hyperpolarization (4-6 mv). The effect of isoprenaline, but not that of hyperpolarization of 4-8 mv.

Transcription

1 J. Physiol. (1974), 239, pp With 4 text-figures Printed in Great Britain THE EFFECT OF GLUCAGON ON THE LIVER CELL MEMBRANE POTENTIAL BY 0. H. PETERSEN From the Institute of Medical Physiology C, University of Copenhagen, Copenhagen, Denmark (Received 11 December 1973) SUMMARY 1. Intracellular recordings of membrane potential were made from superficial cells of isolated mouse liver segments superfused with physiological salt solutions. 2. The mean resting cell membrane potential was mv. 3. Glucagon caused a dose-dependent membrane hyperpolarization which was detectable at 10-9 M and maximal (7 mv) at 10-7 M. The hyperpolarization started within half a minute after exposure to glucagon. Secretin (2 x 10-7 M) had no effect on the membrane potential. 4. Adrenaline (10-6 M) and isoprenaline (106 M) also caused membrane hyperpolarization (4-6 mv). The effect of isoprenaline, but not that of adrenaline, was blocked by propranolol (5 x 10 6 M). 5. Dibutyryl adenosine 3',5'-monophosphate (10-3 M) caused a membrane hyperpolarization of 4-8 mv. 6. In the absence of extracellular K or the presence of Strophanthin-G (10-3 M) the resting potential was decreased and the response to glucagon reduced. During exposure to a solution containing 20 mm-k the resting potential was slightly enhanced and the amplitude of the glucagon-induced hyperpolarization reduced compared with control conditions. 7. It is concluded that the effect of glucagon on the membrane potential is due to an interaction with specific membrane receptors probably leading to activation of the membrane-bound adenyl cyclase. It is probable that the hyperpolarization is mediated by cyclic AMP. The hyperpolarization induced by glucagon is dependent on a normal function of the membrane Na-K pump. TNTRODUCTION The polypeptide hormone glucagon exerts its action on liver cell metabolism by stimulating a membrane-bound adenyl cyclase resulting in an intracellular accumulation of adenosine-3',5'-cyclic monophosphate (cyclic AMP). All metabolic actions of glucagon can be mimicked by cyclic AMP

2 648 O. H. PETERSEN (Robison, Butcher & Sutherland, 1971; Exton, Robison, Sutherland & Park, 1971). Somlyo, Somlyo & Friedman (1971) using a technique of serial micro-electrode penetration have reported that glucagon, cyclic AMP, cyclic GMP and isoprenaline cause a delayed but marked cell membrane hyperpolarization in the perfused rat liver occurring concomitant with a cellular K loss. In the present work on superfused mouse liver segments examined with an indwelling intracellular micro-electrode, the time course of the membrane effect of glucagon, the dose-response relationship and some aspects of the ionic mechanism of the hyperpolarization have been investigated. METHODS The experiments were performed, as previously described for salivary glands and pancreas by Pedersen & Petersen (1973) and Petersen (1973), on isolated segments of mouse liver placed in a Perspex tissue bath (volume 20 ml.). The segments (5 x 5 mm, thickness about 2 mm), which always included an undamaged and untouched surface area, were removed from the mice within 2 min after killing the animals and immediately thereafter placed in the tissue bath through which a Krebs-Henseleit solution, equilibrated with 5 % CO2 and 95 % 02 and warmed to 370 C, flowed at a constant rate (15 ml./min). The control solution contained (mm): NaCl 103, KCl 4 7, CaCl2 2-56, MgCl2 1-13, NaHCO3 25, NaH2PO4 1'15, D-glucose 2-8, Na pyruvate 4.9, Na fumarate 2-7, and Na glutamate 4 9. In some experiments where the K concentration was altered, corresponding alterations in Na concentration ensured constant osmolarity. Membrane potentials were measured using glass micro-electrodes filled with 1.5 M-K citrate having a tip resistance of about 50 M.Q. Parenchymal cells lying near to the undamaged surface of the liver (less than 50,#m in depth) were impaled with the help of a step-motor driven micro-manipulator. Membrane potentials were observed on the oscilloscope screen and recorded on paper by a Devices M2 recorder. Strophanthin-G and (-)-isoprenaline sulphate were obtained from Mecobenzon, Copenhagen; dibutyryl cyclic AMP and (-)-adrenaline from Sigma; glucagon from NOVO; secretin was a gift from Professor V. Mutt, Karolinska Institutet, Stockholm and (± )-propranolol hydrochloride came from I.C.I. RESULTS The resting membrane potential The mean value of the resting cell membrane potential, obtained from fifty impalements of surface liver cells, was mv (S.E.) which is close to values previously reported for rat liver in vivo (-43 mv) (Williams, Withrow & Woodbury, 197 la), perfused rat liver (-35 mv) (Somlyo, Somlyo & Friedmann, 1971) and guinea-pig liver slices (-36 mv) (Haylett & Jenkinson, 1972 a, b). The resting potential was constant for 3-4 hr except for the first min after starting the superfusion when somewhat lower values were found (Fig. 4).

