Inhibition by Glucagon of the Calcium Pump in Liver Plasma Membranes

3 downloads 0 Views 2MB Size Report
Jul 10, 2016 - Sophie LotersztajnS, Richard M. Epandg, Ariane Mallat$, and Franpoise. Peeker$ ... structure-activity relationships of six glucagon deriv-.
Vol. 259, No. 13,Issue of July 10, pp. 8195-8201,1984 Printed in U.S.A.

THEJOURNALOF BIOLOGICAL CHEMISTRY

0 1984 by The American Society of Biological Chemists, Inc.

Inhibition by Glucagon of the Calcium Pumpin Liver Plasma Membranes* (Received for publication, August 15, 1983)

Sophie LotersztajnS, RichardM. Epandg, Ariane Mallat$, and Franpoise Peeker$ From the $Unite Znstitut National de la Sante et de la Recherche Midicab 99, Hhpital Henri Mondor, 94010 Creteil, France and the $Department of Biochemistry, McMaster Uniuersity, Health Sciences Centre, Hamilton, Ontario L8N3Z5,Canada

TheATP-dependentcalciumtransport in plasma related phosphorylated intermediate (20-23). To date, there membrane vesicles prepared fromrat liver was inhib- have only been two preliminary reportson the hormonal ited by 0.1 to 10 IM glucagon. Inhibition of the high sensitivity of the liver plasma membrane (Ca2+-Mg2')-ATaffinity (Ca2+-Mg2+)-ATPase was observedconcomi- Pase. A 13 to15% inhibitionof (Ca2+-Mg2')-ATPaseactivity tantly. This effect was neither mimickedby cyclic was observed in isolated plasma membranes prepared from AMP norbydibutyryl cyclic AMP. A studyofthe isolated hepatocytes which have been incubated with vasostructure-activity relationships of six glucagon deriv- pressin (24, 25) and a 19% activation was reported in mematives demonstrated the specificity of glucagon action branesobtained from insulin-treated hepatocytes (24). It since only one or two analogs markedly altered the should be noted that, until now, no direct hormonal effect of (Ca2+-Mg2+)-ATPase activity. The study also demon- the (Ca2+-Mg2+)-ATPasefrom isolated plasma membranes strated the total absence of correlation between ade- could be demonstrated. Glucagon, although it was not the nylate cyclase activation and (Ca2+-Mg2+)-ATPase inmost effective, is one of the hormones able to alter the calcium hibition inducedby these glucagonderivatives. The decrease in the maximal velocities induced by distribution in liver cells (6). It was therefore tempting to glucagon of both calcium transport and (Ca2+-Mg2+)- examine its possible interaction with the plasma membrane ATPase activity were related toa reduction in the rate (Ca2+-Mg2')-ATPase. of dephosphorylation of the Ca-dependent phosphoryl- In the present study, we report a specific inhibition of the ated intermediate ofthe enzyme. This phosphorylated ATP-dependent calcium transport and of the (Ca2+-Mg2+)intermediate was characterized as a 32P-labeled ATPase of isolated rat liver plasma membranes caused by the 110,000-dalton protein which accumulated to 50 to direct interaction with glucagon. Both transport and enzyme 150%over the basallevel in the presence of glucagon. inhibition occur concomitantly with an increase in the level The present results demonstrate a novel aspect of the of phosphorylated intermediate of the enzyme ascribable to a role of glucagon as a calcium-mobilizing agent. decrease in its dephosphorylation rate. Structure-activity relationships of six glucagon derivatives suggest that inhibition of (Ca2+-Mg2')-ATPaseand activation of adenylate cyclase induced by glucagon are two independent processes. This is Changes in free cytosolic Ca2+concentration have been reinforced by the lack of effect of cyclic AMP on ( C a 2 + - M e ) invoked as the mediator of the action of certain hormones ATPase activity. and other external stimuli on metabolic responses (1-7). A EXPERIMENTALPROCEDURES second messenger role is thus conferred on free cytosolic Ca2+ which is parallel to, or synergic with, that of cyclic AMP (8Animals-Female albino Wistar rats, weighing 100 to 150 g were 13). (Ca2+-Mg2+)-ATPases inplasma membranes, which are obtained from Lessieux (Bray-Lu, France). generally accepted as the enzymatic basis for calcium extruMaterials-Nucleotides (Tris anddisodium salts) and EGTA' were sion pumps, are one of the key systems involved in the control obtained from Sigma. 45Ca (10 to 40 mCi/mg), [Y-~'P]ATP(10 Ci/ of the level of free cytosolic Ca2+.Several recent reports have mmol), and 14C-methylatedmolecular weight markers were purchased Amersham/Searle. Highly purified porcine crystalline glucagon pointed out their role as targets of hormonal action: it has from and insulin were obtained from Novo Laboratories and glucagon been shown that direct additionof insulin inhibitedthe (Ca2+- derivatives were prepared according to Ref. 26 or 27. Other chemicals Mf)-ATPase activity in isolated adipocyte plasma mem- were from Merck (Darmstadt, Germany). branes (14); in human erythrocyte, the enzyme was stimulated Calcium Uptake-For calcium uptake experiments, rat liver plasma in uitro by thyroid hormones (15,16) and in myometrium and membrane vesicles were prepared according to Pilkis et al. (28). The fat pad it was inhibited by oxytocin (17, 18). Also, 1-(~,25- plasma membrane vesicles resuspended in 50 mM Tris-HC1, pH 8.0, 0.25 M sucrose were stored in liquid nitrogen until use.Calcium dihydroxyvitamin D3 increased the (Ca2+-MgZ+)-ATPaseof uptake was assayed according to the procedure of Chan and Junger rat enterocyte basolateral plasma membranes (19). (22) in a 500-pl assay medium containing 50 mM Tris-HC1, pH 8.0, The (Ca2+-Mg2')-ATPase/calciumpump system of rat liver at 37 T , 0.25 M sucrose, 0.2 mM EGTA, 47 to 199 p~ CaC12(1 nM to plasma membranes has been characterized, as well asits 0.25 p~ free Ca*'), 2 pCi/ml of 45Ca, 10 mM MgC12,0.01% bovine * This work was supported by the Institut National de la Santi. et de la Recherche Midicale, the DelCgation Genbrale a la Recherche Scientifique et Technique, the University of Paris-Val de Marne, and the National Science and Engineering Research Council of Canada. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked ''aduertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

