Evidence Indicating that Inactivation of 3-Hydroxy-3-methylglutaryl Coenzyme A Reductase by Low Density Lipoprotein or by 25-Hydroxycholesterol Requires Mediator Protein(s) with Rapid Turnover Rate*

The half-life ( t l , ~ ) of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase of Chinese hamster ovary cells grown in fetal calf serum medium is ap- proximately 2 h. When cells are switched to grow in delipidated serum medium (DeL-M) for more than 24 h, the tl/z of the enzyme is found to be drastically altered to approximately 13 h. Exposure of low density lipo- protein (LDL) (100 pg of protein/ml) or 25-hydroxycho-lesterol (1 pg/ml) to cells grown in DeL-M suppresses reductase activity more rapidly than would be expected solely if reductase synthesis were suppressed, showing that inactivation of reductase activity by sterols, pre- viously demonstrated using only analogs of cholesterol, is a normal mechanism for regulation of HMG-CoA reductase activity by the physiologically important sterol source (LDL). This inactivation effect by LDL or by 25-hydroxycholesterol is shown to be at least in part due to acceleration of reductase degradation rate. Fur-thermore, the inactivation effect by sterols is shown to be largely abolished if cycloheximide (250 pg/ml) is added simultaneously to the growth medium, indicat- ing that continuous synthesis of a class of mediator protein(s) is necessary in mediating the effect of LDL or 25-hydroxycholesterol. T w o These experiments suggest that an important mechanism of reductase inactivation by LDL or by 25-OH cholesterol is to accelerate the reductase degradation rate. Further- more, these results are consistent


From the Department of Biochemistry, Dartmouth Medical School, Hanouer, New Hampshire 03755
The half-life ( t l ,~) of 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase of Chinese hamster ovary cells grown in fetal calf serum medium is approximately 2 h. When cells are switched to grow in delipidated serum medium (DeL-M) for more than 24 h, the tl/z of the enzyme is found to be drastically altered to approximately 13 h. Exposure of low density lipoprotein (LDL) (100 pg of protein/ml) or 25-hydroxycholesterol (1 pg/ml) to cells grown in DeL-M suppresses reductase activity more rapidly than would be expected solely if reductase synthesis were suppressed, showing that inactivation of reductase activity by sterols, previously demonstrated using only analogs of cholesterol, is a normal mechanism for regulation of HMG-CoA reductase activity by the physiologically important sterol source (LDL). This inactivation effect by LDL or by 25-hydroxycholesterol is shown to be at least in part due to acceleration of reductase degradation rate. Furthermore, the inactivation effect by sterols is shown to be largely abolished if cycloheximide (250 pg/ml) is added simultaneously to the growth medium, indicating that continuous synthesis of a class of mediator protein(s) is necessary in mediating the effect of LDL or 25-hydroxycholesterol. T w o different protein synthesis inhibitors (emetine and puromycin) were used and gave essentially identical results. Preincubation of cell culture with cycloheximide for 2 h essentially completely abolishes the effect of 25-hydroxycholesteroI, indicating that the mediator protein(s) turns over rapidly, with t~,~ less than 3 or 4 h. It has been established that a key-regulated reaction in the biosynthesis of cholesterol from acetyl units is the conversion of 3-hydroxy-3-methylglutaryl coenzyme A to mevalonate (for reviews, see Refs. 1-3), catalyzed by the enzyme HMG-COA' reductase (EC 1.1.1.34). The activity of this enzyme appears to be regulated by cholesterol through complex feedback * 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. suppressions. Currently, it is believed that several regulatory mechanisms by which sterols modulate the reductase activity in various biological systems probably exist, e.g. earlier studies by Higgins and Rudney (4,5 ) suggested that cholesterol feeding lowered rat hepatic HMG-CoA reductase activity by two mechanisms, an immediate inactivation of preformed enzyme and a longer term reduction of enzyme synthesis. The work by Edwards and Gould (6) is also consistent with the concept that the initial cholesterol-enhanced diurnal decline of reductase activity in rat liver may be attributed either to direct enzyme inactivation or to an increased rate of enzyme degradation. In cultured human fibroblast cells and other cultured mammalian cells, removal of lipoproteins from the growth medium causes a large increase in reductase activity, which has been shown to be dependent on continuous protein synthesis (7). Adding low density lipoprotein (LDL) to the growth medium specifically and rapidly suppresses reductase activity (8); results of analyses of the first order kinetics of the decline of reductase activity in the presence of LDL or cycloheximide are consistent with the idea that the LDL-derived cholesterol acts solely by decreasing the rate of reductase enzyme synthesis (8). The work by Kirsten and Watson (9) using unfractionated lipoproteins and hepatoma tissue culture cells is largely consistent with this view. In contrast to these results with LDL, certain oxygenated analogs of cholesterol, including 7-ketocholesterol and 25-hydroxycholestero1(25-OH cholesterol) (lo), lower reductase activities in several tissue culture cell lines (11-13) more rapidly than would be expected solely if reductase synthesis were suppressed, suggesting that these sterols may act by inactivating the enzyme by acceleration of enzyme degradation or some other inactivation mechanism(s). Whether this inactivation mechanism operates using LDL as the sterol source has not been extensively studied. The possibility that sterols may act through the phosphorylation-dephosphorylation cycle to modify the reductase protein has been suggested (14-16); however, at present, there is no evidence that this mechanism may play a significant role in tissue culture cells (17,18).
