The Glucagon-sen&ive Adenyl Cyclase System in Plasma Membranes of Rat Liver I' V. AN OBLIGATORY ROLE OF GUANYL NUCLEOTIDES IN GLU&ON ACTION (Received for publication, October 7, 1970) MABTIN F~DBIGLL, LUTZ BIRNBAUMER, STEPHEN L. POEL, AND H. MICEIEL J. KRANS* From ihe Section on Membrane Rqulation, L&oralory of Nutrition and Endocrinology, National Inaitlutt of Arthritis and Metabolio Da'wam, Bethcuda, Mar&and PGQ14 SUMMARY A mrthod ie deecribed for the o azymetlc ayuthmie of S'-odenylyl imidodlphoaphete labeled with "P at the a posl- tion (AMP-PNP-a-P), au umlogue of ATP con-g nitro- gen substituted for oxygen bmeen the terminal phosphates. The nucleotide h only rlowly hydrolyzed during incubation with rat liver pl44m4 membranes and is 4 substrate for rdenyl cyclue in theee membranes. In the presence of 0.2 my AMP-PNP, glucagon md fluo- ride ion stimulete edenyl cyclese 4ctMty; linear rates are maintained for at Ieert 10 min of incubation at 30'. GTP enh4nced the initirl rate of brsrl end glncegon-stimulated deny1 cyclane 4ctivity. Reduction in concentrdion of Mg++ in the uuy medium or incubation of liver membranes for S min at 30. prior to addition of glucegon results in loss of re- rponae of *deny1 cyclue to glucegon and in reduction in the effecta of GTP on basal rctivity. Under these conditions GTP, GDP, or GMP-PCP ue required for glucrgon rtimulr- tion of the enzyme even though the rpwi5c bind& rites for glucagon are uturrted with hormone: a6 little 44 10 IJM GTP or GDP ia required. UTP and CTP exert smaller ef- fecta then the guanyl nucleotides end act only at concentrr- tionr higher than 0.1 LUI. The guanyl nucleotides inhibited the response of rdenyl cyclue to fluoride ion (10 my) over the same concentration rrngeoverrhkhthey 4timal4tethere4pon4eoftheenzyme to glucegon. Thin e&on of the nacleotidea in observed in pl44mr membun~ treated with phoepholipane A under conditions that remit III lore of glucagon biudiug end of hormonal response. It b concluded that guanyl nucleotidea play 4 specific and obligatory role in the activation at adenyl cyclase by glucagon. The nucleotidea bind at sitm, distinct from the gluugon binding rites, that 4ppeu to regulate both the rempoxue of adeny cydue to glucagon, and, portily by 4 related mecha- dsm. the retion4 of fluoride ion on thie ryntem. o Recipient of a fellowship from the Netherlands Organization for the Advancement of Pure Rsrervah Q.W.O. 196@-1970). In the previous study (l), it was shown that gunny1 nucleotides (GTP or GDP) equally and at concentrations as low ns 0.05 +I, 4lt.m the properties of the specific binding sites for glucagon in rat liver plasma membranes. GMP-PCP,' a nonphospborylating 4nalogue of GTP, mimicked the actions of the natural nucleotides on glucsgon binding, suggesting that the nucleotides act by bind- ing to sites, as yet undefined, that in9uencs the structure of the glucagon binding sites. ATP, ADP, UTP, and CTP similarly affected hinding of glucagon but only at concentrations above 0.1 InM. It was of obvious interest to determine whether the actions cf guanyl nucleotides on the glucagon biding sites, which RPIMW tn be related to the adenyl cyclase system (2), have their correlate on the response of adenyl cyclase to glucagon. Studies reported elsewhere (3) in preliminary form, showed that guanyl nucleotidea enhanced the response of adenyl cyclase to glucagon but only at concentmtions considerably higher than those required for their actions on binding of glucagon. The