Increased Particulate and Decreased Soluble Guanylate Cyclase Activity

Transcription

1 Proc. Nat. Acad. Sci. USA Vol. 72, No. 5, pp , May 1975 Increased Particulate and Decreased Soluble Guanylate Cyclase Activity in Regenerating Liver, Fetal Liver, and Hepatomna (guanosine 3':5'-cyclic monophosphate/plasma membranes/microsomes/cycloheximide) HIROSHI KIMURA AND FRID MURAD Division of Clinical Pharmacology, Departments of Internal Medicine and Pharmacology, University of Virginia, Charlottesville, Va Communicated by Alfred Gilman, March 1, 1975 ABSTRACT We determined the activities of soluble and particulate guanylate cyclase [GTP pyrophosphatelyase (cyclizing); C in regenerating rat liver, fetal and neonatal rat liver, and hepatoma. In these tissues we found increased particulate and decreased soluble enzyme activities compared to normal adult rat liver. The particulate activity increased 12 hr after partial hepatectomy, reached maximal activity at 48 hr, and then declined. The soluble enzyme activity decreased within 8 hr and continued to decline. The activity of homogenates did not change. Guanylate cyclase activity was increased in plasma membrane and microsome fractions from regenerating liver. The increase in particulate activity was prevented with cycloheximide. Decreased soluble and increased particulate enzyme activities were found in fetal liver. After birth the soluble activity increased and the particulate activity decreased. Seven to 14 days after birth the activities of soluble and particulate fractions were similar to those of adult rat liver. In hepatoma 3924A, the activity of particulate guanylate cyclase was 9-fold greater and that of the soluble enzyme was 5% that of normal liver. These studies suggest that guanylate cyclase activity and its subcellular distribution may be related to liver growth through some unknown mechanism. Our laboratory (1-3) and Chrisman et al. (4) have recently reported that soluble and particulate guanylate cyplases [GTP pyrophosphate-lyase (cyclizing); C ], which catalyze the formation of cyclic 3': 5'-GMP from GTP, have different properties and different molecular sizes from various rat tissues. In homogenates of normal adult rat liver about 8% of guanylate cyclase is soluble and 2% is particulate (1-3, 5). We found most of the particulate enzyme in liver to be associated with both purified plasma membranes and microsomes (5). In this report we describe an altered distribution of enzyme in regenerating rat liver after partial hepatectomy. The distribution of activity between soluble and particulate fractions of liver homogenates also changed during fetal and neonatal development of rats. These studies suggested that guanylate cyclase activity and its subcellular distribution may be related to tissue growth and development. Thus, hepatoma 3924A was also examined; most of the homogenate activity was associated with particulate fractions. This report describes an alteration of guanylate cyclase activity and its subcellular distribution in vivo. These studies are of interest in view of recent reports of increased levels of cyclic GMIP in proliferating cells (7-1). Some of the present results have been reported in abstract form (11). MATRIALS AND MTHODS Male Sprague-Dawley rats weighing 15-2 g were fed Purina chow and water ad libitum. After ether anesthesia about two-thirds of the liver was removed or sham operations were performed. At the times indicated after surgery (4 hr to 5 days) livers were removed for assay of guanylate cyclase. Immediately after partial hepatectomy.1 mg of cycloheximide in saline per 1 g of body weight was injected intraperitoneally in some animals; larger doses resulted in death of most animals with partial hepatectomy within 24 hr. In the experiments with cycloheximide 1 ACi of -[3H]Ileucine were administered intraperitoneally 3 min before animals were sacrificed. Pregnant rats were sacrificed after 18 and 21 days of gestation. One litter of rats was used for each time period in the fetal and neonatal experiments. Female ACI rats were inoculated subcutaneously with Morris hepatoma 3924A 47 (lays before sacrifice and kindly sup)plie(d by Dr. W. L, Looney, Department of Radiology, University of Virginia. All animals were sacrificed by decapitation. Livers and hepatomas were quickly removed and placed in cold saline. Hepatomas were dissected free of necrotic tissue. Tissues were homogenized with 9 volumes of.25 MvI sucrose containing 5 mm Tris. HCl buffer, ph 8., in a Potter- lvehjem homogenizer at 4. All subsequent procedures were performed at 4. Homogenates were filtered through gauze, and soluble and particulate guanylate cyclases were separated by centrifugation at 15, X g for 6 min. Particulate fractions were suspended in a volume of fresh solution equal to that of the original homogenates. Plasma membranes and microsomes were l)repared with discontinuous sucrose gradient centrifugation by the method of Touster et al. (12) with minor modification as described previously (5). Some homogenates and fractions were treated with 1% Triton X-1 for 6 min at 4 And chromatographed on Bio-Gel P-3 columns (1). Guanylate cyclase activity was determined as reported previously (1, 5). Reaction mixtures of 1 Ml contained 5 mmi Tris-HCl