Proc. Nat. Acad. Sci. USA Vol. 69, No. 4, pp. 800-804, April 1972 Glucagon-Stimulated Phosphorylation of Mitochondrial and Lysosomal Membranes of Rat Liver In Vivo (hormone/inorganic phosphate/phospholipids) RAINER N. ZAHLTEN, ABRAHAM A. HOCHBERG, F. W. STRATMAN, AND HENRY A. LARDY Department of Biochemistry and the Institute for Enzyme Research, University of Wisconsin, 1710 University Ave., Madison, Wis. 53706 Contributed by Henry A. Lardy, January 20, 1972 ABSTRACT The pancreatic hormone, glucagon, stimulates the net uptake of inorganic 32P in vivo into rat liver and its incorporation into proteins of microsomes, mitochondrial membranes, and lysosomes. Incorporation into cytosolic proteins was enhanced only slightly by glucagon. More than 95%( of the protein-bound phosphate is present as plhosphoserine. Both the radioactivity and the total amount of protein-bound phosphate are increased after injection of glucagon. Glucagon treatment enhanced 32p incorporation into the alcohol-ether soluble lipids of mitochondria but did not alter the relative distribution of 32p in various plhospholipid species. Glucagon, the hormone produced by pancreatic islet alpha cells, stimulates glycogenolysis, lipolysis, and gluconeogenesis, and induces the synthesis of several enzymes in mammalian liver (1, 2). Sutherland and coworkers (3) have demonstrated that the effects of glucagon are mediated through cyclic adenosine monophosphate (camp), formed by glucagon-activated adenyl cyclase in the liver plasma membrane (4). One of the actions of adenyl cyclase is to activate protein kinases (5) that specifically phosphorylate certain enzymes (6), histones (7), ribosomal proteins (8, 9), membrane proteins of neurotubules (10), and renal medullary plasma membranes (11). Recently, Guder et al. (12) and Vavrfnkovd and Mosinger (13) found that glucagon treatment stimulates the release of lysosomal enzymes. The lipolytic and proteolytic activity of these released enzymes probably lead to the increased ketogenesis and urea output observed in perfused livers supplemented with glucagon (14). Electron microscopy of liver tissue removed from animals treated with glucagon reveals rapid and striking swelling of lysosomes up to 800% of their normal volume and alterations of liver cell structure (15-17). During an investigation of the function of glucagon, we have observed that this hormone strikingly enhances the phosphorylation of mitochondrial, microsomal, and lysosomal membranes in rat liver. Thus, membrane protein phosphorylation may be a general mechanism for the mediation of the action of glucagon on the liver cell. MATERIALS AND METHODS Animals and Materials. Male rats, 200-220 g (Badger Research Corp, Madison, Wis.), were fasted for 48 hr before the experiments. The nutritional state of the animals was similar to that described by Guder et al. (12), who observed Abbreviations: i.p., intraperitoneally; But2cAMP, N6,2'-Odibutyryl 3',5'-cyclic AMP. the greatest sensitivity of lysosomes to glucagon with maximum release of lysosomal enzymes in adult rats (over 200 g) after 2 days of fasting. Glucagon (a gift from Dr. Wm. Bromer, Eli Lilly Co., Indianapolis) was injected intraperitoneally (i.p.) in doses of 200 Mug/100 g of body weight. Orthophosphate (Pi) containing 32p (New England Nuclear) was diluted with 0.9% NaCl, and less than 10 /Ag Pi containing 2.5 mci in a volume of 2 ml was injected i.p. in each rat. N6,2'-O-dibutyryl 3',5'-cyclic AMP (But2cAMP), obtained from Calbiochem, was injected i.p. (5 mg/100 g of body weight). Liver tissue of equal weight from three animals of each group was pooled and used for the isolation procedures. For the isolation and purification of lysosomes, pooled livers from seven rats of each group were used. Determinations and Analysis. Acid phosphatase (EC 3.1.3.2) (18), protein (19), and Pi (20) were determined by the methods cited. Alkaline hydrolysis of phosphorylated proteins was performed as described by Loeb and Blat (21). For amino-acid analysis, proteins were hydrolyzed aerobically with 12 N HCl for 24 hr at 37. Amino acids were separated with descending paper chromatography of the radioactive protein hydrolyzate on Whatman no. 3 MM paper with the solvent system butanol-acetic acid-water, 60:15:25 (v/v/v). Lipids, extracted with 95% ethanol-ether, 1:1 (v/v), were chromatographed on Silica Gel thin-layer chromatography plates (22) from Camag Inc., New Berlin, Wis. Individual phospholipids were located on the chromatographic plates with iodine vapor and with the aid of authentic compounds as markers. The gel spots containing the individual lipids were transferred to counting vials containing Bray's solution and counted in a Packard Scintillation Spectrometer, model 3310. Isolation and Purification Procedures. Liver homogenates were prepared in 0.25 M sucrose (ph 7.4) at a dilution of 1: 6 (w/v). Mitochondria, microsomes, supernatant fraction (cytosol), inner and outer membrane of mitochondria, and mitochondrial soluble fraction were separated as described by Sottocasa et al. (23). Mitochondrial membrane preparations were checked for purity with the polyacrylamide electrophoresis system of Schnaitman (24). Monoamine oxidase distribution was determined according to Tabor et al. (25) and Allman et al. (26). Rat liver lysosomes were prepared and purified to lysosomal "pellet 2" as reported by Ragab et al. (27) with slight modifications: liver homogenates were propared in 0.25 M sucrose-1 mm EDTA (ph 7.0) with 1:4 800
(w/v) dilution. Additional extensive washings of lysosomal pellets with sucrose gave lower yield but higher purity. Specific activities of the acidic phosphatase of lysosomes were in the range of 625-650 nmol Pi liberated per mg protein per min at 37. Lysosomal membranes were prepared by sonication of "pellet 2" (27). Separation of Proteins. All subcellular preparations were treated with equal volumes of 10% C13CCOOH (v/v) and precipitated at 1700 X g for 10 min; the 5% acid supernatant fraction was saved. Three additional washings of the sediment with 5% CLiCCOOH followed. The precipitate was then suspended in 5 ml 10% C13CCOOH (v/v) and heated for 30 min at 900 in a water bath for extraction of nucleic acids. The suspension was centrifuged at 1700 X g for 10 min, the supernatant fraction was collected, and the precipitate was washed three times with 5% C16CCOOH. The last sediment was treated with 95% ethanol-ether, 1:1 (v/v), heated at 40 for 15 min, and again centrifuged. The supernatant fraction was saved, and the protein precipitate was washed with ether. The protein was finally dissolved in 1 N NaOH and Triton X-100, and the incorporation of 32p was determined. Portions of the purified proteins were incubated for 18 hr at 370 in 1 N NaOH, and the liberated orthophosphate was determined (20). RESULTS Time-Dependent Incorporation of 32p into Proteins of Different Cell Compartments. We performed a study of 32p incorporation into liver proteins of untreated rats as a function of time to establish the experimental conditions under which the influence of glucagon was to be tested. Maximum incorporation of 32p in all isolated protein fractions was achieved 90 min after its injection (Fig. 1). Fig. 1A demonstrates that the most rapid incorporation and highest specific activity was obtained in microsomal proteins. Cytosolic (105,000 X g for 1 hr) proteins and mitochondrial proteins reached only 25-30% of the incorporation of the microsomal proteins by 90 min. Separation of the proteins from mitochondrial membranes and the soluble fraction (Fig. 1B) yielded specific activities different from those observed in total mitochondrial proteins (Fig. 1A). The specific activity of inner membrane proteins was about 60% higher than that of the total mitochondrial proteins and more than double those of the outer membrane and soluble mitochondrial fraction. 