Solubilization, Purification, and Characterization of a Truncated Form ...

THEJOURNALOF BIOLOGICAL CHEMISTRY 0 1992 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 267, No. 12, Issue of April 25, pp. 8628-8635,1992 Printed in U.S.A.

Solubilization, Purification, and Characterization of a Truncated Form of Rat Hepatic SqualeneSynthetase* (Received for publication, November 15, 1991)

Ishaiahu ShechterSO, Ety Klingerll, MaryL. Rucker)),Robert G . EngstromII, Judith A. Spiritoll, M. A. Islam11,Brian R. Boettcher 11, and David B. Weinstein From the $Eleanor Roosevelt Institute, Denver, Colorade 80206, the TDepartment of Biochemistry, Tel Aviv University, Tel Aviv 69978, Israel, and the IlSandoz Research Institute, East Hanover, New Jersey07936

Rat hepatic microsomal squalene synthetase (EC mevalonate pathway. The substrate for the reaction, FPP,is 2.6.1.21) wasinduced 26-fold by feeding rats with dietnot only utilized for the production of sterols and their prodcontaining the hydroxymethylglutaryl-coenzymeA re- ucts but alsoserves as substratefor other prenyl transferases. ductase inhibitor, fluvastatin, and cholestyramine, a Thus,repetitive cis additions of isopentenyldiphosphates bile acid sequestrant. A soluble squalene synthetase result in the formation of 2,3-dehydrodolichyl diphosphate, protein with an estimated mass of 32-36 kDa, as de- whereas trans polymerization results in nonaprenyl diphostermined by gel filtration chromatography on Sepha- phate, an intermediate in ubiquinone synthesis. cry1 S-200 column, was solubilized out of the microThe identificationof Rhodotorucine A-farnesylated mating somes by controlled proteolysis with trypsin. Approxfactor (Kamiya et al., 1979) and the observation that mevaimately 26%of the activity wasrecovered in a soluble lonate-derived products are incorporated cellular into proteins form. The enzyme was purified to homogeneity utilizing a series of column chromatography purification (Schmidt etal., 1984) led to the identificationof a number of steps on DEAE-cellulose, hydroxylapatite, and phenyl- thioether-linked farnesyl and geranylgeranylmoieties with Sepharose sequentially. The purified enzyme showed proteins (Goldstein and Brown, 1990; Glomset et al., 1990). a single band on sodium dodecyl sulfate-polyacryl- Some of these prenylated proteins such as the ras proteins amide gel electrophoresis. Initial kinetic analysis in- play a major role in regulation of normal cellular processes dicated an So.B values for trans-farnesyl diphosphate and in oncogenesis. The localization of squalene synthetase . V,,, with in this central branch pointof the mevalonate pathway, and of 1.0 M M and for NADPH of 40 p ~ The respect to trans-farnesyl diphosphate was calculated the observation that its activity is suppressed in cultured at 1.2pmol/min/mg. NADHalso serves as substrate for fibroblasts grown in the presence of low density lipoproteins the reaction with So.8 value of 800 PM. Western blot led to the hypothesis that its regulation plays a major role in analysis utilizing rabbit antisera raised against the the diversion of the flux of intermediates to either the sterol purified, trypsin-truncated enzyme showed a single or the nonsterol pathway (Faust etal., 1979). band for the isolated solubilized enzyme at 32-33 kDa The reaction catalyzed by squalene synthetase connects the and a band for the intact microsomal enzyme at about two cellular compartments within which the sterol biosyn46-47 kDa. thetic pathway is localized. FPP is produced from mevalonate through a sequence of reactions which utilize and produce cytosolic intermediates whereas squalene and all subsequent The formationof squalene from trans-farnesyl diphosphate intermediates in the productionof sterols are localized in the (FPP)’ is catalyzed by squalene synthetase (farnesyldiphos- endoplasmic cell membrane. The hepatic enzyme has not yet phate:farnesyldiphosphate farnesyltransferase, EC 2.5.1.21) been purified, and the number of proteins required for the in two separate steps. The first is the condensation of two two-step reaction is notknown. Yeast has been theorganism of choice for the studies and purificationof squalene synthemolecules of FPP to form an intermediate presqualene diphosphate through a unique dimerization reaction (Rilling tase. Early attempts to solubilize this enzyme employed deand Epstein, 1969; Poulter and Rilling, 1981). This interme- oxycholate as detergent(Shechterand Bloch, 1971). This diate is then reductively rearranged in a second step for the solubilization was later repeatedby Agnew and Popjak (1978) yeast-derivedenzyme was reaction to form squalene using a reduced pyridine nucleotide usingthesamedetergent.The finally purified to homogeneity by the use of a mixture of as a reductant. nonionic detergents and chromatography purification (KuSqualene synthetase is localized at a branch point in the swik-Rabiega and Rilling, 1987; Sasiak and Rilling, 1988). * This work was supported, in part, by American Heart Association The purified yeast enzyme is a single polypeptideof M , 47,000, Grant 89-614. The costs of publication of this article were defrayed anditcatalyzesthesynthesis of squalenefrom FPP via in part by the payment of page charges. This article must therefore presqualene diphosphate. Recently, the molecular cloning of be hereby marked “advertisement” inaccordance with 18 U.S.C. the yeast squalene synthetase gene was achieved either by Section 1734 solely to indicate this fact. functional complementation of a squalene synthetase-defi8 To whom correspondence shouldbe addressed et 1991) or by screening a genomic The abbreviations used are: FPP, farnesyl diphosphate; DTT, cient mutant (Jennings al., dithiolthreitol; PMSF,phenylmethylsulfonylfluoride; SBTI, soy library (Handler etal., 1991). bean trypsin inhibitor; CHAPS, 3-[(3-~holamidopropyl)diWe report here a technique for the proteolytic release of a methylammonio]-1-propanesulfonate; HMG-CoA, 3-hydroxy-3- truncated, soluble, active fragment of rat hepatic microsomal methylglutaryl coenzyme A; BSA, bovine serum albumin; SDSto withPAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; squalene synthetase and its purificationhomogeneity FPLC, fast protein liquid chromatography; HEPES, 4-(2-hydroxy- out the use of detergents. The catalytic andphysical properties of this soluble enzyme are discussed as well. ethyl)-1-piperazineethanesulfonic acid.