In this report, we have chosen to use Chinese hamster ovary (CHO) cells as a model system to examine the mode(s) of action of LDL or 25-OH cholesterol in suppressing HMG-CoA reductase activities. It has been previously shown (18)(19)(20) that the LDL pathway for regulation of cholesterol metabolism (21) is operational in this cell line.

Materials
Radioactive chemicals were from New England Nuclear. Biochemicals were from Sigma. 25-OH cholesterol was from Steraloids. 25-OH [24-'H]cholesterol was synthesized chemically as previously described (20). Purities of radioactive and nonradioactive 25-OH cholesterol were greater than 95% as previously reported (20). 25-OH cholesterol 3P-oleate was synthesized and purified according to Goodman (22) by Drs. J. Nelson and T. Spencer a t the Chemistry Department of Dartmouth College. Its structure was unambiguously established by IR and NMR analyses. Purity of this compound as analyzed by three different thin layer chromatographic (TLC) solvent systems was found to be greater than 98%. Human LDL was prepared according to published procedure (23) as previously (20). Fetal calf serum was from Microbiological Associates. All other chemicals were of analytical grades.

Methods
Cells-Monolayer cultures of normal (wild type) CHO cells (20,24) were grown in 25-cm2 Falcon flasks or 60-mm Falcon dishes in F-12 medium (linoleic acid deleted) plus 10% fetal calf serum (FCS-M) or plus 10% delipidated fetal calf serum (DeL-M). This is the same cell line as the normal cell line previously employed in this laboratory (20). The original cell line (CHO-Kl) was obtained from American Type Culture Collection in 1976. Over the past few years, this cell line was recloned at three different occasions to achieve a more homogeneous population in morphology. For the current cell line, over 90% of the cell populations exhibits mononuclear morphology under the microscope. Linoleic acid was deleted from the F-12 medium since we found that it served no purpose in supporting the growth of normal CHO cells in FCS-M or in DeL-M. Delipidated serum was prepared according to published procedure (25). 25-OH cholesterol was added to the growth medium from stock dimethyl sulfoxide (Me2SO) solution such that the final Me2530 concentration in the medium was equal to or less than 0.3% as previously described (20). LDL, cycloheximide, emetine, or puromycin was added to the growth medium immediately before use from a 50-fold or 100-fold stock solution in sterile saline. All experiments reported in this paper involving cell culture were carried out using the following format.
Stock flasks of cells grown at high cell density (21 X lo5 cells/cm' surface area) in FCS-M were dissociated by 0.05% trypsin, seeded at 0.03 x lo6 cells/25-cm2 flask or 60-mm dish in 3 ml of FCS-M, and allowed to grow for 63 h at 37 "C in a 5% COn incubator. Cells at this stage are designated as starting cultures. Afterward, the medium was changed and cells were allowed to grow in various media as indicated in specific figure legends and tables. Media were refreshed frequently during cell growth. We found that frequent medium changes produce cells grown in FCS-M and DeL-M with higher in titro reductase activities, as well as many other enzyme activities. Cells were harvested during log phase growth by visual examination. For all experiments reported here, duplicate cultures were used for each assay point. Variation of measurement between duplicate cultures was within 5 to 7%.