32p Incorporation into Proteins After Glucagon Treatment. All rats were killed 120 min after 32p injection. Glucagon was --5 04 1;3 0~ X 2 I' V N 0 30 60 90 120 30 60 90 120 MINUTES FI'G. 1. Time-dependent incorporation of 32JP into proteins of different cell components. Each point represents three pooled livers. (A) Cell fractions: 0, mitochondria; Q, microsomes; A, cytosol (105,000 X g for 1 hr). (B) Mitochondrial fractions: *, inner membrane; *, outer membrane; A, soluble fraction. Glucagon and Protein Phosphorylation 801 injected at different times during this 120-min period so that different groups of rats were exposed to its effect for 15, 30, 60, 90, 105, or 120 min. But2cAMP was always injected at 60 min of the 120-min experiments. Glucagon increased 32p incorporation into the proteins of all hepatic cell components analyzed. 82p incorporation was maximally enhanced when glucagonwas injected 90 min before the ratswere killed (column Ggo, Fig. 2A-F). The most striking effects of glucagon were seen in microsomal, mitochondrial, and lysosomal proteins. When lysosoines were treated with sonic oscillation, sedimented, and washed before the extraction procedure was started, all of the original radioactivity (within experimental counting error) was retained in the sedimented membrane fraction and little or none was found in the soluble fraction. Thus, in both lysosomes (insert, Fig. 2B) and mitochondria (Fig. 2E and F), the greatest effects of glucagon on 32p incorporation was in the membrane protein fraction. Glucagon enhanced 32p incorporation into cytosolic proteins only slightly. But2cAMP at the dose and time tested was much less effective. 32p Distribution in Different Extracts of Isolated Cell Fractions. The data in Table 1 demonstrate that glucagon causes a greater percentage of the 82p taken up by liver mitochondria to be incorporated into compounds extractable with hot 10% trichloroacetic acid (includes nucleic acids) and with ethanolether (mainly lipids). A similar distribution pattern and cor- 4 Z 3 I I0 x 2 CL) UuI 4 3 2 E In ~~~~ ZIO-j FIG. 2. Influence of glucagon and But2cAMP on incorporation of 32p into proteins of hepatic cell components. All rats were killed 120 min after 32p injection. The control (shaded column.) received no glucagon. Each column represents three pooled livers. (A) microsomes; (B) cytosol (105,000 X g for 1 hr); (C) mitochondria; (D) soluble mitochondrial fraction; (E) outer mitochondrial membrane; (F) inner mitochondrial membrane. Lysosomal membranes (insert in B): C control, 90 min after = 32p injection, G 30 min after glucagon administration and 90 = min after 32p injection. For other abbreviations, see Table 1. F 0 00 n22
802 Biochemistry: ZahIten et al. TABLE 1. Distribution of 32p in extracts of hepatic mitochondria and lysosomes* 32p in fraction Total Cold Hot cpm 5% 10% from C13C- C13C- Ethanol- Pro- 1 g wet COOH COOH ether tein Group liver (%) (%) (%) (%) Mitochondriat C30 3.6 90 6.2 3.3 0.6 C60 8.9 80 6.3 13.0 1.0 Coo 16.0 72 9.4 17.0 1.6 C120 11.7 75 7.7 16.0 1.0 G15 14.8 72 9.2 18.0 1.2 Go0 17.2 63 12.0 24.0 1.5 Grso 24.2 67 11.0 21.0 1.3 Ggo 25.4 68 9.8 21.0 1.4 G120 19.0 67 11.0 21.0 1.4 D60 16.0 72 9.8 18.0 1.0 Lysosomest C120 1.6 36 30.0 29.0 5.4 G6o 2.3 34 31.0 31.0 4.6 * Experimental values are for three (mitochondria) or seven (lysosomes) pooled livers. C30 to C120 = no glucagon; subscript indicates no. of min after 32p injection. G11 to G120 = glucagon present for the no. of min indicated by subscript; 32P was always injected 120 min before the rats were killed. Group C120 (no glucagon) serves as control for the glucagon-treated rats. D60 = But2cAMP present for 60 min; 32p present for 120 min. Lysosomes were separated from livers 120 min after 32p was administered to the rats. Glucagon was given to the last group in the table 60 min after 32p injection and 60 min before the rats were killed (G60)- t In mitochondria, cpm X 10-6. In lysosomes, cpm X 10-4. responding effects of glucagon were found for hepatic microsomes and for the inner and outer membrane