8628

Purification of Rat Hepatic Squalene Synthetase EXPERIMENTALPROCEDURES

Materials Radiolabeled Compounds-In early experiments FPP was synthesized in one of our laboratories (IS). Tran~-[~H]farnesol (32.6 mCi/ mmol) was prepared according to themethod of Shechter and Bloch (1971) and was pyrophosphorylated by the method of Popjak (1969). Trar~-[~H]farnesyl diphosphate was purified according to Gafni and Shechter (1979).The experimental results obtainedby the use of this radiolabeled substrate are reported in dpm of radioactivity. In later experiments, [3H]FPP (300,OO disintegrations/minute of l-[3H]farnesyl diphosphate, triammonium salt, 143 mCi/mM) was purchased from New England Nuclear. ~~-[2-'~C]Mevalonate (51 mCi/mmol) and radiolabeled 2-["C]HMG-CoA (R,S) was purchased from the Radiochemical Center, Amersham Corp. ["C]Squalene was biosynthesized in our laboratory from ["C]mevalonate according to the method of Popjak (1969) with slight modifications (Shechter et al., 1980). [14C]-2,3-Oxidosqualenewas biosynthesized in our laboratory by the same procedure as ['4C]squalene except that plant growth retardant Amo 1618 (2-isopropyl-4-(trimethylammoniumchloride]5-methylphenyl-piperidine-1-carboxylate) was added to inhibit 2,3oxidosqualene-lanosterolcyclase (Douglas and Paleg, 1972). Specific radioactivity f w both labeled squalene and 2,3-oxidosqualene was determined experimentally by gas-liquid chromatography. Non-radiolabeled Materials-The following materials were purchased from Sigma: NADH, NADPH, glucose-6-phosphate, glucose 6-phosphate dehydrogenase, 5-bromo-4-chloro-indolylphosphate, nitroblue tetrazolium, 4-chloro-l-naphthol, HMG-CoA, mevalonic acid, bovine serum albumin, SDS, Coomassie Brilliant Blue R and G, cholestyramine, Amido Blue-Black, and Ponceau S. Phenyl-Sepharose 4B-CL was obtained from Pharmacia LKB Biotechnology Inc. and DE52 from Whatman. Goat anti-rabbit alkaline phosphatase conjugate, avidine alkaline phosphatase conjugate, and prestained molecular mass standards were obtained from Bio-Rad. Lovastatin was a gift of Dr. A. Alberts from Merck, Sharp and Dohme Laboratories. Fluvastatin (Sandoz compound XU 62-320) was provided by Sandoz Pharmaceutical. Nitrocellulose paper was purchased from Gelman or Bio-Rad. Non-radioactive FPP was provided by Dr. Larry Perez, Sandoz. Other chemicals and reagents were of analytical grade and purchased from commercial sources. Methods Preparation of Normal and Squalene Synthetase-induced Rat Liver Microsomes-Male 170-190-g Wistar Charles River rats were fed a powdered normal Purina rat chow diet and kept on an altered light cycle from 7 p.m. to 7 a.m. (normal rats). After 7-10 days, the rats were fed the powdered diet with the addition of 5% cholestyramine for 3 days followed by the powdered diet plus 5% cholestyramine and 0.15% of either lovastatin (LO diet) or fluvastatin (FL diet) for 3 more days. The rats were then sacrificed; their livers were removed and quickly cooled in saline. The remaining purification procedures were carried out at 4 "C. The livers were blotted, weighed, minced, and washed in2 volumes (2 ml of buffer/g of tissue) of buffer containing 0.3 M sucrose, 10 mM HEPES, pH 7.4, 1.0 mM DTT, 1.0 mM EDTA (buffer A). When necessary, the protease inhibitors 1.0 mM PMSF, 10 p M leupeptin, and 1.0 p M pepstatin (hereafter called proteaseinhibitors) were added to buffer A, as mentioned under "Results." 2 volumes of buffer A were added, and the livers briefly homogenized with a Polytronand thenwith a moderately tight fitting Teflon/glass Elvehjem homogenizer. After centrifugation of the homogenate for 5 min at 3,000 rpm in a Sorval SS-34 rotor, the speed was increased to 9,000 rpm and continued for 15 min. The surface lipids were then aspirated, and the supernatant removed and recentrifuged at 105,000 X g for 1 h. After aspiration of surface lipids, the supernatant was collected (SI& the pellet was washed three times by suspension in a volume of buffer B (20 mM potassium phosphate, pH 7.4, 1.0 mM DTT, 1.0 mM EDTA with or without the protease inhibitors) at a ratio of 1:l (v/w) followed by homogenization and recentrifugation a t 105,000 X g for 30 min. After the third wash, the microsomes were again resuspended in buffer B to a volume equal to the original total liver weight and frozen in portions at -80 'C. The S106preparation was centrifuged in an Airfuge (Beckman) for 10 min at 160,000 X g (SI,) before it was used for assay. The soluble fractions of livers as well as the corresponding microsomal preparations were stored in liquid NP. Freeze-Thaw Solubilization of Microsomal Squalene SynthetaseMicrosomes prepared without protease inhibitors were homogenized

