the journal of chemistry vol. 263, no. 30, issue of 25, pp ... · vol. 263, no. 30, issue of...

THE JOURNAL OF BIOLOGICAL CHEMISTRY Q 1988 by The American Society for Biochemistry and Molecular Biology, Inc.

Vol. 263, No. 30, Issue of October 25, pp. 15608-15618,1988 Printed in U.S.A.

Effects of Low Density Lipoproteins and Mevinolin on Cholesterol Content and Muscarinic Cholinergic Responsiveness in Cultured Chick Atrial Cells REGULATION OF LEVELS OF MUSCARINIC RECEPTORS AND GUANINE NUCLEOTIDE REGULATORY PROTEINS*

(Received for publication, February 11, 1988)

Linda S. Haigh, George F. Leatherman$, Donald S. O’Hara, Thomas W. Smith, and Jonas B. Galper5 From the Cardiovascular Division, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, Massachusetts 021 15

Cultures of myocytes from embryonic chick atria grown in medium supplemented with fetal calf serum from which lipoproteins had been removed demonstrated a nearly 10-fold increase in sensitivity of beating to the muscarinic cholinergic agonist carbamylcholine compared to cells grown with control serum. This effect was reversed by growth of cells in medium supplemented with lipoprotein-depleted serum (LPDS) reconstituted with the low density lipoprotein fraction from fetal calf serum. In cells grown in LPDS, total cell cholesterol was increased 32% over control levels and returned to control levels in cells grown with LPDS reconstituted with low density lipoprotein. Growth of cells in LPDS plus mevinolin, an inhibitor of endogenous cholesterol synthesis, also reversed the effects of LPDS on cholesterol content and sensitivity of beating to carbamylcholine. The ability of mevinolin (30 pM) to reverse the effect of LPDS on sensitivity of beating to carbamylcholine was inhibited by mevalonic acid, a metabolic precursor to cholesterol, with an ICBo of 7 X lo-’ M. These data suggest that mevinolin reverses the effects of LPDS by altering cellular cholesterol levels. Enhanced responsiveness of embryonic chick heart cells to muscarinic stimulation was associated with a 2-fold increase in the number of muscarinic receptors with high affinity for agonist from 82 f 10 fmol/mg protein in media containing fetal calf serum to 175 f 12 fmol/mg protein in cells grown in the presence of LPDS. The distribution of receptors between high affinity (RH) and low affinity (RL) forms changed from 41% RH and 59% RL in cells grown in control serum to 66.5% RH and 33.5% RL in cells grown in LPDS. Quantitation of the effect of growth in LPDS on the levels of guanine nucleotide regulatory proteins No and Ni which couple the muscarinic receptor to a physiologic response, demonstrated that the relative levels of the 39-kDa a subunits of No and 41-kDa a subunits of Ni determined by ADP ribosylation with pertussis toxin and immunoblotting increased 2-fold

* This work was supported by Grants HL22775 and HL18003 from the National Heart, Lung and Blood Institute, National Institutes of Health and by American Heart Association Grant 81-275. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “adver- tisement’’ in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

$ Present address: Presbytarian Hospital, P.O. Box 2666, Albu- querque, NM 87125.

$ To whom correspondence should be addressed Cardiovascular Div., Brigham and Women’s Hospital, 75 Francis St., Boston, MA 02115.

compared to control cells grown with fetal calf serum. Growth of cells with medium supplemented with LPDS plus mevinolin reduced the levels of as8 and aql to below the levels in control cells. Levels of the B subunit of No and Ni were unaffected by growth with LPDS. Hence, the increase in physiologic response to muscarinic stimulation during growth in LPDS is associated with 1) an increase in cellular cholesterol; 2) an increase in the total number of muscarinic receptors; 3) a 2-fold increase in the number of high affinity muscarinic receptors; and 4) a doubling of the levels of the a subunits of guanine nucleotide regulatory proteins that might couple muscarinic receptors to a physiologic response.

In recent years the dynamic nature of the interaction of hormone and neurotransmitter receptors with the cell membrane has become increasingly apparent (1). Several groups have suggested that the phospholipid and/or cholesterol com- position of the cell membrane may play a critical role in modulating activity of certain membrane-bound enzymes and excitability of cardiac and neuronal tissues (2-3). Hasin et al. (3) demonstrated that a 33% depletion of cholesterol in rat heart myocytes was associated with an increase in the rate of rise of the action potential and a decrease in the sensitivity of beating to both tetrodotoxin and the calcium channel blocker verapamil.

Renaud et al. (4) reported that growth of embryonic chick ventricular cells in medium supplemented with fetal calf serum from which lipoprotein had been removed resulted in increased sensitivity of beating to tetrodotoxin and to the muscarinic agonist oxotremorine. These findings were associated with an increase in total cell cholesterol and an increase in the number of Na+ channels as measured by the binding of an [3H]tetrodotoxin analogue as well as an increase in the number of muscarinic receptors as measured by the binding of the muscarinic antagonist [3H]quinuclidinyl benzilate ( [3H] QNB)’ (4).

Removal of lipoproteins from the growth medium may modulate cellular cholesterol by several mechanisms. High

The abbreviations used are: QNB, quinuclidinyl benzilate; LPDS, lipoprotein-depleted serum; HMG-CoA, hydroxymethylglutaryl-CoA HDL, high density lipoproteins; LDL, low density lipoproteins; VLDL, very low density lipoproteins; NAD+, nicotinamide adenine dinucleotide; MOPS, 4-morpholinepropanesulfonic acid SDS, SO- dium dodecyl sulfate; N-protein, guanine nucleotide regulatory protein; FCS, fetal calf serum.

15608

Cholesterol and Muscarinic Responsiveness in Heart 15609

density lipoproteins (HDL) have been shown to transport cholesterol out of the cell in human skin fibroblasts and in rat aortic smooth muscle cells (5). Hence removal of HDL from the serum might result in an accumulation of cholesterol within the cell. Low density lipoproteins (LDL) transport cholesterol into the cell and regulate the level of endogenous cholesterol synthesis (6). Removal of LDL from the serum would decrease the uptake of cholesterol by the cell, but would also stimulate the synthesis of endogenous cholesterol (6). Hence, in the absence of LDL, the level of cellular cholesterol would depend on the extent of stimulation of endogenous cholesterol synthesis.

In the studies reported here we attempted to determine whether enhanced muscarinic responsiveness could be demonstrated in chick atrial cell cultures grown in media supplemented with lipoprotein-depleted serum and to elucidate the mechanism by which the removal of lipoproteins from the serum component of growth medium mediated this effect. By reconstituting lipoprotein depleted serum (LPDS) with various specific lipoprotein fractions, we determined the factor responsible for modulating the sensitivity of heart cell cultures to muscarinic agonists. Furthermore, mevinolin, a potent competitive inhibitor of hydroxymethylglutaryl-coenzyme A (HMG-CoA) reductase, the rate-determining enzyme in the biosynthesis of cholesterol (7), was used to inhibit endogenous cholesterol synthesis. By growing cells with various concentrations of mevinolin it was possible to directly manipulate cholesterol synthesis and to study the relationship of changes in cellular cholesterol to changes in the physiologic response of the cell to muscarinic agonists, to changes in the number of muscarinic receptors, and the affinity of these receptors for muscarinic agonists and to changes in the levels of guanine nucleotide regulatory proteins which couple muscarinic receptors to a physiologic response.

The physiologic response of the heart to muscarinic stimulation is mediated at least in part via inhibition of adenylate cyclase activity and activation of the muscarinic-gated Kt channel (8, 9). Both of these responses have been shown to be coupled to the receptor via guanine nucleotide regulatory proteins (8-10). Guanine nucleotide regulatory proteins are heterotrimers composed of a, p, and y subunits (11-13). Muscarinic receptors exist in at least two affinity states (14- 16). Data have been presented supporting the hypothesis that association of the receptor with the guanine nucleotide binding protein is required for conversion of the low affinity form of the muscarinic receptor to a high affinity form. Hence, reconstitution of partially purified muscarinic receptors from bovine brain with purified GTP binding proteins (N-proteins) results in conversion of low affinity muscarinic receptors to a high affinity form (17). Furthermore, guanine nucleotides mediate release of the N-protein form the receptor. This release of N-proteins in associated with a decrease in receptor affinity for agonist, consistent with the conversion of the receptor from a high affinity form to a low affinity form (18). We have previously reported that the high affinity form of the muscarinic receptor may be critical for the mediation of a physiologic response to muscarinic stimulation (16, 19).

