The in vivo regulation of rat liver 3-hydroxy-3-methylglutaryl coenzyme A reductase. Phosphorylation of the enzyme as an early regulatory response following cholesterol feeding.

Although substantial evidence supports the conclusion that 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase is the major regulatory enzyme in cholesterol biosynthesis, the molecular events involved in the in vivo regulation of this enzyme have remained obscure. In order to study this problem, rats received a single meal consisting of either rat chow or rat chow containing 2% cholesterol. The rats were killed 60 or 120 min after the beginning of feeding, and liver microsomes were prepared by ultracentrifugation. Two phases of inhibition of microsomal HMG-CoA reductase were observed. The first phase of inhibition, observed 60 min after the beginning of cholesterol feeding, was completely reversed by preincubation of the microsomes with purified phosphoprotein phosphatase. The second phase of inhibition, observed 120 min after the beginning of cholesterol feeding, was not reversed by phosphoprotein phosphatase. These results are consistent with the conclusion that phosphorylation of HMG-CoA reductase is the first step in a series of in vivo regulatory events which produce inactivation and ultimately degradation of the enzyme.

Substantial evidence supports the conclusion that 3-hydroxy-3-methylglutaryl coenzyme A reductase (mevalonate: NADP' oxidoreductase (CoA-acylating), EC 1.1.1.34) is the major regulatory enzyme in the biosynthesis of cholesterol (1,2). In a recent study (3), we described the immunotitration of HMG-CoA' reductase with HMG-CoA reductase antiserum. It was found that regulation of HMG-CoA reductase in various physiological states, i.e. cholesterol feeding, cholestyramine feeding, and diurnal rhythm, occurs both by changes in the concentration of the enzyme and by changes in the activity state of the enzyme.

* This research was supported by United States Public Health
Service Grants HL-16, 796-06, AM-10, 628-14, Fellowship 1F32 HL-005803, and Biomedical Sciences Advancement Program 5-SO6-RR-08139-06 from the National Heart, Lung and Blood Institute. 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 present investigation focuses on the processes which change the activity state of HMG-CoA reductase in uiuo. In order to study this problem, rats received a single meal consisting of either rat chow or rat chow containing 2% cholesterol. The rats were killed 60 or 120 min after the beginning of feeding and liver microsomes were prepared by ultracentrifugation. The results show that, at 60 min after the beginning of cholesterol feeding, the inhibition of rat liver microsomal HMG-CoA reductase, produced in uiuo, can be completely reversed by preincubation of the microsomes in vitro with purified phosphoprotein phosphatase. By contrast, however, at 120 min after the beginning of cholesterol feeding, the inhibition of rat liver HMG-CoA reductase, produced in uiuo, was not reversed by in uitro preincubation of the microsomes with phosphoprotein phosphatase.

EXPERIMENTAL PROCEDURES
Materials-The materials used in this article were obtained from described sources (3) 2), except where noted. The livers were homogenized with a tight-fitting, motor-driven, Teflon pestle, using 2.0 to 2.5 ml of Buffer A/g of liver. The homogenate was centrifuged at 10, OOO X g for 15 min, the supernatant was collected, and centrifugation (10,000 X g) was repeated. The 10, OOO X g supernatant was centrifuged at 303,000 X g (50.2 Ti rotor) for 30 min, the supernatant was removed, the pellets were resuspended in Buffer A, and centrifugation was conducted again (303,000 X g for 30 min). The supernatant was discarded, and the microsomal pellet was resuspended in Buffer A, using a tight-fitting, motor-driven, Teflon pestle. The resulting microsomal suspension was applied to a Sephadex G-25 column (PD-IO), eluting with Buffer A-I (0.1 M sucrose, 0.05 M KCI, 0.04 M imidazole, 0.03 M EDTA, 0.01 M dithioerythritol, at pH 7.2).* Reactivation studies with puritied phosphoprotein phosphatase (see below) were always performed using fresh microsomal suspensions prepared within 5 h of death. Protein was determined by the method of Bradford (4).
Activation of Rat Liver Microsomal HMG-CoA Reductase by Purified Phosphoprotein Phosphatase-Highly purified phosphoprotein phosphatase (0.08 mg/ml, see below) was subjected to gel fdtration (G-25 column), using Buffer A-I. Aliquots (100 pl) of the fresh microsomal suspension (see above) were added to incubation tubes; aliquots of purified phosphoprotein phosphatase (0 to 100 pl) were added, and the volume of each tube was adjusted to 200 pl with Buffer A-I. NADPH (dissolved in 20 pl of Buffer A-I) was added to each tube (final concentration, 4 mM), and the tubes were preincubated in a Dubnoff incubator (with shaking) at 37'C for 20 min, followed by the addition of ['4C]HMG-CoA (132,000 dpm, 300 p~ final concentration), dissolved in 30 pl of distilled water, to each tube. Incubation was then conducted at 37°C for 20 min. Mevalonate The phosphate in Buffer A was removed to prevent inhibition of the purified phosphoprotein phosphatase during the reactivation studies; Buffer A-I was found to be optimal for reactivation.

