Stimulation and Inhibition of cAMP Accumulation ... - Semantic Scholar

0 downloads 0 Views 910KB Size Report
Sep 15, 1986 - and (c) the glucagon and adrenergic systems involved ...... Garrity, M. J., Andreasen, T. J., Storm, D. R., and Robertson, R. P. (1983) J. Biol.
THEJOURNALOF BIOLOGICAL CHEMISTRY 01987 by The American Society for Biochemistry and Molecular Biology, Inc

Vol ,262. No. 32, Issue of November 15, pp. 15514-15520,1387 Printed in U.S.A.

Stimulation and Inhibitionof cAMP Accumulation by Glucagonin Canine Hepatocytes* (Received for publication, September 15, 1986)

Teresa Grady, Maria FickovaS, and HowardS.Tagers From the Department of Biochemistry ana‘ Molecular Biology, The University of Chicago, Chicago, Illinois 60637

Dev Trivedi and Victor J. Hruby From the Department of Chemistry, University of Arizona, Tucson, Arizona85721

We have examined, by use of isolated canine hepatocytes andselected hormone analogs,the mechanisms by which glucagon modifies the accumulation of cellular CAMP. Low concentrations of glucagon (53 nM) enhancedtheaccumulation of hepatocyte CAMP, whereas higher concentrationsof the hormone diminished the effectiveness of lower ones. This biphasic concentration dependence was observed as well for some glucagon analogs, but not for others, and was apparent for cells incubated in the presence or absence of theophylline. Glucagon at high concentrations (210 nM) also inhibited the accumulation of cAMP induced by isoproterenol. The inhibitory effectof glucagon in both of these systems was reversed or attenuated by cell incubations involving the use of pertussis toxin (islet-activating protein) or a peptide antagonistof the glucagon-adenylyl cyclase system. We conclude that (a)glucagon, through its interaction with high and low affinity binding sites, can either stimulate or inhibit the production of hepatocyte CAMP; ( b ) the inhibitory action of the hormone appears to arise from interactions of ligand witha subset of these binding sites and torequirestructuralcharacteristicsinadditionto per se; those that determine receptor binding affinity and (c) the glucagon and adrenergic systemsinvolved in stimulating cAMP accumulation are linked, at least with regard to the negative effect induced by high concentrations of glucagon.

those of glucagon (8-13), where the effect has been attributed to both decreased receptor number and either anuncoupling of receptor from the guanyl nucleotide regulatory protein N,’ (8) or a change in thecoupling of receptor to Ni (11-13), and arginine vasopressin, where the effect has been attributed to theactivation of a pertussistoxinsubstrate(14).Studies demonstrating theheterologous inhibition of adenylyl cyclase systems activated by glucagon, prostaglandin El, and @-adrenergic and muscarinic agonists have also been described (8, 13, 15, 16). In thesecases, changes in thecoupling of receptor to adenylyl cyclase via the guanyl nucleotide-binding proteins N, and Ni (rather than changes in receptor number) are most often regarded as the cause of inhibition of hormone-stimulated adenylyl cyclase. For several reasons, the potential inhibitoryeffects of glucagon on the adenylyl cyclase system identifies an area of biochemical and physiological importance. First, a variety of glucagon analogs have been shown to behave both as partial agonists and asglucagon antagonists with regard to stimulating the adenylyl cyclase of rat liver plasma membranes (1723). Second, although the high dose inhibition of adenylyl cyclase by glucagon analogs has been interpreted in terms of the multivalent bindingof ligand to receptor(201, alternative explanations for glucagon involving participation of Ni have also been proposed (10-12). Third, the effectiveness of glucagon analogs behaving as antagonists in membrane systems is not easily correlated to their potential actionsas glucagon antagonists in cells or in vivo (24, 25).* Fourth, whereas some investigators haveproposed the existence of only a single memThe homologous and heterologous modulation of ligand- class of glucagon receptorsonbothhepaticplasma branes (20, 26) and isolated hepatocytes (27), others have effector systems represents a fundamental process by which the existence of two populations of cells regulate their sensitivities to hormones and other sub- providedevidencefor stances. The homologous desensitization of the @-adrenergic glucagon receptors, eachwith the potential for contributing a receptor/adenylyl cyclase system by catecholamines hasbeen separate biological response (28-31). Fifth,ithasrecently well studied in a variety of systems (1-8) and hasbeen shown been proposed that glucagon might act by stimulating the as well as by stimulating adenylyl to involve, to varying degrees, both an uncoupling of the inositol phosphate pathway receptor from adenylyl cyclase and a decrease in the number cyclase (31). The accompanying manuscript (30) considers in detail the of available @-adrenergic receptorsdue to receptor phosphoof glucagon rylation and sequestration (2,9). Other examples of homolo- importance of ligand structure in the interactions gous inhibition of adenylyl cyclase-linked receptors include and glucagon analogs withhigh and low affinity binding sites of canine hepatocytes and the correlation between occupancy * This work was supported by Grants DK-18347 and DK-20595 (to of high affinity glucagon-binding sites and themodulation of H. S. T.) and DK-21085 (to V. J. H.) from the National Institutes of hepatocyte glycogen metabolism. To investigate further the Health. The costs of publication of this article were defrayed in part biological actions of glucagon, the potential for both positive by the payment of page charges. This article must therefore be hereby and negative effects of glucagon on hepatic adenylyl cyclase, marked “advertisement” in accordance with 18 U.S.C. Section 1734 and the potential forheterologous inhibition of the adenylyl solely to indicate this fact. ~

$ Present address: Inst. of Experimental Endocrinology, Brataslava, Czechoslovakia. 5 To whom correspondence should be addressed Dept. of Biochemistry and Molecular Biology, The University of Chicago, 920 E. 98th St., Chicago, IL 60637.