8629

with a glass/Teflon homogenizer. The suspension was frozen in liquid NSfor 10 min, thawed a t 37 "C, and furtherincubated at 37 "C for 10 min. During that incubation the microsomal suspension was homogenized twice. The resulting suspension was centrifuged at 105,000 X g for 1 h. The freeze-thaw supernatant extract was separated from microsomes and stored a t -20 "C. The microsomal pellet was resuspended in the same buffer and freeze-thaw treated as described above for an additional four to five cycles. Trypsin Solubilization of Squalene Synthetase from Microsomes125 ml of frozen FL microsomes, prepared in the presence of protease inhibitors, (protein concentration= 17.6 mg/ml) were thawed, diluted 1:l with 20 mM potassium phosphate, 2.0 mM DTT, adjusted to pH 7.6, and centrifuged for 30 min at 105,000 X g. The pellet was suspended in 10 mM potassium phosphate, pH7.6, with 1.0 mM DTT, homogenized with a glass/Teflon Elvehjem homogenizer, and the volume was adjusted to 125 ml (protein concentration= 12.2 mg/ml). Trypsin, Sigma Type XIII, ~-l-tosylamide-2-phenylethylchloromethyl keton-treated (7.6 mg in 1.0 ml of 1.0 mM HCl) was added, and thesample was incubated at 15 "Cfor 2 h with occasional stirring. To monitor the progress of the release of squalene synthetase during the trypsinization procedure, 1-ml samples were removed at 0, 1, and 2 h, and 20 pl of soy bean trypsin inhibitor (SBTI, 38.1 mg in 2.50 ml of 1.0 mM HCl) was added to give a 5-fold excess of SBTI compared to trypsin. Portions of the samples of the various time periods were centrifuged for 5 min at 160,000 X g in a Beckman airfuge prior to assay. After 2 h of trypsinization, 2.44 mlof SBTI solution was added to the remaining 122 mlof the suspended microsomes followed by centrifugation for 30 min at 105,000 X g. Protease inhibitor mixture supernatant (96 ml) to a final concentration of was added to the SIOS 1.0 mM PMSF, 10 p M leupeptin, and 1.0 p M pepstatin, and the pH was adjusted 8.0. This trypsin-solubilized enzyme preparation was used for further purification. DEAE-Cellulose Ion-Exchange Chromatography-150 g of Whatman DE52 anion-exchange cellulose was equilibrated with 100 mM potassium phosphate, pH 8.0, and packed into a 2.6 x 35-cm Pharmacia C10 column. The column was washed with 10 mM potassium phosphate, pH 8.0, containing 25% ethylene glycol. The column was then washed with 500 mlof buffer containing 10 mM potassium phosphate, pH 8.0, 1.0 mM DTT, 1.0 mM EDTA and 25% ethylene glycol (buffer C) followed by 100 mlof buffer C plus protease inhibitors. This and subsequent chromatography was performed using a Pharmacia FPLC system controlled by an LCC500 plus liquid chromatography controller. The 96-ml S l o b of trypsin-solubilized enzyme was applied onto the DE52 column using a peristaltic pump at a flow rate of approximately 1 ml/min. From this step on, all buffers used for the purification of the enzyme contained protease inhibitors. The sample was eluted from the column at 1.0 ml/min by first washing with 200 ml of buffer C followed by 500 ml of linear salt gradient up to 0.8 M NaC1. Fraction size was 10 ml. The elution was monitored for protein at 280 nm, and squalene synthetase activity was determined for the different fractions. Fractions containing the peak of activity (-0.15-0.23 M NaCl) were pooled for separation on a hydroxylapatite column. Hydroxylapatite Chromatography-A 1.6 X 28-cm Pharmacia C16 column was packed with hydroxylapatite (Bio-Rad Bio-Gel HTP) and equilibrated in buffer D (10 mM potassium phosphate, pH 6.8, 1.0 mM DTT and 20% ethylene glycol). The 75 mlpool of active fractions from the DE52 column was applied to the hydroxylapatite column using a peristaltic pump a t a flow rate of approximately 1 ml/min. The enzyme was eluted from the column at 1.0 ml/min with 150 ml of Buffer D followed by a linear salt gradient of 600 ml up to collected. 0.50 M potassium phosphate. Fractions of10mlwere Elution was monitored for protein at 280 nm, and squalene synthetase activity was assayed for the different fractions. Fractions containing the peak of activity (-0.22-0.24 M potassium phosphate) were pooled for separation on phenyl-Sepharose. Phenyl-Sepharose Chromatography-A 1 X 8-cm Pharmacia C10 column was packed with phenyl-Sepharose (Pharmacia) and equilibrated in buffer E (20 mM potassium phosphate, pH 7.4,1.0 mM DTT, 1.0 mM EDTA, and 20% saturation of ammonium sulfate). The 18.5-ml pool of the most active hydroxylapatite fractions was diluted 1:l with 40% saturated ammonium sulfate solution at pH 7.4 to give 20% saturation in ammonium sulfate and approximately 12.5% ethylene glycol. The diluted sample was applied to thephenyl-Sepharose column at 0.5 ml/min. The enzyme was eluted a t 0.5 ml/min using 20 ml of buffer E (high salt) followed by a 40-ml decreasing salt and increasing ethylene glycol linear gradients to a final buffer solution containing 20 mM potassium phosphate, pH 7.4, 1.0 mM DTT, 1.0

8630

Purification of Rat Hepatic Squalene Synthetase

mM EDTA, and 50% ethylene glycol (buffer G). Elution continued with an additional 20mlof buffer G, and fractions of 1.0 ml were collected. Elution was monitored for proteinat 280 nm, and enzymatic activity was determined for the various fractions. Fractions containing the peak of activity were pooledand concentrated to one-fifth the volume using an Amicon ultrafiltrationapparatus with aPM-10 membrane. Superose 12 Chromatography-When necessary, purified squalene synthetase obtained from the phenyl-Sepharose column was further purified on Superose 12. A 30-cm Superose 12 column (Pharmacia) was equilibrated with buffer H (20 mM potassium phosphate, pH 7.4, 1.0 mM DTT, 1.0 mM EDTA, 25% ethylene glycol, and 500 mM KCl). 120 pl of the ultrafiltration concentrate of the active fractions eluted from phenyl-Sepharose was diluted to 300 p1 with buffer H, 200 pl of the diluted enzyme were applied to andchromatographed through the Superose 12 column a t 0.25 ml/min. Fractions of 0.2 mlwere collected. Elution was monitored for protein at 280 nm, and enzymatic activity was determined with the microwell assay described below. Enzymatic Assays-Assays for HMG-CoA reductase, squalene epoxidase, and 2,3-oxidosqualene-lanosterolcyclase were performed as described earlier (Eilenberg et al., 1989). Squalene synthetase was assayed as follows. Reactions were carried out in 96-well V-shaped polystyrene microtitration plates. The reactions in total volume of 100 pl contained 100 mM potassium phosphate, pH7.4,5.0 mM MgCl,, 5.0 mM CHAPS, 10 mM DTT, 2.0 mM NADPH, and 10 p~ [3H]FPP. Enzyme concentrations in the assays were adjusted so that less than 10% conversion of the substrate was catalyzed. Incubations were performed in covered plates for 20 min at 37 “C. The reaction was stopped with 10 pl of1.0 M EDTA, pH 9.2. In addition, 10 pl of unlabeled 0.5% squalene in ethanol was then added as carrier. 60 pl of the reaction (50%of the final volume) were applied onto analytical 1.8 X 20-cm, 250-pm thick, Silica Gel-G thin layer chromatography plates whichweredevelopedby 5% toluene in hexane for 13 cm. Radioactivity in zones containing the squalene (Rf= 0.5) werescraped and measured by scintillation spectrometry. The enzymatic activity was reported eitherindisintegration/minute or in nanomoles of squalene formed/minute/milligram of protein takinginto account the loss of one radiolabeled proton during the reaction (Poulter and Rilling, 1981). Protein Assay-The Coomassie protein assay (Pierce Chemical Co.) (Bradford method) using a microtiter plate protocol with BSA as the standard was used in these experiments. Enzyme Kinetic Data Analysis-The enzyme kinetic data was analyzed with the Michaelis-Menton equation ( u = VmaXX [SI/(&, + [SI))using non-linear least squares data analysis (Leatherbarrow, 1990). PAGE-SDS-PAGE was performed according to Laemmli (1970) using 4-20% acrylamide gradient in the separating gel. Thickness of gels was 0.75 mm and sample buffer contained 4 M urea. Coomassie Brilliant Blue R-250 was used for staining. Zmmunoblotting-SDS-PAGEgelswere developed and samples were electroblotted to nitrocellulose for 1-2 h utilizing a Bio-Rad electrophoretic transfer cell following the suggested protocol. The nitrocellulose sheet was then blocked overnight with 5% BSA in 20 r n Tris-HC1, ~ pH 7.4, 0.15 M NaCl. The squalene synthetase antiserum was diluted 1:lOOO into 20 mM Tris-HC1, pH 7.4, 0.5 M NaCl, 0.05% Tween-20, and 2% BSA and thenitrocellulose sheet incubated with this solution for approximately 5-6 h. The sheet was washed five times with 20 mM Tris-HC1, pH 7.4,0.5 M NaCl, and 0.05% Tween-20 allowing 10-15 min for each wash. The goat anti-rabbit alkaline phosphatase-coupled secondary antibody was diluted 1:lOOO with 20 mM Tris-HC1, pH 7.4,0.5 M NaC1, and 2% BSA. The nitrocellulose sheet was incubated overnight with this solution. The sheet was next washed five times with 20 mM Tris-HCI, pH 7.4, 0.5 M NaCl, and 0.05% Tween-20 allowing 10-15 min for each wash. The sheet was further washed two times with 0.1 M Tris-HC1, pH 9.5, 0.1 M NaCl, and 50 mM MgCl,. Finally, the sheet was incubated with the phosalkaline phosphatase substrates (5-bromo-4-chloro-3-indolyl phate and nitroblue tetrazolium) in the Tris-HC1, pH 9.5,MgCl, buffer according to the Bio-Rad protocol. After sufficient color appeared (2-5 min), the reaction was stopped by rinsing the nitrocellulose sheet extensively with deionized water, incubating the sheet with 20 mM Tris-HC1, 10 mM EDTA, pH 7.5, for 5 min, and then air drying the sheet.