In the studies reported here, we determined whether the increased sensitivity of beating to muscarinic stimulation in cultured chick atrial cells grown in LPDS was associated with altered cholesterol content and with changes in levels of high affinity muscarinic receptors. We further tested the hypothesis that this change in receptor distribution and physiologic responsiveness to muscarinic stimulation was accompanied by altered levels of guanine nucleotide regulatory proteins.

EXPERIMENTAL PROCEDURES

Materials ['4C]HMG-CoA, [32P]NAD+, and [3H]QNB were from Du Pont-

New England Nuclear; peroxidase and glucose-6-phosphate dehydrogenase were from Boehringer Mannheim; TPN+, mevalonic acid, and HMG-CoA were from Sigma; embryonated chicken eggs were from Spafas, Inc. (Norwich, CT). Agarose gel electrophoresis slides were from Worthington, Freehold, NJ. Tergitol non-ionic surfactant 15-6- 9 was a gift of Union Carbide, Danbury, CT, and mevinolin was a gift of A. W. Alberts, Merck, Sharp & Dohme Research Laboratories, Rahway, NJ. Pertussis toxin was from List Biochemicals and "'1- protein A from Amersham Corp.; acrylamide and bisacrylamide were from Bio-Rad.

Cultures and Media Heart Cell Cultures-Heart cell cultures were prepared by a modi-

fication of the method of DeHaan (20). Embryonic age was determined by the method of Hamberger and Hamilton (21). Embryos were removed from embryonated Leghorn chicken eggs on day 14 in ouo, hearts removed, and atria separated and incubated with 0.025% (w/v) trypsin in Ca2+-Mp2+-free Hanks' balanced salt solution at 37 "C for 8 min. The trypsin solution was removed and diluted into medium "199 containing 50% heat-inactivated horse serum at room temperature. After successive trypsinizations, suspensions of trypsin- ized cells were sedimented at 1000 rpm and resuspended in growth medium containing either 6% FCS or 6% LPDS, as described in the figure legends, at a density of 1.6 X IO6 cells/cm2, in either 100-mm Petri dishes (Falcon, Oxnard, CA) or on glass coverslips in 100-mm Petri dishes and incubated at 37 "C in a humidified atmosphere of 5% COz, 95% air. Unless otherwise indicated, cells were used from experiments on culture day 4. In experiments involving mevinolin, the drug was added on the 3rd culture day.

Measurement of Changes in Beating Rate-Cells grown on glass coverslips were placed in a chamber and continuously perfused via inlet and outlet ports. The chamber was placed on the stage of an inverted phase contrast microscope enclosed in a Lucite box main- tained at 37 "C. The inlet side of the chamber was connected by polyethylene tubing to two syringe pumps to allow the cultures to be sequentially perfused by separate solutions. Perfusion at 0.98 ml/min did not disturb cell adhesion to the coverslip. Beating was determined by monitoring the movement of the border of a single cardiac cell with a video motion detector and recording the output with a physiologic recorder (Hewlett-Packard Co., Palo Alto, CA) as described previously (22, 23). Each set of data represents the mean of measurements carried out on multiple cells from several different cultures.

Preparation of LPDS-LPDS and purified lipoprotein fractions were prepared by the method of Have1 et al. (24). For the preparation of VLDL, LDL, and HDL a batch of serum was overlaid first with normal saline (density 1.006 g/ml) and subjected to centrifugation in a Ti-70 rotor at 265,000 X g for 16 h at 10 "C. Following removal of the upper VLDL-containing layer, the procedure was repeated by adjusting the density of the lower serum layer first to 1.0635 with KBr and overlayering with normal saline at a density of 1.0635 g/ml for preparation of LDL, followed by repeat centrifugation of the serum layer at a density of 1.21 g/ml for flotation of the HDL fraction. LPDS was that serum fraction remaining after successive removal of all three lipoprotein fractions. Both LPDS and individual lipoprotein fractions were dialyzed against 0.15 M NaCl and 2 mM Na2HOP4, pH 7.4. Agarose gel electrophoresis of LPDS followed by staining with Fat Red demonstrated that all three lipoprotein fractions were un- detectable.

Acetylation of the LDL Fraction-Acetylation of the LDL fraction was carried out as described previously by Basu et al. (25). 1 ml of buffer (154 mM NaC1, 2 mM Na2HP04, pH 7.4) containing 33 mg of LDL protein was mixed with 1 ml of saturated sodium acetate and stirred over ice for 30 min. Acetic anhydride was added in five 2 - 4 (5

buffer (0.15 M NaCl, 2 mM Na2HP04, pH 7.4). Agarose gel electro- mg/pl) aliquots over 1 h. The reaction mixture was dialyzed against

phoresis of the sample followed by staining with Fat Red demonstrated complete conversion of LDL to a band that ran ahead of native LDL as described for acetyl-LDL (25).

Cholesterol Assay-Cholesterol was extracted from the cell pellet by sonication in the presence of 1 ml of 1-propanol. Total cholesterol in the extract was assayed by the method of Heider and Boyett (26). LDL cholesterol was determined by the method of Seidel et al. (27).

HMG-CoA Reductase Assay-HMG-CoA reductase activity was

15610 Cholesterol and Muscarinic Responsiveness in Heart

assayed by the method described by Alberts et al. (7). The washed frozen cell pellet was extracted with buffer (50 mM KzHP04, pH 7.4, 5 mM dithiothreitol, 5 mM EDTA, 0.2 M KCl), Tergitol non-ionic surfactant 15-5-9 (identical to Kryo EOB), 0.25% (w/v), for 10 min at 37 "C. Following centrifugation at 14,000 X g for 1 min, an aliquot of the supernatant containing 10-20 mg of protein was incubated for 2 h a t 37 "c in a reaction mixture containing 10.0 mM KzHPO,, 8 mM dithiothreitol, 5 mM NADf, 20 mM glucose-6-P04, 1-3 units of glucose-6-P04 dehydrogenase, 0.15 mM [14C]HMG-CoA (10 pCi/2.5 pmol/ml) in a volume adjusted with Hz0 to 0.100 ml. The reaction was stopped by addition of 20 pl of 5 N HC1, which converted free ["C]mevalonic acid to the ['*C]mevalonic acid lactone. The ["C] mevalonolactone was separated from unreacted [14C]HMG-CoA by elution with water on a Bio-Rex 5 anion exchange column. The eluate was collected in scintillation vials and counted with 17 cm3 of Instagel (Packard). Assays were carried out in triplicate. Control incubation with buffer alone demonstrated that specificity of conversion was greater than 70%. Activity was calculated as femtomoles of mevalonic acid produced per min/mg protein. Protein concentration following trichloroacetic acid precipitation was determined by the method of Lowry et al. (28), using BSA as standard.

ADP Ribosylation with Pertussis Toxin ADP ribosylation was carried out for the times indicated at 37 "C.

The assay mixture contained 10 p M NAD, 0.3-0.5 pCi of [3ZP]NAD, 2.5 mM ATP, 2 mM GTP, 10 mM isoniazid, 10 mM thymidine, 10 p~ MgC12,25 ng of pertussis toxin, 10 mM phosphocreatine, and 84 units of creatine phosphokinase in a total volume of 25 pl. Atrial cells scraped from 100-mm Petri dishes into buffer (50 mM Tris-HC1, pH 7.6, 10 mM MgClZ, 0.2 M sucrose, 1.0 mM dithiothreitol, 1 mM EDTA) were frozen and thawed once followed by Dounce homogenization and added to the reaction mixture a t 50-100 pg of protein/assay. Protease inhibitors (soybean trypsin inhibitor, lima bean trypsin inhibitor, and leupeptin) were present at the concentrations described previously (29) from the time of initial homogenization of the tissue. The reaction was stopped by the addition of 2% SDS in Laemmli sample buffer (30) followed by boiling for 1 min. Analysis of peptides following ADP ribosylation was performed on 9% acrylamide gels prepared according to Laemmli (30). Dried gels were exposed to Kodak XAR film with or without enhancing screens for 1-2 days at

Immunoblotting-Polyclonal (rabbit) antisera to both bovine 0139

and p subunits of No were obtained from Dr. E. J. Neer (Department of Medicine, Brigham and Women's Hospital). Immunoblotting was carried out by a modification of the method of Towbin et al. (31). Heart proteins were solubilized in 2% SDS in Laemmli sample buffer as described for ADP ribosylation and separated by SDS-polyacrylamide gel electrophoresis on 9% acrylamide gels. The gels were equilibrated in transfer buffer 20 mM Tris-OH, 150 mM glycine, 0.015% SDS, and 20% methanol for 30 min. The proteins were then electrophoretically transferred to nitrocellulose at 30 V overnight. The dried filters were incubated with 3% bovine serum albumin, in 10 mM Tris, pH 7.6, 150 mM NaCl for 60 min at 23 "c to decrease nonspecific binding, followed by incubation overnight a t 4 "C with antiserum diluted 1:200 with the same buffer. Nitrocellulose was then washed two times with 10 mM Tris, 150 mM NaCl, for 10 min, once with 10 mM Tris, 150 mM NaC1, and 0.05% Nonidet P-40 for 20 min, and again with 10 mM Tris, 150 mM NaCl for 10 min. 1z51-staphylo- coccal protein A (Amersham Corp., 100,000-150,000 cpm/ml) in 10 mM Tris, 150 mM NaC1, 3% bovine serum albumin was incubated with blots for 1 h and the wash procedure repeated. After drying, blots were exposed to Kodak XAR film with enhancing screens for 1-5 days at -70 "C.