In Vivo Regulation
of Rat Liver HMG-CoA Reductase formation was linear for the incubation period employed. Mevalonate formation was assayed as described (5).
Purification of Rat Liver Phosphoprotein Phosphatase-Rat liver phosphoprotein phosphatase (phosphorylase phosphatase) was purified by a modification (see below) of the method of Brandt et al. (6). Liver 303,000 X g supernatant from untreated rats was the starting material for this purification, using Buffer I (0.05 M imidazole, 5 mM EDTA, 0.5 mM dithioerythritol, at pH 7.45). The following purification steps (6) were conducted: (i) ammonium sulfate precipitation, (ii) ethanol precipitation, and (iii) ion exchange chromatography, employing a DEAE-Sephadex (A-50) column, using 0.24 M sodium chloride in Buffer I to elute the phosphoprotein phosphatase. The specific activity was 600 to 750 units/mg of protein (6).
Phosphoprotein Phosphatase Assay-The method used for measuring phosphoprotein phosphatase activity was based on the method of Thomas and Wright (7) for glycogen phosphorylase. Bovine serum albumin, in the same amount as the phosphoprotein phosphatase, was incubated with phosphorylase a as a control. Fig. 1 and Table I show the resuits of an experiment in which rats were fed a single meal of either rat chow (control) or rat chow containing 2% cholesterol. Feeding was begun at 8:30 a.m., and the animals were killed 60 min later at 9:30 a.m. Liver microsomes were prepared by homogenization and centrifugation (see "Experimental Procedures"). Microsomal HMG-CoA reductase was assayed after preincubation of the microsomes with varying quantities of purified phosphoprotein phosphatase. Fig. 1 and Table I show a substantial inhibition of microsomal HMG-CoA reductase activity (26.7 f 3.3%), observed 60 min after the beginning of cholesterol feeding. This inhibition was statistically significant ( p = 0.00004). Furthermore, this inhibition was completely reversed by preincubation of   Table IT show the results of an experiment in which rats were fed a single meal of either rat chow (control) or rat chow containing 2% cholesterol. Feeding was begun at 7:30 a.m. and continued until 830 a.m. The animals were killed at 9:30 a.m., 120 min after the beginning of feeding. Liver microsomes were prepared (see "Experimental Procedures"), and microsomal HMG-CoA reductase was assayed after preincubation of the microsomes with varying quantities of purified phosphoprotein phosphatase.  Feeding was begun at 730 a.m. for 1 h, and the animals were killed 120 min later at 930 a.m. Aliquots (100 pl, 3.1 to 5.8 mg/ml of protein) of a given microsomal suspension were preincubated with 0-, lo-, 20-, 50-, or 100-pl aliquots of purified phosphoprotein phosphatase. Four pairs of rats, pooling the livers for each pair, were used for the control, and five pairs of rats, pooling the livers for each pair, were used for the cholesterol-feeding experiments. Therefore, each experimental point for the control is the mean of four experiments f the standard error of the mean (indicated by brackets). Each experimental point for the cholesterol-fed animals is the mean of five experiments f the standard error of the mean (indicated by brackets). HMG-CoA reductase specific activity is plotted as a function of the amount of purified phosphoprotein phosphatase added to the incubation.

Reactivation of rat liver microsomal HMG-CoA reductase by Preincubation of the microsomes with purified phosphoprotein phosphatase
Rats were fed a single meal of either rat chow (control) or rat chow containing 2% cholesterol. Feeding was begun at 7:30 a.m. for 1 h, and the animals were killed 120 min later at 930 a.m. Other experimental details are described in the legend for Fig. 2. All values shown for HMG-CoA reductase specific activity are f the standard error of the mean.   (16,17). Siperstein and co-workers inferred that the principal biosynthetic step regulated was the conversion of HMG-CoA to mevalonic acid (16, 18, 19). The first direct demonstration that cholesterol feeding blocked HMG-CoA reductase was made by Linn (20) and was later confirmed by Shapiro and Rodwell (21).