RESULTS

Induction of Rat Hepatic Squalene Synthetase Activityby HMG-CoA Reductase Inhibitors Feeding rats a diet containingHMG-CoA reductase inhibitors in combination with the bile sequestrant, cholestyramine, resultsin marked elevation of squalene synthetase activity. At enzyme protein concentrations of up to 60 pg/ml, it was observed that animals fed LO diet had an 11-fold elevation of hepatic squalene synthetase activity, whereas animalsmaintained on aFLdiethad more than 25-fold elevation of hepatic squalene synthetase activity in comparison to the hepatic activity of squalene synthetase in animals maintained on normal diet (Fig. 1).Therefore, we chose the FL diet treatment for induction of rat hepatic squalene synthetase activity for the purpose of the purification of the enzyme. Animals fed with diet containing 3% cholesterol showed 57% of liver squalene synthetase activity compared to thatof rats fed normal diet. Development of Multiwell Plate Microassay for Squalene Synthetase Effectof Detergent-We have studied the effect of a variety of detergents on the activity of microsomal squalene synthetase. It was observed that of all detergents tried (non-ionic, ionic, and zwiterionic, data notshown), CHAPSwas the most effective. Microsomal squalene synthetase activity increased about 6-fold in thepresence of 5 mM CHAPS (Fig. 2). Activation of Sepharose 4B Column Filtered Microsomes by DTT-DTT has little effect on squalene synthetase activity in freshly isolated rat liver microsomes. However, Sepharose 4B column-filtered microsomes whichshowed littleor no enzyme activity can be activated in the presence of 10 mM DTT. The highest activity is observed for enzymes which are prepared and assayed in this concentration of thiol (Fig. 3). This effect is similar to the activation of latent HMG-CoA reductase by DTT (Dotan and Shechter, 1982). On the basis of these two observations, a microassay of squalene synthetase in a totalvolume of 100 pl was developed as described under “Experimental Procedures.” One advantage of this microassay is its efficiency in the simultaneous assay of many samples. Samples can be analyzed in a relatively short time in comparison to the traditional assays in 2oooo h

l”---l

15000

P

v

01

E

u

10000

a

-2

5000

X

n

u

0 0

20

40

60

BO

100

I20

Protein concentration (@g/rnl) FIG. 1. Effect of different diets on the activity of rat liver Microsomes were prepared from livers of male squalene synthetase. Wistar Charles River rats maintained on normal Purina chow (01, chow supplemented with 3% cholesterol (O), LO diet (V), and FL diet (V)as described under “Experimental Procedures.” Microsomes were prepared in the presence of a mixture of protease inhibitors and assayed for squalene synthetase activity for 30 min a t various protein concentrations. Radiolabeled [3H]FPP of 32.6 mCi/mmol was used for the enzyme assay, and the results are expressed in dpm squalene produced.

8631

Purification of Rat Hepatic Squalene Synthetase ,

pathway, we have observed persistently the existence of squalene synthetase activity in the Slos cytosolic fraction of the 30000 preparations. Formation of radiolabeled squalene from either [14C]mevalonate or [3H]farnesyl diphosphate in the soluble preparation persisted even when the separation of the soluble 0 20000 cytosolic fraction from the microsomes was performed at a higher speed of centrifugation (160,000 x g) suggesting that the activity observed was not residual contamination of microsomes in the soluble preparation. In addition, unlike the reaction in microsomes, formation of squalene from [3H]FPP and NADPH in the solubilized SlWenzyme was not accom0 5 10 15 panied by formation of radiolabeled 2,3-oxidosqualeneor sterols (Fig. 4B).The above observations prompted us to examCHAPS ( m M ) ine the existence of squalene synthetase inthe soluble extract FIG. 2. Effect of CHAPS on the activityof rat liver squalene more carefully in order to determine if the activity observed synthetase. Microsomes prepared from rats fed FL diet were assayed is catalyzed by an hitherto unknown soluble form of the for squalene synthetase activity as described under “Experimental Procedures” except that the concentration of CHAPS was varied as enzyme or, alternatively, by residual microsomal contaminashown. 10 pg of microsomal protein was used for the enzyme assays tion. Table I shows the activities of four microsomal enzymes and [3H]FPPof 32.6 mCi/mmol. The results are expressed in dpm of the cholesterol biosynthetic pathway in an S I 6 0 preparation squalene produced/pg of microsomal protein in 30 min. and in microsomes. Of the four tested enzymes, only HMGCoA reductase is known to exist in a soluble truncated form with subunit size of52-56 kDa as well as in a microsomal form with a subunit size of 97 kDa (Ness et aE., 1981). Unlike Ea 15000 the activity of squalene epoxidase and 2,3-oxidosqualene-0 v lanosterol cyclase, which are strictly in the microsomes, the aJ activity of squalene synthetase is observed in theSI60 fraction 5 10000 as well as in the microsomes, as was also observed for HMGm 1 CoA reductase. This difference in distribution of activities cr E 5000 between the cyclase and epoxidase on one hand and squalene X synthetase and HMG-CoA reductase on the other strongly 35000