Tissue Preparation for Binding Assay-Cultures of embryonic chick atria 14 days in ouo were harvested by scraping with a rubber policeman and suspended in Locke's solution (154 mM NaCl, 5.6 mM KCl, 2.2 mM CaC12, and 10 mM MOPS), frozen and thawed X 2, and Dounce-homogenized. The homogenate was centrifuged a t 40,000 X g for 40 min and the pellet resuspended in an appropriate volume buffer B (1.8 mM CaClZ, 0.81 mM MgzSO4, 37.5 mM Tris, pH 7.1, 23 "C).

Receptor Binding Assay-Tissue homogenates were allowed to warm to room temperature (23 "C) prior to assay in a 2-ml Microfuge tube containing approximately 100 pl of glass heads; 0.25 ml of tissue homogenate usually containing between 100-200 pg of protein was mixed with 0.25 ml of the same buffer containing 2 nM [3H]QNB and the indicated concentrations of carbamylcholine. Incubation was for

-70 "C.

1 h followed by centrifugation at 14,000 X g for 3 min in a Beckman Microfuge, the supernatant was removed and the pellet, glass beads, and small amount of incubation buffer vortexed for 10 s. The presence of glass beads during the vortex step helped break the pellet into finely dispersed particles and release the unbound ligand nonspecifically trapped within the pellet. Following the vortex step, the pellet and glass beads were washed onto a Whatman GF/C filter by invert- ing the Microfuge tube over a stainless steel filtration port and washing the tube with three 2-ml aliquots of ice-cold buffer B with a repeating syringe. The filter port and the filter were further washed with three 2-ml aliquots of buffer B to remove glass beads from the walls of the filter port and wash them into the center of the filter. The validity of this method has been established as described previously (32). Specific binding was determined by binding in the presence of mM carbamylcholine or M atropine plus 2 nM [3H]QNB. Filters were dried and assayed for radioactivity in a liquid scintillation counter with 10 ml of Scint A (Beckman).

Data were fit to two simultaneous equations by the "LIGAND" program of Munson and Rodbard

where R1 = molar concentration of receptors with affinity Kzl for agonist or K11 for antagonist: Rz = molar concentration of receptors with affinity Kz2 for agonist and Klz for antagonist; TI = total added [3H]QNB concentration; T2 = total added carbamylcholine concentration; FI = free [3H]QNB concentration; and FP = free carbamylcholine concentration. In these analyses, Kll and K12 are kept constant at 1 X 10"' "', and T1 is 1 nM.

RESULTS

Increased Sensitiuity of Beating to Muscarinic Agonists in Chick Atrial Cultures Grown in Media Supplemented with Lipoprotein-depleted Serum-Cells cultured from atria of embryos 14 days in ouo contracted at a rate of 160 * 15 beats/ min (S.E., n = 20, data are the mean of 5 determinations from 4 separate cultures). No difference in either beating frequency, cell morphology, total protein content per dish (0.20 f 0.025 mg/lOO-mm dish), or DNA content (determined by the method of LeBarca (34)) could be detected between control cells and cells grown in LPDS. However, sensitivity of beating to the muscarinic agonist carbamylcholine increased 10-fold in cells grown in LPDS compared to control cells. Beating was totally inhibited at M carbamylcholine with an ICs0 of 3 x M in control cells compared to complete inhibition of beating at M carbamylcholine with an Ic50 of 3 X M carbamylcholine in cells grown in media supplemented with LPDS (Fig. 1). The effect of LPDS was specific for atrial cells, since beating in ventricular cultures of hearts 14 days in ouo decreased only 15% at concentrations of carbamylcholine as high as M in both control and LPDS cultures (data not shown).

To determine whether the removal of a specific lipoprotein component from fetal calf serum was responsible for confer- ring increased sensitivity of beating to carbamylcholine, cells were grown in media supplemented with LPDS that had been reconstituted with specific lipoprotein fractions. Experiments summarized in Fig. 1 demonstrate that reconstitution of LPDS with the VLDL fraction did not reverse the enhanced sensitivity of beating to carbamylcholine seen in cells grown in LPDS alone. Incubation of cells in medium supplemented with LPDS reconstituted with the HDL fraction also had no effect on the enhanced sensitivity of beating to carbamylcholine seen in cells grown in LPDS alone. In contrast, incubation of cells with medium supplemented with LPDS reconstituted with the LDL fraction resulted in complete reversal of the effects of growth of cells in LPDS (Fig. 2). Reconstitution of LPDS with LDL increased the ICso for inhibition of beating


10-7 3 x 10-7 10-6 3 x 10-6 IO-^ [ C A R B A M Y L C H O L I N E ] , M

FIG. 1. Increased sensitivity of beating to carbamylcholine in chick atrial cells cultured in medium supplemented with lipoprotein-depleted serum. Atria of chick embryo hearts 14 days

with fetal calf serum (0); LPDS (0); LPDS plus the VLDL fraction in ouo were cultured on glass coverslips in medium supplemented

prepared as described under “Experimental Procedures” (A); LPDS plus the HDL fraction prepared as described above (0). On the 4th culture day cells were exposed to the indicated concentrations of carbamylcholine and beating rate determined using a video motion detector as described under “Experimental Procedures.” Control data represent the mean of 5 determinations carried out in 3 separate cultures. VLDL data represent the mean of 3 determinations in 3 separate cultures and HDL data the mean of 4 determinations in 3 separate cultures. Lipoprotein fractions were reconstituted with LPDS in the same ratio (v/v) as the relative volumes into which they were separated during floatation (0.31 ml of lipoprotein fraction/ml of LPDS-containing medium).

from 3 X M carbamylcholine in LPDS to 2 X M in reconstituted serum.

To determine whether the effects of adding back LDL to LPDS required the binding of LDL to LDL receptors, we studied the effect of growth of cells in medium supplemented with LPDS reconstituted with the acetylated form of LDL on the sensitivity of beating to carbamylcholine (25). The data summarized in Fig. 2 demonstrate that acetylation of LDL completely abolished its effect on muscarinic responsiveness. Even at concentrations 2-3 times that of LDL found in fetal calf serum, growth with acetyl-LDL had no effect on carbamylcholine sensitivity.

Increased Cellular Cholesterol in Cells Grown in Media Sup- plemented with LPDS Is Reversed by Reconstitution with the LDL Fraction-Renaud et al. (4) suggested that growth in LPDS raised the level of cellular cholesterol. We found that growth of chick atria cells in media supplemented with LPDS resulted in a 28% increase in total cholesterol from 23 k 1 (S.E., n = 3) pg/mg protein to 32 4 2 pg/mg protein (Table I). Reconstitution of LPDS with VLDL and/or HDL had no significant effect on cellular cholesterol levels which remained elevated at 30 -+ 2 pg/mg protein. However, reconstitution with the LDL fraction lowered cholesterol to control levels. Hence, the removal of the LDL fraction of serum is responsible for the effects of growth of heart cells in media supple-

I I 1 I

IO-? 3 x to-? 10-6 3x 10-6

[ C A R B A M Y L C H O L I N E ] . M

FIG. 2. Reconstitution with the LDL fraction decreased sensitivity of beating to muscarinic stimulation in cultured chick atrial cells grown in LPDS. Cells from atria of chick embryo

supplemented with fetal calf serum (0); LPDS (0); LPDS plus the hearts 14 days in DUO were cultured on glass coverslips in medium

LDL fraction (A); LPDS plus acetyl-LDLprepared as describedunder “Experimental Procedures” (W). On the 4th culture day the effect of various concentrations of carbamylcholine on beating rate was determined as described in Fig. 1. The LDL data represent the mean of 4 determinations carried out in 3 separate cultures, the acetyl-LDL data the mean of 2 determinations carried out in 2 separate cultures. The ratio of LDL to LPDS was as described in Fig. 1.