fraction of mitochondria. The relative distribution of label in the different extracts of the lysosomes is quite different from all the other cell components. Although glucagon enhanced total radioactivity in lysosomes, there were no apparent differences in percentage distribution of 32p between the lysosomes of untreated and glucagon-treated animals. About 99% of the total radioactivity of the cytosol (105,000 X g for 1 hr) and the soluble mitochondrial fraction was in the 5% C13CCOOH soluble fraction (data not shown). Distribution of 32p in- Phospholipids. Although the total incorporation of 32p into phospholipids was enhanced by glucagon and by But2cAMP, the percentage distribution among the various phospholipids was not affected by these agents. Table 2 presents the 82p distribution in various phospholipids from the mitochondria and lysosomes. The distribution in microsomes and mitochondrial membranes (not shown) was similar to that of whole mitochondria. In the latter, the highest percentage of 32p incorporated was always found in compounds with the mobility of phosphatidyl ethanolamine, phosphatidyl inositol, and phosphatidyl choline. Sphingomyelin plus lysophosphatidyl-choline contained about half of the radioactivity of each of the three phospho- lipids mentioned above. Phosphatidyl serine and the fraction containing neutral lipids, cardiolipin, and phosphatidic acid incorporated the least 82p. The distribution of 32p in phospholipids of the lysosomal fraction is very different from that of all other cell components. Glucagon enhanced 82p incorporation into the ethanol-ether soluble fraction from lysosomes, but had no apparent influence on the percentage of label present in the different phospholipids. Chromatographic Analysis of Labeled Amino Acids and Determination of Alkali-Labile Phosphate. In all chromatographic separations of the acid hydrolyzed proteins, almost no radioactivity could be detected in the position of inorganic phosphate. Almost all of the radioactivity was confined to the spot of mobility identical with that of authentic phosphoserine used as marker. Quantitative assays for alkali-labile phosphate (mainly phosphoserine) in the proteins of various cell components are reported in Table 3. Glucagon increases the content of alkalilabile phosphate in the extracted proteins of mitochondria, the inner and outer membrane of mitochondria, and in lysosomes. About 95% of all 82p incorporated into proteins was released as alkali-labile phosphate. This corresponds with the amount of radioactivity incorporated into phosphoserine. Any incorporation into phosphothreonine must, therefore, be negligible. DISCUSSION In the present study it has been shown that glucagon treatment in vivo for different periods of time increases the incorporation of 82p into proteins and other components of mitochondria, microsomes, and lysosomes (Fig. 2A-F). The increased incorporation into proteins caused by glucagon is accounted for by 32p in phosphoserine residues. Increased incorporation of 32p into the proteins could be the result of two different processes: (1) increase of the specific activity of the intracellular phosphate pool through facilitated uptake of 82p into the liver in the presence of glucagon; or (2) specific hormonal stimulation of intracellular protein phosphorylating processes. It is known (28, 29) that glucagon decreases serum phosphate and increases uptake of phosphate in liver and heart. Rats subjected to the effects of glucagon for 90 min took up 30% more 32p into their livers than the controls (data not shown). This enhanced uptake of 82p into the liver cell was not sufficient to account for the glucagon-induced increase of the specific activities in the proteins of microsomes (84%), mitochondria (184%), soluble mitochondrial fraction (237%), and outer (340%) and inner (162%) mitochondrial membranes, nor was it sufficient to account for the increased incorporation of 82p into hot C13CCOOH extracts and