Cell Homogenization and HMG-CoA Reductase Assay-Cells were harvested by a rapid and efficient procedure developed in this laboratory and reported in detail elsewhere (26). Briefly, cell monolayers were rapidly washed by 3 X 5 ml of PBS and 1 X 5 ml of hypotonic buffer (Buffer K) (1 mM Tris-HCI, 1 mM EGTA, and 1 mM MgCL, pH 7.6) at 4 "C and exposed to 5 ml of Buffer K at room temperature for 2 min. The buffer was drained, 0.34 &dish of Buffer K at room temperature was added, and the swollen cells were rapidly scraped into a corner by a metal scraper fitted with a silicone rubber blade (Bellco Co., Catalog No. 7731-22000). Cell homogenization took place during scraping. The extent of cell breakage was greater than 99% by microscopic examination. Usually, 100 pl of the cell homogenate was mixed with 25 p1 of Buffer A (consisting of 100 mM concentration of imidazole and 25 mM concentration of dithiothreitol, pH 7.4) and preincubated at 37 "C for 10 min for activation. It is known that this step provides the complete dephosphorylation and activation of the reductase by the endogenous phosphatase activity present in the cell extract (17, 18). In our previous work, the buffer used during the activation step contained 50 nm phosphate (27) which might have inhibited the phosphatase activity. After preincubation, duplicate aliquots (usually 40 pl/aliquot, containing 15 to 40 pg of protein) of cell extracts were taken and assayed for HMG-CoA reductase activity as described previously (18,20,24) with minor modifications. The complete assay system contained in 80 pl: 0.245 pmol of NADP', 10.4 nmol of D,L-['~C]HMG-COA (90,000 dpm), 1.78 pmol of glucose 6-phosphate, 0.073 unit of glucose-6-phosphate dehydrogenase, 2.84 pmol of potassium phosphate (pH 7.4), 0.64 pmol of dithiothreitol, 4 pmol of potassium chloride, 1.78 pmol of EDTA, and 800 nmol of imidazole (pH 7.4). There were also small amounts (20 nmol each) of Tris-HC1, EGTA, and MgC12 present. Assay condition was at 37 " c for 60 min with shaking. Reaction was terminated by 10 pl of concentrated HC1. Eight pl of ['H]mevalonic acid in water (560 nmol, 35,000 dpm) was added as internal standard. The mixture was either stored at 4 "C or incubated at 37 "C for 1 h to cause lactonization of mevalonate. After lactonization, 40 pl/sample was spotted on Whatman LK5D silica plate and analyzed by TLC (acetone/benzene, 1:l). The region RF 0.65 to 0.82 was removed by scraping into a 5-ml size scintillation vial and counted after addition of 3.5 ml of scintillation fluid. Recovery of ['H]mevalonate averaged approximately 50%. The minor modifications in compositions of the cell homogenization buffer and assay mixture as indicated above, plus the accumulated technical experience in handling the cell growth and the assay procedures, resulted in approximately 2to 3-fold increase in specific activities of the reductase from both the FCS-"grown culture and the DeL-Mgrown culture over previously reported values (20) for the same cell line. However, the fold increase in reductase specific activity upon removal of serum lipids from the growth medium (approximately a 14-fold change) remained approximately the same as previously reported values (20). Variation between duplicate assays was within 7%. Control experiments have indicated that cell extracts prepared by this procedure contain HMG-CoA reductase activities as well as many other microsomal or cytoplasmic enzyme activities very similar to those of cell extracts prepared by the conventional procedure involving scraping, centrifugation, hypotonic shock, and Dounce homogenization as previously described (26). Using this new procedure, the recovery of the whole cell homogenate was found to be much higher and much more consistent (26). Protein was determined by the method of Lowry et al. (28).
