Effect of hypoglycemic sulfonylureas on Ca2+ fluxes across lipid bilayers

RIOCHEMICAI.

MEDICINE:

31, 146-F-153 ( 19841

Effect of Hypoglycemic Sulfonylureas on Ca2’ Fluxes across Lipid Bilayers ERIK

GYLFE,

Depurtments

Bo

HELLMAN,

of Medical Biomedicum,

G&TA

ARVIDSON,*

*Medical Cell Biology, l.Jniversit.v of Uppsala, Received

May

AND

JOHN

SANDBLOM?

Chemistry, und tMedica1 Biophysics. S-751 23 Uppsala, Sweden 16, 1983

Evidence has been provided that hypoglycemic sulfonylureas are restricted to the exterior of the pancreatic &cells, and that their insulinreleasing action involves ionic events in the plasma membrane (I). Considerable attention has been paid to the idea that they function as Ca2’ ionophores and/or directly interact with native Ca2+ ionophores in the P-cell plasma membrane (2-6). An alternative mechanism has been suggested to involve sulfonylurea inhibition of K+ conductance with subsequent P-cell depolarization (7,g). The ionophore concept is based essentially on studies where the translocation of 45Ca has been measured into or across a hydrophobic immiscible domain containing sulfonylurea, antibiotic ionophores, or pancreatic islet extracts. The physiological relevance of these observations has been a matter of controversy (7-10). When exploring the exchange diffusion phenomena which normally occur in the presence of carboxylic ionophores, sulfonylureas were found to lack Ca2’ ionophoretic activity in excitable cells and isolated secretory granules (11). In an extension of these studies sulfonylureas have now been compared to different classes of ionophores by measuring both the net efflux of Ca2+ from liposomes and the conductance of black lipid membranes. Whereas the sulfonylureas tested did not exhibit any ionophoretic effects, high concentrations of glibenclamide evoked a general labilization of the liposomal membranes. MATERIALS

AND METHODS

Chemicals. Reagents of analytical grade and deionized water were used. Cholesterol was purified by sialic acid chromatography and 1,2dipalmitoyl-sn-glycero-3-phosphocholine was synthetized and purified to >99% as judged by thin-layer chromatography. Sigma Chemical Company 246 0006-2!M4/84 Copyright All rights

$3 .OO

0 1984 by Academic Press. Inc. of reproduction in any form reserved

HYPOGLYCEMIC

SULFONYLUREAS

AND

Ca”

FLUX

247

(St. Louis, MO.) supplied arsenazo III (grade I), glyceromonooleate, ethylene glycol bis(@aminoethyl ether)-N,N,N’,N’-tetraacetic acid (EGTA), and saponin. Nonactin was kindly provided by Dr. B. Stearns, E. R. Squibb & Sons Inc. (Princeton, N.J.), and gramicidin A was from Dr. E. Gross, National Institutes of Health (Bethesda, Md.). BoehringerMannheim GmbH (Mannheim, FRG) provided A-23187, and hexadecane was from E. Merck, (Darmstadt, FRG). Tolbutamide, glibenclamide, and the nonhypoglycemic ghbenclamide analog S-787510 were gifts from Hoechsl AG (Frankfurt am Main, FRG). 5,6-Carboxyfluorescein from Eastman Kodak Company (Rochester, N.Y.) was purified as previously described (12), resulting in essentially complete removal of contaminants as judged by thin-layer chromatography. Prepurution of liposomes. Liposomes were prepared with the reversephase evaporation method (13) from dipalmitoyl-lecithin and cholesterol (7: 3 mole/mole) at pH 7.0 in medium consisting of 100 mM CaCl, + 0.1 mM Tris-maleate or 1 IO mu 5,6carboxyfluorescein + 15 mM Tris-maleate. As judged by electron microscopy this technique yielded large uni- and oligolamellar vesicles (not shown). The liposomes were filtered through polycarbonate filters (Uni-Pore, Bio-Rad Laboratories, Richmond, Calif.) with a pore size of 400 nm and then dialyzed for at least 24 hr in isotonic Tris-maleate buffer (pH 7.0). The contents of phospholipid-bound phosphorus in the liposomal preparations were assessed at 45 and 37 pmolei ml for the CaCl,- and 5,6-carboxyfluorescein-loaded liposomes, respectively. Experimental procedures. Ca2 * release from the liposomes was recorded by measuring the concentration of the ion in the medium with the metallochromic indicator arsenazo III by dual-wavelength spectrophotometry (14) using a time-sharing multichannel spectrophotometer (15). Release of the fluorescent dye 5,6-carboxyfluorescein was used as an indicator of liposome stability (16). The conductance of black lipid membranes formed from glycerolmonooleate in hexadecane was studied in symmetrical salt solutions using procedures and apparatus previously described (17). Further experimental details are given in the legends to the figures. Since some of the ionophores employed have a very limited solubility in aqueous media it should be noted that the concentrations given are those which would be obtained if they were freely soluble. RESULTS

