Tus Jouanat or Brotoorcat Cuemistay Vol. 246, No. 6, lasue of March 25, pp. 1849-1856, 1971 Printed in U.S.A. The Glucagon-sensitive Adenyl Cyclase System in Plasma Membranes of Rat Liver I. PROPERTIES (Received for publication, October 7, 1970) StrerHen L. Pont, Lutz Birnpaumer, aND Martin Roppevi From the Section on Membrane Regulation, National Institute of Arthritis and Aletabolic Diseases, National Institutes of Health, Bethesda, Maryland 20014 SUMMARY The liver parenchymal cell plasma membrane preparation devised by Neville (Biochem. Biophys. Acta, 154, 540 (1968)) contains an adenyl cyclase system which is stimulated by glucagon and to a much lesser extent by epinephrine. The yield of adenyl cyclase can be greatly increased, at the ex- pense of some contamination by other organelles, by increas- ing the amount of starting material and eliminating the last step in the preparative procedure. Adeny! cyclase activity is 25-fold purified in the membranes compared to a liver ho- mogenate. Liver membrane adenyl cyclase activity is a function of glucagon concentration over a range of 10~'* to 1077 m glu- cagon and is increased more than 10-fold by a maximally stimulating concentration of glucagon. The stimulation produced by epinephrine is less than 10% of that produced by glucagon. Secretin, a polypeptide hormone very similar to glucagon in its primary structure, does not stimulate liver membrane adeny] cyclase activity. Insulin changes neither the basal nor the glucagon-stimulated adenyl cyclase activity. Adenyl cyclase activities, under standard assay conditions, are proportional to time of incubation and to membrane con- centration. The enzyme requires a divalent cation, either Mgt? or Mn*+, but is inhibited by Ca++. At sufficiently high concentrations, both ATP and Mg** inhibit the enzyme. Addition of 1 mm EDTA to the assay medium doubles the glucagon-stimulated activity. Fluoride ion stimulates adenyl cyclase activity but to a lesser extent than does glucagon. Glutaraldehyde, N-ethylmaleimide, and -chloromercuri- benzoate inhibit the adenyl cyclase system. Urea, at a con- centration of 2 Mm, reduces both the maximal glucagon- and fluoride-stimulated activities, but at 0.4 Mm reduces only the apparent affinity of the system for glucagon. Attempts to solubilize and purify the adenyl cyclase activity in an active form from rat liver membranes have been un- successful. Adeny] cyclase in mammalian tissues is a membrane-bound enzyme system which catalyzes the reaction (1, 2) ATP — cyclic AMP! + PP, Certain polypeptide hormones and catecholamines exert at least part of their effects by stimulating the activity of this enzyme in their target cells (3). However, the mechanism, in molecular terms, by which these hormones stimulate adenyl cyclase activity is poorly understood. In a previous communication (4) we reported the presence of a glucagon-sensitive adenyl cyclase system in parenchymal cell plasma membranes prepared from rat liver by the procedure of Neville (5). We have subsequently found that the yield of adenyl cyclase can be greatly increased, at the expense of some contamination by other organelles, by omitting part of the Neville procedure. This modified preparation has several desirable features as a starting material for studies of the mech- anism of hormonal stimulation of adenyl cyclase. The purpose of this paper, the first in a series (6-9), is to describe pertinent general features of the liver membrane prepar:tion and _ its adenyl cyclase system. EXPERIMENTAL PROCEDURE Materials Glucagon (crystalline) was supplied by Lilly. Secretin was a gift from Dr. Victor Mutt (Karolinska Institutet, Stockholm). Epinephrine was purchased from Parke-Davis and was used within 15 min of opening the vial. Highly purified glutaralde- hyde was obtained from Ladd Research Industries, Inc., Burling- ton, Vermont. Lubrol was obtained from Imperial Chemical Industries, Providence, Rhode Island. The sources of all other reagents were specified previously (10-12). Methods Plasma membranes were prepared from livers of male and female Sprague-Dawley rats weighing 140 to 180 g according to the procedures devised by Neville (5). The gradient for Step 13? of the procedure was a 25-ml linear gradient from 26 to 1% sucrose prepared with a Beckman Gradient Former. These membranes are referred to below as “fully purified mem- branes.” “Partially purified membranes” were prepared by ! The abbreviation used is: cyclic AMP, cyclic adenosine 3’,5’- monophosphate. 2 The step numbers used here are those used by Neville (5) in his description of the liver membrane preparation. 