Proc. Nati. Acad. Sci. USA Vol. 74, No. 4, pp. 1431-1435, April 1977 Biochemistry Purification of 3-hydroxy-3-methylglutaryl-coenzyme A reductase from rat liver (affinity chromatography/active and inactive enzyme/immunodiffusion/polyacrylamide gel electrophoresis) DON A. KLEINSEK, S. RANGANATHAN, AND JOHN W. PORTER Lipid Metabolism Laboratory, Veterans Administration Hospital, and the Department of Physiological Chemistry, University of Wisconsin, Madison, Wisconsin 53706 Communicated by David E. Green, January 14,1977

ABSTRACT A procedure for the purification of 3-hy- MATERIALS AND METHODS droxy-3-methylglutaryl-coenzyme A reductase [mevalonate: NADP+ oxidoreductase (CoA-acylating), EC 1.1.1.341 solubilized Materials. Chemicals were obtained from the following from rat liver microsomes is reported. This enzyme has a spe- sources: 3-hydroxy-3-methyl[3-'4C]glutaric acid, New England cific activity of 9,000-10,000 nmol of mevalonate formed per Nuclear; coenzyme A, thioester-linked agarose-hexane-coen- min/mg of protein. This represents a 4100 fold purification over the activity in microsomes, and a specific activity that is ap- zyme A, and dithiothreitol, P-L Biochemicals, Inc.; glucose- proximately 20-fold greater than the highest previously reported 6-phosphate, Nutritional Biochemicals Corp.; 3-hydroxy-3- value. The enzyme is judged to be homogeneous on the basis methylglutaric acid, glucose-6-phosphate dehydrogenase, of sodium dodecyl sulfate/polyacrylamide disc gel electro- NADP+, and NADPH, Sigma Chemical Co.; Bio-Gel, Bio-Rad phoresis, polyacrylamide disc gel electrophoresis, and immu- Laboratories; and cholestyramine, Mead Johnson Laboratories. noanalysis. Data are also presented that indicate the separation All other chemicals and reagents used were of analytical of enzymatically active and inactive species of 3-hydroxy-3- methylglutaryl-coenzyme A reductase on affinity chromatog- grade. on a A column. Treatment of Animals. Male albino Holtzman rats weighing raphy coenzyme 180-200 g were used for all experiments. The animals were 3-Hydroxy-3-methylglutaryl-coenzyme A (HMG-CoA) re- housed in a light-controlled room in which the dark period was ductase [mevalonate:NADP+ oxidoreductase (CoA-acylating); maintained from 1200 to 2400 hours. The animals were fed ad EC 1.1.1.34] catalyzes the reduction of HMG-CoA to mevalonic lib. a 2% cholestyramine Wayne Lab Blox powdered diet for acid, the rate-limiting step of cholesterol biosynthesis in liver a minimum of 4 days to effect maximum liver HMG-CoA re- (1-3). Therefore, a number of researchers have focused their ductase activity. attention on the regulation of this enzyme. However, to study Preparation of Microsomes. The animals were sacrificed the modulation of HMG-CoA reductase under various physi- by decapitation at 1800 hours, the diurnal high point of ological states a method is required to quantitate the amount HMG-CoA reductase activity. The livers were excised and of enzyme present. This can be achieved by either direct iso- immediately placed in ice-cold homogenization medium which lation of the enzyme by a reproducible purification procedure contained 50 mM potassium phosphate buffer (pH 7.0)/0.2 M or by immunoprecipitation using monospecific antiserum to sucrose/2 mM dithiothreitol (buffer I). Livers were homoge- the enzyme. nized in this medium (2 ml/g of liver) in a Waring blendor for A number of procedures have been reported for the solubi- 15 sec, followed by three strokes with a motor-driven Teflon lization and partial purification of HMG-CoA reductase from pestle in a Potter-Elvehjem type glass homogenizer. The ho- the microsomal membrane (4-8). In addition, workers in three mogenate was centrifuged for 10 min at 15,000 X g and the laboratories have reported the preparation of enzyme that yields supernatant solution was centrifuged at 100,000 X g for 75 min. only one band on immunodiffusion or sodium dodecyl sulfate The microsomal pellet was resuspended in buffer I containing (NaDodSO4) disc gel electrophoresis (9-11). However, the 50 mM EDTA and recentrifuged at 100,000 X g for 60 min. enzyme activities of these preparations were very low (10-516 This pellet was used for isolation of the enzyme. All of the above nmol of mevalonate formed