Effect of squalestatin 1 on the biosynthesis of the mevalonate pathway lipids

et Biophysika &a Biochimicaet BiophysicsActa 1215(1994) 245-249

Effect of squa~estati~ 1 on the biosynthesis of the mevalonate pathway lipids Anders Thelin ay*,Elisabeth Peterson a, Julie L. Hutson b, Alun D. McCarthy b, Johau Ericsson a, Gustav Dallner a~c a ~e~~e~ of Bioche~~t~~ Stupor U~‘~er~~~S-106 Stoc~~m, Sweden b G%XO Group Research LTD., Greenford, Middlesev UB6OHE, UK

’ Clinical Research Center, Nouum, Karolinsku Institutet, S-14186 Huddinge, Sweden

Received5 July 1994

Abstract The effects of squalestatin 1 on rat brain and liver homogenates and on Chinese hamster ovary tissue culture cells have been investigated. This compound effectively inhibits squalene biosynthesis in a highly selective manner. Cytoplasmicfamesyl pyrophosphate

and geranylgeranyl pyrophosphate synthases are not affected, which is also the case for microsomal cis-prenyltransferase.In tissue culture cells, squalestatin1 inhibits cholesterol biosyn~esis completely, but does not alter do~chol synthesis or protein i~prenylation to a great extent. Incorporationof f3Hfmevalonateinto ubiquinone-9and -10 increases 3-4-fold, probably as a result of increased synthesis of thii lipid. Squalestatin 1 appears not only to be an effective inhibitor of cholesterol biosynthesis,but also to be more specific than other inhibitors used earlier in various in vitro and in vivo systems. Keywords: Squalestatin 1; Squalene synthase; Ubiquinone; Dolichol; Cholesterol; Protein isoprenylation

1. Introduction Inhibition of the mevalonate pathway and of cholesterol biosynthesis is accomplished at present by the use of various inhibitors of 3-hy~oxy-3-me~yl~uta~l coenzyme A reductase (HMG-CoA reductase) [ 11. The selectivity of these inhibitors is based on the idea that the first committed enzyme of cholesterol biosynthesis, squalene synthase, has a low affinity for farnesyl pyrophosphate (FPP), whereas the branch-point enzymes involved in dolichol and ubiquinone biosynthesis and protein isoprenylation are saturated even when the FPP pool size is decreased [2]. Recent experiments have demonstrated, however, that inhibitors of HMG-CoA reductase also influence a number of the other enzymes of the mevalonate pathway [3-61. The clinical interest of an inhibitor of the peripheral

portion of the pathway for cholesterol biosynthesis is obvious. Recently, it was found that the squalestatin 1 isolated from fungi is a potent inhibitor of squalene synthase, both in in vitro and in vivo systems [7-91. This raises the possibility of a more specific inhibition of cholesterol biosynthesis than can be achieved with inhibitors of HMGCoA reductase. Since the specificity of this type of inhibitor is of crucial importance, we have examined the possible effects of ~ualestatin 1 on some of the branch point enzymes of the mevalonate pathway, as well as on the final products formed in a tissue culture system. The synthesis of dolichol and ubiquinone and the isoprenylation of proteins is of basic importance in cellular metabolism and a modification in these processes would have serious consequences.

2. Materials and methods Abbreviations: IPP, isopentenyl pyrophosphate; GPP, geranyl pyrophosphate; GGPP, geranylgeranyl pyrophosphate; FFP, famesyl pyrophosphate; HMG-t&4, 3-hydroxy-3-methyl-~ut~l coenzyme A; CHO, Chinese hamster ovary. * Corresponding author. Fax: f46 815 36 79. 00052760/94/$07.00 0 1994 Elsevier Science B.V. All rights reserved SSDI OOOS-2760(94)00139-l

2.2. Chemicals Labeled iso~ntenol and all-~~~-f~esol were prepared by oxidation of the alcohol to the corresponding

Fig. 1. Structure of squalestatin

1.