3 GLUCAGON ON LIVER CELL POTENTIAL 649 The effect of glucaqon and cyclic AMP Changing the superfusion fluid from control Krebs-Henseleit solution to one containing 10-7 M glucagon always caused hyperpolarization (Figs. 1 and 3). Glucagon increased the mean membrane potential from - 39*1 mv to - 46*6 mv + 1*4. The hyperpolarization was significant (P < 0.001). Although an exact measurement of latency was impossible, it is clear from Figs. 1 and 3, which are quite typical, that there was no appreciable delay, and certainly not a latency of about 5 min which was found in an earlier study using the serial penetration technique (Somlyo et al. 1971). Fig. 2 shows the dose-response relationship for the hyperpolarizing effect of glucagon. 50[ Glucagon L Adrenaline Isoprenaline 4E Db.-cyclic AMP Fig. 1. Cell membrane potential recordings from superfused mouse liver segments. Effect of glucagon (10-7 M), isoprenaline and adrenaline (106 M) and dibutyryl cyclic AMP (10-3 M). In the time marker trace pulses occur every minute. In order to test the specificity of the glucagon response, the effect of secretin (2 x 10-7 M), a polypeptide hormone which is structurally very similar to glucagon (Mutt & Jorpes, 1967) was tested in two preparations that responded well to glucagon. Secretin caused no change in membrane potential. In three different tissues 6 short periods of superfusion with a solution

4 H. PETERSEN containing 10-3 M dibutyryl cyclic AMP were employed (Fig. 1). In all six cases this caused hyperpolarization (2-7 mv) with a mean value of 4*7 mv + 0*7. The hyperpolarization often had a relatively long duration (Fig. 1) > o / 51-2-=:- 100 od so/ ~~~~~/0J wher 4tlattodfeetcnetain eetse.tedte curv repesnt D the ose-response relationship forthehyeplrzneffect of glucagon.o rat liver plasma membrane adenyl cyclase activity, expressed as percentage of the activity obtained with a supramaximal dose of glucagon, calculated from the data of Pohl, Birnbaurner &r Rodbell (1971). The effect of adrenaline and isoprenaline Adrenaline (Io06m) always caused cell membrane hyperpolarization (Fig. l). Adrenaline increased the mean membrane potential from to mv (P < 0 02). A rather similar effect was caused by isoprenaline (10-- m) (Fig. l). Isoprenaline increased the mean membrane potential from to mv (P < 0 O1). The effect of isoprenaline was abolished by propranolol (5 x l1 06m) while adrenaline still caused hyperpolarization (4-6 mv) in the presence of this concentration of the Plreceptor blocking agent. The effect of Strophanthin-G Changing the superfusion solution to one containing Strophanthin-G (10-3 M) always resulted in depolarization (Fig. 3). The mean value of this