per min/mg of protein). operations were carried out at 4°. In a previous study (12) we succeeded in purifying yeast Solubilization of Enzyme. The method of Heller and Gould HMG-CoA reductase to homogeneity. This enzyme had a (7) was used to solubilize the enzyme. Microsomal pellets were specific activity of approximately 10,000 nmol of mevalonate frozen at -20° for at least 2 hr. After thawing at room tem- formed per min/mg of protein. In this paper we report the perature the microsomes were homogenized in solubilization purification of HMG-CoA reductase from rat liver by a com- buffer that contained 50 mM potassium phosphate (pH 7.0)/0.1 bination of standard protein fractionation steps and coenzyme M sucrose/2 mM dithiothreitol/50 mM KCI/30 mM EDTA. A affinity chromatography. This preparation also has a specific A Potter-Elvehjem homogenizer with a tight-fitting Teflon activity of 9,000-10,000 nmol of mevalonate formed per pestle was used. After standing for 15 min at room temperature, min/mg of protein. This value is approximately 20-fold greater the suspension was centrifuged at 100,000 X g for 60 min at 200. than the best value previously reported. The supernatant solution was collected and used for the puri- As a part of this study we also show that enzymatically active fication of the enzyme. All further operations were carried out and inactive species of HMG-CoA reductase are separated by at room temperature. affinity chromatography. This separation suggests the possibility Assay Systems. Two assay systems were used to measure that cholesterol synthesis may be regulated in vvo by the in- HMG-CoA reductase activity. For measuring microsomal en- terconversion of these species. zyme activity, an adaptation of the NADPH-generating ra- diochemical method previously described (13) was used. The Abbreviations: HMG-CoA reductase, 3-hydroxy-3-methylglutaryl- 0.5-ml reaction volume contained potassium phosphate buffer coenzyme A reductase; NaDodSO4, sodium dodecyl sulfate. (pH 7.0), 50,mol/dithiothreitol, 2,umol/glucose-6-phosphate, 1431 Downloaded by guest on September 24, 2021 1432 Biochemistry: Kleinsek et al. Proc. Nati. Acad. Sci. USA 74 (1977) Table 1. Purification of HMG-CoA reductase NADPH oxidation was determined from the change in ab- from rat liver microsomes sorbance at 340 nm using a glass cuvette with a 2-mm light path. This reaction results in 1 nmol of mevalonate formed per 2 nmol Specific of NADPH oxidized. One unit of HMG-CoA reductase activity Total activity is defined as the quantity of enzyme that produces 1 nmol of Purification protein Total (units/mg Purifi- mevalonate in 1 min at 37'. Protein determinations were car- step (mg) units protein) Yield cation ried out by the method of Lowry et al. (14). Microsomal Polyacrylamide Gel Electrophoresis. The procedure of suspension 4,800 11,255 2.3 100 1 Weber and Osborn (15) was used for NaDodSO4/polyacryl- Soluble extract 198 3,527 17.8 31 8 amide gel electrophoresis. NaDodSO4/5% polyacrylamide gels (NH4)2SO4, were fixed and the NaDodSO4 was leached out in 20% sulfo- 35-50% 65 3,189 49 28 22 salicyclic acid (16) prior to staining with 0.25% Coomassie Heat treatment 13 2,506 193 22 82 brilliant blue. Polyacrylamide disc gels (5%, pH 8.9) were made (NH4)2SO4, with 30% glycerol. The electrophoresis was carried out in 5 mM 0-50% 9 2,050 228 18 97 Tris/35 mM glycine at pH 8.3. Bio-Gel Antisera and Immunodiffusion. Crude antisera was ob- filtration* 0.42 777 1,850 7 787 tained by subcutaneously injecting a 5-kg rabbit at 2-week in- CoA affinity tervals with 1 mg of supernatant protein from the heat-treat- column 0.036 345 9,583 3 4,078 ment step. Blood was withdrawn after 4, 6, and 8 weeks of the initial injection and allowed to serum The data of this table were obtained during the purification of clot; the was concentrated HMG-CoA reductase from 25 rat livers. with a 0-50% (NH4)2SO4 fractionation. The antisera pellet was * Sucrose density gradient centrifugation may be substituted for the dissolved in a 10 mM potassium phosphate (pH 7.0)/0.9% NaCl Bio-Gel filtration step. This procedure was used in the preparation solution and stored at -20°. Immunodiffusion slides consisted of material for immunodiffusion analysis. of a 0.5% agar/0.9% NaCl matrix.

