Effects of mevinolin treatment on tissue dolichol and ubiquinone levels in the rat

102 8 1992 Elsevier

Biocfzimicu et Riuphy.sim Acta, 1 165 ( 1902) 103- I OY B.V. Ail rights reserved nOnS-276(l/uz/%oS.O~)

Science Publishers

BBALTP 54039

Effects of mevinolin treatment on tissue dolichol and ubiquinone levels in the rat

(Revised

(Received manuscript

6 April 1992f received 2Y July lYY2)

Key words: Cholesterol; Dolichyl monophosphate:

Polyisoprenoid: Isoprene:

Liver: Brain: (Rat)

Rats were treated with mevinolin by intraperitoneal injection (15 days) or dietary administration (30 days). The cholcstcrol, dolichol, dolichyl phosphate and ubiquinone contents of the liver, brain, heart, muscle and blood were then investigated. The cholesterol contents of these organs did not change significantly, with the exception of musk. Intrapcritoncal administration of the drug increases the amount of dolicho1 in liver, muscle and blood and decreases the dolichyl-P amount in muscle. The same treatment increases the level of ubiquinone in muscle and blood and decreases this value in liver and heart. Oral administrations decreases doiichol. dolicby~-P and ubiquinone levels in heart and muscle, while in liver the dolichoi lcvcl is elevated and ubiquinone level lowered. In brain the amount of dolichyl-P is incrcascd. I~traperit(~neai injection of mcvinolin also modifies the liver dolichol and dolichyl-P isoprenoid pattern, with an increase in shorter chain polyisoprcncs. The levels of dolichol and ubiquinone in the blood do not follow the changes observed in other tissues. Incorporation of [‘Hlacctatc into cholcstcrol by liver slices prepared from mevinolin-treated rats exhibited an increase, whereas in brain no change was seen. Labeling of dolichol and ubiquinone was increased in both liver and brain, hut incorporation into dolichyl phosphate remained relatively stable. The results indicate that mevinolin affects not only HMG-CoA reductase but, to some extent, also affects certain of the peripheral cnzymcs, resulting in considerable effects on the various mevalonate pathway lipids.

introduction The initial part of the mevalona~e pathway begins with the condensation of acetyl units and results in the synthesis of farnesyl-PP, the common precursor for cholesterol, dolichol, ubiquinone and protein isoprenylation. This pathway is known to be regulated at the level of HMG-CoA reductase, but it has been established that additional peripheral regulation of the synthesis of the various lipids also occurs [I]. There are a number of inhibitors of cholesterol biosynthesis which affect both the initial and terminal portions of the pathway, but only inhibitors of HMG-CoA reductase are sufficiently non-toxic for in viva use 121. These inhibitors are now widely used for lowering blood levels of VLDL and LDL cholesterol.

Correspondence to: P. L6w, Department of Biochemistry, Stockholm University. 106 91 Stockholm, Sweden. Abbreviations: HMCi, 3-hydroxy-3-methylglutaryl; VLDL, very low density lipoprotein; LDL, low density lipoprotein; IP, intraperitoneal; C/M/W, chloroform/methanol/water; HPLC, high performance liquid chromatography.

A number of wets-established experimental facts demonstrate that, in spite of the existence of an additional regulatory step at the level of squalene synthetase, the activity of HMG-COA reductase is the most important regulator of cholesterol biosynthesis [3]. This conclusion is also supported by the fact that the cholesterol level can be effectively lowered by inhibitors of this enzyme. Since the precursor for dolichol and ubiquinone biosynthesis and for protein isoprenylation is also produced by HMG-CoA reductase, the question arises as to what extent such inhibitors influence the levels of these other products. From a medical point of view it is desirable to selectively influence cholesterol biosynthesis, wjthout affecting other products of the mevalonate pathway. There are only a few investigations which deal with the effects of inhibitors of HMG-CoA reductase on dolichol and ubiquinone biosynthesis. In vitro studies with liver microsomes led to the conclusion that long chain cis-prenyl transferase, the major enzyme in polyprenol biosynthesis, has a high affinity for farnesylPP, obviously exceeding the affinity of squalene synthetase for this same compound [4]. This finding is in