The abbreviations used are: N., stimulatory guanyl nucleotidebinding regulatory protein; Ni, inhibitory guanyl nucleotide-binding regulatory protein; Cbm, carbamoyl. * B. Gysin, D. Trivedi, and V. J. Hruby, unpublished results.

15514

Glucagon C A M Pby Modulation of Hepatoc:yte

15515

cyclase system by glucagon, we undertook to study the effects centrifuged (1500 X g) for 30 min at 4 "C. Aliquots of the resulting supernatants were diluted 30-fold with the acetate buffer described of glucagon and glucagon analogs on cAMP accumulationin below and subjected to acetylation with acetic anhydride/triethylaisolated canine hepatocytes, cells that can be prepared with mine (1:2, v/v). Aliquots of the resulting solution (usually 0.1 ml) high yield and viability, are well characterized with regard to were then subjected to radioimmunoassay for cAMP by the use of glucagon-receptor interactions (29, 30), and respond to /3- '251-labeledcAMP and preformed anti-CAMP antibody complexes adrenergic agonists without need for extended cell culture. purchased from Du Pont-New England Nuclear; instructions for the assay were provided by the manufacturer. The buffer (0.1 M sodium Peptides chosen for study include two that exhibit parallel and shallow concentration dependencesfor inhibition of lZ5I- acetate brought to pH 6.2 with acetic acid) and the carrier for sedimentation of the ligand-antibody complexes (preformed comlabeled glucagon binding to hepatocytes (glucagon and glu- plexes prepared from normal rabbit serum and goat anti-rabbit IgG) cagon cleaved by cyanogen bromide), three that exhibit par- were prepared in the laboratory. Control studies showed that the allel andsteepinhibition (des-His1-glucagon, [N"-trinitro- determined level of cAMP was independent of the length of time phenyl-His1,homo-Arg12]glucagon, and[~-Phe~,Tyr',diiodo- during which the trichloroacetic acid suspension of cell proteins were Tyr10,Arg12,Lys17~'s,G1u21]glucagon), and one that exhibits a allowed to stand at 4 "C or the freezing and thawing of the trichloroacetic acid supernatant. Our assay determines cAMP present in rather intermediate grade of inhibition ((Ne-Cbm-His',N" both cells and cell incubation medium. Additional controls (involving Cbm-Lys'2]glucagon) (30). Overall, glucagon-binding sites on the separate measurement of cAMP in cell incubation medium) isolated hepatocytes exhibit apparent dissociation constants demonstrated that extracellular cAMP accounts for a constant fracfor these peptides ranging from about 0.8 to 3000 nM. Our tion (14-18%) of total cAMP present in cell incubations and rises results show that (a) glucagon can either stimulate or inhibit and falls in parallel with total cAMP under all conditions tested. the accumulationof cellular cAMP depending on the concenRESULTS tration of the hormone used and on the resulting degree of saturation of high and low affinity binding sites, and ( b ) the As a control for the studyof the glucagon receptor/adenylyl inhibitory effect of glucagon at high concentrationsisas cyclase system, we initially examined the timecourse for the applicable to the isoproterenol-stimulated accumulation of stimulation by glucagon of cAMP accumulation in isolated cAMP as it is to the glucagon-stimulated accumulation of the canine hepatocytes. Fig. l a shows that the addition of glucacyclic nucleotide. gon to thecell incubation medium resulted in a time-dependent increase in the accumulation of CAMP, with steady-state EXPERIMENTALPROCEDURES levels of the cyclic nucleotide being achievedat approximately Materials-Glucagon was obtained from Lilly, and des-His'-glu- the same ratefor each of the hormone concentrations tested. cagon and porcine insulin were from Novo Pharmaceuticals (CopenSurprisingly, the steady-state levels of cAMP achieved by hagen, Denmark). Methods for the preparation and purification of cyanogen bromide-cleaved glucagon (331, [N*-Cbrn-His',N"Cbm- higher concentrations of the hormone were actually lower than those achieved by lower ones. The results of Fig. 2a Ly~'~]glucagon (17), and [N"-trinitrophenyl-His',h~mo-Arg~~]glucagon (17) have been described previously. [~-Phe~,Tyr~,diiodoconfirm those ofFig. l a and demonstrate more clearly the Tyr'o,Arg'2,Lys'7~'s,G1~21]Glucag~n was prepared by total solid-phase biphasic response of isolated hepatocytes to increasing conpeptide ~ynthesis.~ Theophylline, (-+)-propranolol,(-)-isoproterenol, centrations of glucagon; whereas the level of hepatocyte apamin, phorbol 12-myristate 13-acetate, and cholera toxin were all purchased from Sigma. Pertussis toxin (islet-activating protein) was cAMP increased as the concentrationof glucagon was increobtained from List Biological Laboratories (Campbell, CA). All other mentally increased from 0.01 to 3 nM, the level of cAMP decreased as the concentrationof the hormone was increased chemicals and biochemicals were obtained from standard suppliers. Cell Preparation and Incubation-Hepatocytes were isolated from from 10 to 1000 nM. Fig. 2 ( b and c) shows that the biphasic the livers of mongrel dogs and were incubated as described previously response of hepatocytes to glucagon was also apparent when (29, 32). Cell viability, as determined by the exclusion of the dye cells were incubated in the absenceof theophylline and valitrypan blue, always exceeded 97%. For most assays, cells were incubated at 30 "C for 30 min at a final concentration of 2 X 106/ml in 1 dates the use of the phosphodiesterase inhibitor to elevate ml of supplemented Krebs-Ringer bicarbonate buffer (32) containing hormones or other compounds at theconcentrations identified in the figure legends. Peptides were dissolved at concentrations of about 10 mg/ml in a small volume of 6 M urea prior to dilution into incubation buffer; control studies showed that the vehicle did not alter cAMP accumulation or glucagon association with receptors in any of the systems studied. For experiments involving determination of the effects of glucagon on the @-adrenergicsystem, cells were preincubated for 30 min at 30 "C with 1p~ isoproterenol prior to theaddition of peptides and theophylline. For experiments involving the use of cholera or pertussis toxin,cells were preincubated for 2 h at 37 "C in the presence of the protein toxin (10 pg/ml) and were then cooled to 30 "C prior to the addition of peptides and theophylline; pertussis toxin was dissolved in a small volume of 2 M urea prior to dilution and further use. During experiments involving a preincubation of cells in the presence of a potential effector, a second group of cells was always preincubated in parallel in buffer alone to serve as a control; the effector was presentthroughout the total incubation period. Final incubations occurred in duplicate or triplicate by use of 20-ml glass scintillation vials closed with rubber stoppers. Each experiment was repeated at least twice with similar results. Representative findings (mean -+ S.D.) are reported. cAMP Determinations-At the close of cell incubations, each I-ml aliquot of cells was diluted with an equal volume of 6% (w/v) trichloroacetic acid prepared in water. The resulting suspensions were allowed to stand overnight a t 4 "C, transferred to glass tubes, and

