4

8

Cyclopentanoperhydrophenanthrene nucleus

Summary Oxygenated derivatives of cholesterol (oxysterols) are widely distributed in nature, being found in the blood and tissues of animals and man as well as in foodstuff. They exhibit many biological activities which are of potential physiological, pathological or pharmacological importance. Many oxysterols have been found to be potent inhibitors of cholesterol biosynthesis and one or more oxysterols may play a role as the physiologic feedback regulator of cholesterol synthesis. Oxysterols also inhibit cell replication and have cytotoxic properties, effects which suggest that these sterols may participate in the regulation of cell proliferation and may be potentially useful as therapeutic agents for cancer. Furthermore, there is considerable evidence that oxysterols may be involved in the pathogenesis of atherosclerosis. Although the mechanism of action of oxysterols in all these instances is not well understood, the existence of cytosolic and microsomal proteins which bind oxysterols with high affinity and specificity suggests that this group of compounds may represent a family of intracellular regulatory molecules. Introduction As a group, oxygenated derivatives of cholesterol (or oxysterols) have attracted much attention in recent years on account of their biological activities which are of potential physiological, pathological, or pharmacological importance. They can be broadly defined as compounds which possess (a) a cyclopentanoperhydrophenanthrene nucleus (Fig. l), (b) a hydrocarbon sidechain attached to C17, (c) a hydroxyl group at C3, and (d) one or more additional oxygen functions attached to the nucleus or side chain. The structures of cholesterol and some oxysterols are depicted in Fig. 1. Oxygenated sterols are widely distributed in nature. They may arise from the autoxidation of cholesterol, a process which occurs spontaneously on ex osure to air in the presence of heat, light or radiation(' . Oxysterols are also formed by enzymatic action in vivo and have bccn identified in the blood and tissues of both animals and Their occurrence in foodstuffs is also well documented("'). Although oxygenated sterols were first identified as oxidation products of cholesterol at about the turn of

P

25-hydroxycholesteroI

Cholesterol

26-hydroxycholesterol

HO H

7-ketocholesterol

7-ketocholestanol

Fig. 1. Structurcs of the cyclopentanoperhydrophenanthrene nucleus, cholesterol and some oxysterols.

the century, systematic studies were carried out only in the last 20-30 years. Interest has focused mainly on their wide spectrum of biological activities. These include, among others, cytotoxicity against a variety of cells in vitro and in vivo, inhibition of DNA synthesis, inhibition of cholesterol synthesis, immunological cffects, and effects on cell membrane structure and function. To date, more than 100 oxysterols have been synthesized and their biological activities examined. Several reviews have discussed and catalogued the ~ , ~ )present . review results of these s t u d i e ~ ( ~ ~ ' ,The attempts to highlight those biological activities which have been most extensively studied because of their possible biological or clinical significance. In addition, some recent studies from our own laboratory will also be presented. Inhibition of Cholesterol Biosynthesis It is now generally accepted that eukaryotic cells satisfy their cholesterol requirements either by receptormediated endocytosis of low density lipoprotein

(LDL") particles or by de n o w synthesis of cholesterol from acetate('). Both processes are believed to be regulated by cellular cholesterol content. In studies using human fibroblasts cultured in li oprotein-deficient medium, Brown and Goldstein(K)) established that the activity of the enzyme 3-hydroxy-3-methyglutaryl coenzyme A (HMG-CoA) reductase (EC 1.1.1.34), the rate-limiting enzyme in cholesterol biosynthesis, is suppressed by the addition of LDL to the culture medium. Subsequent studies indicated that the internalized LDL particle must first fuye with lysosomes in which cholesterol esters are hydrolyzed. liberating free cholesterol which presumably then inhibits HMG-CoA reductasc@). The proposal that cholesterol itself is the feedback regulator was challenged by Kandutsch and Chen who, in 1973, first reported that highly purified cholesterol did not inhibit cholesterol biosynthesis or reduce the activity of HMGCoA reductase in primarv cultures of fetal mouse fibroblasts or livcr cells(d")). On the other hand, 'unpurified' cholesterol, as well as several oxygenated sterols (7a-hydroxycholesterol, 7/3-hydroxy-cholesterol and 7-ketocholesterol) significantly depressed HMGCoA reductase activity and cholesterol synthesis. These observations led Kandutsch and his colleagues to hypothesize that the feedback regulator of cholesterol biosynthesis is not cholesterol itself. but possibly an oxygenated sterol('). To date many oxysterols have been reported to inhibit cholesterol biosynthesis or depress the activity of HMG-CoA reductase in cultured cells. some having potencies up to 100 times that of cholesterol. A partial list is rhown in Table 1. Inhibition of cholesterol bios nthesis by oxysterols ha5 also been observed in vivo(lB). Current evidence has not resolved the uncertainty about the nature of the feedback regulator of cholesterol biosynthesis. Although it is truc that bome oxysterols are much more potent than cholesterol in inhibiting HMG-CoA reductase, this diffcrence in potency was demonstrated in experiments in which the sterols were first dissolved in organic solvent and then added directly to the culture medium. with or without emulsification. This method of sterol delivery is not physiologic and the greater potency of oxysterols observed could be due to their greater cellular uptakc or greater water solubility. thus achieving higher concentrations in the cytosol. When cholesterol is taken up by cells through receptor-mediated endocytosis of LDL particles, as occurs physiologically, it is far more effective a$ an inhibitor of its own synthesis, with a potency not very different from that of oxysterols added as emulsions("). This observation, taken together with the fact that the oxysterol content of human LDL is not nearly high enough to account for the inhibitory effect of LDL on HMG-CoA reductase activity of cells in culture(12), would suggest that the physiologic sup* Abbreviationa: AEBS, antiestrogen hinding sitc(s): HMG-CoA. 3hydroxy-3-methylglutaryl coenzyme A: LDL, low density lipoprotein.