aldehyde followed by reduction with [3H]sodium borohydride (Amensham, Great Britain, 16.5 Ci/mmol) according to Keenan and Kryczek, [lo]. R,S-[5-3H]mevalonolactone was prepared as described by Keller [ll]. Isoprenoid pyrophosphates were synthesized using the method described by Popjak et al. [12] Isopentenol is commercially available (Jansen Chimica, Belgium). Geraniol and all-Gina-famesol were kind gifts from Dr. T. Takigawa of the Kururay, Okayama, Japan. Polyprenol standards were prepared from Sorbus suecica as described previously [13]. The presqualene standard was a gift from Dr. D. Poulter, University of Utah, USA. Squalestatin 1 was isolated and characterized as described earlier [7,8]. The structure of this compound is shown in Fig. 1. 2.2. Animals Male Sprague-Dawley rats weighing 120 g were used to prepare liver homogenates, total membranes, supemat~t and microsomes, as well as membranes and supematant from the brain 114,151. 2.3. Tissue culture cells The tissue culture cells Chinese hamster ovary (CHO) variant met-18b-2 were kindly supplied by Dr. J. Faust of Tufts University, Boston, MA. The culturing conditions were as described earlier [16]. Squalestatin 1 was added to the cells on the fifth day, i.e., the time at which the culture became confluent. This inhibitor was present in the medium for 20 h and when the cells were labeled, 500 &i R,S-[5-3H]mevalonate (specific activity 10 Ci/mmol) per 10 ml medium was added for two additional hours. The cells were removed from the plates by trypsinization and then homogenized by short sonication, followed by ultracent~fugation at 1~000 X g for 60 min in order to separate total membranes from supernatant. 2.4. Incubations Famesyl pyrophosphate (FPP) synthase activity was assayed in a 300 ~1 incubation mixture containing 1 mM MgCl,, 10 mM KF, 1 mM dithiothreitol, 20 mM imidazole-Cl (pH 7.01, 50 PM geranyl pyrophosphate (GPP) and 20 PM [3H]isopentenyl pyrophosphate (IPP) (0.15 Ci/mmol). The reactions were started by the addition of supernat~t protein and were allowed to continue for 10 min.

Squalene synthase was measured in the same incubation mixture, but with 75 PM [“H]FPP (54 Ci/mol) instead of IPP and GPP as substrate. This reaction was started by the addition of membrane protein and allowed to proceed for I h at 37°C. For determination of cis-prenyltransferase activity. mcubations were again performed in the same medium but with 100 mM KF, 7.5 PM FPP (instead of GPP) and 1% Triton X-100. The reaction was started by the addition of membrane protein and allowed to proceed for 1 h at 37°C. All trans-geranylgeranyl pyrophosphate (GGPP) synthase was measured in a 300-~1 incubation mixture which contained 25 mM imidazole-Cl (pH 6.01, 1.0 mM dithiothreitol, 100 mM KF, 20 FM [“HIIPP (0.15 Ci,/mmoll. and 50 ,uM truns,trans-FPP. The reactions were started by the addition of supematant protein and allowed to proceed for 60 min at 37°C. 2.5. Lipid extraction and chromatography The assay mixtures were extracted twice with 2 ml water-saturated n-butanol[17]. These organic extracts were combined and evaporated under a stream of nitrogen. For dephosphorylation, the samples were dissolved in 200 ~1 5% n-octyl-&glycopyranoside, followed by the addition of 1.6 ml 0.15 M Tris-Cl (pH 6.51, and 2 mg wheat germ acid phosphatase in 200 ~1 double-distilled water [lS]. The reaction was allowed to proceed for 16 h and the resulting isoprenoid alcohols were extracted with 2 X 3 ml diethylether/petroleum ether, 1:l. The organic phases were combined and evaporated under nitrogen and the residue was dissolved in 30 ~1 methanol/2-propanol/n-hexane, 2:l:l (solvent A). The products were analyzed in an HPLC system equipped with a reversed phase column (HewlettPackard Hypersil ODS Cl81 and a radioflow detector (Radiomatic). Short-chain products (ClO-C30) were separated using a linear gradient starting with methanol/water, 7:3. and te~inating with methanol/water, 9:l (solvent B) in a 30-min run (gradient A). Long-chain products (C30Cl151 were separated using a linear gradient starting with solvent B and finishing with solvent A in a 30-min run (gradient B). Standards were monitored on the basis of their absorption at 210 nm. 2.6. Protein prenylation For analysis of protein isoprenylation the radiolabeled cells were subjected to extensive lipid extraction [19] and the protein residue resolved by SDS-PAGE [20], followed by autoradiography. 2.7. Dot blot Dot-blot analysis was performed as described earlier (211. The antisemm against FPP synthase was kindly supplied by Dr. P. Edwards, UCLA, Los Angeles, CA.