B. Gysin, D. Trevedi, D. G. Johnson, and V. J. Hruby, unpublished results.

0.2

0' 0

I

15

30

I

I

45

60

Incubation Period, rnin FIG. 1. Time courses describing the effects of glucagon on cAMP accumulation in isolated hepatocytes. a, hepatocytes were incubated in the absence of glucagon (0)or in the presence of the hormone a t concentrations corresponding to 3 nM (O),30 nM (O), or 300 nM (A) for the indicated periods prior to terminating the incubations by the addition of trichloroacetic acid. b, hepatocytes were preincubated with 3 nM glucagon for 60 min and were subsequently incubated for the indicated periods (without removal of the hormone) either without any addition (0)or with the further addition of 30 nM (0)or 300 nM (A) glucagon; the figure illustrates data obtained during the second incubation. All incubations occurred at 30 "C in the presence of 5 mM theophylline.

15516

Modulation of Hepatocyte CAMPby Glucagon

and lower steady-state level of the cyclic nucleotide. Taken together,theresults of Figs. 1 and 2defined a complex relationship between glucagon concentration and cAMP accumulation by isolated canine hepatocytes and an effect of high concentrations of glucagon to lessen the steady-state levels of CAMP potentially achievable by hormone stimulation. The ability of the hormone at high concentrations t o decrease steady-state levels of CAMP(relative tolevels achievable by lower concentrations) was not affected by the presence in the incubationmedium of 1 p~ insulin, 0.1 p~ apamin, or 1 p M phorbol 12-myristate 13-acetate and was reversed by washing and reincubating cells that had been previously exposed to glucagon (data not shown).As illustrated in Fig. 2d, low concentrations of glucagon had no effect on the enhancement of cAMP accumulation already stimulated by isoproterenol, whereas higherconcentrations of the hormone permitted the expression of glucagon activity to inhibit isoproterenolstimulated cAMP accumulation. The @-adrenergicantagonist, o0 . l ~ propranolol, however, did not affect the shape of the curve describing the concentration dependencefor the stimulation of the adenylyl cyclase system by glucagon. Thus, whereas Fig. 2a identifies glucagon at higher concentrations to be an antagonist of the glucagon-stimulated cyclase system, Fig. 2d identifies glucagon at the same concentrations to be an antagonist of the @-adrenergiccyclase system aswell. Additional experiments compared the activitiesof glucagon and glucagon analogs in both enhancing and inhibiting cAMP accumulation in isolated hepatocytes. Data presented in Fig. 3 (open circles in a, c, e, and g) show that glucagon a t higher concentrations is as effective at inhibiting cAMP accumula9) 0 tion stimulated by cyanogen bromide-cleaved glucagon, [N"Cbm-His1,N"Cbm-Lys'2]glucagon, des-His1-glucagon, and 0.2 .c [N"-trinitrophenyl-His',homo-Arg'2]glucagon, respectively, as it is a t inhibiting cAMP accumulation stimulated either by X b, d, f, and itself or by isoproterenol. Fig. 3 (closed squares in 2 5 0.I h) shows that, whereas each of the four analogs noted above stimulateshepatocytecAMPaccumulation, onlycyanogen bromide-cleaved glucagon exhibits a biphasic concentration ci dependence in which cAMPaccumulation is enhanced by lower concentrations of the peptide and is diminished by -11 -10 -9 -8 -7 -6 -5 - '-12 0 higherones. Fig. 3 (open squares in b, d, f, and h; and i) Log Molar Glucagon Concentration further shows, again for the four glucagon analogsnoted FIG. 2. Dependence of cAMPaccumulation inisolated hepatocytes on the concentrationof glucagon andon the pres- above, that only cyanogen bromide-cleaved glucagon has an ence of theophylline. a, hepatocytes were incubated with glucagon action at higher concentrations to diminish the hepatocyte at the indicated concentrations for 30 min at 30 "C in the presence cAMP accumulationinduced by glucagon or by isoproterenol. of 5 m M theophylline. b, hepatocytes were incubated with glucagon It can thus be said that these peptides fall into two groups: at the indicated concentrations for30 min at 30 "Cwithout further addition (0)or with the addition of 0.5 mM (0)or 5 mM (W) glucagon and cyanogen bromide-cleaved glucagon in our systheophylline. c, the data of b were normalized to the maximal levelof tem express a n activity to inhibit cAMP accumulation incAMP achieved by glucagon in the presence of different concentra- duced by each alone, by eachother, or by isoproterenol; tions of theophylline (that achieved in the presence of 1 nM hormone). whereas [N"-Cbm-His',N"Cbm-Lys'2]glucagon, des-His1d, hepatocytes were preincubated withoutfurther additions (0)were glucagon, and [N"-trinitrophenyl-His',homo-Arg12]glucagon preincubated in the presence of 2 p~ propranolol (0) or 1 p M exhibitapparently simple concentration dependences for isoproterenol (a) prior to the addition of glucagon and subsequent on incubation during 30 min at 30 'C in the presence of 5 mM theoph- stimulating hepatocyteadenylyl cyclaseand have no effect the inhibition of either glucagon- or isoproterenol-stimulated ylline. cAMP accumulation. It is important to note that our ability was steady-state levels of cellular cAMP and therefore to increaseto detect inhibitoryeffects for the non-inhibiting analogs not limitedby our choice of peptide concentrations; the highthe accuracy of cAMP measurements. ) study of the three To study further theeffect of high concentrations of glu- est concentrations used (10 p ~ during cagon to limit thelevel of cAMP achieved by lower ones, we non-inhibiting analogs noted above are 1400-, 200-, and 50fold greater than their respective dissociation constants for examined how cells thathadalready been stimulatedto accumulatecAMP by glucagon would respondto higher high affinity glucagon-binding sites and loo-, 40-, and 9-fold amounts of the hormone. As shown in Fig. lb, the simple greater than their respective dissociation constants for low affinity glucagon-binding sites on isolated hepatocytes (Table addition of 30 or 300 nM glucagon t o cells that had been previously incubated with 3 nM glucagon resulted in a time- I and Ref. 30). With thepossible exception of the interaction with low afdependent decline in thelevel of cAMP achieved by the lower of [N"-trinitr~phenyl-His',homo-Arg'~]glucagon concentration of hormone and in the achievementof a new finity binding sites, hepatocyte glucagon receptors can thus