In Fig. I the effects of sulfonylureas on Ca2+ efflux from liposomes are compared to those of the Ca”/H’ exchanging ionophore A-23187. In this sensitive system a concentration as small as 20 nM of the ionophore caused a marked release of Ca* + . The rate of efflux, was not influenced by 1 mM tolbutamide. However, in contrast to its nonhypoglycemic analog

248

GYLFE ET AL

l 15-

0

2

4

6

6

0

2

4

6

8

6

6

IOOpM 1OopM s-7a7510

< 4

1 -

0

2

4

6

6

0

2

4

TIME WIN)

FIG. I. Effects of sulfonylureas on Ca’+ efflux from liposomes. The cuvette. which was thermostated at 3s”C, contained 975 ~1255 mM sucrose. 30 mM Tris-maleate (pH 7.0), 20 PM arsenazo III, and 25 ~1 liposomes loaded with 100 mM CaC& and 0.1 mM Trismaleate (pH 7.0). The absorbance difference 675-685 nm of arsenazo III was used to indicate variations in the Ca” concentration of the medium. For calibration purposes the experiments were started by the addition of Ca*’ The effects of tolbutamide and ghbenclamide were evaluated in relation to those of the Ca” ionophore A-23187 and the nonhypoglycemic glibenclamide analog S-787510.

S-787510, 100 FM glibenclamide evoked a prompt and temporary mobilization of the ion followed by a sustained 200% increase of the efflux rate. In attempts to disclose membrane-labilizing effects of sulfonylureas the efflux of .5,6-carboxyfluorescein from liposomes was also studied (Fig. 2). As expected the surfactant saponin effectively released the dye from the liposomes. Although there was a temporary mobilization of 5.6carboxyfluorescein after addition of A-23187 at a concentration as high as 20 PM, this ionophore lacked a persistent effect on the rate of efflux. Also, 1 mM tolbutamide had no sustained effect. The action of 100 PM glibenclamide was similar to that observed when measuring Ca” efflux in exhibiting a prompt initial release of 5,6-carboxyfluorescein followed by a 300% increase of the efflux rate. Sulfonylureas were also compared to ionophores with respect to effects on black lipid membranes (Figs. 3. 4). The neutral K + ionophore nonactin

HYPOGLYCEMIC

50-

SULFONYLUREAS

I

2.5 mglml Saponin

01 0

4

8

AND Ca” FLUX

249

20pM A-23187

0

4

8

4

8

1OOyM 30-

30-

1lRM

QlJbenclamlda

1

1

Tolbutamide

28.

25-

20. 0

4

8

0 TIME

(MINI

FIG. 2. Effects of sulfonylureas on liposome stability. The cuvette, which was thermostated at 35°C contained 95 ~1 255 mM sucrose, 30 mM Tris-maleate (pH 7.0), and 5 ~1 of 100fold diluted liposomes loaded with 110 mM 5,6-carboxyfluorescein and 15 mM Tris-maleate (pH 7.0). The fluorescence of 5,6carboxylfuorescein (485/515 nm) is given in percentage of the value reached after total liposome lysis. The effects of sulfonylureas were evaluated in relation to those of the detergent saponin and the Ca” ionophore A-23187.

induced a 160-fold increase in the conductance when introduced at a concentration of 0.5 PM into a medium containing Na’ , K ‘, Ca*’ , Mg*+ , and Cl-. However, the conductance remained unaffected in the presence of 2 PM A-23 187, 1 mM tolbutamide, or 100 PM glibenclamide. After addition of the channel-forming quasi-ionophore gramicidin A a characteristic conductance pattern emerged with opening and closing single channels (Fig. 4). Whereas tolbutamide (1 mM) or glibenclamide (100 PM) lacked apparent effects on the frequency of channel formation glibenclamide increased the single-channel conductance by 20%. DISCUSSION

The stimulator-y action of hypoglycemic sulfonylureas on insulin release is associated with a sustained depolorization of the P-cell (18,19) and requires the presence of extracellular Ca*+ (20). It has also been demonstrated that blocking of the voltage-dependent Ca*’ channels in the

250 b0

ln a -

ET Al,.

