Proc. Nad. Acad. Sci. USA Vol. 83, pp. 8943-8946, December 1986 Biochemistry Effects of glucagon and vasopressin on hepatic Ca2l release (mitochondria/carbonylcyanide p-trifluoromethoxyphenylhydrazone) NAOMI KRAUS-FRIEDMANN Department of Physiology and Cell Biology, The University of Texas Health Science Center, P.O. Box 20708, Houston, TX 77225 Communicated by Rachmiel Levine, August 22, 1986

ABSTRACT The effects of physiological levels of glucagon Preliminary results of these experiments have been on Ca2" efflux were examined in the perfused rat liver. Two reported. * methods were used to estimate Ca2' efflux: (i) prior labeling of the calcium pools with "5Ca2' and (ii) measurement of perfus- MATERIALS AND ate Ca2' with atomic absorption. According to both methods, METHODS glucagon administration at the physiological level evoked Ca2+ Animals. Fed Sprague-Dawley male rats weighing around release. The released Ca2+ originated mostly from a carbon- 130-200 g were used in all of the experiments. ylcyanidep-trifluoromethoxyphenylhydrazone (FCCP)-deplet- Chemicals. Carbonylcyanide p-trifluoromethoxyphenylhy- able pool and also from an FCCP-insensitive pool from which drazone (FCCP) was from Aldrich; A23187, from Behring Ca2+ could be released with A23187. Maximally effective doses Diagnostics, San Diego, CA; glucagon, from Eli Lilly; and of glucagon and vasopressin had no additive effect on Ca21 vasopressin and albumin (bovine, Cohn fraction V), from release. Prior administration of vasopressin resulted in mark- Sigma. edly reduced Ca2' release by glucagon. These results indicate Methods. Livers were perfused in situ with Krebs-Ringer that glucagon releases Ca2+ from the same pool that vasopres- bicarbonate (KRB) buffer that, if not otherwise indicated, sin does. contained 4% bovine albumin. The perfusion system has been described in detail (5). Drugs and hormones, when indicated, It is generally recognized that a-adrenergic agents and were added to the perfusate through the cannula leading to vasopressin stimulate gluconeogenesis by a Ca2l-dependent the portal vein. To label the hepatic Ca2+ pools with 45Ca2 , mechanism (1-3). Though it has been reported that glucagon livers were perfused in a recirculating system for 90 min with and cyclic AMP as well as catecholamines mobilize intracel- perfusate containing 0.5-1.0 ,Ci (1 Ci = 37 GBq) 45Ca2+ per lular Ca2l from the perfused liver, in these earlier studies the ml. The test period shown in the figures was started by hormones were used in unphysiologically high concentra- switching to Ca2+-free KRB buffer containing 0.2 mM tions (4-6). In more recent investigations in which physio- EGTA. Instead of being recirculated, the perfusate was then logical levels ofglucagon were used, an increase in cytosolic- collected into plastic vials. free Ca2+ was shown to follow hormone administration (7, 8). For measuring the radioactivity of the perfusate, 100-,u Physiological levels ofglucagon were also shown to decrease aliquots were mixed with Protosol. Subsequently, 5 ml of the Ca2+ content of isolated hepatocytes (9). scintillation fluid was added, and the samples were assayed These later studies, while establishing that in isolated liver in a liquid scintillation counter. cells glucagon at physiological concentration causes a redis- The total Ca2+ in the perfusate was measured by standard tribution of Ca2+, did not identify the intracellular Ca2+ techniques with a Perkin-Elmer atomic absorption spectro- pool(s) affected by the hormone. Many previous reports photometer. The liver Ca2+ content was determined after an indicated that the mitochondrial Ca2+ pool was the hormone- overnight ashing at 600°C. Subsequently, the samples were sensitive Ca2+ pool (summarized in ref. 10). However, more diluted with 0.1 M HCl containing 0.01 M SrCl2, and Ca2+ recently, the possible importance of the endoplasmic retic- was determined as in the perfusate. Calculation of Ca2+ ulum Ca2+ pool was emphasized (11). In the light ofthe above content was corrected for extracellular Ca2+. Extracellular reports, it seemed necessary to evaluate the effects of volume was taken to be 28% based on prior determinations physiological amounts of glucagon on hepatic Ca2' distribu- with [14C]sucrose (6). tion and to identify the hormone-sensitive pool(s) from which Results are