1849 1850 doubling the amount of starting material and the volume of each of the steps, and omitting the final rate-zonal density gradi- ent centrifugation (Steps 12 to 15). In some experiments, one-half of partially purified membrane pellet was carried through the final centrifugation and used as fully purified membranes. All membrane preparations were either used immediately or were stored immediately in liquid nitrogen as compact pellets (25,000 *g). For routine use, one or two batches of membranes were suspended in 1 ma KHCO; at a concentration of 8.0 to 18.0 mg of membrane protein per ml and distributed as 0.1- or 0.2-ml aliquots in individual glass tubes. These tubes were stored in liquid nitrogen and thawed individually as needed in order to avoid repeated freezing and thawing of the entire batch of mem- branes. Adenyl Cyclase Assay—Adeny] cyclase activity was measured by the method of Krishna, Weiss, and Brodie (10) as described previously (11, 12). Unless specified otherwise, the following conditions were standard. The assay medium contained 3.2 mam ATP-a-"P (25 to 50 cpm per pmole), 5 mm MgCl, 1 mat EDTA, 25 mM Tris-HCl, pH 7.6, an ATP-regenerating system consisting of 20 mm phosphocreatine and 1 mg per mil of creatine phosphokinase (20 to 50 units per mg), and 10 to 50 wg of mem- brane protein in a volume of 50 yl. Hormones were diluted in the creatine phosphokinase solution. Incubations were initiated by the addition of the membranes and were continued for 10 min at 30°. Reactions were terminated by boiling for 3} min. Protein was measured by the Lowry procedure (13) using bovine serum albumin as standard. “Marker” Enzymes—5'-Nucleotidase (EC 3.1.3.5) was measured by the method of Bodansky and Schwartz (14), glucose 6-phosphatase (EC 3.1.3.9) by the method of Swanson (15) with addition of 1 mm EDTA, alkaline phosphatase (EC 3.1.3.1) by the method of Heppel (16), and acid phosphatase (EC 3.1.3.2) by the same method using 100 mm sodium acetate buffer, pli 5.3. ‘These four enzymes were assayed in 0.1 ml of appropriate medium and incubations were at 30°. Incubation time varied between 10 and 90 min depending on enzyme fraction and activi y measured. Activities were proportional to enzyme concentration. Reactions were stopped by addition of 0,025 ml 50% trichloracetic acid. The precipitate was removed by centrifugation, and the liberated P; was determined on 0.05-ml aliquots by the method of Fiske and SubbaRow (17). Suc- cinate-cytochrome ¢ reductase was measured by the method described by Fleischer and Fleischer (18). Electron Microscopy—Liver membrane suspensions were centrifuged in 0.4-ml centrifuge tubes for 15 min at 25,000 x g. The pellets contained about 1 mg of + -otein and measured about 0.6 X 0.2mm. The pellets were fixed at 0° for 30 min in 3% glutaraldehyde in 0.1 « sodium phosphate buffer, pI] 7.4. The fixative was removed and the pellet rinsed for 1 hour at 0° with buffer. Post-fixation was then performed for 2 hours at 0° in 2% osmium tetroxide in phosphate buffer followed by dehy- dration in cold acetone. The pellets were then embedded in epoxy resin (19) and oriented so that complete top to bottom sections could be obtained. Sectioning was performed with a Reichert Om U 2 microtome. hydroxide solution (20). Electron micrographs were obtained with a Philips EM 300 electron microscope employing a 40 pu objective aperture and an accelerating voltage of 80 kv. Expression of Results and Terminology—In all figures and tables “adenyl cyclase activity” refers to nanomoles of cyclic Glucagon-sensitive Adenyl Cyclase System. I Sections were stained in a lead - Vol. 246, No. 6 AMP formed in 10 min per mg of membrane protein. Basal activity is the activity measured in the absence of glucagon or fluoride ion. Glucagon-, ep:nephrine-, and fluoride-stimulated activities are activities measured in the presence of these com- pounds, RESULTS Membranes prepared by the complete and the modified pro- cedure of Neville (5) are referred to below as fully and partially purified membranes, respectively. The yield of fully purified membranes is 0.5 to 0.8 mg of membrane protein per g, wet weight, of liver. The yield of partially purified membranes is 1.0 to 2.0 mg of membrane protein per g, wet weight, of liver. The purity of membrane preparations was checked by electron microscopy and by assay of marker enzymes. The fully purified plasma membrane preparations are composed mainly of mem- brane sheets (Fig. 1A) usually arranged in pairs joined by various kinds of cell junctions (Fig. 1, B and C). In addition, a rela- tively small number of vesicles are present (Fig. 14). No other