GYLFE

8001

r O.S,uM

2,uM A-23187

Nonactln

new membrane

new membrane

/ /

ZO-

0” Z

I

,Q

Tolbutamlde

100 tit4 Ghbenclamlde 20-

0

1

2

3

0

I

TIME

(MIN)

2

3

ot-0

I

2

3

FIG. 3. Effects of sulfonylureas on the conductance of black lipid membranes. Black lipid membranes were formed from a solution of 14 mgiml glycerolmonooleate in hexadecane. The membrane separated two chambers kept at room temperature (-22°C) containing 125 mM NaCI, 5.9 mM KCI, 1.28 mM CaCI,, and 1.2 mM MgCl?. Test substances were added to the outer chamber followed by mixing and re-formation of the membrane. The effects of sulfonylureas on the conductance were compared to those of the carboxylic Ca” ionophore A-23187 and the neutral K + ionophore nonactin.

plasma membrane results in secretory inhibition (21). The sulfonylurea stimulation of the entry of Ca*’ is evident from an increased uptake of 4sCa (9). Furthermore, the efflux of the isotope from preloaded islets is increased, presumably due to intracellular Ca-Ca exchange resulting from enhanced entry of nonradioactive Ca” (9.10). In analogy to what is seen with sulfonylureas, exposure of pancreatic p-cells to A-23187 (22) or high K’ concentrations (23) leads to stimulation of insulin release by increased entry of calcium. It has therefore been assumed that the sulfonylureas act either as Ca*’ ionophores (2-6) or as initiators of pcell depolarization with subsequent opening of voltage-dependent Ca*’ channels (7-10). The evidence that sulfonylureas act as Ca*’ ionophores has been obtained exclusively from experiments with artificial systems (2-6). The sulfonylureas have been reported to have effects similar to those of the antibiotic ionophore A-23187 in translocating 45Ca by exchange diffusion from an

LO -I--HYPOGLYCEMIC

SULFONYLUREAS

5 PM Gramicidin

A

AND Ca”

251

FLUX

1mM Tolbutamide

new membrane

5; a

20

% 2 z 3

0 B

OL

1

2

1

I* *

3 ”

8

9

IO

II

‘m

IL) Z 2

40

5 fl

s! w I

Gramicidin

A

new membrane

01

I, I

2

3 ’

I 7 TIME

100 $4 Glibenclamide new membrane

8

9

10 ’

I, I

, I2

I3

I.4

I5

(MIN)

FIG. 4. Effects of sulfonylureas on the conductance of gramicidin A channels in black lipid membranes. Black lipid membranes were formed from a solution of 14 mgiml glycerolmonooleate in hexadecane. The membrane separated two chambers kept at room temperature (-22°C) containing 125 mM NaCI, 5.9 mM KCI, 1.28 mM CaCI,, and 1.2 mM MgC&. Gramicidin A and sulfonylureas were added to the outer chamber followed by mixing and re-formation of the membrane.

aqueous buffer into or across a hydrophobic region consisting of toluene/ butanol(2-5). The Ca2’ translocating activity of A-23187 was also found to be increased in the presence of sulfonylureas (5). Moreover, the sulfonylureas potentiated the activity of an islet extract in facilitating exchange diffusion (2) and enhanced the efflux of 4SCa from liposomes (6). All these observations have been made after addition of the sulfonylureas to the organic phase at concentrations much higher than those required for maximal stimulation of insulin release. The ionophore hypothesis has been the subject of various types of criticism (7-l 1). Despite being used at a high concentration, tolbutamide stimulated the efflux of 45Ca only from the pancreatic islets but not from the neurohypophysis or adrenal medulla (11). Moreover, in chromaffin granules the sulfonylureas did not mediate exchange difkion or potentiate the activity of A-23187. The present use of liposomes loaded with high

252

GYLFE ET AL

concentrations of Ca’+ enabled detection of effects of A-23187 on net fluxes of Ca’+ at ionophore concentrations of less than I% of those used for the chromaffin granules. Nevertheless, tolbutamide at a concentration of I mM failed to enhance the rate of Ca’+ efflux from the liposomes. The observation that 100 PM glibenclamide differed from its nonhypoglycemic analog S-787510 in mobilizing Ca’” should not be taken to indicate an ionophoretic action. When present at this high concentration glibenclamide also labilized the liposomal membranes as indicated by a quantitatively similar release of the “stability marker” 5,6-carboxyfluorescein. As a matter of fact the present data are in agreement with observations that glibenclamide decreases the osmotic resistance of erythrocytes (24) and increases the release of particle-bound P-glucuronidase and acid phosphatase in crude lysosomal fractions from pancreatic islets (25). In accordance with the report of Wulf and Pohl(26) exchange diffusion mediated by A-23187 does not influence the conductance of black lipid membranes. However, as indicated from conductance measurements, the black lipid membrane system enabled a very sensitive detection of the ionophoretic activities mediated by nonactin and gramicidin A. The lack of direct effects of the sulfonylureas on the conductance therefore seems to exclude that these drugs act as neutral ionophores or channel forming quasi-ionophores for Na’ . K’ , Ca’+ , Mg”, or Cl - The apparent increase of the gramicidin A channel conductance after exposure to glibenclamide may follow from an enhanced cation activity close to the channels due to the negatively charged groups of the membrane-bound glibenclamide. SUMMARY

