Regulation of 3-Hydroxy-3-methylglutaryl Coenzyme A Reductase Activity in Avian Myeloblasts MODE OF ACTION OF 25-HYDROXYCHOLESTEROL*

25-Hydroxycholesterol inhibits cholesterol biosynthesis by inhibiting the activity of 3-hydroxy-3-meth- ylglutaryl coenzyme A (HMG-CoA) reductase. Addition of 25-hydroxycholesterol to chicken myeloblasts caused a rapid inhibition of HMG-CoA reductase activity, producing approximately an 80% decrease in en- zyme activity after 60 min. The mode of action of 25-hydroxycholesterol was determined by immunoprecip- itating radiolabeled enzyme from 25-hydroxycholes-terol-treated myeloblasts. The decline in enzyme activ- ity due to addition of 25-hydroxycholesterol was not associated with increased levels of [32P]P04 incorpo- ration into the immunoprecipitated reductase polypeptide (Mr = 94,000). Hence, 25-hydroxycholesterol did not appear to regulate reductase activity by enzyme phosphorylation, as observed for other modulators of HMG-CoA reductase. However, 25-hydroxycholes-terol was shown to inhibit reductase activity by caus- ing a 350% increase in the relative rate of reductase degradation and a 72% decrease in the relative rate of reductase synthesis. These alterations in the rates of degradation and synthesis occurred rapidly (within 10-30 min after addition of 25-hydroxycholesterol) and can account

25-Hydroxycholesterol inhibits cholesterol biosynthesis by inhibiting the activity of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase. Addition of 25-hydroxycholesterol to chicken myeloblasts caused a rapid inhibition of HMG-CoA reductase activity, producing approximately an 80% decrease in enzyme activity after 60 min. The mode of action of 25hydroxycholesterol was determined by immunoprecipitating radiolabeled enzyme from 25-hydroxycholesterol-treated myeloblasts. The decline in enzyme activity due to addition of 25-hydroxycholesterol was not associated with increased levels of [32P]P04 incorporation into the immunoprecipitated reductase polypeptide (Mr = 94,000). Hence, 25-hydroxycholesterol did not appear to regulate reductase activity by enzyme phosphorylation, as observed for other modulators of HMG-CoA reductase. However, 25-hydroxycholesterol was shown to inhibit reductase activity by causing a 350% increase in the relative rate of reductase degradation and a 72% decrease in the relative rate of reductase synthesis. These alterations in the rates of degradation and synthesis occurred rapidly (within 10-30 min after addition of 25-hydroxycholesterol) and can account completely for the 25-hydroxycholesterol-induced inhibition of enzyme activity. The rapid decline in the rate of synthesis of HMG-CoA reductase in 25-hydroxycholesterol-treated cells was not associated with concomitant changes in the levels of reductase mRNA; therefore, suggesting that 25-hydroxycholesterol must inhibit the rate of reductase synthesis by translational regulation. We also present evidence that mRNA purified from chicken myeloblasts codes for two reductase polypeptides of M, = 94,000 and 102,000.
3-Hydroxy-3-methylglutaryl coenzyme A reductase catalyzes the rate-controlling step in cholesterol biosynthesis (1). Addition of 25-hydroxycholestero1 to cells causes suppression of cholesterol biosynthesis by inhibiting the activity of HMGI-Health Service Grants HL 19063, HL 25590, HL 30568, and IT32 * This investigation was supported in part by United States Public HL 07412 and by grants from the American Heart Association, Greater Los Angeles Affiliate (649-P5) and the Laubisch Fund. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
$. The recipient of United States Public Health Service Research Career Development Award HL-00426.
CoA reductase (2)(3)(4). Faust et al. (5) and Sinensky et al. (6) immunoprecipitated radiolabeled reductase ( M , = 52,000-68,000) from Chinese hamster ovary cells which were either adapted to grow in the presence of compactin (5) or were mutants auxotrophic for mevalonate (6) and concluded that the primary effect of 25-hydroxycholesterol was to suppress the synthesis of the enzyme. 25-Hydroxycholesterol also increased the rate of reductase degradation in the cells adapted to grow in compactin (5) but had no effect in mevalonate auxotrophs (7). A more recent report demonstrated that the molecular weight of the enzyme subunit from Chinese hamster ovary cells was approximately 90,000 (8). Thus, these data (5-7) derived by immunoprecipitation of a proteolytic fragment of the enzyme may not describe accurately the mode of action of 25-hydroxycholesterol. Consequently, the effect of 25-hydroxycholesterol on the intact enzyme has yet to be established.
