- Email: [email protected]
3-Hydroxy-3-methylglutaryl-coenzyme A reductase
446
Bioc,himrcrr et Bioph.vsrcu Actu.
7 IO
( 19X2)
446-455
Elsevier Biomedical
Press
BBA 51050
3-HYDROXY-3-METHYLGLUTARYL-COENZYME
A REDUCTASE
THE DIFFERENCE IN THE MECHANISM OF THE IN VITRO MODULATION BY PHOSPHORYLATION AND DEPHOSPHORYLATION TO MODULATION OF ENZYME ACTIVITY BY NON-ESTERIFIED CHOLESTEROL S. VENKATESAN Medical
and K.A. MITROPOULOS
Reseurch Council
(Received
Kqv words:
Lipid
Metabolism
Unit, Hummersmith
Hospitul,
Ducune Roud. I.ondon
Wl.!
OHS (U.K.)
July 10th. 1981)
Hydroxymethylglutu~l-CoA
reductrrse; Covalent
modificution;
Cholesterol
trcrnsfer; Cholesterol
s,nthesrs;
(Endoplusnrk
reticulum)
Incubation of rat liver microsomal fraction in the presence of increasing concentrations of a serum preparation and the re-isolation of the treated microsomal vesicles resulted in a progressive increase in the concentration of non-esterified cholesterol, a progressive decrease in the activity of hydroxymethylglutarylCoA reductase and progressive changes in the characteristics of the Arrhenius plots of the enzyme. The changes in the characteristics of the Arrhenius plots of the enzyme in the serum-treated preparations are consistent with a progressive increase in the concentration of non-esterified cholesterol in the environment of the hydroxymethylglutaryl-CoA reductase in endoplasmic reiticulum vesicles. The serum-treated preparations with high non-esterified cholesterol content showed a constant activation energy between 37 and 20°C whereas the enzyme in the non-treated microsomal fraction, the buffer-treated and the lipoprotein-deficient serum-treated preparations showed breaks in the activation energy at about 29°C. The microsomal fraction from rats fed on the standard, cholesterol- or cholestyramine-supplemented diet showed considerable differences in the activity of hydroxymethylglutaryl-CoA reductase and differences in the characteristics of their Arrhenius plots. However, the incubation of the microsomal fraction from the rats in the three experimental conditions with ATP and Mg*+ and the further incubation of the inactivated enzyme with a preparation of cytosolic phosphoprotein phosphatase resulted in Arrhenius plots with similar characteristics to those of the corresponding original microsomal fraction. These results suggest that changes in the concentration of non-esterified cholesterol in the endoplasmic reticular membrane are responsible for the differences in the activity of hydroxymethylghttaryl-CoA reductase in the microsomal fraction from the rats in these dietary conditions.
Introduction
from the enterohepatic circulation, is associated with compensatory changes in the rate of hepatic cholesterogenesis. The rate-limiting step for the biosynthetic sequence is the reduction of 3hydroxy-3-methylglutaryl-CoA to mevalonic acid [ 1,2] catalyzed by the enzyme 3- hydroxy- 3methylglutarylCoA reductase (hydroxymethyl-
The liver can alter the rate of cholesterol biosynthesis to accommodate the changing requirements for cholesterol in the whole animal. Accordingly, the variation in the intake of cholesterol from the diet, or in the rate of return of bile acids OOOS-2760/82/0000-0/$02.75
‘C’ 1982 Elsevier
Biomedical
Press
447
glutaryl-CoA reductase; EC 1.1.1.34). It has been accepted for a number of years that the activity of this enzyme can be controlled at the level of protein synthesis 131. Such a mechanism would imply the same relation between the enzyme concentration and the activity in the microsomal fraction from animals in experimental conditions known to be associated with changes in the rate of hepatic cholesterogenesis and hence hydroxymethylglutaryl-CoA reductase activity. However, the use of an antibody to hydroxymethylglutaryl-CoA reductase and the comparison of enzyme activity and enzyme concentration in the microsomal fraction from rats under various experimental conditions has revealed anomalies (4-61 that have challenged the concept of regulation at the levet of enzyme synthesis. Mechanisms involving control at the level of existing enzyme have been proposed to explain the in vivo changes in the activity of hydroxymethyl~uta~l-CoA reductase through a direct modulation, One of these mechanisms entails covalent modification of the enzyme by a cycle of phosphorylation and dephosphorylation associated with the modulation of activity [7-91. Evidence for such mechanism was obtained mainly in vitro. Another mechanism that attempts to explain the physiological changes in the activity of hydroxymethylglutaryl-CoA reductase through direct modulation proposes [ lO,l 1] that the activity can be influenced by the size of a pool of non-esterified cholesterol in the endoplasmic reticular membrane in the immediate environment of hydroxymethylglutarylCoA reductase. This proposal has been based on the changes in the interaction of the enzyme with its lipid environment in microsomal fractions from rats in various experimental conditions as evidenced from changes in the temperature-induced kinetics of the enzyme. We have demonstrated recently that the specific transfer of non-esterified cholesterol to vesicles of the microsomal fraction results in a decrease in the activity of hydroxymethylglutaryl-CoA reductase [ 121. To assess whether the transfer of non-esterified cholesterol to microsomal vesicles and the decrease in the activity of hydroxymethylglutaryl-CoA reductase that follows is also associated with changes in the interaction of the enzyme with its lipid environment, we have compared the activity and
the temperature-induced kinetics before and after the transfer of non-esterified cholesterol to the microsomal vesicles. Moreover, in order to see whether the in vitro modulation by non-esterified cholesterol and that by the phosphorylation/dephosphorylation cycle share a common mechanism, the Arrhenius plots of the enzyme after inactivation (phosphorylation) or reactivation (dephosphorylation) were compared with such plots of the enzyme in the microsomal fraction from rats fed on cholesterol- or cholestyramine-supplemented and on the standard diet. Materials and Methods The sources of radioactively labelled compounds and other materials used in the present work have been reported elsewhere [ 11,131. The preparations of serum and the lipoprotein-deficient serum were obtained as described previously [ 121. dousing and feeding of rats. Male Wistar rats weighing 180-220 g were used for all experiments. The rats were kept under conditions of controlled lighting and feeding as described previously [14]. The composition of the standard diet 1141and the preparation of the diets supplemented with cholesterol (2%) or with cholestyramine (5%) has been reported elsewhere [ 14,151. Cholestyraminefed rats were given cholestyramine-containing pellets for 4 days before the experiment. Cholesterolfed rats were presented with cholesterol-containing pellets for 12 h before the experiment. All rats were killed between 0600-0800 h on the day of the experiment and the livers were removed at once. Preparation of the microsomal fraction and the treated micros#ma~ preparations. The livers were chilled on ice and perfused with ice-cold 0.25 M sucrose to remove contaminating blood. The pooled livers from the rats in a group were homogenized and the microsomal pellet was obtained as described earlier ( 151.The microsomal fraction was prepared by suspending the microsomal pellet either in 50 mM imidazole/HCl buffer, pH 7.4, containing 250 mM NaCl and 10 mM dithiothreito1 (buffer I) or in 0.25 M sucrose/3 mM imidazole buffer, pH 7.4 (buffer II), with eight strokes in a manually operated Potter-Elvehjem homogenizer. The microsomal fraction in buffer II was incubated with the serum preparation, the lipopro-
448
tein-deficient serum or with the suspending buffer (buffer II) for 1 h at 37°C. The microsomal fraction in buffer I was treated with the serum preparation or with the suspending buffer (buffer I) under identical conditions. All such incubations contained the same amount of microsomal protein (about 3 mg of protein per ml of the mixture) and at the end of the incubation period the mixture was cooled in ice. The incubation mixture was centrifuged at 104000 X g,, for 60 min, the supernatant was removed and the recovered pellet was suspended in the appropriate buffer to the volume of the original microsomal fraction used. Portions of the microsomal fraction in buffer I prepared from rats fed on standard or cholesteroland cholestyraminesupplemented diets were incubated for 20 min. at 37’C, in the presence of 2 mM ATP and 4 mM MgCl, for the inactivation or no ATP plus Mg *+ for the control as described earlier [15]. The reactivation of the inactivated enzyme in the presence of the cytosolic phosphoprotein phosphatase has also been described
1151. Assay
of hydroxymethylglutaryl-CoA
and of the composition and the supernatant
of microsomal
reductase preparations
of treatment. The activity of hydroxymethylglutaryl-CoA reductase was assayed in portions of serum-treated, buffer-treated, lipoprotein-deficient serum-treated, supernatant fractions obtained from these treatments, and the original microsomal fraction in the presence of DL-hydroxymethyl-[3-‘4C]glutaryl-CoA (6 Ci/mol) in a total incubation volume of 0.5 ml. The incubation mixture, incubation conditions and determination of product formation was as described earlier [ 10,161. The assay of the enzyme in portions of inactivated, reactivated and control preparations was as described earlier [ 151. Cholesteryl esters and non-esterified cholesterol [ 171, total phospholipids [ 181 and protein [ 191 in the serum preparation; lipoprotein-deficient serum, microsomal fraction, treated microsomal preparations and supernatant fractions obtained from such treatment were determined by methods described previously.