TABLE I Total cell cholesterol in cultured embryonic chick atrial cells

~~ ~~

Growth medium Cholesterol supplementation (pg/mg protein)”

Control serum LPDS

23 rt_ 1 (S.E., n = 3)

LPDS + HDL 32 f 2

LPDS + VLDL 31 f 1

LPDS + LDL 30 f 2 24 f 2

LPDS + 30 pM mevinolin* 20 & 2 “Atrial cultures from chick hearts 14 days in ouo were grown on

100-mm Petri dishes in medium supplemented with serum as indicated. After the 4th day in culture, cells were harvested, washed, extracted with I-propanol, and total cholesterol determined as described under “Experimental Procedures.”

Mevinolin was added on the 3rd culture day and incubation continued for 24 h before extraction of cholesterol.

mented with LPDS on the enhanced physiologic response of chick atrial cells to muscarinic stimulation and on cholesterol levels. LDL has been shown to regulate cholesterol levels 1) by transporting cholesterol into the cell, 2) by feedback inhibition of LDL cholesterol on HMG-CoA reductase levels, and 3) by feedback control on the number of LDL receptors (6). An increase in cellular cholesterol following removal of LDL from the serum could be related to enhancement of HMG- CoA reductase activity.

Inhibition of Endogenous Cholesterol Synthesis by Mevinolin Decreases Total Cell Cholesterol and the Response of Cultured Chick Atrial Cells to Muscarinic Stimulation-Mevinolin, a fungal metabolite isolated from Aspergillus terreus, bears a


moiety structurally analogous to mevalonic acid, a precursor in the biosynthesis of cholesterol. Mevinolin is a potent competitive inhibitor of HMG-CoA reductase, the enzyme that catalyzes the conversion of HMG-CoA to mevalonic acid (7). Incubation of atrial cells grown in LPDS with mevinolin resulted in a 39% decrease in total cell cholesterol from 32 f 2 pg/mg protein to 20 f 2 pg/mg protein (Table I).

To test the hypothesis that a decrease in cellular cholesterol seen during growth of cells in medium supplemented with LPDS plus mevinolin is associated with decreased sensitivity of beating to carbamylcholine, cells were grown in LPDS for 3 days. On the 3rd culture day, various concentrations of mevinolin were added to the culture media and incubation continued for 24 h. Although the beating rate was somewhat irregular in the presence of maximal doses of mevinolin, average beating rate was 150 f 15 (S.E., n = 5) beats/min, not significantly different from control cells. Total protein/ dish and total DNA/mg protein was unaffected by growth of cells with LPDS plus mevinolin.

As noted previously (Fig. l ) , a t 3 x M carbamylcholine beating was inhibited nearly 100% in cells grown in LPDS, but 3 X 10"j M carbamylcholine inhibited beating only 40% in control cells. Data summarized in Fig. 3a demonstrate that mevinolin decreased the sensitivity of beating to carbamylcholine in cells grown in medium supplemented with LPDS over a narrow concentration range. In Fig. 3a, a percent maximal beating rate of zero represents 100% inhibition of beating. Hence, a t 10 p~ mevinolin and 3 X M carbamylcholine, beating was completely inhibited while at 40 p~ in mevinolin and 3 x M carbamylcholine, beating was 65% of maximum equivalent to a 35% inhibition of beating. Data in Fig. 1 demonstrate that in cells grown in FCS, inhibition of beating at 3 X M carbamylcholine was approximately 35%. This effect of mevinolin on the sensitivity of beating to carbamylcholine was half-maximal a t 15 p~ mevinolin (Fig. 3a). The effect of mevinolin on the beating rate response to carbamylcholine is further illustrated in the experiment summarized in Fig. 3b. In these studies, the ICs0 for carbamylcholine inhibition of beating increased from 3 x

M in LPDS to 3 X M in LPDS plus 30 p M mevinolin; the latter value is quite similar to the value for the IC50 for carbamylcholine inhibition of beating in control cells (Fig. 1). Hence, mevinolin appears to reverse the effect of LPDS on the sensitivity of beating to carbamylcholine.

One possible explanation of these effects of mevinolin is that it acts directly and nonspecifically on the cell membrane. However, mevalonic acid ( M), the product of HMG-CoA reductase, completely reversed the effects of 30 pM mevinolin on sensitivity of beating to carbamylcholine (Fig. 3a, open circles). This reversal was dependent on the concentration of mevalonic acid with a half-maximal effect a t 7 X M (Fig. 4). Since mevalonic acid is a small molecule with no known direct membrane effects and since it specifically reverses the inhibition of cholesterol synthesis by mevinolin, these data support a direct role of cholesterol synthesis on muscarinic responsiveness in cultured chick atrial cells.

Correlations between Changes in Cholesterol Level and Changes in Muscarinic Responsiveness-Treatment of fibroblasts with mevinolin and/or compactin (6-dimethylmevino- lin) results in the induction of high levels of HMG-CoA reductase activity (37). In order to study the relationship between changes in HMG-CoA reductase activity, changes in levels of cellular cholesterol, and changes in the physiologic response of chick atrial cells to muscarinic stimulation, cultures incubated with LPDS for 3 days followed by 24 h incubation with LPDS plus various concentrations of mevi-

u t 1

[ M E V I N O L I N ] . * M

[CARBAMYLCHOLINE 3. M

FIG. 3. Decreased sensitivity of beating to carbamylcholine in atrial cells grown in LPDS plus mevinolin (a); effect of mevinolin on the concentration-effect curve for carbamylcholine inhibition of beating in cells grown in the presence of LPDS (b ) . a, cells from atria of chick embryo hearts 14 days in ouo were cultured on glass coverslips in medium supplemented with LPDS. On the 3rd culture day mevinolin was added at the indicated concentrations with or without 2 mM mevalonic acid and incubation continued for 24 h. Inhibition of beating at 3 X lo-' M carbamylcholine was determined as described in Fig. 1. Data are plotted as a percentage of the maximal beating rate (the beating rate in the absence of carbamylcholine) at each concentration of mevinolin. Hence, 0 implies 100% inhibition of beating. Data represent the means of 10 determinations at 10 p M , 5 at 12.5 pM, 8 at 20 pM, 6 at 30 p ~ , 5 at 40 phi, and 4 at 50 p~ from 2 or more cultures; error burs indicate f S.E. 0, mevinolin; 0, mevinolin plus 2 mM mevalonic acid. 6 , experiment carried out as in Fig. 1; cells grown in medium supplemented with LPDS (0); LPDS plus 30 p M mevinolin (V).

nolin were washed and divided into aliquots for determination of cholesterol (closed triangles, Fig. 5) and HMG-CoA reductase activity (open circles, Fig. 5). Cholesterol levels decreased 38% between 10 and 30 p~ mevinolin with the steepest part of the curve between 15 and 25 pM mevinolin and a half- maximal effect at 19 p~ mevinolin (Fig. 5). The narrow concentration range and IC50 for this effect of mevinolin on cholesterol levels are quite similar to the concentration de- pendence and ICs0 for the effect of mevinolin on the sensitivity of beating to carbamylcholine (Fig. 3a).

HMG-CoA reductase activity increased dramatically, 56- fold, from 4.8 pmol of mevalonic acid produced per mg protein/min in control cultures to 270 pmol of mevalonic acid


‘\ I aePN 1. ’ ’ ’ I ’ l b j

0 10-7 10-6 10-5 10-4 10-3 [MEVALONIC ACID1 (MI

FIG. 4. Mevalonic acid reverses the effect of mevinolin On the sensitivity of beating to carbamylcholine in embryonic chick atrial cells grown in media supplemented with LPDS plus 30 I .~M mevinolin. Cells from atria from embryonic chick hearts 14 days in ouo were cultured on glass coverslips in media supplemented with LPDS for 3 days. On the 3rd culture day, medium was made 30 PM in mevinolin and the indicated concentrations of mevalonic acid added. On the 4th culture day, coverslips were removed and placed in a perfusion chamber on the stage of an inverted phase contrast microscope and the effect of 2 X M carbamylcholine on beating rate determined as described in Fig. 1 and under “Experimen- tal Procedures.” Data are plotted as the percentage of the beating rate at a 2 X M carbamylcholine in the absence of mevalonic acid. Each point represents the mean of 5 determinations in 3 separate cultures (+S.E.).