247

Table 1 Effects of squalestatin 1 on some mevalonate pathway enzymes in microsomes and supemat~t Activity

Enzyme

FPP synthase a GGPP syntbaseb FPP synthaseC GGPP synthased Cjs-Frenyl~~sfe~se Squalene synthase’

*

Control f

Squalestatin 1 f

364f 38 8Irt 9 545 5 3524 26 IlOi 12 9240&1060

354f30 so* to Sl& 5 3&l&32 97* 9 37;t 4

a Liver supematant, pmol/ pg protein per h. b Liver supematant, pmol/mg protein per h. ’ Brain supematant, pmol/ pg protein per h. d Brain supematant, pmoI/mg protein per h. e Liver mi~r~mes, pmoI/mg protein per h. f The values are the means+S.E. of 5 e~~eflts.

2.8. Chemical measurements

1 2 3 Fig.2 Isoprenylated proteins in squalestatin l-treated CHO cells. Cells

Protein was determined by the Lowry procedure using bovine dbumin as standard f22].

were p~~ubated in the presence (lane 11 or absence flane 2) of 2 pM squalestatin 1 for 20 h. ~~H]Mev~o~ate was then added and the cells harvested 2 h later. After exhaustive lipid extraction, the protein residue was subjected to SDS-PAGE and autoradiography. The numbers indicate the molecular weights of the markers used (lane 3).

3. Results Me~urement of the enzyme activities in the vicinity of the branching point of the mevalonate pathway may be performed using homogenate or fractions prepared from the homogenate. In addition to squalene synthesis by the membrane fraction, we have also assayed the first committed enzyme of dolichol synthesis, c~-pr~nyl~ansfer~e. The two final steps of the initial pathway, the synthesis of FPP and GGPP, were measured in the supematant fraction. In liver and brain supematant both FPP and GGPP synthesis were unaffected by the presence of 2 PM squalestatin 1 (Table 1). Supematant from rat brain demonstrates relatively high GGPP synthase activity compared to the rate of FPP formation, probably because this organ shows a high rate of protein isoprenylation and relatively low rate of lipid synthesis, particularly of cholesterol [23,24]. Squalestatin 1 at a concentration of 2 PM completely inhibited squalene synthesis by rat liver microsomes, but did not decrease the ~~-prenyl~~sferase activity. It was necessary to increase the concentration of this inhibitor loo-fold (to 200 FM) before inhibition (20-40%) of the enzymes measured could be obtained (not shown). Protein isoprenylation with FPP and GGPP is a basic event in cellular regulation and is required for the activation of many G-proteins [25]. For this reason the tissue culture cells were incubated with f31-I]mevaIonate and covalendy linked isoprene residues remaining after extensive lipid extraction analysed by SDS-PAGE. As seen in Fig. 2, a group of proteins in the hf, range of 20-30 kDa are associated with radioactivity. These proteins have been identified earlier as GTP-binding proteins. In agreement

with previous investigations, proteins with higher molecular weights were also found to be labeled [26]. In the squalesta~-seated cells the pattern of labeling was un-

Retention time (m~utes) Fig. 3. Cbromatographic profile from liver homogenate incubated in the presence of 13H$nevalonate. Rat liver homogenates were incubated in the absence (A) and presence fBI of 0.1 PM squalestain 1. The products were extracted with butanol, subjected to enzymatic dephospborylation and analyzed using IIPLC. The arrows indicate the position of standards used to identify mevalonate pathway products. F: farnesol; G: geranylgeraniof; P: presqualene; C: cholesterol; S: squaIene; 9: ubiquinone-3; IO: ubiquino~e-~0.