organelles were recognized. The paired membrane sheets with typical intercellular junctions and the bile canaliculus structure (not: shown; sce Reference 21) are unequivocally identified as originating from the plasma membrane of hepatic parenchymal cells (21). The origin of the vesicles is less certain. However, the suggestion has been made that they represent fragments of the “blood front” plasma membrane (22), the region of the plasma membrane not immediately adjacent to another paren- chymal cell. The partially purified membranes differ from the fell purified membranes mainly in the larger number of vesicies relative to sheets (Fig. 1D). In addition, occasional mitochondria and a significant number of small, densely stained, unidentified con- taminants (Fig. 1D) were scen. The marker enzyme data, summarized in Table I, demon- strates 2 substantial increase in specific activity of 5’-nueleo- tidase, a plasma membrane marker (22), in both the partially and fully purified membranes relative to the crude homogenate. The glucose 6-phosphatase and succinate-cytochrome c reductase activities indicate significant microsomal and mitochondrial contamination of the partially purified membrane preparation and that this contamination is substantially reduced in the fully purified membranes. Purification and Yield of Adenyl Cyclase in Membrane Prepa- rations—Table II summarizes the purification and yield of basal, glucagon-, epinephrine-, and fluoride-stimulated adenyl cyclase activities in partially and fully purified plasma mem- brancs compared to a crude homogenate of liver. The glucagon- sensitive adenyl cyclase activity is obtained in greatest yield and specific activity in the partially purified membranes; a 25-fold purification is obtained in this material. The fully purified meribranes have both a lower yield and lower specific activity of glucagon-sensitive activity. Both preparations have several-fold higher specific activity than a 1500 x g pellet of the liver homogenate. Epinephrine-sensitive adenyl cyclase is present in both membrane preparations at much lower ac- tivity than glucagon-sensitive activity but is also substantially enriched in comparison to the homogenate activity, Stinnidation af Adenyl Cyclase Activity in Membrane Prepara- tions by Hormones and Fluoride Ion-—Glucagon produces by far the largest stimulation of the liver membrane adeny! eyclase activity of any agent which we have tested; the stimulation Issue of March 25, 1971 S. L. Pohl, L. Birnbaumer, and M. Rodbell 1851 Fra.1. Electron micrographs of mombrane PMS, plasma membrane sheets; V, vesicle, A, fully purified preparations, membranes, Magnification X 30,000. B, higher magnification showing junctions between two membrane sheets and typical unit membrane structure (arrow). Magnification x 150,000. produced by glucagon is from 10 to 30 times the basal activity. The activities reported here are greater than those in our previous communication (4) primarily because of the addition of EDTA to the assay medium. The effect of EDTA is described below. Epinephrine produces a smull stimulation of the rat liver plasma membrane adenyl cyclase activity (Table TI), an effect which we failed to observe in the preliminary experiments reported earlier (4). However, the epinephrine stimulation is only 1.5 to 2.5 times the basal activity or less than 10% of the stimulation produced by glucagon. A variety of other polypeptide hormones tested, secretin, C, detail of a tight junction showing two membrane sheets with fused outer leaflets (arrow), Magnification x 300,000. D, par- tially purified membranes. Arrow indicates unidentified densely staining contaminant. Magnification x 30,000. adrenocorticotropin, thyrotropin, luteinizing hormone, vaso- pressin, and parathyroid hormone, failed to stimulate the liver membrane adenyl cyclase system, The failure of secretin to stimulate is of particular interest since the primary structures of glucagon and seeretin are similar (23). Insulin at a con- centration of 10 munits per ml failed to change the basal or glucagon-stimulated adenyl cyclase activities. The liver membrane adenyl cyclase system is sensitive to low concentrations of glueagon. As shown in Fig. 2, adenyl cyclase activity is a function of glucagon concentration over a range of 107 to 1077 a with half-maximal stimulation occurring at 1852 about 4 x 107%. The maximal activity obtained with glucagon is greater in the partially purified than in the fully purified membranes. The other parameters of the dose-response re- lationship of adeny! cyclase activity to glucagon concentration are identical in the two