[ M E V I N O L I N ] . p M

FIG. 5. Decreased total cell cholesterol and increased HMG- CoA reductase activity in cells cultured with medium supplemented with LPDS plus mevinolin. Atrial cells from hearts 14 days in ouo were plated on 100-mm Petri dishes a t 4.8 X lo6 cells/ dish. On the 3rd culture day mevinolin was added to replicate dishes at the concentrations indicated and incubation continued for 24 h. Cells were washed 3 times with buffer (150 m M NaC1, 50 mM Tris- HCl, pH 7.4), harvested and the pellet extracted with 1-propanol. Cholesterol was determined as described under “Experimental Pro- cedures.” Data represent the means of 8 determinations carried out in triplicate. Aliquots of cells from replicate cultures were pelleted and frozen in liquid nitrogen. Pellets were extracted with “detergent buffer” and assayed for HMG-CoA reductase as described under “Experimental Procedures.” Data represent the mean of 3 determinations in triplicate. The assay was 70% specific as measured by comparison of hydrolysis of HMG-CoA in the presence of buffer alone. A, cholesterol; 0, HMG-CoA reductase activity. MA, mevalonic acid.

produced per mg protein/min at 10 p M mevinolin. The increase in HMG-CoA reductase activity reached greater than 90% of maximum before any significant change in cholesterol could be detected. Furthermore, only when HMG-CoA reductase activity had plateaued at 10 FM mevinolin did the decrease in cholesterol levels appear. One explanation of these data is that once the cell has exceeded its capacity to increase

HMG-CoA reductase activity, cholesterol levels begin to fall steeply.

Increases in the Number of Muscarinic Receptors and in the Number of High Affinity Muscarinic Receptors Associated with Increases in Cellular Cholesterol-Renaud et al. (4) suggested that growth of cells in LPDS resulted in an increase in the number of muscarinic receptors. We compared the binding at 1 nM [3H]QNB in cells grown in control serum, LPDS, and LPDS plus various concentrations of mevinolin. In control cells specific binding was 180 -+ 20 fmol/mg protein (S.E., n = 4). In cells grown in LPDS binding increased to 235 f 25 ( n = 4) fmol/mg protein and in LPDS plus 30 p M mevinolin binding decreased to 178 f 6 ( n = 4) fmol/mg protein.

Comparison of the concentration-effect curve for the mevinolin-mediated decrease in muscarinic receptor number and the mevinolin-mediated decrease in cholesterol demonstrates striking similarities (Fig. 6). Like the effect of mevinolin on cholesterol (closed triangles), the decrease in receptor number (closed circles) also takes place over a narrow concentration range from 15 to 30 pM mevinolin with an ICbo at 19 pM.

The relative distribution of muscarinic receptors between high and low affinity states in control cells and cells grown in LPDS was determined by comparison of competitive binding between a fixed concentration of [3H]QNB and various levels of carbamylcholine. Since competitive binding is a function of affinity of the receptor for both [3H]QNB and the agonist, it was first necessary to determine whether growth in LPDS altered the KD for [3H]QNB binding to cell homogenates. Scatchard analysis of equilibrium binding of various concentrations of [3H]QNB demonstrated no significant difference in the K D for binding of [3H]QNB to cells grown in medium supplemented either with control serum or LPDS (KD = 0.10 pM f 0.01 ( n = 3), S.E. (data not shown)).

Competition binding studies are summarized in Fig. 7, a and b. Data are the mean of 6 experiments carried out in triplicate. We have previously shown that computer fitting of these data to Equations 1 gave the best fit to a two-site model (16). Here the assumption was made that at steady state K11 = Klz = 10“’ M-’, the association constant for [3H]QNB. The solid line in Fig. 7, a and b, is that line drawn by the computer for the best fit of the data. The dotted lines represent

I I I

0 10 20 30 40

FIG. 6. Decreased total cell cholesterol and [‘HIQNB binding sites in cells grown in media supplemented with LPDS plus mevinolin. Cholesterol levels in the presence of the indicated concentrations of mevinolin were obtained as described in Fig. 5. Cells were harvested, frozen, and thawed and Dounce-homogenized. Specific binding of [3H]QNB (1 nM) was measured following a 1-h incubation of the resuspended membrane pellet using the centrifugation filtration assay described under “Experimental Procedures.” Protein per assay was usually 100-200 fig determined by the method of Lowry (28). Nonspecific binding defined as binding at lo-’ M unlabeled QNB was subtracted from each point. Experimental points are the mean of 5 experiments carried out in triplicate, 0, [3H]QNB bound; A, total cell cholesterol.


a 3 0 1 ,

20 1

log,OICAREAMYCHOLINE1, M

b 3 0 [ 1 1

-

-

5- - -9 -8 -7 -6 -5 -4 -3

lOg~0[CARBAMYCHOLINE], M

FIG. 7. Increased numbers of muscarinic receptors in the high affinity form in cells grown in LPDS. Competition binding studies between 1 nM [3H]QNB and the indicated concentrations of carbamylcholine were carried out as described under “Experimental Procedures.” Data from 6 experiments were simultaneously fit to Equation 1 by nonlinear least squares analysis. Nonspecific binding determined in the presence of 3 X M carbamylcholine and/or

M unlabeled QNB was subtracted from all points. Data points are the mean of 6 determinations carried out in triplicate, error bars represent S.E. The solid lines are the computer-derived optimal fit for the data. The dotted lines represent the computer-derived plot for binding to low affinity sites and the dashed lines represent the computer-derived plot for binding to the high affinity site. a, cells grown in media supplemenmted with fetal calf serum. b, cells grown in media supplemented with LPDS.

the computer-generated binding curve to the low affinity site while the dashed lines represent the computer-generated binding curve for the high affinity site.

The KO values for binding of carbamylcholine to both high and low affinity sites were unchanged by growth of cells with media supplemented with LPDS. The KD values for the high affinity site derived as the mean k S.E. from analyses of 6 different experiments carried out in triplicate as described in Fig. 7, a and b, was 0.18 f 0.06 X M in control cells compared to 0.5 f 0.21 X M in cells grown with LPDS ( p = 0.12). For the low affinity site, the KO was 0.11 f 0.04 X M for control cells and 0.4 -+ 0.15 X M in cells grown with LPDS ( p = 0.08).

However, growth of cells in media supplemented with LPDS had a marked effect on the relative distribution of muscarinic receptors between high and low affinity states. Comparison of the computer drawn plots for high and low affinity binding in Fig. 7, a and b, demonstrates that the ratio of low affinity to high affinity receptors is reversed in cells grown with LPDS compared to control cells. Data for the mean of parameters derived from the analysis of six different experiments summarized in Table I1 demonstrated that while the distribution of receptors was 41% RH and 59% RL in cells grown in media containing FCS, this distribution reversed to 63% RH and 38% R.L in LPDS. Furthermore, the number of high affinity sites doubled from 83 k 11 fmol/mg in cells grown in FCS to

TABLE I1 Summary of the effects of growth in medium supplemented with

LPDS on the distribution of muscarinic receptors between high and low affinity agonist binding states

Data are the mean of parameters f S.E. derived from the analysis of six different experiments carried out in triplicate as described in Fig. 7, a and b. The FCS - LPDS column is the difference between total receptor numbers in FCS and LPDS cultures which was significant at p = 0.00002, the increase in the number of high affinity sites which was significant at p = 0.005, and the change in the number of low affinity sites which was not significant a t p = 0.32 ( p values by Student’s t test).

Binding (fmol/mg protein)

High Low Total affinity affinity

FCS 202 f 15 83 f 11 LPDS

119 f 11 279 f 15 175 f 12

FCS - LPDS 104 f 9

77 f 21 92 f 16 -15 f 14

175 f 12 fmol/mg in cells grown in the presence of LPDS. Total receptor number increased by 77 f 21 fmol/mg in cells grown in the presence of LPDS, while the total number of high affinity sites increased by 92 f 16 fmol/mg. Since the decrease in the number of low affinity sites, 15 k 14 fmol/mg protein, was not statistically significant, these data demonstrate that the appearance of new receptors accounted for nearly all the increase in high affinity binding sites.