1

h

h

Y

0

~

O

A

B

C

D

FIG. 3. Effect of DTT on the activity of Sepharose 4B column filtered rat liver microsomal squalene synthetase.Microsomes were prepared from livers of rats that were fed FL diet in the presence of protease inhibitors as described under “Experimental Procedures” except DTT was omitted from the buffer throughoutthe preparation. 3 ml of microsomes (75 mg/ml) suspended in buffer B which contained protease inhibitors and lacked DTT were applied onto and eluted off a 2.6 X 40-cm Sepharose 4B column using the same buffer. The eluted microsomes were suspended in the same buffer to a concentration of 100 pg/ml and part of the suspension was supplemented with DTT toa final concentrationof 10 mM. The two microsomal suspensions (DTT-; DTT+)were preincubated a t 30 “C for 30 min and then assayed for squalene synthetase activity as described under “Experimental Procedures” in either the presence or the absence of 10 mM DTT in the assay. Activities of DTTenzyme in the absence ( A ) and presence ( B ) of DTT in the assay and activities of DTT+ in the absence (C) and presence (D)of DTT in the assays are expressed in dpm squalene produced/pg of microsomal protein in 30 min. For the assay, [3H]FPP of 32.6 mCi/mmol was used.

which squalene is extracted by organic solvents or radiolabeled protons released during the reaction are exchanged into methanol and distilled (Popjak, 1969; Kuswik-Rabiega and Rilling, 1987). This procedure was essential for monitoring column elutions in the purification of squalene synthetase. An added advantage for this assay is that it can be used not only for cell-free protein isolates but also for in situassays on cultured cells. The assay buffer contains CHAPS which lyses the cells and exposes and activates the enzyme directly in the microwell with no need for additional mechanical manipulations.

Presence of Soluble Squalene Synthetase inRat Liver Cytosolic Preparations In experiments designed to measure activities of rat liver microsomal enzymes involved in the cholesterol biosynthetic

6000

-E a a 4

A

0

5000

4000 3000

2000 1000

0 ?

O 6000

0

B

0

a

5000

n

z

4000

U

3000

m

2000 1000

0 0

1

2

3

4

5

6

7

8

9

1

9

C M F r o m Origin FIG. 4. Theincorporation of [3H]FPP into radiolabeled products in rat liver microsomes and SISO preparations. 55 pg of protein of microsomal (panel A) and 320 pg of soluble SI, protein extracts (panel E ) prepared from livers of rats maintained on FL diet were incubated in a final volume of 1 ml in the presence of 2 mM NADPH, 5 mMMgC12, and lo5 dpm of [3H]FPP,32.6 mCi/mmol at 37 “C for 90 min. The reactions were stopped by the addition of 200 pl of 10% KOH and allowed to stand for an additional 30 min at 37 “C. Extraction of products and separation of radiolabeled farnesol (F-OH), squalene (Sq.), 2,3-oxidosqualene (OS‘.), and lanosterol (Lan) on AgNO3-impregnated silica gel plates was carried out as described by Shechter et al. (1980).

Purification of Rat Hepatic Squalene Synthetase

8632

TABLE I Activities of microsomal enzymes of the cholesterol biosynthetic pathway in microsomal and soluble preparations Crude S10enzyme extract was prepared from rats fed LO diet. The extract was then centrifuged at 105,000 X g, and the microsomal pellet waswashedby additional centrifugation. The Slos cytosolic extract was centrifuged in a Beckman airfuge at 160,000 X g for 10 min to obtain the Slw preparation. Activities of HMG-CoA reductase, squalene synthetase, squalene epoxidase, and 2,3-oxidosqualene-lanosterol cyclase were determined in both the microsomes and the S160 preparations as described under "Experimental Procedures." Specific activity

Enzyme

Microsomes

pmol

HMG-CoA reductase Squalene synthetase Squalene epoxidase 1.2 2,3-Oxidosqualene-lanosterolcyclase

X

SI,

mg" X min"

330 110

3800 2200 940 240

1.0

TABLE I1 Extraction of activity of squalene synthetase from microsomes by freeze-thaw Crude S I 0 preparation was obtained from livers of rats fed with LO diet. The extract was centrifuged a t 105,000 X g for 1 h, and the SI05 supernatant was removed.The microsomes were then frozen in liquid NZ, thawed, and once again centrifuged to obtain the first supernatant extract. This freeze-thaw centrifugation procedure was repeated two more times to obtain second and third SI,, extracts. The activity of squalene synthetase was determined in the original Slobcytosolic preparation as well as in the three freeze-thaw extracts. Enzyme preparation

Specific activity

nmol X mg" S105

360First extract Second extract Third extract

0.3 4.3 6.8 1.8

X

min"

ni

0

h

Total activity

nmol X min"

210 190 47

indicates that the activity of squalene synthetase in the soluble extract is not due to theresidual presence of microsomal particles in the S160 preparation. Additional release of soluble squalene synthetase can be obtained by repeated freeze-thaw of the microsomes (Table 11) similar to that described for HMG-CoA reductase (Heller andGould, 1973;Edwards et al., 1980). Preparation of rat liver microsomes in the presence of protease inhibitorsabolished the activity of cytosolic squalene synthetase and prevented its freeze-thaw extraction from the microsomes (data not shown).

Effect of Protease Inhibitors and Trypsin on the Size of Squalene Synthetase We next attempted to release a solubilized form of squalene synthetase by protease treatment of microsomes. For this purpose, FL microsomes were isolated in the presence of the protease inhibitors described above and washed (three times) with buffer devoid of protease inhibitors. The release of solubilized squalene synthetase from these microsomes by trypsinization is shown in Fig. 5. A time-dependent release of solubilized squalene synthetase was observed. Typically, 2535% of the total microsomal activity was released in a solubilized form after 2 h of trypsinization under the conditions stated (Fig. 5A). Prolonged periods of exposure to trypsin beyond 2 h caused a decrease in the recovery of squalene synthetase activity. Similarly, increase of the temperature during proteolysis to 20 "Cor higher resulted in loss of activity even at short time periods. The specific activity of the soluble enzyme recovered after 2 h of trypsinolysis is approximately 1-1.6 times higher than that in the original microsomal prep-