Table 1. Oxysterols which inhibit HiMG-CoA reductase activiry

Sterol Cholcsterol 5ru-Chole~ten-7/1,7cdiol (7 a-hydroxycholesterol) 5ru-Chole~ten-3P,7P-diol (7/3-hydioxycholesterol) Cholest-5-ene-3/3-01-7-one (7-ketocholesterol) jru-chole~tdn-3~-01-6-one (6-ketocholestanol) Cholestan-3P,Sn,fiB-trloI 3/3-hydroxy-Sru-cholest-8(14)-ene-lS-one 20(S)-Cholest-5 ene-3B,20-diol Choleit-5-ene-3P,22-diol Cholcst-5-ene-3/3,24-d1ol (24-h ydroxycholesterol) Cliolest-S-ene-3~,25-diol (25-h ydrouycholestcrol) Cholest-5-ene-3~,26-diol (26-hydroxycholcsterol) 5ru-lanost-S-ene-3B,32-diol 5a-lanost-S-en-3/3-o1-32-al 5.6-Epoxycholesterol 24(S) ,25-Epoxycholesterol

IDSO values for HMG-CoA reductabe inhibition, p c " >10.0; 2.5 1.9-2.7 1.7-2.5 O.x-1

.s

-

0.2 0.3-1.5 8.2 0.63 0.05-0.17 0.26

0.7-2.5 2.X$

0.9

Only some examples are listed above. For a more comprehensive list as wcll as for the original sources of the IDso values for HMG-CoA reductase inhibition, see Smith and JohnsonL') and references therein. *Values are those observed in mouse fibroblasts. t From Kandutsch ~t al. ( @ . $From Gibbons et d . ( l 8 ) .

pression of HMG-CoA reductase by LDL is mediated by uptake of LDL cholesterol. Several lines of evidence, however, suggest that LDL-derived cholesterol, once internalized, might be converted to an oxysterol which then inhibits cholesterol biosynthesis. Gupta ctaZ.("). for example, showed that the inhibition of HMG-CoA reductase by LDL is prevented by inhibitors of cytochrome P-450, a mixedfunction oxidase. This observation suggests a need to convert cholesterol to an oxysterol via a P-450dependent reaction in order to bring about reductase inhibition. Secondly, genetic evidence indicates that LDL-derived cholesterol and solvent-solubilized oxysterols share at least one common step in inhibiting rcductase activity. Several mutant Chinese hamster ovary cell lines have been isolated that resist regulation by both LDL-derived cholesterol and oxysterols. Single-step revertants of these mutants simultaneously regain responsiveness to both agents(14).One possible explanation for the similar responses to LDL cholesterol and oxysterol is that the LDL-derived cholesterol is converted to an oxystcrol intracellularly. Thirdly, in addition to inhibiting HMG-CoA reductase activity, LDL also suppresses several other enzymes in the cholesterol biosynthesis pathway, reduces LDL receptor activity, and stimulates acyl-CoA: cholesterol