expressed as means + SEM. glucagon releases Ca2+. Thus, the present studies were undertaken to clarify the RESULTS AND DISCUSSION following questions. (i) Does glucagon at physiological concentrations elicit The effects of various amounts of added glucagon on the Ca2' efflux from the perfused liver? release of Ca2+ from perfused rat livers are shown in Fig. 1. (ii) What is the intracellular pool from which glucagon The high dose of glucagon, 5 nM, released 114 + 57 nmol of releases Ca2 ? Ca2+ per g of liver. The low dose, 0.2 nM, released 46 + 10 In order to identify the hormone-sensitive Ca2+ pool, a nmol of Ca2+ per g of liver in one set of experiments and 30 method reported and previously employed by several labo- ± 18 nmol of Ca2+ per g of liver in an additional series of ratories was used (12-14). In this method, mitochondrial experiments. Thus, a maximally effective dose of glucagon Ca2' is released by a mitochondrial uncoupler. The effect of released about 5% of the total liver Ca2+ (to be discussed hormones on Ca2l release after such uncoupler administra- later). tion is measured. A decreased release is taken as an indica- tion that the pool is entirely or partially mitochondrial. Abbreviation: FCCP, carbonylcyanidep-trifluoromethoxyphenylhy- drazone. *Preliminary results were presented at the International Symposium The publication costs of this article were defrayed in part by page charge on Glycolysis and Gluconeogenesis held in Basel, Switzerland, in payment. This article must therefore be hereby marked "advertisement" October 1984 and at the XII Congress of the International Diabetes in accordance with 18 U.S.C. §1734 solely to indicate this fact. Federation, held in Madrid, Spain, in September 1985. 8943 Downloaded by guest on September 24, 2021 8944 Biochemistry: Kraus-Friedmann Proc. Natl. Acad. Sci. USA 83 (1986) 140 was observed. The lowest amount ofglucagon that showed a release was 0.2 nM, while 0.5 nM did not give a response. No 130 additional amounts were examined between these two doses. 120 These experiments prove unequivocally that 0.2 nM glucagon releases Ca2" from the liver. This hormone con- 110 centration is within the physiological range (15-17). These , 100 _ data complement previous demonstrations with quin-2 show- ing that physiological levels of glucagon increase cytosolic o 90 _ Ca2+ in isolated hepatocytes (7, 8) and decrease hepatic Ca2+ E 8080 content (9). These studies apparently disprove previous arguments that physiological levels of glucagon do not affect = 70 hepatic Ca2' distribution (1, 18-20). The effect ofglucagon is ,u: 60 probably mediated by cyclic AMP because cyclic AMP was shown in earlier studies to mimic the effects of glucagon on 50 * ion fluxes (4-6, 9). In the studies with 45Ca2+, the effects of sequential hor- r 40 mone administration on Ca2' release were also examined. 30- Glucagon administration after a previous maximally effec- tive dose of glucagon (Fig. 2) was not followed with Ca2+ 20- release. This might have been partially due to the refractori- 10 ness of the glucagon receptor because cyclic AMP after glucagon did evoke a second smaller burst of Ca2+ efflux 5 0.5 0.25 0.25 (not shown). It is also possible that this dose of glucagon (10 nM) depleted most of the glucagon-sensitive Ca2+ pool(s). Glucagon, nM When nonmaximal doses of hormones were given, sequen- FIG. 1. The effects of decreasing doses of glucagon on Ca2+ tially administered equimolar glucagon and vasopressin both efflux. Livers were perfused as described. Glucagon was added 13 released Ca2+. These experiments showed that we could use min after switching to a Ca2+-free perfusate. The experiments labeled the method of sequential hormone administration in experi- with an asterisk represent two different sets of experiments carried ments aimed at evaluating whether the two hormones release out in identical conditions. Each group represents three to four Ca2+ from the same or different intracellular pool(s). experiments. The result of these studies is presented in Table 1. Glucagon released 107 + 57 nmol ofCa2+ per g ofliver. Thus, Similar results were obtained when Ca2+ release was the amount of Ca2+ released by glucagon was about 20-25% followed in 45Ca2+-loaded livers. A dose-dependent release less than the Ca2+ released by equimolar amounts of vaso-