2,umol/NADP+, 0.5 Mmol/DL-[3-14C]HMG-CoA, 0.15,gmol/ RESULTS glucose-6-phosphate dehydrogenase, 1.25 units/enzyme pro- tein, 300-1200,tg. The reaction was carried out at 370 for 5-10 Purification of Enzyme. HMG-CoA reductase has been min and terminated by the addition of 50 Al of 2.4 M HCL. A purified to near homogeneity from a solubilized extract of rat 200-IdI aliquot of the incubation mixture was spotted directly liver microsomes. A typical purification is presented in Table onto activated Silica Gel G thin-layer plates and the chro- 1. After the solubilization of HMG-CoA reductase from mi- matogram was developed in benzene/acetone (1:1, vol/vol). crosomes, the soluble extract was subjected to a 35-50% satu- HMG-CoA reductase activity in the solubilized fractions was ration with (NH4)2SO4. This resulted in a 2- to 3-fold purifi- assayed spectrophotometrically. The reaction volume was 0.5 cation and a small loss of enzyme activity. The protein pellet ml and it consisted of potassium phosphate buffer (pH 7.0), 50 was dissolved in buffer containing 50 mM potassium phosphate ,umol/dithiothreitol, 2,Mmol/NADPH, 0.30,tmol/DL-HMG- (pH 7.0)/3 mM dithiothreitol/30% (vol/vol) glycerol/1.0 M CoA, 0.15,gmol/enzyme protein, 0.2-400,gg. The reaction KCI at a concentration of 6-8 mg/ml. The enzyme solution was mixture, without substrate, was first incubated for 5 min at 370 heated at 65° for 6 min and denatured protein was'removed and the assay was carried out at 370 in a Gilford 2400S spec- by centrifugation at 100,000 X g for 30 min. The supernatant trophotometer after addition of HMG-CoA. The rate of solution contained 70-80% of the starting enzyme activity and an increase of 3- to 4-fold in specific activity of the enzyme was observed. The heat-treated extract was then concentrated, after 0.5 M KCI V diluting 1:1 with solubilization buffer, by precipitation of protein by 0-50% saturation with The protein I ~~~~~~~~~~00I (NH4)2SO4. pellet E It was dissolved in a minimal volume of 50 mM potassium phos- c 0.30 -' 600 z phate (pH 7.0)/2 mM dithiothreitol/30 mM EDTA/50 mM 0 I KCI/10% sucrose. U- The enzyme solution was loaded onto a Bio-Gel A-0.5m gel 0.20 -I 400 a: filtration column (2 X 44 cm). The peak fractions of enzyme zI 0j activity in the eluant were'purified 8-fold in enzyme activity with a 38% recovery of enzyme protein. The fractions eluted from the gel were concentrated and then dialyzed in collodion 0 200u m bags. The dialyzed protein was applied to a thioester-linked -J 2 agarose/hexane/coenzyme A column. Details of the affinity N z binding and elution conditions are reported in the legend to Fig.' Li 2 4 6 8 10 1. The affinity step routinely resulted in a-95-100% recovery FRACTION NUMBER of enzyme activity with a specific activity of 9,000-10,000 nmol FIG. 1. Affinity chromatography of HMG-CoA reductase on a of mevalonate formed per min/mg of protein. coenzyme A column. The enzyme was added to the column, 0.2 ml of Criteria for Homogeneity of Enzyme. One major staining wet gel packed in a pasteur pipette, at 40 in 25 mM potassium phos- band was obtained when 40 jig of purified enzyme was elec- phate buffer (pH 7.0)/1 mM dithiothreitol/1 mM EDTA/10% sucrose. trophoresed on NaDodSO4/polyacrylamide disc gel (Fig. 2). The flow rate was 1 ml/10 min. The unbound protein fraction was concentrated and saved for immunoanalysis. HMG-CoA reductase Similarly, only a single band was observed when 10-20,ug of was eluted by increasing the ionic strength of the buffer by addition purified enzyme was electrophoresed on 5% polyacrylamide of KCl to a concentration of 0.5 M. The elution flow rate was 1 ml/min disc gel (Fig. 3). A third criterion of homogeneity of the enzyme and fractions of 1 ml were collected. is shown in Fig. 4. A single immunoprecipitin band was ob- Downloaded by guest on September 24, 2021 Biochemistry: Kleinsek et al. Proc. Natl. Acad. Sci. USA 74 (1977) 1433

FIG. 2. NaDodSO4 disc gel electrophoresis of HMG-CoA re- a b ductase. Protein, 40 Mug of affinity-purified enzyme, was electropho- FIG. 3. Polyacrylamide disc gel electrophoresis of HMG-CoA resed in 0.1% NaDodSO4 on a 5% polyacrylamide disc gel. An RF of reductase. Protein, 10,gg (a) and 20 ,gg (b), was stained with 0.25% approximately 0.65 was obtained for the single staining band. Coomassie brilliant blue.

