Insulin-like growth factor-I and insulin increase the stimulatory guanine nucleotide binding protein (Gs) in cultured bovine adrenal cells

53

Molecular and Cellular Endocrinology, 66 (1989) 53-51 Elsevier Scientific Publishers Ireland. Ltd.

MOLCEL

02131

Insulin-like growth factor-I and insulin increase the stimulatory guanine nucleotide binding protein (Gs) in cultured bovine adrenal cells Martine Begeot, Dominique Langlois and Jo& M. Saez INSERM

U 307, HSpital Debrousse, 69322 Lyon Cedex 05, France (Received

Key words: Insulin-like

growth

factor-I;

Insulin;

20 April 1989; accepted

Guanine

nucleotide

binding

1 June 1989)

protein;

Bovine adrenal

cell

Summary The present study shows that pretreatment of BAC cells with insulin or insulin-like growth factor-I (IGF-I) enhances the CAMP response to maximal concentrations of ACTH and cholera toxin. However, the effects of IGF-I at a nanomolar concentration (50 ng/ml) were higher than for insulin at the same concentration but similar for insulin at a micromolar concentration (10 pg/ml). We have investigated whether the effects of the two peptides can be related to some modifications of the guanine nucleotide regulatory binding protein Gs. Insulin enhanced Gs as observed by ADP ribosylation and immunoblotting but the effects were approximately the same at nanomolar and micromolar concentrations; again, the effects of IGF-I (50 ng/ml) were higher. These results indicate that both IGF-I and insulin increase the Gs complex of adenylate cyclase, but IGF-I is more potent than insulin at physiological concentrations.

concerning the stimulatory guanine nucleotide binding protein of the adenylate cyclase (Gs).

Introduction It has been demonstrated that insulin and insulin-like growth factor-I (IGF-I) are able to stimulate growth of bovine adrenal cultured (BAC) cells (Ill and Gospodarowicz, 1982; Simonian et al., 1982). More recently, insulin and IGF-I receptors have been characterized in these cells and the enhancement of CAMP response to ACTH induced by these peptides has been demonstrated (Penhoat et al., 1988). In order to determine the localization of the action of both peptides, we investigate in the present paper the modifications

Address for correspondence: Martine Begeot, 307, Hapital Debrousse, 29 Rue Saeur Bouvier, Cedex 05, France. 0303-7207/89/$03.50

0 1989 Elsevier Scientific

INSERM U 69322 Lyon

Publishers

Ireland,

Materials and methods

Materials ACTH (Synacthen) was obtained from Ciba (Rueil-Malmaison, France); insulin, cholera toxin, ATP from Sigma Chemicals Co. (St. Louis, MO, U.S.A.). IGF-I was a generous gift of Kabi-Vitrum (Stockholm, Sweden) and specific antiserum (ref. 584) for qs and q2 of the stimulatory guanine binding protein (Gs) was a generous gift of Dr. A.G. Gilman (Dallas, TX, U.S.A.). The characterization of the antiserum is described in Mumby et al. (1986). [a- 32P]NAD (23 Ci/mmol) was obtained from New England Nuclear (F.R.G.); iodinated Ltd.

54

cAMP (1100 Ci/mmol) (France).

from Pasteur

Production

Isolation and culture of bovine adrenocortical cells and preparation of crude membranes Bovine adrenal cells were isolated as described previously (Crozat et al., 1986). The cells were cultured in a chemically defined medium, Ham’s F12/DME (1: 1) in the absence or in the presence of insulin at 50 ng and 10 pg/ml or IGF-I (50 ng/ml) for 72 h. The culture medium was changed every 24 h. Preparation of crude membranes from cultured cells is described in a previous paper (Begeot et al., 1988). Protein was determined by the Pierce BCA protein assay method using bovine serum albumin as standard.

Measurement of CAMP CAMP was determined by radioimmunoassay (RIA) as previously described (Langlois et al., 1987).

ADP-ribo~ylation of membrane proteins and gel electrop~oresis ADP ribosylation with cholera toxin is described in a previous paper (Begeot et al., 1988). Polyacrylamide gel electrophoresis on slab gels (2 mm thick, main gel, 10% acrylamide) was performed using the Laemmli system (Laemmli, 1970). After electrophoresis, gels were stained with Coomassie brilliant blue, dried and exposed to Amersham Hyperfilm MP. The radiolabeled protein bands were excised and counted in 299 scintillator in a beta counter.