preparations. Fluoride ion stimulates adenyl cyclase activity in the liver membrane preparations (Table II). The fluoride concentration dependence of this effect is similar in the partially and fully purified membranes (Fig. 3). The effect of fluoride ion will be treated more fully in a separate report (6). The liver membrane preparations may be stored in liquid nitrogen for at least 3 months without loss of glucagon- or fluo- ride-stimulated adenyl cyclase activities. Assay Conditions for Adenyl Cyclase Activity—Fig. 4 presents the time course of the glucagon-stimulated adeny] cyclase reac- tion with varying amounts of liver membrane protein assayed in media with and without ATP-regenerating system. It is apparent that under the standard assay conditions specified under “Methods,” #.e. 10 to 50 xg of membrane protein incu- Tasrz I Activities of marker enzymes in liver homogenates and in partially and fully purified membranes Rat livers were ground in a Dounce homogenizer as described by Neville (5). An aliquot of this crude homogenate was re- ground in a ground glass homogenizer to facilitate pipetting. The remainder of the homogenate was used to prepare partially and fully purified membranes as described under ‘“Methods.”” Specific activity in Glucagon-sensitive Adenyl Cyclase System. I Vol. 246, No. 6 bated for 10 min in a medium containing ATP-regenerating system, the glucagon-stimulated adenyl cyclase activity is proportional to both time of incubation and to membrane con- centration. This is also true for the fluoride-stimulated ac- tivity (data not shown), Without the ATP-regenerating system, the reaction is linear only for relatively short incubation times and at low membrane concentrations. The liver plasma membrane preparation contains a potent ATPase (24). Even with the ATP-regencrating system, the proportionality to time and membrane concentration is lost at membrane protein con- centrations above 50 yg per assay (50 ul). The reason or reasons ‘for this nonproportionality at high membrane concentrations is unknown, but several factors may contribute including ag- gregation of membranes, destruction of glucagon, inadequacy of the ATP-regenerating system, or production of an inhibitor. The pH of the standard assay medium, 7.6, is the optimum for glucagon- and fluoride-stimulated activities in three buffer systems, Although greater enzyme activities were measured at 37° and 42°, the time courses were nonlinear at these tem- peratures; therefore, 30° was chosen as the standard incubation temperature. Addition of 1 mm EDTA to the assay medium Taste III Hormone specificity of adenyl cyclase in liver membranes Partially and fully purified plasma membranes were assayed for adenyl cyclase activity in media containing no stimulant, epineph- rine, secretin, glucagon, or 10 mu NaF. Hormones, when pres- ent, were at 20 ng per ml. Activities are expressed as the mean + half the range of triplicate determinations. Adenyl cyclase activity Enzyme Partially Full Crude purifi purified Addition homogenate mesma meas Partially purified ' Fully purified | umole/min/mg protein wmoles/10 min/mg protein 5'-Nucleotidase.................. 0.032 0.535 0.632 None 0.30 + 0.06 0.16 + 0.01 Glucose 6-phosphatase.......... 0.038 0.021 0.012 Epinephrine 0.42 + 0.02 0.28 + 0.04 Alkaline phosphatase............. 0.0041 0.015 0.016 Secretin 0.34 + 0.02 0.23 = 0.04 Acid phosphatase. . i 0.019 0.008 0.006 Glucagon 3.33 + 0.02 2.76 + 0.02 Suceinate-cytochrome c reductase 0.025 0.028 0.003 NaF 1.92 + 0.08 1.61 + 0.95 Tasie IT Purification of adenyl cyclase Adenyl cyclase activities of the crude homogenate and the partially and fully purified membrane preparations described in Table I were measured under standard assay conditions except that 1 mm cyclic AMP and 10 mm theophylline were added to the medium. When present, glucagon and epinephrine concentrations were 20 ng per m! and NaF was 10 ma. Adenyl cyclase activity is ex: ivressed as nanomoles of cyclic AMP formed i in 10 min per mg of protein. Incr in adeny! cyclase Purification | Basal activity’ due to addition of Yield Preparation path oe Epi- activity Ghcagon neripine NaF pontd neph- NaF | Glucagon netivine NaF t ! y me | | % Homogenate............... 17,000 | 0.034 | 0.244 | 0.036 | 0.194 1 | 4 1 | 109 100 | «100 Partially purified plasma mem- | branes... cece ete eee 7 0.207 6.136 0.506 2.558 25 14 13 MH 3.5 8.0 Fully purified plasma membranes. 