200 A 150 D 160 120L 120 90L 80 60 L 40 301- G li v

- 300 300 0.Q 240 240 C~)CQ 180 180

0. 120 120 60 60 + G v G V CZ . .. 0 V,

100 C 200 80 160 60 120 40 80 V^? -6; 20 V G 40 ,n 5 10 15 20 25 30 0 5 10 15 20 25 30 Time, min FIG. 2. The effects of decreasing doses ofglucagon (G) and vasopressin (V) on the release of preloaded 45Ca2+ livers perfused as described. The final concentrations of the added hormones were: 10 nM (A), 5 nM (B), 1 nM (C), 0.2 nM (D), 0.05 nM (E), and 0 in the control (F). Two to four experiments were carried out with each dose, resulting in similar responses. The result of one experiment in each group is presented. Downloaded by guest on September 24, 2021 Biochemistry: Kraus-Friedmann Proc. Natl. Acad. Sci. USA 83 (1986) 8945

Table 1. Combined effects of glucagon and vasopressin on Table 2. Effects of FCCP on the glucagon-stimulated Ca2l efflux Ca2' release Ca2+ released, Treatment Ca2t, nmol/g of liver Time, nmol/g 1st hormone 2nd hormone (wet weight) Additions min* n of liver (wet weight) FCCP 10 6 90 ± 29 Glucagon 107 ± 45 142 16 FCCP 23 2 102 Vasopressin ± A23187 after FCCP 23 3 69 ± 26 Vasopressin 89 ± 46 Glucagon 16 6 FCCP ± after glucagon 10 6 58 ± 27 Glucagon 13 3 107 ± 45 Glucagon/vasopressin together 149 ± 20 Glucagon Livers were perfused as described. The first hormone (5 ,uM) was after FCCP 16 6 26 ± 9 added after 13 min; the second hormone (5 ,uM), after 23 min. The A23187 results are based on the determination of perfusate Ca2' by atomic after glucagon absorption. The number of experiments in each group was three. and FCCP 23 6 51 ± 16 The final concentrations were: FCCP, 50 ;LM; A23187, 50 /hM; pressin. When the two hormones were added simultaneously, glucagon, 5 ;LM. n, Number of experiments. their effects were not additive, the amount released being *Time of the addition of the test agent to the liver after the switch to similar to the amount released by vasopressin alone. the Ca2+-free medium. Vasopressin administered after glucagon was able to re- lease more Ca2+ than was released when glucagon was might be involved in the glucagon- or cyclic AMP-evoked administered after vasopressin. These results, and the fact hormonal effects. These results support earlier observations that glucagon and vasopressin do not have additive effects, by Chen et al. (12) and others (10, 27). indicate that glucagon releases Ca2+ from the same intracel- The possible involvement of the mitochondria in the lular pool(s) as does vasopressin. Vasopressin is either more hormonally evoked Ca2" release has been questioned on the effective in releasing Ca2+ from that pool(s) or more ef- basis of studies obtained with permeabilized hepatocytes fectively releases Ca2+ from an additional pool less accessi- (11). Although FCCP, when applied to isolated subcellular ble to glucagon. Similar results were reported with sequential fractions, releases Ca2l specifically from the mitochondria administration ofphenylephrine and glucagon in the perfused and not from the endoplasmic reticulum, the possibility that liver (21) and recently with the combination of glucagon and it acts differently in intact cells cannot be excluded. We found vasopressin in isolated hepatocytes (22). that FCCP-in the experimental conditions employed-does The amounts of Ca2+ released by maximally effective lower total hepatic ATP levels considerably (unpublished amounts ofglucagon and vasopressin were close to the values data). While lack ofATP does not trigger Ca2' efflux from the reported to be released by vasopressin and phenylephrine in endoplasmic reticulum directly, it might induce Ca2' release other perfused liver systems. Thus, Kleineke and Soling indirectly by preventing its uptake. Such fast Ca2' release reported that vasopressin releases 97 ± 11 nmol of Ca2+ per might be caused by ATP depletion only, if the turnover rate g of liver (23); Reinhart et al. (24), 80-120 nmol of Ca2+ per of the endoplasmic reticulum Ca2+ pool were very rapid. g ofliver; and Althaus-Salzmann et al. (25), 116 ± 20 nmol of Thus, because of the marked ATP depletion that follows Ca2+ per g of liver for phenylephrine. In these latter two FCCP administration, the interpretation of the results ob- studies, the effect of glucagon was also examined, but the tained with this agent is equivocal. However, based on the results showed considerably less release with glucagon. presented data, the relative importance of the different Reinhart et al. reported a release ofabout 30 nmol ofCa2+ per intracellular Ca2+ pools in the hormonal response has yet to g of liver; Althaus-Salzmann reported almost none. These be defined. latter studies were done in low-Ca2+ perfusate, which might The mechanism by which glucagon releases Ca2+ from contribute to the discrepancy. either the mitochondria or the endoplasmic reticulum is not A pivotal question in connection with Ca2+ release evoked known. Glucagon was not reported to be among the hor- by each stimulant is the identity of the source of its release. mones that affect phosphoinositol metabolism in the liver, a To explore this question, the approach outlined in refs. 12-14 mechanism which is indicated in the hormonal release of was used. In short, the effects of a mitochondrial uncoupler, Ca2+ from the endoplasmic reticulum (11). The elucidation of