Increases in the Levels of Guanine Nucleotide Regulatory Protein in Cells Grown in Medium Supplemented with LPDS-Since the interaction of guanine nucleotide regulatory proteins with the low affinity form of the muscarinic receptor has been shown to be necessary in order for the receptor to assume the high affinity form (17), one explanation for the 2-fold increase in the number of high affinity receptors is that newly appearing low affinity muscarinic receptors interact with increased levels of guanine nucleotide regulatory proteins. We have previously demonstrated that pertussis toxin stimulated ADP ribosylation of embryonic chick heart homogenates in the presence of [32P]NAD followed by polyacrylamide gel electrophoresis and autoradiography was a reliable method for determining the relative levels of the a subunits of No and Ni (29), a 3 9 and a 4 1 in chicken heart homogenates. Ni is associated with coupling the muscarinic receptors to inhibition of adenylate cyclase while the relationship of No to the muscarinic response is unclear. The effect of LPDS and mevinolin on a 3 9 and a 4 1 is summarized in Fig. 8. Autoradiographs of polyacrylamide gels of homogenates of cells ADP-ribosylated with pertussis toxin and [“PI NAD in control cells, cells grown in LPDS, and cells grown with LPDS plus mevinolin revealed two bands which corre- spond to molecular weights of 39,000 and 41,000 (Fig. 8). The intensity of these bands is significantly increased in cells grown in LPDS and decreased below control levels in cells grown with LPDS plus mevinolin (Fig. 8). Comparison of densitometry scans of these autoradiographs corroborated the findings (not shown). The experiment summarized in Fig. 8 is typical of 8 others. Data in Fig. 9 summarize the mean k S.E. of densitometry scans from 9 experiments carried out as described in Fig. 8. Compared to control homogenates, levels of a 3 9 increased 1.93 f 0.04-fold (S.E., n = 9) and a 4 1 increased 2.1 +- 0.06-fold in LPDS. Hence, increased availability of cy39

and a 4 1 in cells grown in media supplemented with LPDS may account for the conversion of low affinity receptors to a high affinity form.

Mevinolin decreased levels of a 3 9 to 0.54 f 0.07-fold (fS.E., n = 9) below the level in control cells grown with FCS and a 4 1 to 0.6 f 0.07-fold below the level in control cells. Hence,


kDa

J4‘ 39

F L L

M +

FIG. 8. Typical experiment demonstrating increased levels of pertussis toxin substrate in cultures of chick atria grown in medium supplemented with LPDS. Autoradiography following polyacrylamide gel electrophoresis of embryonic chick heart proteins ADP-ribosylated with pertussis toxin in the presence of [“P]NAD’. Homogenates of replicate cultures incubated with media supplemented with either FCS, LPDS, or LPDS plus 30 p td mevinolin were prepared and incubated for 30 min a t 37 “C with pertussis toxin and [‘*PINAD’, solubilized, and subjected to polyacrylamide gel electrophoresis and autoradiography as described under “Experimental Procedures.” Each lane contains 60 pg of protein as determined by the method of Lowry (28). Molecular weights were determined by plotting the distance migrated in milli- meters for standard proteins stained with Coomassie Blue uersu.9 the logarithm of their molecular weights. The standards included trypsin- ogen, carbonic anhydrase, glyceraldehyde-3-phosphate dehydrogenase, egg albumin, bovine albumin; molecular weights were taken as 24,000, 29,000, 36,000, 45,000, and 66,000, respectively. Calculated molecular weights for the two bands were 39,000 f 0.04 and 41,000 f 0.03 (S.E., R = 9). F, FCS; L, LPDS; and M, 30 PM mevinolin.

the decreased responsiveness of cells grown in LPDS plus mevinolin to muscarinic stimulation may be due a t least in part to reversal of the effects of LPDS on levels of as9 and

Since differences in levels of cellular cholesterol might effect the relative availability of pertussis toxin substrate for ADP ribosylation, the ADP ribosylation reaction was carried out either in the presence of increased levels of pertussis toxin or in the presence of cholate (1.0-0.1%). Effects of longer incubation times were also examined. These alterations had no effect on the relative levels of ADP ribosylation in cells grown with FCS or LPDS (data not shown).

Quantitation of Levels of asg by Immunoblotting-In order to establish more accurately the relative levels of a39 effects of growth of cells in LPDS and LPDS plus mevinolin were quantitated by immunoblotting. We have previously demonstrated the cross-reactivity of antibody to both bovine brain aR9 and the p subunit of No or Ni with homogenates of embryonic chick hearts (29). Data from a typical experiment in which levels of ang were quantitated by immunoblotting are summarized in Fig. loa. Homogenates of cultured chick atrial cells were subjected to polyacrylamide gel electrophoresis, electrotransferred to nitrocellulose, and incubated with antibody to asg followed by incubation with ‘251-protein A and

a 4 l -

F L L + M

FIG. 9. Histogram of relative levels of a39 and a41 determined from densitometry tracings of autoradiograms in cells grown in LPDS. Peaks from densitometry tracings from 9 experiments similar to those in Fig. 8 were traced individually and the overlapping edge of each peak approximated by symmetry with the free edge. Peaks were cut out and the relative weight under each set of growth conditions determined. Data from a39 and arl were normalized to their values in cells grown with FCS taken as 1.0. Values shown are the mean f S.E. for 9 determinations. 0 = a39; = a41. F, FCS; L, LPDS; M, 30 p~ mevinolin.

autoradiography. The relative levels of a39 were essentially identical to those determined by ADP ribosylation. Data in Fig. 10b summarize the mean f S.E. of densitometry scans from 6 experiments carried out as described in Fig. loa. Compared to control cells, a39 increased 2.1 f 0.4-fold (S.E., n = 5) in cells grown with LPDS and decreased to 0.5 f 0.1- fold of control in cells grown with LPDS plus 30 ptM mevinolin. In each case, measurements by immunoblotting and ADP ribosylation were carried out on protein from homogenates from the same culture.

No Change in the Level of the fi Subunit of No and Ni Demonstrated by Immunoblotting in Cells Grown in Medium Supplemented with LPDS-In order to study the effects of growth of cells with LPDS on the relative levels of the subunit of No and Ni aliquots of cell homogenates used for studies of pertussis toxin substrates and immunoblotting of as9 were subjected to polyacrylamide gel electrophoresis in parallel with an aliquot of purified bovine p subunit. Electro- transfer to nitrocellulose and incubation with antibody to bovine /3 followed by incubation with 1251-protein A and autoradiography demonstrated a single band migrating with standard bovine p subunit. Data summarized in Fig. 11 demonstrate that growth of cells with LPDS had no effect on the relative levels of p. These data are similar to 3 other experiments and demonstrate that although growth of cells in LPDS increases muscarinic receptor number and the a subunits of No and Ni, it has no effect on the level of the p subunit.

DISCUSSION

The data presented demonstrate that under conditions in which cellular cholesterol is increased, the responsiveness of cultured chick atrial cells to muscarinic cholinergic stimulation is significantly increased. This increase in physiologic responsiveness is associated with an increase in the total number of muscarinic receptors. Even more striking is the

15616

a -

” ”

Cholesterol and Muscarinic Responsiveness in Heart

F L L

M + F L L

+ M

FIG. 10. Data from typical experiments demonstrating an increase in levels of a39 in cells grown in media supplemented with LPDS studied by immunoblotting (a); histogram of levels of ‘2nI-protein A binding to immunoblots of a 3 0 ( b ) . a, proteins were prepared, solubilized and immunoblotting carried out as described under “Experimental Procedures.” Equal amounts of protein (100 pg) were loaded onto each lane. F, FCS; L, LPDS; M, 30 p~ mevinolin. Data were typical of 5 other experiments. Arrow indicates position of migration of standard ass from bovine brain obtained from Dr. E. d. Neer. b, autoradiographs similar to that in a were scanned by an LKB laser densitometer, the peaks cut out, and relative weights determined. Data were normalized to the value of an9 in cells grown with FCS taken as 1. Data are the mean of 6 determinations f S.E.

- ”_ -+

FIG. 11. Lack of effect of growth of cells in media supplemented with LPDS on the @ subunit of N studied by immunoblotting. Proteins were prepared, solubilized, and polyacrylamide gel electrophoresis, immunoblotting, and autoradiography carried out as descrihed under “Experimental Procedures.” Equal amounts of protein, 100 pg, were in each lane. F, FCS: L, LPDS; p, purified bovine 1-3 subunit obtained from Dr. E. J. Neer. This experiment is typical of 3 similar determinations.

finding that both the number of high affinity muscarinic receptors and the relative levels of the a subunits of No and Ni are increased 2-fold above control under conditions of elevated cholesterol. The relationship of these effects to changes in cellular cholesterol is supported by the finding that mevinolin, an inhibitor of endogenous cholesterol synthesis, reverses increases in physiologic responsiveness, levels of muscarinic receptors, and the a subunits of No and Ni in

cells grown in media with LPDS. The specificity of these effects of mevinolin is further supported by the finding that mevalonic acid, a small molecule whose only known function is to serve as a precursor in the biosynthesis pathway of cholesterol, reverses the effects of mevinolin on physiologic responsiveness as well as receptor number and levels of a39 and aCll (data not shown).