Determination of Per cent Esterification of Cell-bound 25-OH Cholesterol after ['H]25-OH Cholesterol Incorporation into Cells-At the end of the incorporation period, cells were extensively rinsed and were harvested by scraping into 1.0 ml of PBS as described previously (20). Aliquots of the cell suspension were withdrawn for total 'H incorporation and protein determinations as described (20). The remaining cell suspension (0.8 ml) was extracted with 20 ml of CHCla/CH30H (2:l) at room temperature; 100 pg/tube of 25-OH cholesterol was added as carrier during extraction; 4 ml of H20 was added/tube; and the tube was blended on a Vortex mixer to cause the two-phase separation. After centrifugation at 800 X g at room temperature for 5 min, the separated CHCla phase was washed once with 3 ml of water (saturated with NaCl and CHCL) and twice with 3 ml of the upper phase solvent (CHCL/CH30H/H20 (3:4847)). Aliquots were taken for extraction recovery measurement and analysis by TLC. Extraction recovery of total 'H counts averaged approximately 80%. For TLC analyses, Whatman LK5D silica gel plates were used. Three different solvent systems were used: (a) petroleum ether/ ether/acetic acid (9O:lOl); ( b ) benzene/ethyl acetate (3:l); and (c) benzene. RF values for 25-OH cholesterol were found to be 0.15, 0.49, and 0.09 in these three systems. The corresponding RF values for 25-OH cholesterol-3P-oleate were found to be 0.33, 0.84, and 0.18. To analyze the radioactive TLC chromatogram, the various lipid standards were visualized by iodine staining and marked and then the entire chromatogram was divided into 17 bands and scraped into vials and counted. Monoesters of [3H]25-OH cholesterol were identified unambiguously by their co-migrations in TLC with chemically synthesized 25-OH cholesterol-3P-oleate standard using these three systems. The per cent esterification of r3H]25-OH cholesterol was determined by the "H counts found in the ester band divided by the sum of "H counts in the ester hand plus that found in the free sterol band in TLC analyses. growth medium at 1 pg/m12 causes rapid suppression of this enzyme activity. The fact that the rate of suppression of reductase activity by 25-OH cholesterol occurs much faster than its normal degradation rate (Fig. l B ) , coupled with the fact that 25-OH cholesterol fails to inactivate HMG-CoA reductase activity in cell-free extracts (11,12,29), is consistent with the concept that an important effect of 25-OH cholestero: on this enzyme activity may be to accelerate its degradatior rate (13). The suprising finding occurs when 25-OH cholesterol and cycloheximide are added simultaneously to the growth medium. It is seen that the effect of the sterol on reductase inactivation is almost completely abolished by the presence of the protein synthesis inhibitor (Fig. 1B). This experiment strongly suggests that the effect of 25-OH cholesterol on reductase inactivation depends on continuous protein synthesis, i e . some specific mediator protein(s) possessing very short half-life (the upper limit of t112 is in the order of hours) may be required to mediate the intracellular action(s) of 25-OH cholesterol. It can also be seen (Fig. 1B) that the effect of cycloheximide on 25-OH cholesterol action(s) never reaches

Effect of Cycloheximide and/or 25-OH Cholesterol on HMG-CoA Reductase Activity of CHO Cells-As shown in
Since it is well known that cycloheximide inhibits incubation with the cells, the concentration for 25-OH cholesterol to In experiments not shown, we found that after a 3-h or a 6-h exert its maximal effect in suppressing the reductase activity was at 1 to 5 pg/ml; the concentration needed to cause half-maximal suppression was between 0.1 and 0.2 pg/ml. The same result was seen in at least three separate experiments.
protein synthesis in intact cells within minutes of its addition, this observation suggests that before exposure to 25-OH cholesterol, cells do contain mediator protein(s), but only at relatively low concentration(s). This may explain why the dependency on protein synthesis for the action(s) of 25-OH cholesterol was not demonstrated in hepatoma tissue culture cells by Bell et al. (13). Perhaps under their conditions, the concentration(s) of the pre-existing mediator protein(s) was high enough not to be rate limiting for mediating the intracellular action(s) of the sterol in hepatoma tissue culture cells.
As shown in Fig. lA, cellular protein measurements indicate that cells grown in DeL-M have a doubling time of approximately 20 to 22 h; addition of cycloheximide ceases cell division. Addition of 1 pg/ml of 25-OH cholesterol causes no measurable alteration in cell growth in medium with or without cycloheximide during the entire time course. Also, the effect of cycloheximide on 25-OH cholesterol action(s) can not be correlated with its effect on cellular uptake of this sterol, nor can it be correlated with its effect on per cent cellular esterification of this sterol, as shown in Table I.