The role of enzyme phosphorylation in the mode of action of 25-hydroxycholesterol is also unclear. Some modulators of HMG-CoA reductase are reported to inhibit enzyme activity by inducing phosphorylation of the enzyme (for review, see Ref. 9). However, studies which have examined the effect of 25-hydroxycholesterol on enzyme phosphorylation have used only indirect methods for assessing phosphorylation and further work is needed to clarify this point (10,11). Therefore, in the present study, we have immunoprecipitated the native HMG-CoA reductase ( M , = 94,000) from chicken myeloblasts and determined whether treatment of cells with 25-hydroxycholesterol inhibits enzyme activity by changes in enzyme phosphorylation and/or by changes in the rates of synthesis and degradation of the reductase. We also report the effect of 25-hydroxycholesterol on the levels of reductase mRNA. Sources for all other materials have been previously described (12).
Cells-Chicken myeloblasts (BM-I1 cell line) were grown as described (13) in 75 cm2 culture flasks. The media was supplemented with 10% tryptose phosphate broth, 5% bovine calf serum, and 5% chicken serum (Flow Laboratories). For methionine-free or phosphate-free media the specific component was omitted along with the tryptose phosphate broth, and the media was prepared with dialyzed calf and chicken sera. Cells were adapted t o grow in the presence of mevinolin by first selecting for growth in normal media containing 15 PM mevinolin and then slowly adapting the cells to grow in the presence of higher concentrations of the drug.

Regulation of HMG-CoA Reductase in Avian Myeloblasts
HMG-CoA Reductase Assay-Reductase activity was assayed in microsomes or in cell extracts using a radioassay (14). Cell extracts were prepared by detergent solubilization as described (15). Protein concentration was determined using a modified Coomassie blue dye binding assay (16).
Antibody-Antibody to purified HMG-CoA reductase from rat liver was prepared in rabbits (16) and the IgG fraction was purified on Protein A-Sepharose (17). Normal rabbit serum IgG was prepared by an identical method. The IgG samples were dialyzed and diluted in phosphate-buffered saline (pH = 7.4) to final concentrations of 0.75

Immunotitration-Microsomes
were prepared from 1.5 ml of packed cells. Cells were suspended in buffer A (0.1 M sucrose, 50 mM KC1, 40 mM KH2POa, 10 mM dithiothreitol, and 30 mM EDTA, pH = 7.2), Dounce-homogenized, and centrifuged at 10,000 X g for 10 min at 5 "C. The supernatant was collected and recentrifuged under identical conditions, and the final supernatant was centrifuged a t 100,000 X g for 45 min a t 5 "C. The microsomal pellets were frozen and stored a t -76 "C. For immunotitration experiments, the microsomal pellets were thawed, buffer A was added, and the pellets were briefly sonicated. Samples were preincubated 30 min at 37 "C with varying amounts of either anti-reductase or normal serum IgG and the remaining enzyme activity was determined. Immunoprecipitation-Cells were washed and suspended in methionine-free media (final concentration 1 X lo7 cells/ml). For studies to determine the rate of enzyme degradation, the cells were incubated with [:"S]methionine (specific activity = 1100 Ci/mmol, 26.4 pCi/l X lo7 cells) for 90 min a t 37 "C, washed, and incubated at 37 "C in complete media (1 X lo7 cells/ml) containing unlabeled methionine (5.0 p~) with or without 25-hydroxycholesterol (final concentration 5 pg/ml). 25-Hydroxycholesterol was added in 95% ethanol and control cells received equivalent amounts of 95% ethanol. Aliquots were removed a t specified time intervals and centrifuged, and the cells were solubilized in buffer B (1% Triton X-100, 0.5% deoxycholate, 0.1% SDS, 0.01% NaN3, 0.1 M NaCl, 5 mM EDTA, 0.01 M phosphate buffer, pH 7.5, 0.1 mM leupeptin, aprotinin (26 pg/ml), and 1 mM phenylmethylsulfonyl fluoride) at a concentration of 1 X IO7 cells/ml. The cell lysates were centrifuged a t 100,000 X g for 120 min a t 5 "C and the supernatants were stored at -20 "C.