-

"0

5

5

Modulation of Hepatocyte cAMP by Glucagon

15517

TABLE I Apparent dissociation constants for glucagon and analog interactions with high and low affinitybinding sites and potency of the peptides in stimulating CAMPaccumulation Values for the apparent dissociation constants for peptide interactions with glucagon-binding sites on canine hepatocytes were taken A ~ concen~) from the accompanying paper (30). Values for K O . S ( e(the tration of peptide causing half-maximal stimulation of cAMP accumulation in isolated hepatocytes) were derived from the data shown in Fie. 3. ~

Peptide

K~1hi.h)

nM

nM

KDUW nM

0.83 69 0.22" 100 19' [N"-Cbm-His',N"Cbm-Lyslz]Glucagon 7.3 3100 14 19" Cyanogen bromide-cleaved glucagon 240 49 66' Des-His'-glucagon 1340 420' [N"-trinitrophenyl-His',h~mo-Arg'~]- 180 Glucagon 12 NS' 91 [~-Phe',TyP,diiodoTyr'o,Arg'2,Lys17~18,Gluz1~Glucagon Exhibits abiphasic response in which low peptide concentrations stimulate cAMP accumulation and in whichhigh concentrations inhibit the effectiveness of lower ones. The value shown corresponds tothe lower concentration causing half-maximal stimulation of cAMP accumulation. The higher concentration at which cAMP accumulation is half-maximal is about 48 nM for glucagon and about 10 p~ for cyanogen bromide-cleaved glucagon. Exhibits a monophasic response to increasing peptide concentration a t concentrations 1 1 0 p ~ . e Exhibits no stimulation (NS) of cAMP stimulation at peptide concentrations 1 1 0 p ~ .

Glucagon

rn

50

-11

100

-10 - 9

-8

-7 -6 -5

50 0 -11

I

,

-10 -9

,

,

,

-8 "7 -6 -5

Log Molar PeptideConcentration

FIG.3. Dependence of cAMP accumulation in isolated

'

be considered to be fully saturated when cells were incubated in the presence of 10 1M glucagon analogs. Results obtained by use of the glucagon analog [D-Phe4, Tyr5,diiodo-Tyr'~,Arg'2,Lys'7~'8,Glu2']gl~cag~n differ markedly from those reported above for other less highly modified analogs of the hormone. Fig. 4 (a-c) shows, respectively, that [D - Phe4,Tyr' ,diiodo - Tyr", Arg" ,LYS'~,'~, Glu2']glucagon ( a ) fails to enhance cAMP accumulation in isolated hepatocytes, although it possesses a favorable affinity for the glucagon receptor (Table I and Ref. 30); ( b ) inhibits hepatocyte cAMP accumulation induced by glucagon; and ( c ) has no effect on hepatocyte cAMP accumulation induced by isoproterenol. Thus, whereas the analog might be regarded as a weak or partial agonist in some systems (cf. Ref. 30), it appears to antagonize glucagon action with regard to the stimulation of at least measurable increases in cAMP accumulation. Two additional pointsshould be made regarding the dataof Fig. 4. First, just as propranolol (a @-adrenergicantagonist) has no effect on cAMP accumulation induced by glucagon (Fig. 2 4 , [~-Phe~,Tyr~,diiodo-Tyr'O,Arg~~,Lys'~~'~,Glu~~]glucagon has no