serum albumin, 4 mM potassium oxalate, 20 mM sodium azide, and with either no ATP added or 10 mM ATP. The mixture was preincubated for 5 min at 37 "C. The reaction was started with the addition of 200 to 400pg of plasma membrane proteins which had been preincubated for 5 min a t 37 "C with glucagon at the concentrations

' The abbreviations used are: EGTA, ethylene glycol bis(P-aminoethyl ether)-N,N,N',N'-tetraacetic acid; SDS, sodium dodecyl sulfate.

8195

Inhibition of the Liver Calcium Pump by Glucagon

8196

indicated in Fig. 1. After 15 min at 37 "C, 450-pl samples were removed and vacuum-filteredon 0.5-pm pore Millipore cellulose acetate filters that had been soaked in 0.25 M KC1 for at least 30 min and rinsed twice with 5 ml of 0.25 M sucrose contianing 40 mM NaCI. The membrane vesicles trapped on the filters were washed three times with 5 ml of 0.25 M sucrose containing 40 mM NaCl. The filters were dried and placed in scintillation vials with 10 ml of Beckman Ready-solv. The ATP-dependent uptake was determined from the difference in radioactivity bound to the filter in the presence and absence of ATP. All uptake values were corrected for nonspecific radioactivity bound to the filter when an identical reaction mixture without plasma membranes was filtered. This usually accounted for approximately 20% of the total measured radioactivity. Results were obtained from triplicate determinations. (CaZ+-Mg2')-ATPase Assay-(Ca'+-Mg?-ATPase activity was quantitated as previously described using purified rat liver plasma membranes, prepared according to the procedure of Neville (29) up to step 11. For the reasons previously discussed (20, 21), no external magnesium was added tothe assay medium. The assay medium contained 0.25 mM ATP, 50 mM Tris-HCI, pH 8,0.01%bovine serum albumin, 400 pM EGTA, with either no CaC12 added or with 150 to 399 p M total CaCl2 (1nM to 1p M free Ca2+),3 to 7 pg of protein, and drugs when indicated in a final volume of250 pl. After 10 min a t 30 "C, aliquots were assayed for Pi by a colorimetric determination using malachite green according to themethod of Kallner (30). (Ca"M F ) - A T P a s e activity was calculated by subtracting values obtained in the presence of chelator alone from those obtained with chelator plus Ca2+. Resultswere obtained from triplicate determinations. Phosphorylation Assay-Rat liver plasmamembranesprepared according to the procedure of Neville up to step 11 (29) were used. Phosphorylation was performed (in 1.5-ml polypropylene microcentrifuge tubes) in a 100-plreaction mixture of 50 mM imidazole-HC1, pH 6.9, containing 0.01% serum albumin, 100 N M ATP, 100 mM KCI, 400 p~ EGTA, with either no CaCIZadded or with 393 p~ total CaCh (0.1p~ free Ca2+),glucagon when indicated, and with no exogenously added magnesium. After a 5-min preincubation at 30 "C, initiated by the addition of 80 pg of membrane proteins, 10 p l of [y-32P]ATP(0.4 Ci/mmol, final specific activity) were added to the assay medium and the phosphorylation assay was performed for an additional period of 10 s at 30 "C. The reaction was terminated by the addition of 1 ml of ice-cold 125 mM HClO,, 2 mM H3P04, 25 mM ATP. The reaction mixture was centrifuged at 4 "C for 2 min in a Beckman microfuge a t 10,000 X g and the resulting pellet was washed three times with 1 ml of ice-cold 125 mM HClO,, 2 mM H3P04and twice with 1 ml of icecold HZO. The final pellet was resuspended in 25 p1 of Hz0 and solubilized in 75 p1 of sample buffer for polyacrylamide gel electrophoresis (see below). Then, assays were either subjected to electrophoresis or 32Pbound to the proteins was quantitated by counting in 10 mlof Beckman Ready-solv. Ca-dependent phosphorylation was calculated by subtracting values obtained in the presence of chelator alone from those obtained with chelator plus CaCl2. For hydroxylamine treatment, the acid-washed precipitates were resuspended in 300 pl of freshly prepared 300 mM hydroxylamine, 150 mM sodium acetate, pH 6. The controls were treated with 300 mM Tris-C1, 150 mM sodium acetate, pH 6. After a 10-min incubation at room temperature, thereaction was terminated by the addition of 600 pl of icecold 125 mM HC104, 2 mM H3PO4. The precipitates were processed as described above. Polyacrylamide Gel Ekctr0phresi.s and Autoradiography-SDSpolyacrylamide gel electrophoresis was done as described by Niggli et al. (31). The sample buffer was 100 mM sodium phosphate, pH 6, containing 1%SDS, 10% glycerol, 20 mM dithiothreitol, and 0.002% bromphenol blue. The gel contained 5% (w/v) acrylamide, 0.13% (w/ v) N,N'-methylenebisacrylamide.Electrophoresis was performed at 15 "C,at a constant current of 40 mA. Following electrophoresis, the gels were soaked for 5 min in 10% (v/v) acetic acid, 10% (v/v) isopropyl alcohol, dried, and exposed to KodakXAR-5 film witha DuPont Cronexintensifierscreen at -70 "C. Protein was determined by the method of Lowry et al. (32). Determination of total and free calcium concentrations of reaction mixtures were done as previously reported (20). RESULTS

Glucagon Inhibition of the ATP-dependent Calcium Uptake in Liver Plasma Membrane Vesicles-Optimal conditions for assaying the ATP-dependentcalcium uptake in isolated liver

TABLEI Inhibitory effect of glucagon on ATP-dependent calcium uptake by liver plnsma membrane vesicles at different freecalcium concentrations The ATP-dependent calcium uptake was measured as described under "Experimental Procedures." Glucagon was added to a final concentration of 6.7 p~ and totalCa2+ concentrationwas varied from 47 to 199 vM. ATP-dependent calcium transport

Free Caz+

Control

+Glucagon

9% inhibition ~

nmol/mg protein/l5 min

10-9 M

lo-'