Results Effects of transfer of non-esterified liver microsomal ture-induced reductase.
cholesterol to
vesicles on activity and tempera-
kinetics of hydroxymethylglutaryi-CoA
The incubation of rat liver microsomal fraction in the presence of a preparation of serum followed by re-isolation of the microsomal vesicles resulted in a considerable increase in the content of non-esterified cholesterol and in a large decrease in the activity of hydroxymethylglutarylCoA reductase (Table I). Treatment of the microsomal fraction under similar conditions with buffer (buffer II) resulted in the loss of some phospholipid, non-esterified cholesterol and considerable loss of protein. All these components were recovered in the supernatant fraction obtained on re-isolation of the buffer-treated membranes. Another control preparation involved incubation, under similar conditions, of the microsomal fraction in the presence of a preparation of lipoprotein-deficient serum and re-isolation of the treated membranes. Total activity of hydroxymethylglutaryl-CoA reductase recovered in buffer-treated and lipoprotein-deficient serum-treated preparations was similar and was lower than that in the original microsomal fraction. However, total and specific activity assayed in serum-treated microsomal preparations were only a small fraction of those found in the other treated or non-treated microsomes (Table I). That there was a specific transfer of non-esterified cholesterol from the serum preparation to microsomal vesicles is evident from the ratio of nonesterified cholesterol to cholesteryl esters which decreased considerably in the supernatant fraction obtained on re-isolation of the treated microsomal pellet (0.329 * 0.022) as compared with this ratio in the original serum preparation (0.446). There was also a significant decrease in the ratio of non-esterified cholesterol to phospholipid present in this supernatant as compared with the ratio in the serum preparation. The characteristics of the Arrhenius plots of hydroxymethylglutaryl-CoA reductase in serumtreated, lipoprotein-deficient serum-treated and buffer-treated preparations were different to those of the enzyme in the original microsomal fraction (Fig. 1 and Table I). The plots of the enzyme in the original microsomal fraction, in buffer-treated and
449
TABLE
I
EFFECT OF INCUBATION OF RAT LIVER MICROSOMAL FRACTION IN THE PRESENCE OF SERUM, LIPOPROTEINDEFICIENT SERUM OR BUFFER ON THE COMPOSITION OF THE MICROSOMAL PREPARATION AND ON ACTIVITY AND TEMPERATURE-INDUCED KINETICS OF HYDROXYMETHYLGLUTARYL-CoA REDUCTASE Four groups of rats each containing three animals were killed. The microsomal fraction was prepared from pooled livers of each group and portions incubated for 1 h at 37°C (4.2 mg microsomal protein/ml mixture) with buffer, serum (protein, 20.5 mg; non-esterified cholesterol, 346 nmol; cholesterol esters, 776 nmol; and total phospholipid, 0.530 mg per ml incubation mixture) or lipoprotein-deficient serum (protein, 23 mg; non-esterified cholesterol, 5.7 nmol; cholesterol ester, 37.3 nmol; and total phospholipid, 0.078 mg per ml incubation mixture). At the end of this period all preparations were centrifuged and microsomal pellets obtained were resuspended to the volume of the original microsomal fraction. The original microsomal fractions contained (average *SD.): protein, 7.1 kO.1 mg/ml; non-esterified cholesterol, 524% 18.9 nmol/ml; cholesteryl esters, 20.2* 1.5 nmol/ml; total phospholipid. 2.84*0.07 mg/ml and had activity of hydroxymethylglutaryl-CoA reductase, 4472 197 pmol/min per ml. Values given for treated microsomal preparations are means *SD. obtained from comparison of treated preparations with the corresponding non-treated microsomal fraction. Arrhenius plots of the enzyme in microsomal fraction and various microsomal preparations were obtained on assay in the presence of 90 pM DL-hydroxymethyl-[3-‘4C]glutaryl-CoA (6 Ci/mol) at various temperatures and the lines are the best fit for the points determined. Activation energies were calculated using the slope of the lines from such plots: values given are means 5 S.D. for the four microsomal fractions or the various treated preparations. Treatment
with:
Percentage Protein
None Buffer Serum Lipoproteindeficient serum
of content
or activity
Total phospholipid
Hydroxymethylglutaryl-CoA
reductase
Nonesterified cholesterol
Activity
Break
Activation
(%)
(“C)
energy (kcal/mol)
Above break
Below break 45.71-6.2 43.2* 1.9
ldo 64.9’1.1 81.423.4
100 81.4k3.2 76.3k5.2
100 90.1*2.1 122.022.3
loo 63.3 1-2.6 6.450.5
28.8-‘0.3 28.8kO.9 -
21.0% 1.1 25.4-cO.3
92.9k3.4
78.025.5
98.1k5.5
7O.lk3.1
29.0*0.2
31.3-cO.6
43.5 k2.3 57.8k4.5
lipoprotein-deficient serum-treated preparations showed breaks at about 29°C whereas the enzyme in serum-treated preparations exhibited a constant activation energy between 37 and 22°C. The enzyme in the original microsomal fractions had activation energies lower above and higher below the corresponding values in buffer-treated preparations. The constant activation energy of enzyme in serum-treated preparations was higher than that above and lower than that below the break calculated for the enzyme in lipoproteindeficient serum-treated preparations.
31
3: 32.5
3; 33.0 fx1oL
28
25 33.5
22 ZO
2ooc OK
Fig. I. Arrhenius plots of hydroxymethylglutaryl-CoA reductase activity in the liver microsomal fraction (0). buffertreated microsomal preparation (A). lipoprotein-deficient serum-treated microsomal preparation (0) and serum-treated microsomal preparation (a). Arrhenius plots were obtained as in the legend to Table I and activation described energies (kcal/mol) are shown beside each line. Other experimental details are given in the text.
450
relationship between non-esterif~ed choiesteroi concentration and activity of hydroxymethyigluta~lCoA reductase. To see the effect of varying transfer of non-esterified cholesterol on the interaction of hydroxymethylglutaryl-CoA with its lipid environment in the membrane, the microsomal fraction was incubated in the presence of various concentrations of the serum preparation and a supplement of lipoprotein-deficient serum. Under these conditions, the concentration of serum nonesterified cholesterol in the incubation mixture varied considerably whereas the serum protein and microsomal protein remained constant (Table II). The progressive increase in concentration of serum non-esterified cholesterol in the incubation mixture resulted in a progressive increase in concentration of non-esterified cholesterol in reisolated microsomal preparations and in a progressive decrease in activity of hydroxymethylglutarylCoA reductase together with considerable changes
in the characteristics of the Arrhenius plots of the enzyme. As may be seen in Table II, in treated preparations there is a progressive increase in activation energy above and a decrease below the break with increasing concentration of nonesterified cholesterol. The two preparations having the highest concentration of cholesterol show a constant activation energy between 37 and 21°C. A plot of the activation energy below the break versus the concentration of cholesterol gave a negative correlation (correlation coefficient -O.~), whereas a similar plot of the activation energy above the break (this plot excludes the highest concentration of cholesterol) gave a linear correlation (correlation coefficient 0.950). In agreement with previous observations [12] there was also a significant negative correlation between the concentration of non-esterified cholesterol and the logarithm of activity of hydroxymethylglutaryl -CoA reductase (correlation coefficient -0.965).
TABLE II EFFECT OF INCUBATION OF LIVER MICROSOMAL FRACTION WITH VARIOUS DILUTIONS OF SERUM SUPPLEMENTED WITH LIPOPROTEIN-DEFICIENT SERUM ON THE NON-ESTERIFIED CHOLESTEROL CONCENT~TION AND ON ACTIVITY AND TEMPERATURE-INDUCED KINETICS OF HYDROXYMETHYLGLUTARYL-CoA REDUCTASE The microsomal fraction was prepared from pooled livers of a group of rats and portions were incubated for 1h at 37’C (3.07 mg microsomal protein/ml mixture) with various mixtures of serum (protein, 78.6 mg; non-esterified cholesterol, 1062.4 nmol; cholesteryl esters, 1587 nmol; and total phospholipid, 1.344 mg/ml) and lipoprotein-deficient serum (protein, 75.7 mg; non-esterified cholesterol, 16.3 nmol; cholesteql esters 35.3 nmol; and total phospholipid 0.143 mg,/ml) in a total volume of 12 ml to give the serum concentration shown and the same concentration of serum protein (23.5 mg/ml incubation mixture). At the end of this period all preparations were centrifuged and the microsomal pellets obtained resuspended to the volume of the original microsomal fraction. Serum non-esterified cholesterol in incubation (nmol/mg microsomal protein)
Non-esterified cholesterol (nmol/mg protein)
None (original microsomal fractions) 53.0 62.1 1.6 63.9 11.6 66.9 15.7 70.0 21.4 73. I 35.6 79.3. 52.5 81.7 69.4 94.3 103.0
Hydroxymethylglutaryl-CoA Activity (pmol/min per) mg protein.