Immunoblotting After gel electrophoresis as described above, membrane proteins were transferred onto nitrocellulose sheets. To block non-specific protein binding, nitrocellulose was incubated in Blotto solution (Mumby et al., 1986) for 1 h at room temperature. The blots were then incubated overnight at room temperature in Blotto solution containing a 1 : 500 dilution of the specific anti-a antiserum (A-584) which recognizes both forms, the 52 kDa and the 45 kDa, of the QISprotein (Mumby et al., 1986). The nitr~ellulose sheets were then washed 4 times for 30 min with Blotto solution. The blots were then incubated at room temperature for 1 h

with radioiodinated protein A (5~~0 cpm/ml) in Blotto solution, were then washed 3 times in Blotto solution and 2 times with Tris 50 mM, CaCl, 2 mM, NaCl 80 mM. Dried blots were exposed to Amersham Hyperfilm MP. The radiolabeled protein bands were excised and counted in 299 scintillator in a beta counter.

Statistical analysis Statistical analyses were performed with Student’s t-test for comparison of two groups. Differences were considered significant when P was < 0.05. Results and discussion In the first series of experiments we studied the effects of pretreatment of BAC cells with insulin or IGF-I on their ability to produce CAMP in response to several effecters. Insulin was used at 50 ng/ml and 10 pg/ml. The last concentration occupied most of its own receptors and about 50% of IGF type I receptors whereas IGF-I was used at a concentration of 50 ng/ml which occupied most of its own receptors but not insulin receptors (Penhoat et al., 1988). Neither insulin nor IGF-I modified the CAMP production under basal conditions (Table 1) but pretreatment with insulin at nanomolar concentrations enhanced significantly the CAMP response (Table 1) to maximal concentrations of ACTH and cholera toxin. Moreover, the response of BAC cells pretreated with insulin at micromolar concentrations and IGF-I (50 ng/ml) was significantly higher than for cells pretreated with 50 ng/ml of insulin but in these conditions no significant mo~fication of the ED,, for ACTH was observed (data not shown). These trophic effects of insulin and IGF-I in BAC cells on CAMP response were similar to those produced by these peptides on cultured Leydig cells (Bernier et al., 1986; Lin et al., 1986; Perrard-Sapori et al., 1987). Since both ACTH- and cholera toxin-induced CAMP productions were enhanced, it can be assumed that insulin or IGF-I had a positive trophic effect on the stimulatory guanine nucleotide binding protein Gs complex. Therefore, the amount of this protein was evaluated by both ADP ribosylation and immunoblotting after insulin or IGF-I treatment.

55

Fig. 1. Autoradiographic analysis of membrane proteins in 72 h cultured bovine adrenal cells in the absence (control) or in the presence of insulin at different concentrations or IGF-I; after (A) ADP-ribosylation with cholera toxin, or (B) immunoblotting with . spectftc antibody against a4s and a s2 peptide (gift of Prof. Gilman, Dallas, TX, U.S.A.). Lane 1: control; 2: insulin 50 ng/ml; 3: insulin 10 ug/ml; 4: IGF-I 50 ng/ml. Numbers on the left-hand side indicate the molecular masses (kDa).

As shown in Fig. lA, BAC cells contain the two a-subunits 45 and 52 kDa, but the latter is predominant. The third minor band with lower molecular weight is probably due to some proteolytic degradation of the above subunits, since it was not seen in all experiments. Both peptides enhanced the amount of both a-subunits, essentially the 52 kDa form, but the effect of IGF-I was significantly higher than those induced by insulin at both concentrations (Table 2). The stimulatory effect of insulin and IGF-I on (YSwas confirmed by immunoblotting (Fig. 1EZ) using an antibody which recognizes the 45 and 52 kDa subunits

TABLE

(Mumby et al., 1986). However, the 45 kDa Gs a-subunit is weakly labeled, contrarily to the 52 kDa subunit. Wide tissue variation in the ratios of the 52 and 45 kDa Gs a-subunits was reported by immunoblotting, and the adrenal cortex contains more of the 52 kDa than the 45 kDa form (Mumby et al., 1986). The effects of insulin at nanomolar concentrations were approximately the same as those produced by insulin at micromolar concentrations and the differences were not statistically significant; again, the effects of IGF-I were higher (Table 2). The stimulatory effects of both peptides, in particular of IGF-I, on (YSwere higher when the proteins were measured by immunoblot than by