hepatocytes on the concentration of glucagon and glucagon analogs and effects of glucagon and glucagon analogs on cAMP accumulation induced by glucagon, glucagon analogs, and isoproterenol. a and b, c and d , e and f , and g and h are paired; each pair illustratesthe results of a separateexperiment. Experiments were designed to test the abilities of glucagon and glucagon analogs to enhance cAMP accumulation in isolated hepatocytes, the ability of glucagon to affect cAMP accumulation induced by glucagon analogs, the ability of glucagon analogs to affect cAMP accumulation induced by glucagon, and theability of glucagon and glucagon analogs to affect cAMP accumulation induced by isoproterenol. a, dependence of cAMP accumulation on the concentration of glucagon (0)and effect of glucagon on cAMP accumulation induced by 0.1p~ cyanogen on cAMP accumulation induced bromide-cleaved glucagon (0).b, dependence of cAMP accumulation trophenyl-His1,homo-Arg12]glucagon on the concentration of cyanogen bromide-cleaved glucagon (B) and by 3 nM glucagon (0).i, dependence of cAMP accumulation on the effect of cyanogen bromide-cleaved glucagon on cAMP accumulation concentration of glucagon analogs in cells preincubated with 1 p~ induced by 3 nM glucagon (0).c, dependence of cAMP accumulation isoproterenol (30 min, 30 "C) prior to the addition of other comon the concentration of glucagon (0)and effect of glucagon on cAMP pounds: glucagon (01,cyanogen bromide-cleaved glucagon (O), [N"trinitrophenyl-His1,homo-Arg12]glucagon (B), and [N"-Cbm-His',N" accumulation induced by 0.1 p~ [N"-Cbm-Hisl,N"Cbm-Lyslz]glucagon (0).d, dependence of cAMP accumulation on the concentration Cbm-Ly~~~]glucagon (0).All cell incubations for the measurement of (B)and effect of [NO-Cbm- cAMP occurred during 30 min at 30 " C in the presence of 5 mM of [N"-Cbm-Hisl,N"Cbm-Lyslz]glucagon His1,N"Cbm-Lys'2]glucagonon cAMP accumulation induced by 3 theophylline and glucagon or glucagon analogs at theconcentrations nM glucagon (0).e, dependence of cAMP accumulation on the con- indicated. Data have been normalized to account for differences in centration of glucagon (0)and effect of glucagon on cAMP accumu- cAMP levels achieved in the different experiments. "Maximal acculation induced by 1pM des-His1-glucagon(0).f , dependence of cAMP mulation" of cAMP is defined as thelevel obtained during incubation accumulation on the concentration of des-His1-glucagon (B) and of cells with the effectors at concentrations previously determined to effect of des-His'-glucagon on cAMP accumulation induced by 3 nM give maximal or near-maximal response: glucagon, 3 nM; glucagon glucagon (0).g, dependence of cAMP accumulation on the concen- cleaved by cyanogen bromide, 0.1 /IM;[N"-Cbm-His',N"Cbm-Ly~~~] 0.1 p M ; des-His'-glucagon, 1 pM; [Nu-trinitrophenyltration of glucagon (0)and effect of glucagon on cAMP accumulation glucagon, ; isoproterenol, 1 pM. Maximal (0). His1,homo-Arg'2]glucagon,3 p ~ and induced by 3 p~ [N"-trinitrophenyl-His',homo-Arglz]glucagon h, dependence of cAMP accumulation on the concentration of [NO- values ranged from 0.17 to 0.28 nmol of cAMP/2 X lo6 cells in the trinitrophenyl-His',homo-Arglz]glucagon(B) and effect of [N"-trini- experiments shown.

Modulation of Hepatocyte CAMPby Glucagon

15518

O+-

I

I

I

0'

d U

I

I

I

I

I

1

1

\

I

I

I

I

03 I

0

-

1

C

0.2

0.I

0 -12

I

-11

I

-10

I

I

I

I

I

-9

-8

-7

-6

-5

tide-bindingprotein Ni andtostimulate adenylyl cyclase activity by inactivating an inhibitory component(34, 35)). A comparison of Fig. 5 ( a and b ) shows that theadenylyl cyclase system once activated by cholera toxin is no longer affected by glucagon in eithera positive or negative way. A comparison of Fig. 5 (a and c ) shows, however, that treatment of hepatocytes by pertussistoxineliminatesthe negative effect of glucagon which results in the inhibitionof cAMP accumulation by higher concentrations of the hormone. As Fig. 6a shows that the inhibitory action of glucagon on the isoproterenol-stimulated accumulation of cAMP is also prevented by prior treatment of hepatocytes by pertussis toxin, we can conclude that (a) the stimulatory and inhibitory activitiesof glucagon on the hepatocyte adenylyl cyclase system are mediated by different mechanisms, and ( b )the inhibitory activity of the hormone on both glucagon- and isoproterenol-stimulated cAMP accumulation arisesfrom a common mechanism thatis affected by a pertussis toxin substrate. Last, Fig. 6b shows thattheantagonist[~-Phe~,Tyr',diiodoTyr10,Arg32,Ly~17*18,Glu21]glucag~n inhibits the ability of glucagon at high concentrations to decrease isoproterenol-stimulated cAMP accumulation in isolated hepatocytes. This predictable result (obtained by use of an analog which, by its bindingto glucagon receptors, prevents glucagon-receptor interactions and which has no direct effect on isoproterenolstimulated cAMP accumulation) identifies the source of the glucagon-directed inhibition of cAMP accumulationa t higher

Log Molar Peptide Concentration FIG. 4. Dependence of cAMP accumulation in isolated v) hepatocytes on the concentration of glucagon and [DPhe4,Tyr' ,diiodo - Tyr'O, Arg" ,Lys'7J8,Gl~2']gl~~agon and effect of glucagon and [~-Phe",Tyr~,diiodo-Tyr'~,Arg'~, Lys17*'S,G1~2']gl~~agon on glucagon- and isoproterenol-induced cAMPaccumulation. a, dependence of cAMP accumulation