5 CLUCACON ON LIVER CELL POTENTIAL 651 initial depolarization (Fig. 3) was 7-0 mv After 5-8 min of superfusion with Strophanthin-G, glucagon (10-7 M) caused a comparatively small (4.0 mv + 0.6) and relatively short-lasting hyperpolarization (Fig. 3). O_5L Glucagon Strophanthin-G l- ~~~~ - so~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Glucagon Fig. 3. The effect of glucagon (10-7 M) in the absence and presence of Strophanthin-G (10-3 M). The upper tracing was obtained during exposure to control solution. In the lower record the superfusion solution was changed from control to one containing Strophanthin-G and back to control again. Time mark intervals: 1 min. The effect of varying [K]0 Three experiments with K-free solutions were carried out. The results are summarized in Fig. 4. During exposure to a K-free solution the resting potential decreased, and as seen in Fig. 4 the response to glucagon (10-7 M) was first reduced and finally abolished. Reintroducing the control solution ([K] = 4.7 mm) during intracellular recording hyperpolarized the cell membrane by 10-3 mv The maximal hyperpolarization was attained about 10 min after readmission of K (4.7 mm). In the presence of a solution containing 20 mm-k the mean resting cell membrane potential was mv In this condition glucagon (10-7 M) always evoked a relatively small hyperpolarization (4.4 mv + 0.9).

6 652 O. H. PETERSEN Z 1-40_ ; i i + C~~~~~~~~~ E I ;Xl6-l C C C K-free K-free Fig. 4. The effect of a K-free solution on the glucagon-induced hyperpolarization. The results shown are mean values + S.E. from three experiments on three mouse livers in which the tissues were exposed to glucagon for 2-3 min every 40 min, 5 times in total, the two last times in the presence of a K-free solution. The K-free solution was introduced 20 min before the first stimulation in the K-free solution. Filled circles represent resting potentials while open circles represent the maximal membrane potential recorded during exposure to glucagon (10-7 M). For comparison mean values are also shown from other experiments involving exposure to control solution (triangles) at a time after starting the superfusion of the liver segments corresponding to the last glucagon stimulation period in the experiments with K-free solutions. DISCUSSION The essential finding in this study is that the polypeptide hormone glucagon causes a rapid hyperpolarization of the plasma membrane of the liver cell. The dose-response curve for the glucagon effect on the membrane potential is remarkably similar to the dose-response curve for the effect of glucagon on adenyl cyclase activity from rat liver plasma membrane (Fig. 2) previously described (Pohl, Birnbaumer & Rodbell, 1971). The maximal effect of glucagon on glucose metabolism is attained at a hormone concentration of 10-9 M while at M there is hardly any effect (Robison

7 GLUCAGON ON LIVER CELL POTENTIAL 653 et al. 1971). This means that the lower part of the dose-response curves for the glucagon effect on adenyl cyclase activity (Pohl et al. 1971) membrane potential (Fig. 2) and glucose metabolism (Robison et al. 1971) are very similar. The dose-response curves for the glucagon effect on adenyl cyclase activity and membrane potential are also very similar to the curve describing the relationship between binding of labelled glucagon and concentration of labelled hormone (Rodbell, Krans, Pohl & Birnbaumer, 1971). The time course of action of glucagon on the membrane potential (Fig. 1) is generally rather similar to the time course of cyclic AMP accumulation in rat livers perfused with glucagon. An effect of glucagon on the cyclic AMP level was apparent about half a minute after start of glucagon perfusion and maximal after about 3 min (Exton et al. 1971). These findings might indicate that the membrane effects of glucagon described here were caused by intracellular cyclic AMP, but the alternative interpretation that a conformational change in the cell membrane associated with the hormone-receptor interaction, leading to the activation of the adenyl cyclase, necessarily involves a membrane permeability change, cannot be excluded at the present stage. The finding that epinephrine and isoprenaline also cause membrane hyperpolarization may be taken as further support for the idea of cyclic- AMP-induced hyperpolarization, since these substances are known to exert their effect on glucose metabolism through the second messenger cyclic AMP. This effect is usually attributed to activation of a,8-adrenergic receptor (Robison et al. 1971). The effect of the fl-receptor activating agent isoprenaline was as expected blocked by propranolol, but this drug did not abolish adrenaline-induced hyperpolarization. A major part of the adrenaline effect is therefore probably not related to intracellular cyclic AMP accumulation, but is probably due to x-receptor activation resulting in an enhanced cell membrane K permeability, as has been worked out in detail by Haylett & Jenkinson (1972a, b). More evidence in favour of the hypothesis that the effect of glucagon on the membrane potential is mediated by cyclic AMP came from the experiments showing that exogenous dibutyryl cyclic AMP caused cell membrane hyperpolarization (Fig. 1). Glucagon acts on the liver cell membrane by interacting with specific binding sites (Rodbell et al. 1971). It seems likely that the glucagon effect on the membrane potential is also due to interaction with specific receptor sites, since secretin, which is structurally very similar to glucagon (Mutt & Jorpes, 1967) did not exert any effect on the membrane potential. It has previously been shown that secretin does not stimulate the liver plasma membrane adenyl cyclase and does not interfere with binding of labelled glucagon to the plasma membrane (Pohl et al. 1971; Rodbell et al. 1971).