103 agreement with the experiment of Tavares et al. [5], in which decreased cholesterol synthesis in animals fed cholesterol was parallelled by an increase in dolichol biosynthesis. Using human fibroblasts it was found that compactin decreased cholesterol biosynthesis, but increased the incorporation of [ “Hlmevalonate into ubiquinone [6]. In various experiments involving dietary manipulation and drug treatment, dolichol and ubiquinone synthesis were affected in a manner opposite to that of cholesterol biosynthesis [7,8]. Dietary mevinolin was reported recently to decrease the ubiquinone content of rat liver and heart [9]. In humans it was reported that compactin did not influence serum ubiquinone levels [lo]. Pravastatin treatment of patients with hypercholesterolemia resulted in a decrease in the elevated ubiquinone blood levels, which, however, did not fall below the control value, whereas blood dolichol levels were unchanged [ 111. Cholesterol is partly taken up from the diet and partly synthesized de novo, mainly by the liver and transported and distributed via the blood. This lipid is utilized by a number of peripheral organs [12]. In contrast, dolichol and ubiquinone are not taken up from the diet to any great extent 113-151. These lipids are synthesized in all tissues [16] and are not redistributed via the circulation, where their concentrations are lOOOO-fold lower than that of cholesterol 1171. Consequently, it may be inappropriate to use serum dolichol and ubiquinone levels as a measure of the tissue biosynthesis of these lipids. An increase or decrease in the synthetic capacities of liver and other organs may not necessarily cause any change in the blood values, which appear to be relatively constant. It appears to be of interest to study the effects of inhibitors of HMG-CoA reductase on ubiquinone and dolichol biosynthesis in the liver and determine the influence of such treatment on the amounts of these lipids in other tissues. Such studies are best performed in vivo, since the results here may differ considerably from those obtained with in vitro systems, including tissue cultures. Materials

and Methods

Chemicals Mevinolin (lovastatin) was kindly supplied by Dr. A.W. Alberts (Merck, Sharp and Dohme, Rahway, NJ, USA). [3H]Acetate was purchased from the Radiochemical Centre (Amersham, UK). Dolichol-23 was isolated from human pituitary gland [18] and phosphorylated according to Danilov and Chojnacki [19]. Both of these lipids were used as internal standards. The internal standard for the analysis of cholesterol was ergosterol (Steraloids, Wilton, USA) and for ubiquinone ubiquinone-6 (Sigma, St. Louis, USA). HPLC solvents from E. Merck were used.

Animals In all experiments male Sprague-Dawley rats weighing 180-200 g (90 days old) were used. The animals were either injected intrapertioneally with mevinolin suspended in 75% ethanol (5-25 mg/ml) or fed with rat chow mixed with mevinolin (500 mg/kg). The controls for the rats injected with mevinolin received alcohol alone. Comparison between untreated rats and those injected with alcohol, showed no differences in lipid content or incorporation of radioactivity. The animals received no food or drug during the 12 h immediately before they were decapitated and blood and organs removed. The weights of the liver, brain and heart were not significantly different in treated rats and controls. Tissues were homogenized in water (1 g/5 g) in an Ultra Turrax blender and supplemented with known amounts of the appropriate internal standards. Heparin was added to whole blood samples which were analyzed without separation of cells and plasma. Preparation of slices After removal, the liver or brain was cut into 0.5 x 0.5 x 1.0 cm pieces and placed in cold Krebs-Henseleit buffer. Slices of 0.5 mm thickness were subsequently prepared with a Polaron H 1200 Vibrating microtome (Biorad, Watford). These slices were incubated in conical flasks containing Krebs-Henseleit buffer supplemented with 5 mM glucose. This system was then preincubated for 15 min before [“HIacetate was added (2 mCi/O.S g t issue). During the incubation the mixtures were maintained at 37”C, exposed to carbogen gas and shaken slowly. Lipid analysis For dolichol, dolichyl phosphate and cholesterol analyses, the samples were mixed with 1 volume methanol and 0.5 volume 60% KOH (w/v>. The samples were then hydrolyzed for 45 min at 100°C. Chloroform and methanol were added to obtain a final C/M/W ratio of 3 :2: 1. The water-rich phase was then removed and the lower chloroform-rich phase was washed three times with Folch upper phase [20]. Dolichyl phosphate was separated from the neutral lipids using a DEAE-Sephadex ion exchanger [21]. The two lipid fractions were analyzed with HPLC. When investigating ubiquinone, the samples were extracted with chloroform/methanol, 2 : 1, without alkaline hydrolysis and charged lipids were removed using a silica column [22]. Reversed-phase HPLC The chromatographic separations were performed on a Shimadzu LC4A system using a Waters WISP autoinjector. The sample was injected onto a HewlettPackard Hypersil ODS Cl& 3 pm column with a size