ethanol-ether soluble lipids (Table 1). An influence of glucagon on phosphorylation processes is clearly demonstrated by the fact that this hormone decreased the percent of 82p found in the fractions containing inorganic phosphate. Glucagon-dependent activation of protein-phosphorylating processes via ATP (5) are, therefore, very likely the main mechanism for increasing the 82p incorporation into the membrane proteins reported here, as was shown in the case of glucagon-stimulated histone phosphorylation in vivo (7). Langan (7) also found that glucagon does not produce significant changes in the specific activity of the liver phosphate pools. Although glucagon enhanced the incorporation of IT into the lipid fraction of mitochondria and of lysosomes,
TABLE 2. Distribution of 32p in various lipids of mitochondria and lysosomes* Glucagon and Protein Phosphorylation 803 Neutral Sphingolipids + Phospha- myelin + cardiolipin + tidyl Phospha- Phospha- Phospha- lysophosphosphatidic ethanol- tidyl tidyl tidyl phatidyl acid amine serine inositol choline choline Group Total cpm (%) (%) (%) (%) (%) (%) Mitochondriat C30 0.1 3.2 32.0 4.7 22 27 11 C60 1.1 3.5 26.0 3.6 21 32 15 C9o 2.6 2.6 29.0 3.8 19 31 15 0120 1.9 3.2 31.0 3.5 19 30 13 G,5 2.7 1.9 30.0 2.8 21 31 14 G30 4.1 1.8 29.0 3.8 21 30 14 Gwo 5.1 1.0 31.0 4.6 19 31 14 Ggo 5.4 2.3 29.0 3.7 20 31 14 Go20 4.0 1.9 28.0 3.6 21 32 14 Dw0 2.8 1.7 28.0 2.9 21 32 15 Lysosomesj C120 4.6 17.0 2.7 11.0 20 33 17 G60 7.1 24.0 1.8 11.0 16 31 16 * Conditions as described in legend to Table 1. t Total cpm in ethanol-ether fraction of mitochondria from 1 g wet liver (cpm X 10-i). Total cpm in ethanol-ether fraction of lysosomes from 1 g wet liver (cpm X 10-3). it had no differential effect on the distribution of the isotope in various phospholipids isolated (Table 2). The extreme difference in distribution of 32p in phospholipids of the lysosomes (Table 2), compared with mitochondria, can be explained on the basis of a completely different phospholipid composition of the lysosomal membrane, as was reported by Henning et al. (30). The fact that glucagon enhances 82p incorporation in the microsomal and soluble mitochondrial proteins without influencing the total amount of alkali-labile phosphate (Table 3) indicates a rapid turnover of phosphate in these proteins induced by the hormone. But2cAMP, which generally induces glucagon-like effects in the liver (1, 14), is, under our experimental conditions, much less effective than glucagon. The increases of specific activities of phosphate-labeled proteins in mitochondria (Fig. 2C), as well as in outer and inner membrane (Fig. 2E and F), after injection of But2cAMP are much less impressive than that occurring after injection of glucagon. The possibility of different time curves for camp accumulation and degradation in the liver cell after the use of glucagon or But2cAMP makes a valid comparison of time-dependent effects of these TABLE 3. Alkali-labile phosphate in isolated protein fractions Cell component Control Glucagon (nmol Pi liberated per mg of protein) Cytosol (105,000 X g for 1 hr) 6.3 * 8.3 Microsomes 3.2 3.3 Lysosomes 4.0 19.0 Mitochondria 5.6 8.8 Soluble mitochondrial fraction 6.9 6.6 Outer mitochondrial fraction 5.5 8.3 Inner mitochondrial fraction 4.4 7.9 All rats were killed 90 min after receiving 321p; glucagon was given 30 min before they were killed. Each number represents three pooled livers (for lysosomes, seven livers). agents on protein-phosphorylating processes difficult. Such time differences could explain our failure to detect a significant effect of But2cAMP on 32p incorporation into microsomal proteins (Fig. 2A) at the only time tested. The glucagon-stimulated phosphorylation of mitochondrial and lysosomal proteins, which intracellularly is probably mediated through camp-activated protein kinases (5), can be integrated into the whole pattern of known glucagon effects in rat liver. The catabolic effects of glucagon in rat liver (14), the release of lysosomal enzymes (15-17), and the facilitated uptake of pyruvate into mitochondria (31) may all be influenced by the state of phosphorylation of various. cellular membranes. Phosphorylation of serine residues, mediated through protein kinases, of lysosomal membrane proteins could change membrane properties with increased permeability through enhanced osmotic sensitivity and subsequent swelling (15). The glucagon-induced release of lysosomal enzymes initiates autophagocytosis in rat liver (12, 13) and leads to the well-known increase of urea output (32) and ketogenesis (14). Phosphorylation of mitochondrial membranes or, alternatively, the proteolytic or lipolytic action of lysosomal enzymes could enhance the permeability of mitochondrial membranes and thereby increase pyruvate uptake and carboxylation (31). Electronmicrographs (15-17) confirm digestion of mitochondria in the neighborhood of swollen lysosomes after glucagon injection. Supported in part by grants from the National Institutes of Health (Grant AM-10334) and the National Science Foundation (Grant GB-6676X). F. W. S. held the Babcock Fellowship from the Department of Biochemistry, University of Wisconsin, during this research. 1. Exton, J. H., Malette, L. E., Jefferson, L. S., Wong, E. H. A., Friedman, N., Miller, T. B., Jr. & Park, C. R. (1970) in Recent Progress in Hormone Research, ed. Astwood, E. B. (Academic Press), Vol. 26, pp. 411-457. 2. Greengard, 0. (1969) Biochem. J. 115, 19-24. 3. Sutherland, E. W. & Rall, T. W. (1960) Pharmacol. Rev. 12, 265-299.
804 Biochemistry: ZahIten et al. 4. Sutherland, L. W., Rall, T. W. & Menon, T. (1962) J. Biol. Chem. 237, 1220-1227. 5. Kuo, J. F. & Greengard, P. (1970) Biochim. Biophys. Acta 212, 434-440. 6. Walsh, I). A., Perkins, J. P. & Krebs, E. G. (1968) J. Biol. Chem. 243, 3763-3765. 7. Langan, T. A. (1969) Proc. Nat. Acad. Sci. USA 64, 1276-1283. 8. Kabat, 1I. (1970) Biochemistry 9, 4160-4175. 9. Loeb, J. E. & Blat, C. (1971) FEBS Lett. 18, 124-126. 10. Goodman, 1). B. P., Rlasmussen, H., DiBella, F. & Guthrow, C. L., Jr. (1970) Proc. Nat. Acad. Sci. USA 67, 652-659. 11. J)ousa, T., Sands, H. & Hechter, 0. (1971) Fed. Proc. 30, 200. 12. Guder, W., Hepp, K. 1). & Wieland, 0. (1970) Biochim. Biophys. A cta 222, 593-605. 13. VavrfilkovA, H. & MIosinger, B. (1971) Biochim. Biophys. Acta 231, 320-326. 14. Menahan, L. A. & Wieland, 0. (1969) Eur. J. Biochem. 9, 55~-62. 15. Deter, IR. L. (1971) J. Cell Biol. 48, 473-489. 16. Rlosa, F. (1971) J. Ultrastruct. Res. 34, 205-213. 17. Sawant, P. L., D)esai, I. 1). & Tappel, A. L. (1964) Biochim. Biophys. Acta 85, 93-102. 18. Gianetto, R. & DeDuve, C. (1955) Biochem. J. 59, 433-445. 19. Layne, E. (1957) in Methods in Enzymology, eds. Colowick, Proc. Nat. A cad. Sci. USA '69 (1972) S. P. & Kaplan, N. 0. (Academic Press), Vol. III, pp. 450-451. 20. Sumner, J. B. (1944) Science 100, 413-414. 21. Loeb, J. E. & Blat, C. (1970) FEBS Lett. 10, 105-108. 22. Skipski, V. P., Peterson, R. F. & Barclay, M. (1964) Biochem. J. 90, 374-378. 23. Sottocasa, G. L., Kuylenstierna, B., Ernster, L. & Bergstrand, A. ( 1967) in Methods in Enzymology, eds. Estabrook, R. W. & Pullman, M. E. (Academic Press), Vol. X, pp. 448-463. 24. Schnaitman, C. A. (1969) Proc. Nat. Acad. Sci. USA 63, 412-419. 25. Tabor, C. W., Tabor, H. & Rosenthal, S. M. (1954) J. Biol. Chem. 208, 645-661. 26. Allman, 1). W., Bachman, E., Orme-Johnson, N., Tan, W. C. & Green, D. E. (1968), Arch. Biochem. Biophys. 125, 981-1012. 27. Ragab, H., Beck, C., Dillard, C. & Tappel, A. L. (1967) Biochim. Biophys. Acta 148, 501-505. 28. DeVenanzi, F., Altares, C. D. & Forero, J. (1964) Diabetes 13, 609-614. 29. Foa, P. P. (1968) in "Glucagon," Reviews of Physiology, (Springer-Verlag), Vol. 60, pp. 165-166. 30. Henning, R., Kaulen, H. D. & Stoffel, W. (1970) Hoppe- Seyler's Z. Physiol. Chem. 351, 1191-1199. 31. Adam, P. A. J. & Haynes, R. C., Jr. (1969) J. Biol. Chem. 244, 6444-6450. 32. Miller, L. L. (1960) Nature 185, 248.