Table 2. Examples o,f oxysterols with antiproliferative or cytotoxic acyltransferase activity (EC 2.3.1.26). ,411 these effects are mimicked by an oxysterol(2S-hydrochole~terol)(~~). effects Cells lines tcstcd 0x ys Ler ol These observations provide only circumstantial evidence for the oxysterol hypothesis. To date, direct Chinese hamster V79 lung fibroblasts 5,6a-Epoxy-5n-cholesranevidence is lacking. The oxysterols which have been rat Morris liepatoma HCT cells 3p-01 mousc fibroblasts proposed to perform a regulatory role in cellular mouse fibroblasts Cholest-5-ene-3B,7~-diol cholesterol homeostasis include side-chain oxygenated rat Morris hepatoma HTC cells derivatives of cholesterol such as 26-hydroxycholes- Cholest-5-ene-38,7P-diol murine lymphoma EL4 cells terol, 24-hydroxycholesterol, 25-h droxycholesmouse fibroblasts human skin fibroblasts as well as terol(''), and 24(S), 2S-epo~ycholesterol~~) Cholest-5-ene-38,25-dioI human arterial smooth muscle cells oxygenated cholesterol precursors such as Sar-lanost-8liumaii marrow mononuclear cclls ene-3P,32-diol and 3P-hydroxy-5ar-lanost-8-en-32-al(l"). rabbit aortic smooth muscle cells The mechanism by which oxysterols inhibit HMGmouse fibroblasts CoA reductasc remains unclear, but it is known that the iiioiise lymphocytes rat Morris hepatoma HTC cells 3phydroxycholest-5-eneinhibitory effect on reductase activity is indirect mouse fibroblasts 7-one because it is demonstrable only with intact cells. The 3/?-hydroxv-5w-u-chole~in- mouse fibroblasts isolated enzyme is not susceptible to inhibition by rabbit aortic smooth muscle cells 6-one oxysterols. Kandutsch and coworkers("' identified a Chinese hamster V7Y lung fibroblasts 3/3-hydroxy-5a-cholestancytosolic protein that binds oxysterols but not choles7-onc human marrow mononuclear cells 3/3-hydroxy-5-ene-22-0ne terol with high affinity and proposed that this protein Chinese hamster V79 lung fibroblasts Cholestan-3/3,5~,6lr;-triol might mediate the action of oxysterols in much the same Syrian hamstcr embryo cells way as steroid hormone receptors mediate steroid mouse fibroblasts hormone action. Some evidence supporting this prorabbit aortic smooth muscle cells posal comes from the observation that the binding and Peiig For a more complete list. see Smith and affinities of different oxysterols for this protein beem to and Taylorc3'. parallel their potencies for inhibiting HMG-CoA reductase(2u). This cytosolic protein has now been purified to homogeneity(21) and its cDNA ha5 been showed that whcn lipid-deficient culture medium was cloned("), but there is yet no direct evidence that this used. the inhibition of cell growth by oxysterols could oxysterol binding protein participates in sterol-mebe reversed by adding cholesterol to the culture diated regulation. In a related development, Rajavasmedium, suggesting that reduced cholesterol availhisth et al. (21) demonstrated that 25-hydroxycholesterol ability might explain thc growth inhibition by oxystertreatment of the hepatoma cell line HepC2 led not only 01s. Furthermore, when the inhibition of HMG-CoA to a suppression of cholesterol synthesis but also to reductase was bypassed by the addition of mevalonatc increased synthesis of a 19-kD protein which bound (the product derived from HMG-CoA through respecifically to the sterol regulatory element of the ductase action), growth inhibition was also reversed. A HMG-CoA reductasc gene promoter. Whether this cholesterol requirement for cell proliferation is not oxysterols-induced protein regulates HMG-CoA resurprising since cell rcplication requires the synthcsis of ductase gene activity is currently unclear since transfeccell membranes of which cholesterol is an integral part. tion studies failed to reveal any significant effect of the over-expression of this protein on the expression of a Additional studies, however, indicated that the rereductase promoter-reporter gene construct. lationship between cell proliferation and cholesterol biosynthesis is more complex. Experiments carried out with compactin, a fungal product which is a potent Antiproliferative and Cytotoxic Effects competitive inhibitor of NMG-CoA reductase, are As early as 1911, oxygenated derivatives of cholesterol particularly illuminating. When used at sufficiently high were noted to be cytotoxic to mammalian cells("). Since concentrations. compactin totally inhibits HMG-CoA then many oxysterols have been found to inhibit cell reductase activity, and also inhibits DNA synthesis and growth or kill cells both in vitro and in vivo (Table 2). It cell growth(26).Under such conditions, the addition of is worth pointing out that, under similar experimental cholesterol alone will not restore cell growth, but the conditions. cholesterol itself is not cytotoxic and has no addition of cholesterol plus a small amount of antiproliferative effect; clearly. the introduction of an mevalonate will allow DNA synthcsis and cell growth to additional oxygen function into the cholesterol molresumc. These observations have been interpreted to ecule dramatically alters its biological properties. mean that DNA synthesis and cell proliferation need The mechanisms of the antiproliferative and cytothe presence of mevalonate or a mevalunate-derived toxic effects of oxysterols are incompletely understood. product in addition to cholesterol. The nature of this The ability of oxysterols to inhibit HMG-CoA remevalonate product is a fascinating question which is ductase appears to pla a role in at least some instances. still being investigated. In 1974, Chen et ul.(') and Brown and G ~ l d s t e i n ' ~ ~ ) A clue to this question came from the studies of