167.0 87.5 83.1 78,3 75.8 58.6 40.4 22.9 6.7
reductase
Temperature-induced Break SC)
28.9 28.8 29.1 29.7 31.2 31.4 32.5
kinetics a
Activation energy (kcal/mol) Above break
Below break
13.3 34.7 37.4 40.4 41.3 41.9 45.8
51.1 75.9 13.6 68.2 58.6 65.5 66.3 53.1 45.2
a Arrhenius plots of the enzyme in the original microsomal fraetion and various treated preparations were obtained as described in Table I.
451
Transfer of non-esterified cholesterol and modulation of hydroxymethylglutaryl-CoA reductase in liver microsomal vesicles from rats fed cholesterol- or cholestyramine-supplemented diet. Feeding rats for 12 h on a 2% cholesterol-supplemented diet or for 4 days on 5% cholestyramine-supplemented diet resulted in a corresponding decrease and increase in activity of hydroxymethylglutaryl-CoA reductase as compared with that in liver microsomal
Standard
det
amine dvzt
Chole! Lte
Fig. 2. Specific activity of hydroxymethylglutaryl-CoA reductase in liver microsomal fraction from rats fed cholesterolor cholestyramine-supplemented or standard diet and in various preparations obtained on treatment of the microsomal fraction. Three rats in each group were killed. The pooled livers from each group were homogenized and the homogenate was divided into two equal parts to prepare the microsomal pellets. These were suspended in either buffer I or buffer II. Portions of the microsomal fraction were incubated for I h at 37°C (4-5 mg microsomal protein/ml mixture) with the same buffer, serum (protein, 21.R mg: non-esterified cholesterol, 379 nmol; cholesteryl ester, 816 nmol: phospholipid, 0.593 mg per ml incubation mixture) or lipoprotein-deficient serum (protein, 15.0 mg: non-esterified cholesterol, 4.8 nmol: cholesteryl ester, 13.6 nmol; phospholipid, 0.04 mg per ml of incubation mixture). At the end of this period all preparations were centrifuged and the treated microsomal pellets were resuspended in appropriate buffer to the volume of the original microsomal fraction. Portions of treated microsomal preparations and microsomal fractions from the three groups of rats were assayed for activity. The open bars show activity in microsomal preparations suspended in buffer II and the shaded bars show activity in preparations suspended in buffer I.
fraction from rats fed standard diet. The effects of these dietary manipulations on activity of the enzyme were similar whether the microsomal pellet obtained from the same liver homogenate was suspended in buffer I or buffer II (Fig. 2). However, activity of the enzyme in the microsomal fraction in buffer I was in all cases 2-3-fold higher than that in the microsomal fraction suspended in buffer II. The microsomal fractions in buffer II from rats in the three experimental conditions were incubated with the same buffer, serum or lipoprotein-deficient serum under conditons similar to those described in Table I. The specific activity of hydroxymethylglutaryl-CoA reductase in the various treated microsomal preparations is shown in Fig. 2. Activity of the enzyme was in all cases considerably lower in serum-treated than in other treated preparations or the original microsomal fraction. Thus, activity in serum-treated preparations from rats fed standard, cholesteroland cholestyramine - supplemented diet was correspondingly 11.4, 15 and 11.2% of the activity in the respective lipoprotein-deficient serum-treated microsomal preparations. In all three experimental conditions, concentration of non - esterified cholesterol was higher in serum-treated than in the other preparations or the original microsomal fractions. The microsomal fraction obtained by suspending microsomal pellets from rats in the three experimental conditions in buffer I were also treated under similar conditions with the same buffer or with serum. In all cases, specific activity of hydroxymethylglutaryl-CoA reductase was higher in buffer-treated preparation than in the corresponding original microsomal fraction (Fig. 2). Moreover, in all cases the serum-treated preparation obtained from microsomes suspended in buffer I showed activity higher than that in the original microsomal fraction. Modulation of enzyme activity by non-esterified cholesterol in in vivo and in vitro covalent modification. For experiments involving in vitro covalent modification of hydroxymethylglutaryl-CoA reductase (phosphorylation and dephosphorylation), the microsomal pellet obtained from liver homogenates from rats fed standard, cholesterolor cholestyramine-supplemented diet were suspended
37
34
30
25
33
20 oc 34 OK
$X104
37
34
IO
30
2F
22
33
T 3‘ "K
4x10
15
37 '
u
30 33
25
22
oc 34 OK
fX104
Fig. 3. ductase
Arrhenius plots of hydroxymethylglutaryl-CoA reactivity in liver microsomal preparations from rats fed
in buffer I as described earlier [ 151. These microsomal fractions were incubated with the same buffer for 20 min. at 37’C (control preparations) or under the same conditions in the presence of 2 mM ATP plus 4mM MgClz (inactivated enzyme). The Arrhenius plots of the enzyme in control preparations from rats in the three experimental conditions have different characteristics (Fig. 3) but are similar to the corresponding plots of the enzyme in microsomal fraction suspended in buffer II (see Ref. [lo]). Thus, the enzyme in control preparation from rats fed cholesterol-supplemented diet shows a constant activation energy between 37 and 22°C whereas Arrhenius plots of the enzyme from rats fed cholestyramine-supplemented or standard diet show a break in the activation energy at about 29°C (Fig. 3). Incubation of microsomal fraction from rats in the three experimental conditions in the presence of ATP plus Mg2+ resulted in all cases in a considerable decrease in activity of hydroxymethylglutaryl-CoA reductase. However, the Arrhenius plot of the enzyme in the inactivated preparation showed in every case similar characteristics to the plot of the corresponding control preparation (Fig. 3). Further incubation of inactivated preparations in the presence of cytosolic phosphoprotein phosphatase [ 151 resulted, in all cases, in reactivation of the enzyme to the levels of the correspond-
cholesterolor cholestyramine-supplemented or standard diet. The microsomal pellets were suspended in buffer I and portions of microsomal fractions from the three experimental conditions (10.7 mg microsomal protein) were incubated at 37°C for 20 min in a total volume of 2 ml containing buffer I (control) or buffer I plus 2 mM ATP and 4 mM MgClz(inactivated). At the end of this period 2 ml of 50 mM EDTA and 2 ml of 150 mM NaF containing 25% glycerol were added to each flask and 0.2 ml portions were removed to assay the activity of the enzyme in the presence of 180 PM DLhydroxymethyl-[3-‘4C]glutaty-CoA (6 Ci/mol) at various temperatures Each line was obtained from the best fit for the 5-7 points determined and the correlation coefficient was in all cases better than -0.990. Activation energies (kcal/mol) are as indicated beside each line and were calculated from the slope of the line. A. Microsomal preparations from rats fed standard diet; B. microsomal preparations from rats fed cholesterolsupplemented diet: C. microsomal preparations from rats fed cholestyramine-supplemented diet.