1

EFFECTS ADRENAL

OF INSULIN AND IGF-I CYCLIC AMP PRODUCTION

ON

BOVINE

Cells were incubated for 3 days without or with the indicated concentrations of insulin or IGF-I. The medium was changed every 24 h. The CAMP production was measured as described in Materials and Methods. The results are expressed as percent of control cells and correspond to seven or eight independent experiments, each done in quadruplicate. The means ( f SEM) of absolute values (pmol/106 cells) in control cells are respectively: basal, 0.14*0.03; ACTH (10m9 M), 6.1 f0.8; cholera toxin (1 rg/ml), 2.4kO.4. CAMP production

Percent

of control

Basal Control Insulin (50 ng/ml) Insulin (10 j.tg/ml) IGF-I (50 t&ml)

100 109 f 16 94+12 lOl*lO

* P < 0.05 versus previous

2

EFFECTS OF INSULIN AND LATED OR IMMUNOBLOTTED LATORY PROTEINS (Gs)

Cholera M)

100 1535 8* 195fll * 214*12

Percent toxin

(1 pg/mB 100 127+15 * 161*14 * 172 f 17

values in the same column.

IGF-I ON ADP-RIBOSYNUCLEOTIDE REGU-

Cells were incubated for 3 days without or with the indicated concentrations of insulin or IGF-I. The medium was changed every 24 h. The cholera toxin-ADP ribosylation and the immunoblotting with specific anti-as antiserum were performed as described in Materials and Methods. The results, expressed as percent of control cells, are the means of four experiments.

( f SEM)

ACTH (IO-9

TABLE

Control Insulin (50 ng/ml) Insulin (10 rg/ml) IGF-I (50 ng/ml)

of control

( f SEM)

ADP-ribosylated proteins

Immunoblotted proteins

100 151+ 9 * 163+12 21Ort16 *

100 218+ l* 226 f 12 406f 6*

* P -c 0.05 versus previous

values in the same column.

56

Fig. 2. Autoradiographic analysis of membrane proteins in 72 h cultured bovine adrenal cells in the absence (control) or in the presence of IGF-I at different concentrations; after (A) ADP-ribosylation with cholera toxin, or (B) immunoblotting with the same antibody as in Fig. 1. Lane 1: control; 2: IGF-I 1 ng/ml; 3: IGF-I 10 ng; 4: IGF-I 50 ng/ml. Numbers on the left-hand side indicate the molecular masses (kDa).

ADP-ribosylation. It is difficult to explain this apparent discrepancy and some hypotheses could be that (1) ADP-ribosylation can underestimate the amount of Gs compared to immunoblotting as reported in S49 lymphoma cells (Ransnas and Insel, 1988); (2) the presence of an endogenous ADP-ribosylation can reduce the effects of peptides on ADP-ribosylated proteins, but this last possibility was not studied. The present data and those previously reported (Penhoat et al., 1988) indicate that insulin or IGF-I are required for the maintenance of several specific functions of BAC cells: angiotensin II receptors, ACTH-induced CAMP production, steroidogenic capacity. For all these parameters, insulin at nanomolar concentrations was less potent than insulin at micromolar concentrations which in turn was equipotent to IGF-I at a saturating concentration (50 ng/ml). In contrast, IGF-I was more potent than insulin on (YS,and no significant differences were observed between insulin at nanomolar and micromolar concentrations. This apparent discrepancy could be explained by one of the following hypotheses: (1) Insulin at 10 pg/ml occupied only 50% of IGF type I receptors while IGF-I occupied more than 90% of its own receptors (Penhoat et al., 1988). Thus, one can postulate that full occupancy of IGF type I receptors is required for the trophic effect on (YS. Against this hypothesis is the fact that 10 ng/ml of IGF-I (about 50% occupancy) was as efficient as 50 ng/ml (Fig. 2). (2) The most likely hypothesis is that some of the effects of IGF-I were exerted through IGF type II receptors,