Schmidt et LIZ.(”) showing that when mammalian cells were cultured in the presence of radioactively-labeled mevalonate, there was incorporation of radioactivity into proteins. Further studies(=) revealed that a mevalonate product (the 15-carbon farnesyl intermediate in the cholesterol biosynthetic pathway) was incorporated into proteins among which were the ~21’‘‘~ proteins, a family of GTP-binding proteins that arc believed to regulate cell growth. Farnesylation allows attachment of the ras proteins to the inner surface of the plasma membrane, which appears to be required for these proteins to act as regulators of cell growth. The cholesterol biospnthetic pathway therefore appears to be linked to cell proliferation in two ways: first to provide sufficient bulk cholesterol for cell membrane synthesis and, second, to generate regulatory molecules (such as the farnesyl moiety covalently linked to the ras proteins) which control, in some as yet undefined manner, DNA synthesis and cell proliferation. Inasmuch as oxysterols are capable of inhibiting HMG-CoA reductase activity, DNA synthesis. and cell proliferation, an intriguing question that arises is whether oxysterols might function as physiologic regulators of cell proliferation. This question remains to be examined. The cytotoxic effects of oxysterols cannot be totally explained by inhibition of cholesterol synthesis since they cannot always be reversed by the addition of cholesterol to the culture medium(29) and may not correlate with their ability to inhibit HMG-CoA reductase(”). In these instances, it has been speculated that, since oxysterols are known to be inserted into cell membranes and to alter membrane structure and function(2), it is possible that membrane dysfunction eventually causes cell death. One property of oxygenated sterols which has not yet been adequately explored as a possible mechanism by which they exert their antiproliferative and cytotoxic effects is their ability to bind specifically and with high affinity to the antiestrogen-binding site (AEBS). The AEBS, first identified by Sutherland and coworkers(”) about 10 years ago. is a microsomal protein which binds synthetic antiestrogens, such as tamoxifen and clomiphene, with high affinity. It does not bind estrogens such as estradiol or other steroid hormones. The AEBS is ubiquitously distributed in animal and human tissues, with the highest concentrations found in the liver(32).Its physiological function is unknown and it has not yet been purified to homogeneity. There is ample evidence that the known properties of the AEBS clearly distinguish it from the estrogen receptor(32).The estrogen receptor, unlike the AEBS, is a cytosolic protein which binds estrogens with high affinity and which has a much more restricted tissue distribution with the highest concentrations found in classical estrogen target organs such as the uterus. Moreover, the estrogen receptor is not known to bind oxysterols. Murphy et aZ.(33)first reported that the oxysterol

7-ketocholesterol inhibited the binding of radiolabelled tamoxifen ([3H]tamoxifen) to the AEBS in cockerel liver, suggesting that oxygenated derivatives of cholesterol might bind to the AEBS. The relative binding affinity of 7-ketocholesterol, estimated from the competition studies, however, was rather low, being < O . S % of that of tamoxifen. Following upon this observation, our laboratory examined a large number of ox sterols for their ability to bind to the AEBS in rat li~er(‘~>’~). These studies revealed that several oxyster01s bound rat liver AEBS with high affinity. Among these, 7-ketocholestanol displayed the highest affinity, having a Kd of 2.7 n M . These studies also established, on the basis of ligand specificity and subcellular distribution, that the AEBS is clearly distinguishable from the cytosolic oxysterol binding protein identified by Kandutsch and coworkers(19). The high-affinity binding of oxysterols to thc AEBS raises the possibility that these binding sites may mediate some of the known biological effects of oxysterols. Our laboratory compared the relative binding affinities of six oxysterols and four antiestrogens with their ability to kill MCF7 breast cancer cclls in culture(36).The results showed that, in general, the binding affinities paralleled the cytotoxic potencies: the higher the binding affinity for AEBS, the more toxic the compound (Fig. 2). Other laboratories have also carried out similar experiments using other AEBS ligands. Tang et studied a number of synthetic analogues of tarnoxifen and Fargin et LIZ.(”) examined derivatives of diphenylamine. Both studies revealed that the binding affinities of these compounds for the AEBS in general paralleled their cytotoxic potency against cultured mammalian cells. The relationship of the AEBS to ligand-induced cytotoxicity remains to be clarified but several possibilities deserve consideration. It is possible that the AEBS mediates the cytotoxic effects of oxysterols, antiestrogens, and diphenylamine derivatives in much the same way as hormone receptors mediate hormone action. Second. AEBS (whose physiological function is currently unknown) could be involved in some cellular function essential for survival; its occupancy by oxysterols or other ligands might block this activity and thus lead to cell death. Third, it is also possible that the AEBS is a molecule that has been selected by evolution to bind and detoxify potentially harmful compounds which the animal may encounter. The observed correlation between cytotoxic potency and binding affinity for the AEBS then reflects the selection by evolutionary pressure of a molecule which preferentially binds and eliminates toxic substances introduced from the external environment or generated in vivo. The observation that the liver, which is the first organ to encounter absorbed nutrients, contains the highest concentrations of AEBS is consistent with the hypothesis that the AEBS may help to remove toxic substances ingested with the diet. Oxysterols are found in food(435)

addition. oxysterols have other immunological effects including the activation of complement and the expression of class II MHC antigens(','"). Finally. oxysterols have been reported to inhibit the activity of c a l m ~ d u l i n ( ~a~ calcium-binding ), protein which is involved in numerous calcium-dependent cellular processes. The effects of oxysterols on cell membrane structure and function have been mentioned earlier. In all these instances, the physiological significance of oxysterol action is unknown.