21 | 0.172 4.166 0.273 2.651 17 | 9 13 1.0 0.2 1.0 ¢ Basal activily subtracted. + The glucagon-stimulated activities in this experiment are somewhat greater than those in other experiments described here pos- sibly because it was performed several months later during which time the supplier of the ATP-a-*P made a significant change in the preparative procedure. Issue of March 25, 1971 S. L. Pohl, L. Birnbaumer, and M. Rodbell 1853 ADOED GLUCAGON (ug/ml) 3 OO 0 f 1 10 2. g e s 2 5 20 wey S tet CE 5 4 t ! or e 3 . 3 fi in g 5 . =~ 10 S/ fe ‘ o ww c 5 ff s o as| » sol- 4 | a ee a % - Bat » GTS S80 2 es Fig, 2 (left). Effect of glucagon concentration on adeny] cyclase activity in partially purified (©) and fully purified (@) plasma membranes. Fra. 3 (center). Effect of fluoride ion concentration on adenyl cyclase activity in partially purified (©) and fully purified plasma membranes (@). Fia. 4 (right). Time course of the glucagon-stimulated adenyl caused a doubling of the glucagon-stimulated activity and lesser increases in the basal and fluoride-stimulated activities (Table Ty). The effects of varying ATP and magnesium concentrations on adeuy] cyclase activity are shown in Figs. 5 and 6, respec- tively, Since the substrate concentration cannot be optimized for both glucagon- and fluoride-stimulated activities, 3.2 ma was chosen arbitrarily for standard assay conditions because it gives near optimal activities for both stimulants. In the presence of 3.2 ma ATP, 5 mu Mgt+ gives optimum basal, glucagon-, and fluoride-stimulated activities. Manganous ion may be substituted for magnesium (6) but calcium cannot and, at concentrations above 10~‘ m, is inhibitory in the presence of magnesium. A cyclic phosphodiesterase inhibitor, theophylline or caffeine, was not included in the standard assay medium because of our earlier finding (4) that the cyclic phosphodiesterase activity of the liver membrane preparation is so low as not to interfere with the adenyl cyclase assay under routine assay conditions with both partially and fully purified membranes. “Solublization” Experimenta—A variety of methods were employed to alter the membranes in such a way that the adeny] cyclase activity would remain in the supernatant for at least 1 hour in a centrifugal field exceeding 100,000 x g. These in- cluded sonic oscillation, succinylation, and exposure to high salt concentrations, urea, and detergents. All of these treat- ments either abolished the adenyl cyclase activity or failed to solubilise the enzyme. Using the Lubrol procedure devised by Levey (25), approximately 15% of the fluoride-stimulated activity was obtained in a high speed supernatant. However, this activity was insensitive to glucagon. Effects of Glutaraldehyde, Sulfhydryl Reagents, and Urea— As shown in Table V, glutaraldehyde treatment completely inac- tivates the adenyl cyclase system of the rat liver membrane preparation. Adenyl cyclase activity is also inhibited by treat- ment of the membranes with N-ethylmaleimide and p-chloro- mercuribensoate, agents which react with sulfhydryl groups (26).° In other experiments, the extent of inhibition by sulf- TIME (min) cyclase reaction. Partially purified membranes were incubated for varying times in adenyl cyclase assay medium containing 10 ug per ml of glucagon prepared with (A) and without (8) the ATP- regenerating system. Numbers on curves indicate micrograms of membrane protein used for each study. Assay conditions were otherwise the same as described under “Methods.” © and e, serve only to distinguish adjacent curves. Tasie IV Effects of EDTA on adenyl cyclase activity Partially purified membranes were incubated in standard assay medium with or without 1 mx EDTA. When present, glucagon was 10 ug per ml and NaF was 10 mu. Adenyl cyclase activity Addition —-EDTA +EDTA None 0.17 0.30 Glucagon 1.65 3.33 NaF 1.59 1.92 = .f = » ~~ bee g = 3 5 < | % 5 52 a g Bo] farmer tn, 3 * “Sp pa at < G10 203040 60 7 ©6009 5 10,, 15 20" 40 AIP (mi) Mg’ (mM) Frio. 5 (left). Effect of varying ATP concentration on basal, glucagon-, and fluoride-stimulated adenyl cyclase activities. Partially purified membranes were incubated in media containing no stimulant (0), 10 xg per ml of glucagon (©), 10 mu NaF (A), and varying concentrations of ATP. Other conditions were as described under ‘‘Methods.”’ Fria. 6 (right). Effect of varying Mg** concentrations on basal > glucagon-, and fluoride-stimulated adeny! cyclase activities. Partially purified membranes were incubated in media containing no stimulant (0), 10 wg per ml of glucagon (O), 10 mu NaF (4), and varying concentrations of MgCls. Other conditions were as described under “Methods.” 