104 of 60 x 4.8 mm. For neutral lipids a linear gradient was used from the initial m~thano~/water, 9 : 1, in reservoir A to 100% methanol/Z-propanol/hexane, 2: 1: 1, in reservoir B. The flow rate was 2 ml/min and the program time 33 min. The eluate was monitored at 210 nm, except when analyzing ubiquinone, in which case monitoring was performed at 275 nm. In the case of dolichyl-P a convex gradient (setting - 1) was run from water/methanol/2-propanol, 5 : 60 : 40, to 40% 2-propanol/hexane, 3: 7, both solvents containing 20 mM phosphoric acid, during a period of 17 min and at a flow rate of 2 mI/min. Al1 the peaks were well resolved and q~antitation was based on integration of peak areas, using internal standards to compensate for sample losses. The recovery was 65-85%. When radioactive samples were analyzed, the individual peaks were collected and the solvents evaporated. Scintillation cocktail was then added and the samples counted. Internal standards were used to compensate for sample losses in this case as well.

TABLE

Tissue

I

Lipid cholesterol

Liver Brain Heart Muscle Blood

3.1 i 0.2 16.2i_ 1.1 1.1 -F-o.2 0.9 i 0.2 2.4 * 0.4

h

dolichol 60.7 1h.S 7.5 3.6 0.00

’

dolichyl-P

k7.1 +1.x + 1.2 to.6 * 0.02

’

18.4 k2.3 7.7 * 0.‘) 2.2 *0.7 2.5 +0.6 0.1 Y* 0.04

ubiquinone

’

152 _+I7 52.6 i 6.1 236 +27 26.1 i_ 3.2 i.2* 0.2

” Values ’ pmol/g

are the means $ S.D. of 7-X experiments. wet weight tissue. or @mot/g blood. nmol/g wet weight tissue, or nmoljg blood.

total amount of this lipid is phosphorylated. Ubiquinone distribution differs from that found for other lipids, i.e., it is present at a high level in heart, lower levels in liver and, on a gram basis, only 10 and 20% as much in muscle and brain, respectively. ChoCesterol

Results

In the present study four organs, i.e., liver, brain, heart and muscle were examined and blood was also analyzed (Table I). The cholesterol content of brain is very high, a well known fact from earlier investigations [23]. The dolichol contents of various organs increase during the lifetime of the animals [24]. In this study rats at an early adult age (2 months) were used. The liver dolichol content is 4, 8 and 16-fold higher than in the brain, heart and muscle, respectively. A considerin all abie portion of the dolichol is phospho~lated organs. In muscle and brain as much as 30-40% of the

The effects of mevinolin treatment on cholesterot levets in various organs are limited. Daily treatment for 15 days with intraperitonea~ (IP) injections of 1, 3 or 5 mg mevinolin or oral administration (500 mg/kg chow) for 30 days did not influence the cholesterol contents of the organs or of the blood, with the exception of muscle (Fig. 1). In muscle, high doses of mevinolin given IP resulted in a 3O-40% increase in cholesterol content. Dolichol and dolichyl phosphate