preliminary studies were a~rried out with concentrations of ATP, the substrate for edenyl cyclaw, that affeoted glucagon binding in the same manner as the guanyl nucleotides. Reduction of ATP concentration to low levels (0.2 rnM or less) resulted in mpid hydrolysis of the nucleo- tide even in the presence of ATP-regenerating systems, creating dif?icultiea in interpretation of kinetic data. In this study, the enrymatic synthesis of "P-labeled AMP- PNP is described. This analogue of ATP contains nitrogen in place of oxygen between the terminal phosphates and was found to be resistant to hydrolysis by ATPases in liver membranes. It will be shown that it is a substrate for adenyl cyclase. It will also be shown that guanyl nucleotides, at concentrations as low 4s 10 no, play an obligatory role in the activation of adenyl cy- &se by glucagon but inhibit, possibly by a related mechanism, the response of the ensyme to fluoride ion. EXPERIMENTAL PROCEDURE Only thotw materisle and methods not described in the pm- 1 The abbreviations used we.: CMP-PCP, c'-gusnylyl-diphw- phonate; AMP-PNP, 5'-adenylyl-imidodiphosphate; PNP, di- phoephoimide; ADP-NH,, 5'-adenylyl-pbosphoamide; cyclic AMP, cyclic3',5'-AMP+. 1877 1878 Qlucagon-seMilivc A&q11 Cyclaec. V Vol. 246, No. 6 ceding paper8 (1,2,4, S) are documented in tbe present communi- twhkm. MO&&-AMP-PNP(Na,) and PNP(NsJ were kindly sup. plied by Dr. Ralph Yount (Washington State University, Pull- man). AMP-PNP WM contaminated with 16% ADP-NHl. Zhharkhiu di extract WM prepared from E. wfi (strain B) ruxording to the procedure of Nirenberg (6) and represents the lOS,OOO X p 8upematant (S-100 fraction) obtained in Step 4 Sf tbi8pKMXdur8. Pmpudfon and PtuQltdon oj AMP-PNP-uJ*P: Zameonlk and Stepb8n8on have shown (7) that metbylene diphwpbonate ir inoorporated into ATP by purlfled E. cdi lysyl-tRNA 8ynthe- ta8e (EC 6.1.1 .tlO) aocording to the following reactions: ATP + Lye + enrymeh # (AMP*Ly8-enrymeL,,) + PP PCP + (AMP.Ly8~enrymo~J + AMP-PCP + enzymeL,, + Ly8 which in the presence of txce8a PCP ewentially go ta completion, particularly if pyropbo8pbatrrse is present during incubation. We wet-8 informed by Dr. Zsmscnik that AMP-PNP can he pm- pared by tb8 8ame ma&ion. Instead of E. c&d lyayl-tRNA syntbetaae we ued a crude 6. coZi extract containing a number of amino acid tRNA 8yntbetaae8 and which k rich in pyropbo8pha. tase activity. A mixture of 19 naturally occurring amino acids (except leucine), each at 1 mu, wa8 used a8 8ubetrate8 for the ma&ion. Tbe following prooedure was wed to obtain AMP- PNP-a-"P of high rpeuitic activity. Solution8 containing 0.15 to 0.3 rmole 01 ATP-a-*P (3 to S Ci per mmob) were evapo- Fla. 1 (is/l). ERect of incubation of liver plasma membranes, under adcnyl cyclaee assay conditions, on levels of ATP and AMP-PNP. n, liver plasma mombranes (0.7 mg of protein per ml) were incubated at 30. for varyin periods of time in medium contrlnlnll: 0.2 mu ATPu-aP (140 cpm per pmole), 3 my MgClr, 23 my Trir-HCI, pH 7.6, and an ATP-regenerating system con- rlstlng of 20 mu cnatine phoepbate and 1 mg per ml of creatine kinase. b. liver plasma membranes (0.4 mg of protein per ml) were incutiatad ai 30' for 20 min in medium coninining 6.28 rn; AMP-PNPu-*P @Xl eom WC omole). 4.0 my MnCL 1 my EDTA. 1 mx cyclic AM*, O.