buffer, ph 7.6, 1 mmi theophylline, 15 mm creative phosphate, 2 lg of creatine kinase ( units/ mm GTP, and 4 mmi M1nCI2 and were incubated at mg), 1 37 for 1-2 mim. Under these conditions less than 5% of added cyclic GMIP was hydrolyzed, and product formed was linear with protein concentration and time. Cyclic GMP formed was determined with a radioimmunoassay method (13, 14). Some values presented are means of duplicate incubations from representative experiments. Protein was determine(l by the method of Lowry et al. (15). Livers from rats injected with L- [H ]leucine (New ngland Nuclear Corp., specific activity 35.8 Ci/mmol) were homogenized as described above. Homogenate (.5 ml) was added to 5 ml of 6% trichloroacetic acid and centrifuged. Precipitates were washed twice with cold 5% trichloroacetic acid, dissolved with 1 ml of 1965

2 1966 Biochemistry: Kimura and Murad 5 C._) 4 L- a) a I 3 I. g o 2 I.. C. in) L lur +=P < Days after partial hepatectomy FIG. 1. Liver guanylate cyclase activity at various times after partial hepatectomy. Rats were sacrificed at the times indicated after partial hepatectomy, and the guanylate cyclase activities of homogenates, 15, X g supernatant, and particulate fractions were determined as described in Materials and Methods. ach point is the mean 4-SM of four to eight rats. Pluses (+) indicate values that are significantly different (P <.5) from the activities on day zero. 1 N NaOH, and analyzed for tritium by scintillation counting. Other reagents were obtained as described previously (1, 5, 13, 14). RSULTS We determined the activity of guanylate cyclase in regenerating rat liver at various times after partial hepatectomy (Fig. 1). While guanylate cyclase activity in homogenates of regenerating liver did not change, distribution of the enzyme between supernatant and particulate fractions was altered. Supernatant activity decreased within 8 hr and remained lower for 5 days; longer times were not examined. The particulate activity increased within 12 hr. The increased activity persisted for several days and returned to control values. The observations were qualitatively similar when activities were expressed as nmol of cyclic GMP formed per min/g of liver (not shown). At 2 days after partial hepatectomy 45% of the homogenate activity was associated with the particulate fraction and 6% with the soluble fraction. As reported previously (1-3, 5) normal liver has about 8% of the guanylate cyclase activity of homogenates associated with the soluble fraction. When preparations from regenerating liver 2 days after partial hepatectomy were assayed after treatment with 1% Triton X-1, 74% of the homogenate activity was derived from particulate fractions. About 3-4% of the homogenate activity from normal adult rat liver assayed after Triton X-1 treatment is derived from particles (2). Particulate guanylate cyclase of normal liver is found predominantly in plasma membranes and microsomes (5). The activities of purified plasma membrane and microsome fractions increased with partial hepatectomy (Table 1). 5 TABL 1. Guanylate cyclase activity of subcellular fractions of liver from sham-operated and partially hepatectomized rats Cyclic GMP formed (pmol/min per mg of protein) Partial % Sham- hepa- Inoperated tectomy crease Crude nuclear preparation Fraction A Crude microsomal preparation Fraction A Fraction C Rats were sacrificed 24 hr after the operation. Crude nuclear and crude microsomal precipitates were purified with discontinuous sucrose gradient centrifugation as described in Materials and Methods. The properties of each fraction are described in a previous report (5). Fraction A from crude nuclear pellets is homogeneous plasma membranes; Fraction A from crude microsomes is plasma membranes with some Golgi apparatus; Fraction C is homogeneous microsomes. Values presented are means of duplicate determinations. As summarized in Table 2, the administration of cycloheximide (.1 mg per 1 g of body weight) prevented the increase in particulate guanylate cyclase activity and the increased incorporation of - [3H ]leucine into protein with partial hepatectomy. In this experiment the decrease in soluble activity with partial hepatectomy was not observed. Cycloheximide alone increased the particulate activity 2-fold, suggesting that particulate guanylate cyclase activity may be inhibited or repressed in normal liver. These studies indicate that the increase in particulate guanylate cyclase activity with partial hepatectomy required increased de novo protein synthesis and that new enzyme was synthesized. The same Q o 1 Peak PeaklI CL6 } Partial - o 4 Hepatectomgy_) 24 Proc. Nat. Acad. Sci. USA 72 (1975) Control ffluent (ml) FIG. 2. Guanylate cyclase activity of control and regenerating liver after Bio-Gel P-3 chromatography. Control liver and a 48-hr regenerating liver were used. Homogenates were incubated at 4 with 1% Triton X-1 for 6 min and centrifuged at 15, X g for 6 min. One and a half milliliters of the supernatant fractions (38. mg and 3.5 mg of protein from control liver and regenerating liver, respectively) were applied to a Bio-Gel P-3 column (2.6 X 34 cm) pre-equilibrated with 2 mm Tris- HCl buffer, ph 7.6, containing.1 mm DTA. Peak I eluted at the void volume.