The work of Renaud et al. (4) demonstrated that growth of cardiac cells in LPDS resulted in an increased sensitivity of beating to muscarinic agonists. The cells used in those studies (4) appeared to be predominantly ventricular. In our own experiments, the effects of growth in LPDS are seen princi- pally in cells cultured from atria. Ventricular cultures are only slightly responsive to carbamylcholine, showing a 15% decrease in beating rate at doses as high as M carbamylcholine irrespective of the type of serum used. These findings are in agreement with the work of Josephson and Sperelakis (38), who demonstrated that aggregates of chick ventricular cells gave neither a negative chronotropic response nor an increase in K’ permeability in response to muscarinic stimulation.

One interpretation of these effects of growth of cells in media containing LPDS is that during the course of the dialysis step for removal of KBr used to adjust the density of serum before floatation of the lipoprotein fraction, factors controlling the expression of muscarinic function are removed from the serum. However, when whole serum containing lipoproteins was mixed with KBr and subjected to dialysis, no increase in sensitivity of beating to carbamylcholine could be demonstrated in cells grown in media supplemented with this dialyzed serum. Hence, these effects appear to be specific for the removal of the lipoprotein fraction. Furthermore, the finding that reconstitution of LPDS with a dialyzed LDL fraction reverses the effects of growth of cells in LPDS on sensitivity of beating to carbamylcholine suggests that it is the removal of LDL from serum which is responsible for the enhanced muscarinic responsiveness in cells grown in medium supplemented with LPDS.

Basu et al. (25) demonstrated that acetylation of the amino groups in LDL resulted in loss of the ability of LDL to compete with Iz5I-LDL for binding to LDL receptors. Fur- thermore, acetyl Iz5I-LDL was not taken up or degraded by fibroblasts. The finding that reconstitution of LPDS with acetyl-LDL has no effect on sensitivity of beating to muscarinic agonists also supports the conclusion that the effect of LPDS on muscarinic responsiveness is specific for LDL. These data further suggest that the effect of LDL on muscarinic responsiveness requires binding of LDL to the LDL receptor.

Comparison of the concentrations of LDL cholesterol in reconstitution studies reported here with studies of levels of LDL cholesterol in human fibroblasts also support the conclusion that the effect of LDL on the muscarinic response requires binding to the LDL receptor. Previous work with human fibroblasts has demonstrated that half-maximal accumulation of cholesterol in the cell occurred at an LDL- cholesterol concentration in the range of 30 pg/ml (35). This value is similar to the value for half-maximal binding of LDL to its receptor, 10-15 pg/ml LDL protein, which is equivalent to 13-20 pg/ml LDL-cholesterol (36). In the reconstitution studies reported here the final concentration of LDL-cholesterol in the growth medium was 29 pg/ml. Although data are not directly comparable with the studies in fibroblasts, these data suggest that levels of LDL in the reconstituted medium should be capable of saturating sufficient LDL receptor sites to alter cellular cholesterol content.

Beating rate in cardiac tissue is a complex function of


membrane potential, ionic conductances, and specific factors which modulate those conductances. Hence, any attempt to correlate a specific change in membrane function with a change in beating rate is difficult. However, the data presented here relate changes in the response of beating rate to muscarinic stimulation to changes in cholesterol levels. The increase in physiologic responsiveness and muscarinic receptor number in cells grown with LPDS was reversed by mevinolin over a very narrow concentration range (10-30 pM). Mevinolin mediated a 10-fold decrease in sensitivity of beating to muscarinic agonists, a 38% decrease in total cell cholesterol and a 28% decrease in the number of muscarinic receptors. Mev- inolin affected all three of these values with an IC50 for mevinolin in the range of 14-19 p ~ . Furthermore, each of these effects of mevinolin was reversed by mevalonic acid in a dose-dependent manner. Since mevalonic acid is a small molecule with no known effect on membrane function, it is highly likely that these effects are all related in some manner to cholesterol content.

Since beating is a nonlinear function of changes in membrane conductances and membrane potential, a small change in these parameters could lead to a substantial change in beating rate response. Although the steepness of the concentration-effect curve for the mevinolin effect on beating might be consistent with such an explanation, the steepness of the concentration-effect curve for all three parameters, cholesterol, receptor number, and physiologic response, might also be explained by comparing it to the concentration-effect curve for the mevinolin induction of HMG-CoA reductase activity. At 10 p~ mevinolin, the cell has reached the limit of its capacity to increase HMG-CoA reductase activity. At concentrations greater than 10 WM mevinolin, beating rate, total cholesterol, and muscarinic receptor number all begin to decrease over a narrow range of mevinolin concentrations. Hence, at the point where HMG-GOA reductase can no longer continue to increase, cholesterol synthesis and cholesterol levels begin to fall off rapidly and with them the muscarinic responsiveness of the atrial cultures. These data support the conclusion that the steepness of the concentration-effect curve to mevinolin reflects the inhibition of HMG-CoA reductase activity beyond a critical level at which cellular cholesterol synthesis cannot keep pace with cholesterol turnover.

Cholesterol levels have been shown to modulate membrane fluidity. Thus, a decrease in the mobility of fluorescent probes can be demonstrated in cholesterol-rich blood cell membranes (39). Patients with abetalipoproteinemia demonstrate a decreased fluidity of the red cell membrane associated with an increased cholesterol content (40). It has been shown that the interaction of guanine nucleotide regulatory protein with the low affinity form of the muscarinic receptor mediates the conversion of the low affinity form of the receptor to a high affinity form (17). One possible explanation for the increase in the number of high affinity receptors associated with an increase in cholesterol levels in cultured chick atrial cells reported here might be an enhancement or stabilization of the receptor-regulatory protein interaction associated with a decrease in membrane fluidity. A change in membrane fluidity could also affect the rate of membrane turnover and hence the steady-state level of muscarinic receptors. Finally, the availability of cholesterol precursors has been shown to influ- ence the levels of ubiquinone, dolichol phosphate, isopentenyl tRNA, and steroid hormones (41). Hence, effects of the availability of cholesterol precursors on protein synthesis, glyco- sylation of receptor proteins (42), and energy metabolism could also play a role in modulation of muscarinic receptor

number and the responsiveness of the heart to muscarinic stimulation.

Several lines of evidence suggest that the increase in muscarinic responsiveness observed in cells grown in the presence of LPDS was not due only to enhanced interaction of muscarinic receptors and guanine nucleotide regulatory proteins in a cholesterol-rich membrane. Although binding of [3H]- QNB to muscarinic receptors could be affected by differences in cellular cholesterol, we have presented data demonstrating that the KO for binding of [3H]QNB to atrial cell membranes is unaffected by growth of cells with LPDS. Hence, increased levels of [3H]QNB binding in membranes of cells grown with LPDS-containing media cannot be due to alterations in the binding of properties of [3H]QNB. These data do not rule out the possibility that growth of cells in LPDS results in increased accessibility of sites for [3H]QNB binding.

The increase in the levels of and measured by both ADP ribosylation and the a39 increase measured by immunoblotting of a39 suggest that growth with LPDS results in an increase in the levels of these proteins, rather than increased accessibility of a39 and a41 to the receptor and/or the effector. This is supported by our findings that: 1) addition of cholate to the ADP ribosylation reaction did not alter the levels of ADP ribosylation of CY,, and ai in cells grown with LPDS; 2) increased levels of pertussis toxin and/or more prolonged incubation of homogenates with pertussis toxin had no effect on levels of ADP ribosylation; 3) although the p subunit of No and Ni has been shown to be necessary for ADP ribosylation of a. and ai by pertussis toxin (43), we have demonstrated that growth with LPDS did not alter levels of 0; and 4) the levels of as measured by immunoblotting, a method that is not affected by the localization of a39 or its interaction with the lipid environment demonstrates exactly the same changes in the relative levels of as demonstrated by ADP ribosylation. Furthermore, in studies reported here no cell fractionation which might lead to preferential loss of as9 and alll from homogenates of cells grown with FCS compared to cells grown with LPDS was carried out in the preparation of tissue for either pertussis toxin studies or immunoblotting.

One explanation for the 2-fold increase in the number of high affinity receptors is that growth of heart cells in media supplemented with LPDS results in the interaction of increased levels of both low affinity receptors and guanine nucleotide regulatory proteins. Alternatively, a 2-fold increase in the number of high affinity receptors could result from the interaction between newly appearing low affinity muscarinic receptors and preexisting guanine nucleotide regulatory proteins.