Black lipid membranes and liposomes loaded with Ca*+ or 5,6-carboxyfluorescein were used for exploring the mechanism of action of insulin-releasing sulfonylureas. Unlike the CaZf/H+ exchanging ionophore A-23187, tolbutamide did not stimulate the net efflux of Ca’+ from the liposomes. Glibenclamide caused a sustained release of Cal’, but this effect could be attributed to labilization of the liposomal membrane as indicated by a quantitatively similar loss of the stability marker 5,6carboxyfluorescein. Unlike the neutral ionophore nonactin or the channel forming quasi-ionophore gramicidin A, the sulfonylureas did not alter the conductance of black lipid membranes in medium containing Naf , K+, Ca*+, Mg2+, and Cl-. It is concluded that the sulfonylureas tested lack ionophore properties but that glibenclamide can labilize membranes. ACKNOWLEDGMENTS Financial support was provided by the Swedish Diabetes Association, the Swedish Medical Research Council (12x-6240. 12x-562. 14x-4138). the Swedish Natural Science

HYPOGLYCEMIC

SULFONYLUREAS

AND Ca” FLUX

253

Research Council, and the Swedish Council for Planning and Coordination of Research. The authors are indebted to Marianne Lindfors and Anna-Carin Gjteg for skillful technical assistance.

REFERENCES I. Hellman, B., and Taljedal, 1. B., in “Insulin II” (A. Hasselblatt and F. v. Bruchhausen Eds.), pp. 175-194. Springer-Verlag. Berlin. 1975. 2. Anjaneyulu, R., Anjaneyulu, K., Couturier, E., and Malaisse, W. J.. Biochem. Pharmacol. 29, 1879 (1980). 3. Couturier, E., and Malaisse, W. J., Diabetologiu 19, 335 (1980). 4. Couturier, E., and Malaisse, W. J., Arch. fnt. Pharmacodyn. 245, 323 (1980). 5. Couturier, E., and Malaisse, W. J., &it. J. Pharmucol. 71, 315 (1980). 6. Deleers, M., Couturier, E., Mahy, M.. and Malaisse. W. J., Arch. Znt. Pharmacodyn. 246, 170 (1980). 7. Henquin, J.-C., Diabetologia 18, 151 (1980). 8. Henquin, J.-C., and Meissner. H. P., Biochem. Pharmacol. 31, 1407 (1982). 9. Hellman, B., Mol. Pharmncol. 20, 83 (1981). 10. Hellman, B., Pharmaco/. Res. Commun. 14, 701 (1982). 11. Gylfe, E., and Hellman, B., Mol. Pharmacol. 22, 715 (1982). 12. Ralston, E., Hjelmeland, L. M., Klausner, R. D.. Weinstein, J. N.. and Blumenthal, R., Biochim. Biophys. Acta 649, 133 (1981). 13. Szoka, F., Jr., and Papahadjopoulos, D., Proc. Nat/. Acad. Sci. USA 75,4194 (1978). 14. Scarpa, A., Brinley. F. J.. Tiffert, T., and Dubyak, G. R., Ann. N. Y. Acad. Sci. 307, 86 (1978). 15. Chance, B., Legallais, V., Sorge, J., and Graham, N., Anal. Biochem. 66,498 (1975). 16. Gregoriadis. G., and Davis, C., Biochem. Biophys. Res. Commun. 89, 1287 (1979). 17. Neher. E., Sandblom, J., and Eisenman. G., J. Membrune Biol. 40, 97 (1978). 18. Matthews, E. K., Dean, P. M., and Sakamoto, Y., in “Pharmacology and the Future of Man: Proceedings, 5th International Congress on Pharmacology,” Vol. 3, “Problems of Therapy” (G. T. Okita and G. M. Acheson, Eds.). pp. 221-229. Kruger, Basel, 1973. 19. Meissner, H. P., and Atwater, I. J., Horm. Metabol. Res. 8, 11 (1976). 20. Curry, D. L., Bennett, L. L., and Grodsky. G. M.. Amer. J. Physiol. 214, 174 (1968). 21. Malaisse, W. J., Devis, G., Pipeleers. D. G., and Somers, G.) Diabetologia 12, 77 (1976). 22. Hellman, B., Biochim. Biophys. Acta 399, 157 (1975). 23. Henquin, J.-C., and Lambert, A. E., Diabetes 23, 933 (1974). 24. Ariens, E. J., Acta Diabetol. Lat. B,(Suppl. I), 143 (1969). 25. Gylfe, E., Diabetologia 7, 400 (1971). 26. Wulf. J.. and Pohl, W. G., Biochim. Biophys. Acta 465, 471 (1977).