&" ilbu-min, &d 25 my ?ri&HCl, pH 7.6: Heactions were terminated by addition of 1 volume of 4 Y formic acid. Zero times were prepamd by addition of liver plasma mem- branes after formic acid. The percentage of ATP or AMP-PNP remsininx was determined bv chromstonraohv on thin laver eheetd of-PEI -cellulose F (see-"Experime&ai &cedure8"). - FIO. 2 Whl). Effect of addition of unlabeled ATP or AMP- PNP on io&tion of cyclic AMP-`P from ATP+aP. Liver plasma membranes (0.4 mg of protein per ml) wem incubated for 10 min at 30' in 0.06 ml of medium containing 0.63 my ATP+*P (31 apm per pmole), 6.0 my M&l*, 12 PM glucagon, 1.0 my EDTA, 26 mx Trls-HCl, pfi 7.6, an ATP-regenerating system consisting of 20 mn creatine pho8phati and 1 mg per ml of creatine kinaee, and the indicated aoncentrations of unlabeled ATP (O-0) or AMP-PNP (O - - -0). The reaction8 were terminated and the cyollo AMP-tip formed wu drtormlnrd u derorlbed in Table I. ra&l to dryness at room temperature under a stream of nitro- gen. The following solutions were added to the residue: 15 pl of 1 Y Tris-HCI, pH 7.6,15 ~1 of 0.1 Y MgCI,, 30 ~1 of amino acid mixture (see above), 20 ~1 of 50 mu PNP, and 40 ~1 of E. coli extract. The reaction mixture was incubated for 1 hour at W', followed by the addition of 10~1 of a suspension of partially puri- fied rat liver plasma membrane (4) containing 10 mg of mem- brane protein per ml and further incubation for 15 min at 30". The latter 8tep mrved to degrade residual labeled ATP by nu- oleotlti pmmnt ln liver membrane8 (8) ; AMP-PNP is only slightly bydrolyred in tbe presence of liver membranes (see `%lCUJUlt(l"). All eubeequent procedures were carried out at 5". The above reaction mixture was applied directly to a column (0.4 X 4.0 cm) of DEAE-cellulose (Wbatman DE-52) that had heen previously washed with 10 ml of 2 m ammonium formate and then washed with 20 ml of distilled water. The column WBO eluted with a linear gradient formed from 15 ml of water and 15 ml of 0.5 Y formate, pH 7.4, at a flow rate of 0.5 ml per min; 0.5.ml fractions were collected. Labeled AMP-PNP appeared in Fraction8 42 to 47 which were pooled and treated as follow8 to remove salt. Washed analytical grade Dowex 50 (H+ form), 0.5 g (wet weight), was added to the pooled fractions. After stirring for 2 min in an ice h&h, the mixture wns filtered, the residue was washed twice with 10 ml of cold water, nnd the combined filtrates were lyophi- limd. Labeled AUP-PNP was dissolved in 0.5 ml of 25 my Tris-HCl, pH 7.6, nnd 8tored at -10". Radio purity of A%$P- PNP'-a-wP WM determined hy thin layer chromatography (de- scribed below) nnd ranged from 96 to 987,; the major contami- nant is ADP-NH*. BBsed on the amount of radicktctive ATP added to the incuhtion mixture, yields of labeled AMP-PNP ranged from 60 to 75%. Sepodon o$ Nucleolida by Thin Layer Chruntdogmph~De- terminations of chnnges in conccntntion of lnbclcd ATI' or AMP-PNP during incubation with liver meml&anes were carried out by thin layer chromatography (ascending) on precoated, alu- minum backed sheeta of PEI-cellulose F (Brinkmann 6@20310-4). A solution of 0.5 s ,LiClS M formic acid wns used as devel- oping salvent. With this solvent system the RF value for ABIP- PNP is 0.68; ATP, 0.51; AI)&NH*, 0.78. ADP doe8 not s&a- rate. from AZIP-PNP in this system but readily separates on PEI-cellulose chromatography using 1 Y Tris-HCl, pH 7.6. La- beled nuoleotider were cochromatogmphed with unlabeled nu- cleotider which were detected under ultmviolet light. The areas containing the nucleotides were cut out and placed in counting vial8 containing 10 ml of Bray's scintillation fluid (9). Radioactivity wa8 determined in a liquid scintillation