0013-7227/78/1036-2207S02.00/0 Endocrinology Copyright © 1978 by The Endocrine Society

Vol. 103, No. 6 Printed in U.S.A.

The Effect of Potassium and Valinomycin on Insulin and Grlucagon Secretion in the Perfused Rat Pancreas* GORDON EPSTEIN, RUDOLPH FANSKA, AND GEROLD M. GRODSKY Metabolic Research Unit and the Departments of Biochemistry and Biophysics, University of California, San Francisco, California 94143 ABSTRACT. We examined the response of the a- and /3-cells of the perfused rat pancreas to an elevation of potassium ion concentration and to the potassium ionophore, valinomycin (1 X 10~6 M). Elevation of potassium from a baseline concentration of 5.8 mM to 11.3, 18.3, and 53.3 mM caused a biphasic release of immunoreactive insulin characterized by a spiked first phase and a low sustained second phase and a spiked release of immunoreactive glucagon followed by a very low sustained secretion that ended with the termination of the stimulation. The second phase of potassium-stimulated insulin release increased in the presence of basal glucose (3.3 mM). Sequential higher concentrations of potassium caused a dose-response increase in both phases of insulin release, but did not increase glucagon secretion. Valinomycin inhibited the insulin secretion caused by potassium, arginine, and glucose and the

TNSULINOTROPIC agents, such as manJL nose, sulfonylureas, and D-glucose, elicit electrical activity in the pancreatic /?-cell; each causes depolarization and a burst or continuous pattern of spike potentials across the plasma membrane (1-3). These spike potentials have been strongly correlated with the secretory process (4). Substances that are believed to depolarize cells specifically, such as ouabain and potassium ion, also cause the secretion of insulin (2, 3, 5). Matthews (6) and Atwater and Beigelman (7) have proposed that changes in membrane permeability to potassium bring about the opening of voltage-dependent calcium channels, and the resultant increased intracellular cytosolic calcium provides a requisite signal for insulin release. Due to the relatively small Received March 7, 1978. Address requests for reprints to: Dr. Gerold M. Grodsky, Metabolic Research Unit, University of California, San Francisco, California 94143. * This work was supported in part by NIAMDD Grant AM-01410, the Kroc Foundation of Santa Ynez, CA, and the Juvenile Diabetes Foundation of New York City.

glucagon secretion caused by potassium and arginine. Potassium, when applied on the second phase of arginine-stimulated insulin and glucagon release, enhanced the secretion of insulin but caused a dramatic inhibition of glucagon that did not correlate directly with insulin release and which may possibly result from the entrance of an excessive amount of calcium into the islet cells. The secretion caused by potassium has been attributed to its depolarizing qualities. If the principal effect of valinomycin is a result of its postulated hyperpolarization of the islet cells, it seems that the electrical behavior of the a and /?-cells with regard to secretion is similar; depolarization enhances secretion and hyperpolarization inhibits it. Finally, the established inhibition of glucagon secretion by glucose must be induced by a mechanism other than depolarization. (Endocrinology 103: 2207, 1978)

number of a-cells in the islet, it has not been possible to determine the electrical characteristics of the a-cell or to speculate on the role of membrane potential and potassium permeability in the mechanism of glucagon secretion. In this study, we investigated the insulin and glucagon response of the perfused rat pancreas to elevated potassium concentrations and to increased potassium permeability produced by the potassium ionophore, valinomycin (1 X 10~6 M). The effect of valinomycin on glucose-, arginine-, and potassiumstimulated insulin release and arginine- and potassium-stimulated glucagon release was also examined. Bhathena et al. (8) have reported that potassium inhibits secretion from the a-cell, in contrast to its positive effect on the y8-cell. Our experiments, however, suggest that the potassium ion can be a positive signal for glucagon secretion as well as for insulin. It also seems that, under specific conditions, potassium can inhibit the secretion of glucagon, although this may or may not be due to a direct effect of

2207

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 07:11 For personal use only. No other uses without permission. . All rights reserved.

2208

EPSTEIN, FANSKA, AND GRODSKY

the ion. Valinomycin was shown to block the secretion of both insulin and glucagon. An increase in potassium permeability caused by valinomycin has been shown to produce hyperpolarization in other systems (9, 10). If the principal effect of valinomycin and potassium is to hyperpolarize and depolarize, respectively, then it would seem that depolarization is conducive to secretion of insulin and glucagon and that hyperpolarization will prevent the release of both hormones.