11 1.5

2.0

I

I

2.5

3.0

J

3.5

log,oEC50values (antiproliferativeeffect)

Fig. 2. The ICw values of antiestrogens and oxysterols for inhibition of [3H]tamoxifen binding to the AERS and ECso values o f the antiproliferative effect against MCF7 cells. The compounds tested were: clomiphene (l), CI628 (2), tamoxifen (3). nafoxidine (4), 7-ketocholestanol ( 5 ) , 6-kctocholcstanol (Q, 7/3-hydroxycholestero1 (7), 7-ketocholesterol (S), 7ct-hydroxycholestero1(9) and 4-choleslen3-one (10). Relative TC50and EC5,,values were calculated with values for clomiphene set at 1W. Notc that logarithmic scales were used for both axes. To eliminate binding of ['H]tamoxifen to estrogen reccptors, 1L{M dicthylstilbestrol was added to all assay tubes. (Reproduced. with permission of the publisher, from Lin and Hwang(jb)).

and are known to be absorbed readily by the gastrointestinal tract(3). Other Biological Actions of Oxysterols A number of other biological activities of oxysterols have been described. These include mutagenicity and possibly carcinogenicity, immunological effects, and several other miscellaneous activities which have received limited attention. Highly purified cholesterol is neither mutagenic nor carcinogenic. Cholesterol preparations which have been autoxidized have been found to be mutagenic in bacteria as well as in mammalian cell cultures(2). In addition, some oxysterols, such as the isomeric 5,6epoxides of cholesterol and cholestan-3/3, 5m, 6/?-triol, exhibit transforming activity against certain cell lines and may therefore be potential carcinogen^('^"^. Oxygenated sterols have been reported to be immunosuppressants, inhibiting lymphocyte proliferation and transformation, the mixed lymphocyte reaction, the activity of natural killer cells, the secretion of interleukin-2 by splenocytes and the appearance of interleukin-2 receptors in human T lymphocytes. In

Possible Biological Significance of Oxysterols Three major questions may be asked about the biological activities of oxysterols: (1) Are any of the biological activities of oxysterols of physiological relevance? (2) Do oxysterols play any pathogenetic role in human disease? (3) Are oxysterols of potential therapeutic value in the treatment of human disease? Physiologically. a number of oxygenated sterols are known to be produced in vivo through enzyme action. For example, 20m-hydroxycholesterol and 7achydroxycholesterol are intermediates of steroid hormone and bile acid synthesis, respectively. In other instances the role of oxysterols formed in vivo remains undefined. The evidence that one or more oxysterols may serve as the feedback regulator of cholesterol biosynthesis has been summarized above, but specific oxysterols serving this particular regulatory function have nut yet been identified. Although possible physiological roles of oxysterols are speculative, there is something attractive about the idea that oxysterols may represent another family of cholesterol derivatives which may serve, for example, as intracellular signalling or regulatory molecules. Cholesterol, an altogether remarkable molecule whose synthesis requires some 30 different steps. is not only an essential component of cell membranes. but also serves as the precursor of steroid hormones and bile acids. Since enzymes capable of converting cholesterol into oxysterols exist in vivo, it would be surprising if nature did not make use of this opportunity to generate a family of molecules to subserve some physiological function. In this connection it is intriguing to note that at least two high-affinity binding sites for oxyrterols have now been identified in mammalian cells: the cytosolic oxysterol binding protein which binds 25hydroxy-cholerterol with high affinity(19,21.2')and the microsomal AEBS which binds 7-ketocholestanol, but not 25-hydroxycliolesteroI, with high affinity(35).One could speculate that different oxysterols might have different iiitracellular 'receptors' with which they interact to bring about different physiological effects. Do oxysterols play any role in human disease? There is little doubt that animals and humans are exposed continually to both exogenous oxysterols in the diet and to endogenous oxysterols produced in vivo. Is this