453
ing control preparation without any change in the characteristics of the Arrhenius plot of reactivated enzyme (results not shown). Discussion
Incubation of rat liver microsomal fraction with a preparation of serum results in transfer of nonesterified cholesterol and a decrease in activity of hydroxymethylglutaryl - CoA reductase in re isolated microsomal vesicles. We have shown previously [ 121 that, under the present conditions, transfer of non-este~fied cholesterol and changes in activity of hydroxymethylglutaryl-CoA reductase are incubation time-dependent and both are dependent on concentration of serum in the incubation mixture. In complete agreement, the present results show an inverse relationship between activity of the enzyme and concentration of non-esterified cholesterol in serum-treated microsomal preparations. Similarly, incubation of microsomal fraction from rats fed cholesterol- or cholestyramine-supplemented diet with serum preparation results in low activity of the enzyme as compared with the corresponding original microsomal fraction, buffer- treated or lipoproteindeficient serum-treated microsomal preparations. Transfer of non-esterified cholesterol to microsomal vesicles and changes in activity of hydroxymethylglutaryl-CoA reductase are also associated with changes in interaction of the enzyme with its lipid environment in the membrane. This is demonstrated in the changes in temperature-induced kinetic properties of the enzyme in serum-treated microsomal preparations as compared with those for the enzyme in lipoprotein-deficient serumtreated preparations. An increase in concentration of non-esterified cholesterol in the membrane in the environment of hydroxymethylglutaryl-CoA reductase is expected to result in an increase in activation energy above the break and a decrease below the break. Assuming that the lipoproteindeficient serum-treated preparations are a suitable control for serum-treated preparations, the increase in concentration of non-esterified cholesterol in the microsomal fraction is expected to result in a proportional increase in cholesterol concentration in the endoplasmic reticular membrane and in progressive changes in the characteristics of
Arrhenius plots of the enzyme that are observed in Table II. At high concentrations of cholesterol the net result of these changes is a constant activation energy for the range of temperatures employed (Tables I and II). Arrhenius plots with similar characteristics are observed for the enzyme in the microsomal fraction from rats fed the cholesterolsupplemented diet (see also Refs. 10, 20, 21), from rats after intravenous injection of a load of mevalonic acid [ 111 or from rats fed for 12 h with standard diet supplemented either with 10% safflower oil and 2% cholesterol or 10% tristearin and 2% cholesterol [22]. In all these cases the changes in characteristics of Arrhenius plots can be attributed to increased concentration of non-esterified cholesterol in the endoplasmic reticular membrane in the environment of the enzyme which can also be responsible for the decreased activity of hydroxymethylglutaryl-CoA reductase observed in the above experimental conditions. A decrease in concentration of non-esterified cholesterol in endoplasmic reticular membrane in the environment of the enzyme is expected to result in opposite changes in the characteristics of Arrhenius plots, i.e. a decrease in activation energy above and an increase below the break. Such changes are observed in the Arrhenius plot of the enzyme in microsomal fraction from rats fed ChoIestyramine-supplemented diet (see also Ref. lo), from rats following ligation of the bile duct [23] or from rats following intraveous injection of Triton WR-1339 [24]. All these experimental conditions are also associated with an increase in activity of hydroxymethylglutaryl-CoA. For microsomal fraction suspended in buffer II, total activity of hydroxymethylgluta~l-CoA reductase recovered with buffer-treated or lipoprotein-deficient serum-treated preparation was significantly lower than that in the original microsomal fraction. Such changes were not associated with an increase in content of cholesterol in treated microsomal preparation or can be attributed to solubilization and release of enzyme into the supernatant fraction obtained on re-isolation of the microsomal membranes. However, the observed increase in activation energy above the break and the decrease below the break in Arrhenius plots of the enzyme in buffer-treated preparations, as compared with corresponding values in the original
454
microsomal fraction, suggests a higher concentration of non-esterified cholesterol in the environment of the enzyme in this preparation. It is possible that redistribution of cholesterol during this in vitro treatment between microsomal vesicles results in modulation of activity. These effects were not observed when the microsomal fraction in buffer1 was treated with the same buffer or lipoprotein-deficient serum. Moreover, treatment of this microsomal fraction with the serum preparation did not result in transfer of non-esterified cholesterol to the endoplasmic reticular membranes. The demonstration that hydroxymethylglutarylCoA reductase can exist in vitro in active and inactive forms obtained through a cycle of phosphorylation and dephosphorylation of the enzyme suggests a mechanism for rapid short-term regulation in vivo [25]. We have shown previously that feeding rats on a diet supplemented with cholesterol or with cholestyramine results in changes in the isothermic kinetics of hydroxymethylglutaryl-CoA reductase in the microsomal fraction subsequently prepared from liver homogenates and suspended either in buffer I [15] or buffer II [lo]. Moreover, changes in the isothermic kinetics of the enzyme are produced on treatment with ATP plus Mg*+ and are reversed on further incubation of inactivated enzyme with a cytosolic phosphoprotein phosphatase [ 151. The present results demonstrate that changes in activity of the enzyme brought about in vitro by phosphorylation and dephosphorylation are not associated with changes in interaction of the enzyme with its lipid environment. Thus, the Arrhenius plots of inactivated or reactivated enzyme in microsomal preparations from rats in the three experimental conditions show characteristics very similar to those of the corresponding control preparations. The possibility that only the active molecules of hydroxymethylglutarylCoA reductase (non phosphorylated) interact with the membrane cannot be excluded. The present results, however, define the difference between the mechanism of in vitro modulation by non-esterified cholesterol and modulation by covalent modification. Moreover, since the considerable differences in activity of the enzyme in microsomal fraction from rats fed on - sup cholesterol - or cholestyramine standard,
plemented diets are associated with changes in the interaction of the enzyme with the membrane, changes in concentration of non - esterified cholesterol in the endoplasmic reticular membrane are responsible for the differences in activity. Recent work from a number of laboratories [ 15,261 is consistent with the conclusion that the phosphorylation/dephosphorylation cycle is not responsible for the differences in activity of hydroxymethylglutaryl-CoA reductase in microsomal fraction from rats; at least in the dietary conditions studied in the present paper. The changes in the concentration of the enzyme that are associated with these conditions [6,27] presumably follow the initial direct modulation by non-esterified cholesterol and are either due to changes in the rate of association of newly synthesized enzyme to a membrane with altered composition [ 11,241 or to a change in the rate of enzyme synthesis controlled at the pre-translational level [2,3]. References I Dietschy. J.M. and Brown, M.S. (1974) J. Lipid Res. 15, 508-516 2 Rodwell. V.W.. McNamara. D.J. and Shapiro, D.J. (1973) Adv. Enzymol. 38, 373-412 3 Rodwell, V.W.. Nordstrom, J.L. and Mitschelen, J.J. (1976) Adv. Lipid Res. 14, l-74 4 Higgins, M.J.P., Brady, D. and Rudney, H. ( 1974) Arch. B&hem. Biophys. 163, 271-282 5 Higgins. M. and Rudney. H. (1973) Nature New Biol. 246. 60-6 I 6 Edwards, P.A., Lemongello. D., Kanne, J., Schechter. I. and Fogelman, A.M. (1980) J. Biol. Chem. 255, 3715-3725 7 Beg. Z.H., Stonik, J.A. and Brewer, H.B.Jr. (1978) Proc. Natl. Acad. Sci. U.S.A. 75. 3678-3682 8 Ingebritsen, T.S.. Lee, H.-S., Parker, R.A. and Gibson, M.D. (1978) Biochem. Biophys. Res. Commun. 81. l268I277 9 Nordstrom, J.L., Rodwell, V.W. and Mitschelen. J.J. ( 1977) J. Biol. Chem. 252, 8924-8934 IO Mitropoulos, K.A. and Venkatesan, S. (1977) B&him. Biophys. Acta 489, 126- 142 II Mitropoulos. K.A., Balasubramaniam. S., Venkatesan, S. and Reeves, B.E.A. (1978) Biochim. Biophys. Acta 530. 99-111 12 Mitropoulos, K.A., Venkatesan, S., Reeves, B.E.A. and Balasubramaniam, S. (1981) B&hem. J. 194, 265-271 13 Mitropoulos. K.A., Venkatesan, S., Balasubramaniam. S. and Peters, T.J. (1978) Eur. J. B&hem. 82, 419-429 14 Mitropoulos. K.A., Balasubramaniam, S. and Myant. N.B. (1973) Biochim. Biophys. Acta 326,428-438 15 Mitropoulos. K.A., Knight, B.L. and Reeves, B.E.A. ( 1980) B&hem. J. 185, 435-441
455
16 Mitropoulos, K.A. and Balasubramamam, S. (1976) Biothem. J. 160,49-55 17 Balasubramaniam, S., Mitropoulos, K.A. and Venkatesan, S. (1978) Eur. J. Biochem. 90. 377-383 18 Bartlett, G.R. (1959) J. Biol. Chem. 234. 466-468 19 Lowry, O.H.. Rosebrough, N.J.. Fax, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193. 265-275 20 Sabine. J.R. and James. M.J. (1976) Life Sci. 18. I I85- I 192 2 I Sipat. A.B. and Sabine. J.R. (1981) Biochem. J. 194.889-893 22 Mitropoulos. K.A., Venkatesan. S. and Balasubramaniam. S. (1980) Biochim. Biophys. Acta 619. 247-257 23 Mitropoulos, K.A.. Venkatesan, S. and Balasubramaniam,
S. (I 979) in Biological
Effects
of Bile Acids (Paumgartner,
G., Stiehl, A. and Gerok, W., eds.), pp. 137-144, MTP Press, Lancaster 24 Mitropoulos, K.A., Venkatesan, S. and Balasubramaniam, S. (1978) Biochem. Sot. Trans. 6, 878-883 25 Gibson, D.M. 2649-2664
and
Ingebritsen,
T.S. (1978)
Life Sci. 23,
26 Brown M.S., Goldstein, J.L. and Dietschy, J.M. (1979) J. Biol. Chem. 254, 5144-5149 27 Hardgrave, J.E., Heller, R.A., Herrera, M.G. and Scallen, T.J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 3834-3838
Bioc,himrcrr et Bioph.vsrcu Actu.
