with 7a-hydroxylated C2,steroids as intermediates, and its importance for the accumulation of cholestanol in Cerebrotendinousxanthomafosis. J Clin Invest 75:448455, 1985 29. Erickson SK, Cooper AD, Matsui SM, Could RG: 7-Ketocholesterol: its effects on hepatic cholesterogenesis and its hepatic metabolism in vivo and in vitro. J Biol Chem 252:s 186-5 19 I , 1977 30. Altschul R, Spender EY: Biological effect of 7-ketocholesterol. Rev Canad Biol I 1:250-258, 1952 31. Bing RJ, Sarma JSM, Chan SI: Inhibition of cholesterol uptake by the arterial wall in the intact animal. Artery 5: 14-28, 1979 32. Kantiengar NL, Lowe JS, Morton RA: The effects of administering cholesterol and cholesta-3:5-dien-7-one to cockerels. Biochem J 60:34-39, 1955 33. Gibbons GF, Mitropoulos KA, Myant NB: Biochemistry ofCholesterol. New York, Elsevier Biomedical Press, 1982, p 277 34. Siperstein MD, Gyde AM, Moms H P Loss of feedback control of hydroxymethylglutaryl coenzyme A reductase in hepatomas. Proc Natl Acad Sci USA 68:3 15-3 17, I97 I 35. Masotti L, Casali E, Galeotti T: Lipid peroxidation in tumour cells. Free Radical Biol Med 4:377-3867, 1988 36. Ahmad MS, Mushfiq M, Khan NZ: Reaction of cholesta-3,Sdien-7-one with peracids. Indian J Chem 34B:936-938, 1976 37. Mendelsohn D, Mendelsohn L, Staple E: The in vitro catabolism of cholesterol: a comparison of the formation of cholest-4-en-7~~-01-3one and 5/3-cholestan-7-a-ol-3-onefrom cholesterol in rat liver. Biochem J 5:1286-1289, 1966