For experiments to determine the rates of enzyme synthesis the cells were washed and incubated in methionine-free media (1 X lo7 cells/ml) at 37 "C.
[35S]Methionine was added (26.4 pCi/l x lo7 cells) simultaneously, unless otherwise noted, with either 25-bydroxycholesterol (final concentration 5 pg/ml) or an equivalent volume of ethanol in the control cells. Aliquots were removed at specified time intervals and the cell lysates were prepared as stated above.
In [32P]P04 labeling experiments the cells were washed and suspended in phosphate-free media (final concentration 3.8 X lo7 cells/ ml) and incubated with [32P]P04 (100 pCi/ml) for 60 min a t 37 "C. The cells were divided into two groups; one received 25-hydroxycholesterol (final concentration 5 pglml) and the control group received an equivalent volume of ethanol. Aliquots were removed a t specified time periods and washed twice in phosphate-buffered saline, and the cell lysates were prepared as stated above except that buffer B also contained 50 mM NaF and the concentration of cells in buffer B was M / d .

X IO7 cells/ml. In both ["S]methionine and [32P]
P04 experiments the total cellular incorDoration of radiolabel was determined by precipitation of 10 pl of ceil lysate in 2.0 ml of 10% trichloroacetk acid a t 5 "C. The precipitates were collected by filtration on Whatman GF/C filters which were rinsed with 1.0 N HC1 prior to use. After filtration, each precipitate was washed three times with 5 ml of 10% trichloroacetic acid, twice with 5 ml of ethanol, and dried, and the radioactivity was determined by liquid scintillation spectrometry after the addition of 10 ml of Aquasol (New England Nuclear).
All immunoprecipitations were conducted using the same protocol. Radiolabeled cell lysates were thawed, 3.8 pg of normal rabbit serum IgG were added, and the samples were incubated overnight a t 5 "C. Pansorbin (formalin-treated Staphylococcus aureus) was washed and resuspended in buffer B (containing 5 mg/ml of bovine serum albumin) to make a 10% v/v suspension, and 70 pl of Pansorbin were added to each sample. After incubation for 120 min a t 5 "C the samples were centrifuged in a Brinkmann microfuge, the supernatants were collected, and 3.8 pg of anti-reductase IgG were added. Samples were incubated overnight a t 5 "C. Pansorbin was prepared as stated above except that the bacteria were washed in nonradioactive 25-hydroxycholesterol-treated cell lysate prior to making a 10% suspension in buffer B. The nonradioactive lysate was prepared by treating cells (1 X lo7 cells/ml) with 25-hydroxycholesterol (5 pglml) for 120 min at 37 "C followed by solubilization with buffer B (1 X lo7 cells/ml) and centrifugation a t 100,000 X g for 120 min at 5 "C. The supernatant was termed the 25-hydroxycholesterol-treated cell lysate. This treatment was used to reduce the nonspecific binding of radiolabeled proteins to the Pansorbin. Each radiolabeled sample received 70 p1 of the treated Pansorbin and was incubated 120 min a t 5 "C. with 1.0 ml of buffer C (1% Triton X-100, 0.5% deoxycholate, 0.1% After centrifugation the Pansorbin pellets were washed sequentially SDS, 0.01% NaN3,O.l M NaC1,5 mM EDTA, 0.01 M phosphate buffer, pH = 7.5), 1.0 ml of buffer D (0.5 M LiC1, 0.1 M Tris, pH = 8.0), and 1.0 ml of buffer C. In phosphorylation studies buffers B, C, and D also contained 50 mM NaF. The final Pansorbin pellets were processed for electrophoresis as stated below.