M M

0.50

2.5 x 10-7 M

0 0.50 1.25 1.25

0 0.65 0.55

0 48 56

plasma membrane vesicles prepared according to Pilkis et al. (28) have been precisely defined by Chan and Junger (22). Using these conditions, we have measured the ATP-dependent calcium uptake as a function of free calciumconcentration in the reactionmedium. The rateof calcium uptake exhibited saturablekineticsreaching a maximum at afreecalcium concentration of approximately 0.1 p ~ Direct . additionof 6.5 pM glucagon to the plasma membrane vesicles resulted in a 55% decrease inmaximal calcium uptake(TableI).The inhibitory effect of glucagon was concentration-dependent, half-maximal inhibition occurring in the presence of 2.5 ~ L M glucagon (Fig. 1). Glucagon Inhibition of (Ca2+-Mg2+)-ATPaseActivity-In a previous report (20), we have characterized a ( C a 2 + - M P ) ATPase in rat liver plasma membranes. The kinetic properties of this enzyme suggested that it was closely related to ATPdependent calcium transport. In the present study,we tested the action of glucagon on the activity of this enzyme under optimal conditions,namely in the absence of exogenously added magnesium and using liver plasma membranes prepared according to Neville (29).' Addition of 10 pM glucagon to liver plasma membranes resulted in a 25 to 35% maximal decrease in (Ca2+-Mg2+)-ATPase activity, with half-maximal inhibition occurring in thepresence of 0.7 -+ 0.3 pM glucagon (Fig. 2). Kinetic studies of (Ca2+-Mg2')-ATPaseactivity in the presence and absence of glucagon showed that the hormone did not affect the apparent affinity of the enzyme for either the substrate ATP (not shown),or Ca2+, but caused a decrease in the maximal velocity of the reaction (Table 11). Since cyclic AMP iswell established as a mediator of glucagon action, we tested its effect on (Ca2+-Mg2')-ATPase activity. As shown in Fig. 2, neither cyclic AMP nor dibutyryl cyclic It should be noted that optimal expression of the ATP-dependent Ca uptake activity in vesicles required millimolar concentrations of magnesium. In contrast, exogenous magnesium has to be eliminated from the medium for the assay of (Ca2+-Mg2+)-ATPaseactivity in order to lower the extremely high background of Mg2+-ATPaseactivity in these membranes, which hinders measurement of the (Ca2+M e ) - A T P a s e activity. Such a discrepancy between assay conditions for Ca uptake and (Ca'+-M%+)-ATPase activities was not particular to the liver system but was also found in adipocyte (33), macrophage (34), spermatozoa (35), lymphocyte (36), duodenum (37), and pancreatic islet cells (38). The maximal velocities of the Ca uptake and (Ca'+-MF)-ATPase activities (1.25 nmol of Ca/mg/l5 min of 0.75 pmol of Pi/mg/lO min, respectively) yield a very low ratio of Ca transported per ATP hydrolyzed. Addition of millimolar concentrations of magnesium to theCa uptake assay,which renders the calcium pump system sensitive to its Mg-dependent inhibitor (21), the leakiness of the vesicles to Ca, the presence of right-side-out vesicles which do not accumulate calcium, and the inefficiency of membrane recovery during filtration could all contribute to a significant underestimation of calcium translocation.