counter. RESULTS In previous studies (4), measurements of adenyl cycl8rze activ- ity in rat liver plasma membrane8 were carried out with 3.2 my ATP a8 sub&rate. At this concentration, ATP affecta binding of glucagon to about the 88rne extent as 1.0 PM GTP (1). Beduc- tion of ATP concentmtion to 0.2 lllld as a means of minimking ita effect8 on glucagon binding resulted in rapid hydrolysis of the nucleotide despite the presence of an ATP-regenemting system, a8 ill&rated in Fig. 1. AMP-PNP-a-"P, at 0.2 mu, wa8 hy- droiysed by liver membranes to a slight extent (shout 16%) dur- ing 20 min of inauhation with liver membrane8 in the nhsence of an ATP-regenemtlng 8y8t@m. The rlight hydrolydr otmerved xalleofMuch25,1871 Y. Ratbell, L. BirnbawMt, 67. L. Pahl, and H. M. J. Kmna 1879 TABLB I TABLB II OTP ea jumdion oj qdic AMP jmm AMP-PNP Livu zMmbr85w (0.4 mg of protdzl par ml) were izloubBtBd for ?? ??* o ? av ill zzlaKum aozlw 0.2 my Abe-PNP- a98 epmpr?~),41)~~~,1l)myEDTA,1~0my~i0AMp, 0.2% 8lbuzlzin, 26 my TriB-ml, pH 7.6, NW3 tlm india~tod uldl. tbom. FiMlwlulwwo.lBzzll. TbamBotionowmtorzoiMtod by tin 8dditioa of 0.10 ml of mbzt&o oont8inizlg 40 ZnM ATP, 123 my oyolio Ab_P-`H (spprodzwdy WOO opm), and 1% adiumdodooylmllfate,followed~- boiling for q min. cyolh AMP* formed WY hdatad aooordh( to KrimllM, weia, azd Brod& (ll) Y damibod pmdody (11, IS). f3pgeif~ oj &ion q? "~limNlolal Baawl Liwr plvnu uxmbmna (0.36 mg of probin pu ml) were inau- M in 0.a Id of medium coat&dng 0.24 lily AMP-PNPW.=P (260 cpm par pzwlr), 2.6 my MgCl,, 1 my EDTA, 1 my cyolio AMF, Owe albuzoin, 26 zzw TrWiCl, pH 7.6, I ry glum, and the indhted dditions. Incubations bare for 10 min at 80'. The remtio~ wro torzniM and cyclio AMP- formad wu deknnined Y daoribed io Table I. Adam Now GTP (0.01 ml) GDP (0.01 my) GMP-PCP (0.1 mu) UTP (0.1 mu) CTP (0.1 mm) M4ltla NOW GTP @.a zzuz) 1000 000 ii (r ? 000 I 2 400 G f L 200 OLUCAOON 5 IO IS 20 MINUTES @`IO. 8. IQTect of fluoride ion, &wagon, tid combination of glwagon md OTP on formation of aydio AMP from AMP-PNP. Liwr plamzn~ menub- (oas mg of pztaoizz per ml) mm inou. bated 8t W ia 0.011 ml of zwdiuzn aontddzhg 0.2 z85z AMP-PNP- asP~oOcpm~pnds),4.Omy~1,1myHITA,1my oyolio AMP, 02% albumin, aa my TrWICl, pH 7.6, md 01th lOmyN~pl~6~~~nor6pl~\#yonpluO.Olmu yP.~ rowt~oB8 wB5B kmrlB8tod Bzld tbo oyolio AMP-V dhnvhodwda~IbodinTablo1. nmy baw born dus to nuobotide triphoophab pyrophoeph. lpizobs rqmtul in liwr membmaem (10). Add&b of myiug oo11d(mfntioIu of AMP-PNP and ATP to tbs UIJ meHum aontriniry ATPeW (0.6 my) reulted in proposfionrlraduatioPlint&smouatofkbsbdoyclioQ',s'-AMP ~b~~apohrs(Ft.2). *GnAinaw mBAsAL P / *-4 CTP / b-b OLUCACON /' c-a OTP . OLUCAOON / / MINUTES Fro. 4. E&et of tizw of addition of glwagon o nd GTP on formation of ayclio AMP from AMP-PNP. Liver plamn8 mem- branea (0.4 mg of protein per ml) mre iwubhd at 3(r in 0.M ml of medium conthing 0.24 mm AMP-PNPa-aP (l&3 cpm per pmolo), 4.0 mu MgCl~, 1 mu EDTA, 1 my cyclic AMP, 022% albumin, and 25 my Tria-HCl, pH 7.6. The following additiona worn nuQ at either rem time or after 5 min of iwubhon: I * glwwn, 0.01 my GTP, md I ry glucagon phm 0.01 my GTP. Tbo reaction0 were termin8ted 8t tba indiatad time6 and the ayolia AMP-aP formed w80 datermined am dawibed in Table I. that ATP md AMP-PNP behsved ident&@ M eubstrrrte for ~Yld=* IzloubB~ of AMP.PNP.