-

0

1 I

I

I

,

I

I

on the concentration of glucagon (0) and[~-Phe~,Tyr',diiodo, ,Lys17.18,G1u21]glucagon (0).b, effect of glucagon (0)and [~-Phe~,Ty~,diiodo-Tyr~~,Arg~~,Lys~'~~~,G1u~~]gl~cag0n (0)on cAMP accumulation induced by glucagon; cells were preincubated with 3 nM glucagon for 30 min a t 30 "C prior to the start of the experiment. e, effect of glucagon (0)and [~-Phe',Tyr~,diiodo-Tyr'O,Arg~~,Lys~'J~, GluZ1]glucagon(0)on cAMP accumulation induced by isoproterenol; cells were preincubated with 1 p~ isoproterenol for 30 min a t 30 "C prior to the startof the experiment. All incubations occurred during 30 min a t 30 "C in the presence of 5 mM theophylline. Tyr10 Arg12

o'

effect on cAMP accumulationinduced by isoproterenol (a padrenergic agonist) (Fig. 4c). Second, whereas both glucagon and the glucagon analog noted above inhibit glucagon-stimu0.0I - 1 1 -10 -9 -8 -7 -6 -5 lated cAMP accumulation(Fig. 4b), the mechanismsby which these two peptides act probably differ considerably: glucagon apparently participatesby inhibiting cAMP accumulation per Log MolarGlucagon Concentration se; the analog apparently acts by competing for the binding FIG. 5. Dependence of cAMP accumulation on Jucagon of glucagon to adenylyl cyclase-linked glucagon receptors. concentration in isolated hepatocytes treated with cholera To investigate the mechanism by which glucagon a t high toxin or pertussis toxin. a, dependence of cAMP accumulation on concentrations exerts its inhibitory affect on cAMPaccumu- glucagon concentration in cells preincubated in buffer alone for 2 h at 37 "C. b, dependence of cAMP accumulation on glucagon concenlation in isolated hepatocytes, we next determined the hormone sensitivitiesof hepatocytes pretreated with either chol- trations in cells preincubated for 2 h at 37 "C in the presence of cholera toxin (10 pglml). e, dependence of cAMP accumulation on era toxin (a protein known to cause the ADP-ribosylationof glucagon concentration in cells preincubated for 2 h at 37 "C in the the guanyl nucleotide-bindingprotein N, andtoactivate presence of pertussis toxin (10 pglml). All incubations with glucagon adenylyl cyclase directly(33,34)) or pertussis toxin (a proteinoccurred during 30 min at 30 "C in the presence of 5 mM theophylline; known to cause the ADP-ribosylation of the guanyl nucleo- the toxins were present in the final incubation.

Modulation of Hepatocyte CAMPby Glucagon

high and low affinity glucagon-binding sites (29, 30) and the full effects of glucagon on the accumulation of hepatocyte O.3 cAMP (this report) arevery similar and extend well beyond what would be expected for simple and homogenous biological processes. Whereas high affinityhormone-bindingsites havebeen implicated in the actionsof glucagon on hepatocyte glycogen metabolism (30), the dissection of the roles of high and low affinity binding in the stimulationof cAMP accumulation is Y-x o not possible under the circumstances that (a) glucagon and cyanogen bromide-cleaved glucagon exhibit biphasic stimulatory and inhibitoryeffects on cAMP accumulation;(6) [NeCbm-His1,N"Cbm-Lys'2]glucagon, des-His1-glucagon, and [N"-trinitrophenyl-His',homo-Arg'2]glucagon exhibit only 0.1 monophasic stimulatory effects; (c) [~-Phe*,Tyr',diiodoTyr'o,Arg'2,Lys'7~'8,Gluz']glucag~n exhibits noeffect on cAMP accumulation; and ( d ) in all cases, occupancy of low affinity O-9 -8 -7 -6 -5 sites as well as that of high affinity sites increase monotoniLog molar peptide concentration cally as effector concentrations are increased from very low FIG. 6. Effects of pertussis toxin and [D-Phe',Tyr",diiodo- to high values. Whereas it islikely that high affinity binding Tyr'o,Arg'2,Lys'7~'8Glu2']glucagon on the inhibition by glu- sites participate in the stimulation of adenylylcyclase by cagon of isoproterenol-induced cAMP accumulation in iso- glucagon, we cannot exclude the participation of low affinity lated hepatocytes. a, data from cells preincubated for2 h at 37 "C binding sites or of both populations of binding sites in the in buffer alone (0)or in buffer containing 10 rg/ml pertussis toxin (0).Subsequent to preincubation at 37 "C, cells were incubated first stimulatory effects of the hormone. It isalso noteworthy that peptide concentrations causing half-maximal stimulation of for 30 min at 30 "c in the presence of 1 p M isoproterenol and then for 30 min at 30 "C with glucagon at the indicated concentrations in cAMP stimulation (Table I) are higher than those necessary the presence of 5 mM theophylline; all agents were present in the to cause half-maximal inhibition of ['4C]fructose incorporafinal incubation. b, data from cells preincubated for 30 min at 30 "C tion into hepatocyte glycogen (30); it thus appears that relain the presence of 1 WM isoproterenol and then incubated for 30 min at 30 "C in the presence of 5 mM theophylline with glucagon at the tively fewer so-called spare receptors exist in the glucagonindicated concentrations (0)or with glucagon at the indicated con- stimulated generation of CAMP. Notwithstanding the recent finding that some actions of glucagon might be mediated centrations plus 1 p~ [~-Phe~,Tyr~,diiodo-Tyr'~,Arg'*,Lys~'J~,Glu~~] glucagon (0). through the inositol phosphate pathway (31), the latter result does not negate the possibility that very small increases in concentrations of the hormone as arising from specific inter- cAMP concentration might actually be sufficient to induce actions of glucagon with its plasma membranereceptor. major effects on cellular metabolism. Despite reservations inassigning glucagon's ability to stimDISCUSSION ulate cAMP accumulation specifically to its interaction with high or low affinity binding sites,our data are consistent with Our results on the stimulation and inhibitionof the adenthe involvement of low affinity receptors in the activity of ylyl cyclase system in isolated canine hepatocytes have demglucagon which inhibitstheaccumulation of hepatocyte onstrated ( a ) the stimulation of cAMP accumulation by low CAMP. Since (a) high affinity receptors would be more than concentrations of glucagon, ( b ) the homologous inhibition of 90% saturated a t 10 nM hormone; and (b) theinhibitory glucagon-stimulated cAMP accumulationby high concentraeffects of glucagon on cAMP accumulation extend through tions of glucagon, (c) the heterologous inhibition of isoproterenol-stimulated cAMP accumulation by high concentrations glucagon concentrations rangingfrom 10 to1000 nM (concenof glucagon, ( d ) the potential involvement of a pertussis toxin trations thatwould increase markedly occupancy of low affinof high affinityreceptors),the substrate in the homologous and heterologous inhibition of ity receptors but not that inhibitory action of glucagon seems most likely to arise from cAMPaccumulation by glucagon, and ( e ) the use of [DPhe4,Tyr5,diiodo-Tyr1o,Arg'z,Lys17~''Gluz']glucagon as a probe occupancy of low affinity glucagon-binding sitesonthe for thestimulatoryandinhibitory effects of glucagon on hepatocyteplasmamembrane.Whereasrelatedarguments to theinhibition of cAMPaccumulation hepatic cAMP accumulation. Inhibitory effects of glucagon mayalsoapply on the adenylyl cyclase system have also been reported for caused by cyanogen bromide-cleaved glucagon, final resolution of this issue as well as that pertaining to binding sites assays involving the use of rat hepatic plasma membranes (10) and hepatocytes (11).Most important, the biphasic cel- applicable in the stimulation of cAMP accumulation must lular response to glucagon reported here for the accumulationawait the availability of specific ligands that bind independof cAMP mimics in many ways that recently reportedfor the ently to one set of binding sites or the other. The failure of [N"-Cbm-His',N"Cbm-Lys'z]glucagonand other analogs to glucagon-stimulated formation of hepatocyte inositol phosphates (31). It thus appears thatglucagon may act by either cause inhibition of cAMP accumulation at high concentraof two separate signal transducing pathways and that inhibi-tions (concentrations that would essentially saturate low aftory aswell as stimulatoryeffects of glucagon have the poten- finity as well as high affinity binding sites) ( a ) demonstrates tial for contributing to its activity in multiple systems. Use of the existenceof specific structural requirements for initiation isolated canine hepatocytes has been advantageous in permit-of the inhibitory response and ( b ) suggests an incomplete coupling of receptors filled with these peptides to the relevant ting analysis of glucagon's effects on hepatocyte cAMP accumulation under conditions identical to those used for the signal transducingsystem. The causeof this incompletecoureceptor binding studies reported in the accompanying paperpling and the nature of the related structural requirements (30). In fact, the range of glucagon concentrations (about remain to be determined. 0.05-1000nM) required to describe both the saturation of The potentialinvolvement of a pertussis toxin substrate in