1854 Tasie V Effects of glutaraldehyde and sulfhydryl reagents on adenyl cyclase activity For treatment with glutaraldehyde, partially purified mem- branes, 28 yg of membrane protein, were incubated in 0.02 ml of a medium containing 50 mu cacodylate-nitrate buffer, pH 7.4, and 4.8% sucrose with or without 1% glutaraldehyde. After 10 min at 30°, adenyl cyclase assay reagents were added in a volume of 0.03 ml and adeny) cyclase activity was determined as described under “Methods.” For treatment with sulfhydryl reagents, fully puri- fied membranes, 18 yg of membrane protein, were incubated in 0.02 ml of a medium containing 25 mu Tris-HC], pH 7.4, and no addition, 1 mu N-ethylmaleimide, or 0.05 mu p-chloromercuri- benzoate. After 15 min at 22°, adenyl cyclase assay reagents were added in a volume of 0.08 ml. EDTA was omitted from the stand- ard assay medium, and 2-mercaptoethanol was included at a final concentration of 1 mu. When present, glucagon was 10 pg per mi and NaF was 10 mw. Adeny! cyclase activity in the presence of Treatment Basal | Glucagon NaF Control... 00... cece cece cece neces 0.78 4.86 2.96 Glutaraldehyde.................0... 0.02 0.01 0.04 Control... 0... cece ccc eee eee 0.08 2.41 3.21 N-Ethylmaleimide................... 0.03 1.69 2.20 p-Chloromercuribenzoate............. 0.00 1.26 1.49 25 r 7 r 7 @e + UREA E 20 7 F 2 yw bbe 4 < 3 > Sloe 4 a > a 0.5/ - ,_t °> tt 10"8 10"? GLUCAGON (M} Fia. 7. Effect of the addition of 0.4 « urea to the standard adenyl cyclase assay medium on the dose-response relationship between glucagon and adeny] cyclase activity. hydryl reagents was found to be nearly complete with longer incubation times, and addition of 2-mercaptoethanol or di- thiothreitol was found to block the effect of these agents. Incubation of the membranes in 2 a urea for 10 min at 30° reduces both the glucagon- and fluoride-stimulated activities by 60 to 80%. Addition of 0.4 m urea to the adenyl cyclase assay medium decreases the apparent affinity of the system for glucagon (Fig. 7). Other effects of urea on the adeny! cyclase system will be described in a later paper (7). DISCUSSION Adenyl cyclase activity is widely distributed in mammalian tissues (1, 3) and is controlled by a variety of polypeptide hor- Glucagon-sensitive Adenyl Cyclase System. I Vol. 246, No. 6 mones and catecholamines (3). Despite the central position of the enzyme system in hormone action, little is known of the mechanism by which hormones control its activity. It is mem- brane-bound and loses its response to hormones when attempts are made to solubilize and purify its components. For this reason, we have studied adenyl cyclase systems in intact mem- branes and have employed indirect means of characterizing the components (12, 27, 28), Previous studies of the adenyl cyclase system in fat cell ghosts indicated that at least five different hormones stimulate the activity of a single catalytic component through discrete re- ceptors (27, 28), and that the affinity of an allosteric site for magnesium may be involved in the mechanism of hormonal stimulation (12). Further study of the fat cell ghost system was hampered by the small amounts of this material that can prac- tically be obtained and by its relative instability. The finding of a glucagon-sensitive adenyl cyclase in the plasma membrane preparation devised by Neville was initially of interest because it added further evidence for the cellular location of adeny! cyclase in plasma membranes (4, 29-31). In addition, it can be prepared in large quantities and is stable for long periods of time in the frozen state. Furthermore, liver plasma membrane preparations have been characterized more extensively (5, 22, 32, 33) than any other source of plasma mem- branes containing hormonally sensitive adenyl cyclase. Fully purified liver plasma membranes consist of paired membrane sheets with junctions and the bile canaliculus struc- ture which unequivocally identify the origin of these sheets as the plasma membrane of hepatic parenchymal cells. Vesicles are also present which, according to Benedetti aud Emmelot (22), may be derived from the plasma membrane because they are isopyenic with the membrane sheets and because preparations containing vesicles have low specific activities for enzyme and chemical markers of other organelles. This suggestion is sup- ported by the report (33) that purified liver plasma membranes contain undetectable amounts of a unique glycolipid found in other types of liver membranes. Our finding that the fully and partially purified membranes, the latter being rich in vesicles, have the same specific activity of §’-nucleotidase, a plasma mem. brane marker (34), adds to the evidence that the vesicles are derived from plasma membrane. Adenyl cyclase activity is considerably