In contrast to the case with cholesterol, mevinolin had pronounced effects on dolichol distribution and El IZZI

control 1 mg mevinoliniday 3 mg mevinolinlday 5 mg mevinolin/day

liver

brain

heart

muscle

blood

Fig. 1. Effects of mcvinolin treatment on the cholesterol content of organs and blood. Rats were treated by daily intraperitoneal injection (0, I, 3 or 5 mg mevinolin) for 15 days or supplied ad libitum with a diet containing 500 mg mevinolin/kg chow for 30 days. Control animals received the basic rat chow. After these time periods, samples of blood, liver, brain, heart and pieces of psoas muscle were removed and their lipid contents analyzed. The percentage changes compared to control values were calculated. The values shown are from 7 or 8 experiments. The S.D. varied between 3 and 16% and are not shown in the figure for the sake of clarity. * These values differ significantly from the control (P < 0.05).

105

0 lz!

control 1 mg mevinolin/day

3 mg mevinolin/day 5 mg mevinolin/day

-

liver

brain

Fig. 2. Effects of mevinolin

heart treatment

on dolichol

muscle content.

patterns, effects dependent on the organ examined and the route of drug administration. The dolichol content of liver increased extensively with all treatments employed and with 3 mg mevinolin IP this content was doubled (Fig. 2). When 5 mg was injected, the dolichol content decreased somewhat. In brain and blood the effects were limited, but both heart and muscle responded quite specifically to different treatments. In both these organs oral administration resulted in a 20-25% decrease in the dolichol content, whereas in muscle IP injection caused an increase. The highest concentration of mevinolin injected IP increased muscle dolichol content by 100%. The response of dolichyl phosphate levels to inhibitor treatment was more limited than the response of the free alcohol (Fig. 3). No large changes were observed in liver and brain. Heart and muscle dolichyl

blood

For details,

see the legend to Fig. 1

phosphate contents were decreased, especially after oral administration, which lowered the amount of this lipid as much as 30-40%. Most interestingly, mevinolin treatment not only influenced the polyisoprenoid contents of various tissues, but also influenced the distribution pattern, particularly in liver. The levels of dolichols with 17 and 18 isoprene units were increased, while those with 19, 20 and 21 units were decreased (Fig. 4A). The changes were more pronounced with higher concentrations of the inhibitor and were found with both routes of administration. In the case of dolichyl phosphate the modifications in the isoprenoid pattern were similar to those described for the free alcohol, although more accentuated (Fig. 4B). The changes were as great as 100% for certain isoprenes and this pattern was very similar in all experiments.

u

control 1 mg mevinolin/day 3 mg mevinolin/day 5 mg mevinolin/day

liver Fig. 3. Tissue dolichyl

brain phosphate

heart contents

after mevinolin

muscle treatment.

blood For details,

see the legend to Fig. 1

106 cant increase, particularly with the highest dose injected IP. The changes in muscle were very similar to those found for dolichol, i.e., oral administration decreased the ubiquinone level by 20%, whereas IP injection for 15 days elevated this level to as much as 250% of the control value.

B

17

I8

20

19

21

Fig. 4. Pattern of individual polyisoprenoids in the liver after intraperitoneal or oral administration of mevinohn. (A) The individual dolichols were separated by reversed phase HPLC and the total amount of dolichol (17-22 isoprene units) was taken as 100%. (B) Dolichyl phosphates 17-21 were separated as described in Materials and Methods and values calculated as in A.

Ubiquinone Mevinolin treatment influenced the ubiquinone content of all tissues investigated, with the exception of brain, where the amount of this lipid was constant (Fig. 5). In liver and heart there was a moderate decrease in the content of this lipid, to levels between 70 and 90% of the control values. In the blood there was a signifi-