I-

!

15519

4

15520

Modulation of Hepatocyte cAMP by Glucagon

the inhibitory action of glucagon on glucagon-stimulated CAMPaccumulation is consistent with both previous suggestions for a role of N, in the modulation of the adenylyl cyclase system by glucagon (10-12, 36) and the determination that the inhibitory actions of angiotensin (37) and vasopressin (14) on the adenylyl cyclase system requires the participation of pertussis toxin substrates. Whereas it is possible that the inhibitory action of glucagon on cAMP accumulation arises from a direct inhibition of adenylyl cyclase, alternative mechanisms might involve instead an uncoupling of occupied receptors from the CAMP-generatingenzyme or a quite separate activation of a CAMP-directed phosphodiesterase. The heterologous inhibition of isoproterenol-stimulated hepatocyte cAMP accumulation by glucagon and its prevention by the treatment of cells with pertussis toxin indicate further a tight coupling of the &adrenergic and glucagon systems. Nevertheless, considering both previous results (17-24, 31) and results reported here, it is difficult to predict whether a particular glucagon analog will uniformly behave as anagonist or antagonist based on its action in a single membrane, cell, or animal system. Taken together, our results identify complexities that may arise when evaluating the activities of glucagon analogs as potential glucagon antagonists (i.e. bloodglucose-lowering agents) in uivo. First, in the absence of @-adrenergicresponses in intact animals, a glucagon analog might appear to act asa glucagon agonist or glucagon antagonist, depending on its concentration. Second, in the presence of important P-adrenergic responses, a glucagon analog might behave as a glucagon agonist, a glucagon antagonist, or a@-adrenergicantagonist. Third, in the presence of important @-adrenergicresponses that areinhibited by glucagon, a glucagon antagonist similar to the inhibitory analog [~-Phe*,Tyr',diiodoTyr10,Arg12,Ly~'7~'sGlu21]plucagon might actually appear to act as a glucagon agonist by means of its relief of P-adrenergic responses from inhibition by glucagon. It remains to be determined whether the effect of high concentrations of glucagon to decrease potential cellular responses to glucagon and to@adrenergic agonists represents what might be called a pharmacologic "fail-safe" mechanism to prevent the generation of excessive levelsof hepatic CAMP.Dissection of these complex processes at the physiological, cellular, and molecular levels will clearly provide important information on the regulation of glycogenolytic and gluconeogenic metabolism in vivo, on the efficacy of glucagon antagonists both in vivo and in vitro, and on the mechanisms by which the filling of glucagon and @-adrenergicreceptors by ligand affect homologous and heterologous responses to hormones and hormone analogs.