tained when purified enzyme was reacted against crude anti- Five percent or less of active HMG-CoA reductase did not sera. bind to the column, and this protein was recovered in the un- Evidence for Active and Inactive Enzyme Species. The bound eluant. However, this amount of protein did not account protein fraction that did not bind to the coenzyme A affinity for the line of confluence between unbound and bound frac- gel contained a species that was immunologically identical to tions on immunoanalysis. Furthermore, the staining profiles the protein fraction that bound to, and was eluted off, the af- of the two fractions on NaDodSO4/polyacrylamide gels (Figs. finity column. This is demonstrated by the partial identity of 2 and 5) indicated that these two species were identical in their antigenic determinants in well c with those in wells b and d. migration characteristics. Hence it is concluded that the ma- Because the enzyme solution had been centrifuged at 100,000 jority of the protein unbound on affinity chromatography was X g prior to affinity chromatography the enzymatically inactive inactive HMG-CoA reductase. species that did not bind to the affinity column was not dena- tured protein. In addition, the protein fraction not bound to the affinity column was not inactivated because of cold lability (17) DISCUSSION because this fraction showed no reactivation of catalytic The procedure for the purification of HMG-CoA reductase properties after preincubation at 37°. reported in this paper yields a protein with a specific activity Downloaded by guest on September 24, 2021 1434 Biochemistry: Kleinsek et al. Proc. Nati. Acad. Sci. USA 74 (1977)

FIG. 4. Ouchterlony double-diffusion precipitation of affinity protein fractions. The center well is antisera prepared against the heat-treated enzyme preparation. Wells a and c contain 1.5 ,ug and 7 jg, respectively, of affinity-bound enzyme. Wells b and d represent 8 jug and 16 gg, respectively, of the protein fraction that did not bind to the affinity column.

of 9,000-10,000 nmol of mevalonate formed per min/mg of protein. Polyacrylamide disc gel and NaDodSO4/polyacryl- amide disc gel electrophoresis of the purified enzyme show only a single major staining band. In addition, immunodiffusion of the enzyme against antisera to crude enzyme yields only one FIG. 5. NaDodSO4 disc gel electrophoresis of the protein (30 ,g) precipitin band. These results are evidence for homogeneity not bound to a CoA affinity gel column. An RF of approximately 0.65 of our preparation. was obtained for the major staining band. Prior studies from three other laboratories have reported the preparation of HMG-CoA reductase from rat liver. However, NaDodSO4 gel electrophoresis, identical migratory behaviors each of these preparations had an activity that was 5% or less of the proteins were noted. of the activity we report for the rat liver and yeast enzymes. In The separation of active and inactive species of HMG-CoA the initial reported preparation, Higgins et al. (9) solubilized reductase is a confirmation and an extension of previous studies microsomal-bound HMG-CoA reductase with sodium deoxy- (18-21) that indicated that HMG-CoA reductase activity of cholate and then purified the enzyme to an activity of 10 nmol crude liver preparations could be increased or decreased by of mevalonate formed per min/mg of protein. More recently, incubation with appropriate subeellular fractions. Presumably Heller and Shrewsbury (11) and Tormanen et al. (10) reported these studies involved a change in enzyme activity through the isolation of