A. Thelin et al. /Biochimica et Biophysics Acta 1215 (1994) 245-249

248

Table 2 Effects of squalestatin 1 on [3H]mevalonate terol, ubiquinone and dolichol in CHO cells Squalestatin

1 ( PM)

None 0.5 1 2

incorporation

into choles-

% of control a Cholesterol

Ubiquinone

Dolichol

100 (1900) b 38 8 1.5

100 (9.2) ’ 286 387 345

100 (6.1)’ 118 100 86

a The values are the means of four experiments. b The values in brackets represent lipid labeling (dpm/

pg protein per h).

changed compared to the control but there was a considerable variation in the intensity of labeling of individual proteins in different experiments. Some of the protein bands of the squalestatin l-treated cells exhibited lower intensities than those of the untreated cells but the picture varied in individual experiments. In spite of these variations, the conclusion was that protein isoprenylation is not affected to a great extent by this inhibitor. Upon incubations with [ 3H]mevalonate tissue homogenates exhibit metabolic labeling of lipids and lipid intermediates. After incubation of liver homogenate with [3H]mevalonate both FPP and GGPP were found but the dominant products were cholesterol and its metabolites presqualene and squalene (Fig. 3A). In this system, which was not supplemented with other precursors and cofactors, there was also some synthesis of both ubiquinone-9 and -10. When these incubations were performed in the presence of squalestatin 1, the peaks of presqualene, squalene and cholesterol completely disappeared (Fig. 3B). Large accumulation of FPP occured but the amount of GGPP was unchanged. This latter finding is in line with the lack of effect on protein isoprenylation, since the only known function of GGPP is as a substrate for protein prenyltransferase. Interestingly, a 3-fold increase in the labeling of ubiquinone-9 and -10 in the homogenate was apparent in the presence of the inhibitor. Two additional peaks (at 13 and 15 min) appeared in this chromatogram which were not identified. Probably they are intermediates in the biosynthesis of ubiquinone. In addition to isoprenylation and cholesterol biosyn-

Table 3 Activity of mevalonate pathway enzymes of CHO cells treated with squalestatin 1 Enzyme

cis-Prenyltransferase Squalene synthase FPP synthase GGPP synthase

in membranes

and supematant

dpm/ Kg protein per h a Control

Squalestatin

120+ 6 2485 24 106500 + 9400 267+ 19

168* 14 329* 36 27 400 + 3800 152* 13

a The values are the means+S.E.

of 5 experiments.

Fig. 4. Dot blot analysis of supernatant FPP synthase from CHO cells. CHO cells were grown in the presence (1) or absence (2) of squalestatin 1 for 20 h, and were thereafter subjected to sonication followed by ultracentrifugation at 100000X ,g for 60 min. Dot blot was performed on the supernatant fraction thus obtained using antisera against FPP synthase as described in Section 2. After autoradiography quantitation was performed by densitometry. Bar 3,4 as 1,2 but the supernatant was diluted IO-fold. The bars represent the mean + S.E. of 5 rats.