Electrophoresis-The Pansorbin pellets were suspended in 50 pl of 15% SDS, 10% P-mercaptoethanol, 63 mM Tris, pH = 6.8, and incubated at 25 "C for 40 min. After centrifugation, the supernatants were collected and solid urea was added to make a final concentration of 8 M urea, 7.5% SDS, 5% P-mercaptoethanol, and 31.5 mM Tris. Samples were heated at 100 "C for 5 min, 8 pl of 0.01% bromophenol blue, 50% glycerol were added, and the samples were applied to a SDS-polyacrylamide slab gel containing 8 M urea in both the 5% polyacrylamide stacking gel and the 7.5% polyacrylamide separating gel. Electrophoresis buffers were those described by Laemmli (18). Electrophoresis was conducted for 16 h a t 50 V (constant voltage), and the gels were processed for fluorography using sodium salicylate (19) and exposed to Kodak XAR-5 film at -76 "C. Molecular weight standards were: myosin, M , = 200,000; phosphorylase b, M , = 92,500; and ovalbumin, M , = 43,000. To determine the radioactivity present in a protein band, the appropriate segment was cut out of the gel and incubated overnight in 2.0 ml of 30% H202 at 70 "C, and the radioactivity was determined by liquid scintillation spectrometry after adding 10 ml of Aquasol.
RNA Purification-Cells (1 X lo7 cells/ml) were incubated a t 37 "C for 120 min in media containing 25-hydroxycholesterol (5 pg/ml) or with an equivalent volume of ethanol for the control sample. Enzyme activity was assayed after 120 min and the 25-hydroxycholestero1treated cells contained only 20% of the reductase activity present in control cells. The cells (3.4 X 10') from each group were collected and processed for purification of total RNA as described (20). Polyadenylated RNA was purified by oligo(dT)-cellulose chromatography (20). I n Vitro Translation of Poly(A+) RNA-Rabbit reticulocyte lysates were prepared (21) and treated with S. aureus SI nuclease (22). In uitro translations were performed using 50 p1 of reticulocyte lysate, 6 p1 of [3sS]methionine (1100 Ci/mmol), and 6.5 pl of poly(A+) RNA. Assays were performed a t 37 "C for 60 min and the reactions were stopped by addition of 550 pI of buffer B. Samples were immunoprecipitated as described above. Total incorporation of ["S]methionine into protein was determined by trichloroacetic acid precipitation as previously described.

RESULTS
Anti-reductase Antibody-The specific activity of HMG-CoA reductase in microsomes prepared from chicken myeloblasts grown in sera was 0.144 f 0.025 (S.D.) nmol/min/mg of microsomal protein (n = 9). Enzyme activity increased when cells were incubated for 48 h in media devoid of lipoproteins (data not shown). Antibody prepared to the rat liver enzyme cross-reacted and inactivated approximately 80% of the reductase activity present in myeloblast microsomes ( Fig.  1). Addition of normal rabbit serum IgG had no effect on enzyme activity (data not shown). Similar results were obtained when the anti-reductase antibody was added to microsomes prepared from fresh chicken liver (data not shown). The inability of the antibody to inactivate 100% of the reductase activity has also been observed in microsomal preparations from rat liver (16) and human liver (14).
Immunoprecipitation Antibody was made to the purified HMG-CoA reductase from rat liver (16) and the IgG fraction was isolated (17). The IgG concentration was 0.35 p g / p l . Microsomes from myeloblasts were prepared as stated in the text and preincubated for 30 min a t 37 "C with varying amounts of anti-reductase IgG. Reductase activity was determined by a radioassay (14) and the protein concentration was determined by a Coomassie blue dye binding assay (16). To clearly identify the M, = 94,000 band on the fluorograph as the reductase polypeptide, immunoprecipitations using [:''S]methionine-labeled cell lysates were carried out in the presence or absence of unlabeled pure rat liver HMG-CoA reductase (16). The presence of nonradioactive reductase should compete with the radiolabeled enzyme for binding to the anti-reductase IgG, and we would predict that the band on the fluorograph which represents HMG-CoA reductase should be selectively removed. Suboptimal amounts of antireductase IgG (1.9 pg) were used in these studies and, hence, due to the longer exposure time for fluorography, the nonspecific bands appear more prevalent in the fluorograph shown in Fig. 3 (lane I). These data clearly demonstrate that only the 94,000-dalton polypeptide was specifically removed from the sample which contained 31 pg of unlabeled rat liver reductase (Fig. 3, compare lanes 2 and 3). Therefore, we conclude that the subunit molecular weight of the HMG-CoA reductase from chickens is actually 94,000 daltons and not 18,000 daltons as previously reported by Beg et al. (23) for the enzyme purified from chicken liver. Also shown in Fig. 3 is an immunoprecipitate from ["S]methionine-labeled rat hepatocytes (lune 4). The molecular weights of the rat liver and chicken myeloblast enzymes were identical when analyzed on the SDS, 8 M urea gels.