Effect of Cycloheximide and/or LDL on HMG-CoA Re-
ductase Activity-An experiment using LDL as the sterol source was next carried out. As shown in Fig. 2, the rate of suppression of reductase activity by LDI, is also faster than its normal degradation rate, and its effect is almost completely abolished by the simultaneous presence of cycloheximide. The data shown in Fig. 1 and Fig. 2 reinforce the concept that 25-OH cholesterol is an intracellular analog of LDL-derived cholesterol (20,30). They also indicate that inactivation of reductase by sterol(s), previously demonstrated using only analogs of cholesterol (11-13), is a normal mechanism for regulation of HMG-CoA reductase activity by the physiologically important sterol source (LDL). In this experiment, the t1,2 of the reductase activity in cells grown in parallel in FCS-M was determined and found to be 1.8 h, which is in fair agreement with previously reported tl12 values by other investigators in various systems (see Ref. 1 for a review). Since the specific activity of reductase in zero time culture grown in FCS-M or in DeL-M was found to be 0.17 or 2.29 nmol. min"mg", respectively, assuming cycloheximide causes no significant change in enzyme degradation, it follows that the decrease in rate of degradation of this enzyme (a 6.3-fold change) constitutes an important mechanism accounting for the observed increase in reductase specific activity (a 13.5-fold change) upon removal of serum lipids from the growth medium.

Effect of cycloheximide on cellular uptake of ['H]25-OH cholesterol and on cellular esterification of (3H]25-0H cholesterol
Starting cultures in 60-mm dishes as described under "Methods" were rinsed with 1 X 5 ml of PBS/dish and were switched to grow in 5 ml of DeL-M for 24 h. Cycloheximide was then added to cell culture at 1 mM at 37 "C for the indicated periods of time. Control cultures received equal amounts of PBS for 4 h. Afterward, cells were treated with an additional 2 ml/dish of DeL-M containing ['H]25-OH cholesterol so that the final concentration of 25-OH cholesterol was 1 pg/ml (4.07 X IO4 dpm/pg). Two h later, cells were harvested for cellular protein content and radioactivity determinations as described (20). The per cent esterification of cell-bound ['HH]25-OH cholesterol was determined as described under "Methods." Values shown represent the mean k variation from the mean between duplicate dishes. Effects of Other Specific Protein Synthesis Inhibitors on 25-OH Cholesterol Action($-Emetine and puromycin are two well known protein synthesis inhibitors. The mode of action of either agent is known to be different from that of cycloheximide (31, 32). As shown in Fig. 3, A and B, the results using emetine as the inhibitor confirm the results using cycloheximide. The fact that emetine is somewhat more effective in the 47-h grown culture (Fig. 3B) than in the 24-h grown culture (Fig. 3A) is consistent with the idea that the preexisting mediator protein(s) in cells before exposure to 25-OH cholesterol may be present in higher concentration in the 24h grown culture than in the 47-h grown culture. It is possible that the mediator protein@) is present in very high concentrations in FCS-M grown culture and is rapidly depleted when cells are switched to grow in DeL-M. Further investigations are needed to explore this possibility.