Incorporation of /-iH/acetate Tissue slices of constant thickness were prepared from liver and brain and incubated with [jH]acetate in order to estimate the rate of the biosynthesis of these different lipids (Table II). Since the animals had received no drug during the 12 h immediately prior to incubation, the tissue concentration of mevinolin was low [2.5]. When liver slices were incubated with this precursor and the extent of labeling in isolated cholesterol determined, a IO-fold increase in labcling was found in livers of animals fed 0.05% mevinolin for 30 days. The rates of incorporation into ubiquinone and dolichol also exhibited a three-fold increase. The effect of mevinolin on labeling of dolichyl-P was quite limited. In the case of brain the response was different. Neither the cholesterol nor dolichyl-P level was affected by mevinolin treatment. Incorporation of radioactivite precursor into ubiquinone was doubled and incorporation into dolichol increased by 4-fold. In both liver and brain an increased labeling of shorter dolichols was observed (data not shown). Discussion The liver, particularly in the rat, is the major organ for the synthesis of cholesterol. Inhibitors of HMG-CoA reductase affect the hepatic enzyme, which is the reason why these substances are utilized as medical drugs. It is plausible to suppose that the inhibition is not

cl

control 1 mg mevinolin/day 3 mg mevinolin/day 5 mg mevinolin/day

liver Fig. 5. Ubiquinone

brain contents

heart

of blood and tissues after mevinolin

muscle treatment.

blood For details,

see the legend to Fig.

I

107 TABLE

II

Incorporation of [3H]acetate lipids using tissue slices

into liL>erand brain meualonate pathways

Slices from tissues of rats fed 0.05% mevinolin for 30 days and from control animals were incubated in the presence of 0.5 mCi [‘H]acetate/ml for 45 and 90 min. The homogenates were then extracted according to Materials and Methods and the lipids were isolated by HPLC and radioactivity determined. The values are the means + SD. of values obtained with slices from five rats. Lipid

Liver, cholesterol Liver. ubiquinone Liver, dolichol Liver, dolichyl-P Brain, cholesterol Brain, ubiquinone Brain, dolichol Brain, dolichyl-P

Incubation time (mitt)

45 90 45 90 45 90 45 90 4s 90 45 90 45 90 4s 90

Radioactivity (dpm/g tissue x lo-“) control

treated

450 i50 750 k60 6.7i 0.11 12.2* 1.1 2.4k 0.4 5.0 k 0.6 3.2* 0.3 6.1 * 0.9 17.0* 2.2 30.2k 3.6 O.hi 0.1 1.7* 0.2 0.9* 0.1 2.5k 0.3 1.5* 0.3 2.7+ 0.4

4570 +430 9490 k660 18.1 * 2.1 40.5* 3.9 7.4* 0.8 10.2* 1.2 4.1 k 0.5 9.2* 1.2 15.1 k 2.3 27.5k 3.3 1.3* 0.2 2.8k 0.3 5.2i 0.7 8.8k 1.0 1.7+ 0.3 2.3k 0.4

specific for the liver and that it also affects other lipids of the mevalonate pathway. These questions were studied here by analyzing the products of the mevalonate pathway, not only in the liver, but also in certain other organs. In our studies with male rats, mevinolin treatment had no significant influence on the cholesterol contents of the various organs, with the exception of muscle. In vitro incorporation of L3H]acetate into cholesterol by liver slices was increased after mevinolin treatment as a result of the induction of HMG-CoA reductase, in agreement with previous investigations [261. In brain induction of cholesterol synthesis was not observed. The differences among various tissues can probably be explained by variations in the uptake of mevinolin by the organs. Previous investigations have established that this drug is present at highest concentration in the liver, but appears to a variable extent in a number of other tissues as well 125,271. After mevinolin treatment, the dolichol level increases in liver and, after IP administration, also in muscle. On the other hand, there is a decrease in the muscle and heart levels after peroral administration. No experiments have been performed in the present or in previous studies to determine the efficiency of drug uptake by the liver using different routes of administration. Obviously, the concentration of this drug in the