exposure harmful? In this regard, there is a substantial found to contain up to 12.3 '70 oxysterols not found in body of data implicating oxysterols in the pathogenesis fresh butter. of atherosclerosis. All these observations, taken together with reports It is now firmly established that the higher the plasma that oxysterols have been found in human atheroscleroconcentration of LDL, the greater the risk for tic lesions as well as in lipoprotein particled') and in developing atherosclerosis. The mechanism by which food("."), suggest that oxysterols are likely to play a role high LDL levels lead to the development of the fatty in the pathogenesis of atherosclerosis. Already. there streak and later to atheroma formation remains are indications that antioxidants (which suppress incomplctely understood. There is considerable evioxidation of LDL) might bc effective in the prevention dence that the LDL particle has to be oxidatively of atherosclerosis in animal studied4*). If a pathogenmodified before it can be rapidly taken up by etic role of oxysterols is established, efforts directed macrophages in the arterial wall, thereby converting towards reducing the oxysterol content of foodstuffs or these cells into the foam cells which are characteristic of the development of therapeutically useful antioxidants the fatty streak('*). It is believed that oxidation results may prove effective in reducing the incidence of in the fragmentation of unsaturated fatty acids in the atherosclerosis. LDL particle; the fragments in turn covalently modify With respect to the therapeutic potential of oxysterthe LDL apolipoprotein (apo B-100) rendering the 01s in human disease, three areas are being actively LDL particle recognizable by the 'scavenger' receptor explored. First, in view of the ability of oxysterols to of the macrophages. This allows rapid internalization of inhibit cholesterol biosynthesis, attempts are being the oxidized LDL particle by macrophages and the made to synthesize oxysterols capable of lowering serum cholesterol in the intact organism without formation of foam cells. The evidence that oxidation of producing significant toxic effects. Several 15-oxygeLDL particle occurs in vivo and is probably an nated sterols such as 5~-cholest-8(14-en-3/3-ol-15-one important step in atherogenesis has been thoroughly appear promising in this regard(@ . Secondly, the reviewed by Steinberg et al.(42). One consequence of the oxidation of LDL particle, cytotoxic property of oxysterols is being explored to which has received less attention, is the conversion of determine if it could be made to act selectively against some of its cholesterol content to oxy~terols(~~).cancer cells. Some oxysterols appear to be toxic to Whether this increased oxysterol content in the LDL malignant cells but non-toxic to their normal counterparticle contributes to atherogenesis is unclear. But it is p a r t ~ ( ~these ~ ) ; agents may be useful as chemotherapeutic agents against cancer. Thirdly, the immunosuppresknown that oxidized LDL, unlike native LDL, is sive properties of oxysterols clearly deserve to be cytotoxic to cells both in vitro and in viva("). It is examined further for their therapeutic potential in possible that this cytotoxicity of the oxidized LDL organ transplantation and the treatment of autoimmune particle is due, at least in part, to its increased oxysterol content and could cause endothelial injury which, in disorders. turn, would predispose to the development of atherosclerosis. Acknowledgements Some, but not all, studies on experimental atheroThe author thanks Miss Asha Das for her superb sclerosis support the possibility that oxysterols may be secretarial assistance and the National University of involved in a t h e r o g e n e s i ~ ( ~ Imai , ~ ~ ) .et U Z . ( ~ ) reported Singapore for its support. that purification of stock USP grade cholesterol prevented its atherogenicity when fed to rabbits, References whereas feeding of the impurities (which were largely 1 SMITH.L. L. (1Y87). Cholesterol autoxidation 1481-1986. Chem. Phys. oxysterol5 ) produced severe arterial damage. On their Lipids 44, 87-125. study in pigeons, Jacobson et al.(") observed that 2 SMITH.L. L. AND JOHNSON. B. H. (1489). Biological activities of oxysterols. pigeons fed cholesterol plus cholestan-3@,5a, 6@-triol Free Rud. Bid. Med. 7,285-332. 3 PENC,S. K . AVD TAYLOR, C. B. (1984). Cholesterol autoxidation, health and developed more severe coronary atherosclerosis than arteriosclerosis, Wld. Rev. Nutr. Diet. 44, 117-154. control pigeons given cholesterol alone. By contrast, 4 FINOCCHIARO, E. T.AND ~IICHARDSON; T. (1983). Sterol oxides i n foodstuffs: Higley et u Z . ( ~ ~ ) reported that rabbits given purified a review. J. Food Protect. 46, 917 -925. 5 PIE, J. E . , SPAHIS,K . AND SIXLLAN, C. (199U). Evaluation of oxidative cholesterol showed 6-fold more arterial lesions than degradation of cholesterol in food and food ingredients: identification and rabbits fed oxysterols. This discrepancy could possibly quantitation of cholesterol oxides. J . Agric. Food Cizem. 38. 973-979. 6 KANDCTSLH, A. A , , CHEN,H. W. AND HEINICLR.H. J . (1978). Biological be due to different oxysterols used in the different activity of some oxygenated sterols. Science 201, 498-501. studies and will require clarification. 7 BECK,J. P. AND CRASIESDE PAULEI,A , , eds (1988). Biological activities of Further support for an atherogenic role for oxysterols oxygenated sterols. Colloque INSERM. uol. 166, Paris. 8 BROWK, M. S. A N D GOLDSTEIN, J. L. (1986). A receptor-mediafed pathway comes from an epidemiologic study of Indian migrants fnr cholestcrol homeostasis. Science 232, 34-47. to London and to the West Indies who had higher than 9 KANJIIJTSCH, A . A. AND CHEN,H. W.(1973). Inhibition of aterol synthesis in expected morbidity and mortality from atherosclerosis cultured mouse cells by 7a-hydroxycholestcrol, 7/3-hydroxycholesterol and 7ketocholealerol. J. B i d . Chem. 248. X408-8417. but did not show the usual risk factors for this 10 KANIJUTSC~I. A. A,, HEINIGFR, H. J. AND CHEN,H. W. (1977). Effects of 25disorder(47).These people consumed large amounts of bydroxycholesterol and 7-ketocholesterol, inhibitors of sterol synthcsis, administered orally to mice. Biochim. Biophys. Acta 486. 260-272. ghee, a butter preparation used for cooking which was

:i'