thesis the other major functions of the mevalonate pathway are the synthesis of ubiquinone and dolichol. In order to examine the possible effects of the inhibitor on branches of the mevalonate pathway other than that leading to cholesterol, we studied the labeling of end-products in tissue culture cells (Table 2). As is the case in most other systems, labeling of cholesterol with [ “H]mevalonate exceeded the labeling of ubiquinone and dolichol by 200300-fold. Squalestain 1 at a concentration of 0.5 PM inhibited the cholesterol synthesis in CHO cells by 60% and total inhibition was achieved with 2 PM. Dolichol biosynthesis was slightly decreased, which was not the case for ubiquinone. Labeling of this latter lipid upon squalestatin 1 treatment increased 3-4 times similar to what was observed using liver homogenate. The individual branch-point enzymes in cultured cells treated with the inhibitor were also measured (Table 3). The membrane fraction prepared from treated CHO cells exhibited moderately increased cis-prenyltransferase and squalene synthase activities. The supernatant fraction from the same cells displayed a somewhat decreased GGPP synthase activity, in contrast to FPP synthase, which in the treated cells was as low as 25% of the control. This decrease in FPP synthase activity raised the question as to whether this enzyme in the treated cells was inhibited or down-regulated. In order to quantitate the amount of enzyme protein, we performed dot-blots using anti-FPP synthase antisera (Fig. 4). Increasing concentrations of supernatant protein from the control cultured cells resulted in increased precipitation, a picture similar to that seen with the supernatant fraction from treated cells. Thus, this experiment indicates that the amount of FPP synthase in the treated cells was not decreased.

1

4. Discussion Squalestatin 1 is a most interesting inhibitor of cholesterol biosynthesis, since beyond the branching point no

A. Thelin et al. / Biochimica et Biophysics Acta 1215 (1994) 245-249

inhibitor with su~cient specificity and low toxicity in vivo is yet available. Inhibitors of HMG-CoA reductase affect a number of enzymes of the mevalonate pathway, in addition to inhibiting ubiquinone biosynthesis [5,6,27]. Squalestain 1 was a very efficient inhibitor of cholesterol synthesis in the tissue and cell culture systems we have studied here and the inhibition at the presqualene level indicates an inhibition of the head-to-head condensation of FPP units. The high specificity of this inhibitor is demonstrated by its lack of influence on FPP and GGPP synthesis, protein prenylation, cis-prenyltransferase and ubiquinone and dolichol biosynthesis. After continuous treatment in vivo mevinolin is known to up-regulate not only the reductase, but also several enzymes of the cytosol, microsomes and peroxisomes. In our systems no significant up-regulation of the enzymes measured by squalestatin 1 could be found. FPP synthase activity was greatly decreased upon trea~ent of the cells and precipitation with antisera demonstrated that this decrease did not reflect changes at the transcriptional level. Probably, this effect reflects product inhibition of the enzyme. A very interesting effect of squalestatin 1 is the 3-4-fold increase obtained in ubiquinone labeling. Since the size of the FPP pool is increased several-fold [9], one would expect a dilution of the labeled precursor and, thus lowered labeling of ubiquinone. One possible explanation for the increased labeling seen is that, in spite of the generally established assumptions, the ~~~~~-prenyltr~sferase is not saturated in the control cells. It appears more probable, however, that in treated cells the rate of ubiquinone biosynthesis is increased, which in a long-term experiment would result in an elevation in the amount of ubiquinone. Since our experiments involved short-time incubations, they provide no info~ation concerning the effect of squalestatin 1 on the amount of this lipid. Interest in ubiquinone has increased greatly since it was recently found that this lipid is not only a mitochondrial redox component, but is also our only endogenous antioxidant, not originating from the diet 1281.Decreases in tissue levels of ubiquinone have been observed in ageing, cardiomyopathy, muscle diseases, oxidative stress and carcinogenesis [29,30]. No dietary conditions or drug treatments for increasing the level of this lipid are yet available. If after long time treatment in vivo squalestatin 1 proves to decrease blood cholesterol levels and increase blood and tissue levels of ubiquinone, one may have a drug with unusually advantageous properties, Acknowledgements

This work was supported by the Swedish Medical Research Council and the Ring Gustav V and Queen Victoria Foundation.