7 IO
( 19X2)
446-455
Elsevier Biomedical
Press
BBA 51050
3-HYDROXY-3-METHYLGLUTARYL-COENZYME
A REDUCTASE
THE DIFFERENCE IN THE MECHANISM OF THE IN VITRO MODULATION BY PHOSPHORYLATION AND DEPHOSPHORYLATION TO MODULATION OF ENZYME ACTIVITY BY NON-ESTERIFIED CHOLESTEROL S. VENKATESAN Medical
and K.A. MITROPOULOS
Reseurch Council
(Received
Kqv words:
Lipid
Metabolism
Unit, Hummersmith
Hospitul,
Ducune Roud. I.ondon
Wl.!
OHS (U.K.)
July 10th. 1981)
Hydroxymethylglutu~l-CoA
reductrrse; Covalent
modificution;
Cholesterol
trcrnsfer; Cholesterol
s,nthesrs;
(Endoplusnrk
reticulum)
Incubation of rat liver microsomal fraction in the presence of increasing concentrations of a serum preparation and the re-isolation of the treated microsomal vesicles resulted in a progressive increase in the concentration of non-esterified cholesterol, a progressive decrease in the activity of hydroxymethylglutarylCoA reductase and progressive changes in the characteristics of the Arrhenius plots of the enzyme. The changes in the characteristics of the Arrhenius plots of the enzyme in the serum-treated preparations are consistent with a progressive increase in the concentration of non-esterified cholesterol in the environment of the hydroxymethylglutaryl-CoA reductase in endoplasmic reiticulum vesicles. The serum-treated preparations with high non-esterified cholesterol content showed a constant activation energy between 37 and 20°C whereas the enzyme in the non-treated microsomal fraction, the buffer-treated and the lipoprotein-deficient serum-treated preparations showed breaks in the activation energy at about 29°C. The microsomal fraction from rats fed on the standard, cholesterol- or cholestyramine-supplemented diet showed considerable differences in the activity of hydroxymethylglutaryl-CoA reductase and differences in the characteristics of their Arrhenius plots. However, the incubation of the microsomal fraction from the rats in the three experimental conditions with ATP and Mg*+ and the further incubation of the inactivated enzyme with a preparation of cytosolic phosphoprotein phosphatase resulted in Arrhenius plots with similar characteristics to those of the corresponding original microsomal fraction. These results suggest that changes in the concentration of non-esterified cholesterol in the endoplasmic reticular membrane are responsible for the differences in the activity of hydroxymethylghttaryl-CoA reductase in the microsomal fraction from the rats in these dietary conditions.
Introduction
from the enterohepatic circulation, is associated with compensatory changes in the rate of hepatic cholesterogenesis. The rate-limiting step for the biosynthetic sequence is the reduction of 3hydroxy-3-methylglutaryl-CoA to mevalonic acid [ 1,2] catalyzed by the enzyme 3- hydroxy- 3methylglutarylCoA reductase (hydroxymethyl-
The liver can alter the rate of cholesterol biosynthesis to accommodate the changing requirements for cholesterol in the whole animal. Accordingly, the variation in the intake of cholesterol from the diet, or in the rate of return of bile acids OOOS-2760/82/0000-0/$02.75
‘C’ 1982 Elsevier
Biomedical
Press
447
glutaryl-CoA reductase; EC 1.1.1.34). It has been accepted for a number of years that the activity of this enzyme can be controlled at the level of protein synthesis 131. Such a mechanism would imply the same relation between the enzyme concentration and the activity in the microsomal fraction from animals in experimental conditions known to be associated with changes in the rate of hepatic cholesterogenesis and hence hydroxymethylglutaryl-CoA reductase activity. However, the use of an antibody to hydroxymethylglutaryl-CoA reductase and the comparison of enzyme activity and enzyme concentration in the microsomal fraction from rats under various experimental conditions has revealed anomalies (4-61 that have challenged the concept of regulation at the levet of enzyme synthesis. Mechanisms involving control at the level of existing enzyme have been proposed to explain the in vivo changes in the activity of hydroxymethyl~uta~l-CoA reductase through a direct modulation, One of these mechanisms entails covalent modification of the enzyme by a cycle of phosphorylation and dephosphorylation associated with the modulation of activity [7-91. Evidence for such mechanism was obtained mainly in vitro. Another mechanism that attempts to explain the physiological changes in the activity of hydroxymethylglutaryl-CoA reductase through direct modulation proposes [ lO,l 1] that the activity can be influenced by the size of a pool of non-esterified cholesterol in the endoplasmic reticular membrane in the immediate environment of hydroxymethylglutarylCoA reductase. This proposal has been based on the changes in the interaction of the enzyme with its lipid environment in microsomal fractions from rats in various experimental conditions as evidenced from changes in the temperature-induced kinetics of the enzyme. We have demonstrated recently that the specific transfer of non-esterified cholesterol to vesicles of the microsomal fraction results in a decrease in the activity of hydroxymethylglutaryl-CoA reductase [ 121. To assess whether the transfer of non-esterified cholesterol to microsomal vesicles and the decrease in the activity of hydroxymethylglutaryl-CoA reductase that follows is also associated with changes in the interaction of the enzyme with its lipid environment, we have compared the activity and
the temperature-induced kinetics before and after the transfer of non-esterified cholesterol to the microsomal vesicles. Moreover, in order to see whether the in vitro modulation by non-esterified cholesterol and that by the phosphorylation/dephosphorylation cycle share a common mechanism, the Arrhenius plots of the enzyme after inactivation (phosphorylation) or reactivation (dephosphorylation) were compared with such plots of the enzyme in the microsomal fraction from rats fed on cholesterol- or cholestyramine-supplemented and on the standard diet. Materials and Methods The sources of radioactively labelled compounds and other materials used in the present work have been reported elsewhere [ 11,131. The preparations of serum and the lipoprotein-deficient serum were obtained as described previously [ 121. dousing and feeding of rats. Male Wistar rats weighing 180-220 g were used for all experiments. The rats were kept under conditions of controlled lighting and feeding as described previously [14]. The composition of the standard diet 1141and the preparation of the diets supplemented with cholesterol (2%) or with cholestyramine (5%) has been reported elsewhere [ 14,151. Cholestyraminefed rats were given cholestyramine-containing pellets for 4 days before the experiment. Cholesterolfed rats were presented with cholesterol-containing pellets for 12 h before the experiment. All rats were killed between 0600-0800 h on the day of the experiment and the livers were removed at once. Preparation of the microsomal fraction and the treated micros#ma~ preparations. The livers were chilled on ice and perfused with ice-cold 0.25 M sucrose to remove contaminating blood. The pooled livers from the rats in a group were homogenized and the microsomal pellet was obtained as described earlier ( 151.The microsomal fraction was prepared by suspending the microsomal pellet either in 50 mM imidazole/HCl buffer, pH 7.4, containing 250 mM NaCl and 10 mM dithiothreito1 (buffer I) or in 0.25 M sucrose/3 mM imidazole buffer, pH 7.4 (buffer II), with eight strokes in a manually operated Potter-Elvehjem homogenizer. The microsomal fraction in buffer II was incubated with the serum preparation, the lipopro-
448
tein-deficient serum or with the suspending buffer (buffer II) for 1 h at 37°C. The microsomal fraction in buffer I was treated with the serum preparation or with the suspending buffer (buffer I) under identical conditions. All such incubations contained the same amount of microsomal protein (about 3 mg of protein per ml of the mixture) and at the end of the incubation period the mixture was cooled in ice. The incubation mixture was centrifuged at 104000 X g,, for 60 min, the supernatant was removed and the recovered pellet was suspended in the appropriate buffer to the volume of the original microsomal fraction used. Portions of the microsomal fraction in buffer I prepared from rats fed on standard or cholesteroland cholestyraminesupplemented diets were incubated for 20 min. at 37’C, in the presence of 2 mM ATP and 4 mM MgCl, for the inactivation or no ATP plus Mg *+ for the control as described earlier [15]. The reactivation of the inactivated enzyme in the presence of the cytosolic phosphoprotein phosphatase has also been described
1151. Assay
of hydroxymethylglutaryl-CoA
and of the composition and the supernatant
of microsomal
reductase preparations
of treatment. The activity of hydroxymethylglutaryl-CoA reductase was assayed in portions of serum-treated, buffer-treated, lipoprotein-deficient serum-treated, supernatant fractions obtained from these treatments, and the original microsomal fraction in the presence of DL-hydroxymethyl-[3-‘4C]glutaryl-CoA (6 Ci/mol) in a total incubation volume of 0.5 ml. The incubation mixture, incubation conditions and determination of product formation was as described earlier [ 10,161. The assay of the enzyme in portions of inactivated, reactivated and control preparations was as described earlier [ 151. Cholesteryl esters and non-esterified cholesterol [ 171, total phospholipids [ 181 and protein [ 191 in the serum preparation; lipoprotein-deficient serum, microsomal fraction, treated microsomal preparations and supernatant fractions obtained from such treatment were determined by methods described previously.