120

0

60

T r y p s i n o l y s i s Time ( m i n ) FIG. 5. Trypsin solubilization of squalene synthetase from microsomes. Thetotal enzyme activity refers tothe enzymatic activity of the microsomal suspension at theindicated times following incubation with trypsin at 15 "C. The supernatant activity is the activity in the supernatant fraction following centrifugation with a Beckman airfuge for 5 min at 160,000 X g. After the 2-h incubation, approximately 25%of the total activity originally present in the microsomes had been solubilized (panel A ) . Other conditions are given under "Experimental Procedures." In this experiment the final specific activity of the solubilized enzyme was 1.6X the initial specific activity of the microsomal enzyme (panel B ) .

aration (Fig. 5). Fig. 6A shows the elution of squalene synthetase activity of a crude enzyme preparation (Slo)that was prepared without protease inhibitors from a Sephacryl S-200 gel filtration column. As shown, there is no unique elution of activity, and the size distribution extends from the void volume to approximately 32 kDa. The presence of a mixture of protease inhibitors in the buffers, throughout the enzyme preparation procedure, resulted in a single peak of activity which was eluted at the void volume under the same chromatography condition (Fig. 6B). However,removal of the protease inhibitors by sequential washing of the microsomes followed by mild trypsinolysis as described above resulted in a soluble enzyme preparation (Fig. 6C) which eluted from the column under the same conditions at anestimated size of 3236 kDa. Furthermore, this proteolysis product retained enzyme activity and showed remarkable stability in comparison to detergent-solubilized enzyme (data not shown). We concluded thatthistreatment led to proteolysis of squalene synthetase and the release of an enzymatically active fragment into solution. The above experiments served as a basis for the production of a truncated form of squalene synthetase which was then further purified.

Purification of Trypsin-solubilized Squalene Synthetase The purification of trypsin-solubilized squalene synthetase was achieved through aseries of standard column chromatography techniques employing ion-exchange and hydrophobic column chromatography followed by FPLC gel filtration on a Superose 12 column when necessary. Fig. 7A shows the first column chromatography purification step of the trypsin-solubilized enzyme on a DEAE-cellulose (DE52) ion-exchange column. Fractions 42-47, obtained from the DE52 column, were pooled and loaded onto a hydroxylapatite column. The

Purification of

Rat Hepatic Squalene Synthetase

8633 80

60

,

,

I

I

-

h

I

-

0

x

40

-

20

-

x

0 -

4

N

v

0

N

h

N I

0 M

x

N

v

x 0

N

4

h

CY

I

41 x

N

v

x 0

N 4

0 SO

40

50

60

I

70

0

Fraction Number FIG. 6. The effect of protease inhibitors and trypsinolysis on the elution of squalene synthetase in gel filtration chromatography. One ml of crude S10 preparation, without protease inhibitors (90 mg/ml), of livers of rats maintained on FL diet (panel A ) was chromatographed on a 1.5 X 50-cm Sephacryl S-200 column. A similarextract, prepared in the presence of protease inhibitors, was chromatographed under the same conditions (panel B ) . Microsomes, obtained from the S10 preparation in the presence of protease inhibitors, were washed and trypsinized as described under “Experimental Procedures.” After stopping the trypsinolysis reaction with the soy bean trypsin inhibitor and theprotease inhibitors, the entire mixture was applied onto the Sephacryl S-200 column and chromatographed underthe sameconditions (panel C). Fractions of1.0 ml were collected and assayed for squalene synthetase activity. Enzyme activities in the different fractions is expressed as a percent of maximal activity for each run.

activity of squalene synthetase was eluted off the hydroxylapatite column with a salt gradient (Fig. 7 B ) , and fractions containing the major peak of activity (fractions 54-58) were pooled. This enzyme was then loaded onto a phenyl-Sepharose column and eluted as described under “Experimental Procedures.” The purified enzyme was eluted at fractions 99-109 (Fig. 7C). The purified fractions were pooled and concentrated. Thisconcentrated enzyme was either used directly for production of rabbit anti-sera or further chromatographed on gel filtration columns using Superose 12 FPLC chromatography. No further increase in specific activity was observed following the gel filtration chromatography of freshly isolated enzyme, and, therefore, it was used only to repurify aged enzyme. Using Superose 12 FPLC filtration chromatography for the estimation of the purified enzyme size indicated 35 kDa (data not shown) as was estimated by Sephacryl S-200 chromatography (see above). Fig. 8 shows an SDS-PAGE of the proteins obtained at thevarious purification steps. Table I11 summarizes the purification and yields at the various purification steps. Typically, over 160-fold purification was achieved using the above procedure, and thepurified enzyme was obtained at a yield of 5-10%. The enzyme migrates as a single peak on SDS-PAGE with an estimated size of 32-33 kDa (Fig. 8). Isoelectric focusing indicates aPI value of 5.6. The purified enzyme was characterized with respect to initial velocity kinetics. The SOafor FPP was calculated to be 1.0 p M with V,,, of 1.2 pmol/min/mg. So, for NADPH was calculated to be 40 ~ L whereas M NADH, which also served as

20

40

60

80

100

120

Fraction number FIG. 7. Column chromatography purifications of trypsin-

solubilized squalene synthetase. Elution of squalene synthetase activity (0)and protein, as determined at A2W (O), are shown. Panel A illustrates the separationof the enzyme on DEAE-cellulose; panel

B displays the hydroxylapatiteseparation, and panel C gives the phenyl-Sepharose elution profile. The conditions used for these separations are given under “Experimental Procedures.”

+

32kDa

FIG. 8. SDS-PAGE of various purification stepsof squalene synthetase. Proteins separated on an SDS-PAGEgel (4-20% acrylamide gradient) were stained with Coomassie Brilliant Blue R-250. Samples on the various lanes are asfollows: 1, Sigma molecular mass standards: bovine serumalbumin (67 kDa), ovalbumin (44 kDa), glyceraldehyde-3-phosphate dehydrogenase (36 kDa), bovine erythrocyte carbonic anhydrase (30 kDa), trypsinogen (24 kDa), soybean trypsin inhibitor (21 kDa), and a-lactalbumin (14 kDa); 2, enzyme obtained from Superose 12 FPLC chromatography; 3, enzyme from phenyl-Sepharose column; 4 , enzyme from DEAE-cellulose column; 5, S160 of trypsinized induced microsomes; 6, induced microsomes. TABLE I11 Purification of squalene synthetase and yields at various purification steps Squalene synthetase-specific activity was determined for the enzyme preparations at each of the purification steps (see “Experimental Procedures”). The total activity recovered and the percent yield was calculated for each of the preparations. Purification steD

Specific activity

Yield

nmol X mg” X min” 7.7 8.7 38 4.9 260 1200

100.0 25.0 12.0 5.9 7.8

Fold

~

Induced microsomes Trypsin supernatant DEAE-cellulose Hydroxylapatite Phenyl-SeDharose

%

1 1.1

33.8 155.8

Purification of Rat Hepatic Squalene Synthetase

8634

a substrate, has a much higher SOavalue and was calculated to be 800 PM. The kinetic values were determined with20 mM potassiumphosphate, pH 7.4, buffer rather than 100 mM potassium phosphate. Theapparent Soa for NADPH was increased by increasing the concentration of the potassium phosphate buffer.