which is known to release mitochondrial Ca2 , on the the exact mechanism by which glucagon alters Ca2' and glucagon-evoked Ca2+ efflux were measured. The results other ion distributions will be necessary for the full under- (Table 2) show that in the experimental conditions used, standing of its mode of action. glucagon releases Ca2" both from the mitochondrial and the microsomal pools. This conclusion is based on the observa- The excellent technical assistance of Sandra Higham and Arden tion that the amount of Ca2+ released after FCCP treatment Chan is fully appreciated. This work was supported by Grant is considerably less than the amount released without prior PCM-8114547 from the National Science Foundation. FCCP administration. Moreover, glucagon administration before the addition of FCCP the 1. Reinhart, P. H., Taylor, W. M. & Bygrave, F. L. (1984) considerably reduces Biochem. J. 223, 1-13. amount of Ca2+ released by the latter. Because glucagon 2. Williamson, J. R., Cooper, R. H. & Hoek, J. B. (1981) Bio- releases some Ca2+ after the mitochondrial pool was depleted chim. Biophys. Acta 639, 243-295. by FCCP, it seems that glucagon also releases Ca2+ from an 3. Tolbert, M. E. M., Butcher, F. R. & Fain, J. N. (1973) J. Biol. FCCP-insensitive pool-probably from the endoplasmic Chem. 248, 5686-5692. reticulum. 4. Friedmann, N. & Park, C. R. (1968) Proc. Natl. Acad. Sci. In a publication by Somlyo et al. (26), it was estimated that USA 61, 504-508. the mitochondrial Ca2+ content corresponds to 5% of the 5. Friedmann, N. & Rasmussen, H. (1970) Biochim. Biophys. total Ca2t in the liver. In the present study, the total Acta 222, 41-52.

FCCP-releasable Ca2t is also about 5% of the total cell Ca2 . 6. Friedmann, N. (1972) Biochim. Biophys. Acta 274, 214-225. Thus, based on the presented data, the mitochondrial calcium 7. Charest, R., Blackmore, P. F., Berthon, B. & Exton, J. H. Downloaded by guest on September 24, 2021 8946 Biochemistry: Kraus-Friedmann Proc. Nadl. Acad. Sci. USA 83 (1986) (1983) J. Biol. Chem. 258, 8769-8773. J. H. (1978) J. Biol. Chem. 253, 4851-4858. 8. Sistare, F. D., Picking, R. A. & Haynes, R. C., Jr. (1985) J. 19. Reinhart, P. H., Taylor, W. M. & Bygrave, F. L. (1984) Biol. Chem. 260, 12744-12747. Biochem. J. 220, 43-50. 9. Mauger, J. P. & Claret, M. (1985) FEBS Lett. 195, 106-110. 20. Studer, R. K., Snowdowne, J. W. & Borle, A. B. (1984) J. 10. Kraus-Friedmann, N. (1984) Physiol. Rev. 64, 170-259. Biol. Chem. 259, 3596-3604. 11. Burgess, G. M., Irvine, R. F., Berridge, M. J., McKinney, 21. Kimura, S., Kugai, N., Tado, R., Kojima, I., Abe, K. & J. S. & Putney, J. W., Jr. (1984) Biochem. J. 224, 741-746. Ogata, E. (1982) Horm. Metab. Res. 14, 133-138. 12. Chen, J. L. J., Babcock, D. F. & Lardy, H. A. (1978) Proc. 22. Combettes, L., Berthon, B., Binet, A. & Claret, M. (1986) Natl. Acad. Sci. USA 75, 2234-2238. Biochem. J. 237. 13. Beilomo, G., Jewell, S. A., Thor, H. & Orrenius, S. (1983) Proc. Natl. Acad. Sci. USA 79, 6842-6846. 23. Kleineke, J. & Soling, H. D. (1985) J. Biol. Chem. 260, 14. Joseph, S. K., Coll, K. E., Cooper, R. H., Marks, S. & 1040-1045. Williamson, J. R. (1983) J. Biol. Chem. 258, 731-741. 24. Reinhart, P. H., Taylor, W. M. & Bygrave, F. L. (1982) 15. Galbo, H., Holst, J. J. & Christensen, N. J. (1975) J. Appl. Biochem. J. 208, 619-630. Physiol. 38, 70-76. 25. Althaus-Salzmann, M., Carafoli, E. & Jacob, A. (1980) Eur. J. 16. Luyckx, A. S., Dresse, A., Cession-Fossion, A. & Lefebvre, Biochem. 106, 241-248. P. J. (1975) Am. J. Physiol. 229, 376-383. 26. Somlyo, A. P., Bond, M. & Somlyo, A. V. (1985) Nature 17. Jaspan, J. B. & Rubenstein, A. H. (1971) Diabetes 26, (London) 314, 622-625. 887-902. 27. Joseph, S. K. & Williamson, J. R. (1983) J. Biol. Chem. 258, 18. Blackmore, P. F., Brumley, F. T., Marks, J. L. & Exton, 10425-10432. Downloaded by guest on September 24, 2021