249

References [l] Endo, A. (1992) J. Lipid. Res. 33, 1569-1582. [2] Goldstein, J.L. and Brown, MS. (1990) Nature 343, 425-430. [3] Folkers, K., Langsjoen, P., Willis, R., Richardson, P., Xia, L.-J., Ye, C-Q. and Tamagawa, H. (1990) Proc. Natl. Acad. Sci. USA 87, 8931-8934. [4] I.&w, P., Andersson, M., Edlund, C. and Dallner, G. (1992) Biochim. Biophys. Acta 1165, 102-109. [S] Ericsson, J., Runquist, M., Thelin, A., Andersson, M., Chojnacki, T. and Dallner, G. (1993) J. Biol. Chem. 268, 832-838. [6] Teclebrhan, I-I., Olsson, J., Swiezewska, E. and Dallner, G. (1993) J. Biol. Chem. 268, 23081-23086. [7] Dawson, M.J., Farthing, J.E., Marshall, P.S., Middleton, R.F., Oneill, M.J., Shuttlewo~h, A., Stylli, C., Tait, R.M., Taylor, P.M., Wildman, H.G., Buss, A.D., Langley, D. and Hayes, M.V. (1992) J. Antibiot. 45, 639-647. [8] Sidebottom, P.J., Highcock, R.M., Lane, S.J., Procopiou, P.A. and Watson, N.S. (1992) J. Antibiot. 45, 648-658. [9] Baxter, A., Fitzgerald, B.J., Hutson, J.L., McCarthy, AD., Motteram, J.M., Ross, B.C., Sapra, M., Snowden, M.A., Watson, N.S., Williams, R.J. and Wright, C. (1992) J. Biol. Chem. 267, 1170% 11708. [lo] Keenan, R.W. and Kruczek, M.E. (1976) Biochemistry 15, 15861591. [ll] Keller, R.K. (19861 J. Biol. Chem. 261, 12053-12059. 1121 Popjak, Cl., Comforth, J.W., Cornforth, R.H., Ryhage, R. and Goodman, D.S. (1962) J. Biol. Chem. 237, 56-61. 1131Chojnacki, T. and Vogtman, T. (1984) Acta Biochim. Pol. 31, 115-126. [14] Runquist, M., Ericsson, J., Thelin, A., Chojnacki, T. and Dallner, G. (1992) Biochem. Biophys. Res. Cornmutt. 186, 157-165. 1151Ericsson, J., Thelin, A., Chojnacki, T. and Dallner, G. (1992) J. Biol. Chem. 267, 19730-19735. 1161Faust, J. and Krieger, M. (1987) J. Biol. Chem. 262, 1996-2004. [17] Ericsson, J., Appelkvist, E.L., Thelin, A., Chojnacki, T. and Dallner, G. (1992) J. Biol. Chem. 267, 18708-18714. [18] Carson, D.D. and Lennarz, W.J. (1981) J. Biol. Chem. 256, 46794686. [19] Thelin, A., I&w, P., Chojnacki, T. and Dallner, G. (1991) Eur. J. Biochem. 195, 755-761. [ZO] Laemmli, U.K. (1970) Nature 227,680-685. 1211Hawkes, R. (19861 Methods Enzymol. 121,484-491. [22] Lowry, O.H., Rosebrough, N.J., Farr, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193, 265-275. 1231Schaber, M.D., O’Hara, M.B., Garsky, V.M., Mosser, S.C., Bergstrom, J.D., Moores, S.L., Marshall, M.S., Friedman, P.A., Dixon, R.k and Gibbs, J.B. (1990) J. Biol. Chem. 265, 1470114704. 1241Audersson, M., Elmberger, P.G., Edhmd, C., Kristensson, K. and Dallner, G. (1990) FEBS Lett. 269, 15-18. [25] Sinensky, M. and Lutz, R.J. (1992) BioEssays 14, 25-31. [26] Clarke, S. (1992) Annu. Rev. B&hem. 61, 355-386. 1271Runquist, M., Ericsson, J., Thelin, A., Chojnacki, T. and Dallner, G. (19931 J. Biol. Chem. 269,5804-5809. 1281Beyer, R.E. and Ernster, L. (1990) in Highlights in ~iquinone Research (Lenaz, G., Bamabei, O., Rabbi, A. and Battino, M., eds.), pp. 191-213, Taylor and Francis, London. 1291Kalen, A., Appelkvist, E.L. and Dallner, G. (1989) Lipids 24, 579-584. [30] Eggens, I., Ekstrom, T.J. and Aberg, F. (19901 J. Exp. Pathol. 71, 219-232.