HMG-CoA reductase suecific activitv
Studies from several laboratories are consistent with the conclusion that cholesterol feeding may suppress HMG-CoA reductase activity by decreasing the quantity of enzyme protein present (22)(23)(24)(25)(26) or by the inhibition of pre-existing enzyme (24-26). In a recent study, we described the immunotitration of HMG-CoA reductase with HMG-CoA reductase antiserum (3). It was found that regulation of HMG-CoA reductase, following cholesterol feeding for 7 days, occurred both by a change in the concentration of the enzyme and by a change in the activity state of the enzyme.
Although in vitro evidence exists supporting the concept that HMG-CoA reductase activity is inhibited by phosphorylation and activated by dephosphorylation (27-36), there has been a lack of evidence that this mechanism actually is physiologically relevant and is operative in vivo. In fact, the study of Brown et al. (37) seemed to exclude phosphorylation/ dephosphorylation of HMG-CoA reductase as an in vivo regulatory phenomenon, when liver microsomes were examined 12 to 72 h after a variety of physiological manipulations (fasting, cholesterol-feeding, cholestyramine-feeding, etc.).
However, it is clear from the present study that the reason that Brown et al. did not observe evidence for the in vivo regulation of HMG-CoA reductase by phosphorylation was the long time interval used (12 h) between the regulatory intervention, e.g. cholesterol-feeding and the killing of the animals. In the experiments conducted in the present investigation, we demonstrate that cholesterol feeding produces: (i) a rapid inactivation in vivo of liver HMG-CoA reductase, observed 60 min after the beginning of cholesterol-feeding, which can be completely reversed by in vitro preincubation of the microsomes with phosphoprotein phosphatase (Fig. 1 and Table I) and (ii) a further inactivation of HMG-CoA reductase, observed 120 min after the beginning of cholesterol feeding, which is not reversed by in vitro preincubation of the microsomes with phosphoprotein phosphatase (Fig. 2 and Table 11). Thus, the timing of the experiments is crucial. These results are consistent with the conclusion that phosphorylation of HMG-CoA reductase is the first step in a series of in vivo regulatory events which produce inactivation and, ultimately, degradation of the enzyme.
In a related investigation, we studied the effect of a single dose (100 mg) of mevalonolactone, administered to rats by intragastric tube, on liver microsomal HMG-CoA reductase (38, 39). This experimental model allows the study of the effect of a sudden pulse of newly synthesized sterol on the regulation of HMG-CoA reductase (40,41). When microsomes were isolated in the absence of KF (38, 39), a 38.2 & 2.1% inhibition of HMG-CoA reductase activity was observed 20 min after mevalonolactone administration ( p = O.OooOo2, compared to control). Furthermore, this inhibition was completely reversed by preincubation of the microsomes with purified phosphoprotein phosphatase. When microsomes were prepared in the presence of 50 m~ KF, a 35.9 & 2.8% inhibition of HMG-CoA reductase was observed 20 min after mevalonolactone administration ( p = O. OOO1, compared to control). As above, this in vivo inhibition was reversed (after removal of the KF) by treatment of the microsomes with phosphoprotein phosphatase (38,39). Therefore, it made no difference, as far as the per cent inhibition of HMG-CoA reductase, whether the microsomes were prepared in the presence or absence of KF.
As in the present study with cholesterol feeding, two phases of inhibition were observed with mevalonate feeding (38,39). As mentioned above, the first phase of inhibition, observed 20 min after mevalonolactone administration, was completely reversed by preincubation of the microsomes with purified phosphoprotein phosphatase. The second phase of inhibition, observed 60 min after mevalonolactone administration, was not reversed by phosphoprotein phosphatase. The reactivation of liver microsomal HMG-CoA reductase by phosphoprotein phosphatase, in vitro, was blocked by either potassium fluoride or by phosphoprotein phosphatase inhibitor.
In summary, the findings obtained in the present article are consistent with the conclusion that phosphorylation of rat liver HMG-CoA reductase, the rate-limiting enzyme in cholesterol biosynthesis, is an early in vivo regulatory response, following cholesterol feeding, and that this initial response is followed by other as yet unknown regulatory events.