which do not bind insulin (Rechler and Nissley, 1985). Although very often both IGF receptor subtypes coexist in the same cells (Rechler and Nissley, 1985) until now, no IGF type II receptors have been described in adrenal cells. Very few data have been reported concerning the regulation of the G-proteins. An increase in the amount of the a-subunits ((Ys, ai and cue) as well as the P-subunit has been reported in 3T3-L cells during differentiation (Watkins et al., 1987). However, since several factors are required to induce such differentiation the role of each on G-proteins remains unknown. In some human growth hormone-secreting pituitary adenomas an altered Gs has been reported and this was associated with a lack of effect of NaF and growth hormone releasing hormone on the adenylate cyclase activity (Vallar et al., 1987). More recently, it has been shown that experimentally induced diabetes leads to the loss of Gi in rat liver and that this effect can be reversed by treating diabetic animals with insulin (Gawler et al., 1987). However, since diabetes causes a marked decrease of plasma IGF-I (Maes et al., 1986) it is not clear whether the effect of the lack of insulin is direct or caused by the decrease of IGF-I. Moreover, negative regulation of both Gs and Gi by retinoic acid in osteosarcoma cell line ROS 17/2 has been described and this was associated with an inhibition of the hormone-stimulated adenylate cyclase (Imai et al., 1988). The trophic differentiating effect of IGF-I on membrane-bound receptors and in the steroidogenie response has been observed in granulosa

cells (Adashi et al., 1986; Veldhuis and Rodgers, 1987), in Leydig cells (Bernier et al., 1986; Lin et al., 1986; Perrard-Sapori et al., 19871, in ovarian th~al-interstitial cells (Cara and Rosenfield, 1988) and adrenals (Penhoat et al., 1988). Our results show that, in addition, IGF-I and insulin increase the Gs complex of adenylate cyclase. Since in all these models the mitogenic effect of IGF-I is very low, if any, the peptide should be considered as a potent differentiating factor rather than a mitogen. Acknowledgements

This work was supported by grants from the Fondation pour la Recherche Medicale Franc;aise and Universite Claude-Bernard, Lyon (UER de Biologie Humaine). The authors thank Dr. A.G. Gilman (Dallas, TX, U.S.A.) for his generous gift of specific anti-as antiserum, Dr. J. Carew for reviewing the English manuscript and Ms. J. Bois for her expert secretarial assistance. References Adashi, E.Y., Restrick, C.E., Svoboda, M.E. and Van Wyk, J.J. (1986) J. Biol. Chem. 261, 3923-3926. Begeot, M.. LangIois, D., Penhoat, A. and Saez, J.M. (1988) Eur. J. Biochem. 174, 317-321. Bernier, M., Chatelain, P., Mather, J.P. and Saez, J.M. (1986) J. Cell. Physiol. 129, 257-263.

Cara, J.F. and Rosenfield, R.L. (1988) Endocrinology 123, 733-739. Crozat, A., Penhoat, A. and Saez, J.M. (1986) Endocrinology 118, 2312-2318. Gawler, D., Milligan, G., Spiegel, A.M., Unson, C.G. and Houslay, M.D. (1987) Nature 327, 229-232. Ill, C.R. and Gospodarowicz, D. (1982) J. Cell. Physiol. 113, 373-384. Imai, Y., Rodan, S.V. and Rodan, G.A. (1988) Endocrinology 122, 456-463. Laemmli, U.K. (1970) Nature 227, 680-685. Langlois, D., Saez, J.M. and Begeot, M. (1987) Biochem. Biophys. Res. Commun. 146, 517-523. Lin, T., Haskell, J., Vinson, N. and Terracio, L. (1986) Endocrinology 119, 1641-1647. Maes, M., Underwood, L.E. and Ketelslegers, J.M. (1986) Endocrinology 118, 372-382. Mumby, SM., Kahn, R.A., Mining. D.R. and Gilman, A.G. (1986) Proc. Natl. Acad. Sci. U.S.A. 83, 265-269. Penhoat, A., Chatelain, P.G., Jaillard, C. and Saez, J.M. (1988) Endocrinology 122, 2518-2525. Perrard-Sapori, M.H., Chatelain, P.G., Jaillard, C. and Saez, J.M. (1987) Eur. J. B&hem. 165, 209-214. Ransnls, L.A. and Insel, P.A. (1988) J. Biol. Chem. 263, 9482-9485. Rechler, M.M. and Nissley, S.P. (1985) Annu. Rev. Physiol. 47, 425-442. Simonian, M.H., White, M.L. and Gill, G.M. (1982) Endocrinology 111, 919-927. VaIlar, L., Spada, A. and Gianattanasio, G. (1987) Natum330, 566-568. Veldhuis, J.D. and Rodgers, R.J. (1987) J. Biol. Chem. 262, 7658-7664. Watkins, D.C., Northup, J.K. and Malron, CC. (1987) J. Biol. Chem. 262, 10651-10657.