HMG-CoA reductase. Each of these preparations covalent modification of the enzyme. Hence it will be of interest was solubilized from microsomes by a slow freeze-thaw tech- to learn from future work the difference between active and nique and then purified by different procedures. Heller and inactive HMG-CoA reductase species and the mechanism of Shrewsbury (11) reported a specific activity of 462 nmol of' regulation of their interconversion. mevalonate formed per min/mg of protein, while Tormanen et al. (10) obtained a specific activity of 516. This investigation was supported in part by grants from the Wis- Obviously, the HMG-CoA reductase preparations reported consin Heart Association and the National Heart and Lung Institute previous to this paper were (i) impure or (ii) a mixture of active (no. HL-16364) of the National Institutes of Health, U.S. Public Health and inactive enzyme or (iii) they contained significant amounts Service. of an inhibitor (possibly cholesterol) bound to the enzyme, in- 1. Siperstein, M. D. & Fagan, V. M. (1966) J. Biol. Chem. 241, asmuch as the best of these systems had an activity of 5% or less 602-609. of the value we report. A partial explanation for the discrepancy 2. White, L. W. & Rudney, H. (1970) Biochemistry 9, 2725- has been supplied by the data in this paper. Affinity chroma- 2731. tography on a coenzyme A column separated HMG-CoA re- 3. Shapiro, D. & Rodwell, V. W. (1972) Biochemistry 11, 1042- ductase into active and inactive species. The fraction that passed 1045. through the affinity column contained little or no 4. Linn, T. C. (1967) J. Biol. Chem. 242, 990-993. HMG-CoA 5. Kawachi, T. & Rudney, H. (1970) Biochemistry 9, 1700-1705. reductase activity. However, this fraction gives an immu- 6. Brown, M. S., Dana, S. E., Dietschy, J. M. & Siperstein, M. D. noprecipitin band that is continuous with that given by the (1973) J. Biol. Chem. 248, 4731-4738. active enzyme bound to the coenzyme A affinity column. 7. Heller, R. A. & Gould, R. G. (1973) Biochem. Blophys. Res. Furthermore, when both of these species were subjected to Commun. 50,859-865. Downloaded by guest on September 24, 2021 Biochemistry: Kleinsek et al. Proc. Nati. Acad. Sci. USA 74 (1977) 1435

8. Ackerman, M. E., Redd, W. L. & Scallen, T. J. (1974) Biochem. 15. Weber, K. & Osborn, M. (1969) J. Biol. Chem. 244, 4406- Biophys. Res. Commun. 56,29-35. 4412. 9. Higgins, M. J. P., Brady, D. & Rudney, H. (1974) Arch. Biochem. 16.- Dunker, A. K. & Rueckert, R. R. (1969) J. Biol. Chem. 244, Biophys. 163,271-282. 5074-5080. 10. Tormanen, C. D., Redd, W. L., Srikantaiah, M. V. & Scallen, T. 17. Heller, R. A. & Gould, R. G. (1974) J. Biol. Chem. 249,5254- S. (1976) Biochem. Biophys. Res. Commun. 68,754-762. 5260. 11. Heller, R. A. & Shrewsbury, M. A. (1976) J. Biol. Chem. 251, 18. Beg, Z. H., Allman, D. W. & Gibson, D. M. (1973) Biochem. 3815-3822. Biophys. Res. Commun. 54, 1362-1369. 12. Qureshi, N., Dugan, R. E., Nimmannit, S., Wu, W. H. & Porter, 19. Brown, M. S., Brunschede, G. Y. & Goldstein, J. L. (1975) J. Biol. J. W. (1976) Biochemistry 15,4185-4190. Chem. 250, 2502-2509. 13. Nepokroeff, C. M., Lakshmanan, M. R., Ness, G. C., Dugan, R. 20. Berndt, J. & Gaumert, R. (1974) Hoppe-Seyler's Z. Physiol. E. & Porter, J. W. (1974) Arch. Biochem. Biophys. 160, 387- Chem. 355,905-910. 393. 21. Shapiro, D. J., Nordstrom, J. L., Mitschelen, J. J., Rodwell, V. J. 14. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. & Schimke, R. T. (1974) Biochim. Biophys. Acta 370, 369- (1951) J. Biol. Chem. 193,265-275. 377. Downloaded by guest on September 24, 2021