Preparation of Monospecific Antisera for Squalene Synthetase The purified enzyme was used for the preparationof antisqualene synthetase monospecific rabbit antisera. Using this sera, we immunoanalyzed for squalene synthetase using Western blots. Fig. 9 shows such an analysis of the microsomal and trypsin-treated enzymes as well as thephenyl-Sepharosepurified enzyme. A single band was observed a t 45-47 kDa for the microsomal enzyme and at about 32 kDa for the purified trypsin-treated enzyme. Asmallerfragment (-28 kDa), whichwas removed duringsubsequentpurification steps, was also observed with the trypsin-treatedsample. Fig. 9 also illustrates the induction of enzyme mass in microsomes isolated from animals treated with fluvastatinwhen compared to animals maintained on a normaldiet. DISCUSSION

One important aspect that allowed the successful isolation and purification of the active truncated form of squalene synthetase is the development and implementation of a microenzyme assay in a 96-well plate that does not employ an organic extraction step. This allowed an efficient procedure for the analysis of large numbers of column fractions in a relatively short time. In addition, this assaypermits the determination of squalene synthetase activity in tissue cultures without prior collection of the cells and extraction of enzyme. A key observation that allowed the development of a microassay was the activation of the enzyme by CHAPS and DTT. While DTT has limited effect on freshly isolated microsomes, CHAPS was found to be effective in the activation of both microsomal and solubilize enzyme (data not shown for the solubilized enzyme). The combination of both DTT and CHAPS in the assay proved to yield reproducible results in the determination of squalene synthetase activity in the various preparations. Early attempts tosolubilize yeast membrane squalene synthetase involved the use of detergent (Shechter and Bloch, 1971). The purification of the yeast-derived squalene synthe1

2

3

4

5

6

+ 47 k D a +

32 kDa

FIG. 9. Western blot analysis of various samples of rat hepatic squalene synthetase. Samples from an SDS-PAGE gel (420% acrylamide gradient) were transferred to a nitrocellulose sheet followed by detection with an alkaline phosphatase coupled immunoassay as described under “Experimental Procedures.” The samples shown in the following lanes are: 1, Bio-Rad prestained molecular mass standards: phosphorylase b (97 kDa), bovine serum albumin (67 kDa), ovalbumin (44 kDa), bovine carbonic anhydrase B (30 kDa), soybean trypsin inhibitor (21kDa), andlysozyme (14kDa); 2, sample of trypsinized induced microfrom phenyl-Sepharose column; 3, SIM) somes; 4, induced microsomes (12 pg of protein); 5, normal microsomes (12 pg of protein); 6, normal microsomes (120 pg of protein).

tase was later achieved by employing nonionic detergents for the solubilization of the enzyme out of the membrane (Kuswik-Rabiega and Rilling, 1987; Sasiak and Rilling, 1988). Early attempts in our laboratory to isolate and purify the hepatic enzyme also involved the use of detergents. However, the observation that this enzyme, in the absence of protease inhibitors, showed partition of activity between the soluble and themicrosomal fractions (Table I, Fig. 4)prompted us to consider its purification in the absence of detergents. While proteolysis of the yeast enzyme was observed and proved to be a problem for its purification (Sasiak and Rilling, 1988), we have taken advantage of this process for the isolation of an active truncated form of the enzyme. The evidence that shows the release of this enzyme in an active form to a soluble fraction, either at theearly stage of preparation of the soluble extract (Table I, Fig. 4)or by the freeze-thaw procedure (Table 11), does not indicate or prove the truncation of the protein to a smaller size enzyme. However, the reduction in protein size was assumed at thatstage due to similar observations for HMG-CoA reductase which was shown to be released in a soluble truncated form from microsomes by endogenous proteolysis in thefreeze-thaw procedure (Heller and Gould, 1972; Edwards, et al., 1980). This proteolysis can be inhibited by the presence of leupeptin inthe preparation and preservation of the intact 97-kDa enzyme (Ness et al., 1981). Support for the possibility that hepatic squalene synthetase can also be proteolyzed out of the microsomes came first from preliminary studies inwhich freeze-thaw-treated microsomes were centrifuged in a sucrose gradient in the presence of 5 mM CHAPS. Activity was found in a protein of a smaller size (data not shown). However, quantitative inconsistencies prompted us to devise a better procedure for the proteolytic release of this enzyme. Employing mild trypsinolysis, we were able to release this enzyme out of the membrane in a soluble active form (Fig. 5). Additional similarities between squalene synthetase and HMG-CoA reductase were indicated by Eilenberg et al. (1989) that have shown that these two enzymes, unlike squalene epoxidase and 2,3-oxidosqualene-lanosterolcyclase, are not inhibited by digitonin. Since digitonin is known to interact with sterols in the membrane it was assumed that change in the membrane structure does not affect the catalytic site of either HMG-CoA reductase or squalene synthetase and brought out thepossibility that thecatalytic site of squalene synthetase is not integrated within the membrane. Additional consideration for the hypothesis that squalene synthetase has a distinct cytosolic domain similar to HMGCoA reductase comes from examination of the catalytic reaction itself. The substrates for HMG-CoA reductase are all water-soluble. The structuraldetails of HMG-CoA reductase indicate that, although it is associated with and embedded in the ER, itscatalytic site is located in an hydrophilic domain of the protein which protrudes into thecytosol. Thus, a release of a 52-56-kDa active truncated form of reductase is accomplished by proteolysis of this domain out of the membrane (Liscum et al., 1983, 1985). For squalene synthetase, all the substrates and metal ions required for the reaction (FPP, NADPH, Mg+) are water soluble, and most of the products of the reaction (PPi, NADP+) are water soluble as well. Thus, it is reasonable to assume that thecatalytic siteof this enzyme also protrudes into the cytosol to allow access of these substrates andremoval of products. A strong indication that this is indeed the case comes from the release itself of an active form of the enzyme by mild proteolysis (Fig. 5) accompanied by a decrease in the size of the enzyme (Figs. 6 and 9).Since squalene is not water soluble, there still remainsthe question of its removal at each catalytic cycle. Under in vivo conditions,