K-
Immunoprecipitates from cells labeled with ['*PIPOI also revealed the presence of the reductase polypeptide ( M , = 94,000) and clearly demonstrate that the enzyme is phosphorylated in vivo (Fig. 2, lane 4). Another phosphorylated band ( M , = 108,000) was also immunoprecipitated with anti-reductase IgG and was absent in samples treated with normal serum IgG (Fig. 2, lunes 4 and 5). Immunoprecipitations performed in the presence or absence of unlabeled pure rat liver reductase, as described above, demonstrated that only the M, = 94,000 polypeptide was selectively removed from the immunoprecipitate (data not shown). The identity and function of the M , = 108,000 band are unknown.
No proteolysis was detected during immunoprecipitation of the radiolabeled reductase. Identical amounts of enzyme were immunoprecipitated in the presence or absence of the protease inhibitors (leupeptin, aprotinin, and phenylmethylsulfonyl fluoride). We conclude that the endogenous proteases are inhibited by the detergents (SDS, Triton X-100, and deoxycholate) present in buffer B.
Effect of 25-Hydroxycholesterol on HMG-CoA Reductase Activity-Addition of 25-hydroxycholesteroI at concentrations ranging from 0.1 to 10 pg/ml caused approximately an 83% decrease in enzyme activity after 120 min (Fig. 4). Trypan blue dye exclusion tests revealed no apparent toxicity due to addition of 25-hydroxycholestero1 or ethanol after a 120-min exposure, and a concentration of 5 pg/ml was used in all subsequent experiments.
The 25-hydroxycholesterol-induced inhibition of enzyme  (15). Enzyme activity was determined by radioassay (14), and protein concentration was measured by a Coomassie blue dye binding assay (16). The results are the mean values from two determinations. Results varied by less than 15%.
activity was rapid and an approximate 80% inhibition was observed by 60-90 min after addition (Fig. 5). The rate of enzyme inactivation by the hydroxysterol was the same for cells incubated in the complete media or in media containing dialyzed serum and no tryptose phosphate broth (Fig. 5). Maximal inhibition after 120 min was approximately 83% of the initial reductase activity. Similar results were also observed in rat hepatocytes treated with 25-hydroxycholesterol (data not shown).
Effect of 25-Hydroxycholesterol on the Rate of Phosphorylation of HMG-CoA Reductase-Phosphorylation and subsequent inactivation of HMG-CoA reductase has been reported to be a mechanism for rapidly regulating reductase activity (9). To determine if this regulatory mechanism was involved in the 25-hydroxycholesterol-induced inhibition of enzyme activity, cells were preincubated for 60 min with [szP]P04 to radiolabel [y-32P]ATP and the cells were then incubated for 20,40, or 60 min in the same media in the absence or presence of 25-hydroxycholestero1 (5 pg/ml). The amount of [32P]P04 incorporated into the reductase polypeptide was determined by immunoprecipitation. If the rapid decline in reductase activity were due to enzyme phosphorylation, we would predict a significant increase in the amount of ' "P associated with the ( M , = 94,000) polypeptide. Short incubation periods (20,40, and 60 min) with 25-hydroxycholesterol were used in this experiment to ensure that we would observe any rapid but possibly transient increase in reductase phosphorylation.