3 Proc. Nat. Acad. Sci. USA 72 (1975) TABL 2. ffect of cycloheximide on liver guanylate cyclase Alteration of Liver Guanylate Cyclase Activity 1967 Cyclic GMP formed (pmol/min per mg of protein) I- [3H]leucine incorporated Homogenate Soluble Particulate (cpm/mg of protein) Sham-operated i t 2.3 Sham-operated + cycloheximide i * Partial hepatectomy t t.62* * Partial hepatectomy + cycloheximide t.38* * Rats were sacrificed 24 hr after operation. Cycloheximide,.1 mg per 1 g of body weight, was injected intraperitoneally immediately after operation and 1 jsci of i,[3h]leucine was injected intraperitoneally 3 min prior to decapitation. Incorporation of i-[3h]leucine into acid-insoluble materials was determined as described in Materials and Methods. Guanylate cyclase was determined in homogenates and in soluble and particulate fractions after centrifugation at 15, X g for 6 min as described in Materials and Methods. Values are means + SM for three or four animals. Asterisks indicate values that are significantly different (P <.5) from the sham-operated controls. conclusion was drawn from gel filtration studies. Homogenates from control and regenerating livers were solubilized with 1% Triton X-1 and applied to Bio-Gel P-3 columns as described previously (1). Two peaks of guanylate cyclase activity were eluted (Fig. 2). We have reported previously that Peak I has the properties of the particulate enzyme and Peak II the properties of the soluble enzyme (1-3). In control animals most of the activity eluted in the second peak. The converse was true with homogenates from regenerating liver. These experiments suggested that the altered distribution of guanylate cyclase in regenerating liver was not due to a translocation of the soluble enzyme to particulate fractions. Also the particulate enzyme from control and regenerating liver had similar Hill coefficients for GTP of 1.6; the soluble enzyme has a Hill coefficient of 1.2 (5). This would not be expected if the redistribution of activities was due to translocation. We also examined hepatoma 3924A, which has a growth rate of about 7 cm/month and a volume doubling time of 4.35 days (16). This tumor also has very high levels of cyclic GMP in vivo. The activity of the particulate enzyme from hepatomas was increased, while that of the soluble fraction was decreased (Table 3). About 69% of the homogenate activity was recovered in the particulate fraction. Such a high proportion of activity in particulate fractions has only been reported in small intestinal mucosa (1, 17). The activity of particulate guanylate cyclase was also increased about 2-fold in livers of tumor-bearing animals for reasons that are not apparent. Guanylate cyclase activity is about equally divided between supernatant and particulate fractions of homogenates of fetal rat liver (Fig. 3). After birth the homogenate activity (per g of liver) increased and subsequently declined. * However, the specific activity of homogenates (pmol/min per mg of protein) did not change significantly with age. This may be the result of an increased number of liver parenchymal cells and protein and a decrease in hematopoietic cells (18). After birth the activity of supernatant fractions rapidly increased while the particulate activity gradually decreased. Thus, 1-2 weeks after birth the activity and its distribution in soluble and particulate fractions of homogenates were similar to values in adult animals. * We also found that guanylate cyclase activity in homogenates of rat lung increased after birth. Values were nmol/ min per g before birth and fourteen days after birth (mean + SM, n = 4). DISCUSSION After finding guanylate cyclase activity in soluble fractions, plasma membranes, and microsomes of liver (1-3, 5), we designed in vivo and in vitro experiments to determine whether these activities were regulated by different mechanisms. Such information should help in the understanding of cyclic GMP synthesis and its accumulation in tissues. Several recent studies suggested that cyclic G.MP may be related to cell proliferation (7-1) and synthesis of nucleic acids (9, 19, 2) and protein (21, 22). Therefore, we examined some commonly used models of liver proliferation. These included regenerating liver, fetal liver, and a rapidly growing hepatoma line (3924A). In these tissues increased particulate and decreased soluble 'D a i> L- a) c co a..2.7c, am [ 5 [ 4 [ L Homogenate +=P < ,/\\ /- Supernatant / \'l \\I ~ Homogenate, V*,. Particulate --<--. +=P < Age (days) FIG. 3. Guanylate cyclase activity in fetal and newborn rat liver. Guanylate cyclase activities in homogenates and supernatant and particulate fractions after centrifugation at 15, X g for 6 min were assayed as described in Materials and Methods. Values are the means + SMT of three to eleven animals at each age, designated as minus values prior to birth. The pluses (+) indicate values that are significantly different (P <.5) from values 4 days before birth.