The finding that both the levels of high affinity muscarinic receptors and ( ~ 3 ~ and a41 increase 2-fold in cells grown with LPDS and the increase in total new [3H]QNB binding sites accounts for the increase in newly appearing high affinity binding sites is consistent with the view that new high affinity receptors are formed by the association of newly appearing low affinity muscarinic receptors with newly appearing guanine nucleotide regulatory proteins. Finally, such an interpretation suggests that the expression of muscarinic receptors and the a39 and a 4 1 proteins might be coordinately regulated in these cells.

Although to date only 1 isoform of a. has been demonstrated, recent studies have demonstrated that all or ai is actually a family of at least 3 proteins, all of which might be pertussis toxin substrates which migrate with a similar elec- trophoretic mobility on a one-dimensional polyacrylamide gel (44, 45). Hence, since the distribution of various ais in the


chick heart cells has not been established, any discussion of changes in levels of a 4 1 must be interpreted as a change in one or more members of a family of ai proteins.

Finally, the possibility that growth in media containing LPDS nonspecifically increases a host of membrane proteins is ruled out by the finding that levels of the P subunit of No and/or Ni are unaffected by growth with LPDS.

Recently, Parsons and Stiles (46) have presented data sug- gesting that down-regulation of adenosine receptors in rat fat cells was associated with an up-regulation of both P-adrener- gic receptors and as in the same system (46). Whether enhanced parasympathetic responsiveness of the chick atrial heart cell grown in the presence of LPDS is associated with reciprocal changes in sympathetic responsiveness, P-adrener- gic receptors, and N, in these cells is a subject of current interest in our laboratory.

Acknowledgments-We thank A. W. Alberts, M. Greenspan, and D. Clapham for discussion of the data; P. Davies and F. Sacks for assistance; S. Shamah and G. Saunders for technical assistance; and B. Zillman and M. Gillan for typing the manuscript. We are grateful to Dr. E. J. Neer for generously providing us with the antibody.

REFERENCES

2.

3.

4.

5.

6.

7.

8. 9.

10.

11.

12.

13.

14.

15.

1. Catt, K. J., Harwood, J. P., Aguilera, G., and DuFau, M. C. (1970)

. Rubaclava, B., and Rodbell, M. (1973) J. Biol. Chem. 248,3831- Nature 2 8 0 , 109-116

00""

Hasin, Y., Shimoni, Y., Stein, D., and Stein, Y. (1980) J. Mol. Cell. Cardiol. 12, 675-683

Renaud, J. F., Scanu, A. M., Kazazoglou, T., Lombert, A., Romey, G., and Lazdunski, M. (1982) Proc. Natl. Acad. Sci. U. S. A.

Stein, O., and Stein, Y. (1973) Biochim. Biophys. Acta 326 , 232-

Goldstein, J. L., and Brown, M. D. (1977) Annu. Rev. Biochem.

Alberts, A. W., Chen, J., Kuron, G., Hunt, V., Huff, J., Hoffman, C., Rothrock, J., Lopez, M., Joshua, H., Harris, E., Patchett, A., Monaghan, R., Currie, S., Stapley, E., Albers-Schonberg, G., Hensens, O., Hirshfield, J., Hoogsteen, K., Liesch, J., and Springer, J. (1980) Proc. Natl. Acad. Sci. U. S. A. 7 7 , 3957- 3961

303 I

79 , 7768-7772

244

46,497-530

Gilman, A. G. (1984) Cell 3 6 , 577-579 Logothetis, D. E., Kurachi, Y., Galper, J., Neer, E. J., Clapham,

D. E. (1987) Nature 325 , 321-326 Pfaffinger, P., Hunter, M. J., Nathanson, N., and Hills, B. (1985)

Nature 3 1 7 , 536-538 Manning, D. R., and Gilrnan, A. G. (1983) J. Biol. Chem. 258 ,

Hildebrandt, J. D., Codina, J., Risinger, R., and Birnbaumer, L.

Kurose, H., Katada, T., Amano, T., and Ui, M. (1983) J. Biol.

Rosenberger, L. B., Roeske, W. R., and Yamamura, H. J. (1979)

Berrie, C. P., Birdsall, N. J. M., Burgen, A. S. V., and Hulme, E.

7059-7063

(1984) J. Biol. Chem. 2 5 9 , 2039-2042

Chem. 258,4870-4875

Eur. J . Pharmacol. 5 6 , 179-180

C. (1979) Biochem. Biophys. Res. Commun. 8 7 , 1000-1005 16. Galper, J. B., Dziekan, L. C., O'Hara, D. S., and Smith, T. W.

17. Florio, V. A., and Sternweiss, P. C. (1985) J. Biol. Chem. 2 6 0 ,

18. Kurose, H., Katada, T., Amano, T., and Ui, M. (1983) J. Biol.

19. Galper, J. B., Dziekan, L. C., and Smith, T. W. (1984) J. Biol.

20. DeHaan, R. L. (1967) Deu. Biol. 16 , 216-249 21. Hamberger, A., and Hamilton, H. L. (1945) J. Morphol. 10,861-

22. Beidert, S. W., Barry, W. H., and Smith, T. W. (1979) J. Gen.

23. Galper, J. B., Dziekan, L. C., Miura, D. S., and Smith, T. W.

24. Havel, R. J., Eder, H. A., and Bragdon, J. H. (1955) J. Clin.

25. Basu, S. K., Goldstein, T. L., Anderson, R. G. W., and Brown,

26. Heider, J. G., and Boyett, R. L. (1978) J. Lipid Res. 19, 514-518 27. Siedel, J., Rollinger, W., Roshclaa, P., and Ziegenhorn, J. (1985)

in Methods of Enzymatic Analysis (Bergmeyer, H, ed) Vol. 8, Sect. 1.4, pp. 139-148, VCH Verlagsgesellschaft mbH, Wein- heim, West Germany

28. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193 , 265-275

29. Liang, B. T., Hellmich, M. R., Neer, E. J., and Galper, J. B. (1986) J. Biol. Chem. 261,9011-9021

30. Laemmli, U. K. (1970) Nature 227 , 680-685 31. Towbin, H., Staehlin, T., and Gordon, J. (1979) Proc. Natl. Acad.

32. Galper, J . B., Haigh, L. S., Hart, A. C., O'Hara, D. S., and

33. Munson, J., and Rodbard, D. (1980) Anal. Biochem. 107 , 220-

34. Lebarca, C., and Paigen, K. (1980) Anal. Biochem. 402 , 344-352 35. Brown, M. S., Faust, J . R., and Goldstein, J. L. (1975) J. Clin.

36. Brown, M. S., and Goldstein, J. L. (1974) Proc. Natl. Acad. Sci.

37. Brown, M. S., Faust, J. R., Goldstein, J. L., Kaneko, I., and Endo,

38. Josephson, I., and Speralakis, N. (1982) J. Gen. Physiol. 79,69-

39. Ladbrook, B. D., Williams, R. M., and Chapman, D. (1968)

40. Cooper, R. A., Durocher, J. R., and Cestie, M. H. (1977) J. Clin.

41. Brown, M. S., and Goldstein, J. L. (1980) J. Lipid Res. 2 1 , 505-

42. Lennarz, W. (1983) Methods Enzymol. 9 8 , 1191-1197 43. Neer, E. J., Lok, J. M., and Wolf, L. G. (1984) J. Biol. Chem.

44. Bray, P., Carter, A,, Guo, V., Puckett, C., Kamholz, J., Spiegel, A., and Nirenberg, M. (1987) Proc. Natl. Acad. Sci. U. S. A.

45. Jones, D. T., and Reed, R. R. (1987) J. Bid. Chem. 2 6 2 , 14241-

46. Parsons, W. J., and Stiles, G. L. (1987) J. Biol. Chem. 2 6 2 , 841-

(1984) J. Bi01. Chem. 2 5 7 , 10344-10356

3477-3483

Chem. 258,4870-4875

Chem. 2 5 9 , 7382-7390

867

Physiol. 74,479-494

(1982) J. Gen. Physiol. 80, 231-256

Invest. 34,1346-1353

M. S. (1976) Proc. Natl. Acad. Sci. U. S. A. 73,3178-3182

Sci. U. S. A. 76 , 4350-4353

Livingston, D. J. (1987) Mol. Pharmacol. 3 2 , 230-240

239

Invest. 55,783-791

U. S. A. 71 , 788-792

A. (1978) J. Biol. Chem. 253 , 1121-1128

86

Biochim. Biophys. Acta 150,333-340

Invest. 60,115-121

517

2 5 9 , 14222-14229

84,5115-5119

14249

84 7

the journal of chemistry vol. 263, no. 30, issue of 25, pp ... · vol. 263, no. 30, issue of...

Documents