Addition of 25-hydroxycholesterol to the myeloblasts had no effect on the total rate of incorporation of ["2P]P04 into the cells (Fig. 6A). We conclude that 25-hydroxycholesterol did not affect the specific activity of the [y-"PIATP since the radioactive content of total cellular proteins was not affected by the hydroxysterol. The total cellular incorporation continuously increased in Fig. 6 25-Hydroxycholesterol was added (final concentration 5 gg/ml) to cells (1 X 107/ml) suspended either in complete media (A) or in media containing dialyzed sera and no tryptose phosphate broth (0). Aliquots were removed at the specified time periods. The myeloblasts were treated as stated in the legend to Fig. 4 for determination of reductase activity. Enzyme activity is represented as a percentage of the time zero value (125 pmol/min/mg of protein). The results are the mean values from two determinations. Results varied by less than 15%.
were observed for the incorporation of [32P]POq into the reductase polypeptide ( M , = 94,000) immunoprecipitated from control cells (Fig. 6B, lanes 2-5). However, after addition of 25-hydroxycholesterol to the cells, there was no increase in the amount of [32P]P04 incorporated into the enzyme (Fig.  6B, lanes 6-8). This is shown quantitatively in Fig. 6C, where the M , = 94,000 bands were cut out of the gel shown in Fig.   6B and the radioactivity was determined. The apparent decrease in the rate of phosphorylation of the reductase polypeptide in 25-hydroxycholesterol-treated cells was presumably due to the 25-hydroxycholesterol-induced changes in the rates of reductase degradation and synthesis (see below). Therefore, 25-hydroxycholesterol clearly did not enhance the level of reductase phosphorylation, and we conclude that 25hydroxycholesterol does not inhibit enzyme activity by promoting enzyme phosphorylation.
Effect of 25-Hydroxycholestero1 on the Rates of Reductase Degradation and Synthesis-Although the addition of 25hydroxycholesterol did not affect the rate of total cellular protein degradation (Fig. 7 A ) , it significantly increased the relative rate of degradation of HMG-CoA reductase, (Fig. 7B, compare lanes 3-5 and 6-8). The apparent half-life of the enzyme in control cells was approximately 196 min, while in the 25-hydroxycholesteroI-treated cells the apparent half-life was only 55 min (Fig. 7C). Hence, 25-hydroxycholestero1 caused an approximate 350% increase in the relative rate of reductase degradation.
25-Hydroxycholesterol did not affect the rate of total cellular protein synthesis (Fig. 8A). After addition of 25-hydroxycholesterol a rapid and significant decrease in the rate of enzyme synthesis was observed, and the effect was apparent 10 min after addition (Fig. 8B, compare lanes 2-7 and 8-13). When the radioactivity present in the M , = 94,000 polypeptide was determined (Fig. 8C), addition of 25-hydroxycholesterol decreased the relative rate of reductase synthesis by approximately 50%. The rates of synthesis were calculated from data on the 10-and 20-min time points in order to minimize any interference due to the effect of 25-hydroxycholestero1 on enzyme degradation. The 50% inhibition of the rate of enzyme synthesis after 20 min may underestimate the actual value because the 25-hydroxycholesterol and [35S]methionine were added simultaneously and the 20-min incubation period would also include the time required for the 25-hydroxycholesterol to bind to the cells. To examine this point, cells were treated with 25-hydroxycholesterol for 120 min, then pulsed with ["Slmethionine for either 10 or 20 min and the reductase was  purified from cells treated with 25-hydroxycholesterol for 120 min and containing only 20% of enzyme activity present in the control cells. RNA from either the 25-hydroxycholesteroltreated or control cells was translated in uitro in the rabbit reticulocyte lysate translation system and the translation products were immunoprecipitated (Fig. 9, lane I). Two polypeptides (MI = 102,000 and 94,000) were specifically precipitated with anti-reductase IgG and immunoprecipitation of these polypeptides could be selectively eliminated by addition of unlabeled pure rat liver reductase (Fig. 9, lanes 2). When the two bands (M, = 102,000 and 94,000) were cut out of the gel and the radioactivity was determined, the two polypeptides contained almost equal amounts of radioactivity. The ratio of the radioactivity present in the 102,000-dalton polypeptide/ 94,000-dalton polypeptide was 1.00 * 0. 16  Comparison of the relative amounts of reductase mRNA present in 25-hydroxycholesterol-treated and control cells was measured in the translation system using concentrations of mRNA at which the amount of reductase synthesized was proportional to the concentration of mRNA (Fig. 10). These data clearly demonstrate that there was no significant difference in the amounts of reductase mRNA present in 25hydroxycholesterol-treated and control cells even though there was a 72% difference in the relative rates of reductase synthesis in the cells. We conclude that the inhibitory effect of 25-hydroxycholesterol on the rate of reductase synthesis resulted from translational regulation.