4 1968 Biochemistry: Kimura and Murad Proc. Nat. Acad. Sci. USA 72 (1975) TABL 3. Guanylate cyclase activity of hepatoma 3924A and host livers Cyclic GMP formed nmol/min per g of tissue pmol/min per mg of protein Normal liver Host liver Hepatoma 3924A Normal Host Hepatoma % O% % liver liver 3924A Homogenate 3.21 ± i * ± ± ± 1.86* 15, X g supernatant 2.88 ± ± ±.7* ± ± ± 3.4* 15, X g precipitate.44 ± ±.9* ±.23* ± ±.77* ± 3.13* Recovery ACI rats were inoculated with hepatoma 3924A and used after 47 days. Values presented are means ±SM of three or four animals in each group. Asterisks indicate values that are significantly different (P <.5) from normal liver. Higher activity per mg of protein but not per g of tissue of hepatoma homogenates compared to normal liver was due to less protein per g of tumor tissue (8% of weight versus 17% in normal liver). guanylate cyclase activities compared to normal adult rat liver were observed. The decrease in soluble activity and the increase in particulate activity in regenerating liver suggested initially that a translocation of the enzyme might be occurring. This, however, seems quite unlikely, since gel filtration of Tritonsolubilized homogenates (Fig. 2) resulted in two peaks of enzyme activity with different ratios of Peak II to Peak I and properties characteristic of either the particulate or soluble enzyme (1, 5). Also, in many experiments, a decrease in the soluble activity was not coincident with an increase in the particulate activity and vice versa. Furthermore, cycloheximide prevented the increase in particulate enzyme and protein synthesis with partial hepatectomy. If translocation of enzyme from the soluble to particulate fractions was occurring, it seems that this process required protein synthesis. The increase in particulate activity of regenerating liver was attributed to increased guanylate cyclase activity in purified preparations of plasma membranes and microsomes. Cycloheximide alone increased particulate guanylate cyclase activity (Table 2). This apparent paradoxical effect has also been observed with tyrosine transaminase after actinomycin D (23). The interpretations may be similar in that some protein factors could inhibit these enzymes or the enzymes are normally repressed. Thus, the decrease in synthesis of some specific protein(s) may increase the activities of these enzymes. Other interpretations may be provided and additional studies are needed. We previously reported that hepatoma 3924A has very high levels of cyclic GMP, 26 pmol/mg of protein (16). In this cell line, about 7% of the enzyme activity was found in particulate fractions with a 9-fold increase in its specific activity compared to that of control liver. The studies with hepatoma indicate that particulate guanylate cyclase activity best correlates with the elevated cyclic GMP levels in this tissue. It is also of interest that particulate guanylate cyclase is increased in livers of tumor-bearing animals (Table 3). Some humoral or metabolic effect may be the explanation for this observation, and additional studies are needed. The existence in tissues of two forms of guanylate cyclase with different properties and regulatory mechanisms suggests that there may be different physiological roles for the enzymes. The characteristics of hepatomas, fetal liver, and regenerating liver include an increased rate of glycolysis (24-26) and increased nucleic acid and protein synthesis (27-29). After partial hepatectomy, a number of enzymes related to nucleic acid synthesis increase within hr and reach maximal activities in hr (3-32). These activities are also high in fetal liver and decrease after birth (32, 33). Particulate guanylate cyclase activity was similarly high in proliferating liver and hepatomas. Soluble guanylate cyclase activity correlates with the loss or acquisition of differentiated functions in hepatoma, regenerating