Purification of Rat Hepatic Squalene Synthetase there should be a mechanism for the removal of squalene and its transfer to thelumen of the membrane, therefore a carrier protein was considered earlier for this function (Shechter and Bloch, 1971). The removal of squalene in the assay is accomplished, most likely, by the presence of the detergent. This may explain the observed activation of the enzyme byCHAPS (Fig. 2). The activity of squalene synthetase in fibroblasts is suppressed in the presence of low density lipoprotein (Faust et al., 1979). The presence of the bile acid sequestrant cholestyramine in the diet of rats causes the increase in activity of hepatic squalene synthetase (Cohen et al., 1986). The results presented in Fig. 1 are in agreement with the reported data. As shown, cholesterol-fed rats show 60% suppression of squalene synthetase activity, whereas an increase of 11- and 25fold in activity was observed for animals maintained on diets containing lovastatin and fluvastatin, respectively. While discussion on the significance of this observation in relationship to cholesterol homeostasis is beyond this article, this observation was nevertheless useful for the induction of hepatic squalene synthetase for the purpose of its isolation. A point for further consideration should be the regulation of the turnover of the enzyme protein. For HMG-CoA reductase, itwas shown that thedegradation of the enzyme protein is under regulatory control (Edwards et al., 1983; Gill et al., 1985;Peffley and Sinensky, 1985; Nakanishi et al., 1985; Skalnik et al., 1988), and the mechanism for the degradation involves the membrane-bound domain of the protein (Gill et al., 1985; Jingami et al., 1987; Skalnik et al., 1988). Unlike HMG-CoA reductase, the detailed structure of squalene synthetase in not yet known. However, the datapresented in this work suggest the existence of both amembrane and a cytosolic domain for the squalene synthetase protein as well. Thus, it will be interesting in future studies to elucidate the degradation control of this enzyme and to investigate whether other similarities to HMG-CoA reductase exist. Acknowledgments-We thank Dr. Larry Perez for preparation of the unlabeled farnesyldiphosphate. We also thank Dr. Faizulla Kathawala for useful discussions throughout these studies. REFERENCES Agnew, W. S., and Popjak, G. (1978) J. Biol. Chem. 253,4574-4583 Cohen, L. H., Griffioen, A. M., Wanders, R. J. A., Van Roermund, C. W. T., Huysmans, C. M. G., and Princen, H. M. G. (1986) Biochem. Biophys. Res. Commun. 138,335-341

8635

Dotan, I., and Shechter, I. (1982) Biochim. Biophys. Acta 713, 427434 Douglas, T. J., andPaleg, L. G. (1972) Plant Physiol. 49, 417-420 Edwards, P. A., Lemongello, D., Kane, J., Shechter, I., and Fogelman, A. M.(1980) J. Biol. Chem. 255,3715-3725 Edwards, P. A., Lan, S-F., Tanaka, R., and Fogelman, A. M. (1983) J. Biol. Chem. 258, 7272-7275 Eilenberg, H., Klinger, E., Przedecki, F., and Shechter, I. (1989) J. Lipid Res. 30, 1127-1135 Faust, J. R., Goldstein, J. L., and Brown, M. S. (1979) Proc. Natl. Acad. Sci. U. S. A. 76,5018-5022 Gafni, Y., and Shechter, I. (1979) Anal. Biochem. 92, 248-252 Gill. G . . Faust. J. R.. Chin. D. J.. Goldstein. J. L.. and Brown. M. S. (1985) Cell 41,249-258' Glomset, J. A., Gelb, M. H., and Farnsworth, C. C. (1990) Trends Biochem. Sci. 15, 139-142 Goldstein. J. L.. and Brown. M. S. (1990) Nature 343. 425-430 Handler, B., Soltis, D., Mowa, R., Kocher, H. P., Ha, S . , Kathawala, F., and Nemecek, G. (1991) Abstr. Pap. Am. Chem. SOC.201, 1-2 Heller, R. A., and Gould, R. G. (1973) Biochem. Biophys. Res. Commun. 50,859-865 Jennings, S. M., Tsay, Y. H., Fisch, T. M., and Robinson, G. W. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 6038-6042 Jingami, H., Brown, M. S., Goldstein, J. L., Anderson, R. G. W., and Luskey, K. L. (1987) J. Cell Biol. 104, 1693-1704 Kamiya, Y., Sakurai, A., Tamura, S., Takahashi, N., Tsuchiya, E., Abe, K., and Fukui, S. (1979) Agric. Biol. Chem. 43, 363-369 Kuswik-Rabiega, G., and Rilling, H. C. (1987) J . Biol. Chem. 262, 1505-1509 Laemmli, U.K. (1970) Nature 227,680-685 Leatherbarrow, R. J. (1990) Trends Biochem. Sci. 15, 455-458 Liscum, L., Cummings, R. D., Anderson, R. G. W., DeMartino, G. N., Goldstein, J. L., and Brown, M. S. (1983) Proc. Natl. Acad. Sci. U.S. A. 80, 7165-7169 Liscum, L., Finer-Moore, J., Stroud,R. M., Luskey, K. L., Brown, M. S., and Goldstein, J. L. (1985) J. Biol. Chem. 260,522-530 Nakanishi, M., Goldstein, J. L., and Brown, M. S. (1988) J . Biol. Chem. 263,8929-8937 Ness, G. C., Way, S. C., and Wickham, P. S. (1981) Biochem. Biophys. Res. Commun. 102,81-85 Peffley, D., and Sinensky, M. (1985) J. Biol. C k m . 260, 9949-9952 Popjak, G. (1969) Methods Enzymol. 15,394-454 Poulter, C. D., and Rilling, H. C. (1981) in Biosynthesis of Isoprenoid C O ~ D O (Porter. U ~ ~ S J. W.. and Suuraeon. S. L.. eds) DD. .. 413-442. Wiliey, New'York ' Rillina. H. C.. and EDstein. W. W. (1969) . . J. Am. Chem. SOC.19. 104111042 Sasiak, K., and Rilling, H. C. (1988) Arch. Biochem. Biophys. 260, 622-627 Schmidt, A. R., Schneider, C, J., and Glomset, J. A. (1984) J. Biol. Chem. 259,10175-10180 Shechter, I., and Bloch, K. (1971) J. Biol. Chem. 246. 7690-7696 Shechter; I.; Fogelman, A. M., and Popjak, G. (1980) J. Lipid Res. 2 1,277-283 Skalnik, D. G., Narita, H., Kent, C., and Simoni, R. D. (1988) J. Biol. Chem. 262,6836-6841 "

I

I

I