DISCUSSION
We previously reported that antibody prepared to purified rat liver HMG-CoA reductase cross-reacted and inactivated the enzyme present in human liver (13) and human monocyte macrophages (24). This same antibody preparation also crossreacted with the reductase in chicken myeloblasts and specifically immunoprecipitated a M , = 94,000 polypeptide. These data suggest that portions of the reductase polypeptide are conserved evolutionarily and, in this regard, it is of interest to note that the chicken and rat enzymes had identical molecular weights.
Identification of the M , = 94,000 polypeptide as HMG-CoA reductase was based on competition experiments using pure rat liver reductase and on selective changes in this polypeptide after either 25-hydroxycholestero1 or mevinolin treatment. The subunit molecular weight of the reductase purified from chicken liver was previously reported by Beg et al. (23) to be approximately 18,000. The apparent discrepancy in the sub-unit molecular weights of the enzyme from chickens probably stems from proteolysis during purification of the chicken liver enzyme since the procedure used to solubilize the reductase has been reported to be inhibited by addition of the protease inhibitor, leupeptin (25). We detected no proteolysis during immunoprecipitation of the enzyme and, therefore, we conclude that the endogenous subunit molecular weight of the HMG-CoA reductase from chickens is 94,000.
The molecular weight of the reductase immunoprecipitated from rat liver was identical to the molecular weight of the enzyme from chicken myeloblasts (Mr = 94,000). Previous investigators have reported the molecular weight of the rat liver enzyme to be approximately 52,000 (16,26). Proteolysis during enzyme solubilization and purification may also explain the apparent discrepancies in molecular weight. We have recently reported that the molecular weight of the enzyme subunit from rat liver is 94,000 (27,28).
Phosphorylation of HMG-CoA reductase has been proposed as a putative mechanism for rapidly regulating reductase activity (9). However the effect of 25-hydroxycholesterol on reductase from chicken myeloblasts was not correlated with enhanced levels of enzyme phosphorylation. Our results actually show a slight decrease in reductase phosphorylation (Fig. 6C) after 25-hydroxycholesterol treatment, but this probably results from the rapid 25-hydroxycholesterol-induced alteration in the relative rates of reductase degradation and synthesis. Work by Erickson et al. (10) also noted that phosphorylation was not associated with 25-hydroxycholesterolinduced changes in rat liver HMG-CoA reductase. These investigators found that the decreased levels of reductase activity following 25-hydroxycholesterol administration to rats were not reversed by treatment of the microsomes with phosphatase. Similarly Cavenee et al. (11) reported the failure of 25-hydroxycholesterol to suppress reductase activity in enucleated Chinese hamster ovary cells even though the (ATP-Mg2')-dependent system for inactivating the reductase was present. They also concluded that 25-hydroxycholesterol did not decrease enzyme activity by inducing phosphorylation of the reductase. Both of these reports (10, 11) utilized relatively indirect methods for determining phosphorylation; however, our results from immunoprecipitation of ["P]P04labeled HMG-CoA reductase provide conclusive evidence for the absence of enhanced enzyme phosphorylation in 25-hydroxycholesterol-treated cells.
Bell et al. (29) and Erickson et al. (30) had proposed, from measurements of changing enzyme activity, that 25-hydroxycholesterol enhanced the rate of reductase degradation. Our direct studies confirm and extend their observations. The 350% increase in the relative rate of reductase degradation following 25-hydroxycholesterol treatment (Fig. 7) was similar to results reported by Faust et al. (5) for Chinese hamster ovary cells. In the latter study the conditions used for immunoprecipitation did not prevent proteolysis of the reductase and the rate of degradation was determined primarily from a M , = 62,000 proteolytic fragment of the enzyme.