Results Effects of transfer of non-esterified liver microsomal ture-induced reductase.
cholesterol to
vesicles on activity and tempera-
kinetics of hydroxymethylglutaryi-CoA
The incubation of rat liver microsomal fraction in the presence of a preparation of serum followed by re-isolation of the microsomal vesicles resulted in a considerable increase in the content of non-esterified cholesterol and in a large decrease in the activity of hydroxymethylglutarylCoA reductase (Table I). Treatment of the microsomal fraction under similar conditions with buffer (buffer II) resulted in the loss of some phospholipid, non-esterified cholesterol and considerable loss of protein. All these components were recovered in the supernatant fraction obtained on re-isolation of the buffer-treated membranes. Another control preparation involved incubation, under similar conditions, of the microsomal fraction in the presence of a preparation of lipoprotein-deficient serum and re-isolation of the treated membranes. Total activity of hydroxymethylglutaryl-CoA reductase recovered in buffer-treated and lipoprotein-deficient serum-treated preparations was similar and was lower than that in the original microsomal fraction. However, total and specific activity assayed in serum-treated microsomal preparations were only a small fraction of those found in the other treated or non-treated microsomes (Table I). That there was a specific transfer of non-esterified cholesterol from the serum preparation to microsomal vesicles is evident from the ratio of nonesterified cholesterol to cholesteryl esters which decreased considerably in the supernatant fraction obtained on re-isolation of the treated microsomal pellet (0.329 * 0.022) as compared with this ratio in the original serum preparation (0.446). There was also a significant decrease in the ratio of non-esterified cholesterol to phospholipid present in this supernatant as compared with the ratio in the serum preparation. The characteristics of the Arrhenius plots of hydroxymethylglutaryl-CoA reductase in serumtreated, lipoprotein-deficient serum-treated and buffer-treated preparations were different to those of the enzyme in the original microsomal fraction (Fig. 1 and Table I). The plots of the enzyme in the original microsomal fraction, in buffer-treated and
449
TABLE
I
EFFECT OF INCUBATION OF RAT LIVER MICROSOMAL FRACTION IN THE PRESENCE OF SERUM, LIPOPROTEINDEFICIENT SERUM OR BUFFER ON THE COMPOSITION OF THE MICROSOMAL PREPARATION AND ON ACTIVITY AND TEMPERATURE-INDUCED KINETICS OF HYDROXYMETHYLGLUTARYL-CoA REDUCTASE Four groups of rats each containing three animals were killed. The microsomal fraction was prepared from pooled livers of each group and portions incubated for 1 h at 37°C (4.2 mg microsomal protein/ml mixture) with buffer, serum (protein, 20.5 mg; non-esterified cholesterol, 346 nmol; cholesterol esters, 776 nmol; and total phospholipid, 0.530 mg per ml incubation mixture) or lipoprotein-deficient serum (protein, 23 mg; non-esterified cholesterol, 5.7 nmol; cholesterol ester, 37.3 nmol; and total phospholipid, 0.078 mg per ml incubation mixture). At the end of this period all preparations were centrifuged and microsomal pellets obtained were resuspended to the volume of the original microsomal fraction. The original microsomal fractions contained (average *SD.): protein, 7.1 kO.1 mg/ml; non-esterified cholesterol, 524% 18.9 nmol/ml; cholesteryl esters, 20.2* 1.5 nmol/ml; total phospholipid. 2.84*0.07 mg/ml and had activity of hydroxymethylglutaryl-CoA reductase, 4472 197 pmol/min per ml. Values given for treated microsomal preparations are means *SD. obtained from comparison of treated preparations with the corresponding non-treated microsomal fraction. Arrhenius plots of the enzyme in microsomal fraction and various microsomal preparations were obtained on assay in the presence of 90 pM DL-hydroxymethyl-[3-‘4C]glutaryl-CoA (6 Ci/mol) at various temperatures and the lines are the best fit for the points determined. Activation energies were calculated using the slope of the lines from such plots: values given are means 5 S.D. for the four microsomal fractions or the various treated preparations. Treatment
with:
Percentage Protein
None Buffer Serum Lipoproteindeficient serum
of content
or activity
Total phospholipid
Hydroxymethylglutaryl-CoA
reductase
Nonesterified cholesterol
Activity
Break
Activation
(%)
(“C)
energy (kcal/mol)
Above break
Below break 45.71-6.2 43.2* 1.9
ldo 64.9’1.1 81.423.4
100 81.4k3.2 76.3k5.2
100 90.1*2.1 122.022.3
loo 63.3 1-2.6 6.450.5
28.8-‘0.3 28.8kO.9 -
21.0% 1.1 25.4-cO.3
92.9k3.4
78.025.5
98.1k5.5
7O.lk3.1
29.0*0.2
31.3-cO.6
43.5 k2.3 57.8k4.5
lipoprotein-deficient serum-treated preparations showed breaks at about 29°C whereas the enzyme in serum-treated preparations exhibited a constant activation energy between 37 and 22°C. The enzyme in the original microsomal fractions had activation energies lower above and higher below the corresponding values in buffer-treated preparations. The constant activation energy of enzyme in serum-treated preparations was higher than that above and lower than that below the break calculated for the enzyme in lipoproteindeficient serum-treated preparations.
31
3: 32.5
3; 33.0 fx1oL
28
25 33.5
22 ZO
2ooc OK
Fig. I. Arrhenius plots of hydroxymethylglutaryl-CoA reductase activity in the liver microsomal fraction (0). buffertreated microsomal preparation (A). lipoprotein-deficient serum-treated microsomal preparation (0) and serum-treated microsomal preparation (a). Arrhenius plots were obtained as in the legend to Table I and activation described energies (kcal/mol) are shown beside each line. Other experimental details are given in the text.
450
relationship between non-esterif~ed choiesteroi concentration and activity of hydroxymethyigluta~lCoA reductase. To see the effect of varying transfer of non-esterified cholesterol on the interaction of hydroxymethylglutaryl-CoA with its lipid environment in the membrane, the microsomal fraction was incubated in the presence of various concentrations of the serum preparation and a supplement of lipoprotein-deficient serum. Under these conditions, the concentration of serum nonesterified cholesterol in the incubation mixture varied considerably whereas the serum protein and microsomal protein remained constant (Table II). The progressive increase in concentration of serum non-esterified cholesterol in the incubation mixture resulted in a progressive increase in concentration of non-esterified cholesterol in reisolated microsomal preparations and in a progressive decrease in activity of hydroxymethylglutarylCoA reductase together with considerable changes
in the characteristics of the Arrhenius plots of the enzyme. As may be seen in Table II, in treated preparations there is a progressive increase in activation energy above and a decrease below the break with increasing concentration of nonesterified cholesterol. The two preparations having the highest concentration of cholesterol show a constant activation energy between 37 and 21°C. A plot of the activation energy below the break versus the concentration of cholesterol gave a negative correlation (correlation coefficient -O.~), whereas a similar plot of the activation energy above the break (this plot excludes the highest concentration of cholesterol) gave a linear correlation (correlation coefficient 0.950). In agreement with previous observations [12] there was also a significant negative correlation between the concentration of non-esterified cholesterol and the logarithm of activity of hydroxymethylglutaryl -CoA reductase (correlation coefficient -0.965).
TABLE II EFFECT OF INCUBATION OF LIVER MICROSOMAL FRACTION WITH VARIOUS DILUTIONS OF SERUM SUPPLEMENTED WITH LIPOPROTEIN-DEFICIENT SERUM ON THE NON-ESTERIFIED CHOLESTEROL CONCENT~TION AND ON ACTIVITY AND TEMPERATURE-INDUCED KINETICS OF HYDROXYMETHYLGLUTARYL-CoA REDUCTASE The microsomal fraction was prepared from pooled livers of a group of rats and portions were incubated for 1h at 37’C (3.07 mg microsomal protein/ml mixture) with various mixtures of serum (protein, 78.6 mg; non-esterified cholesterol, 1062.4 nmol; cholesteryl esters, 1587 nmol; and total phospholipid, 1.344 mg/ml) and lipoprotein-deficient serum (protein, 75.7 mg; non-esterified cholesterol, 16.3 nmol; cholesteql esters 35.3 nmol; and total phospholipid 0.143 mg,/ml) in a total volume of 12 ml to give the serum concentration shown and the same concentration of serum protein (23.5 mg/ml incubation mixture). At the end of this period all preparations were centrifuged and the microsomal pellets obtained resuspended to the volume of the original microsomal fraction. Serum non-esterified cholesterol in incubation (nmol/mg microsomal protein)
Non-esterified cholesterol (nmol/mg protein)
None (original microsomal fractions) 53.0 62.1 1.6 63.9 11.6 66.9 15.7 70.0 21.4 73. I 35.6 79.3. 52.5 81.7 69.4 94.3 103.0
Hydroxymethylglutaryl-CoA Activity (pmol/min per) mg protein.