8 654 O.H.PETERSEN Somlyo et al. (1971) have described a delayed (5 min) hyperpolarizing effect of both glucagon, cyclic AMP and isoprenaline. Cyclic AMP also caused a delayed increase in K efflux from the perfused rat liver, having a time course similar to the membrane hyperpolarization. Although it is clear that the technique employed in the present work with an indwelling micro-electrode is superior to the serial micro-electrode penetration technique, especially with respect to measurement of latency, there may be a real discrepancy between the results presented here and those of Somlyo et al. (1971). It is obvious, however, that the present results are in reasonable agreement with the time course of activation of the adenyl cyclase in the liver cell membrane by glucagon (Exton et al. 1971). The ionic mechanism of the glucagon-induced hyperpolarization has not been entirely clarified by the results obtained in the present study. 5-8 min after introduction of ouabain (1 mm) the effect of glucagon was reduced (Fig. 3) and in the absence of extracellular K the effect of glucagon was nearly abolished (Fig. 4). This might indicate that the hyperpolarization induced by glucagon was caused by activation of an electrogenic Na-K pump. In excitable tissues it is now generally accepted that under certain conditions the membrane Na-K pump can act electrogenically, i.e. contribute to the membrane potential (Thomas, 1972). There is evidence that this may occur in pancreatic tissue (Petersen, 1973) as well as in liver (Williams, Withrow & Woodbury, 1971 a, b; Haylett & Jenkinson, 1972a). The finding that ouabain causes a rapid and marked depolarization (Fig. 3), and that increasing [K]o from 4.7 to 20 mm causes a slight hyperpolarization is in agreement with this. There is, however, an alternative explanation which might account for the glucagon-induced hyperpolarization, namely an enhanced K-permeability. In the presence of ouabain and K-free solutions arrest of the Na-K pump might result in a rapid reduction of the normally existing Na and K gradients. The results from the experiments in which a 20 mm-k solution were employed show that in this condition the amplitude of the glucagon-induced hyperpolarization was reduced, which would be compatible with the hypothesis of an enhanced membrane K permeability. There is relatively little information about the effect of polypeptide hormones on electrical properties of cell membranes, and from the available evidence no general rules are apparent. Cholecystokinin-pancreozymin (CCK-Pz) depolarizes the pancreatic acinar cell membrane and reduces the membrane resistance (Matthews & Petersen, 1973; Nishiyama & Petersen, 1974). Adrenocorticotrophic hormone (ACTH) has normally no influence on the membrane potential of adrenocortical cells, but during exposure of these cells to K-free solutions ACTH depolarizes the cell membrane and evokes action potentials (Matthews & Saffran, 1973). It is