Inhibition of the Liver Calcium Pump by Glucagon

8197

AMP (1nM to 1 mM) affected (Ca2+-Mg2')-ATPaseactivity, nor did addition to the assay medium of 10 mM theophylline, a potent inhibitor of phosphodiesterase, potentiate theeffect of glucagon. The specificity of the glucagon effect was indicated by the fact that boiled glucagon and vasoactive intestinal peptide, which is structurally related to glucagon, had no effect on (Ca2+-M$+)-ATPase activity when added at concentrations varying from 1nM to 10 p~ (not shown). We havealso tested the effect of a number of glucagon analogs on (Ca2+-M$+)ATPase. These glucagon derivatives have been used to identify the structural features of glucagon which are important for the stimulation of liver adenylate cyclase (26, 27). Two derivatives were potentagonistsininhibitingthe(Ca2+M$+)-ATPase (Fig. 3).Carbamoyl glucagon induced 20% inhibition at 10 p~ while glycinamide glucagon caused a r I maximal 35% inhibition at the same concentration (Fig. 3). 7 6 5 However, the glycinamide glucagon had only about half of the GLUCAGON (-log M) potency of glucagon. It is surprising that this derivative reeven FIG. 1. Effect of glucagon on the ATP-dependent calcium tains so much activity in inhibiting (Ca2+-Mg2')-ATPase uptake by liver plasma membrane vesicles. The ATP-dependent though four of the amino acid residues of glucagon are modiof four negative charges calcium uptake was measured as described under "Experimental fied with the resulting disappearance Procedures" in the presence of 199 p M total CaC&(0.25 pM free Ca2+). at neutral pH. In contrast, guanidyl glucagon which has the same charge as the native hormone at neutral pH and has only one group modified shows no activity. Arginine-substituted glucagon, Na-trinitrophenyl glucagon, and thecyanogen bromide glucagon also had no agonist effect on (Ca2+-Mg2+)ATPase (Fig. 3). We therefore examined the ability of the inactive derivatives to antagonize the inhibition induced by glucagon. Arginine-substituted glucagon andNa-trinitro) ineffective in blocking phenyl glucagon (1nM to 10 p ~ were the glucagon action (not shown). In contrast,10 p~ guanidyl or cyanogen bromide derivatives totally suppressed the inhibition of (Ca*+-MC)-ATPaseby glucagon (Fig. 4), while at 5 p~ these derivatives cause a 3- to 4-fold shift to the right in the glucagon inhibition curve. The potencies of glucagon derivatives for inhibiting (Caz+M$+)-ATPase and activating adenylatecyclase are reported in Table 111. It appears that arginine-substituted, guanidyl, and cyanogen bromide glucagons significantly stimulated ad60 enylate cyclase but were ineffective in inhibiting(Ca*+-M$+), IllATPase. In contrast,while glycinamide glucagon had little or 9 8 7 6 5 3 no effect on adenylate cyclase (Table 111; Refs. 27, 39, and GLUCAGON OR NUCLEOTIDE 40), it was the most potent derivative in inhibiting (Ca2+MgVATPase. Itshould also benoted that N"-trinitrophenyl (-log M) FIG. 2. Effect of glucagon (m) and cyclic AMP (0)or dibu- glucagon which behaves as a glucagon antagonist in