Materials and Methods Adult, overnight-fasted, Long Evans rats were used in all studies. The in vitro perfused rat pancreas technique has been previously described in detail (11, 12). In order to limit the contribution of extrapancreatic glucagon secretion, a modified preparation was used in which most of the duodenum, stomach, and spleen were removed (13). The perfusion medium was a bicarbonate-buffered saline solution containing 1% human serum albumin and 3% dextran. Unless otherwise specified, the ionic composition of the perfusate (subsequently called normal ionic composition) was: Na+2 (145 meq), K+ (5.8 meq), Ca2+ (4.2 meq), Mg+2 (2.4 meq), H2PO~4 (1.5 meq), HCO~3 (29.0 meq), and Cl" (150.0 meq). The perfusate was introduced into the celiac artery, and complete effluent was collected from the portal vein every minute in chilled tubes containing 12 mg EDTA. Flow rate was maintained at 4 ml/min, and the arterial pressure exceeded 80 mm Hg only when 53.3 mM potassium were perfused. The usual arterial pressure was 40-60 mm Hg. Potassium chloride was dissolved in perfusate ind delivered by side-arm syringe at a rate of 0.1 nl/min to produce the desired concentrations (11.3, L8.3, or 53.3 mM). Valinomycin (Calbiochem, 3erkeley, CA) was dissolved in absolute ethanol ind infused via side-arm syringe at a rate of 0.024 nl/min to produce a final concentration of ethanol n the perfusate of 0.6% and a concentration of 'alinomycin equal to 1 X 10~6 M. (The presence of his amount of ethanol was found to be necessary o keep the valinomycin in solution.) Control studes were run in which ethanol alone was introduced o determine if the ethanol affected insulin or gluagon secretion. When basal glucose (3.3 mM) was iresent, it was added during the preparation of the »erfusate and was therefore present throughout quilibration and experimental periods.

Endo i 1978 Vol 103 i No 6

All experiments were preceded by an equilibration period of 15 min at normal ionic concentrations (time, —15 min). The pancreas was exposed to potassium at varying concentrations (11.3, 18.3, or 53.3 mM), glucose (16.7 mM), and arginine (10 or 20 mM). Perfusion of the ethanol or the valinomycin plus ethanol was begun 5 min before the start of a stimulation and continued until 5 min after the termination of the stimulation. To obtain severely diabetic rats (14), streptozotocin was dissolved immediately before use in cold 10"3 M citrate buffer (pH 4.2), and a dose of 120 mg/kg (approaching the LD5o) was injected into the saphenous vein under light sodium pentobarbitol anesthesia (4.5 mg/100 g). Initially, standard amounts of streptozotocin (80 mg/kg) were used, but it was found that the combined arginine and potassium stimulus still caused some, although reduced, insulin secretion. After 3 days, the animals, which were all ketoacidotic (mean blood glucose, 300 mg/100 ml; moderate to high urine ketones, as measured with Ketostix reagent sticks; Ames Co., Elkhart, IN), were sacrificed and their pancreases were perfused. Immunoreactive insulin was measured by a solid phase system, using the Micromedic pipetting station (15). Immunoreactive glucagon was measured by the dextran charcoal RIA method of Unger (16).

Results Insulin and glucagon secretion by the pancreas in response to varied potassium concentrations As illustrated in Fig. 1, the pancreas was exposed to three consecutive stimulations of potassium (11.3,18.3, and 53.3 mM) for 15 min each, with intervals of 10 min at a baseline potassium concentration (5.8 mM). No glucose was present in this protocol. A control study in which no potassium was added is also shown. As previously reported, glucagon in control experiments is elevated immediately after initiating the pancreatic perfusion and declines spontaneously thereafter (17). (Insulin remained undetectable and, therefore, is not illustrated.) The elevation of potassium to 18.3 and 53.3 mM caused a rapid secretion of insulin, followed by a low sustained second phase, which continued for the duration of the stimulation. In both phases of secretion, insulin displayed a dose responsiveness to po-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2015. at 07:11 For personal use only. No other uses without permission. . All rights reserved.

K+ AND VALINOMYCIN ON ISLET SECRETION Potassium

s;

ll.3mM

l8.3mM

l.3mM

33.3mM

70

7.0 r

60

6.0

50

£

5.0

40

£

4.0

I ^

30 20 10

1



-

-10

10

L

20 30

V

2209 IS.3mM

53.3mM

2.0 1.0 0

40 50 60 70

-10

10

20

L=L

30 40 50 60 70

MINUTES

FIG. 1. Pancreatic insulin and glucagon response (mean ± SE) to three potassium stimulations. , Perfusions in which potassium was applied (n = 5); (glucagon only), control (n = 4). (A fall in glucagon secretion, which occurred in all experiments at 16, 41, and 66 min, is difficult to see when data are averaged.)

tassium. Potassium also caused a positive release of glucagon characterized by a rapid early phase and a low sustained secretion that terminated promptly after discontinuation of the stimulation. No dose-response pattern for glucagon secretion was apparent. The response to the same three potassium concentrations in the presence of basal glucose (3.3 mM) is shown in Fig. 2. The insulin secretion was found to have a slightly higher second phase than when glucose was absent (P

The effect of potassium and valinomycin on insulin and glucagon secretion in the perfused rat pancreas.

0013-7227/78/1036-2207S02.00/0 Endocrinology Copyright © 1978 by The Endocrine Society Vol. 103, No. 6 Printed in U.S.A. The Effect of Potassium and...
793KB Sizes 0 Downloads 0 Views