liver varies, since its effect on dolichol and ubiquinone biosynthesis are dose-dependent. Inhibition of HMGCoA reductase reduces the mevalonate pool, which affects primarily cholesterol biosynthesis. Dolichol synthesis is not sensitive to changes in pool size, since long chain cis-prenyl transferase, the major enzyme in polyprenol biosynthesis, has a high affinity for its substrate and appears to be saturated under various conditions [4]. Consequently, increased cis-prenyl transferase activity elicited by the increased substrate concentration (farnesyl-PP and isopentenyl-PP) cannot explain the elevation in dolichol level. In addition, in the case of protein isoprenylation it was found that mevinolin inhibits the transfer of farnesol in vitro, a finding which has, however, no in vivo significance, since the affinity of the transferase for this substrate is even higher than that of the cis-prenyltransferase participating in dolichol biosynthesis [281. It appears probable that mevinolin has a direct effect on the condensation reactions participating in the elongation of the polyisoprenoid chain. Alternatively, mevinolin treatment influences the biosynthesis of the transferase, thereby increasing its amount. In contrast to the free alcohol, the dolichyl-P content in the liver is unaffected by mevinolin treatment, but this content decreases in heart and also somewhat in muscle. These findings strengthen the idea that the biosynthesis of these two forms of the lipid are regulated separately, which may be particularly important under certain conditions 129,301. An alternative explanation for these findings could be that dolichyl-P is also synthesized more rapidly, but that a constant level is maintained by a concommitant increase in the activity of dolichyl phosphatase. A large decrease in dolichyl-P content may be incompatible with continued cellular function, since this lipid may be rate-limiting in protein glycosylation [31,32]. An interesting effect of mevinolin treatment were the decreases in the average chain lengths of dolichol and dolichyl-P. In previous in vitro studies it was found that the concentration of substrate, such as mevalonate, isopentenyl pyrophoshate and farnesyl pyrophosphate, present in the incubation medium modulates the isoprenoid pattern of newly synthesized polyprenols [29,33,341. Furthermore, the condition of the membrane and the level of Mg2+ play regulatory roles [341. An increased chain length was found in liver after in vivo treatment of rats with di(2-ethylhexyl) phthalate [35] and in liver preneoplastic noduli [36]. In human hepatocellular carcinoma a decrease in chain length could be observed 1371. The functional consequences of such a decrease in the chain length of dolichyl-P is not at present known. Previous investigations concluded that shorter dolichyl-P species are also effective as sugar carriers in N-linked protein glycosylation 1381.

108

Mevinolin treatment has interesting effects on the level of ubiqui~o~e, which decreases in liver and heart and increases or decreases in muscle, depending on the route of drug administration. In addition, the blood level of this lipid is increased. Since the synthesis of both dolichol and ubiquinone utilize the same substrate pool and since these two processess are affected in opposite manners by the same treatment, the most probable explanation for the decreased ubiquinone content is a direct effect of the inhibitor on transprenyitransferase. Previously, it was found that mevinolin affects protein synthesis and degradation, primarily that associated with the membranes of the endoplasmatic reticulum [39,40]. Mevinolin either inhibits rrans-prenyitransferase and/or decreases its de novo synthesis. We have no explanation for the increased rate of ubiquinone biosynthesis after intraperitoneal injection and for the decreased rate after peroral administration. It is possible that differences in the metabolites produced may explain these opposite effects. Labeling with [ “HIacetate demonstrated that HMG-CoA reductase is induced during this treatment, giving rise to increased 1abeIing in cholesterol when the inhibitor was removed. This induction is not sufficient to explain the increased in~o~oration into liver dolichol and ubiquinone and the unchanged incorporation into dolichyl-P. The terminal portions of the biosynthetic pathways for these compounds, involving cisand trans-prenyltransferase as the major enzymes, are also affected and the enzymes seem to be regulated, or at least affected, by the inhibitor. Mevinolin is present in a high concentration in the liver, but it is also taken up, although to considerably lower extent, by the brain [25]. Additional evidence that this drug passes the blood-brain barrier is provided by direct measurements of the mevinoli~ concentration in cerebrospinal fluid, which was found to be ten times less than in the serum [41]. The presence of the drug in the brain is also demonstrated by the finding that the rate of incorporation of [3H]acetate into dolichol and ubiquinone was altered by mevinolin treatment. Both of these lipids are synthesized locally and cannot be supplied by way of the circujation, in contrast to cholesterol. Dietary and newly synthesized cholesterol are transported and redistributed to a large extent through the circulation. ConsequentIy, the blood cholesterol level reflects the changes occurring in the liver. On the other hand, dolichol and ubiquinone are synthesized in all organs, taken up from the diet to a limited extent only and not redistributed via the circulation [16]. Their presence in the blood probably serves specific purposes. Very small amounts of ubiquinone, a thousand times less than in the case of cholesterol, are secreted to the blood by the liver which would seem to speak against a general, less specific function for this lipid. For example, ubiquinone serves as an endogenous an-