enriched, relative to liver homogenates, in both partially and fully purified mem- branes. The similarity in the two preparations of the dose response curves for glucagon and fluoride ion suggests that omission of the final purification step does not introduce adenyl cyclase with different properties from that in the fully purified membranes. The glucose 6-phosphatase and succinate-cyto- chrome c reductase activities indicate a considerable contamina- tion of the partially purified membranes with other organelles, but because of the much higher yield of adenyl cyclase activity we have considered the partially purified membranes preferable to the fully purified membranes for our studies. It should be emphasized that the present studies do not indi- cate, particularly with the partially purified membrane prepara- tion, that adenyl cyclase is present exclusively in plasma mem- branes of parenchymal cells. There is histc-hemical evidence (35) that glucagon- and epinephrine-sensitive adenyl cyclase systems are present in plasma membranes of both reticuloen- dothelial and parenchymal cells. It is possible that the higher specific activity of the glucagon-stimulated adenyl cyclase in Issue of March 25, 1971 partially purified membranes represents contributions of both parenchymal and reticuloendothelial cells. Unfortunately, quantitative assessment of the contributions of glucagon-sensitive adenyl cyclase from the two cell types cannot he made from the histochemical study since gluiaraldehyde, used as fixative, inactivates adenyl cyclase in isolated membranes and causes substantial loss of activity in blocks or sections of liver used for the, histochemical studies (35). Studies of the hormone re- sponsiveness of isolated parenchymal and reticuloendothelial cells may provide a definitive answer to this question. Glucagon, of the several hormones and other agents tested, produced the greatest stimulation of adenyl cyclase activity in the liver membrane preparations. In contrast to our previous report (4), epinephrine also produced a stimulation of adenyl cyclase activity although to a much lesser degree than did glu- cagon. Bitensky et al. (36, 37) have provided evidence for two different adenyl cyclase systems in liver homogenates, one sensitive to glucagon and the other to epinephrine. Our finding that glucagon- and epinephrine-sensitive activities are not. in- creased by the same factor in partially and fully purified mem- branes compared to the homogenate is consistent with the hypothesis that there are two adenyl cyclase systems present in liver. Marinetti et al. (38, 39) have reported recently a much greater epinephrine stimulation of adenyl cyclase in rat liver membranes prepared in a different manner than in this study. They also reported several other observations which are opposed to our findings in the liver membrane preparation and the findings of other investigators (3) in different tissues. For example, they report inhibition by fluoride, activation by calcium ion, in- hibition by insulin, and activation of adenyl cyclase by p-chloro- mercuribenzcate. Finally, their preparation appears to be 100- to 1000-fold less sensitive to glucagon than our preparation. Fluoride ion stimulates the liver plasma membrane adenyl cyclase system as it does all other mammalian systems (3). However, under our standard assay conditions, the activity obtained with a maximally stimulating concentration of glucagon is about two times that obtained with a maximally stimulating concentration of fluoride ion. Fluoride-stimulated activity is not exceeded by hormonal stimulation in several other systems (3, 12, 40, 41) and the suggestion has been made that an ap- propriate concentration of fluoride ion produces maximal ex- pression of enzyme activity (1, 41). Inclusion of 1 mu EDTA into the incubation medium in the present study enhanced the response of liver adenyl cyclase to glucagon. In the following report (6) it will be seen that a number of agents selectively alter the response of liver adenyl cyclase to fluoride ion and glucagon. The adenyl cyclase assay conditions and method of measure- ment have been adequately described in publications from this and other laboratories (4, 10-12, 25, 38, 41) and do not require extensive comment. However, it is important to note that under our standard assay conditions, the measured activity is pro- portional to time of incubation and to membrane concentration. The liver plasma membrane preparation contains both a potent ATPase (24) and a system which rapidly inactivates glucagon in the assay medium (7). Consequently, any significant change in the composition