167.0 87.5 83.1 78,3 75.8 58.6 40.4 22.9 6.7
reductase
Temperature-induced Break SC)
28.9 28.8 29.1 29.7 31.2 31.4 32.5
kinetics a
Activation energy (kcal/mol) Above break
Below break
13.3 34.7 37.4 40.4 41.3 41.9 45.8
51.1 75.9 13.6 68.2 58.6 65.5 66.3 53.1 45.2
a Arrhenius plots of the enzyme in the original microsomal fraetion and various treated preparations were obtained as described in Table I.
451
Transfer of non-esterified cholesterol and modulation of hydroxymethylglutaryl-CoA reductase in liver microsomal vesicles from rats fed cholesterol- or cholestyramine-supplemented diet. Feeding rats for 12 h on a 2% cholesterol-supplemented diet or for 4 days on 5% cholestyramine-supplemented diet resulted in a corresponding decrease and increase in activity of hydroxymethylglutaryl-CoA reductase as compared with that in liver microsomal
Standard
det
amine dvzt
Chole! Lte
Fig. 2. Specific activity of hydroxymethylglutaryl-CoA reductase in liver microsomal fraction from rats fed cholesterolor cholestyramine-supplemented or standard diet and in various preparations obtained on treatment of the microsomal fraction. Three rats in each group were killed. The pooled livers from each group were homogenized and the homogenate was divided into two equal parts to prepare the microsomal pellets. These were suspended in either buffer I or buffer II. Portions of the microsomal fraction were incubated for I h at 37°C (4-5 mg microsomal protein/ml mixture) with the same buffer, serum (protein, 21.R mg: non-esterified cholesterol, 379 nmol; cholesteryl ester, 816 nmol: phospholipid, 0.593 mg per ml incubation mixture) or lipoprotein-deficient serum (protein, 15.0 mg: non-esterified cholesterol, 4.8 nmol: cholesteryl ester, 13.6 nmol; phospholipid, 0.04 mg per ml of incubation mixture). At the end of this period all preparations were centrifuged and the treated microsomal pellets were resuspended in appropriate buffer to the volume of the original microsomal fraction. Portions of treated microsomal preparations and microsomal fractions from the three groups of rats were assayed for activity. The open bars show activity in microsomal preparations suspended in buffer II and the shaded bars show activity in preparations suspended in buffer I.
fraction from rats fed standard diet. The effects of these dietary manipulations on activity of the enzyme were similar whether the microsomal pellet obtained from the same liver homogenate was suspended in buffer I or buffer II (Fig. 2). However, activity of the enzyme in the microsomal fraction in buffer I was in all cases 2-3-fold higher than that in the microsomal fraction suspended in buffer II. The microsomal fractions in buffer II from rats in the three experimental conditions were incubated with the same buffer, serum or lipoprotein-deficient serum under conditons similar to those described in Table I. The specific activity of hydroxymethylglutaryl-CoA reductase in the various treated microsomal preparations is shown in Fig. 2. Activity of the enzyme was in all cases considerably lower in serum-treated than in other treated preparations or the original microsomal fraction. Thus, activity in serum-treated preparations from rats fed standard, cholesteroland cholestyramine - supplemented diet was correspondingly 11.4, 15 and 11.2% of the activity in the respective lipoprotein-deficient serum-treated microsomal preparations. In all three experimental conditions, concentration of non - esterified cholesterol was higher in serum-treated than in the other preparations or the original microsomal fractions. The microsomal fraction obtained by suspending microsomal pellets from rats in the three experimental conditions in buffer I were also treated under similar conditions with the same buffer or with serum. In all cases, specific activity of hydroxymethylglutaryl-CoA reductase was higher in buffer-treated preparation than in the corresponding original microsomal fraction (Fig. 2). Moreover, in all cases the serum-treated preparation obtained from microsomes suspended in buffer I showed activity higher than that in the original microsomal fraction. Modulation of enzyme activity by non-esterified cholesterol in in vivo and in vitro covalent modification. For experiments involving in vitro covalent modification of hydroxymethylglutaryl-CoA reductase (phosphorylation and dephosphorylation), the microsomal pellet obtained from liver homogenates from rats fed standard, cholesterolor cholestyramine-supplemented diet were suspended
37
34
30
25
33
20 oc 34 OK
$X104
37
34
IO
30
2F
22
33
T 3‘ "K
4x10
15
37 '
u
30 33
25
22
oc 34 OK
fX104
Fig. 3. ductase
Arrhenius plots of hydroxymethylglutaryl-CoA reactivity in liver microsomal preparations from rats fed
in buffer I as described earlier [ 151. These microsomal fractions were incubated with the same buffer for 20 min. at 37’C (control preparations) or under the same conditions in the presence of 2 mM ATP plus 4mM MgClz (inactivated enzyme). The Arrhenius plots of the enzyme in control preparations from rats in the three experimental conditions have different characteristics (Fig. 3) but are similar to the corresponding plots of the enzyme in microsomal fraction suspended in buffer II (see Ref. [lo]). Thus, the enzyme in control preparation from rats fed cholesterol-supplemented diet shows a constant activation energy between 37 and 22°C whereas Arrhenius plots of the enzyme from rats fed cholestyramine-supplemented or standard diet show a break in the activation energy at about 29°C (Fig. 3). Incubation of microsomal fraction from rats in the three experimental conditions in the presence of ATP plus Mg2+ resulted in all cases in a considerable decrease in activity of hydroxymethylglutaryl-CoA reductase. However, the Arrhenius plot of the enzyme in the inactivated preparation showed in every case similar characteristics to the plot of the corresponding control preparation (Fig. 3). Further incubation of inactivated preparations in the presence of cytosolic phosphoprotein phosphatase [ 151 resulted, in all cases, in reactivation of the enzyme to the levels of the correspond-
cholesterolor cholestyramine-supplemented or standard diet. The microsomal pellets were suspended in buffer I and portions of microsomal fractions from the three experimental conditions (10.7 mg microsomal protein) were incubated at 37°C for 20 min in a total volume of 2 ml containing buffer I (control) or buffer I plus 2 mM ATP and 4 mM MgClz(inactivated). At the end of this period 2 ml of 50 mM EDTA and 2 ml of 150 mM NaF containing 25% glycerol were added to each flask and 0.2 ml portions were removed to assay the activity of the enzyme in the presence of 180 PM DLhydroxymethyl-[3-‘4C]glutaty-CoA (6 Ci/mol) at various temperatures Each line was obtained from the best fit for the 5-7 points determined and the correlation coefficient was in all cases better than -0.990. Activation energies (kcal/mol) are as indicated beside each line and were calculated from the slope of the line. A. Microsomal preparations from rats fed standard diet; B. microsomal preparations from rats fed cholesterolsupplemented diet: C. microsomal preparations from rats fed cholestyramine-supplemented diet.