@P with liver meIn- ruult8d llltbBfonMtionof l&&d qdia 3',S'-AMP whiah wan stimuMed by glucagon md fluoride ion (`Iable I). Tbh provided direct ovidemo~ tlmt AMP-PNP w M 6ubetxate for rrdsnyl cyclaae~ It will be noted alao that GTP (02 mu) otimulated bmal aathity by~%and~~a8d-foldrtimubtionoftbe~oftbe Bzl8po zlptozn to glue8gozl. The guzmyl nuclBotidB izlbibitBd tbeflumidereapmmbyin%. Glucqp (6.0 oly) and fiwride ion (10 mx) huhted tbo Vol. 246, No. 6 0 $1 II , I I I I I I 0 to-9 10-6 10'7 10-6 10-5 10'4 10-3 [GTP OR GMP- PCP] M Fro. 6. Effect of varying concentrations of GTP and GMP- PCP on rdenyl oyalaee aotivity determined in the absence and in the presence of glucagon and fluoride ion. Liver membranes wra incubated in the presence of 6 ry glucagon &o&en lines) with GTP (epes circler) or GMP-PCP (elmed circlea). Adenyl ayolsee activity in response to NaF (10 my) was investigated only in the preeence of GMP-PCP. Bas8l activity was studied in the presence of GTP or GMP-PCP, both of which gave identical aativltiea. Other incubation conditions oonslsted of 0.21) rn~ AMP-PNP&`P, 2.0 my M 0.3% albumln, md 26 my F Is, 1 rnx EDT& 1 my oyolia AMP, rL-HCI, pli 7.8. production of bbeled oyclio 3',6*-AMP at a line8r rata for 8t bad 10 min at 20" (Frg. 8); GTP enlmnced the initial r&e of gluwn-stimubbied 8ctivity. GTP or GDP, eqrmlly at 100 PM, stimulated the responss of 8thXl)`l OjdM6 t0 f&l~OlI ~&'8th8tlO-fdd When t&3 Mp aowfmtration WIU zduced to 2.6 my. Thin is shown izt Table 11 (cj. Table I where the Mp concentration was 4.0 mu). The rob of m8gnesium ion iu the process of activation of the ensyme by glucsgon and gusnyl nucleotides ls under ourrent investig- tion. We have shown in a previous study (4) that the Mg++ conu&r8tion is critic81 for the response of the ensyme to gluca- gon. UTP and CTP, at 160 CM exerted smaller stimulstory e&eta on the response of 8denyl cycl8se tu gluc8gon. An showu in Fig. 4, GTP (or not shown, the other grmnyl nu- obotid66) A0 &imuhil kul8denyl oyokue activity (sbssnos of glucagon or fiuorkia ion) in the prenenca of 4.0 my Mg++; uzhr tbew ounditlons b8s81 rurtlvity in the pmsenos of GTP w8s about equ8l to tb8t observed with gluc8gon (6 PM) alone. As ahowzz, iztoubzhitm of the membranes for S min st 30" prior to addition of GTP or gluc8gun rusulied in 8 marked decm8se in the e&at of GTP on lnuntl activity lrnd in eeaentially no eife& of glucagun alone; addition of GTP war required to obtain 8 gluts- gon response. The loss of response of 8denyl cyclase to glucagon dter the S-mm incubation period ~8s not due to 8 olmnge in bind. ing of glua8gou; upt8ke of W-glucagon (4.6 n.u) ~8s found to be identical before 8nd 8fter incubation of the membnrnes under tb conditions deuxibed iu Fig. 4. The b8sis of the loss of responss Of 8denyl cycllvll! to glucsgon or tlw dimhbbed dhtof GTP on besal activity attar incubation is not known but may bs r&ted, respectively, to destruction of endofpmous, membrane-bound gu8nyl nucleotides by nuoleotid- ues md of membrane-bound gluoagon by gluoagon-inaotivatlng TABL~C 111 EJect oj lrculmcnf oj liacr plamna membmnee toilA pho5pholipa5r A on inhibitory e&m oj QMP-PCP on #wride-rlim~claltd adenyl eyelaw activily Liver plasma membranes (1.2 mg of protein per ml) were iacu- bated in the aheence and preeenoe of 160 unite per ml of phospho- lipase A in medium containing 1 mu CaCl~ and 25 mx Tris-HCI, plf 7.6. Incubation WM for 6 min at 30'. Treatment was ter- minated by addition of 0.1 volume of 25 rnx EDTA and immediate cooling to 0.. Adenyl cyclsse activity was subsequently deter- mined on 10.pl aliquote in the abasnce or in the presence of 0.1 mx GMP-PCP in 0.a ml of medium containing 0.24 mn AMP-PNP- e-*nP (195 apm per pmole), 2.5 my MgClr, 1 my EDT& 1 my cyclic AMP, O.Wo albumin, 26 mu Tris-HCI, pH 7 6. and 10 mx NaF. Incubationa were for 10 min at 30o. The resctioos were ter- minated and cyclic AMP-MP fnrmed was determined M described in Table I. cydkAMPfonmd Additba durbsg tratment Inhibit&m due I loGMP-PCP Control CUP-PCP #adU/al# pdir x None. . . . . . . . . . . . . . . . . . . . 