of the assay medium requires at least that the linearity of enzyme activity with time be checked. Mammalian adenyl cyclase systems are highly complex and probably multimolecular. An understanding, in molecular terms, of the mechanism or mechanisms by which polypeptide S. L. Pohl, L. Birnbaumer, and M. Rodbell 1855 hormones control these systems depends upon identification, characterization, and isolation of the components of which such systems are made. The liver plasma membrane preparation, with its glucagon-sensitive adenyl cyclase system, provides a useful starting material for such studies. Acknowledgmenta—We are most grateful to Dr. David Neville for supplying some of the plasma membrane preparations, to Dr. Joan Blanchette-Mackie for performing the electron mi- croscopy, and to Mr. Thomas Demar for excellent. technical assistance. REFERENCES 1. SurHervanp, E. W., Ratt, T. W., axp Menon, T., J. Biol. Chem., 287, 1220 (1962). 2. Rau, T. W., anp Sutwertann, E. W., J. Biol. Chem., 287, 1228 (1962). .3. Ronison, G. A., Burcner, R. W., ann SUTHERLAND, E. W Annu. Rev. Biochem., %, 149 (1969). - Pout, S. L., Birnnaumer, L., aNp Rope, M., Science, 164, 566 (1969). . Nevitie, D. M., Biochim. Biophys. Acta, 184, 540 (1968). . Brrnpaumer, L., Pons, 8. L., anp Roprens, M., J. Biol. Chem., 246, 1857 (1971). . Ropes, M., Krans, H.M.J., Pont, 8. L., ann BinNnAUMER, L., J. Biol. Chem., 246, 1861 (1971). - Roppry, M., Krans, H.M.J., Pow, S. L., anp BrinNpaumer, L., J. Biol. Chem., 248, 1872 (1971). . Ropseiy, M., Birnaaumer, L., Pon, S. L., aNp Knans, H.M. J., J. Biol. Chem., 248, 1877 (1971). ; 10. Krisuna, G., Weiss, B., anp Brovir, B. B., J. Pharmacol. Exp. Ther., 168, 379 (1968). M1. Ropseti, M., J. Brol. Chem., 24%, 5744 (1967). 12. Brrnsaumer, L., Pons, 8S. L., ann Ropneuy, M., J. Biol. Chem., 244, 3468 (1969). 13. Lowry, O. H., Rosesroucu, N.J., Farr, A. L., ann RANDALL, R. J., J. Biol. Chem., 198, 265 (1951). 14. Bopansxy, O., ann Scuwantz, M. K., J. Biol. Chem., 288, 3420 (1963). 15. Swanson, M. A., in S. P. Corowicx anp N. O. Kapuay (Edi- tors), Methods in enzymology, Vol. IT, Academic Press, New York, 1955, p. 541. 16, Herren, L. A., in 8. P. Corowick anp N. O. Kapuan (Edi- tors), Methods in enzymology, Vol. II, Academic Press, New York, 1955, p. 530. 17. Fiske, C, H., anp SuppaRow, Y., J. Biol. Chem., 66, 375 (1925). 18. FLetscuer, S., aND FLEerscuer, B., in S. P. Cotowick AND N. 0. Kaptan (Editors), Methods in enzymology, Vol. Xx, Academic Press, New York, 1967, p. 427. 19. Lurt, H. J., J. Biophys. Biochem. Cytol., 9, 409 (1961). 20. Kannovaxy, M. J., J. Biophys. Biochem. Cytol., 11, 729 (1961), 21. Nevitie, 1, M., Jn., J. Biophys. Biochem. Cytol., 8, 413 (1960). 22. Buxeo errr, E. L., aNp EMMELOT, P., in A. J. DALTON AND F, Haauenav (Editors), Ultrastructure in biological systems, Vol. 4, Academie Press, New York, 1968, p. 33. . Murt, V., ano Jorpes, J. E., Recent Progr. Hormone Res., 23, 483 (1967). . Eumeor, P., Bos, P. J., Beneperti, FE. L., ann Rtuxe, P. H., Biochim. Biophys. Acia, 90, 126 (1964). . Levey, G.S., Bivchem. Biophys. Res. Commun., $8, 86 (1970). . Rrorpan, J. F., anp Vauues, B. L., in S. P. Conowick anp N. O. Kapian (Editors), Methods in enzymology, Vol. XT, Academic Press, New York, 1967, p. 541. . Brrnpaumer, L., aND RopBewu, M., J. Biol. Chem., 244, 3477 (1969). . Roppeit, M., Brrnnaumer, L., ano Pout, S. L., J. Biol. Chem., 245, 718 (1970). 29, Surnuer.anp, E. W., anp Rau, T. W., Pharmacol. Rev., 12, 265 (1960). eo fo NN AO ff Sk © B 8 N 1856 | 30. Davonsn, P. R., ann Surner.anp, E. W., J. Biol. Chem., 238, 3016 (1963). 31. McKee, D. W., anp Jarert, L., J. Cell Biol., 44, 417 (1970). 32. Pruzcer, R. C., ANperson, N.G., anv Snyper, F., Biochem- tetry, 7, 2868 (1968). 38. Rar, T. K., Sxirrsxi, V. P., Bancuay, M., Essner, E., ann AucuipaLp, F. M., J. Biol. Chem., 244, 5528 (1960). 3. Wionett, C. C., anp Unxetess, J. C., Proc. Nat. Acad. Sci. U. 8. A., 61, 1080 (1968). , 35. Rerx, L., Persoup, G. L., Hicains, J. A., Greenaanrp, P., anp Barruert, R. J., Science, 168, 384 (1970). Glucagon-sensttive Adenyl Cyclase System. I Vol. 246, No. 6 36. Birensxy, M. W., Russz.z, V., anv Roserrson, W., Bio- chem. Biophys. Res. Commun., 31, 706 (1968). 37, Birenaxy, M. W., Russeuy, V., anp Banco, M., Endocrinol- ogy, 88, 154 (1970). 38. Maninerti,G. V., Ray, T. K., anp Tomas, V., Biochem, Bio- phys. Res. Commun., 38, 185 (1969). 30. Rav, T. K., Toast, V., anp Maninertt, G. V., Biochim. Bio- phys. Acta, 211, 20 (1970). 40. fcurauu, M., anp Nam, E., J. Biol. Chem., 945, 3225 (1970). 41. Taunton, O. D., Rors, J., anp Pastan, I., J. Biol. Chem., 244, 247 (1960).