453
ing control preparation without any change in the characteristics of the Arrhenius plot of reactivated enzyme (results not shown). Discussion
Incubation of rat liver microsomal fraction with a preparation of serum results in transfer of nonesterified cholesterol and a decrease in activity of hydroxymethylglutaryl - CoA reductase in re isolated microsomal vesicles. We have shown previously [ 121 that, under the present conditions, transfer of non-este~fied cholesterol and changes in activity of hydroxymethylglutaryl-CoA reductase are incubation time-dependent and both are dependent on concentration of serum in the incubation mixture. In complete agreement, the present results show an inverse relationship between activity of the enzyme and concentration of non-esterified cholesterol in serum-treated microsomal preparations. Similarly, incubation of microsomal fraction from rats fed cholesterol- or cholestyramine-supplemented diet with serum preparation results in low activity of the enzyme as compared with the corresponding original microsomal fraction, buffer- treated or lipoproteindeficient serum-treated microsomal preparations. Transfer of non-esterified cholesterol to microsomal vesicles and changes in activity of hydroxymethylglutaryl-CoA reductase are also associated with changes in interaction of the enzyme with its lipid environment in the membrane. This is demonstrated in the changes in temperature-induced kinetic properties of the enzyme in serum-treated microsomal preparations as compared with those for the enzyme in lipoprotein-deficient serumtreated preparations. An increase in concentration of non-esterified cholesterol in the membrane in the environment of hydroxymethylglutaryl-CoA reductase is expected to result in an increase in activation energy above the break and a decrease below the break. Assuming that the lipoproteindeficient serum-treated preparations are a suitable control for serum-treated preparations, the increase in concentration of non-esterified cholesterol in the microsomal fraction is expected to result in a proportional increase in cholesterol concentration in the endoplasmic reticular membrane and in progressive changes in the characteristics of
Arrhenius plots of the enzyme that are observed in Table II. At high concentrations of cholesterol the net result of these changes is a constant activation energy for the range of temperatures employed (Tables I and II). Arrhenius plots with similar characteristics are observed for the enzyme in the microsomal fraction from rats fed the cholesterolsupplemented diet (see also Refs. 10, 20, 21), from rats after intravenous injection of a load of mevalonic acid [ 111 or from rats fed for 12 h with standard diet supplemented either with 10% safflower oil and 2% cholesterol or 10% tristearin and 2% cholesterol [22]. In all these cases the changes in characteristics of Arrhenius plots can be attributed to increased concentration of non-esterified cholesterol in the endoplasmic reticular membrane in the environment of the enzyme which can also be responsible for the decreased activity of hydroxymethylglutaryl-CoA reductase observed in the above experimental conditions. A decrease in concentration of non-esterified cholesterol in endoplasmic reticular membrane in the environment of the enzyme is expected to result in opposite changes in the characteristics of Arrhenius plots, i.e. a decrease in activation energy above and an increase below the break. Such changes are observed in the Arrhenius plot of the enzyme in microsomal fraction from rats fed ChoIestyramine-supplemented diet (see also Ref. lo), from rats following ligation of the bile duct [23] or from rats following intraveous injection of Triton WR-1339 [24]. All these experimental conditions are also associated with an increase in activity of hydroxymethylglutaryl-CoA. For microsomal fraction suspended in buffer II, total activity of hydroxymethylgluta~l-CoA reductase recovered with buffer-treated or lipoprotein-deficient serum-treated preparation was significantly lower than that in the original microsomal fraction. Such changes were not associated with an increase in content of cholesterol in treated microsomal preparation or can be attributed to solubilization and release of enzyme into the supernatant fraction obtained on re-isolation of the microsomal membranes. However, the observed increase in activation energy above the break and the decrease below the break in Arrhenius plots of the enzyme in buffer-treated preparations, as compared with corresponding values in the original
454
microsomal fraction, suggests a higher concentration of non-esterified cholesterol in the environment of the enzyme in this preparation. It is possible that redistribution of cholesterol during this in vitro treatment between microsomal vesicles results in modulation of activity. These effects were not observed when the microsomal fraction in buffer1 was treated with the same buffer or lipoprotein-deficient serum. Moreover, treatment of this microsomal fraction with the serum preparation did not result in transfer of non-esterified cholesterol to the endoplasmic reticular membranes. The demonstration that hydroxymethylglutarylCoA reductase can exist in vitro in active and inactive forms obtained through a cycle of phosphorylation and dephosphorylation of the enzyme suggests a mechanism for rapid short-term regulation in vivo [25]. We have shown previously that feeding rats on a diet supplemented with cholesterol or with cholestyramine results in changes in the isothermic kinetics of hydroxymethylglutaryl-CoA reductase in the microsomal fraction subsequently prepared from liver homogenates and suspended either in buffer I [15] or buffer II [lo]. Moreover, changes in the isothermic kinetics of the enzyme are produced on treatment with ATP plus Mg*+ and are reversed on further incubation of inactivated enzyme with a cytosolic phosphoprotein phosphatase [ 151. The present results demonstrate that changes in activity of the enzyme brought about in vitro by phosphorylation and dephosphorylation are not associated with changes in interaction of the enzyme with its lipid environment. Thus, the Arrhenius plots of inactivated or reactivated enzyme in microsomal preparations from rats in the three experimental conditions show characteristics very similar to those of the corresponding control preparations. The possibility that only the active molecules of hydroxymethylglutarylCoA reductase (non phosphorylated) interact with the membrane cannot be excluded. The present results, however, define the difference between the mechanism of in vitro modulation by non-esterified cholesterol and modulation by covalent modification. Moreover, since the considerable differences in activity of the enzyme in microsomal fraction from rats fed on - sup cholesterol - or cholestyramine standard,
plemented diets are associated with changes in the interaction of the enzyme with the membrane, changes in concentration of non - esterified cholesterol in the endoplasmic reticular membrane are responsible for the differences in activity. Recent work from a number of laboratories [ 15,261 is consistent with the conclusion that the phosphorylation/dephosphorylation cycle is not responsible for the differences in activity of hydroxymethylglutaryl-CoA reductase in microsomal fraction from rats; at least in the dietary conditions studied in the present paper. The changes in the concentration of the enzyme that are associated with these conditions [6,27] presumably follow the initial direct modulation by non-esterified cholesterol and are either due to changes in the rate of association of newly synthesized enzyme to a membrane with altered composition [ 11,241 or to a change in the rate of enzyme synthesis controlled at the pre-translational level [2,3]. References I Dietschy. J.M. and Brown, M.S. (1974) J. Lipid Res. 15, 508-516 2 Rodwell. V.W.. McNamara. D.J. and Shapiro, D.J. (1973) Adv. Enzymol. 38, 373-412 3 Rodwell, V.W.. Nordstrom, J.L. and Mitschelen, J.J. (1976) Adv. Lipid Res. 14, l-74 4 Higgins, M.J.P., Brady, D. and Rudney, H. ( 1974) Arch. B&hem. Biophys. 163, 271-282 5 Higgins. M. and Rudney. H. (1973) Nature New Biol. 246. 60-6 I 6 Edwards, P.A., Lemongello. D., Kanne, J., Schechter. I. and Fogelman, A.M. (1980) J. Biol. Chem. 255, 3715-3725 7 Beg. Z.H., Stonik, J.A. and Brewer, H.B.Jr. (1978) Proc. Natl. Acad. Sci. U.S.A. 75. 3678-3682 8 Ingebritsen, T.S.. Lee, H.-S., Parker, R.A. and Gibson, M.D. (1978) Biochem. Biophys. Res. Commun. 81. l268I277 9 Nordstrom, J.L., Rodwell, V.W. and Mitschelen. J.J. ( 1977) J. Biol. Chem. 252, 8924-8934 IO Mitropoulos, K.A. and Venkatesan, S. (1977) B&him. Biophys. Acta 489, 126- 142 II Mitropoulos. K.A., Balasubramaniam. S., Venkatesan, S. and Reeves, B.E.A. (1978) Biochim. Biophys. Acta 530. 99-111 12 Mitropoulos, K.A., Venkatesan, S., Reeves, B.E.A. and Balasubramaniam, S. (1981) B&hem. J. 194, 265-271 13 Mitropoulos. K.A., Venkatesan, S., Balasubramaniam. S. and Peters, T.J. (1978) Eur. J. B&hem. 82, 419-429 14 Mitropoulos. K.A., Balasubramaniam, S. and Myant. N.B. (1973) Biochim. Biophys. Acta 326,428-438 15 Mitropoulos. K.A., Knight, B.L. and Reeves, B.E.A. ( 1980) B&hem. J. 185, 435-441
455
16 Mitropoulos, K.A. and Balasubramamam, S. (1976) Biothem. J. 160,49-55 17 Balasubramaniam, S., Mitropoulos, K.A. and Venkatesan, S. (1978) Eur. J. Biochem. 90. 377-383 18 Bartlett, G.R. (1959) J. Biol. Chem. 234. 466-468 19 Lowry, O.H.. Rosebrough, N.J.. Fax, A.L. and Randall, R.J. (1951) J. Biol. Chem. 193. 265-275 20 Sabine. J.R. and James. M.J. (1976) Life Sci. 18. I I85- I 192 2 I Sipat. A.B. and Sabine. J.R. (1981) Biochem. J. 194.889-893 22 Mitropoulos. K.A., Venkatesan. S. and Balasubramaniam. S. (1980) Biochim. Biophys. Acta 619. 247-257 23 Mitropoulos, K.A.. Venkatesan, S. and Balasubramaniam,
S. (I 979) in Biological
Effects
of Bile Acids (Paumgartner,
G., Stiehl, A. and Gerok, W., eds.), pp. 137-144, MTP Press, Lancaster 24 Mitropoulos, K.A., Venkatesan, S. and Balasubramaniam, S. (1978) Biochem. Sot. Trans. 6, 878-883 25 Gibson, D.M. 2649-2664
and
Ingebritsen,
T.S. (1978)
Life Sci. 23,
26 Brown M.S., Goldstein, J.L. and Dietschy, J.M. (1979) J. Biol. Chem. 254, 5144-5149 27 Hardgrave, J.E., Heller, R.A., Herrera, M.G. and Scallen, T.J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 3834-3838