173 I 113 35 Phospholipese A. . . . . . . . . . SC 53 3.3 process or processes; both destructive activities are present in liver membranes (2). In the previous study (l), it was shown that GTP and GDP aiter biding and dlssoaWon of bound glucsgon over a concen- tdC%I ntIt@ Of 0.06 pII &J M) CM; Gh'fP-mP 8b3 dbd thses proce%aea but at higher oonoentmtions. Maximal effects of the anslogue were observed 8t 160 fiy. Iuvestig8tions of the re- sponse of the Bdenyl oyolsee system to glucagon, illustmted in Fig. 6, revealed that guanyl nucleotides exert their eftecte on this system at eomewhst lower concentr8tioncr than were observed in the biding studies. Siii6oant effects of GTP and GMP-PCP were ohse.rved at 0.01 ~AI and 0.05 PM, reslxxtively. As was observed in the binding studies, maximal effects of GMP-PCP on glucagon zesponea wem%een at loo &ru. Since GTP inhibite the nxponse of the ensyme system to fluoride ion (Table I), it was of interest to determine the effecta of GMP-PCP on this process. As shown in Fig. 5, the aneloguo of GTP inhibited the fluoride response st conoentmtions similar to those at which it stimulated the response of the enryme to gluc8gon. Treatment of liver membranes with phospholip8ss A resulte in complete lees of gluc8gon binding or motivation of adenyl oycl8ss but in retention of the fluoride responss (2,5). Under conditions iu which the glucagon response of sdenyl oyclase ~8s abolished and the fiuoride response w8s reduced by 56% phospholipnse A trestment did not alter the inhibitory elfects of GMP-PCP ou the responss of adeny cyohxse to fluoride ion (Table III). Studies were also carried out to determine whether gu8nyl nu- cleotides alter the concentration of gluc8gon required for half- maximal response of adenyl cycbxse, which ~8s previously shown to be 0.004 NY (4). In the presence of 0.1 mur GTP, adenyl cy- ol8se activity in response to 0.005 prd gluc8gon w8s 66% of th8t given by glucsgon at 5 PM, indicating that GTP does not influence the apparent affinity of adenyl cyclase system for the hormone. It c8n be concluded from these findings that guanyl nualeotides and gluc8gon interact with the sdenyl cyclsse system in 8 non- a0mpetitive faohion. Iam1eofMar&25,1871 Ad. Rdtll, L. Birnbaumet, S. L. Pohl, and H. M. J. Krana D1llC!t7t3BION The purpo8e of the pre8ent 8tudy wa8 to ascertain the effect8 of guanyl nuoleotidea on the re8ponw of adenyl cycla8e to gluca- gon. The um of AMP-PNP M ~~b&ute for adenyl cyclase faoilitated &ii study. Compared to ATP, AMP-PNP wa8 only slowly hydroly8ed during incubation with liver membranes, thu8 drcumventing the usual problem8 encountered in maintaining mbtmto OoncentWion during khetio enaly8b of rdenyl oychnm a&My. For thi8 rvmum, addition of ATP-regenerating systems, which may complicate snaly8i8 becawe of their possible influence on the adenyl cychwe system, can be avoided. Initial mte8 of adenyl cycla8e activity were maintained for at lea8t 10 min; ehange8 in rate that occur& can be attributed to factors other than change8 in substrate concentration. Indeed, the