the adetyryl cyclic AMP (A) on the liver plasma membrane (Ca2+- nylate cyclase system (26)did not affect the inhibition of MgZ+)-ATPase.Assays were run as described under "Experimental (Ca*+-M$+)-ATPase by glucagon. These data demonstrate Procedures" in the presence of 393 p M total CaClz (0.1 p~ free Ca"). of glucagon are required for the Results are expressed as per cent of the control activity (1.2 -t 0.2 thatdifferentproperties inhibition of (Ca2+-Mg2')-ATPase andfor the stimulation of pmol of P,/mg of protein/lO min). adenylate cyclase. Insulin Action-We have tested the effect of insulin added in vitro to the (Ca2+-Mg2')-ATPase incubation medium, at TABLE 11 concentrations ranging from 100 to 500 microunits ml-': no Inhibitory effect of glucagon on (Ca2+-M$+)-ATPase activity at effect of insulin on (Ca2+-M$+)-ATPase activity could be different free calcium concentrations detected. Wehave also looked for a possible role of insulin as (Ca*+-MgZ+)-ATPasewas measured as described under "Experi- an antagonist of glucagon action. However, insulin added in mental Procedures." Glucagon was added to a final concentration of vitro at concentrations ranging from 10 to 500 microunits 1 p M and total CaClz concentration was varied from 150 to 399 pM. ml" did not suppress the inhibition of (Ca2+-M$+)-ATPase (Ca"-MgZ')-ATPase activity induced by glucagon, at all concentrations of glucagon (0.1 to Free Ca2+ Control +Glucagon % inhibition 10 p ~ studied. ) p m l P,/rngprotein/lO rnin Lin et al. (24) have reported a 19% activationby insulin of thepresent enzyme. Infact,this effectwas observed by 10-9 M 0 0 lo-' M 0.25 0.25 0 measuring the (Ca"-M$+)-ATPase activity in plasma mem10-~M 0.75 0.45 40 branes obtained from hepatocytes which have been incubated 0.48 M 0.78 39 isolation for 3 min with 120microunits ml" of insulin prior to

1

I

Inhibition of Liver the

8198

Calcium Pump by Glucagon

1

CYANOGEN BROMIDE Na-TRINITROPHENYL GLUCAGON GUANIDYL ARG SUBSTITUTED

FIG. 3. Effects of glucagon derivatives on theliver plasma membrane (Ca2+-Mg2+)-ATPase. Activity was assayed as described under "Experimental Procedures" in the presence of 393 p~ total CaC& (0.1 p~ free CaZ+) and in the presence of varying concentrations of glycinamideglucagon (e), carbamoyl glucagon (O), and guanidyl glucagon, arginine-substituted glucagon, cyanogenbromideglucagon, or N*-trinitrophenyl-glucagon (A). For brevity and simplicity, A~p~*'~.*~-trigIycinamideglucagonyl-(glycinamide) is referred to as glycinamide glucagon; Ne-carbamoyl glucagon as carbamoylglucagon; (homoarginine") glucagon as guanidylglucagon; di(d-(5-nitro-2-pyrimidyl)orniglucagon as arginine-substituted glucagon, and (Des-AsnZ8,ThP') (homoserine lactonez7) glucagon as cyanogen bromide glucagon.

\

1

fi

-GLYCINAMIDEGLUCAGON

1

7

6

5

TABLE 111 Potencies of glucagon derivatives for stimulating adenylate cyclase and inhibiting CaA TPase Potencies relative to glucagonb Derivative" CaATPase

A

6o

i

5 P M GUANIDYL GLUCAGON

0 10 V M " " A 5 V M CYANOGEN

1

BROMIDE GLUCAGON

0 1OPM

I