11 KRIEGER, M., GIXDSTEIV,J. L. AND BROWN,M. S. (1978). Receptormediated uptake ol low-density-lipoprotein reconstituted with 25hydroxycholesteryl oleate suppresses 3-hydroxy-3-methylglutaryl-coenzyme A reductase and inhibits growth of normal fibroblasts. Froc. Nut1 Acud. Sci. U S A 75, 5052-5056. 12 CHANG.T-Y. (1983). Mammalian HMG-CoA reductase and its regulation. In: The Enzymes (edited by P. D. Boyer) XVI, 491-521. 13 GUFTA.A , , SEXTON.R . C. AXD R C D N ~ Y H,. (1986). Modulation of regulatory oxystcrol forniation and low density lipoprotein suppression of 3hydroxy-3-glutaryl coenzyme A (HMG-CoA) rcductasc activity by ketoconazole. A role for cytochrome P-450 in thc regulation of HMG-CoA reductase in rat intestinal epithelkal cells. .l. Riol. Chern. 261, 8348-8356. 14 CHAW,7 - Y . AUD CHANG. C. C. Y. (1982). Revertants of a Chinese hamster ovary cell mutant resistant to suppression by an analogue of cholesterol: isolation and partial biochemical characterization. Biochemi.srry 21.5316-5323. 15 GIBBONS.G. F. (1983). The role of oxysterols in the regulation of cholesterol biosynthesis. Biochem. SOC.Trans. 11. 649-651. 16 SAucrER: S. E.; KANDIITSCH, A. A.. GAYEN,A. K.. SWAHW, D. K. AND SPENCF.R, T. A. (1989). Oxyrterol regulators of 3-hydroxy-3-methylglutaryl coenzyme A reductase in liver. Effect of dietary cholesterol. J. Bid. Chem. 264, 6863-6869. 17 SAUCIER. S. E., KANDUTSCH. A. A , , T>iYLoR, F. R. , SPENCER, T. A,, PHIRWA,S. AND GAYEN, A. K. (1985). Identification of regulatory oxysterols, 24(S), 2.5-epoxychole~teroland 25-hydroxycliolesterol. in cultured fibroblasts. .I. Riol. Chrm. 260, 14571-14 579. 18 GIBBONS, G. F., PUT.l.IWGER, C. R.. CHEN, H. w., CAVENEE. W.K. AND KANDIJTSCH, A . A . (1980). Regulation of cholesterol hiosynthesis i n cultured cells by probable natural precursor sterols. J. Biol. Chem. 255: 395-400. 19 KANDIJTSCH, A. A , , CHEN,H . W. AND SHOWN, E. P. (1977). Binding of 25hydroxycholesterol and cholesterol to different cytoplasmic proteins. Proc. Narl h a d . Sci. USA 74, 2500-2506. 20 TAYLOR.F. R., SAUCIER, S. E., SHOWN,E. P . , PARISH.E. J. AND A. A. (1984). Corrclation bctween oxysterol binding to a cytosolic KANDUTSCH. binding protein and potency in thc repression of hydroxymethylglutaryl coenzyme A reductase. J . Bid. Chem. 259. 12382-12387. 21 DAWSON, P. A , , VAN DER WESrHUYZPN, D. R., GOLDSTEIN. J. L. AND BROWN,M. S. (1989). Purification of oxysterol binding protein from hamstcr liver cytosol. J. Biol. Chcm. 264, 9046-9052. 22 DAWSON, P. A , , RIDGWAY. N. D.. SLAUOHTER. C. A, . BROWN, M. S. AND GOLDSTEIN, J. L. (1989). cDNA cloning and expression of oxysterol-binding protein. an oligomer with a potential lcucine zipper. 1. Bid. Chenz. 264. 16798-16 803. 23 RAIAVASHTSTH, T. B . , TAVZOR, A. K., ANDALIRI, A , , SVENSOY, K. L. AND Lrisis, A. J. (1989). Identification of a zinc finger protein that binds lo the sterol regulatory element. Science 245, 640-642. 24 CHPN.H. W . , KANDUTSCH. 4. A. A N D W.wMomH, C. (1974). Inhibition of cell growth by oxygenated derivatives of cholesterol. Nuzure 251. 419-421. 25 BROWN,M. S. AND G O L DS T EJ. I~ L. , (1974). Suppression of 3-hydroxy-3methylglutaryl coenzyme A reductase activity and inhibition of growth of human fitroblasts by 7-ketocholesterol. J. Biol. Chem. 249, 7306-7314. 26 BROWN,M. S. AND GOLDSTLIE;, J . L. (1980). Multivalent feedback regulation of HMG-CoA reductase. a control mechanism co-ordinating isoprenoid synthesis and cell growth. J. Lipid Res. 21. 505-517. 27 SCHMIDT, R. A,, SCHNEIUER, C. J. AKU GLOMSET, J. A . (1984). Evidence for post-translational incorporation of mevalomic acid into Swiss 3T3 cell proteins. J . B i d . Chem. 259, 10175-10180. 28 SCHAFER, W . R., KIM. K.,STERNE, R. 'I'HORNER, J., KIM,S-H. AND RINE.J. (1989). Genetic and pharmacological auppression of oncogenic mutations in ras genes of yeast and humans. Science 245, 379-385. 29 HIETTER,H.. TRIFILIEFF,E., RICHERT,L., BECK,J. P., LUU, B. AND OIJRISSON, G. (1984). Antagonistic action cholesterol towards the toxicity of hydroxysterols on cultured hyptoma cells. Biochem. Biophys. Res. Cornmun. 120, 657-664. 30 PARISH, E. J., CHITRAKORN. s., I L I L I , B., SCHMIDT, G. 4hiU OURISSON, G. ~