main advantage8 of using AMP-PNP in this study wa8 that effect8 of nucleothbm on adeny cyclam activity could be d, at low concentration8 of rubstrrrte, with certainty that the effects were not due to changes in 8ub8trate concentration. A major findiig WM that guanyl nucleotides (GTP, GDP, and GMP-PGP) stimulated, by an immediate action, the response of adenyl cychme to glucagon. Hormonal responm WM dependent upon the presence of guanyl nucleotides either when the mag- neaium ion concentration wa8 lowered from 4.0 to 2.0 my or after incubation of the liver membmnes. The bssi for the8e condi- tionn leading to dependence of hormone response on guanyl nu- oh&de8 is under current investigation. An important point L that the response of adenyl cyclsse to glucagon required aa little a8 10 ny GTP. Only the guanyl nucleotides stimulated the hormonal response of adenyl cycla88 at concentrationa le88 than 0.1 nur, indicating that the hormone-activation process is reia- tively epecific for guanyl nucleotidea. The low concentrations required, the equivalent effects of GTP and GDP, and the eimilar action of GMP-PCP, a nonphosphorylating unalogue of GTP, on the rwponre of adenyl oyolsw to glucagon sugge8t that the guanyl nucleotideu regulate this process through binding but not through phosphoryhtion. The premnt studies were initiated because of the previous finding8 (1) that guanyl nucleotides affected the biding site8 for glucagon in a manner that re8ulte in enhanced dissociation of bound hormone, and in decreaeea in both the apparent affinity of the biding Bite8 for glucagon and in the amount of gluaagon bound. The guanyl nucleotides stimulated relea8e of the total quantity of glucagon bound, indicating that all of the biding dter wem a6ected by nucleotides. It was ale0 shown that the biidiig aitss were specific for glucagon and that these site.8 have the 8ame apparent affinity for glucagon a8 doe8 the proeesa in- volved in activation of adenyl cyclaee (2). It WBB ale0 found that inhibition of biding by euch agenta M urea, detergents, and phoepholipaee A lead8 to inhibition of activation of adenyl cycla8e by glucagon (2). Taken in f&o, the previous studies indicate that all of the bindmg dtes share ahamoteristiar with the initial procea, through which glucagon acti\-ate8 adenyl cyclase and can be appropriately termed `%criiinator," a8 defined pre- vioualy to be that material which is involved in the primary ac- tion of glucagon (2). It remain8 unknown how binding of gluca- gon to it8 discriminator, which appears to be a separate entity fmm adenyl cycla8e (S), i8 potentially translated into activation of adenyl ~yolam. The finding8 that guanyl nudeotides alter both the biding of ghrcagon and the response of the enzyme to the hormone offer another mean8 of relating biding to the hormone-activation 1881 proce88. There appear8 to be 8ome relationship between the effects of guanyl nucleotides on glucagon biding and hormone activation a8 evidenced by the following correlations. (a) Both proeeme8 are relatively epecific for guanyl nucleotides; UTP, CTP, and ATP act only at coneFntmtions greater than 0.1 mM; (a) GTP and GDP are equally effective, GMP-PCP is relatively lees effective on hormone binding and activation; and (c) the nucleotider exert their effect8 on both processes independently of the concentmtion of glucagon in the medium. Such correlate8 do not eetablidr how them etiecta of the guenyl nucleotides are interrelated, but indicate that both processes