itz, R. J. (1984)J . Biol. Chem. 259,9742-9749 4. Su, Y.-F., Harden, T. K., and Perkins, J. P. (1980)J. Biol. Chem. 255, 7410-7419 5. Toews, M. L., Waldo, G. L., Harden, T. K., and Perkins, J. P. (1984)J. Biol. Chem. 259,11844-11850 6. Homburger, V., Lucas, M., Cantau, B., Barabe, J., Penit, J., and Bockaert, J. (1980)J. Biol. Chem. 255. 10436-10444 7. Kassis, S., Zaremba, T., Patel, J., and Fishman, P. H. (1985)J. Biol. Chem. 260,8911-8917 8. Noda, C., Shinjyo, F., Tomomura, A., Kato, S., Nakamura, T., and Ichihara, A. (1984)J. B i d . Chem. 259,7747-7754 9. Sibley, D. R., Strasser, R. H., Caron, M. C., and Lefkowitz, R. J. (1985)J. Biol. Chem. 260, 3883-3886 10. Heyworth, C. M., and Houslay, M. D. (1983)Bioehem. J. 214, 93-98 11. Heyworth, C. M., Hanski, E., and Houslay, M. D. (1984)Biochem. J. 222,189-194 12. Heyworth, C. M., Grey, A.-M., Wilson, S. R., Hanski, E., and Houslay, M. D. (1986)Biochem. J. 235, 145-149 13. Rich, K. A., Codina, J., Floyd, G., Sekura, R., Hildebrand, J. D., and Iyengar, R. (1984)J.Biol. Chem. 259, 7893-7901 14. Wilson, P. D., Dixon, B. S., Dillingham, M. A., Garcia-Sainz, J. A., and Anderson, R. J. (1986)J. Biol. Chem. 261, 1503-1506 15. Garrity, M.J., Andreasen, T. J., Storm, D. R., and Robertson, R. P. (1983)J. Biol. Chem. 258,8692-8697 16. Hsia, J. A., Hewlett, E. L., and Moss, J. (1985)J. Biol. Chem. 260,4922-4926 17. Bregman, M. D., Trivedi, D., and Hruby, J. (1980)J.Bid. Chern. 255,11725-11731 18. Epand, R.M., Rosselin, G., Hoa, D. H. B., Cote, .T. E., and Laburthe, M. (1981)J. Biol. Chem. 256,1128-1132 19. Hruby, V. J. (1982)Mol. Cell. Biochem. 44.49-64 20. England, R. D., Jenkins, W. T., Flanders, K. C., and Gurd, R. S. (1983)Biochemistry 22, 1722-1728 21. Krstenansky, J. L., Trivedi, D., and Hruby, V. J. (1986)Biochemistry 25,3833-3839 22. McKee, R. L., Pelton, J. T., Trivedi, D., Johnson, D. G., Coy, D. H., Sueiras-Diaz, J., and Hruby, V. J. (1986)Biochemistry 25, 1650-1656 23. Gysin, B., Trevedi, D., Johnson, D. G., and Hruby, V. J. (1986) Biochemistry 25,8278-8284 24. Corvera, S., Huerta-Bahena,J., Pelton, J. T., Hruby, V. J., Trivedi, D., and Garcia-Sainz, J. A. (1984)Biochim. Biophys. Acta 804,434-441 25. Johnson, D. G., Goebel, C. V., Hruby, V. J., Bregman, M. D., and Trivedi, D. (1982)Science 215,1115-1116 26. Rojas, F. S., Swartz, T. L., Iyengar, R., Garber, A. J., and Birnbaumer, L. (1983)Endocrinology 113, 711-719 27. Horwitz, E. M., Jenkins, W . T., Hoosein, N. M., and Gurd, R. S. (1985)J. Biol. Chem. 260,9307-9315 28. Sonne, O., Berg, T., and Christoffersen, T. (1978)J. Biol. Chern. 253,3203-3210 29. Bonnevie-Nielsen, V., and Tager, H. S. (1983)J. Biol. Chem. 258,11313-11320 30. Hagopian, W . A., Tager, H. S., Gysin, B., Trivedi, D., and Hruby, V. J. (1987)J. Biol. Chem. 262,15506-15513 31. Wakelam, M. J. O., Murphy, G. J., Hruby, V. J., and Houslay, M. D. (1986)Nature 323,68-71 Acknowledgments-We thank Dr. Beat Gysin for supplying us with 32. Bonnevie-Nielsen, V., Polonsky, K. S., Jaspan, J. B., Rubenstein, the glucagon antagonist [~-Phe~,Tyr~,diiodo-Tyr~~,Arg~*,Lys~'*~*, A. H., Schwartz, T. W., and Tager, H. S. (1982)Proc. Natl. Gl~*~]glucagon and Crystal Sherman-Jones for expert assistance in Acad. Sci. U. S. A. 79,2167-2171 the preparation of the manuscript. 33. Northup, J. K., Smigel, M. D., Sternweis, P. C., and Gilman, A. G. (1983)J. Biol. Chem. 258, 11369-11376 REFERENCES 34. Gilman, A. G. (1984)Cell 36,577-579 1. Mukharjee, C., Caron, M. G., and Lefkowitz, R. J. (1975)Proc. 35. Katada, T., Bokoch, G. M., Northup, J. K., Ui, M., and Gilman, Natl. Acad. Sci. U. S. A. 72,1945-1949 A. G. (1984)J. Biol. Chem. 259,3578-3585 2. Stadel, J. M., Strulovici, B., Nambi, P., Lavin, T. N., Briggs, M. 36. Cerione, R. A., Staniszewski, C., Caron, M. G., Lefkowitz, R. J., Codina, J., and Birnbaumer, L. (1985)Nature 318, 293-295 M., Caron, M.C., and Lefkowitz, R. J. (1983)J. Biol. Chem. 37. Pobiner, B. F.,Hewlett, E. L., and Garrison, J. C. (1985)J. Biol. 258,3032-3038 Chem. 260,16200-16209 3. Sibley, D. R., Peters, J. R., Nambi, P., Caron, M. G., and Lefkow-