(1989). Studies of the o.njsterol inhibition of tumor cell growth. Steroid.v 53, 579-596. 31 SUTHERLAND. R . L., MURPHY.L. C., Foo, M. S.. G R E E ~M. , D.. WHYBOURNE. A. M. AND KRoZOWSKI, z. s. (1980). High-affinity anti-oestrogen binding site distinct from the estrogen rcccptor. Nafure 288, 273-275. 32 LAZIER,C. B. AND BAP.~T, B. V. (1988). Antiestrogen binding sites: general and comparative propertiea. J. S t r ~ i dRiodwrn. 31. 665-669. 33 MURPHY, P. R., BRECKENRIDGE, W. c.A N D LA7IER>c. B. (1985). Binding Of oxygenated cholesterol metabolites to antiestrogen binding sites from chicken liver. Biochem. Biophys. Res. Comtnun. 127, 786-792. 34 HWANG. P. L. H . AND MARTIN,A. (1989). Interactions of sterols with anticstrogcn-hinding sites: structural requirements for high-affinity binding. J. Lipid Res. 30, 239-24.5. 35 HWANC, P. L. H . (1990). High-affinity binding sites for oxygenated sterols in rat liver microsomes: possible indentity with antiestrogen binding sites. Biochinz. Biophys. Acfu 1033. 154-161. 36 Lr r , L. .AND HWANG, P. L. (1991). Antiproliferative eflects of oxygenated sterols: positive correlation with binding affinities for the antiestrogen binding site. Biochinz. Biophys. Arm, in press. 37 TANG,B. L., Tr o , C. C.. SIM. K. Y., Nti, M. L. A W U KON. 0. L. (1Y89). Cytostatic effect of antiestrogens in lymphoid cells: relationship to high affinity antiestrogen-binding sites and cholesterol, Biochim. Biophys. Acta 1014, 162-172. 38 FARGIN. A , . BAYARD. F.. FAYF,J. C.. TRAORE. M.. POIROT. M., KLAEAE. A. A N D Pmm. J. .I. (1988). Further evidence for a biological role of anti-estrogenbinding sites in mediating the growth inhibitory action of diphenylmethane derivatives. Chem.-Bio[. Interactions 66, 101-109. 39 RAAPHORST. G. P., AZZAM.E. I.. LANGLOIS. R . AND VAN LIER,J. E. (1987). Effect of cholesterol N and 9, cpoxides on ccll killing and transformation. Biorhrm. Phurmacol. 36,2369-2372. 40 MOOG?C., JI, Y . H . , WALIZINGPR, C.,Luu, B. AND BISCHOII,P. (19Lx)). Studies on the immunological properties of oxysterols: in vivo actions of 7,25dihydroxycholesterol upon murine peritoneal cells. Immunology 70, 344-350. 41 TIPION,C.L., LLLNG, P. C., JOHNSON, J. S . . BROOKS.R. J. AND BEITZ,D. C. (1987). Cholesterol hydroperoxides inhibit calmodulin and suppress atherogcnesis in rabbits. Biochem. Biophys. Res. Commun. 146, 1166-1172. 42 STETNBERG. D., PATHASARATHY, s., CAREW. T. E . , KHOO, J. c. AND WITZTUM,J. L. (1989). Beyond cholesterol. Modifications of low-density lipoprotein that increase ils atherogenicity. N e w Engl. J. Med. 320, 915-924. 43 HUBBARD, R. W . , ONO,Y. A N D SANCHLZ, A. (1989). Atherogenic effect of oxidized products of cholesterol. Prog. Food Nufr. Sci. 13. 15-44. 44 IMAI.H..WEKIHESSEN.N. T., SUBR~MANYAM, V.. LEQUESNE,P. W., SULOWAY, A. H. m u KANISAWA, M. (1980). Angiotoxicity of oxygciiated steroids and possible precursors. Science 207, 651-654. 45 .lACOBSON, M. S., PRICE,M. G.. SHAMOO, A. E. AND HEALD,F. P. (1985). Atherosclerosis in white carneau pigeons. Effects of low-level cholesrane-trio1 feeding. Atherosclerosis 57. 209-217. 46 HIGLEY.N . A . , BEERY,J. T., TAYLOR, S. L., PORTER, J. W.. DZILJRA. J. A. AND LALICH, J. J. (1986). Comparative athcrogenic effccts of cholesterol and cholesterol oxidcs. Atherosclemvis 62. 91-104. 47 JwnRsoN. M. s. (1987). Cholesterol oxides in Indian ghee: possible cause of unexplained high risk ol atherosclerosis in Indian immigrant populations. Lancet 2, 656-658. 48 SCHKOEPCPK, G. J . (1981). Sterol biosynthesis. Annu. Rev. Biochem. 50, 585-621. 49 HIETTER,H . , BTSCHOFF, p., BECK.J . P.. OURISSON. G. AND L U U , B. (1986). Comparative eflects ol 7phydroxycholesterol towards murine lymphomas, lyniphoblarts and lymphocytes: selective cytotoxicity and blastogenesis inhibition. Cancer Biochem. Biophys. 9. 75-83.

1

Peter L. Hwang is at the Department of Physiology, National University of Singapore, 10 Kent Ridge Cresccnt, Republic of Singapore 0511.