The experiment using puromycin as the inhibitor was not completely satisfactory, since the cells treated with puromycin for 8 h or longer started to detach from the surface of the tissue culture flask approximately 20 to 30% of the cells were lost into the medium at the end of the 10-h period. Nevertheless, the result shown in Fig. 3C qualitatively confirms the data shown in Fig. 1B and Fig. 3B. As a control experiment, the efficiency of these three different protein synthesis inhibitors in intact CHO cells was measured and found to be very high in each case (Table 11). Taken together, these data rule out the possibility that the result shown in Fig. 1B   of LDL or 25-OH cholesterol on inactivation of reductase activity is to accelerate its rate of degradation, then preincubation of cells with LDL or 25-OH cholesterol followed by cycloheximide exposure should provide tlr2 values shorter than the value with cycloheximide alone. This has been shown in Fig. 4. Preincubation of LDL for 3 h accelerates the reductase degradation rate by 2.7-fold; similar treatment with 25-OH cholesterol accelerates the reductase degradation rate by 3.3-fold. These experiments suggest that an important mechanism of reductase inactivation by LDL or by 25-OH cholesterol is to accelerate the reductase degradation rate. Furthermore, these results are consistent with the following working Effect ofprotein synthesis inhibitors on r3H]leucine incorporation to protein in intact CHO cells Starting cultures in 60-mm dishes as described under "Methods" were rinsed with 1 X 5 ml of PBS and allowed to grow for another 24 h in 3 ml of fresh DeL-M. Afterward, cells were exposed to 1.0 ml of fresh DeL-M with or without various inhibitors at the indicated final concentration for 30 min at 37 "C. This treatment was followed by adding 10 pCi/dish of ~-[4,5 'HJleucine (50 Ci/mmol in H20) to cell culture for 1 h at 37 "C. Afterward, cells were washed with 5 X 5 ml of PBS at 4 "C, followed by 5 X 5 ml of 5% trichloroacetic acid at room temperature. The trichloroacetic acid-insoluble material was digested by 1.0 ml/dish of 0.5 M NaOH at room temperature for 30 min. Aliquots were taken for cellular protein content and trichloroacetic acid-insoluble radioactivity determinations. Duplicate dishes were used for each measurement. Variation between duplicates was within 5% from the mean.  Fig. 3, B and C. Afterward zero time flasks were harvested and assayed; the remaining cells were fed with 5 ml/flask of fresh DeL-M containing no addition (0), 0.89 m~ concentration of cycloheximide (X), 100 p g of protein/ml of LDL (O), or 1 pg/ml of 25-OH cholesterol (A). Three h later, certain flasks of cells exposed to LDL or 25-OH cholesterol were further exposed to cycloheximide at 0.89 mM. At the indicated time, cells were harvested and assayed for HMG-CoA reductase activities.
hypothesis. Preincubation of cells with LDL or 25-OH cholesterol sensitizes HMG-CoA reductase protein which in turn is degraded much more rapidly. The sensitization process by LDL or by 25-OH cholesterol requires the presence of mediator protein(s). Once the sensitization process is complete, the mediator protein(s) is no longer needed to cause the rapid turnover of the fully sensitized reductase protein. The reason why cells preincubated with LDL for 3 h and then exposed to cycloheximide showed a decreased rate of enzyme turnover Certain flasks of cells exposing to cycloheximide for 2, 4, or 6 h were further exposed to 1 pg/ml of 25-OH cholesterol for 8 h (A), 6 h (@), or 4 h (V). At the indicated time, cells were harvested and assayed for HMG-CoA reductase activities.
relative to the noncycloheximide-treated cells in the presence of LDL, but not so for the 25-OH cholesterol-treated cells ( Fig. 4), may be due to the fact that the time lag for the LDL action($ was shown to be approximately 3 h longer (Figs. 2 and 4) than that for 25-OH cholesterol; i.e. during the preincubation period (3 h), the sensitization process with LDL may be less complete than with 25-OH cholesterol.
Effect of Preincubation of Cycloheximide on the Effect of 25-OH Cholesterol in Reductase Inactiuation-The data presented in Figs. 1 to 3 strongly suggest that the mediator protein(s) turns over rapidly. This is confirmed by the preincubation of cells with cycloheximide before exposure to 25-OH cholesterol. As shown in Fig. 5, preincubation of cycloheximide for 2 h essentially completely abolishes the effect of the subsequent exposure to 25-OH cholesterol; the 4-h and the 6-h preincubation data confirm the 2-h preincubation data. Since a time lag of approximately 1 to 2 h for the 25-OH cholesterol action($ has always been seen (as typically shown in Fig. 5), the tIr2 of the mediator protein(s) can not be accurately measured from this type of experiment, but can be estimated to be less than 3 to 4 h.