tioxidant for lipoproteins and its blood level is regulated independently from tissue levels 142,431. Therefore, measurement of dolichol and ubiquinone in the blood may not necessarily reflect the IeveIs of these lipids in various tissues, which, in fact, was found to be the case in this study. It appears that mevinolin influences mainly, but not exclusively, the HMG-CoA reductase of various organs. In addition to this enzyme, both long-chain cisand trans-prenyl transferases may also be affected, which could lead to an increase or decrease in dolichol and/or ubiquinone contents, depending on the experimental ~nditions. The inhibitor may exert a direct effect on these enzyme proteins or, alternatively, cause gene regulation of enzyme levels. In order to clarify the exact mechanism(s) of inhibition of the various portions of the mevalonate pathway, it will be necessary in the future to perform detailed investigations on the purified enzymes, on the enzyme levels in various organs and on the genetic regulation of these enzymes.

The valuable technical assistance of Kristina Hoimberg is gratefully acknowledged. This work was supported by the Swedish Cancer Society. References 1 2 3 4 5 6 7 8 9 10

11

12

13 14 15

Goldstein, J.L. and Brown MS. (1990) Nature 343, 425-430. Alberts, A.W. (1988) Am. J. Cardiol. 62, 105-155. Rudney, H. and Sexton R.C. (1986) Annu. Rev. Nutr. 6, 245-272. Ericsson, J.. The&n, A., Scallen, T.J. and Daltner G. (19911 FASEB J. 5, 798. Tavares, LA., Cc&bear, T. and Hemming, F.W. (19811 Arch, Biochem. Biophys. 207.427-436. Faust, J.R., Goldstein J.L. and Brown MS. (1979) Arch. B&hem. Biophys. 192, X6-99. Appelkvist, E.-L. and Dallner G. (1987) in Pharmacology (Rand, M.J and Raper, C., eds.), pp. X57-861. Elsevier, Amsterdam. Kalin, A., Appelkvist E.-L. and Dallner, G. (1990) Chem. Biol. Interact. 73, 221-234. Willis, R.A., Folkers, K., Tucker, J.L., Ye, C.Q., Xia, L.J. and Tamagawa, H. (1990) Proc. Natl. Acad. Sci. USA 87, 892X-8930. Mabuchi, H., Haba, T., Tatami, R., Miyamoto, S., Sakai, Y., Wakasugi, T., Watanabe, A., Koizume, J. and Takeda R. (1981) N. Engl. J. Med. 305.478-482. Elmberger, PG., Kalen, A., Lund, E., Reihnkr. E., Eriksson, M.. Berglund, L., Angelin, B. and Datlner G. (1991) J. Lipid Res. 32, 935-940. Turley, F.D. and Dietchy, J&4. (1988) in The Liver, Biology and Pathobiology (Arias, LM., Jakoby, W.B., Popper, I-i. and Schachter, D., eds.), pp. 617-641, Raven Press, New York. Chojnacki, T. and Dallner. G. (1983) J. Biol. Chem. 258, 916-922. Keller, R.K., Jehle, E. and Adair, W.L. (1982) J. Biol. Chem. 257. 8985-8989. Appelkvist, E.-L., KalEn, A. and Dallner, G. (1991) in Biomedical and Clinical Aspects of &enzyme Q (Folkers, K., Littarru, G.P. and Yamagami, T., eds.), Vol. 6, pp. 141-150, Elsevier, Amsterdam.