9 GLUCACON ON LIVER CELL POTENTIAL 655 thus apparent that CCK-Pz, ACTH and glucagon have entirely different effects on the membrane potential of their respective target cells. The importance of the effect on the membrane potential or permeability is not entirely clear in the cases of ACTH and glucagon, while the effect of CCK-Pz on the pancreatic acinar cell membrane results in an altered intracellular Ca distribution causing exocytosis (Matthews, Petersen & Williams, 1973). CCK-Pz, ACTH and glucagon all activate a membrane-bound adenyl cyclase in their respective target tissues (Robison et al. 1971; Bonting & de Pont, 1974), but it is doubtful whether cyclic AMP is responsible for the effect on the membrane potential in all three cases. In the case of CCK-Pz, cyclic AMP is probably not the mediator (Matthews & Petersen, 1973; Williams, 1974) whereas it seems likely, from the present experimental results, that glucagon is working through cyclic AMP when hyperpolarizing the liver cell membrane. REFERENCES BONTING, S. L. & DE PONT, J. J. H. H. M. (1974). Adenylate cyclase and phosphodiesterase in rat pancreas. In Secretory Mechanism8 of Exocrine Gland8, ed. THORN, N. A. & PETERSEN, 0. H. Copenhagen: Munksgaard (in the Press). EXTON, J. H., ROBISON, G. A., SUTHERLAND, E. W. & PARK, C. R. (1971). Studies on the role of adenosine 3',5'-monophosphate in the hepatic actions of glucagon and catecholamines. J. biol. Chem. 246, HAYLETT, D. G. & JENKINSON, D. H. (1972 a). Effects of noradrenaline on potassium efflux, membrane potential and electrolyte levels in tissue slices prepared from guinea-pig liver. J. Phy8iol. 225, HAETT, D. G. & JENKINSON, D. H. (1972b). The receptors concerned in the actions of catecholamines on glucose release, membrane potential and ion movements in guinea-pig liver. J. Phy8iol. 225, MATTHEws, E. K. & PETERSEN, 0. H. (1973). Pancreatic acinar cells: ionic dependence of the membrane potential and acetylcholine-induced depolarization. J. Physiol. 231, MATTHEws, E. K., PETERSEN, 0. H. & WILLIAMS, J. A. (1973). Pancreatic acinar cells: acetylcholine-induced membrane depolarization, calcium efflux and amylase release. J. Phy8iol. 234, MATTHEws, E. K. & SAFFRAN, M. (1973). Ionic dependence of adrenal steroidogenesis and ACTH-induced changes in the membrane potential of adrenocortical cells. J. Phy8iol. 234, MUTT, V. & JORPES, J. E. (1967). Contemporary developments in the biochemistry of the gastrointestinal hormones. Recent Prog. Horm. Res. 23, NISHIYAMA, A. & PETERSEN, 0. H. (1974). Pancreatic acinar cells: membrane potential and resistance change evoked by acetylcholine. J. Physiol. 238, PEDERSEN, G. L. & PETERSEN, 0. H. (1973). Membrane potential measurement in parotid acinar cells. J. Physiol. 234, PETERSEN, 0. H. (1973). Electrogenic sodium pump in pancreatic acinar cells. Proc. R. Soc. B 184, POHL, S. L., BIRNBAUMER, L. & RODBELL, M. (1971). The glucagon-sensitive adenyl cyclase system in plasma membranes of rat liver. I. Properties. J. biol. Chem. 246,

10 656 O.11. PETERSEN ROBISON, G. A., BUTCHER, R. W. & SUTHERLAND, E. W. (1971). In Cyclic AMP, pp New York and London: Academic Press. RODBELL, M., KRANS, H. M. J., POHL, S. L. & BIRNBAUMER, L. (1971). The glucagonsensitive adenyl cyclase system in plasma membranes of rat liver. III. Binding of glucagon: method of assay and specificity. J. biol. Chem. 246, SOMLYO, A. P., SOMLYO, A. V. & FRIEDMANN, N. (1971). Cyclic adenosine monophosphate, cyclic guanosine monophosphate, and glucagon: effects on membrane potential and ion fluxes in the liver. Ann. N.Y. Acad. Sci. 185, THOMAS, R. C. (1972). Electrogenic sodium pump in nerve and muscle cells. Physiol. Rev. 52, WILLIAMS, J. A. (1974). Intracellular control mechanisms regulating secretion by exocrine and endocrine glands. In Secretory Mechanisms of Exocrine Glands, ed. THORN, N. A. & PETERSEN, 0. H. Copenhagen: Munksgaard (in the Press). WILLIAMS, J. A., WITHROW, C. D. & WOODBURY, D. M. (1971a). Effects of ouabain and diphenylhydantoin on transmembrane potentials, intracellular electrolytes, and cell ph of rat muscle and liver in vivo. J. Physiol. 212, WILLIAMS, J. A., WITHROW, C. D. & WOODBURY, D. M. (1971b). Effects of nephrectomy and KCl on transmembrane potentials, intracellular electrolytes, and cell ph of rat muscle and liver in vivo. J. Physiol. 212,