Administration of 25-hydroxycholesterol also caused a 50-72% inhibition of the relative rate of reductase synthesis (Fig.  8C). The rate of synthesis was determined using short labeling times (10-20 min) in order to minimize the effect of the 25hydroxycholesterol-induced 3.5-fold increase in the rate of degradation. 25-Hydroxycholesterol altered the rate of synthesis as early as 10 min after addition. After 60 min the amount of reductase in the cells appears to be in a steady state (Fig. 8C) which may reflect the approximately 17% of the reductase activity which could not be suppressed by addition of 25-hydroxycholestero1 (Figs. 4 and 5). Faust et al. (5) reported that 25-hydroxycholesterol inhibited the rate of reductase synthesis by approximately 98% in Chinese hamster ovary cells which were adapted to grow in the presence of compactin. The reason for the higher level of inhibition of reductase synthesis noted by these investigators is not known, but it may be due to proteolysis during immunoprecipitation or to different cell types.
The combined effects of the approximately 350% increase in the relative rate of reductase degradation and the 50-72% inhibition in the relative rate of reductase synthesis would be predicted to cause approximately a 86-92% inhibition in enzyme activity. This prediction approximates the observed 83% inhibition of the total reductase activity following 25hydroxycholesterol treatment (Figs. 4 and 5). Therefore, the relative changes in the rates of reductase degradation and synthesis can account completely for the decrease in HMG-CoA reductase activity due to 25-hydroxycholesterol treatment.
The 25-hydroxycholesterol-induced inhibition of the rate of reductase synthesis was not correlated with concomitant changes in the levels of reductase mRNA. The amount of reductase mRNA was not decreased in cells treated with 25hydroxycholesterol for 120 min and containing only 20% of the initial reductase activity (Fig. 10). To ensure that the in vitro translation system was able to detect %fold changes in reductase mRNA, several concentrations of mRNA were tested (Fig. 10). The amount of the total immunoprecipitable reductase ( M , = 102,000 and 94,000) was proportional to the amount of mRNA added to the translation system, and there were no significant differences in the amount of reductase mRNA present in the control or 25-hydroxycholesteroltreated cells. These data clearly suggest that 25-hydroxycholesterol inhibited the rate of reductase synthesis by translational regulation. Translational regulation of HMG-CoA reductase was also reported by Koizumi et al. (31) for increases in enzyme activity in rat hepatocytes due to addition of compactin. These investigators noted that the increases in enzyme activity could be prevented by addition of puromycin, an inhibitor of protein synthesis, but not by addition of either actinomycin D or a-amanitin. Post-transcriptional control of lipoprotein-induced HMG-CoA reductase was also reported in hepatoma cells (32) and human lymphocytes (33). All of these studies (31-33) utilized inhibitors of RNA synthesis and can only indirectly monitor the levels of reductase mRNA. Therefore, our results clearly demonstrate for the first time that HMG-CoA reductase activity can be controlled by translational regulation. The mechanisms involved in this regulation are still unknown.
I n vitro translation of the poly(A+) RNA produced two reductase polypeptides (Mr = 102,000 and 94,000). The presence of two polypeptides might be due to proteolysis during translation of the RNA. This appears unlikely since no precursor-product relationship could be demonstrated and the relative amounts of the M , = 102,000 and M, = 94,000 polypeptides synthesized were always approximately equivalent in all in oitro translations which utilized a wide range of RNA concentrations. The higher molecular weight polypeptide could represent a contaminating polypeptide with similar antigenic sites. This would seem unlikely since we used polyclonal antibody in our experiments, but further work is needed to characterize both polypeptides.
The possibility that the two polypeptides ( M , = 102,000 and 94,000) represent two forms or subunits of HMG-CoA reductase appears intriguing. Chin et al. (8) recently cloned the HMG-CoA reductase from Chinese hamster ovary cells and reported that their cloned DNA hybridized to two differ-ent sizes of RNA present in the cells. These investigators postulated that this may reflect either two different forms of the enzyme or perhaps two different sized mRNAs which code for the same polypeptide.
Chin et al. (8) reported that the molecular weight of the reductase synthesized by in vitro translation of RNA from Chinese hamster ovary cells was predominantly 90,000, which was identical to the molecular weight of the reductase immunoprecipitated from radiolabeled cells. This differs from our results with chicken myeloblast RNA where one of the immunoprecipitated translation products had a higher molecular weight (Mr = 102,000) than the reductase immunoprecipitated from cells (M, = 94,000). These data may suggest that in chicken myeloblasts, one form of HMG-CoA reductase may be synthesized as a larger sized proenzyme which is posttranslationally processed to the M , = 94,000 polypeptide.