DISCUSSION
Data presented in this report show that in CHO cells, the decrease in rate of degradation of HMG-CoA reductase constitutes an important mechanism accounting for the observed increase in specific activity of the enzyme upon removal of serum lipids from the growth medium. Whether this is a general phenomenon for other cell types grown in similar conditions remains to be investigated. The tl,z values of this enzyme from cells grown in DeL-M determined in different experiments range from 9 to 17 h (see Figs. 1 to 4) regardless of whether cycloheximide or emetine is used as the protein synthesis inhibitor. We do not yet fully understand why this value fluctuates over such a large range; it seems to be related to the freshness and/or quantity of the growth medium and to cell density during the experiment. It is known that alteration of nutrients in the growth medium can alter rates of protein degradation in tissue culture cells (33). This possibility remains to be clarified by future experiments. Evidence is presented (Figs. 1-4) which strongly suggests that both 25-OH cholesterol and LDL accelerate the normal degradation rate of this enzyme by at least 3-to 4-fold (Figs. 1-4), which is consistent with the concept that 25-OH cholesterol is an intracellular analog of LDL-derived cholesterol (11-13). These data also indicate that inactivation of reductase activity, at least in part via acceleration of reductase degradation rate, is a normal mechanism by which LDL exerts its action in suppressing the HMG-CoA reductase activity in log phase CHO cells. Previously, evidence provided by Brown et al. (8) suggested that the major suppression effect by LDL on reductase activity in confluent human fibroblast cells was due to inhibition of specific enzyme synthesis. The difference in results between their study and ours may be due to the fact that the different cells at different stages of growth were used. It is known that control of enzyme synthesis as well as control of enzyme degradation in cultured cells may vary with different growth conditions (33, 34). It is likely that inhibition of reductase synthesis by LDL also occurs under our condition; however, further experiments are needed before it can be unambiguously demonstrated. The pre-existing data (7-13) and the data presented here support the conclusion that regulation of reductase activity by exogenous cholesterol occurs at the level of specific enzyme degradation and specific enzyme synthesis. In addition, it is entirely possible that the LDL-cholesterol or 25-OH cholesterol may cause a direct inactivation of reductase via specific allosteric inhibition; this type of inhibition can not be demonstrated by data presented in this report and remains to be explored.
An unexpected finding demonstrated in this report discloses the need for a class of mediator protein(s) to mediate the intracellular effect of 25-OH cholesterol or LDL in accelerating the reductase inacti~ation.~ The mediator protein(s) is shown to depend on protein synthesis to maintain its competency; the synthesis of the mediator protein(s) may be inducible by serum lipids, by LDL-derived cholesterol, or by the cholesterol analog (25-OH cholesterol) present in the growth medium. Alternatively, it is possible that the mediator protein(s) is constitutive in the cell and may only require an exogenous signal generated from the incoming sterol to manifest its regulatory effect(s) on reductase. The turnover rate of the functional mediator protein(s) in cells grown in DeL-M is shown to be very rapid, with tl,Z estimated to be less than 3 or 4 h. Recently, Kandutsch and co-workers (36, 37) have described two interesting classes of 25-OH cholesterol-binding proteins in various tissue culture cell lines. It is unlikely that these binding proteins bear any functional relationship to the mediator protein(s) reported here since these binding proteins all possess very long half-lives (37). The nature of the mediator protein(s) is unknown at present. It may be a specific binding protein for cholesterol (or its analogs) or may be an enzyme modifying cholesterol into an active suppressor. Other possibilities are not excluded at present. The possibility that it may vealed to us that in 1975, P. Edwards made the observation that the During preparation of this manuscript, a literature survey resuppression effect by lecithin-cholesterol dispersion of HMG-CoA reductase activity in rat hepatocytes may depend on denouo protein synthesis (35). of HMG-CoA Reductase Activity 6179 play a role in controlling the permeability of 25-OH cholesterol into cells has been eliminated (Table I). It is important to realize that the experiment shown in Table I merely indicates that the difference in effects of 25-OH cholesterol with or without cycloheximide can not be accounted for by the difference observed in per cent cellular esterification of this particular sterol. This experiment does not eliminate the possibility that the mediator protein(s) may play a role in mediating the effect of 25-OH cholesterol or LDL-derived cholesterol in stimulating the activity of the enzyme acyl CoA:cholesterol acyltransferase (30). It is also possible that the mediator protein(@ may play a role in mediating the suppression effects of LDL-derived cholesterol or its oxygenated analog (25-OH cholesterol) on other cholesterogenic enzymes (20). Experiments designed to explore these possibilities are currently in progress in this laboratory.