are. sffected spe- cifically by guanyl nucleotide8 through eite8 that are independent of the eitee reacting with glucagon. Since, under appropriate incubation conditions, both glucagon and guanyl nucleotide8 are required for activation of adenyl cyclaee, it appear8 that there are two regulatory sites, requiring the binding, respectively, of glucagon and guanyl nucleotides for activation of the enlyme. In a previous study (g), it wa8 ehown that gbacagon and fluoride ion activate adenyl cyclnae in liver membranes through pmces8ea that have difIerent characteristics, indicating that fluoride ion doe8 not opemte through the discriminator. It was of interest, therefore, to find that guanyl nucleotides, specifically and at low concentrations, inhibited the re8ponse of adenyl cyclaee to fluo- ride ion. GMP-PCP inhibited the fluoride rmponse over the 8ame range of concentmtions 08 it 8timuhQed the response of adenyl cyclam to glucagon. Furthermore, GMP-PCP inhibited the fluoride reeponee in liver membranes treated with phospho- lipam A under condition8 that abolished both glucagon binding and activation of adenyl cyclase by the hormone. The specificity of action of guanyl nucleotides and the similar concentration range8 over which the nucleotides alter glucagon biding, hor- monal activation of adenyl cyclase, and inhibition of fluoride re- aponae euggest that the effects nre related, possibly by a common action. However, the site and mode of action nf guanyl nucleo- tidea on this complex adenyl cych~ system remains obscure. These rtudies were initiated on the premiee, widely held, that hormone receptorskceive and transmit information imparted by the hormone to ita target cell depending only upon the circulating level8 of the hormone. Thii premise seems untenable for gluca- gon in view of the finding that guanyl nucleotides play an obliga- tory role in regulating the m8ponse of liver adenyl cyclase to ghrcagon. Further studies of the actions of the gunny1 nucleo- tide8 on the liver and other ndenyl cyclase systems may provide new insighta into not only how hormones regulate target cell metabolism at the receptor level but also how a target cell me- t&&e reguh~tes the initial response to a hormone. REFERENCES 1. RODBELL, M.,KRANB, Ii. M. J., POHL, 8. L., AND B~RN~AUMER, L., J. Biol. Chem., 246, 1301 (1971). 2. RODBCLL,M.,KRANB, N. M. J., POHL, 8. L., AND BIRNIIAUYER, L., J. Biol. Chm., 246, 1872 (1911). 3. RODBELL, M., BIRNBAUHER, L., PORL, S. L., AND KRAN~, H. M. J., Ada Diobefol. Lot., 7 (Suppl. l), 9 (1970). 4. POHL, 8. L., BIRNBAUYER, L., AND RODBELL, hl., J. Biol. Clcrn., 346. 1349 (1971). 6. BXRNBAUNER. L.. Pnm, S. L.. AND RODRELL. hf.. J. Biol. Chm., 3uJ,lf!&7 (1971j. - 6. NI~~NBIF~O, M. W., in 8. P. COLOWXK AND N. 0. KAPLAN (Editon), &felhda in mrymology, Vol. VI, Aondemio Pre8.8, New York, 1983, p. 17. 7. ZAYECNIS,P.C., AND STCPRENSON,M. L., in II.M.KAL?KAR, H. KENOW, A. MUNCH-PCTERBON,M. OTTESEX, AND Y. H. ;& ~AdrnylCyclars V Vol. 246, No. 6 ..`. . . I. !, .: ;`b:; TATS& (Edltom), 9% TO& oj mamuda :..dtw@nnUa~auymn,dedrppioProm,N~Yort, jw th lametim 10. LX~~MAN, X., LAHUIIO, A. I., rm Lrws, W. E., J. Bid. CAmIt., al_# 726 (1267). 11. KBUENA, G., Wmm), B., AHD Bnoxm, 8. B., J. Phwmuuol. a B!%&?ii L AIID lboum~, P., ia A. 3. Dumm AND P.&&ton),-i"(I-, Bsp. zb., 162,279 (ma). Vd. 1, Adda Prr, Now York, 1006, p. 22. 12. mDBBLL, M., J. 81ol. Ckm., w, 5744 (lQ67). . 0. Bur, Q. A., Aad. -., 1,279 (MW). l2. Xaumu, G., AND BmoAUM~, L., And. Bloclkn., Y, 292 (Im. `.