109 16 Elmberger, P.G., KaI&n, A., Ap~I~ist, E.-L. and Dallner, G. (1987) Eur. J. Biochem. 168, 1-11. 17 Elmberger, P.G., Engfeldt, P. and Dallner, G. (1988) J. Lipid Res. 29, 1651-1662. 18 Radominska-Pyrek A., Chojnacki, T. and Pyrek, J.S. (1979) Biochem. Biophys. Res. Commun. 86, 395-401. 19 Danilov, L.L. and Chojnacki, T. (1981) FEBS Lett. 131, 310-312. 20 Folch, J., Lees, M. and Stanley, G.H.S. (1957) J. Biol. Chem. 226, 497-509. 21 Behrens, N.H. and TCbora, E. (1978) Methods Enzymol. 50, 402-435. 22 Elmberger, PG,, Eggens, I. and Dallner, G. (1989) Biomed. Chromatogr. 3, 20-28. 23 Rouser, G., Kritchevsky, G., Yamamoto, A. and Baxter, CF. (1972) Adv. Lipid Res. IQ, 261-360. 24 Kalin, A., Appelkvist, E.-L. and Dallner, 6. (1989) Lipids 24, 579-584. 25 Bocan, T.M.A., Ferguson, E., McNally, W., Uhlendorf, P.D., Mueller, S.B, Dehart, P., Sliskovic, D.R., Roth, B.D., Krause, B.K. and Newton, R.S. (1992) Biochim. Biophys. Acta 1123, 133-144. 26 Bjorkhem, I. (1986) Biochim. Biophys. Acta 877, 43-49 27 Germershausen, J.I., Hunt, V.M., Bostedor, R.G., Baily, P.J., Karkas, J.D. and Alberts, A.W. (1989) Biochem. Biophys. Res. Commun. 158, 667-675. 28 Reiss, Y., Seabra, M.C., Armstrong, S.A., Slaughter, C.A., Goldstein, J.L. and Brown, MS. (1991) J. Biol. Chem. 266, 1067210677.

29 Ekstrom, T.J., Chojnacki, T. and Dallner, G. (1987) J. Biol. Chem. 262, ~9~-4097. 30 Keller, R.K. (1987) Trends Biochem. Sci. 12, 443-445. 31 Carson, D.D. and Lennarz, W.J. (1979) Proc. Nat]. Acad. Sci. USA 76, 5709-5713. 32 Potter, J.E., James, M.J. and Kandutsch, A.A. (1981) J. Biol. Chem. 256, 2371-2376. 33 Sagami, H., Lennarz, W.J. and Ogura, K. (1989) Biochim. Biophys. Acta 1002, 218-224. 34 Matsuoka, S., Sagami, II., Kurisaki, A. and Ogura, K. (19911 J. Biot. Chem. 266, 3464-3468. 35 Edlund, C., Ganning, AX. and Dallner, G. (1986) Chem. Biol. Interact. 57, 255-270. 36 OIsson, J.M., Eriksson, L.C. and Dallner, G. (1991) Cancer Res. 51, 3774-3780. 37 Eggens, I., Elmberger, P.G. and Liiw, P. (1989) Br. J. Exp. Pathot. 70, X3-92. 38 Low, P., Peterson, E., Mizuno, M., Takigawa, T., Chojnacki, T. and Dallner, G. (1986) Biosci. Rep. 6, 677-683. 39 Sinensky, M. and Logel, J, (1983) J. Biol. Chem. 258. 8547-8549. 40 Li, A.C., Tanaka,R.D., Callaway, K., Fogelman, A.M. and Edwards, P.A. (1988) J. Lipid Res. 29, 781-796. 41 Botti, R.E., Triscari, J., Pan, H.Y. and Zayat, J. (1991) Clin. Neuropharmacol. 14, 256-261. 42 Frei B., Kim, M.C. and Ames, B.N. (1990) Proc. Natl. Acad. Sci. USA 87,4879-4883. 43 Stocker R., Bowty, V.W. and Frei, B. (1991) Proc. Natl. Acad. Sci. USA 88, 1646- 1650.