liver, and fetal and neonatal liver, respectively. These correlations have led us to speculate that particulate guanylate cyclase activity may be associated with tissue growth and that the soluble activity is associated with acquired functions of differentiation. Some recent reports support this hypothesis. Rudland et al. (1) described an activation of particulate but not soluble guanylate cyclase from cell cultures of mouse fibroblasts with a fibroblast growth factor. Other reports have also described effects of cyclic GMP on cell proliferation (7-9). Because most of the guanylate cyclase is associated with particulate fractions from a number of tissues (1-3), it seems likely that the particulate enzyme should also be important in other processes. Additional studies are needed to support our hypothesis. The authors thank Mrs. Joanne Holmes, Mrs. Pamela Napier, and Mr. Joseph Krisanda for their technical assistance. Dr. Hiroko Kimura helped us in performing partial hepatectomies in rats. These studies were supported with a grant from the National Institutes of Health (AM 15316), a U.S. Public Health Service Diabetes-ndocrinology Research Center grant (AM 1742) and a grant from the Virginia Heart Association. F.M. is the recipient of a U.S. Public Health Service Research Career Development Award (AM 7456). 1. Kimura, H. & Murad, F. (1974) J. Biol. Chem. 249, Kimura, H. & Murad, F. (1975) Metabolism 24, Kimura, H. & Murad, F. (1975) Advan. Cyclic Nucleotide Res. 5, in press. 4. Chrisman, T. D., Garbers, D. L., Parks, M. A. & Hardman, J. G. (1975) J. Biol. Chem. 25, Kimura, H. & Murad, F. (1975) J. Biol. Chem., in press. 6. Hardman, J. G., Chrisman, T. D., Gray, J. P., Suddath, J. L. & Sutherland,. W. (1973) Proceedings of the Fifth Interntational Congress of Pharmacology, San Francisco, 1972 (Karger, Basel), Vol. 5, pp Hadden, J. W., Hadden,. M., Haddox, M. K. & Goldberg, N. D. (1972) Proc. Nat. Acad. Sci. USA 69,

5 Proc. Nat. Acad. Sci. USA 72 (1975) 8. Kram, R. & Tomkins, G. M. (1973) Proc. Nat. Acad. Sci. USA 7, Seifert, W. & Rudland, P. S. (1974) Nature 248, Rudland, P. S., Gospodarowicz, D. & Seifert, W. (1974) Nature 25, , Murad, F. & Kimura, H. (1975) Fed. Proc. 34, Touster, O., Aronson, N. N., Dulaney, J. T. & Hendrickson, H. (197) J. Cell Biol. 47, Kimura, H. & Murad, F. (1974) J. Biol. Chem. 249, Kimura, H., Thomas,. & Murad, F. (1974) Biochim. Biophys. Acta 343, Lowry,. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. (1951) J. Biol. Chem. 193, Thomas,. W., Murad, F., Looney, W. B. & Morris, H. P. (1973) Biochim. Biophys. Acta 297, Ishikawa,., Ishikawa, S., Davis, J. W. & Sutherland,. W. (1969) J. Biol. Chem. 244, Oliver, I. T., Ballard, F. J., Shield, J. & Bentley, P. J. (1962) Develop. Biol. 4, Averner, M. J., Brock, M. L. & Jost, J. P. (1972) J. Biol. Chem. 247, Weinstein, Y., Chambers, D. A., Bourne, H. R. & Melmon, K. L. (1974) Nature 251, Alteration of Liver Guanylate Cyclase Activity Watson, J., pstein, R. & Cohn, M. (1973) Nature 246, Lanzani, G. A., Giannattasio, M., Manzocchi, L. A., Bollini, R., Soffientimi, A. N. & Macchia, V. (1974) Biochem. Biophys. Res. Commun. 58, Garren, L. D., Howell, R. R., Tomkins, G. M. & Crocco, R. M. (1964) Proc. Nat. Acad. Sci. USA 52, Warburg,. (193) Metabolism of Tumors (translated by F. Dickens) (Arnold Contable, London). 25. Weber, G., Henry, M. C., Wagle, S. R. & Wagle, D. S. (1963) Advan. nzyme Regul. 2, Suda, M., Tanaka, T., Sue, F., Harano, Y. & Morimura, H. (1966)GannMonogr. 1, Hecht, L. I. & Potter, V. R. (1956) Cancer Res. 16, Bresnick,. (1971) Methods Cancer Res. 6, Stowell, R.. (1949) Cancer 2, Bollum, F. J. & Potter, V. R. (1959) Cancer Res. 19, Tsukada, K. & Lieberman, I. (1964) J. Biol. Chem. 239, lford, H. L. (1972) Gann Monogr. 13, Burch, H. B., Lowry,. H., Kuhlman, A. M., Skerjance, J., Diamant,. J., Lowry, S. R. & vondippe, P. (1963) J. Biol. Chem. 238,